Eastern Vakataka Kingdom
Eastern Vakataka Kingdom
Eastern Vakataka Kingdom
How to cite:
LACEY, HARRIET,RUTH (2017) R amt.ek and its landscape: An archaeological approach to the study of
the Eastern V ak a.taka kingdom in central India., Durham theses, Durham University. Available at Durham
E-Theses Online: http://etheses.dur.ac.uk/12377/
Use policy
The full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or
charge, for personal research or study, educational, or not-for-prot purposes provided that:
• a full bibliographic reference is made to the original source
• a link is made to the metadata record in Durham E-Theses
• the full-text is not changed in any way
The full-text must not be sold in any format or medium without the formal permission of the copyright holders.
Please consult the full Durham E-Theses policy for further details.
Academic Support Oce, Durham University, University Oce, Old Elvet, Durham DH1 3HP
e-mail: [email protected] Tel: +44 0191 334 6107
http://etheses.dur.ac.uk
2
Abstract
This thesis investigates the development of the landscape surrounding the Eastern
Vākāṭaka ritual centre of Rāmṭek in central India. The research aims to contextualise
the site of Rāmṭek through the use of landscape archaeology, to explore its
relationship to rural settlement and thus go beyond the existing preoccupation with
the isolated study of its monumental remains. The results of the survey are used to
construct a hypothetical case study for the development of the Early Historic
landscape in this region. This narrative of landscape development is connected to the
region’s socio-economic development under the Vākāṭakas, which will be related to
the wider context of Early Historic to Early Medieval change in India. The survey
develops existing methodologies to suit the environment encountered on fieldwork
and subsequently a preliminary approach to data analysis is presented.
Through landscape survey and ceramic seriation, broad phases of development can
be determined. Based on a significant increase in material evidence from the Early
Historic period, it is argued that this phase witnessed changes in religious, political
and socio-economic spheres. Whilst these developments are only securely related to
the over-arching Early Historic period, there is evidence to suggest that the
Vākāṭakas influenced development following their establishment of the ritual site
and occupation of the area as a dynastic centre. The survey results demonstrate a
prosperous local economy as opposed to deurbanisation and economic decline,
which is popularly associated with the period of Vākāṭaka rule. The Eastern
Vākāṭaka data is then referred to the wider context of the nature of Early Historic to
Early Medieval urbanism in the Indian subcontinent. It is argued that ‘urbanism’
may have been expressed differently in this period resulting in low-density networks
of productive settlements or conurbations.
1
Rāmṭek and its landscape: An archaeological approach to
the study of the Eastern Vākāṭaka kingdom in central
India
Two Volumes
Volume 1
March 2016
2
Table of Contents
Abstract ................................................................................................... 1
List of Tables......................................................................................... 10
List of Abbreviations............................................................................. 26
Acknowledgements ............................................................................... 28
Chapter 2. The history of research and state of the art: The Eastern Vākāṭaka
dynasty ............................................................................................................... 37
2.2.4. The Decline of the Vākāṭakas and the subsequent Early Medieval to
Medieval development of Vidarbha .................................................................... 60
3.2. The physical geography of the survey area, Vidarbha, Maharashtra ........... 76
3.2.2. Climate, flora and the management of agriculture and water ............... 80
4
4.4. Changing theory and the progression of landscape archaeology in South
Asia .................................................................................................................... 125
4.5. Archaeological theory and practice: Progress in the field and the Rāmṭek
survey .................................................................................................................... 132
4.6. Factors affecting the success of archaeological survey in South Asia ....... 133
4.6.1. Methodology and the nature of the archaeological remains ............... 134
4.7. Conclusion: Landscape archaeology and its potential in the study of Early
Historic India......................................................................................................... 143
Chapter 5. The Rāmṭek survey: Research design and fieldwork methodology .... 146
5.3.4. GIS database and visualisation of the recorded evidence ................... 154
5.4. Ground survey conducted during the three field seasons ........................... 155
5
5.4.4. Definition and types of ‘Site’ .............................................................. 159
5.4.5.2. Season 2: The full survey (October 2012 to March 2013) .......... 163
6
7.3. Ware families ............................................................................................. 224
7.3.6.4. Black slipped red ware family (Red BLS) ................................... 253
8.3. Early Historic activity and the Eastern Vākāṭakas ..................................... 272
7
Chapter 9. Ritual activity in the Rāmṭek landscape: Temple architecture and
sculptural remains ..................................................................................................... 301
9.2. Evidence of Pre-Vākāṭaka ritual activity in the Rāmṭek landscape ........... 304
10.2. Eastern Vākāṭaka presence at Rāmṭek: expansion and investment ........ 358
11.3. Eastern Vākāṭaka impact and investment on the Early Historic landscape
of Rāmṭek .............................................................................................................. 378
Bibliography........................................................................................ 381
8
Appendix 6: Ceramic typology ........................................................... 679
9
List of Tables
Table 2-1. The excavated sites within Maharashtra which display Vākāṭaka period
remains. Nine Early Historic/Early Medieval excavations are fully published while
others are mentioned in review articles. See Sali (1998) and Sawant (2008-2009) for
an overview of further explored sites reported to have Vākāṭaka period remains. ..... 70
Table 3-1. Main sites investigated during the excavations at Mansar ...................... 106
Table 3-2. The periodisation outlined for the structural remains at Mansar following
excavation (Joshi & Sharma, 2005) .......................................................................... 107
Table 3-3. Villages reported as having been explored within Rāmṭek tehsil............ 117
Table 5-1. The types of imagery utilised in this study. Given the financial restrictions
of the project, use was made of free satellite imagery. ............................................. 151
Table 5-2. Outline of the characteristics recorded for each sherd during the first stage
of individual sherd analysis....................................................................................... 176
Table 6-1. The number of sites recorded during the ‘Informed’ survey versus the
systematic quadrat walking, with mention of an additional 47 sites which were
previously known to researchers. .............................................................................. 194
Table 6-2. The broad site categories, to which the 444 recorded sites were assigned.
................................................................................................................................... 195
Table 6-3. The number of each feature type identified over the course of the
‘informed’ survey. ..................................................................................................... 197
Table 6-4. The relevant information for each settlement mound site identified during
the ‘informed’ survey. ............................................................................................... 198
Table 6-5. The types of site identified during systematic quadrat walking. ............. 203
Table 6-6. The count of ceramic sherds per quadrat for transect one, and the number
of quadrats in which that quantity of sherds was collected. ...................................... 204
10
Table 6-7. The count of ceramic sherds per quadrat for transect two, and the number
of quadrats in which that quantity of sherds was collected. ...................................... 204
Table 6-8. The count of ceramic sherds per quadrat for transect three, and the number
of quadrats in which that quantity of sherds was collected. ...................................... 205
Table 6-9. The working project chronology, which gives both the broad period and
the project phasing from I – VI. ................................................................................ 211
Table 7-1. Type codes for the ceramic sherds in the Rāmṭek assemblage, and the
count of each type. .................................................................................................... 216
Table 7-2. Outline of the ware types identified during ceramic analysis.................. 217
Table 7-3. The major categories of ceramic decoration encountered in the Rāmṭek
assemblage; further illustrated examples are available in appendix seven. .............. 221
Table 7-4. Sites with the highest proportion of storage wares. ................................. 224
Table 7-5. The broad ware family groupings in the Rāmṭek assemblage, with their
corresponding specific ware types. ........................................................................... 225
Table 7-6. The count of sherds falling within each ware family in the Rāmṭek
assemblage ................................................................................................................ 226
Table 7-7. Count of ceramic sherds within each ware type in the Black and Red ware
family. ....................................................................................................................... 228
Table 7-8. Table displaying the two vessel types present in the Black and Red ware
family, and the count of each within the specific Black and Red wares. .................. 229
Table 7-9. Count of ceramic sherds within each of the four different micaceous ware
families. ..................................................................................................................... 230
Table 7-10. Vessel types which are exclusively found in the micaceous ware families
and their proposed phase based on published parallels. ............................................ 232
Table 7-11. Individual ware types contained within the Mica1 ware family, and the
count of sherds within each type. .............................................................................. 233
11
Table 7-12. The main vessel types identified within the Mica1 ware family, and the
count of examples per type. The vessel types displayed here have over 3 examples in
this ware family. ........................................................................................................ 234
Table 7-13. Individual ware types contained within the Mica2 ware family and the
count of sherds within each type. .............................................................................. 235
Table 7-14. The main vessel types identified within the MICA2 ware family, and the
count of examples per type. The vessel types displayed here have over 4 examples in
this ware family. ........................................................................................................ 236
Table 7-15. Individual ware types contained within the Dark Mica ware family and
the count of sherds within each type. ........................................................................ 238
Table 7-16. The main vessel types identified within the Dark Mica ware family, and
the count of examples per type. The vessel types displayed here have over 2 examples
in this ware family. .................................................................................................... 238
Table 7-17. The count of sherds within the ware type ‘PATCHY’, which constitutes
the ‘Other Mica’ ware family. ................................................................................... 239
Table 7-18. Count of ceramic sherds within each ware type in the Red ware family.
................................................................................................................................... 240
Table 7-19. The main vessel types identified within the Red ware family, and the
count of examples per type. The vessel types displayed here have over 4 examples in
this ware family. ........................................................................................................ 242
Table 7-20. Sites which have reported ‘sprinklers’ during excavations and the
relevant phasing in which this sherd type was encountered. .................................... 243
Table 7-21. Count of ceramic sherds within each ware type in the Red Burnished Red
ware family................................................................................................................ 244
Table 7-22. The main vessel types identified within the Red Burnished Red ware
family, and the count of examples per type. The vessel types displayed here have over
4 examples in this ware family. ................................................................................ 245
12
Table 7-23. Count of ceramic sherds within each ware type in the Grey/Buff ware
family. ....................................................................................................................... 246
Table 7-24. The main vessel types identified within the Grey/Buff ware family, and
the count of examples per type. The vessel types displayed here have at least 10
examples in this ware family..................................................................................... 248
Table 7-25. Count of ceramic sherds within each ware type in the Gritty ware family.
................................................................................................................................... 249
Table 7-26. The main vessel types identified within the Gritty ware family, and the
count of examples per type. The vessel types displayed here have over 2 examples in
this ware family. ........................................................................................................ 250
Table 7-27. Count of ceramic sherds within each ware type in the Storage ware
family. ....................................................................................................................... 251
Table 7-28. All vessel types identified within the Storage ware family, and the count
of examples per type. ................................................................................................ 252
Table 7-29. Count of ceramic sherds within each ware type in the Grey/Orange ware
family. ....................................................................................................................... 252
Table 7-30. All vessel types identified within the Red BLS ware family, and the count
of examples per -type. ............................................................................................... 253
Table 7-31. Count of ceramic sherds within each ware type in the Unknown ware
family. ....................................................................................................................... 254
Table 7-32. Sites used during the initial ceramic seriation ....................................... 256
Table 7-33. Example table which gives the distribution of red and micaceous wares in
sites with over 100 sherds. The data is sorted by quantity of red ware from largest to
smallest. This clearly demonstrates the proposed different phasing of the Red ware
family and Mica2, with overlap occurring between Red and Mica1 wares. See
appendix eight for the full seriation table. ................................................................ 258
Table 7-34. Ware family distributions at the sites of Hamlapuri (79), Bhijewada
(210), Mandri (348) and Kelapur (353). Counts are shown as a percentage of the
13
assemblage at each site with the total sherd count given in the final column. Ware
families with minimal numbers of sherds have been excluded and the table has been
conditionally formatted to illustrate the spread of wares. ......................................... 260
Table 7-35. The working project phasing including information from published
reports and the ceramic seriation............................................................................... 261
Table 8-1. The sites from where Black and Red ware sherds were collected during
survey. ....................................................................................................................... 270
Table 8-2. The ceramic assemblages of each site with Black and Red ware sherds.
The table displays the major ware families and the quantities of each appearing at the
sites. The ware group ‘Patchy’ has been omitted as only two sherds were present at
Site 213 (Satak). ........................................................................................................ 270
Table 8-3. Examples ceramic sequences from the known Vākāṭaka sites in the survey
area ............................................................................................................................ 274
Table 8-4. Sites in the survey area with significant quantities of Red and Mica1
wares, as at Hamlapuri and Mansar (included within the table). .............................. 276
Table 8-5. Areas of sherd collection around Nagardhan town giving the relative
amounts of Red, Mica1 and Mica2 sherds. ............................................................... 281
Table 8-6. Sites with the highest proportion of Mica2 wares, demonstrating the
common association with Red and Mica1 wares. ..................................................... 282
Table 8-7. Sites which are not represented by all three ceramic types, but which have
more than 10 sherds of at least one type. .................................................................. 283
Table 8-8. The sites with the largest quantities of Red Burnished ware in the survey
area. ........................................................................................................................... 289
Table 8-9. Sites with the largest quantities of Grey/Buff wares in the Rāmṭek
assemblage ................................................................................................................ 296
Table 9-1. Brick dimensions referred to in reports of the Mansar excavations (Joshi &
Sharma, 2000; 2005). ................................................................................................ 309
14
Table 9-2. Dimensions of bricks from MNS II (the ‘Palace’ site) at Mansar measured
during survey. All pieces are proposed to be Vākāṭaka bricks and demonstrate the
range of sizes in bricks associated with Vākāṭaka structures. It is noted that some of
the complete bricks are comparable to brick sizes noted in Table 8.1 which have been
assigned different dates by the excavators. ............................................................... 310
Table 9-3. Miniature Narasiṃhas known from the survey area in publications. ...... 318
15
List of Figures
Figure 2-1. Genealogy of the Vākāṭaka dynasty with key dates, after Bakker (1997).
..................................................................................................................................... 46
Figure 2-2. General map of the find-spots of the Vākāṭaka inscriptions. The red
markers correspond to the inscriptions of the Western branch while blue refers to the
Eastern branch. (See the British Museum project for further details. Upcoming
publications contain proprietary information
http://www.britishmuseum.org/research/research_projects/all_current_projects/beyon
d_boundaries/inscriptions_and_sites.aspx) ................................................................. 64
Figure 3-2. The position of Vidarbha within the state of Maharashtra. The locations of
Rāmṭek and other major archaeological sites are indicated. ....................................... 77
Figure 3-4. Map of the survey area indicating the mines, land-use, earthworks and
tanks that appear on topographic maps of the area. A number of the mines indicated
appear to have expanded since the topographic map was produced. All land outside of
the forest cover is simply designated as ‘open scrub’ and in practice is largely
agricultural land or areas of culturable waste. ............................................................ 79
Figure 3-5. Images showing the extent of quarrying and manufacturing activities in
the survey area: a) Satak manganese mine. b) Brick manufacturing north of Rāmṭek ...
..................................................................................................................................... 79
Figure 3-6. The hill-top site of Rāmṭek with the seven Vākāṭaka period temples
indicated. ..................................................................................................................... 84
Figure 3-7. Square pavilion temple containing the Varāha sculpture on Rāmṭek hill 86
Figure 3-8. Decorative details on the Narasiṃha temples: a) Decorative frieze on the
exterior of the Kevala-Narasiṃha temple. b) Lotus pattern on the columns of the
Rudra-Narasiṃha temple. ........................................................................................... 88
16
Figure 3-9. The ‘Kevala-Narasiṃha’ inscription, thought to originally have been
located in the Trivikrama temple. ............................................................................... 89
Figure 3-10. a) The Trivikrama sculpture, which is located on Rāmṭek hill but is no
longer housed in the remaining temple structure. b) Part of the original Trivikrama
temple which has been partly reconstructed. .............................................................. 91
Figure 3-11. Decorative elements on the exterior of the Bhogarāma’ temple ............ 92
Figure 3-12. Side view of the Guptarāma temple on Rāmṭek hill .............................. 93
Figure 3-13. The Sindūravāpī tank on Rāmṭek hill, which displays columns similar to
those in Vākāṭaka constructions. ................................................................................. 95
Figure 3-14. Vaulted roof of the Kālikā Devi temple in Rāmṭek town....................... 96
Figure 3-15. Karpūravāpī at the base of Rāmṭek hill on the north side. ..................... 96
Figure 3-16. A group of Medieval temples situated on the south west side of Ambālā
Lake. ............................................................................................................................ 97
Figure 3-17. a) The small soapstone snake in the Wellsted Collection of the British
Museum, b) A selection of the ceramic vessels in the Wellsted Collection ............. 102
Figure 3-18. Wellsted’s map of Mansar tank and hill, showing the presence of buried
monumental constructions (1934). ............................................................................ 103
Figure 3-19. Mansar hill as it stands today with several of the larger sites exposed by
excavation and reconstructed. ................................................................................... 104
Figure 3-20. The Mansar Śiva at Delhi National Museum, labelled as Śiva Vamana
(Śiva in the form of a dwarf) and measured at 84cm in height, 65cm wide, and 37cm
deep. .......................................................................................................................... 105
Figure 3-21. Stellate-plan liṅga temple on Mansar hill (site 10: MNS V)................ 108
Figure 3-22. South side of MNS III, the Pravareśvara temple, with its elaborate brick
shrines ....................................................................................................................... 110
17
Figure 3-24. MNS IV (site 9) described as a Buddhist vihara by the excavators of
Mansar. ...................................................................................................................... 113
Figure 3-25. Remains of the site at Hamlapuri (photograph taken during the first
season of fieldwork). The paddy field banks and the ‘island’ with the trees represent
the original height of the ground. Ceramic and brick remains are found throughout the
sections and across the surface of the fields.............................................................. 115
Figure 3-26. Temple at the site of Kelapur in an area often flooded by Khindsi tank.
Surrounding the temple are surface finds of ceramic sherds .................................... 116
Figure 4-2. A traditional plough used by farmers around Rāmṭek, with short farrows
which would not penetrate deep into the ground. ..................................................... 139
Figure 5-1. The 10km survey area surrounding Rāmṭek displaying the main, known
sites and the spread of modern villages..................................................................... 157
Figure 5-2. The randomly generated transects across the 10km survey area............ 171
Figure 5-3. The randomly selected quadrats within the transects across the original
10km survey area ...................................................................................................... 172
Figure 5-4. Rim types 147 and 148, which display a similar overall vessel shape of a
double lipped rim, but were split into two types based on coherent groupings of
variation..................................................................................................................... 181
Figure 5-5. A typical string cut base, commonly seen on type 11 within the ceramic
assemblage. The vast majority of bases identified were of this type. ....................... 182
Figure 6-1. GoogleEarth colour imagery of the low mound at site 262. No shadow is
cast from the mound which has been flattened by agriculture but the surface brick
scatter is visible as orange colouration of the ground. .............................................. 190
Figure 6-2. Site 276 as seen on GoogleEarth colour imagery. The ploughing of the
field is visible and obscures the shape and tone of the mounded area. See in
comparison to Figure 6-3. ......................................................................................... 190
18
Figure 6-3. Site 276 identified during survey of the Orbview 3 imagery; the shape of
the mound is more clearly visible than on the colour imagery and the variation is soil
colour is recognised................................................................................................... 191
Figure 6-4. The identified mound sites in the survey area featuring brick and ceramic
remains. ..................................................................................................................... 200
Figure 6-5. Map giving the visibility across sampled quadrats in the three transects
surveyed during systematic field-walking. ............................................................... 202
Figure 6-6. Map displaying the sherd density per sampled quadrat with clear low
levels of findings across the majority of quadrats. .................................................... 205
Figure 6-7. The progression of modern damage at the site (364) of Naharwani from
top left as shown in (a) 2002 GoogleEarth imagery, (b) 2009 GoogleEarth imagery,
(c) 2014 GoogleEarth imagery (d) A photograph of the site during survey (2012). 207
Figure 6-8. From left: Modern digging activities south of the fort. Site 73 is shown as
(a) cultivated land on 2012 GoogleEarth imagery and (b) following excavation on
2014 imagery. ............................................................................................................ 208
Figure 6-9. Photographs show (a) the extent of the disturbance south of the fort at site
73 during the second field season in 2013, and (b) the collapsed section of wall at the
Bhonsle Fort. ............................................................................................................. 208
Figure 7-1. The two types of Black and Red ware rims encountered in the Rāmṭek
assemblage. ............................................................................................................... 229
Figure 7-2. The most common vessel shape present in the Mica1 ware family. ...... 234
Figure 7-3. The most common vessel shape in the Mica2 ware family .................... 237
Figure 7-4. The most common vessel shape in the Dark Mica ware family. ............ 239
Figure 7-5. Type 11, which is the only vessel shape identified in the Patchy
micaceous ware. ........................................................................................................ 239
Figure 7-6. A common jar form in the Red ware family. ......................................... 242
19
Figure 7-7. The most common type present in the Red Burnished Red ware family.
................................................................................................................................... 245
Figure 7-8. The most common type present in the Grey/Buff ware family. ............. 248
Figure 7-9. The most common type found in the Gritty ware family. ...................... 251
Figure 7-10. Two sherds included in the ware group NON-ID: a) Sherd 3317 with a
golden micaceous slip, b) Sherd 4221 which is also micaceous but has a light pink
fabric and a friable texture. ....................................................................................... 254
Figure 8-1. Section in the mound at Dudhala (site 236) displaying the ashy layer, with
the location of find 20 indicated. .............................................................................. 267
Figure 8-2. Map of the survey region displaying the sites with the presence of Black
and Red ware sherds.................................................................................................. 269
Figure 8-3. Map of the survey region displaying the sites with the presence of Red
ware sherds. ............................................................................................................... 273
Figure 8-4. Map of the survey region displaying the sites with the presence of Mica1
sherds. ....................................................................................................................... 277
Figure 8-5. Map showing the relative proportions of Red and Mica1 wares at sites in
the survey area........................................................................................................... 277
Figure 8-6. The mound configuration around Nagardhan on Orbview 3 imagery ... 278
Figure 8-7. Section below the Bhonsle fort displaying layers of brick and ceramic 281
Figure 8-8. Map of sites in the survey area displaying Mica2 sherds. ...................... 282
Figure 8-9. Difference in site distribution between those with Mica1 wares and those
with Mica2 wares. A number of sites are represented by both types, however the
distribution of ceramics changes across the two ware phases. .................................. 283
Figure 8-10. Remains of a ring well in a large area of digging south of Nagardhan fort
(site 73)...................................................................................................................... 287
20
Figure 8-11. Map of sites across the survey area with the presence of Red Burnished
wares ......................................................................................................................... 290
Figure 8-12. Site 307: Stone well in Bori village with octagonal casing and hooped
stones on a parapet for a pulley system. .................................................................... 292
Figure 8-13. Site 162: Round brick and stone well with a parapet, located in
agricultural fields. ..................................................................................................... 292
Figure 8-14. Site 407: Loose stones near to an old stone well adjacent to a Hanuman
temple in Chorkumari village, which show the hole used for a pulley system to fetch
water. ......................................................................................................................... 293
Figure 8-15. View of brick and stone bāvlī to the east of Ambālā tank. In the left hand
side of the picture the long, underground arched bāvlī terminates in a round brick and
stone well with a parapet. .......................................................................................... 294
Figure 8-16. The bāvlī in a temple complex in Rāmṭek town (site 119), which has
since been filled in and concreted over. .................................................................... 294
Figure 8-17. Example earthwork in the Rāmṭek landscape. This appears to be an old
tank embankment at Manapur village, with a modern shrine situated on the top (site
227). .......................................................................................................................... 295
Figure 8-18. Map of sites across the survey area displaying the presence of Grey/Buff
wares. ........................................................................................................................ 296
Figure 9-1. From left a) 2014 GoogleEarth imagery of Beldongri mound (site 242)
and b) 2013 GoogleEarth imagery of the second mound at Dudhala to the west of the
main mound which is partially cut away. ................................................................. 306
Figure 9-2. The inner and outer phases of the pilaster mouldings on MNS II. ......... 308
Figure 9-3. Damage caused by rains to the southern side of MNS II. Inner courses of
brick have been revealed across this structure and at MNS III. ................................ 309
Figure 9-4. Red Sandstone architectural fragments on Mansar hill associated with the
stone and brick monumental constructions. .............................................................. 311
21
Figure 9-5. Beldongri mound (site 242) which is a brick construction with large sized
bricks comparable to those used in other Gupta period constructions. ..................... 312
Figure 9-6. Section on south west side of Beldongri mound. Courses of in situ brick
has been reveals indicated an angular construction. ................................................. 312
Figure 9-7. a) Eroded torso of sculpture number 211, which resembles the shape of
the large Narasiṃha on Rāmṭek hill; b) Kevala-Narasiṃha sculpture from Rāmṭek hill
for comparison (photo credit: Prof Hans Bakker) ..................................................... 313
Figure 9-8. Sculpture no. 40 and 43: 5th century liṅgas with pārśvasūtra and
brahmasūtra at Mansar. ............................................................................................ 316
Figure 9-9: Left to right: Sculpture no. 668: Small sandstone liṅga with no discernible
markings in Nagardhan town. Sculpture no. 88: Liṅga in Hivri village associated with
brick and ceramic remains (site 202) ........................................................................ 317
Figure 9-10. Sculpture no. 25: Small Narasiṃha at Koteshwar temple in Nagardhan.
................................................................................................................................... 319
Figure 9-11. Small Narasiṃha at Nagpur Museum. This sculpture was discovered at
Rāmṭek and is dated to the 5th century AD ............................................................... 320
Figure 9-12. (a) Sculpture no. 21: Ganesh sculpture at Hamlapuri, published in
Bakker (1997). (b) Sculpture no. 23: possible 5th century Liṅga found in the modern
temple alongside the image of Ganesh ..................................................................... 321
Figure 9-13. Sculpture no. 534: Fragmentary sculpture of a boy’s head, possibly
originating from the 5th Century AD. ........................................................................ 322
Figure 9-14. Sculpture no. 295: Fragmentary sculpted head with elaborate curled
hairstyle. ................................................................................................................... 322
Figure 9-15. Sculpture no. 294: Sculpture of a head with matted hair at Rāmṭek .... 322
Figure 9-16. (a) Sculpture no. 17: Gaṇa-type figure at Koteshwar temple in
Nagardhan. (b) 5th century corpulent Gaṇa type figure at Nagpur Museum. ........... 323
22
Figure 9-17. Lajjā Gaurī figures identified at Hamlapuri: a) Sculpture no. 69. b)
Sculpture no. 70, the second Lajjā Gaurī figure, which has been placed on the cabūtrā
upside down. ............................................................................................................. 326
Figure 9-18. Sculpture no. 271: A small sandstone Viṣnupada, typical of those
identified in the Rāmṭek survey, at site 159 near Kawadak. ..................................... 327
Figure 9-19. Sculpture no. 2: Five-headed Naga from the Kevala-Narasiṃha temple at
Rāmṭek. ..................................................................................................................... 329
Figure 9-20. (a) Sculpture no.538: 5-hooded metal Naga in a modern shrine in
Rāmṭek, (b) Examples of metal Nagas in a small shrine in Mandri (site 340) ......... 330
Figure 9-23. Āmalaka fragments from across the survey area. Left to right, top row:
(a) Red sandstone āmalaka in the enclosure of the Varāha temple on Rāmṭek hill (b)
Sculpture no.142. Left to right, bottom row: (c) Sculpture no.418: An āmalaka
worshipped as the saptamatrikas (d) Sculpture no.321: An āmalaka worshipped as
the saptamatrikas ...................................................................................................... 334
Figure 9-24. (a) Sculpture no. 58: A vyāla fragment found at Manapur temple. (b)
Sculpture no. 274: A vyāla fragment incorporated into a cabūtrā north of Rāmṭek
(site 169), which features a number of architectural pieces. ..................................... 335
Figure 9-25. Doorjamb fragments discovered during survey: a) Sculpture no. 272,
part of a doorjamb on a cabūtrā north of Rāmṭek (site 169), which features a number
of architectural pieces. b) Sculpture no.16, architectural door piece featuring a
Makara located at Koteshwar temple, Nagardhan (site 50). ..................................... 335
23
Figure 9-26. Chatrī located on Nagardhan mound, west of the fort (site 75),
displaying a common plinth motif. ........................................................................... 337
Figure 9-28. Examples of Medieval or Maratha period temples in the survey area: a)
The temple at Japala (site 393), which also features an associated stepped reservoir.
b) A temple at Lohadongri village (site 284), whose superstructure has recently been
renovated and painted. .............................................................................................. 339
Figure 9-29. The distribution of temples and shrines across the survey area. Modern
villages have been indicated on the map and modern temples were most commonly
recorded at modern settlement sites. Temples deemed to be older (Post-
Vākāṭaka/Medieval) also demonstrate a correlation with some of the modern villages
suggesting a continuation of settlement. Shrines and temples are also found
throughout the agricultural areas. .............................................................................. 339
Figure 9-30. South facing Hanuman temple with modern exterior in Nagardhan (site
62). ............................................................................................................................ 340
Figure 9-31. Sculpture no. 266: Example of the trimūrti of Ganesh, Annapurna and
the Naga which form part of the Pañchāyatana. ...................................................... 341
Figure 9-32. Individual temple niche icon of Annapurna (sculpture 175) located in a
Hanuman temple in Lohara (site 315). ...................................................................... 342
Figure 9-33. Left to right: Sculpture no. 275: Painted Nandi at Japala temple (site
393). Sculpture no. 202 Nandi at a Hanuman temple in Bori village (site 311) ....... 343
Figure 9-34. Sculpture no. 147, ‘Yoni’ with small liṅga and oval, shallow pīṭha.... 343
Figure 9-35. Left to right: Sculpture no. 56: Standing Hanuman with tail curled over
his head, hands in front of the body and a gada under his left arm; Sculpture no. 203:
Striding Hanuman in the “heroic” vīra pose. ............................................................ 344
Figure 9-36. Site 166: Small modern shrine containing stones, located in an
agricultural field. ....................................................................................................... 345
24
Figure 9-37. (a) Sculpture no. 282: Unusual folk-style temple alcove piece of
unknown figure at Japala temple (site 393) potentially Medieval (Bakker pers. comm)
(b) Sculpture no. 77: Sculpture of unknown figure on a cabūtrā between Mansar and
Rāmṭek, which is said to be Hanuman by villagers (site 174). ................................. 346
Figure 9-38. Sculptures no. 169 and 170: Folk-style depictions of two figures. Similar
sculptures are very common across the survey area ................................................. 346
Figure 9-39. Sculpture no. 449: Red sandstone sculpture of a female torso in a
provincial style at Chichala (site 258) ....................................................................... 347
Figure 9-41. Left to right: (a) Sculpture no. 157: Five-headed Naga at a Hanuman
temple in Kandri, north of Mansar (site 426). (b) Sculpture no. 513: Single-headed
Naga on village outskirts (site 370) .......................................................................... 348
Figure 9-42. (a) Sculpture no. 304: Unusual Nagakal at site 198 in Hivra village. (b)
Sculpture no. 54: Naga stone at the Kālikā Devi temple in Rāmṭek. ........................ 349
Figure 9-43. A Jain temple in the complex north of Rāmṭek town (site 99). ........... 350
Figure 9-44. Sculpture no. 696: Possibly Jain image kept in the Jain Mandir in
Rāmṭek. ..................................................................................................................... 351
Figure 9-45. Left to right: a) Sculpture no. 303: Potential hero stone piece with
depiction of the sun and moon. b) Sculpture no. 152: Sandstone hero stone fragment.
................................................................................................................................... 352
Figure 9-46. From top left to right: (a) Sculpture no. 232: Memorial stone (b)
Sculpture no. 137: Image which may be a memorial stone. (c) Sculpture no. 333: Red
sandstone memorial stone. (d) Sculpture no. 342: Memorial stone .......................... 353
25
List of Abbreviations
26
The work contained in this thesis has not been previously submitted elsewhere for
any other degree or qualification and unless otherwise referenced is the author’s own
work.
The copyright of this thesis rests with the author. No quotation from it should be
published without the author's prior written consent and information derived from it
should be acknowledged.
27
Acknowledgements
I wish to extend my sincerest gratitude to everyone who has helped and supported me
over the course of this PhD. I would like to especially thank Derek Kennet for his
supervision, humour and endless patience. It is through his perseverance and support
that this thesis has concluded. My deepest thanks go to Hans Bakker for his
contributions and friendship; I have benefited greatly from his insights and
incomparable knowledge of the Vākāṭaka dynasty. For his support, leadership and
sponsorship I would like to thank Michael Willis, without whom this research could
not have taken place.
Funding for this PhD was generously provided by The Leverhulme Trust and The
British Museum. I would also like to thank the Indian National Trust for Art and
Cultural Heritage for providing a grant for fieldwork.
Thank you to Niko Galiatsatos for steering me through my initial encounter with
ArcGIS and for his patient guidance. I would also like to thank Jason Hawkes for his
advice and kind words; additional thanks go to Jason for arranging the digitisation of
my ceramic illustrations with Vicky Herring, and to Vicky for doing such a fantastic
job.
I would like to show my appreciation for the staff and students of the Department of
Archaeology in Durham. A number of individuals were always on hand to help or
give encouragement; particular thanks go to Tom Fitton, Mark Manuel, Chris Davis,
28
Dan Lawrence, Brian Buchanan, Jo Shoebridge Jo Matias, Davina Craps, Maria
Lahtinen and the residents of office 329! Thank you to Ben Saunders, whose
unstinting support for me went far beyond anything I might have reasonably
expected.
Finally, I would like to thank all of my friends and family who have been with me
every step of the way. Thank you to Dan for keeping me sane and always making me
laugh, and to Tina and Ronan for keeping me going. To Lizzie, Nicola, Cerri, Leanne,
Jess and David, thank you for putting up with my long absences and always making
me feel loved. To Andrew, thank you for everything.
There are not enough words to express my gratitude to my parents to whom this
thesis, and everything I do, is dedicated. Thank you to my mum and dad for their
love, generosity and encouragement, with special additional thanks to my dad for the
endless proof-reading. Although he probably deserves an archaeology degree of his
own by now, a dedication will have to do.
29
Chapter 1. Introduction
Chapter 1. Introduction
1.1. Introduction
This research uses surface survey methods to evaluate the landscape around the site
of Rāmṭek in order to address the central hypothesis, which seeks to challenge the
historical narrative of Eastern Vākāṭaka expansion by looking for pre-existing Early
Historic populations. This study considers the Eastern Vākāṭakas’ economic and
ritual activity in the immediate landscape at Rāmṭek, in order to assess their impact
on this area.
The Vākāṭakas are often considered to have been a powerful kingdom in central India
during the 5th century AD, second only to the Guptas, and are assumed to have
operated from a state level. However, the nature of their organisation and kingship is
not well established archaeologically, and it is not known how the polity operated
through the landscape (Kulke, 2004:3; Smith, 2003). Landscapes are central in
understanding cultural and political systems as they can be considered the arena
through which social structures are “communicated, reproduced, experienced and
explored” (Duncan, 1990:17). The Eastern Vākāṭakas moved to the Rāmṭek area to
establish a new centre for ritual and royal authority and to execute their religious and
30
Chapter 1. Introduction
Available evidence suggests that the 3rd to 5th centuries AD witnessed crucial changes
in secular and religious spheres, which has been linked to the spread of a shared
ideology across large areas of India. The Guptas are perceived to have led the
expansion of shared ideals, however the Vākāṭakas appear to have been instrumental
in the spread of this ideology across the region under their control (Misra, 1987:643).
The transformations at hand include the revitalisation of Brahmanism evidenced by
the widespread and swift re-adoption of Sanskrit across large areas of India as the
language of political discourse, scientific and ritual texts and literature (Goyal,
2006:28). Religious iconographies became increasingly standardised and were
supported by vibrant artistic visualisation. Gupta sculpture is considered to represent
the zenith of ancient Indian art and a vivid flourishing of art also occurred in central
India under the Vākāṭakas (Goyal, 2006:27). At this time, the building of temples
intensified, Buddhism, Hinduism and Jainism gained permanent endowments and
religious centres appear to have developed as socio-economic institutions (Altekar,
1960:199). Economic transformations during the period of Vākāṭaka rule include the
widespread use of permanent land-grants and the establishment of village settlements
(Bakker, 2010:472; Goyal, 2009:30). The striking conformity to a set of ideals in
land-grants across different dynasties indicates mutual legal systems and
organisational apparatus alongside the shared ideology visible in the prevailing
religious and ritual systems across India. Willis believes that this consistency could
be construed as a form of political unity, which saw socio-economic and political
relationships structured by ritualism (Willis, 2009).
31
Chapter 1. Introduction
formation at this time, and how organisational networks were used to navigate and
formulate relationships between the ritual and political polity and the landscape, local
population and production.
The Vākāṭaka ritual centre of Rāmṭek was established during this transformative time
in Indian history. Rāmṭek and its environs provide an opportunity to study the longue
durée of a central Indian landscape and to investigate the impact that the
establishment of such a major religious site would have had on its surroundings.
Rāmṭek forms the centre of the survey universe in an attempt to explore the
organisation and development of the Eastern Vākāṭaka dynasty, given that the site is
representative of royal investment and monumentality. Surface survey is aimed at
assessing the impact of monumental investment in the landscape and the nature of the
relationship between the ritual site and its hinterland. Power is connected to
landscapes as they both reflect and direct the nature of state organisation, perceptions
of authority and the establishment and legitimisation of political, economic and ritual
systems. Despite their prominence in central Indian history, the Eastern Vākāṭakas
appear to have been contained within a small territory in a traditionally remote area.
Rāmṭek is considered a case study for landscape development under the Eastern
Vākāṭaka polity, and it is hoped that survey data may be linked to concepts of
kingship and the ideology held by the elite and general population.
This research originated with the British Museum’s project “Politics, Ritual and
Religion: State formation in early India”, which aims to assess the impact of the
Gupta dynasty and its extended networks on India. This project is directed towards
reuniting the study of the Guptas and their partner kingdoms, following a history of
separate study due to disciplinary divides and regional constraints in Indian research.
Research by Hans Bakker at Rāmṭek-Mansar and by Michael Willis at Udayagiri
under the British Museum, has so far attempted to broaden our understanding of royal
centres of ritual and the developments of temples as socio-economic institutions
under the Guptas and Vākāṭakas. The research at the Gupta site of Udayagiri can be
viewed in light of Shaw’s Sanchi survey and archaeological research at Vidisha,
which provided evidence for the chronological development of the region and
therefore the wider context necessary for considering Udayagiri as part of socio-
economic development in the area. This thesis is aimed at applying this concept of
32
Chapter 1. Introduction
1. What was the nature of ritual and economic activity across the landscape
prior to the major recorded period of c.4th to 5th centuries AD and is it
possible to trace the landscape’s habitational development over the longue
durée?
2. What is the relationship between the nascent political and ritual centre of
Rāmṭek and the surrounding landscape? How did Vākāṭaka rule and the
establishment of the ritual site alter or impact on the nature of settlement?
3. Does data indicate that the developments witnessed in the 5th century were
an independent phenomenon generated by the arrival of the Vākāṭakas
after high investment in the landscape, or did they involve the
continuation of existing socio-economic/religious patterns?
4. How does Vākāṭaka investment in the landscape compare to that of the
Post-Vākāṭaka periods?
In order to attempt to address the research questions throughout the course of the
research, the following three components were initially defined for exploration:
33
Chapter 1. Introduction
There is great potential for the application of landscape techniques to Indian historical
archaeology and the focused survey of an area not only assists researchers in
improving understanding of a specific locale, but may also inform about regional
socio-political organisation. This research is concerned with the archaeological
investigation of Rāmṭek and the surrounding landscape primarily through field
survey, producing data which can be used to create an archaeological map of the
area’s development to be considered against what is known of the Vākāṭaka Kingdom
from wider scholarship.
34
Chapter 1. Introduction
the analysis of the ceramics. One method of data collection and analysis for the study
of the ceramic assemblage is presented here and this represents an initial attempt to
approach the landscape and characterise the resulting material remains in the absence
of previous research in the hinterland of the central site.
Chapter six evaluates the success of the methodology and the general overview of the
results, referring to the different datasets utilised and the approaches employed. This
chapter also introduces the working chronology of the project prior to a more in-depth
consideration of the major categories of remains identified during survey. The
potential for land-scape survey to characterise development in an area is
demonstrated by the successful collection of archaeological data related to rural
activity. Chapter seven focuses on the results of the ceramic analysis, which provides
the basis for discussions of the economic nature of the landscape. Following the
collection of data, remains were considered spatially and temporally to indicate the
Vākāṭaka impact on the landscape compared to other periods. Chapters eight and nine
focus on the economic and ritual landscape respectively in order to inform us about
the scope of Eastern Vākāṭaka polity. These two spheres of activity are
interconnected, particularly as religious ideology is said to have structured the
political and economic relationships of the Gupta Vākāṭaka age (Willis, 2009:9).
However, separate discussions are utilised to ascertain the level of investment by the
Vākāṭakas given the quantity of findings.
35
Chapter 1. Introduction
Chapter ten will synthesise the evidence regarding the economic and ritual function
of the site of Rāmṭek and its hinterland in order to theorise about the extent of
Vākāṭaka influence on the landscape. This chapter will compare the fieldwork data to
wider concerns, such as the nature of state formation and organisation and kingship,
to achieve a more holistic view of the Vākāṭakas and their activity in the Rāmṭek area.
This will conclude with a comment on the original hypothesis of this research and test
the historical narrative of Vākāṭaka expansion.
In this study, it is argued that there was great Early Historic development around
Rāmṭek compared to the previous periods. The evidence at the key sites and the
pattern of rural habitation seem to imply an agricultural economy focused on royal
and ritual activities. Despite being the central area of an important dynasty there does
not appear to have been a highly urbanised and centralised core, and it is possible that
the Eastern Vākāṭaka dynasty was organised loosely to exploit the landscape. This
could be akin to a ‘node and network’ state organisation or a ‘kingly reading’ of
landscape, which establishes a political ideology and legitimates authority alongside
enabling flexibility in territorial control.
The evidence from the survey is limited by the constraints of the project, and
although it may be used to reflect on our questions surrounding the Vākāṭakas and
Early Historic central India, the conclusions presented here are necessarily
preliminary and conjectural. It should be acknowledged that our understanding of the
period is still limited and there remains considerable scope for further research; the
study presented here acts as the necessary foundation for more in-depth research and I
welcome challenges or additions to the conclusions presented here.
36
Chapter 2. History of Research
2.1. Introduction
37
Chapter 2. History of Research
understanding of the Gupta-Vākāṭaka period in central India, and aids in situating the
current study in the history of research.
Agricultural settlements are suggested to have emerged as early as the 8th century
BC, and a clear Iron Age tradition existed which followed on from, and occasionally
overlapped with, Chalcolithic remains (Chakrabarti & Lahiri, 1994:23; Parashar-Sen,
1999:163). The Megalithic tradition - the use of specialised burial monuments
associated with a common ideology - developed alongside Iron Age society from the
early first millennium BC, persisting up to the Early Historic period (Park & Shinde,
2013:3823). The earliest emergence of Megaliths occurred in South India
38
Chapter 2. History of Research
contemporary with Neolithic traditions, with a later spread into Vidarbha. 86 out of a
known 91 Megalithic habitation sites in Maharashtra were located in Vidarbha
(Mohanty & Walimbe, 1993). The higher numbers of Megaliths towards the East of
Vidarbha, compared to more limited evidence in the Western Deccan, has been
suggested to be associated with the area’s rich natural resources (Gullapalli,
2009:443; Parashar-Sen, 1999:163; Thakuria, 2014:652). A rise in site number during
the Vidarbha Iron Age indicates there was an increasing population and recent
research into Iron Age/Megalithic habitations has identified a complex and varied
economy. Iron Age communities appear have practised both agricultural and hunter-
gathering activities, and were engaged in developed goods production (Mohanty &
Joshi, 1996).
39
Chapter 2. History of Research
The size of several major settlements, such as Adam, viewed alongside the
archaeological evidence of relatively advanced agriculture, craft specialisation,
defence strategies and long distance trade, may indicate that the beginnings of early
state formation and urbanisation emerged prior to the Early Historic period
(Morrison, 1995:208). The process of Early Historic urban growth across India is
connected to the political consolidation and economic change associated with the
formation of kingdoms known as janapadas in the first half of the first millennium
BC (Agrawala, 1953; Thapar, 1990). Janapadas developed to have more defined
boundaries as local chiefs became more akin to territorial monarchs in competition
for territory (Ghosh, 1973:32). In North India around the 6th to 4th centuries BC
some of the janapadas developed into larger, organised polities called
mahajanapadas. On the Gangetic plain during this period, one of the largest
mahajanapadas, Magadha, appears to have become dominant and brought large areas
of North India under the control of the Mauryan Dynasty (c. 320 to 180 BC) (Banerji,
1933). Magadha appears to have been “agriculturally rich” and supported by a high
population density, and its initial spread of power and control across neighbouring
areas was achieved through conquest and annexation (Thapar, 2000b:467).
40
Chapter 2. History of Research
operated independently. Mauryan contact with central India may have been related to
attempts to access raw materials and control the major trade routes through the
Peninsula, such as the Dakṣiṇāpatha (Thapar, 2000b:474). Comprehensive studies of
the 54 sites with Mauryan Aśokan edicts across India demonstrate that their extent is
not analogous with the territory and control of the Mauryan Empire (Fussman, 1988).
Rather, it seems that as well as being declarations of Buddhist values, the Aśokan
edicts were monumental symbols of authority linked to state-level interactions and
attempts by the Mauryan Empire to integrate diverse communities (Falk, 2006).
Both the Chalcolithic to Iron Age transition and the Iron Age/Megalithic to Early
Historic transition display a great deal of continuity. As the central Indian Early
Historic does not appear to represent a complete transformation from previous
Megalithic society, it should be explored within the context of existing social
complexity (Morrison, 1995:207-208). According to the excavators at Adam, the pre-
Early Historic periods presented brick structures, tiles, ring/brick wells and a public
drainage system, while at Bhon there were also extensive structural activities,
including a brick canal and an early stupa thought to date to the 3rd century BC
(Deotare, et al., 2007:184; Mahapatra, 1994:63; Mahapatra, 1995:45). However, there
was an expansion of settlement in Vidarbha during the Early Historic period and
archaeological evidence reveals an increasing number of features which suggest the
evolution of “unambiguously” urban sites for the first time (Chattopadhyaya,
2003b:28; Parashar-Sen, 2008:329). Archaeological, numismatic and epigraphic
evidence suggests that throughout the overarching ‘Mauryan’ period dominant local
powers ruled Vidarbha, thus suggesting a form of local administration was
responsible for coordinating exchange and construction.
Following the collapse of the Mauryans, several powerful ruling families emerged
across the subcontinent during the first centuries AD (Ray, 1987:97). Within central
India and Vidarbha, the Sātavāhanas appear to have risen from a local power base to
form a significant polity from the mid to second half of the 1st century BC until the
early 3rd century AD (Morrison, 1995:209). The Sātavāhana period was the first time
that indisputable evidence of large urban sites, state authority and established
religious practice can be extended to central India (Morrison, 1995:210). Much is still
unknown about Sātavāhana administration, economy, territorial extent and
chronology. Further archaeological research is necessary for a better understanding of
41
Chapter 2. History of Research
It is clear that there was a significant relationship between Buddhist monastic sites
and major routes of movement and trade; Buddhism appears to have been highly
integrated into the Sātavāhana economic system and was involved in social and
institutional change (Ray, 1986:87). Under the Sātavāhanas, the spread of both
Brahmanism and Buddhism promoted new types of worship and burial practices, and
it is thought that at this point that the Megalithic tradition declined (Ray, 1999:203).
Despite Sātavāhana allegiance and patronage to Brahmanism, Buddhism prospered
during their reign through the donation of non-permanent land-grants to religious
beneficiaries, leading to a great increase in the number of monastic sites (Sinopoli,
2001:171). Political, social and economic motivations have been suggested for the
state support of Buddhism, in particular connected to the creation and strengthening
of regional identities and a shared ideology (Morrison, 1995:210). Land-grants to
Buddhist institutions may have aided in the expansion of trade, intensification of
agriculture and legitimisation of Sātavāhana rule (Ray, 1999:203; Sinopoli,
2001:172). Sātavāhana patronage to Buddhism is more visible in western
Maharashtra than in Vidarbha, and it is in the West that there were more numerous
pre- or early Sātavāhana Buddhist sites (Parashar-Sen, 1999:169). Evidence of
42
Chapter 2. History of Research
Hinayana Buddhism in Vidarbha from around 3rd to 1st century BC is known from
sites such as Pauni (Deo & Joshi, 1972). Yet despite the clear Buddhist affiliation of
some sites no Sātavāhana inscriptions have been identified in Vidarbha, in contrast to
the numerous epigraphic records found in the Western Deccan (Deo & Joshi, 1972;
Meister, et al., 1988).
Despite the Sātavāhana dynasty claiming hegemony over the Vidarbha region, as
demonstrated in records of their territory in inscriptions, it is unclear as to what extent
the Sātavāhanas had control and directly influenced development in the region. The
large number of Sātavāhana coins found in Vidarbha is not sufficient to corroborate
claims of active or direct authority over this region. Furthermore, numismatic
evidence, which refers to local ruling families such as the ‘Mitras’ and ‘Bhadras’,
occurs in the same stratified contexts as Sātavāhana coins. Therefore, it seems that
smaller local powers continued to have some authority after the Sātavāhanas came to
rule (Chattopadhyaya, 2003b:30). The extent to which pre-existing urban elements
were furthered by Sātavāhana activity is also debatable. A developed form of
urbanism in Vidarbha under the Sātavāhanas has only been determined at a limited
number of sites, such as Adam and Pauni, and few others show significant evidence
of long-distance or maritime trade (Sawant, 2012:152). Within Vidarbha, it may be
that Sātavāhana rule was too short-lived to promote any significant socio-political or
economic change from the preceding period, or that the Sātavāhanas exploited the
resources of the region but did not directly develop Vidarbha as a part of their empire
(a similar proposition to the potential relationship between the Mauryans and central
India) (Sinopoli, 2006:331; Sugandhi, 2008).
The chronology and organisation of the Sātavāhana dynasty is still contentious and
epigraphic evidence suggests the picture of their rule was complex with different
members of the family maintaining contemporary rule over “provincial centres”
(Falk, 2009). Numismatic evidence suggests that at least three rulers of the
Sātavāhana dynasty reigned within Vidarbha after c. 225 AD when Sātavāhana
dominance in other regions collapsed (Shastri, 1987:43). Following the decline of the
Sātavāhana state, local rulers again grew in power, leading to a complex of
fluctuating “political entities” prior to the rise of the Vākāṭaka dynasty (Majumdar &
Altekar, 1967:94; Smith, 2001b:19). Limited archaeological data is available to
elucidate the socio-economic and political position in Vidarbha prior to Vākāṭaka
43
Chapter 2. History of Research
rule, leading to little understanding of the conditions which led to their rise to power
(Sharma, 1987:179-182).
The Vākāṭakas were the dominant power in central India from the mid-3rd to the end
of the 5th century AD and were the first central authority to rule over and develop
Vidarbha as a ‘core’ dynastic seat. The Vākāṭakas were the key partner kingdom of
the Gupta Empire, which was gaining power in the Gangetic plain to the north during
the same period (Bakker, 1989b:467-468). The Gupta-Vākāṭaka period is
characterised by state formation and imperial and regional development. Great socio-
economic and political transformations occurred alongside agricultural expansion,
investment in structural religious architecture and “Brahmanisation” of peripheral
areas (Kapur, 2006:14). However, research is still debating the nature and character
of state formation during this period, including that of the most powerful dynasty, the
Guptas. Virkus has suggested that the Gupta Empire was “loosely structured”, rather
than an expansive, tightly centralised imperial power. This decentralised structure is
suggested to have been a “deliberate policy” to extend indirect control over a wide
area rather than being connected to instability (Virkus, 1992:255). The success of the
Gupta Empire in maintaining this control was connected to a system of great
“political and administrative uniformity” so that regional socio-economic and
political structures were subsumed within a strong state apparatus of institutions
directed by the Gupta kings (Virkus, 1992). The inscriptions of the Guptas and their
partner kingdoms suggest that the subordinate kings operated autonomously,
implementing their own state apparatuses, but generally were restricted to governing
smaller territories. The nature of the relationship between the Vākāṭakas and the
Guptas has been a source of ongoing debate; certainly, the Vākāṭakas, Guptas and
other contemporary dynasties including the Kadambas of Karnataka and the
Viṣṇukuṇḍins of Andhra Pradesh were tied by matrimonial relations (Kulke, 2004:1).
44
Chapter 2. History of Research
our understanding of the broader political context of the Vākāṭaka kingdom prior to
their establishment in Vidarbha is not as developed. The Vākāṭakas are thought to
have expanded southwards from around the mid to late 3rd century AD, with their
initial move from their dynastic origin suggested to relate to the expansion of the
Guptas to the north (Bakker, 1997:10; Shastri, 1997:221).
The different availability and quality of evidence for the Western and Eastern polities
has partly directed our current understanding of each branch, and there are few
comparative studies beyond those attempting to construct king lists and the overall
dynastic chronology. Studies on the Western Vākāṭakas provide less information
45
Chapter 2. History of Research
about their administration and attention is mostly drawn to their artistic prowess at
Ajanta, while economic studies are almost exclusively focused on the Eastern
Vākāṭakas due to the higher numbers of copper-plate inscriptions. Alongside a lack of
understanding of Eastern and Western Vākāṭaka relations, fairly little is known about
the relationship between the Vākāṭakas and other regional powers, but inscriptions do
not suggest any rivals existed in Vidarbha (Bakker, 1997:14; Shastri, 1997:221).
Vindhyaśakti I
Pravarasena I (saṃrāj)
275-335 AD
Vatsagulma ? ? Nandivardhana
Sarvasena I Gautamīputra
Rudrasena I
Vindhyaśakti II Pṛthivīṣeṇa I
360-400 AD AD 360-395
Devasena Narendrasena
450-460 AD 457-470 AD: Independent rule;
461-475 AD: Subordinate to Vatsagulma
Hariṣeṇa Pṛthivīṣeṇa II
460-478 AD c. 475 AD
478-495 AD: independent rule
Figure 2-1. Genealogy of the Vākāṭaka dynasty with key dates, after Bakker (1997).
Sarvasena I was the founder of the Western Branch based at Vatsagulma in District
Akola of Vidarbha. Vatsagulma appears to have been the only state capital and was a
pre-existing site for pilgrimage (Shastri, 1997:197). Mirashi notes that under the
46
Chapter 2. History of Research
Vākāṭakas, the site was elevated to become a “great centre of learning and culture” as
well as one of religious significance (Mirashi, 1963:xxix). This area of Vidarbha
formed an important link between the North and the South of India, and the vital
trade routes passing through would have brought traders with Buddhist interests
(Meister, et al., 1988:62). The Western Vākāṭakas appear to have practised religious
tolerance and did not restrict patronage within their kingdom so that merchants and
officials were able to fund Buddhist monasteries (Bakker, 1997:45). Furthermore, the
Śaivite king Hariṣeṇa has been accredited with the development of the famous
Buddhist caves at Ajanta (Meister, et al., 1988). However, no structural temples
attributed to the Western Vākāṭakas has been conclusively identified (Meister, et al.,
1988:75).
Based on the quality and quantity of archaeological and epigraphic remains it seems
that the Western branch was less influential or subordinate to the Eastern branch of
the Vākāṭakas until around the middle of the 5th century AD (Bakker, 1997:15). The
epigraphic evidence related to the Western Vākāṭakas is particularly limited. A set of
plates were issued by Vindhyaśakti II from his capital at Vatsagulma and record the
donation of a village, Nanded, in Maharashtra (Mirashi, 1963:xxix). However, the
only evidence for the successors to his throne can be gleaned from genealogical
portions from an inscription of Devasena (Bakker, 1997:31). Devasena issued the
Bidar plates which record a land-grant and reveal the Western Vākāṭakas’ “growing
importance and self-confidence” later in their dynastic history (Bakker, 1997:33).
Devasena was also responsible for the Hisse-Borala inscription which is particularly
significant in providing the sole chronological reference which can be related to
known periodisation (Bakker, 1997:32). Dated to 380 of the Śaka Era, around 457-8
AD, the inscription records the construction of a water storage area called Sudarśana
(Shastri, 1992a). The increasing prosperity of the Western Vākāṭakas around this time
is recorded in the Ajanta cave inscription of Varāhadeva, which notes the dedication
of a cave to the Buddhist Sangha (Bakker, 1997:34). Interesting information has been
gleaned from the Ajanta inscription as it is claimed that Devasena entrusted the
governance of his kingdom to officials (Mirashi, 1963). Western Vākāṭaka texts are
dissimilar to the Eastern Vākāṭaka inscriptions as they tend not to extol the kings’
leadership but rather indicate that courtiers and ministers managed day-to-day
government so that the king led a more leisurely life (Bakker, 1997:44).
47
Chapter 2. History of Research
Devasena was succeeded by Hariṣeṇa (c. 460 AD) who is the last known king of the
Western Vākāṭakas. Hariṣeṇa practised a policy of expansion and his only inscription
demonstrates a tonal change to praise of the king’s military prowess and his ability to
conquer the surrounding lands (Bakker, 1997:34). Hariṣeṇa’s expansion allowed him
to take control of the North-South caravan route and to increase the Western
Vākāṭaka wealth by “plundering” rich territories (Bakker, 1997:45). Furthermore, his
authority appears to have been extended over the agricultural and resource-rich areas
of the Eastern Vākāṭaka kingdom, where king Narendrasena is theorised to have lost
his sovereignty (Spink, 2005:7). The period of Hariṣeṇa’s reign is when the Western
Vākāṭakas are particularly known for their religious and artistic investments through
the excavation of caves at Ghatotkacha, Bagh and Ajanta. Several scholars have
claimed that the Mahayana developments at these sites were stylistically comparable
and therefore contemporaneous (Meister, et al., 1988:74). Bakker theorises that after
the loss of power in the Eastern kingdom, artisans sought new patronage and may
have migrated towards Vatsagulma, where they contributed to the development of the
Ajanta caves (Bakker, 2008c). There are similarities between the temple decoration
and iconography at Rāmṭek and Ajanta (Bakker, 1997:41; Meister, et al., 1988:75).
Unfortunately, the success of the Western Vākāṭakas was not to last and they appear
to have weakened under pressure from the growing power of the Aśmakas. With the
collapse of Hariṣeṇa and Western Vākāṭaka authority, the Eastern Vākāṭakas once
more became preeminent.
48
Chapter 2. History of Research
The Mandhal temples are also the earliest identified to date, as stratigraphy indicates
occupation from the last quarter of the 4th century AD (Bakker, 1997:17).
Pṛthivīṣeṇa I succeeded his father around 360 AD and was gloriously eulogised in the
land-grant tradition, which appears to have gained importance at this point.
Pṛthivīṣeṇa I is praised for his various qualities such as courage, compassion,
truthfulness, charity, integrity, wisdom and political acumen, and is recorded as being
a pious devotee of Maheśvara (Bakker, 1997:14). Pṛthivīṣeṇa’s apparent policy of
peace is shown in Vākāṭaka texts which refer to him as having had “a supply of
treasure and army which had been accumulating for a hundred years” (Mirashi,
1963:xxi). Pṛthivīṣeṇa appears to have established peace with the Gupta king
Candragupta II following a period of conflict within the imperial territory (Bakker,
1997:15).
The Eastern Vākāṭakas' status as a regional kingdom is thought to have been raised by
the marriage of Candragupta’s daughter Prabhāvatī Gupta to Rudrasena II. Prabhāvatī
Gupta later became the regent of the Eastern Vākāṭaka kingdom following
Rudrasena’s death (c.405AD to c.419AD) and continued to rule alongside her son
Pravarasena II until approximately 20 years into his reign (Kulke 2004:8). During this
period, scholars present the Eastern Vākāṭaka and Gupta relationship as generally
amicable despite the balance of power resting with the Guptas; Bakker believes that a
cordial relationship “contributed enormously to the material and cultural flourishing”
of the Eastern Vākāṭakas (1997:15). As Prabhāvatī Gupta was the daughter of
Candragupta and a princess of the Nāga dynasty, this marital relationship created
political ties between the Vākāṭaka kingdom in Vidarbha, the Guptas in the north and
the previous territory of the Nāgas around Udayagiri-Vidisha (Bakker, 2010:465).
Kulke has proposed that Gupta influence encouraged key developments within the
Vākāṭaka kingdom, such as the increased use of copper-plate inscriptions and the
foundation of Rāmṭek as a state sanctuary. Bakker has suggested that Prabhāvatī
Gupta was influenced by the Gupta reshaping of Udayagiri into a holy place under
Candragupta II, and sought to construct her own “holy mountain” on Rāmṭek hill
(Bakker, 2002:3; Willis, 2009).
It is unknown how long Padmapura remained capital, but in the 13th year of
Prabhāvatī Gupta’s regency, inscriptions point to a new capital at Nandivardhana,
49
Chapter 2. History of Research
identified with modern Nagardhan close to the central religious site of Rāmṭek.
Despite their initial prominence, the Eastern Vākāṭakas established their new state
capital at Rāmṭek in, what is often considered, a newly-developing, inland and
‘peripheral’ location, away from the long-standing established sites located along
river systems and trade routes which previously formed the centre of Sātavāhana
control. The nature of this move within the Eastern Vākāṭaka kingdom is not known,
but it has been speculated to be connected to the economic changes occurring under
their rule. Nandivardhana remained the capital until at least the 11th year of the reign
of Pravarasena II, when the charters began to be issued from nearby Pravarapura. The
majority of inscriptions are known from Pravarasena II and seem to attest to a period
of peace and prosperity, as evidenced by royal investment into the monumental
constructions on Mansar hill. Bakker suggests that the source of the Eastern Vākāṭaka
prosperity came from their exploitation of the land around Nagpur and the
Wainganga plain where their territory was focused (Bakker, 1997:23).
This period of peace appears to have ended contemporary with Skandagupta’s war of
succession in the North and alongside shifts in power across the regional dynasties.
The Rāmṭek inscription accounts event where Pravarasena II ventured across the
Narmada to forcibly return his sister, Atibhāvatī, to her dynastic home following the
death of her husband, Ghatotkacagupta, a rival to Skandagupta. Prior to this
expedition to the North, Pravarasena seems to have attempted to secure his realm
through developing an alliance with the ruler of Kuntala, an enemy of Vatsagulma
(Bakker, 1997:27). Shortly after, Pravarasena II was succeeded by his son
Narendrasena, who appears to have followed a more forceful and expansionist policy.
The Balaghat inscription of his son Pṛthivīṣeṇa II claims that Narendrasena
“subjugated the enemies” and gained areas of eastern Malwa, which had previously
been under Gupta administration (Mirashi, 1963:xxv). At the beginning of his reign,
Narendrasena seems to have been taking advantage of the advances of previous rulers
and the breakdown of Gupta relations, however according to the Mandhal Inscription
of Pṛthivīṣeṇa II, “the royal fortune of Narendrasena's house was taken away by a
kinsman” (Mirashi, 1963:xxv). This is assumed to refer to Devasena and later
Hariṣeṇa of the Western branch. Archaeological and architectural evidence goes some
way to support this loss of power as following high levels of visible earlier material in
50
Chapter 2. History of Research
the Eastern kingdom, there was a swift reduction in remains concurrent with
Narendrasena’s reign (Bakker, 1997:40).
During the 4th and 5th centuries AD, copper-plate land-grants occurred in significant
numbers as part of a land donation process (Ali, 2004). Formal land-grants were not a
new phenomenon; they existed in both the Mauryan and Sātavāhana periods, and
there are references to very early grants which may have involved a pre-existing
51
Chapter 2. History of Research
Early studies of the Vākāṭakas frequently referred to their great imperial status based
on epigraphical rhetoric which applauded the virtues of the kings and the kingdom’s
successes (Bakker 1997:14). However, it would be wrong to make assumptions about
the extent and power of the Vākāṭaka kings using this information alone and in reality
the available evidence supporting such a proposition is relatively limited (Misra,
1987:644). If the sites where inscriptions were found are an indicator of the extent of
the kingdom occupied by Eastern Vākāṭakas then it appears that their reach was fairly
contained within the fertile land of the Wainganga plain, which underpinned their
economy (Bakker, 1997:23).
It has been established that the dynastic king led the government, and that the
Vākāṭakas followed a hereditary right to kingship (Altekar, 1960; Ghoshal, 1954).
The Vākāṭakas appear to have followed the guidance of the Arthaśāstra, which
encouraged the king to “adopt the customs, dress and language of the region he
rules”, respect the local rituals and deities and then grant land, benefits and
exemptions to learned and pious men (Willis, 2009:159). Copper-plate inscriptions
are informative about political administration as they mention various categories of
officials, but there are scarcely any references to the local or regional administration
52
Chapter 2. History of Research
Etymology suggests that the Vākāṭakas founded village settlements in the Eastern
territory on a fairly large scale which may have resulted in a rise in agricultural
production (Misra, 1987:645). The attention paid to record-keeping and the
delineation of village boundaries in the copper-plate charters suggests the importance
of land and territory (Maity, 1970:45). Researchers have established that the
Vākāṭakas had at least a basic system of land measurement (the nivartana) and a
means of “territorial administration” and revenue collection (Misra, 1987:644). There
has been much discussion attempting to determine settlement hierarchy based on size
as inferred from the various suffixes on place names in the inscriptions. A schema has
been loosely identified as the charters suggest a territory was divided into rāshtras
(divisions) then vishayas (districts), which were subdivided into āhāras or bhogas
(cities, towns and villages). Cities and towns generally ended with -nagara or –pura
while villages were primarily -grama (Mirashi, 1963:xxxv). However, there has been
difficulty identifying sites and relating epigraphy to archaeological evidence (Misra,
1987:643).
There is little information about village administration in Vākāṭaka records and few
references to local people within the settlements beyond the identification of
cultivators (Altekar, 1960:194). The mention of śūdra cultivators indicates that
agriculture was the primary economic activity of those villages mentioned in
inscriptions, while some of the village names indicate production (Misra, 1987:645).
It has been suggested that the villages in the inscriptions match Kauṭilya’s vision for
villages to be engaged in production for cities (Misra, 1987). Some idea of revenue is
suggested from inscriptions due to the privileges and exemptions granted to the
donees of the copper-plates; it appears that agriculture and animal husbandry were the
main economic activities in the Vākāṭaka villages, which were inhabited by
Brahmanas and cultivators (Misra, 1987:645). Many charters expressly laid down
specific conditions under which the land-grants were maintained for the Brahmana
donees. While residents were given the right to areas such as village pastures, the
53
Chapter 2. History of Research
Poona and the Riddhapur plates of Prabhāvatī Gupta and the Siwani and Chammak
copper-plates of Pravarasena II imply that mines, mineral wealth and reserve forests
were reserved for the state (Altekar, 1960:196; Maity, 1970:132). Both the Siwani
and Chammak plates refer to exemption from forced labour along with the
endowment of the land (Maity, 1970:197). Forced labour is frequently referred to in
Gupta period inscriptions and appears to have been a fairly common source of state
income or taxation.
The nature of ‘urbanism’ is not well understood during the period of Vākāṭaka rule
and it appears that fundamental changes occurred within the economy. An
astonishing paucity of coins in the archaeological record originally led scholars to
consider that the Vākāṭakas did not mint their own coinage; a claim bolstered by the
absence of references to currency in contemporary land-grants. Shrimali has
suggested that the proliferation of grants indicate that land was the foundation of the
economy and revenue took the place of monetary taxes (Shrimali, 1987:6; 1992;
Sircar, 1969:33). Shrimali believes that the land-grants would have resulted in the
increase of small-scale agricultural village settlements and a comparatively non-urban
economy (1992:103). Archaeological data appears to demonstrate a relatively low
number of coins in circulation towards the end of the Early Historic onwards, and a
steep decline in the deposition of low denomination coinage around the 4th/5th century
(Kennet, 2013:343). There is speculation that the Vākāṭakas used the coins of
contemporary or preceding rulers (Goyal, 1998:176; Shrimali, 1992:103). However,
the proposition that a monetised economy was absent seems unlikely given the
Vākāṭakas’ standing as a powerful dynasty in central India, their apparent prosperity
and their interaction with the Guptas (Raven, 2004:19). Gupta numismatic evidence is
well documented, and subordinate local powers issued their own coinage (Shastri,
1992b:256).
Raven has posited that “regional monetary traditions” can influence the decisions of a
new dominating kingdom wishing to introduce its own currency to legitimate their
authority. Thus, the Vākāṭakas may have issued coins closely resembling the lead,
copper and potin coins that already existed in the Sātavāhana territory they inherited
(Raven, 2004:20). Shastri has suggested that lightweight, base metal coins were used
in routine ‘day-to-day market transactions’ and may have supplemented a cowrie
currency as used in the Gupta Empire. Shastri opined that base metal coinage may in
54
Chapter 2. History of Research
fact represent “deep penetration” of a monetary economy and large-scale use of coins
(Shastri, 1992:291). However, as yet there is no evidence of precious metal coins,
which would be necessary in higher trade, and the relative scarcity of Vākāṭaka coins
in archaeological contexts remains unexplained. Shrimali has noted that it is possible
that the Western Vākāṭakas inherited trade links while the Eastern Branch needed to
adapt to their more peripheral location and construct a rural economy. It has been
suggested that trade was declining in Eastern Vidarbha around the 5th century AD and
trade is not mentioned in the available inscriptions (Raven, 2004:19; Shrimali,
1992:108). This is contrary to the earlier Buddhist phenomenon where the expansion
of Buddhist centres in West and Central India was “intimately linked” to a developed
trade and patronage system. Few sites under the Eastern Vākāṭakas have been
recorded as displaying distinctly urban traits and alongside the shift in settlement
pattern to more rural areas, there is a strong case for the “promotion of a small-scale
agrarian village economy” (Shrimali, 1992:107). Bakker has suggested that the
Eastern Vākāṭakas had a rural economy without much emphasis on long distance
trade as economic surplus would have been channelled into supporting royal and
religious activities (Bakker, 1997:23).
Bakker believes the Eastern Vākāṭaka policy of donating land to Brahmanas was
instrumental in the revitalisation of the Brahmanical tradition and the spread of
Vedic/Śāstric rituals in the region under their control. The land-grants clearly
demonstrate a developed system of religious patronage and the instillation of a
“devotional ethos” (Bakker 2010:472). The intensification of temple-building on a
large scale during this period is connected to economic and political changes and the
Brahmanical revival. The cult of the temple became increasingly popular and temples
began to be constructed in durable materials (Altekar, 1960:199). Buddhism,
Hinduism and Jainism began to accumulate permanent endowments at this time and
religious and political elites emerged as advocates and protectors of the new temple-
centred order. Concurrent with the peak of temple architecture and patronage,
extensive networks of religious interaction appear to have developed across the
country. However, the relationships between the dominant Brahmanical tradition,
Buddhism, Jainism, and periphery “tribal” (now Adivāsi) communities are not well
understood.
55
Chapter 2. History of Research
Under the Guptas and Vākāṭakas, the Brahmanical iconography developed, resulting
in the worship of Gods and Goddesses which are identifiable in modern Hinduism.
As religious iconographies were standardised, they were supported by vibrant artistic
visualisation and Gupta art and sculpture in particular is considered to represent the
zenith of ancient Indian art (Goyal, 2006:27). This artistic peak is reflected in the
vivid flourishing of art which occurred in central India under the Vākāṭakas.
Vākāṭaka sculpture bears close similarity to the northern Gupta tradition but with
certain nuances specific to the region of their control (Bajpai, 1989:103). It is this
artistic “efflorescence” in sculpture, architecture and literature, which often leads to
the portrayal of the Gupta-Vākāṭaka period as a Classical or ‘Golden Age’ in Indian
history (Thapar, 1966:136).
Sharma’s urban decay theory (1987b) has been influential in the interpretation of
Vākāṭaka remains as the first stage of a twofold process of decline was theorised to
encompass their reign, with the second following their collapse. Sharma originally
interpreted the absence or degeneration of urban attributes in Early Historic contexts
as demonstrative of socio-cultural deterioration and the desertion of sites (Sharma,
1987b). To support this theory, the period of decline was proposed to coincide with
descriptions of the breakdown of social order during the tumultuous ‘Kali Age’ in
ancient literary sources, such as the Purāṇas (Jha, 2000; Sharma, 1987a; 1987b:139).
Considerable emphasis has been placed on the reduction of international trade with
56
Chapter 2. History of Research
the Mediterranean after the 3rd century AD, which is theorised to have resulted in
restricted income for Indian urban centres (Kennet, 2013:337).
Several scholars have suggested that this focus on the external influence of dwindling
international trade undervalues the strength of the indigenous Indian economy and the
extent of inter-regional trade (Kennet, 2013; Smith, 2002a; Thakur, 1993:98). Recent
archaeological evidence suggests that trade with Western Asia and the Gulf did not
suffer during the period in question and that the regional trade apparatus persisted
(Kennet, 2013:349; Nandi, 2000:19; Smith, 1999). Furthermore, rather than decline
and demonetisation, a change in coin minting, use and deposition could simply reflect
a different form of monetisation or may be more related to coin availability and
methodological and interpretative inadequacies (Bhandare, 2015:198; Kennet,
2013:345). It is also possible that denominisation during the Sātavāhana period may
have flooded the market with base metal coins, and thus a residual presence of
Sātavāhana coins in later stratigraphic layers may have resulted in the inaccurate
dating of Gupta/post-Gupta phases as earlier and the subsequent erroneous
periodisation of artefacts (Hawkes, 2014b:212; Kennet, 2004b:13).
57
Chapter 2. History of Research
The urban decay argument has been consistently interlinked with a wider debate on
economic feudalisation (Hawkes, 2014c:56; Jha, 2000:6; Kennet, 2013:333).
Feudalism is generally considered to encompass agricultural economies featuring
landlords and a subordinate peasant class (Heitzman, 1997:14; Sharma, 1984:17;
Sircar, 1969:32-33). The concept of Indian Feudalism was connected to the broader
political attempts of post-independence Marxists to fit India’s development into
Marx’s European paradigm and thus disengage it from previous ideologies of the
Asiatic or ‘despotic’ mode of production (Chattopadhyaya, 2003a:193). ‘Indian
feudalism’ includes features such as the transfer of royal powers to landed
subordinate rulers, forced labour (visti) and subjugation of peasants (Chattopadhyaya,
2003a:186; Jha, 2000:4). The feudalisation theory suggests that the land-grants of the
5th century eroded the king’s sovereignty as rights over the land were relinquished
through donations (Sharma, 1991:340). It has been suggested that village economies
subsequently became closed, resulting in the decline of towns in the Vākāṭaka
kingdom and a reduction in artisanal production and exchange (Ali, 2012:8). As self-
sufficient villages expanded and the urban sector contracted, it has often been claimed
that Brahmanas and skilled workers/artisans migrated to rural areas, which finds
support in the grants of rural villages to Brahmanas and temples (Chattopadhyaya,
2003a:183; Sharma, 1987b:155).
58
Chapter 2. History of Research
power and a strong royal hold over the remaining territory (Kapur, 2006:31). There is
little to suggest that land-grants undermined political or royal authority, particularly
as the majority recorded donations to religious institutions (Heitzman, 1997:14). This
tradition was employed by dynasties across the subcontinent during the Gupta and
post-Gupta period, which would have been counterintuitive if the practice worked to
destabilise a state’s political and economic basis (Sahu, 2012:154). Furthermore, the
land-grants generally involved small areas of land used by communities to maintain
institutions and a relatively limited number of inscriptions have come to light despite
the size of the kingdoms and their long duration of reign. As opposed to being a
substantial economic investment, the land-grants may have represented a high level
transaction, which would not have significantly diminished state resources (Singh,
2009:167). As markers of social and economic change the importance of the grants
may have been overstated, but unfortunately we have no basis for comparison in
order to establish the social value of such transactions.
It has commonly been assumed that land-grants were assigned to peripheral areas in
need of subjugation in order to extend the reach of agrarian society and bring land
under the plough (Chattopadhyaya, 2003a:191). However, recent research into the
distribution and content of land-grants suggests that the majority were donated in
areas which were already settled and cultivated. For example, the Chammak copper-
plate of Pravarasena II donated land to one thousand Brahmanas in an area with an
established settlement (Maity, 1970; Singh, 2009:164). Rather than generating new
agricultural areas, grants to large groups of Brahmanas may have been made to
consolidate or stabilise state power, encourage the maintenance of an existing
agricultural zone and integrate local groups into Brahmanical society (Kapur,
2006:29; Kulke, 1978:133; 1993:5; Lubin, 2005:97). The land-grants supported the
permanent foundation of religious institutions, which would legitimise royal authority
and connect rural areas to regional networks (Sahu, 2001:14; Sharma, 1991:342). The
distribution of discovered Vākāṭaka copper-plate charters indicates that the majority
were geographically localised within dynastical core areas; for example, Pravarasena
II did not issue any grants beyond Vidarbha, which suggests a process of territorial
integration and legitimation during Vākāṭaka state formation (Kapur, 2006:21). The
introduction of Brahmanas into existing local social and economic structures would
have altered the structure of the village community, as the Brahmanas were clearly
59
Chapter 2. History of Research
economically dominant and socially superior due to royal patronage and assistance
(Singh, 2009:167-168). Therefore the spread of Brahmanas related to the
reinforcement of the religious and social order which underpinned the authority of the
Eastern Vākāṭaka kings (Bakker, 1997:24).
2.2.4. The Decline of the Vākāṭakas and the subsequent Early Medieval to
Medieval development of Vidarbha
The Vākāṭakas reached their zenith around 510 AD, but then swiftly disappeared
within about 40 years. No successors are known following Hariṣeṇa of the Western
Vākāṭakas or Pṛthivīṣeṇa II of the Eastern Vākāṭakas (Altekar, 1960:191). In the east,
it appears that Pṛthivīṣeṇa II came under attack from the Nalas and sought support
from the Viṣṇukuṇḍins during this conflict, but his reign does not appear to have
persisted long after this event (Bakker, 1997:56). The circumstances of the Western
Vākāṭaka decline are not well understood, although a weak rule by Hariṣeṇa is often
cited as being responsible. As noted above, following the collapse of Vākāṭaka rule,
the traditional model of development in the region outlines the abandonment of sites
and overall decline. Archaeological reports from a number of Early Historic sites,
including Adam, Kaundinyapura, Mandhal and Pauni, suggest a gap in occupation
until the Medieval period. However, our archaeological understanding of the regional
Early Historic to Early Medieval transition period in this region is fraught with
uncertainty and it is unsure whether a lack of evidence is a true reflection of past
activity.
Historically, authority in Vidarbha following the end of Vākāṭaka rule fell first to the
Viṣṇukuṇḍins and the Rāṣṭrakūṭa kings (c. 750 AD), before the Cālukyas took over in
973 AD (Bakker, 1990:62). By the second half of the 12th century, the Cālukya
kingdom was subject to increasing pressure from the Kalacuri family and by the 8th
decade of the 12th century the kingdom had collapsed (Bakker, 1990:64). It is likely
that a complex of local powers ruled in the absence of a central authority but there is
limited supporting archaeological evidence. On the contrary, archaeological evidence
of the succeeding power, the Yādavas, is common across the region. The Yādavas
were originally feudatories of the Cālukyas, and took over in the mid-12th century,
with their state encompassing parts of the Konkan, Desh and Vidarbha (Sahu,
60
Chapter 2. History of Research
2012:148). The Yādavas reinvigorated structural temple architecture in the region and
invested significant patronage into temple construction (Jamkhedkar, 1985-1986:25).
The end of Yādava rule came under increasing pressure and incursions by Islamic
armies from Delhi and the dynasty appears to have succumbed to the Islamic powers
by 1294 AD (Bakker, 1990:75; Sohoni, 2011:51). Along with their collapse, Hindu
temple construction was abandoned and it appears that often the temples were
disassembled, with the occasional incorporation of architectural pieces into Islamic
buildings (Sohoni, 2011:51). Islamic rule in Maharashtra was challenged by a cultural
and religious group of “Hindu elites” called the Marāṭhās, who had ambitions to
reclaim Maharashtra and whose power intensified over the first half of the 18th
century (Sohoni, 2011:51). The Marāṭhās expanded to occupy a territory roughly
covering Maharashtra state, and under their influence Hindu architecture was
rejuvenated, incorporating clear aesthetic influences from the Islamic tradition
(Bakker, 1990:62; Sohoni, 1998:21). Vidarbha initially appears to have remained
slightly isolated from the dominant centres of Marāṭhā power, possibly due to its
geographical separation, but was brought under their authority by the Bhonsle family
(Dikshit & Singh, 1971:732). Nagpur became the centre of the Bhonsle dynasty and
attention was drawn to Rāmṭek and the development of the religious site on the hill
(Bakker, 1990:63). The Bhonsle territories were prosperous and so liberal patronage
was devoted to temple construction, resulting in a developed local architectural
tradition (Sohoni, 1998:145). Nagpur remained out of British control for slightly
longer than the rest of Maharashtra, but around 1857 AD indigenous rule was brought
to a close (Sohoni, 1998:39).
Despite certain lacunas in our understanding of the Vākāṭakas, scholars are generally
agreed that the dynasty played a ‘pivotal role’ in the politico-cultural history of the
Deccan (Shastri 1997:213). As can be inferred from the established history of the
Vākāṭaka dynasty, there has been varying degrees of research into the epigraphic, art-
historical, numismatic and archaeological evidence. The majority of research displays
a fundamental dependence on written sources and notable works have been produced
which translate and interpret the epigraphic records of the dynasty. Additionally, art-
61
Chapter 2. History of Research
historical research has been relatively abundant due to the exceptional nature of the
sculptural and architectural remains. Archaeological investigations provide
indispensable contextual information for epigraphic and art-historical sources.
However, archaeological investigations concerning the Vākāṭakas have been subject
to a number of limitations in interpretation and scope, which means there is still great
potential for archaeology to contribute new information. Moreover, there are further
opportunities for scholarship to synthesise the available evidence, and for researchers
to continue conducting collaborative archaeological and historical research.
Epigraphic evidence was the earliest known source of information about the
Vākāṭakas following the discovery of a copper-plate grant of king Pravarasena II in
Siwani (1836). Since then, around 41 inscriptions have come to light, mostly in the
form of copper-plate grants with occasional stone inscriptions at key sites. The wealth
of evidence contained within the inscriptions has attracted plentiful research. Some of
the most active scholars researching the Vākāṭakas have primarily focused on the
translation and interpretation of epigraphic sources with reference to other written
records, such as the literary works of Kālidāsa or ancient chronicles such as the
Purāṇas, to complement their research. The Vākāṭaka dynasty did not find mention in
early texts on the history of India as they were originally believed to be of Yavana or
Greek origin, despite the early decipherment of genealogical portions of the Ajanta
Cave XVI inscription in 1862 (Mirashi, 1963:I). Once their indigenous origin was
realised, scholarly attention was drawn to establishing their dynastical history,
genealogy and chronology (Altekar, 1960; Majumdar & Altekar, 1967). An early
reading of the Siwani plates the text was published in 1836 by Prinsep (1836) and the
text was again translated by Fleet, alongside the Chammak and Nachna plates (1888).
In 1914, V.A. Smith published an article on the Poona Plates of Prabhāvatī Gupta
(discovered in 1912), also translated by K.B. Pathak and K.N. Dikshit (Pathak &
Dikshit, 1925; Smith, 1914).
Following these early translations, the study of the Vākāṭakas was brought into the
sphere of general research by the work of scholars such as K.P. Jayaswal (Jayaswal,
1933) and Mirashi, whose 1939 his translation of the Basim Grant of Vindhyaśakti II
62
Chapter 2. History of Research
demonstrated the existence of the Eastern and Western Vākāṭakas (Mirashi, 1963:ii;
Mirashi & Mahajan, 1952). Mirashi later published a hugely influential volume of
translations and interpretations of 27 Vākāṭaka epigraphs (1963). In addition to these
significant readings of the texts aimed at reconstructing the Vākāṭakas' political
history, administrative details and chronology, Mirashi proposed the identification of
Rāmṭek with the literary place of the Rāmagiri, memorialised in Kālidāsa’s epic
poem, the Meghadūta (c. 5th century AD). The mythological entity the Rāmagiri is
considered to reflect a historical site; its identifications at Rāmṭek was largely based
on its geographical position, topographical descriptions and inscriptional evidence
and has fed into ongoing debates concerning the Vākāṭaka relationship with the
Gupta Empire (Bajpai, 1992:85; Mirashi, 1964:137-139). Further important 20th
century publications of the Vākāṭakas, primarily outlining their history and
chronology are works by Majumdar & Pusalker (1954) and Majumdar & Altekar
(1967).
Since Mirashi’s 1967 volume, Shastri has been responsible for the interpretation and
publication of a number of inscriptions. Shastri’s publications on the political history
of the Vākāṭakas have encouraged a holistic view of the dynasty incorporating
interdisciplinary scholarship on archaeology, numismatics and iconography in
conjunction with inscriptional evidence. Shastri’s seminal volumes on the Vākāṭakas
provide a comprehensive historical outline and a re-appraisal of earlier evidence and
research (Shastri, 1992a; 1997). These texts should be referred to during research into
the historical narrative and known socio-political development of the dynasty.
63
Chapter 2. History of Research
Despite the economic nature of the vast majority of the copper-plate inscriptions, the
nature and meaning of the land-grants and their significance within the Vākāṭaka
economy remains inconclusive. This lacuna in existing studies was recognised by
Shrimali who attempted to elucidate the Eastern Vākāṭaka economy through a study
of their copper-plate land-grants (1992:102). Shrimali’s research represents an
important development in Vākāṭaka epigraphic research, as it demonstrates a move
towards systematic and contextualised analysis of the available evidence. Shrimali’s
research reflects on the landscape of the Vākāṭakas and considers the significance of
the inscriptions as geographical and material evidence, rather than simply in terms of
their content. Preliminary work is being carried out at the British Museum to map
inscription find-spots and their content (Willis, et al., 2012). Through such methods
it may be possible to consider how the two Vākāṭaka branches and other dynasties
compare in their production of epigraphic evidence, and relate the copper-plates to
their context. Further research has begun to explore the archaeological contexts of
copper-plate charters by locating their find spots across Vidarbha in order to subject
them to archaeological survey (Hawkes, proprietary and forthcoming).
Figure 2-2. General map of the find-spots of the Vākāṭaka inscriptions. The red markers correspond to
the inscriptions of the Western branch while blue refers to the Eastern branch. (See the British
Museum project for further details. Upcoming publications contain proprietary information
http://www.britishmuseum.org/research/research_projects/all_current_projects/beyond_boundaries/ins
criptions_and_sites.aspx)
64
Chapter 2. History of Research
It is important to recognise the biases that may be present in historical sources; there
are general uncertainties related to provenance, intended impact and the nature of
texts (Shastri, 1997:149). The overwhelming majority of inscriptions belong to the
Eastern Vākāṭakas, which tend to result in deeper focus on the political organisation
and economic structure of this branch. Studies of the economy almost exclusively
deal with the Eastern Vākāṭakas, while epigraphic studies of the Western Vākāṭakas
mainly focus on donative inscriptions and the patronage and chronology of the Ajanta
caves. Bias towards the Eastern Vākāṭakas in epigraphic records may be responsible
for a tendency to over-emphasise their power and importance. The majority of
inscriptions are attributed to Rudrasena II, queen Prabhāvatī Gupta or Pravarasena II
(c. 398-455 AD) of the Eastern Vākāṭakas, which limits our understanding of
chronology (Kulke, 2004:4). A further issue arises in that the Vākāṭaka inscriptions
refer to regnal years and lack irrefutable palaeographic evidence to indicate a more
accurate date (Majumdar & Altekar, 1967). The Purāṇic texts have been used to date
the dynastic origins, and subsequent episodes in Vākāṭaka history have been based on
postulations concerning lengths of reign (Shastri, 1997:163). Currently, the backbone
of the Eastern Vākāṭaka chronology is currently found in references to the marriage
of the king Rudrasena II to Prabhāvatī Gupta, daughter of Chandragupta II (c. 376-
412 AD), as Gupta records are based on a relatable timescale (Goyal, 1992:298).
Without a dependable chronology, many theories concerning the meaning and
consequences of conjectured political events, relations to other dynasties and
potential power struggles remain indeterminate (Shastri, 1997:178).
Finally, despite their original purpose as records of economic systems and political
history, the inscriptions often do not include specific references to concepts such as
monetary systems, territorial boundaries or the nature of central government. Studies
of inscriptions are inevitably dominated by reconstruction of aspects of state
administration. However, isolated references to state officials, administrative terms or
categories of site may not be sufficient to allow reconstruction of the nature of state
formation, as is occasionally attempted in the literature (Mahajan, 1992; Shastri,
1992c). As such, attempts to reconstruct Vākāṭaka political and economic systems
often rely on conjecture and challenges in interpretation have resulted in protracted
debate. There are issues with the variety and depth of the information available in
65
Chapter 2. History of Research
inscriptions; it may be that there is simply not enough detail in the inscriptions alone
for effective discussion about certain aspects of Vākāṭaka rule.
There has been comparatively little numismatic research regarding Vākāṭaka coinage
given that for a long time no coins had been incontrovertibly discovered. All early
scholarly accounts suggested that the dynasty did not mint coins and often opined that
the Vākāṭakas allowed other currency to circulate in their kingdom. K.P. Jayaswal
was initially the only key historian to suggest coins had been discovered bearing
Vākāṭaka legends, but these claims were refuted by Altekar and Mirashi (Jayaswal,
1933; Majumdar & Altekar, 1967; Mirashi, 1963). Rather, attention was more
typically drawn to the currency of the Guptas and the growing collection of their
precious metal coinage.
Since the earlier claims that the Vākāṭakas did not issue their own coins, Shastri has
published several base metal/copper coins from Wardha which he attributed to the
dynasty. These coins, held in private collections, are described as “minute…
irregular… and light-weight” (Shastri, 1990; 1992b:256). Chandrashekhar Gupta also
believes that Vākāṭaka coins have been found in Vidarbha, including at the site of
Paunar, but were not accurately identified due to the absence of a legend and were
therefore determined to be ‘Viṣṇukuṇḍin type’ coins (Gupta, 1992:143-144).
Kulkarni has followed Shastri in publishing further details of base-metal coins
attributed to the Vākāṭakas (Kulkarni, 2003). There have been other claims of coins,
apparently with complete legends, but these have not been verified (Meshram, 2010;
Pathak & Thakur, 2005). Raven has produced a thorough overview of the current
state of affairs and has discussed the possibility that the Vākāṭakas issued coins which
resembled the lead, copper and potin coins that already existed in the Sātavāhana
territory which they inherited (Raven, 2004:20).
66
Chapter 2. History of Research
divided over Vākāṭaka coinage and scant evidence has been uncovered to aid
interpretation. The question of Vākāṭaka coinage also exists within the broader
context of research which suggests that India was demonetised during the Gupta and
Post-Gupta period. Deyell (1990) published an influential response to this arguing
that north India continued to be heavily monetised and that Post-Gupta coinage
increased in amount, albeit in a debased form and more limited in type (Ali, 2012:9).
However, subsequent scholars have noted that Deyell’s quantification still implies a
change in monetisation and the numismatic history of the Gupta-Vākāṭaka period is
still not well understood (Kennet, 2004b:13).
For a number of scholars, the defining characteristic of the Vākāṭaka period is the
artistically refined and intricate sculptures and temples, thus iconographic and art-
historical research is fairly extensive. The period of Vākāṭaka rule is often considered
a classical era, which produced distinctive and “progressive” sculpture (Brown,
2004:60; Misra, 1992:173). Analysis of sculpture and architecture principally furthers
our understanding of the more intangible aspects of the Vākāṭaka period; that of
religion, ideology and theology. Monumental architecture and sculptural art can also
be indicative of the prosperity of a kingdom, patronage and religion, as well as active
attempts by regimes to legitimise their authority over the landscape.
Art-historical studies have tended to be specialist as the art of the Eastern and
Western Vākāṭakas is considered to have been moulded by their different socio-
political backgrounds and religious affiliations. The Eastern Vākāṭaka rulers were
followers of Śaivism and Vaiṣṇavism, as reflected in the religious architecture and
sculpture at sites such as Rāmṭek, Mansar and Mandhal (Bakker, 1997:58; Deva &
Sharma, 2009; Thapar, 1979:36). Vākāṭaka art has often been referenced by scholars
67
Chapter 2. History of Research
when discussing Gupta sculpture and architecture (Harle, 1974; Harle, 1991).
Williams has considered the artistic offerings of both the Western and Eastern
Vākāṭakas, as despite their different socio-economic and religious influences, they do
share common qualities (Williams, 1982; Williams, 1983). The Western Vākāṭakas
gave liberal patronage to Buddhism, despite no indication that they themselves were
devotees, and the impressive Buddhist remains at Ajanta were developed under their
auspices (Spink, 2005). There is significant art-historical research on the Western
branch and their cave temples and architecture (Fergusson & Burgess, 1880:280-295;
Meister, et al., 1988; Yazdani, 1930-1955). Spink is a prolific scholar on the caves of
Ajanta, and has published numerous volumes on the architecture, imagery,
chronology, political background and patronage of the caves (Spink, 2005; 2014).
Additionally, while not tightly connected to wider archaeological data. Spink has
begun to associate the developments at Ajanta to the history of the Western Vākāṭaka
dynasty and theories of socio-economic growth and political power.
With regard to the Eastern Vākāṭakas, Krishna Deva produced a catalogue of the
sculptures excavated from Mansar, which demonstrates the characteristic style of the
sculptural tradition (Deva & Sharma, 2009). Jamkhedkar’s detailed discussion of the
sculpture and architecture of the major Eastern Vākāṭaka sites, including Rāmṭek,
demonstrates the typical iconographic traditions and architectural remains of the
period (Jamkhedkar, 1985-1986; 1991a; 1991b). Alongside numerous journal
contributions, Bakker has provided an extensive overview of the Hindu iconography
of the Eastern Vākāṭakas with a detailed catalogue of the known sculpture, and has
outlined a comprehensive history of the Vākāṭaka kingdom (Bakker, 1997:2).
Bakker’s 1997 study on the development of Vākāṭaka Hindu iconology attempts to
use art in conjunction with texts and archaeological evidence to establish the “social
and cultural context” in which art is formed (Bakker, 1997:3). This study aims to take
a further step in scholarship towards holistic consideration of the available evidence
to establish the broader context of art-historical research (along with Willis (2009),
who draws on landscape archaeology, iconography, epigraphy and ethnography in his
study of Udayagiri). Additionally, Bakker has produced edited volumes bringing
together various scholars to combine discussion of iconography, epigraphy and
archaeological evidence (Bakker, 2004; 2008c).
68
Chapter 2. History of Research
Archaeological research into the Vākāṭakas refers to both excavations and regional
explorations. Prior to the recognition that the Vākāṭakas were an independent Indian
dynasty, early surveys were carried out at sites now known to be connected to the
Vākāṭakas (Beglar, 1878; Cousens, 1887; Cunningham, 1878; Cunningham, 1972
(1866)). One of the first surveys directed towards establishing the nature of known
Vākāṭaka sites was carried out by Wellsted in 1933, focusing mainly on Mansar and
Nandpur (Wellsted, 1934). The findings of this survey have since been corroborated
by a short survey in 2006 led by Bakker. The primary archaeological work on
Vākāṭaka sites conducted across the second half of the 20th century has been by the
Maharashtra state department, the ASI, and several university departments. These
institutions have carried out a number of explorations and small test excavations at
Vākāṭaka sites or sites with contemporary remains and thus the potential for Vākāṭaka
horizons. Explorations at sites reported in IAR include those conducted at Adam,
Arambha, Bhokardhan, Brahmapuri, Hamlapuri, Kaundinyapura, Mandhal, Mansar,
Mulchera, Nagara, Nagardhan, Naikund, Pachkheri, Paithan, Paunar and Pauni.
Past archaeological studies of Vākāṭaka sites have tended to have a tight site-focus or
have involved the isolated study of major ritual or urban centres. Therefore, there is
much scope for the investigation of rural or transient remains to achieve a regional
perspective and situate sites in a broader archaeological and landscape context. An
appreciation of the wider perspective supports development of our understanding of
vital concepts such as chronology, socio-economic organisation and the process of
urbanisation. Landscape archaeology is advancing in Vākāṭaka studies and research at
Rāmṭek would benefit from consideration of its wider political and socio-economic
setting. Landscape archaeology contributes considerably towards our understanding
of the impact and significance of major urban or religious centres, and can connect
sites to evidence of state formation processes, the regional growth of kingdoms and
networks of change.
Excavated sites containing remains contemporary to the Vākāṭaka period are fairly
numerous, but fewer sites have been identified as having material culture that is
categorically attributable to the Vākāṭakas. It is commonly theorised that earlier
settlements were abandoned alongside the establishment of new ones and so
69
Chapter 2. History of Research
Vākāṭaka sites are often reported as having been ‘single culture’ (see overviews of
Mansar, Nagara Nagardhan-Hamlapuri and Washim (Deotare, et al., 2013)). Within
the Vākāṭaka territory as described by Bakker (1997) are nine fully published Early
Historic/Early Medieval sites: Bhokardhan, Brahmapuri (Kohlapur), Kaundinyapura,
Nasik, Nevasa, Maheshwar and Navdatoli, Paunar, Pauni and Prakash (Kennet,
2004b:11).
Table 2-1. The excavated sites within Maharashtra which display Vākāṭaka period remains. Nine Early
Historic/Early Medieval excavations are fully published while others are mentioned in review articles.
See Sali (1998) and Sawant (2008-2009) for an overview of further explored sites reported to have
Vākāṭaka period remains.
70
Chapter 2. History of Research
71
Chapter 2. History of Research
While several sites such as Pauni, Paunar and Kaundinyapura have been fully
published, others lack a complete site report. Small-scale excavations, surface
explorations and interim reports are often published in the Indian Archaeology
Review (IAR), and Sali has published a review article (1998). The published
evidence to data demonstrates that Vākāṭaka structural remains tend to be temples or
religious edifices, as at Mandhal and Mansar, and few town-sites have been fully
excavated or identified. Commonly site reports mention evidence of Post-Sātavāhana
activity, but these remains are either not conclusively Vākāṭaka or are described in
limited detail. Even at the major site of Pauni, the Vākāṭaka period settlement remains
are described as low quality structures made from re-used building materials with a
paucity of associated ceramic evidence (Nath, 1998:14). Paunar is one of the few
excavated and published settlement sites with Vākāṭaka habitation remains which
were described as being more numerous, “varied… and rich in conception” than the
remains of other periods (Deo & Dhavalikar, 1968:114).
Past archaeological investigations into the Early Historic to Early Medieval transition
have been heavily influenced by models such as Sharma’s ‘Urban Decay’ theory of
economic decline (1987b). Published excavation data has often been interpreted
through the lens of urban decline, and so the critical exploration of processes relating
to the Vākāṭakas, such as political administration and economy, has not been the
focus of research. Archaeologically there is still much to learn about the formation,
development and influence of the Vākāṭaka kingdom and the political and socio-
economic landscape of Vidarbha (Kapur, 2006:13). Furthermore, there is still great
potential for a complementary archaeological perspective to connect the extensive
epigraphic, architectural and art-historical evidence to the economic, political and
ideological history of the Vākāṭakas.
There have been excellent studies connected with the Vākāṭakas, and certain aspects
of their history have received thorough academic attention. The majority of
information has been gleaned from epigraphical sources and the numerous
inscriptions of the Eastern Vākāṭaka kings in particular have been used in attempts to
unravel the dynastical history and chronology. However, the overall period is not well
72
Chapter 2. History of Research
As the centre of the Eastern Vākāṭaka kingdom, Rāmṭek and its nearby dynastic
‘capitals’ are considered to present an archaeological case study for peripheral
landscape development. It is hoped that any patterns on the specific development of
this area emerging from the survey data, may be related to larger debates across India
on the expansion of settlement and agriculture, used to comment on the possible
nature of Eastern Vākāṭaka polity and kingship and compared to the evidence for
contemporary developments to the west. The overall development of the Rāmṭek
landscape based on existing evidence usually presents regional prehistoric settlement
followed by limited initial Early Historic activity in the specific area, outside of the
purported Buddhist investments at Mansar. Sātavāhana and Buddhist activity appears
to have been primarily drawn to areas south of Rāmṭek near the major river systems.
It is frequently inferred that the arrival of the Vākāṭakas at Rāmṭek stimulated the
area, and the establishment of the religious centre and capital cities, altered the
structure of rural settlement and encouraged a burgeoning agrarian economy to
support a growing population.
73
Chapter 2. History of Research
the collapse of the Vākāṭaka dynasty, until much later when the area was under
Yādava control from the 12th century onwards the renewal of building activity and
investment in the landscape. In reality, the regional development may have been far
more complex and archaeological survey is aimed at enabling us to investigate the
expected model of development. A holistic landscape archaeology approach may be
able to identify further evidence of pre-Vākāṭaka settlement activity. Landscape
survey supports investigations into the nature of settlement and attempts to identify
Early Historic habitation and religious evidence in order to contextualise the proposed
transformations under Eastern Vākāṭaka rule.
74
Chapter 3. The Survey Region
3.1. Introduction
Having considered the regional historical development of Vidarbha, this chapter will
initially consider the geographical setting of the survey area by summarising the
physical environment of Maharashtra, with particular attention to Vidarbha, where
Rāmṭek is located. A brief synopsis is presented here, primarily compiled from the
works of Deshpande (1971) and Dikshit (1986); the reader is referred to these texts
for a more complete overview. A summary of the expected contemporary cultural
situation and anthropogenic landscape, as a result of the historical and geographical
context, will be noted as the modern setting presents specific challenges which
inform the fieldwork methodology. An understanding of the current landscape is
appropriate for a more complete consideration of the past.
The second part of this chapter provides a summary of the known archaeological,
sculptural, numismatic and epigraphic remains at the three sites of great significance
within the defined survey area; Rāmṭek, Mansar and Nagardhan. The archaeological
study of these sites has varied in intensity, with Mansar attracting the most
archaeological attention in recent years due to its monumental religious and secular
constructions. Additionally, the known archaeology of the rural hinterland will be
introduced as a number of small sites have been reported from explorations in
District Nagpur, with some falling within Rāmṭek Tehsil. These details from the
local archaeological context of the survey area provide context for the later survey
findings and begin to indicate how the landscape has been interpreted to show the
relationship with the ruling dynasty.
This chapter concludes with a comment on the expected nature of the archaeological
landscape of Rāmṭek tied to the historical narrative being tested in this study. Once
this has been established, it is possible to theorise how this model would be
manifested on the ground and how landscape archaeology techniques may be
employed to explore this.
75
Chapter 3. The Survey Region
This survey focuses on the landscape in the immediate vicinity of the site of Rāmṭek
(around 21°23’56N and 79°17’51E), which is located in the Nagpur District of the
Vidarbha region of Maharashtra. The state of Maharashtra extends over an area of c.
307,713 km² in central India and displays great physical, economic and cultural
diversity (Deshpande, 1971:3; Rajaguru & Joshi, 2013:31). The Vidarbha region
(Lat. 19 21’N and Long. 76 80’ 30’’ E) is located towards the east of Maharashtra.
While Western Vidarbha is physically characterised by high plains and hills
surrounding the central Purna Valley, Eastern Vidarbha is predominantly of low
elevation with occasional hills. The eastern portion of Vidarbha largely encompasses
the Wardha-Wainganga basin and includes the Rāmṭek Upland on the northeast side
of the Nagpur plain, formed by the Kanhan River (Dikshit, 1986:187). As it is
enclosed by hills on three sides, the Wardha-Wainganga basin appears to have
developed as a “distinct regional entity” within Maharashtra (Dikshit & Singh,
1971:732).
76
Chapter 3. The Survey Region
Figure 3-2. The position of Vidarbha within the state of Maharashtra. The locations of Rāmṭek and
other major archaeological sites are indicated.
77
Chapter 3. The Survey Region
The east of Vidarbha, including Nagpur District, has prospered due to the large and
small-scale mining industries supported by the mineral rich environment with access
to coalfields, iron ore, mica and clay deposits (Chakravorty, 2001:7; Chari, et al.,
1975:249; Deshpande, 1971:45). There are several large manganese mines (M.O.I.L.
Nagpur) across the Rāmṭek area, with the most extensive site located on the
peripheries of Mansar (Krishnan, 1982:153). Active and abandoned mines within the
survey area can be found in Beldongri, Nagardhan, Lohadongri, Kachurwahi,
Waregaon, Khandala, Bhandarbodi, Mandri, Manegaon, Borda-Kumhari,
Bhodewada, Musewadi and Kandri (Chari, et al., 1975:221-228). Brick manufacture
results in extensive digging for clay, and there is further extraction of ‘Murrum’
(soils formed from weathered rock), which is commonly used as a foundation
material (Ketkar, 1970; Muley, et al., 2010:4896). Extensive mining in the survey
area has caused large amounts of disturbance and has restricted access to sections of
the Rāmṭek landscape. It is unclear how long small-scale mining has been occurring
but archaeological reports from the area demonstrate the use of mica in ceramics (see
chapter seven) and often refer to the use of local clay in production. Production sites
and material remains in the region demonstrate the use of local resources in the past,
such as sandstone in monumental constructions (Smith, 2001b:80), and the inter-site
exchange of resources such as iron (Mitra, 1983a; Thapar, 1980b).
78
Chapter 3. The Survey Region
Figure 3-4. Map of the survey area indicating the mines, land-use, earthworks and tanks that appear on
topographic maps of the area. A number of the mines indicated appear to have expanded since the
topographic map was produced. All land outside of the forest cover is simply designated as ‘open
scrub’ and in practice is largely agricultural land or areas of culturable waste.
Figure 3-5. Images showing the extent of quarrying and manufacturing activities in the survey area: a)
Satak manganese mine. b) Brick manufacturing north of Rāmṭek
79
Chapter 3. The Survey Region
80
Chapter 3. The Survey Region
In the survey area, agricultural access to water is dominated by the modern canal
network from Khindsi reservoir, built on the River Sur, which supports the
agricultural landscape south of Rāmṭek hill (Dikshit, 1986:187). In addition, there is
a tradition of smaller tanks, both modern and historical, and wells are prominent
features in villages (Sutcliffe, et al., 2011:784). Tributaries of the Wainganga, Pench
and Dawanthari rivers flow around the Rāmṭek hills and this creates an appropriate
environment for effective water management (Dikshit, 1986:187).
The modern anthropogenic landscape has been moulded by the physical environment
and the historical development of Vidarbha. Vidarbha’s position in central India
ensured there was a mix of cultural influences through the interaction of different
groups of people from surrounding regions (Dikshit, 1986:104). Marāṭhā culture has
in particular infused into Vidarbha from the west, and there has been significant
contact with Hindi-speaking people to the north in Madhya Pradesh (Dikshit &
Singh, 1971:732). The current culture, which has resulted from this history of
interaction and exchange, is popularly known as “Varhadi” culture (Sawant,
2012:14). The religious affiliation of the modern population reflects the historical
development as it is dominated by Hinduism (Dikshit, 1986:105). Almost the entire
population of Buddhists around the Nagpur area are ‘Neo-Buddhists’; as followers of
the late Babasaheb Ambedkar, Neo-Buddhists are distinguished from the original
followers of Buddhism and are akin to Hindus in their social organisation and moral
codes (Dikshit, 1986:105). A little over 5% of the population consists of Adivāsis,
distinguished by their beliefs, customs, dialects, and absence of firm adherence to
institutionalised religions (Dikshit, 1986:109). The most recent census (Deol, 2011)
81
Chapter 3. The Survey Region
showed that Maharashtra is the second most populous state and is growing at a faster
rate than the national average. Across Vidarbha, the majority of the population lives
in small villages in regular, close spacing. The spatial distribution of villages may be
connected to the relatively easy access to water and fertile land in this region, and the
focus on labour-intensive rice cultivation which requires close management of the
agricultural land (Deshpande, 1971:35, 97; Dikshit, 1986:137). The wider state of
Maharashtra has a burgeoning urban population and within Nagpur District urban
development is concentrated at Nagpur and Kamptee which have direct rail and road
links to Rāmṭek (Deshpande, 1971:201). Rāmṭek is a developing area in terms of
infrastructure and industry and is well connected for trade (Deshpande, 1971:149;
Diddee & Wakhare, 2002:213).
The known archaeological, sculptural and epigraphic evidence from Rāmṭek, Mansar
and Nagardhan, and the rural archaeological material known from within the defined
survey area, forms the basis of this project. The impetus for broader landscape
82
Chapter 3. The Survey Region
The history of research at the Rāmṭek temple complex has primarily focused on
epigraphy or the art-historical study of the Vākāṭaka temples, with comparatively
limited archaeological enquiry around the site. Seven Vākāṭaka shrines and temples
lie on the Rāmṭek hilltop and side, in addition to a water tank thought to originally
date from the Vākāṭaka period. The temples are those of Varāha, Trivikrama,
Kevala-Narasiṃha, Rudra-Narasiṃha, Bhogarāma, Guptarāma (the cave temple) and
a cave-reclusory; four of the temples still contain the original idol (Bakker,
1989b:467). (Bakker, 1991:30). Early recording of the architecture and sculpture at
Rāmṭek was conducted by Beglar in 1873-4, followed by Cousens (1887), but both
were denied access to the inner temples and did not recognise the antiquity of the
Vākāṭaka constructions (Jamkhedkar, 1992:155). Further early notes may be found in
Hiralal’s Visit to Rāmṭek (Hiralal, 1908), the Nagpur Gazetteer. Mirashi made
references to Rāmṭek in the Corpus Inscriptionum Indicarum (Vol. V; See also
Meghadūta meṃ Rāmagiri arthāt Rāmṭek, 1959) and was the first scholar to identify
the Gupta-Vākāṭaka period origin of the Trivikrama temple.
The leading scholars working on Rāmṭek’s architectural remains, who have informed
this research, are Jamkhedkar and Bakker. Jamkhedkar has studied the architecture
(e.g. Meister, et al., 1988) and chronological development of the temples, with
limited archaeological test excavations, while Bakker has primarily focused on
iconographic and architectural research. Jamkhedkar is credited with identifying the
true antiquity of the Vākāṭaka temples, which established the date of the emergence
of structural temple architecture at Rāmṭek around 600 years earlier than previously
thought (Bakker, 1997:2; Jamkhedkar, 1985-1986:26; Jamkhedkar, 1987:217). Prior
to this, academic interest was drawn to the architectural developments of the Imperial
83
Chapter 3. The Survey Region
Yādavas (12th to 14th century AD), who were thought to have been the initiators of
structural temple construction in Maharashtra (as in Deglurkar, 1974:108). The
Vākāṭaka developments at Rāmṭek reflect the early spread of free-standing temple
architecture, which was particularly patronised by the Gupta dynasty across northern
India from the 4th century AD. The tradition of free-standing architecture, as
opposed to rock-cut architecture which occasionally persisted alongside free-
standing constructions (for example at Udayagiri), expanded the use of permanent
building materials such as brick and stone (Blurton, 1992:52). Regional architectural
styles developed on the periphery of the Gupta Empire. Accordingly the Vākāṭaka
temples at Rāmṭek combine distinctive regional traits, such as the preferential use of
local red sandstone for plinths and walls and decorative features (as opposed to the
heavy basalt used in load-bearing pillars), with aesthetic parallels to Gupta
constructions (Jamkhedkar, 1985-1986:26; Meister, et al., 1988:63; Sohoni,
1998:147).
Figure 3-6. The hill-top site of Rāmṭek with the seven Vākāṭaka period temples indicated.
84
Chapter 3. The Survey Region
Both Jamkhedkar and Bakker have published sequences for the architectural
development of Rāmṭek (please see Jamkhedkar (1985-1986) and Bakker (1997) for
a complete overview) and concur that the Vākāṭaka constructions originate from a
limited period of religious construction on the hill. Jamkhedkar identified a tentative
architectural progression supported by inscriptional evidence, and proposed that the
temples were constructed between 400-450 AD (1985-1986:26). Similarly, Bakker
has proposed a 5th century date for the monuments based on the identification of
characteristic 5th century AD iconography, decorative motifs and architectural plans,
and the palaeography and genealogical content of the ‘Rāmṭek Kevala-Narasiṃha
inscription’ (Bakker, 1990:66; Bakker & Isaacson, 1993:64).
Bakker sees the east-facing Bhogarāma temple, the west-facing Trivikrama temple
and the three oldest sanctuaries of Varāha and Narasiṃha, oriented west in a north-
south line, as one of three areas which constituted the full Vākāṭaka religious
complex spread over Rāmṭek hill. This area is located c. 350m east of the main
promontory of the hill and has been suggested to represent a memorial compound
with temples built to “transfer merit” to the dead (Bakker, 2007b:19).
Both Jamkhedkar and Bakker assign the earliest date to the Varāha temple, with
Bakker considering the structure to date to the first quarter of the 5th century AD
(Bakker, 1992a:10). The Varāha temple is simple, square and open, and is
constructed from a four-pillared maṇḍapa resting on a restored platform. The four
square rucaka pillars have plain bracket capitals and are decorated by lotus reliefs
supporting a roof over a large theriomorphic Varāha sculpture (Bakker, 1990:66).
The current ceiling of the structure consists of intersected squares apparently dating
from the Bhonsle period, however the original ceiling appears to have once had a
lotus medallion rather than a crowning āmalaka (Meister, et al., 1988:65)
85
Chapter 3. The Survey Region
one tusk, which is said to represent the Earth goddess (Bakker, 1997:138). As such,
the choice of Varāha could be connected to symbolism of royal glory as this deity
rescued the Earth by lifting her up from the sea. This positioning of Varāha as a
central icon of worship is significant in comparison to trends in the Gupta Empire, as
a large figure of a boar-headed Varāha is carved in relief in Cave 5 of Udayagiri
(Willis, 2009:41). In addition to being completely theriomorphic, the Rāmṭek Varāha
differs from the Udayagiri image in the absence of an anthropomorphic Nāga king at
his feet, which has been interpreted as a metaphorical display of power against the
Nāga dynasty in the Gupta kingdom (Bakker, 2002). Instead, at Rāmṭek the Varāha
rests upon sculpted coils belonging to the ‘cosmic serpent’, which is perhaps more
appropriate given Bakker’s association of Rāmṭek hill with Prabhāvatī Gupta, who
bore matrilineal ties to the Nāga dynasty (Bakker, 1997).
Figure 3-7. Square pavilion temple containing the Varāha sculpture on Rāmṭek hill
The two strikingly similar Narasiṃha shrines to the south-east of the Varāha temple
are considered to be the subsequent developments on the hilltop and have been
identified as memorial shrines (Bakker, 1992a:12). Both temples are red sandstone
flat-roofed structures, containing large anthropomorphic images of Narasiṃha with a
86
Chapter 3. The Survey Region
lion’s head (Bakker, 1990:66). The simpler of the two temples, the Rudra-Narasiṃha
temple, has plain basalt pillars and no windows or decoration save for lotus flower
motifs on the interior red sandstone pillars. The Rudra-Narasiṃha temple appears to
have been constructed earlier due to these plain stylistic traits and is suggested to
have been built in commemoration of king Rudrasena at the same time as or just after
the Varāha temple around the first quarter of the 5th century AD (Bakker, 1992a:10).
An interesting feature of the temple is that it is surrounded on three sides by
projections and structures, which according to Bakker resemble fire-pits (kuṇḍa)
(Bakker, 1997). Jamkhedkar notes that it is uncertain whether the ten projections on
the exterior of the temple, with what he describes as the remains of eight cisterns
below the platform, are original. It is apparent that much of the temple wall and the
platform is reconstructed from debris of the original temple (Meister, et al.,
1988:67).
87
Chapter 3. The Survey Region
Figure 3-8. Decorative details on the Narasiṃha temples: a) Decorative frieze on the exterior of the
Kevala-Narasiṃha temple. b) Lotus pattern on the columns of the Rudra-Narasiṃha temple.
The Narasiṃha temples demonstrate that the Bhāgavata religion prospered in the 5th
century AD (Bakker, 1990:67-69). The two large basalt Narasiṃha sculptures are
almost identical, although the Rudra-Narasiṃha is slightly smaller. They depict the
deity in an unusual Mahārājalīlāsana (a dignified royal and kingly pose), with a
lotus cap and a cakra in hand as a symbol of sovereignty (Bakker, 1997:141). The
representation of the deity alone (Kevala) rather than killing the demon
Hiraṇyakaśipu conveys a calm authority and has slightly later parallels at Deogarh
and Garhdhanora (Bakker, 1997). A semi-anthropomorphic Narasiṃha is also carved
in relief at Udayagiri (Bakker, 2002). A 5th century Naga sculpture is still associated
with the Kevala-Narasiṃha temple and is conjectured to have served a “protective”
function (Bakker, 2007b:38).
A stone slab inscription is situated on the interior south wall of the Kevala-
Narasiṃha temple, which was identified during restoration work in 1981, and
interpreted by several scholars (Bakker, 2013; Bakker & Isaacson, 1993;
Jamkhedkar, 1987; Meister, et al., 1988:66). Bakker believes that this inscription
would originally have been located in the Trivikrama temple and may have been re-
situated in the Kevala-Narasiṃha temple when it was reconstructed in the Bhonsle
88
Chapter 3. The Survey Region
period (18th century). It is possible that the adjacent ruins of the Trivikrama Temple
could have provided material for the reconstruction (Bakker, 2013:169).
The content of the 15 line epigraph appears to support this supposition. The
inscription records the return of Prabhāvatī Gupta’s daughter Atibhāvatī to Rāmṭek
and her subsequent pious acts, such as the construction of the Sudarśana reservoir.
This inscription makes reference to the village Kadalīvāṭaka, which was connected to
the building activity, and where an image of a God was established (Bakker,
1997:29). Significantly, the inscription records the erection of a temple in
commemoration of Prabhāvatī Gupta and Bakker’s interpretation indicated that this
temple was dedicated to Viṣṇu Trivikrama (Bakker, 2013:175).
Figure 3-9. The ‘Kevala-Narasiṃha’ inscription, thought to originally have been located in the
Trivikrama temple.
The Viṣṇu Trivikrama temple, to the north-east of the first three structures, is placed
slightly later in construction than the Kevala-Narasiṃha temple, around the second
half of the 5th century AD (Bakker, 1992a:11). Today only parts of the maṇḍapa are
still standing on the foundation block and the structure has a Bhonsle period flat roof.
The temple remains feature dwarf pillars holding the beams of the maṇḍapa and
simple moulding decorates the temple walls with recessed niches and pilasters, floral
89
Chapter 3. The Survey Region
patterns and gaṇas (Meister, et al., 1988:66). Jamkhedkar used the similarities
between the exterior decorations on the Trivikrama temple and the Kevala-
Narasiṃha temple, including its a frieze of gaṇas, to support a similar date of
construction for both (Jamkhedkar, 1985-1986:26). Bakker believes that the specific
style of the Trivikrama temple’s ornamentation and its location outside the main
complex are evidence of its later date. Furthermore, the pillars are comparable to
examples from Nachna and Deogarh in the late 5th and early 6th centuries AD
(Bakker, 2013:169; Williams, 1983). As the Trivikrama temple is thought to be the
original home of the Kevala-Narasiṃha inscription, this text indicates that the
temple’s construction would most likely have been between 453 and 457 AD
(Bakker, 2013:175).
The badly damaged image of Viṣṇu is no longer housed by the remaining temple,
although it appears to be in its original location where the missing garbhagṛha would
have been and is orientated to face west (Bakker, 2013:169; Meister, et al., 1988:66).
This red sandstone sculpture depicts Viṣṇu striding through the universe and would
originally have had eight arms, but all are now broken off. The base of the sculpture
is significant as there is an important relief, which although worn, appears to depict
King Bali under the right leg of Trivikrama with a smaller figure supposed to
represent Bali’s wife. Bakker has compared the iconography with a lintel from
Pawaya which depicts “Bali’s sacrifice”. It is this myth which has been identified in
the dedicatory verse of the Kevala-Narasiṃha inscription and lends support to the
proposition of the Trivikrama temple being the original location of this text (Bakker,
2013:173)
90
Chapter 3. The Survey Region
Figure 3-10. a) The Trivikrama sculpture, which is located on Rāmṭek hill but is no longer housed in
the remaining temple structure. b) Part of the original Trivikrama temple which has been partly
reconstructed.
91
Chapter 3. The Survey Region
A second area identified by Bakker is the location of the Guptarāma temple and a
reclusory cave, c. 250m south of the promontory of the hill and lower on the hill side
(Bakker, 1997:86). These two Vākāṭaka structures are said to attest to the continuing
tradition of cave architecture in the area (Jamkhedkar, 1985-1986:26). Guptarāma is
particularly significant in that it is a ‘cave-cum-structural’ temple (Jamkhedkar,
1985-1986:26). The Guptarāma temple is partly carved into a natural cave in the
rock-side and integrated with a structural pillared portico (Bakker, 1990:68). The
temple currently contains a liṅga and a Nandi but these are not the original idols
(Meister, et al., 1988:70). A badly damaged four-armed Viṣṇu image was discovered
nearby and is thought possibly to represent the original sculpture which would have
been housed in this temple (Bakker, 1997:87). Approximately 200m away from
Guptarāma is the reclusory cave of Siddhanātha. This cave features two rock cut
rooms and the ceiling of the cave is decorated with pseudo pillar capitals of a simple
‘cross bracket’ type, resembling those of the Vākāṭaka temples (Bakker, 1990:68).
(Bakker, 1997:42; Meister, et al., 1988:70).
Bakker has connected the presence of rock-cut caves at Rāmṭek to the suggestion that
development of the ritual site was influenced by Prabhāvatī Gupta’s knowledge of
the religious monuments in the Gupta Empire, specifically the Udayagiri. Bakker has
92
Chapter 3. The Survey Region
suggested that these sites represent two attempts to excavate temples on the hill side,
before the decision was made to build structural temples (Bakker, 2002:3). The
tradition of rock-cut architecture may also have been continued at Mansar, which
according to Deo featured rock cut caves which have since been in-filled (1975-76,
275). The large brick Śiva temple (MNS III) features brick lined and plastered caves
with approaches on the northern and southern sides of the structure, which were used
as shrines (Joshi & Sharma, 2000:129).
Bakker proposes that the main sanctuary of the full Vākāṭaka complex would have
been located on the most prominent western point of the hill, which is now occupied
by Yādava temples. Given this location’s prominence and the peripheral formation of
the remaining Vākāṭaka temples, it would have been the logical location for the main
sanctuary, identified as the ‘Rāmagiristhāna’ in the ‘Rāmṭek inscription’ (Bakker,
1992a:11; Bakker, 1997:86). There is no evidence of the main sanctuary due to the
later overlays, but Bakker has theorised that it may have been dedicated to
Rāmacandra (Viṣṇu) with Viṣṇu’s footprints as a likely icon (Bakker, 1990:69). The
name Rāmṭek (‘Rāmagiri’ in the Meghadūta) implies that the chief 5th century
temple would have been associated with Rāma. Mirashi notes that inscriptions of
Prabhāvatī Gupta indicate her worship of the footprints of Rāmachandra and several
93
Chapter 3. The Survey Region
grants were issued from Rāmagiri-svāmin indicating a temple on the hill (Mirashi,
1963:xxiii). The 13th century Yādava inscription in the Lakṣmaṇa temple on Rāmṭek
hill also refers to the feet of Rāma (Bakker, 1991:29; 1997:64-65). This proposition
originates from deduction, however a small sculpted sandstone piece portraying a
pair of footprints in bas-relief, was found around Nagardhan (Bakker, 1991:30). This
has been dated to the 5th century through its association with other sculptures of
Vākāṭaka provenance and has been postulated to be an imitation of the original icon
on the hill (Bakker, 1991:30). The production of replicas of main deities for personal
worship is known in the area, as shown by the discovery of several small Narasiṃhas
mimicking the pose of those in the two Rāmṭek Narasiṃha temples (Bakker,
1991:31; Joshi & Sharma, 2005:14). There is no known evidence for the first
millennium worship of Rāmacandra, but Bakker has suggested that Rāma could have
been accepted as an incarnation of Viṣṇu during the Gupta age, without necessarily
being the object of an independent cult (Bakker, 1991:30).
The citadel now on the promontory of Rāmṭek hill results from later periods of
building from the Yādava period onwards. The Yādavas developed the promontory
of Rāmṭek hill with temples and water tanks, and the walled temple complex which
incorporates Bhogarāma and Varāha temples continued to be elaborated during the
Vijayanagara period (15th to 16th centuries) and from the Mārāṭha period onwards
(Post-18th century) (Bakker, 1989b:467). The later constructions make it difficult to
assess the level of immediately post-Vākāṭaka Rāṣṭrakūṭa or Cālukya period activity
between the 5th and 12th centuries (Bakker, 1990:70).
The known constructions of the Yādava period on Rāmṭek hill include the
Sindūravāpī and the Siteei Nhānī (Sītāsnānasthala) tanks and three main temples
dedicated to Hanumat, Lakṣmaṇa and Rāma (Bakker, 1990:71). The Yādava temples
do not resemble the earlier Vākāṭaka shrines, as they are constructed from different
stone and lack sculptural decoration (Bakker, 1989a:94). They are influenced by a
Northern style of temple architecture, which largely correspond to the Phamsana and
Bhumija type of the Nagara tradition (Bakker, 1989a:97; Sohoni, 1998:48). The
stone Sindūravāpī tank appears to be the earliest post-Vākāṭaka structure on the hill
and has cloisters which appear pre-Yādava as the pillars are reminiscent of the
Vākāṭaka tradition (Bakker, 1989a:93; 1989b:475). Sītecī Nhāṇī tank on Rāmṭek hill
is conjectured to be an ancient structure which has been reconstructed over time and
94
Chapter 3. The Survey Region
has Vijayanagara period design elements, and a gate and chhatra which appear to be
Mārāṭha period additions (Bakker, 1989b:476).
Figure 3-13. The Sindūravāpī tank on Rāmṭek hill, which displays columns similar to those in
Vākāṭaka constructions.
A brief test excavation between the two Narasiṃha temples was conducted by
Jamkhedkar (1981) to investigate the past activity on Rāmṭek hill, but little was
recovered and the excavation was not continued. No structures from this intervening
period are visible on the hill and in the immediate area only the Kālikā Devi temple
in Rāmṭek town has been tentatively identified as a 7th/8th century Cālukya
construction (Jamkhedkar, 1985-1986:26). Architecturally this temple displays a
relatively plain vaulted roof, suggestive of a southern style of architecture (Meister &
Dhaky, 1986). The Devi temple’s roof is supported by pillars which appear to have
some continuity with the earlier Vākāṭaka constructions on the hill. Bakker believes
that the “barrel-vaulted roof” featuring caitya windows is reminiscent of 8th century
architectural traditions (Bakker, 1989a:93). However, Willis believes the simple
moulding and decoration could be indicative of a slightly later date for construction
(pers. Comm). Fragments of early sculpture can be found on a platform outside the
temple and an image of Ganesh has stylistic qualities which could make it
comparable to gaṇa images found at Mansar (Bakker, 1989a:94).
95
Chapter 3. The Survey Region
Figure 3-14. Vaulted roof of the Kālikā Devi temple in Rāmṭek town.
Situated on the lower north side of Rāmṭek hill is a temple characteristic of the 13th
century Hemādpanti style, which demonstrates religious activity towards the end of
the Yādava period; Karpūravāpī (site 92) consists of a temple and a stone tank with
arcades on four sides (Bakker, 1989b:475; 1990:71). On the western side of the
temple are three śikharas topped with large āmalakas. The stones of the śikharas are
slotted together with no use of mortar. The largest central tower has curved sides
while the two smaller towers appear to have a ‘pyramidal’ shape (Bakker, 1989a:96).
Figure 3-15. Karpūravāpī at the base of Rāmṭek hill on the north side.
96
Chapter 3. The Survey Region
Ambālā Lake, at the east end of Rāmṭek hill, has a small ruined temple
(Ambikanatha) on the south-eastern side, which appears to be of the early Yādava
period and shows some affinity with the Kālikā Devi temple (Bakker, 1989a:95).
This structure may be that referred to in the Rāmṭek stone inscription of the Yādava
king Ramachandra (last quarter of 13th century AD) (Bakker, 1989b:474). Bakker
reports that the temple housed an image of Balarama, carved in red sandstone and
styled so as to suggest an early date. He theorises this could represent one of the
original images of the Bhogarāma temple on the hill (Bakker, 1989a:96). When the
Mārāṭha Bhonsle chiefs established their capital at Nagpur in the 18 th century,
Rāmṭek underwent development as a key ritual centre in their territory, resulting in
the renovation of old temples and building of numerous new structures on and nearby
Rāmṭek hill (Bakker, 1989a:98). The majority of the temples around Ambālā Lake
appear to originate from the Mārāṭha period as demonstrated by the Islamic influence
in their design (Sohoni, 1998:149). Bakker notes that although clearly much later
than the hill temples, those around Ambālā combine Indo-Islamic elements with a
“remarkable continuity” in aspects of style, such as the use of plain square pillars
topped with a cross bracket (Bakker, 1989a:99).
Figure 3-16. A group of Medieval temples situated on the south west side of Ambālā Lake.
97
Chapter 3. The Survey Region
Roughly 6km south of Rāmṭek is the village of Nagardhan, which has been identified
as Nandivardhana, the capital or ‘residence’ of King Pravarasena II. Nandivardhana
is the second known capital of the Eastern Vākāṭakas following Padmapura. The site
is first mentioned in the Pune copper-plates of Prabhāvatī Gupta (Year 13 of her
regency) and the charters of Pravarasena II were issued from Nandivardhana until the
Belora plates of the 11th year of his reign. At some point before the 16th year of
Pravarasena II’s reign, the capital is thought to have shifted to the new state
sanctuary at Pravarapura, from where most of the subsequent charters were issued.
98
Chapter 3. The Survey Region
Several scholars, such as Hiralal and Mirashi, connected the ancient place name of
Nandivardhana with the modern village of Nagardhan or ‘Nandardhan’ in Rāmṭek
Tehsil (Mirashi, 1963:7). The area has rich archaeological potential and findings of
5th century AD remains have been taken to support the site’s etymology and
epigraphic evidence (Bakker, 1991:29). The standing remains which dominate
Nagardhan today are post-Vākāṭaka and include a large Bhonsle period fort (18th
century) on the southern outskirts of the village, which has seen extensive renovation
by the State Archaeology Department. It has been suggested that the Medieval
structure was raised at the site of an ancient fort, but there is a lack of extant
structural or architectural remains for any Vākāṭaka ‘residence’ (Bakker, 1997:84).
Early assessments of Nagardhan were carried out by Wellsted during his 1933 survey
of Mansar and its surroundings. Wellsted did not agree with tentative identifications
of the site as Nandivardhana stating that while brick remains were present, none
appeared to be sufficiently large to be identified as ‘Gupta’ type bricks. In a note,
Wellsted remarks that the Nagardhan bricks are almost exactly half the size of
average ‘Gupta bricks’, leading to their misidentification as fragmentary Gupta
period brick (Wellsted, 1934:161, Note 2). Wellsted considered there to be stronger
grounds for identifying Nandivardhana at the more northerly site of Nandpur (in the
Pench wildlife reserve), based on its perceived strategic position and remains of brick
fortification (Wellsted, 1934:160). During the 2006 short survey by Bakker and
Cork, no archaeological evidence was located around the location of Nandpur. Cork
concluded that while Nagardhan and Nandpur may have been contemporary,
Nagardhan was most likely the main settlement site attached to Rāmṭek and Mansar,
and thus the ancient site of Nandivardhana (Cork, 2006). The site of Nandpur has not
been subjected to sufficient archaeological investigation to fully evaluate Wellsted’s
claims that it was a large Vākāṭaka site and unfortunately, the site was inaccessible
during the current research.
Rāmṭek is thought to have been approximately at the centre of the Eastern Vākāṭaka
‘sphere of influence’ and a site of significant religious importance, so it would be
expected that the Vākāṭaka capitals would be located in its immediate vicinity
(Wellsted, 1934:166). It is proposed that Nagardhan may have been the royal
‘residence’ attached to the state sanctuary at Rāmṭek (Rāmagiristhāna) (Bakker,
1992b:83). Rāmṭek is a prominent landscape feature clearly visible from Nagardhan
99
Chapter 3. The Survey Region
and furthermore would have been situated within the agricultural zone south of the
hill. Sutcliffe suggested that the Khindsi reservoir could have supported a maximum
cultivatable area of 100km² (Sutcliffe, 2007; Sutcliffe, et al., 2011:785). Mirashi
supported the identification of Nagardhan with Nandivardhana and noted that it too
occupied a beneficial location in the wider landscape as the royal capital would have
been surrounded by “strongly fortified” sites such as Ghughusgadh and Bhivgadh
(22km north west of Nagardhan in the Pench reserve) (Mirashi, 1963:xxiii).
The ancient site of Nandivardhana is generally considered to have included the area
now occupied by the adjacent village of Hamlapuri; if so it would have stretched
across a distance of 3km (Shastri, 1997:219). Finds that support Nagardhan’s
identification as a Vākāṭaka settlement have been reported over years of preliminary
survey. The Department of Archaeology and Museums noted that the modern village
of Nagardhan is clearly situated on top of a mounded habitation site and reported
pottery, architectural fragments and sculptures alongside multiple brick structures
and sections of a brick fortification wall measuring 10.5m wide in places (Sali,
1998:13). In connection to this, IAR reports that Chandrashekhar Gupta identified
several sculptures from Nagardhan in the early 1980s, which he assigned to the
Vākāṭaka period (Mitra, 1983c:49-50). The sculptures included images of Ganapati
and Lajjā Gaurī. Agarwal and Nath explored Nagardhan under the Archaeological
Survey of India and recorded various Early Historic finds including terracotta
figurines, a stone pestle, a spindle-whorl and ceramic sherds of incurved bowls,
carinated handis and miniature pots with “mica dusting” (Mitra, 1983c:96).
The 2006 exploration by Bakker included a visit to the Nagardhan area and a number
of potential mounds were reported as being visible in the fields to the northwest of
the main village. However, the one chosen for investigation yielded no
archaeological remains and focus was drawn to nearby Hamlapuri (Cork, 2006).
Finally, the site was recently re-explored during PhD research conducted by
Bhaisare, who has reiterated that the extant fort was built on the remains of the
Vākāṭaka ‘durg’ (Bhaisare, 2012). To complement these reports, further
archaeological evidence is needed to more intensively explore, and excavate the
Nagardhan area. Further investigations may corroborate the conjectured
identification of Nagardhan as Nandivardhana and increase our understanding of the
nature of the ancient site; it is unknown whether a ‘capital’, assumed on the basis of
100
Chapter 3. The Survey Region
epigraphic places of issue, encompassed major settlement areas or whether the focus
of settlement in the Rāmṭek landscape was elsewhere.
Mansar (21°24’N and 79°17’E) lies six km west of Rāmṭek and was identified as
Pravarapura following extensive archaeological research over the last decade.
Mansar was known to have archaeological remains in the early 20t-h century and
there were reports of Gupta-type bricks and ruins assumed to be part of a Buddhist
monastery (Wellsted, 1934:161). Notes in the appendix of the Nagpur Gazetteer of
1908 state that the village was known due to its tank, called Manisar (‘jewel tank’),
which attracted pilgrims to the site (Russell, 1908). These early reports mention the
presence of brick bats on the low hills by Mansar village and state that exposed
building material from the foundations of brick buildings was being re-used by
villagers (Joshi & Sharma, 2005:2). The first full exploration of the area was carried
out by Wellsted in 1933 who recorded extensive stone facing on Mansar tank and
architectural fragments displaying 5th century AD characters (Wellsted, 1934:162).
The site was dated through the identification of large ‘Gupta-type’ brick remains,
supported by sculptural and epigraphic evidence, and Wellsted recorded traces of an
“extensive town-site” to the east and south of Mansar tank (Wellsted, 1934:161).
101
Chapter 3. The Survey Region
finds including a small soapstone snake image and a large pot containing ashes,
leading to the conclusion that this site was a burial shaft (Wellsted, 1934:164).
Wellsted’s collection is now housed at the British Museum (Willis, 2008). Bakker
has proposed that the burial shaft was a funereal monument to Prabhāvatī Gupta
given her Nāga descent.
Figure 3-17. a) The small soapstone snake in the Wellsted Collection of the British Museum, b) A
selection of the ceramic vessels in the Wellsted Collection
In the 1930s, Wellsted stated that the decay and notable level of destruction from
brick quarrying at Mansar was already so advanced that excavation would be
unfruitful, with the possible exception of the ‘monastery’ site and Hill B (Wellsted,
1934:164). This suggests that some of the remains may not be locatable through
modern survey - a predicament encountered in Bakker’s 2006 exploratory field
season which attempted to corroborate Wellsted’s findings at Mansar and evaluate
the impact of modern development over the intervening 70 years. The team explored
the area around the hill and tank, largely confirming the major observations by
Wellsted. Brick remains were found to be scattered over the hills surrounding
Mansar were scattered with brick remains, which suggests there were dispersed
buildings around the Mansar hills. However, no substantial remains were discovered
which would be indicative of a large settlement site and there is no longer evidence
for brick walls at the area identified as settlements to the east of Mansar hill (Cork,
2006:2). The estimated GPS location of ‘Site T’ and the altar site were explored, but
102
Chapter 3. The Survey Region
these features appear to have been totally destroyed by mining activity (Cork,
2006:2). Sutcliffe confirmed the major details of Wellsted’s original map of Mansar
tank as the rough masonry facing is still present, although there is less than reported
in 1934, and the embankment was confirmed to be man-made (Sutcliffe, 2007).
Three projections from the embankment were recorded rather than two as in the
original map and modern sluices were identified on the east and western end of the
tank (Sutcliffe, 2007:5). Based on the size of the tank, and estimated inflows of water
and rainfall, it was concluded that the lake was likely to have been ornamental,
although it may have been useful as a “domestic” water source (Sutcliffe, 2007).
Figure 3-18. Wellsted’s map of Mansar tank and hill, showing the presence of buried monumental
constructions (1934).
103
Chapter 3. The Survey Region
Figure 3-19. Mansar hill as it stands today with several of the larger sites exposed by excavation and
reconstructed.
Another important find from Mansar was a set of Vākāṭaka copper plates (the
Rāmṭek Plate of Pravarasena II), of which one has been published by Mirashi (1963).
A Vākāṭaka copper-plate inscription usually consists of five plates and it was
reported that all five were originally recovered from the manganese mine, but only
one was recovered from the contractors. In a typical copper-plate, the first two plates
104
Chapter 3. The Survey Region
contain the genealogical information and the donor’s name; the third plate has the
details of the land-grant including boundaries and the recipient of the land; The
fourth plate contains the conditions, privileges and immunities of the grant; finally
the fifth plate states the regnal year of the charter and the benedictory verses. The
Mansar plate is the fourth and shows that the donation was made for the “religious
merit, life, strength and prosperity” of the donor (Mirashi, 1963:73). While not
definitively attributable to a king, its palaeography is thought to be consistent with
those of Pravarasena II (Mirashi, 1963).
Figure 3-20. The Mansar Śiva at Delhi National Museum, labelled as Śiva Vamana (Śiva in the form
of a dwarf) and measured at 84cm in height, 65cm wide, and 37cm deep.
An initial sounding was excavated at Mansar by Nagpur University, but few details
are available regarding this test excavation, and the mounds have since been
extensively excavated and the structures partly renovated (Sali, 1998:8). Nath led the
first excavation in 1994 under the Archaeological Survey, which focused on
Wellsted’s ‘Monastery site’ in order to ascertain the extent, layout and nature of the
remains and to provide further context for the sculptural finds (Manjhi, et al.,
2000:55). Excavations on the western side of the mound revealed part of a large
burnt-brick structure with mud masonry, and three phases of structural activity were
tentatively identified. In the first phase, Nath identified a core area with a moulded
adhiṣṭhāna built over a “knoll”. In the second phase the adhiṣṭhāna appears to have
been enlarged without making any significant alterations in ground plan however the
105
Chapter 3. The Survey Region
exterior was decorated with pilaster mouldings. Finally in the third phase walls were
raised around the adhiṣṭhāna and filled with brick bats, topped with a porch and
“subsidiary shrines”. On the west side of the structure the main stairs were sealed and
filled in and the original core area of the structure was supported by the construction
of “ramp-walls”. The excavators noted that a thick burnt deposit suggested large-
scale burning. Sculptural remains at the site, including Uma-Maheśvara, Lajjā Gaurī
and a plaque with footprints, are reported as indicating the Brahmanical religious
affinity of the site (Manjhi, et al., 2000:56). Other finds included ceramics, several
coins, clay figurines, beads, ornaments and iron objects, and the site was dated to
between 300 and 650 AD on the basis of the numismatic evidence, primarily
attributed to the Kshatrapas and Vākāṭakas (Manjhi, et al., 2000:57). The ceramic
types were said to be characterised by medium fabric red ware, coarse gritty
micaceous red ware and fine polished red ware in a range of shapes including vases
with flared rims and carinated handis (Manjhi, et al., 2000:56).
Extensive excavations were carried out from 1997 by the ASI, under the supervision
of the Buddhist organisation who owns the land (Joshi & Sharma, 2000; 2005; Mani,
et al., 2003). Nath and Wellsted’s ‘monastery’ site was renamed MNS II and re-
identified as a secular building; the ‘palace’. Wellsted’s ‘Hill B’ was excavated
(MNS III) and further brick constructions across the hill were subject to smaller scale
investigations.
Joshi and Sharma’s findings differed to those of Nath and resulted in the proposition
of a long chronology for Mansar. Prehistoric evidence has been obtained during
106
Chapter 3. The Survey Region
explorations in the area, including stone tools and a large rock formation determined
to be an anthropomorphic figure, taken to be representative of religious development
at the site by early man (Joshi & Sharma, 2005:4). Megalithic cairns were also
reported along the bank of Mansar tank (Joshi & Sharma, 2005:127). The chronology
of structural sites on Mansar hill has been dated from the Mauryan-Sunga period
onwards with great development under the Sātavāhanas, followed by occupation by
the Vākāṭakas.
Table 3-2. The periodisation outlined for the structural remains at Mansar following excavation (Joshi
& Sharma, 2005)
Mansar has exceptional archaeological remains and the two main structures (MNS II
and III) are important examples of a limited number of brick terraced structures
dating to the Gupta-Vākāṭaka period. The excavators have also reported that the two
star-shaped temples at the site were constructed at least as far back as the 2nd century
AD under the Sātavāhanas, which would make them possibly the earliest known
temples with this plan in India (Sharma, 2008). Early stellate temples are known
from Tala (c. 500 AD) and the Mansar temples are particularly similar in plan to the
Rāma temple in Sirpur and the Dhobinī temple near Damakheda (Bakker, 2008b).
Unfortunately, the original details of the architectural plan have been obscured by
modern reconstruction.
107
Chapter 3. The Survey Region
Figure 3-21. Stellate-plan liṅga temple on Mansar hill (site 10: MNS V).
The excavators assigned the construction of the original ‘palace’ site (MNS II) to
Period II under the Sātavāhanas, with an enlargement under the Vākāṭakas in Period
III. It is hypothesised that when the Vākāṭaka king Pravarasena II moved his capital
from Nandivardhana to Pravarapura, an abandoned Sātavāhana palace was re-
occupied and subject to additions and enlargements to the original plan (Joshi &
Sharma, 2005:22). The excavators at Mansar believe that the only change under the
Vākāṭakas at MNS II was the enlargement of the adhiṣṭhāna and a change to the
overall plan by the addition of four “bastion like square projections” to the corners of
the platform (Joshi & Sharma, 2005:20). MNS II is described as standing on a solid
brick platform of 51m (east-west) by 44m (north-south), with the ‘palace’ itself
comprising a large complex of rooms approached from the west by a flight of wide
steps (Joshi & Sharma, 2005:8). The palace was surrounded by wide brick
fortification walls on all sides, parts of which still exist to a height of 5m. The area
within the fortification walls is estimated to be 124m east to west and 110m north to
south, containing areas of courtyards, veranda and rooms with a forty-two pillared
maṇḍapa dated to the Sātavāhana period (Joshi & Sharma, 2005:9).
MNS III has a stone foundation and a brick superstructure with carved red sandstone
architectural pieces, the construction of which has been assigned by the excavators to
around 340 AD (Joshi & Sharma, 2005:14). Maurya-Sunga origins were identified
by the excavators at MNS III, and the published articles report the remains of
108
Chapter 3. The Survey Region
Buddhist structures including two brick built stupas and a chaitya, which are no
longer visible (Joshi & Sharma, 2005:128). The excavators suggest that a pre-
existing structure at MNS III was later developed into a large Vākāṭaka temple in
early Period IIIA (the second quarter of the 4th century AD), with a second phase of
construction in Period IIIB (Joshi & Sharma, 2005:14). A further significant find at
MNS III was the so-called ‘Man of Mansar’, described as a more than life-size lime
figure of a man with a crushed head, on a sacrificial altar made of bricks (Joshi &
Sharma, 2000). Associated with the figure were an iron snake and several pots. This
has been described as a substitute for a human sacrifice (Puruṣamedha) made for the
construction of the building in order to spiritually protect and safeguard it (Bakker,
2007a:8).
109
Chapter 3. The Survey Region
Figure 3-22. South side of MNS III, the Pravareśvara temple, with its elaborate brick shrines
At MNS II there does appear to have been at least three phases of construction
(initial structure, a period of enlargement, and further modification and enlargement)
but Bakker suggests these took place over a limited period of time under the
Vākāṭakas and that the modification was due to instability of the monument. The
later phases of building appear to be primarily ballasts and extensions to support
collapsing portions of the building. Raven has compared the alterations at MNS II to
those observed in building phases at Pawaya and Ahichhatra. At both sites the
platform was modified and refurbished to reflect later fashions in pañcāyatana plan
(Raven, 2008:8). Other similarities between the sites are observed in the construction
110
Chapter 3. The Survey Region
of the adhiṣṭhānas from “receding courses of bricks”, the use of foundation boxes
filled with brick-bats and barricades or retaining walls. Bakker has stated that while
the original basement construction appears to have been enlarged, he finds there to
be no significant difference in decoration and architectural style between phases
(Bakker, 2002:8; 2008b). The identical appearance of the pilaster mouldings, lime-
plastering and painted decoration of the two proposed phases of construction at MNS
II have been explained by excavators as Vākāṭaka attempts to imitate the original
Sātavāhana structure. Bakker has also noted that the temple friezes on the temples at
Rāmṭek display similar ornamental pilasters and recesses (Bakker, 2008b). Similar
decorations are known from the remains of a large brick temple uncovered at Nagara,
located towards the edge of the Eastern Vākāṭaka territory, which is (inconclusively)
dated by Jamkhedkar to the end of the 5th century AD (Bakker, 2008b). Raven and
Bakker have supported the proposition that MNS II was originally a Vākāṭaka
construction, referring to Nath’s 1994-95 excavation report in which the structures
and ceramics identified were dated to c.300-650 AD rather than an earlier origin
(Bakker, 2002:7; Manjhi, et al., 2000; Raven, 2008:9). Raven also notes that no
material, numismatic or inscriptional evidence is mentioned from Nath’s excavation
of MNS II which would verify a pre-Vākāṭaka phase (Raven, 2008:7).
Furthermore, Bakker disputes the early dating for MNS III and assigns its
construction to the 5th century AD. Maurya-Sunga origins were proposed based on
the reported remains of Buddhist structures however Bakker has been unable to
identify these remains and rather considers Mansar to be a complex of Hindu
temples, similar to the 5th century constructions at Ahichhatra and Pawaya (Bakker,
2002:11; Raven, 2008). Also noted is the similarity between the carved brick designs
and architectural motifs at MNS III and MNS II, which indicates a similar date of
construction. Unfortunately there are no plans or photographs related to the Buddhist
structural remains and the Vākāṭaka monument has been reconstructed, so these
cannot be assessed in the current research. The excavators refer to artefactual
evidence of Buddhist activity at the site in the form of a sealing with a Buddhist text
common in Gupta and post-Gupta times. This sealing is dated to the 6th or 7th
centuries AD on palaeographic grounds but its provenance has been questioned by
secondary scholars (Kropman, 2008:5). The proposition of earlier structures by the
excavators is significant for the aims of this thesis which hopes to examine the
111
Chapter 3. The Survey Region
Following the Vākāṭaka occupation at the site, the excavators identified Viṣṇukuṇḍin
(Mahavihara) period remains. A Buddhist stupa was discovered over the debris of the
Pravareśvara temple alongside a Viṣṇukuṇḍin copper coin belonging to
Mahendravarman II (470-580 AD) (Joshi & Sharma, 2005:23; Mani, et al.,
2003:133). It has been proposed that following the destruction of the Vākāṭaka
palace, the entire complex was converted into a Mahavihara under the influence of
the Viṣṇukuṇḍins, and a stupa was erected at the ‘palace site’ (Joshi & Sharma,
2005:23-24). Furthermore, a building described as a vihara was built adjacent to the
Pravareśvara temple although this identification is also challenged by Bakker who
has been unable to determine Buddhist features in the building’s layout. Rather he
conjectures that the structure may have been a secular building connected to the royal
residence (Bakker, 2008c). The excavators were unable to establish how long a
Mahavihara may have been operational, but assert that Mansar became a “great
centre of Buddhist learning” in the 6th century AD (Joshi & Sharma, 2005:25).
112
Chapter 3. The Survey Region
Despite the reported architectural evidence, only a few artefacts of Buddhist origin
were discovered and there were no finds of Buddhist sculpture leading to suggestions
that the antiquities at the site have been plundered prior to the excavations (Joshi &
Sharma, 2005:25).
Figure 3-24. MNS IV (site 9) described as a Buddhist vihara by the excavators of Mansar.
The full site report and the finds and ceramic catalogue from Mansar are awaited.
Limited photographs are available from visitors to the site and the report of Bakker’s
exploratory 2006 survey notes the state and extent of the structures as they were
revealed (Cork, 2006). Following the fieldwork by Bakker, a colloquium was held at
the British Museum in 2008 to discuss the site and consider it against the wider
context of the Vākāṭaka dynasty. Scholars had access to some of the sculptures,
architectural pieces and inscriptions and produced an e-book which primarily
discussed the iconography, art-history and architecture, with new research into the
excavated artefacts and inscriptions (Kropman, 2008:1).
Following the extensive surveys of the ASI in the 19th century and early 20th century
brief archaeological explorations have been conducted across the rural landscape of
Rāmṭek and Nagpur District. Wellsted’s survey in 1933 recorded visible
archaeological remains in the vicinity of Rāmṭek and Mansar and was focused on one
113
Chapter 3. The Survey Region
site of particular interest. This site, thought to be a ruined brick and rough-stone
‘fort’, is situated on the hilltop immediately west of the Khindsi dam site and 300ft
above the plain (Wellsted, 1934:159). Wellsted also identified the remains of a
settlement on the north side of the hill, of unknown extent, which at the time was
stated to be normally submerged by the reservoir (Wellsted, 1934:159). Preliminary
village-to-village survey has been carried out across Rāmṭek Tehsil and the
neighbouring districts, and the Directorate of Archaeology and Museums in
particular have identified several sites noted as having potential Vākāṭaka
archaeological remains within the research area (Sali, 1998:5).
Hamlapuri, roughly 2km east of Nagardhan, is perhaps the best-known rural site in
the Rāmṭek landscape and has been subjected to several preliminary investigations.
Due to reports of archaeological remains extending between the two sites, Hamlapuri
is frequently identified as an extension of the ancient capital at Nagardhan (Bakker,
1997:84). Pieces of red sandstone sculpture were discovered during brief
explorations under the Nagpur Central Museum, and by Gupta during explorations of
Hamlapuri and neighbouring Nandpuri in 1981 (Mitra, 1983c:49-50; Rao, 1985:137).
The State Department has reported pottery sherds, architectural fragments and
Vākāṭaka sculptures, alongside a ‘Vākāṭaka fort’, but no accompanying details were
provided (Sali, 1998:12). A few test trenches were excavated in 1982 to establish the
potential of the site, and uncovered the remains of structures and associated artefacts
(Sali, 1998:9). Again, limited published details are available but clearly potential for
discovery of religious and habitational remains in the hinterland of Rāmṭek.
Hamlapuri has been identified as an “active religious centre” due to the number of
Vākāṭaka sculptures identified at the site, which include images of Ganesh,
Mahiṣamardinī, Viṣṇu and Lajjā Gaurī, but there are no structural remains clearly
identifiable as a temple or monastery (Jamkhedkar, 1985:18). An isolated find of
three bronze Buddha images was discovered near to Hamlapuri, two of which appear
to be local while one is considered to reflect the Gupta style from Mathura
(Jamkhedkar, 1991a:87-88; Rao, 1985:131). The bronzes have been dated to around
the 6th century AD (Rao, 1985:131) and are cited as evidence to support continuing
Buddhist activity in the area alongside the Hindu Vākāṭakas (Joshi & Sharma,
2005:25).
114
Chapter 3. The Survey Region
The 2006 survey by Bakker identified that the agricultural area surrounding the
Ganesh shrine has been dug down through the original ground surface exposing
pottery and brick fragments in the field banks (Cork, 2006:3). At the time it was
noted that Hamlapuri’s archaeological potential had been mostly destroyed, but in
comparison to Mansar there was a high number of surface ceramics visible (Cork,
2006:3). The most recent exploration at Hamlapuri was conducted by Bhaisare
(2012), who states that this site is invaluable in attempts to establish diagnostic
Vākāṭaka material remains as past excavations revealed it to be ‘single culture’.
Bhaisare’s publication of surface ceramics at Hamlapuri states that the majority of
identified are micaceous red or micaceous black ware and generally display
continuity from the Sātavāhana period in “degraded” forms and coarser fabric. The
ceramics are used to reiterate that the site is single culture, meaning that the
assemblage should be reliably datable to the Vākāṭaka period (Bhaisare, 2010-
2011:105)
Figure 3-25. Remains of the site at Hamlapuri (photograph taken during the first season of fieldwork).
The paddy field banks and the ‘island’ with the trees represent the original height of the ground.
Ceramic and brick remains are found throughout the sections and across the surface of the fields.
In addition to Hamlapuri is another site of interest within the survey area, Kelapur,
2km north of Rāmṭek on the edge of the Khindsi tank. This area is known due to its
identification in the Kevala-Narasiṃha inscription as the village of Kadlīvātaka
involved in the construction of the Sudarśana water reservoir and Khindsi dam
115
Chapter 3. The Survey Region
(Bakker, 2013:175; Bakker & Isaacson, 1993:69). There has been little exploration at
Kelapur but the site is reported as having few remains, due perhaps to its location in
the basin of the River Sur, which regularly floods (Bakker & Isaacson, 1993:69).
Figure 3-26. Temple at the site of Kelapur in an area often flooded by Khindsi tank. Surrounding the
temple are surface finds of ceramic sherds
116
Chapter 3. The Survey Region
Table 3-3. Villages reported as having been explored within Rāmṭek tehsil.
In the previous chapter, it was established that historically the Eastern Vākāṭakas are
thought to have inherited a region associated with Early Historic Sātavāhana activity,
but that they subsequently moved and established a dynastic centre in the ‘peripheral
area of Rāmṭek-Mansar. The archaeological information contained within this
chapter regarding the two known monumental Vākāṭaka sites in the survey region
demonstrates the high investment in these areas during the 5th century AD. Although
contested, the published evidence does imply pre-existing Sātavāhana and Buddhist
investment at the site of Mansar and could be taken to support the picture of the
Eastern Vākāṭakas imposing their ritual and ideology at sites. Buddhist finds have
also been recorded from Hamlapuri which further suggests organised religious
activity was present around Rāmṭek, both prior to the construction of the Hindu
temples on Rāmṭek hill and contemporary with Vākāṭaka authority. It remains
unclear as to what extent the Vākāṭakas inherited a landscape with socio-economic
117
Chapter 3. The Survey Region
and ritual development and to what extent they developed the landscape. It is hoped
that new archaeological data will improve our understanding of the Vākāṭaka impact
on this area.
The major sites in the Rāmṭek landscape are not currently known to demonstrate
significant religious or settlement activity between the end of Vākāṭaka rule and later
Yādava occupation, but it is unclear whether the apparent absence of activity is a true
phenomenon or due to a lack of evidence. If the level of activity did reduce following
the Vākāṭakas, the survey would be expected to uncover relatively short-lived Early
Historic sites and few remains from the intermediate period. The accepted model of
development, after the introduction of Yādava control in the region, suggests that we
would expect to find evidence of renewed architectural development, expansion of
towns and village areas and investment in infrastructure, such as irrigation facilities.
The extant evidence at Rāmṭek and Nagardhan demonstrates Yādava through to
Mārāṭha period activity, particularly evidenced by religious structures. It is likely that
there will be further small temples and sculptural pieces in the village areas. Many of
the standing remains are likely to be of the Mārāṭha period, following the heavy
influence of the Mārāṭhas in Nagpur District.
There is clear potential for archaeological survey to explore the landscape and
establish how this expected model relates to evidence on the ground. However, it
should be noted that disturbance of sites and re-use of archaeological material in the
area was already reported prior to 1934, and it is likely that the situation has
deteriorated. Alongside a swiftly developing local environment, this provides a
further argument for recording of the current visible archaeological features. Given
the potential for survey to increase our understanding of the period of Vākāṭaka rule
in central India, the following chapter will move on to outline the development of
landscape archaeology and its potential to contribute to the study of the Early
Historic period.
118
Chapter 4. Landscape archaeology
4.1. Introduction
As a study aimed at using surface survey to situate the Eastern Vākāṭaka site of
Rāmṭek in its landscape and evaluate the impact and nature of the governing polity,
this research looks towards projects which have successfully investigated centre-
hinterland dynamics. This chapter summarises the ways in which archaeologists in
South Asia have targeted questions of long-term regional development though
investigations of the landscape, and highlights the value of archaeological survey in
answering questions we might have about the social, political and religious
organisation and development of an area. This overview proves the potential and
successes of Indian survey, however the practical limitations of this project must also
119
Chapter 4. Landscape archaeology
Landscape archaeology has been defined as the study of the land and the relationship
between people and their environment. Landscape archaeology questions how the
land has been ordered and changed for “economic, social, religious, symbolic or
cultural processes”, and how in turn it has influenced human thought and action
(Wilkinson, 2004:334). It has developed both theoretically and methodologically to
become a significant and multifaceted sub-discipline within archaeology, which can
encompass a combination of techniques and different categories of evidence for a
holistic assessment of the land (Wilkinson, 2004:335). Landscape archaeology and
field survey is employed to study the spatial and temporal distribution of sites, with
the aim of producing data related to social, political, economic and religious models
of the landscape (Wilkinson, 2004). Archaeological questions of a regional nature,
which can be addressed through landscape archaeology, are necessary for the
contextualisation of sites, the understanding of site relationships and investigation of
the social, political, economic or religious aspects of past landscapes.
While archaeological research has always had an inherent interest in the landscape
through studies of space, the treatment and perception of ‘landscape’ has
significantly altered (Knapp & Ashmore, 1999:1). The concept of ‘landscape’ is
complex and changeable, which keeps it consistently relevant to disciplines such as
archaeology, geography, history and anthropology (Gosden & Head, 1994:115-116).
Initially, the landscape was viewed as a ‘backdrop’ for material culture (Knapp &
Ashmore, 1999:1). However, shifting theoretical mind-sets have encouraged the
consideration of different aspects of the landscape, and prompted the emergence of
several methodological approaches. The growing concern with a regional perspective
led away from the isolated site focus of early studies, and rather than being viewed as
120
Chapter 4. Landscape archaeology
Traditionally, survey was perceived as a basic tool to aid in the location of sites or to
prepare for excavation (Ammerman, 1981:63). However, as a result of increasing
interest in landscape studies, the use and development of archaeological survey has
accelerated and it now forms the foundation of regional research. In terms of
practicality and logistics, survey offers great benefits over excavation; it is non-
destructive, requires lower fiscal investment, fewer resources and man-power, and
offers flexibility to account for project limitations. Although unable to provide the
total history of an individual site without the combined use of excavation, survey has
the ability to cover wide areas of land and produce primary data directed towards a
regional perspective (Dunnell & Dancey, 1983:271; Schiffer, et al., 1978:1). In
current archaeological practice, there is a plethora of survey projects being carried
out across the world, incurring on-going methodological and theoretical development
and making significant contributions to the course of archaeological inquiry.
When much of the world’s archaeology was unknown, early landscape investigations
primarily took the form of general reconnaissance surveys. These extensive and
unsystematic surveys were aimed at maximising coverage of the landscape and
recording sites for the production of gazetteers and monuments records and resulted
in the abundant discovery of sites (Ammerman, 1981:65). From the beginning of the
20th century, the number of landscape surveys increased worldwide following
methodological developments, such as the use of distribution maps in Europe
(Crawford, 1912; 1922). As techniques evolved, surveys became progressively
systematic and more survey projects were conducted (Ammerman, 1981:65). The
innovation of regional settlement pattern research during the late 1940s and early
1950s heightened interest in the regional context of sites (Ammerman, 1981:65).
(Anschuetz, et al., 2001:168). Mesoamerican and Near Eastern surveys are often
cited as archetypal examples of settlement pattern studies focusing on creating
121
Chapter 4. Landscape archaeology
122
Chapter 4. Landscape archaeology
Bourdieu’s theory of habitus (1977) and Giddens’ Structuration theory (1979) have
been very influential within landscape archaeology in this context, as they emphasise
the active role of humans as agents in creating the space around them. By
considering how past people actively shaped and experienced their social world, both
theories recognise the mutual dependency between people and their systems, material
culture and landscapes (Fisher, 2014:355). Studies of the relationship between
ritual/political centres and their landscapes require consideration of both the general
population and that of the elite. Elite agency may be determined through the
archaeological remains of ritual and economic investment as monumental
123
Chapter 4. Landscape archaeology
constructions and control over land-use shapes the landscape. These built landscapes
can structure cultural systems and be used as symbols for ideologies, authority and
the meaning of space (Fisher, 2014:358).
Practices like monument construction are the result of having enough resources,
social mobility and authority to invest in the land (Glatz, 2014). Monumentality is
often viewed by archaeologists as a proxy for social complexity and attempts by the
elite to control the population and perpetuate political power (Osborne, 2014:5).
However, the reception of monuments by the population is as important as the
physical constructions; monumentality is created by the association of the structure
or object with particular meanings, which occurs through people’s interaction with it
(Osborne, 2014:8). Relationships between people and landscapes and monuments are
formed through “experience, perception and imagination” (Caraher, 2014:43; Smith,
2003:169). As at Rāmṭek, monuments are visible and permanent constructions but
rather than simply being “commemorative symbols of hegemony”, monuments are
active in conferring power and ideology to a particular audience (Glatz, 2014:112;
Kolb, 2014:154).
124
Chapter 4. Landscape archaeology
The following section considers the principal survey projects carried out in the
Indian subcontinent and is intended to reflect on developments which have particular
relevance to the context of this research, rather than presenting an exhaustive history
of the theoretical progression of landscape archaeology (Boivin & Fuller, 2002). The
current study hopes to contribute to the expansive body of survey projects in India
and is grounded in the tradition of landscape research and past survey approaches.
Early landscape studies in India reflected broad trends elsewhere and had clear
comparisons with exploratory surveys in countries such as Greece and Italy (Cherry,
1982). Formal surveys in India began in the 19th century under the first director
Alexander Cunningham of the Archaeological Survey of India (ASI, formed in
1861), who believed that the “careful and systematic” exploration and recording of
India’s monuments was a “moral obligation” of the colonial government
(Cunningham, 1972 (1866):iii; Trautmann & Sinopoli, 2002:497). Archaeological
observations were frequently documented in geographic exploration, which
accompanied the spread of colonial activity, and prehistoric archaeology benefited
from the explorations of the Geological Survey of India (Banning, 2002:2; Paddayya,
2002:131). The main aim of early surveys was to identify major urban and ritual
centres described in historical sources and so much of the ASI’s attention was
directed towards seeking out key Buddhist centres mentioned in the accounts of the
Chinese pilgrims, Faxian and Xuanzang (5th and 7th century AD respectively)
(Trautmann & Sinopoli, 2002:499). The ASI reconnaissance surveys were a great
accomplishment which contributed vast amounts of information, but were limited by
a fundamental reliance on historical texts (Trautmann & Sinopoli, 2002:500). The
primary focus on historical archaeological material, in conjunction with South Asia’s
rich textual record, resulted in a strong historiographical tradition where texts often
underpin the interpretation and chronological understanding of archaeological data
125
Chapter 4. Landscape archaeology
Indian settlement pattern studies of the 1960s and 1970s directed archaeological
enquiry towards establishing the location of sites in the landscape and the physical
factors influencing settlement location (for example, see Fairservis’ research in
Baluchistan (1956; 1959) which strove to relate aspects of Harappan civilisation to
ecological adaptation). The integrated investigation of sites in the landscape was an
important development within Indian archaeology, resulting in large-scale research
into regional distribution data, environmental contexts and inter-site relationships.
Theoretical approaches in Indian landscape archaeology, advocated by scholars such
as Malik, were rooted in a similar background to Processualism and stressed the
importance of pursuing a “holistic anthropological approach” (1973:7). Sankalia’s
influential assessment of New archaeology created wider discussion of theoretical
developments among Indian archaeologists as regional settlement distribution studies
were becoming integral to South Asian landscape archaeology across the late 1970s
to 1980s (Lal, 1984a:2; Paddayya, 1995:136).
126
Chapter 4. Landscape archaeology
127
Chapter 4. Landscape archaeology
128
Chapter 4. Landscape archaeology
numerous rural village shrines, and revealed the complexity of the sacred landscape
(Lahiri, 1996:245).
Shaw’s research in Sanchi aimed to consciously adapt survey approaches to suit the
environment and to improve the village-to-village approach, creating an ‘integrated’
survey methodology (Shaw, 2007:65). The Sanchi survey extended archaeological
investigation to the landscape surrounding Sanchi Hill, covering a relatively small
area within a 10-15 km radius from the major Buddhist complex (Shaw, 2007:72).
Shaw combined village-to-village survey and landscape archaeology with art-
historical and textual study, documentation of the modern cultural landscape and
interaction with communities to be sympathetic towards the local context (Shaw,
2007:67, 72). Shaw drew on the theory of ‘historical phenomenology’ within
landscape archaeology which accounts for memory, social meaning and historical
sources within investigations of the landscape (Thomas, 1991). In addition to
monuments and settlement sites, Shaw documented transient remains, ritual sites and
“natural cult spots”, and suggested that purely systematic surveys may not have the
same success in identifying certain site categories (Blinkhorn, et al., 2010; Shaw,
2007:64).
129
Chapter 4. Landscape archaeology
broader cultural and physical environment of religious sites (For example, see Fritz’s
(1986) research into the symbolic layout of Vijayanagara). This survey is centred on
Rāmṭek as a ritual site but does aim to elucidate the wider socio-political and
economic context of the hinterland in order to assess Rāmṭek’s integration into, and
effect on, the landscape. This aim develops out of studies such as Schopen’s study of
the material evidence of Buddhism (1997), Hawkes’ survey at Bharhut (Hawkes,
2006; 2008; 2009; 2010) establishing the relationship between Buddhism and society
in central India (c. 400 BC to 1200 AD), and Fogelin’s intensive survey of the
monastic centre of Thotlakonda in Andhra Pradesh aimed at establishing the spatial
and temporal connections between the site and its archaeological setting (Fogelin,
2003).
Alongside the increased use of systematic methodologies has been the aim of
undertaking intensive systematic survey through urban hinterlands to contextualise
sites in their social, political and economic landscape. The study of
urban/monumental remains and elite activities has traditionally dominated Indian
historical archaeology but holistic intensive, systematic surveys consider rural
communities through the recording of rural sites, transient remains and surface
artefacts as well as monumental architecture and large settlement mounds
130
Chapter 4. Landscape archaeology
131
Chapter 4. Landscape archaeology
4.5. Archaeological theory and practice: Progress in the field and the
Rāmṭek survey
132
Chapter 4. Landscape archaeology
There are considerable areas of land not yet explored in South Asia, which may
support the continued relevance of extensive survey. In some areas it may be more
appropriate to work towards establishing wide patterns before concentrating on
intensive investigations (Shaw, 2007:65). Village-to-village survey is a standard PhD
methodology for students in India and those in small archaeological projects with
restrictions such as permission and institutional support, strict time frames and
resources. It is important to consider the benefits of survey projects such as this one,
and whether the end result justifies a less advanced approach. The resulting data
must be evaluated in light of the methodologies employed and contextualised in
wider research to ensure its usefulness. This research concerns a period and region
which is not well understood. Preliminary data can be achieved through ‘informed’
survey and reconnaissance, which is preferable to a complete lack of data. It is
recognised that the survey results cannot form a totalising narrative and
methodological limitations mean that the data must be considered with a degree of
caution.
Scholars have highlighted attributes of the Indian environment which may affect the
success of survey and support the need for flexible approaches to archaeological
surveys. The following section provides a brief overview of some of the limitations
previously encountered by landscape archaeologists, which may impact on the
approach to survey in the Rāmṭek landscape.
133
Chapter 4. Landscape archaeology
The nature and period of the archaeological material in a study area should be a
leading consideration when designing landscape surveys, given the diversity and
complexity of archaeological remains across regions and the potential for different
timescales and trajectories of development (Paddayya, 2002:139). The nature of the
archaeological remains and the factors involved in site formation in particular
regions of India may require adaptation of research designs rather than the uncritical
application of methodologies and theoretical approaches. It may also be that a
“judicious combination of extensive and intensive survey strategies” is more suited
to investigate some archaeological research questions (Cherry, 1982:16).
134
Chapter 4. Landscape archaeology
research, researchers are aware of addressing the potential for a lack of integration
between method and theory, which in the past appears to have hampered the
development of approaches such as non-site based archaeology (Gkiasta, 2008:36;
Shaw, 2007:60; Stark, 2006:422). As rural archaeology has sometimes not been a
focus of research, landscape archaeology has not produced substantial data on Gupta
and post-Gupta rural settlement distributions compared to prehistoric and
protohistoric research (Chattopadhyaya, 1990:19). All excavations of Early Medieval
sites to date have been of ‘urban’ sites rather than rural habitations, despite historical
sources demonstrating that settlement at the time was more focused around villages
(Hawkes, 2014c:66). Furthermore, major religious and urban sites were frequently
considered separately from settlement patterns, economy and routes of movement
resulting in decontextualized sites and inadequate understanding of inter-site
relationships (Coningham, et al., 2006:55; Shaw, 2007:62). Excavated historical
cities, such as Pauni, often lack an accompanying hinterland survey, unlike Harappan
cities (see Possehl (2002) for an overview of landscape archaeology relating to the
Harappan civilisation alone) (Lahiri, et al., 2002:110). However, there is an
increasingly strong body of modern projects aimed at establishing the context of
many historical sites and most of the current foreign projects conducted across the
subcontinent are surveys (Bauer, et al., 2007; Johansen, 2009; Smith, 2000; 2002b;
Suvrathan, 2013).
135
Chapter 4. Landscape archaeology
prominent examples of large sites where no excavation has taken place to achieve a
better understanding of the Early Medieval material, such as Kalyana, which is
proposed to have been the capital of the Cālukyas and the Kalacuris (Dayal,
2005:62). Epigraphic evidence challenges the picture of “economically isolated”
villages and implies that there was a large number of Early Medieval rural
settlements (Ali, 2012:9; Dayal, 2005:61). There is growing awareness of the
possibility for a great number and density of Early Medieval settlements in some
parts of India; for example, recent investigations in Maharashtra, such as Khaladkar’s
survey in Sholapur district, and Kulkarni’s research in Osmanabad District,
demonstrate the presence of Early Historic sites and substantial Early Medieval and
Medieval settlements (Jamkhedkar, 2010:5).
Additional potential barriers to overcome are related to the sheer size of India and the
wealth of material remains across the country. The imposition of modern state
divisions within India does not necessarily reflect regional networks of activity in the
past, and regionalisation and linguistic differences can foster a lack of inter-regional
comparisons in archaeological research (Willis, et al., 2012). Furthermore, while a
number of state archaeological departments are very active, others have completed
less archaeological work, which increases the likelihood that differences in site
distributions may be more reflective of coverage and intensity of research.
136
Chapter 4. Landscape archaeology
Although vegetation cover and terrain can cause difficulties in systematic survey,
these variables do not necessarily render such methodologies impossible.
Challenging and inconsistent environmental contexts are not unique to India as
shown by the use of intensive systematic survey in Thailand (Pryce, et al., 2011) and
other diverse, global environments (see Kowalewski (2008) for an overview). By
applying “modified” intensive survey in tropical environments, it was concluded that
survey is feasible in most areas, except for extremely mountainous or jungle
conditions (Pryce, et al., 2011:63). This indicates that the environment alone may not
be prohibitive to systematic methodologies, and that obstacles in India could be
overcome. However, there must be sufficient resources available to counteract
limiting factors and certain methodologies may not be feasible in small-scale
projects. For example, within the scope of Smith’s survey of Mahasthangarh,
systematic transect survey was impossible due to irrigation and agriculture, and so
pedestrian/cycle survey was used throughout. This method was cost-effective and
appropriate to an environment with year-round low visibility and seasonal flooding,
and it was sufficient to gain a broad cross section of the archaeological landscape
(Smith, 2001a:63).
The modern anthropogenic landscape and the preservation and visibility of sites, is
another challenge to effective survey. The modern use and representation of
archaeological material, such as the continued use of religious sites or the re-
installation of sculpture and artefacts in villages, affects the ways in which
archaeologists must think about site recovery during landscape survey (Shaw,
2007:75). The ‘village shrine’ is a rural phenomenon noted during surveys by Shaw
(2007) and Lahiri (1996), and architectural and sculptural material is particularly
found on small village platforms or cabūtrās. The movement of archaeological
material to predictable places, such as village cabūtrās, supports the use of village
explorations to identify parts of the archaeological record, which relates to the sacred
landscape.
137
Chapter 4. Landscape archaeology
Parts of India are highly agricultural and a large amount of its archaeology is situated
within the ploughzone (Jhaldiyal, 1997:19). As seen in ploughzones elsewhere,
agriculture can result in the exposure of archaeological material, but traditional
ploughing in India is shallow and there is the potential for earlier material to remain
buried. Traditional farming methods are still widely used and involve non-
mechanical ox-drawn ploughs that typically only reach a depth of about 15cm
(Paddayya, 1996:78). As the plough furrows may not be deep enough to penetrate the
topsoil and expose certain artefacts, results of surface collection must be tempered
with the understanding that absences in the available evidence may be a result of
visibility or the under-representation of earlier periods. This is in contrast to the
situation witnessed in Europe, for example, as deep ploughing was a key factor in
uncovering archaeological remains and prompting enormous survey projects from
the 1960s. The depth of deposition and a lack of deep ploughing in India can affect
the likelihood of locating small or transient sites. This may have influenced how
researchers have perceived settlement patterns given that few examples of dispersed
settlement are known from South Asia and therefore have not been thoroughly
investigated.
138
Chapter 4. Landscape archaeology
Figure 4-2. A traditional plough used by farmers around Rāmṭek, with short farrows which would not
penetrate deep into the ground.
Modern use of the landscape can affect the success of survey in a number of ways as
irrigation and farming activities can alter the context and character of archaeological
sites. Site destruction through land modification for agriculture is common as ancient
material may be used along with soil to create field banks and archaeological
deposits may be levelled for cultivation (Paddayya, 1996:82). Sites have been
“erased” by intensive agriculture and reclamation of land for cultivation, alongside
the escalating use of mechanical farming machinery, land-levelling and irrigation
projects (Paddayya, 1990:55). Diverse agriculture, coupled with often irregular and
complicated land divisions, exacerbates difficulties in site identification on the
ground and from aerial imagery (Terrenato, 2000:24). Intensive cropping and dual
cropping systems limit the time and area available for field-walking. Industrial and
commercial activities pose a threat to the survival of past remains, as sites may be cut
through to quarry natural resources (Mohan, 1996; Rajan, 2003:41). In some cases
the archaeological material itself may be quarried, as noticed by Paddayya in
Karnataka where filling material for construction was excavated from Neolithic
ashmounds (Paddayya, 1996:81). Survey should be crucial as if archaeological sites
are not protected by the government, they are either in danger of being covered by
modern infrastructure or disassembled for the re-use of archaeological building
materials (1999:425).
139
Chapter 4. Landscape archaeology
Small-scale explorations are widely carried out by universities and are actively
encouraged for students as part of doctoral research in order to fill the lacunas in the
archaeological record (Possehl, 2002:8). However, difficulties can arise from the
proliferation of small projects which one should be aware of when embarking on a
similar project. The resulting information may not see a full publication unless the
research is undertaken through a larger project and it can be difficult for external
researchers to determine the number of existing survey projects published in journals
(e.g. Panja, 1996; Venkata Subbaiah, 1989). The Land, Water, Settlement Project has
noted that while valuable data is achieved from student surveys, it can be
complicated to consolidate, appraise and compare the information. Results can vary
with the use of different methodologies and the site densities of intensely surveyed
areas are difficult to compare to that of the wider landscape (Singh, et al., 2010b:45).
Moreover, inaccuracies in existing data have been noted during dissemination by
recent landscape projects. In a survey of the Rakhigarhi hinterland, Singh et al.
identified several errors in existing site registers and publications, both in site
140
Chapter 4. Landscape archaeology
141
Chapter 4. Landscape archaeology
As a free resource, GoogleEarth has great potential for small-scale projects or those
conducted in developing areas (Beck, 2006; Thakuria, et al., 2013:20). GoogleEarth,
launched in 2005, is an accessible, open resource whose availability means it can be
used in the field in conjunction with ground-survey. GoogleEarth is not as refined as
other satellite imagery for use in a GIS, but is freely available and has worldwide
coverage, which can help in overcoming project budget restrictions (Britt Bousman,
2006; Déodat & Lecocq, 2012; Madry, 2007; Ur, 2006). GoogleEarth enables broad
142
Chapter 4. Landscape archaeology
remote survey over large areas, whereas the same coverage may be prohibitively
expensive with other satellite imagery, and therefore it can encourage cross-regional
research and collaboration (Britt Bousman, 2006:34). The potential of GoogleEarth
imagery in India has been demonstrated at Sisupalgarh, where it was used to identify
walled settlements and assess their spatial distribution (Thakuria, et al., 2013:22).
143
Chapter 4. Landscape archaeology
The number of survey projects in India is increasing but the full potential of the data
remains to be fully exploited (Lal, 1984b:61). In an observation in South-East Asian
archaeology, which is equally valid for South Asia, Stark noted that there is a
particular need for survey in combination with excavations to form regional ceramic
typologies (Stark, 2006:423). An absence of landscape data can hinder
contextualisation of the excavation data, which cannot then be efficiently compared
to regional patterns. Surprisingly little is known about the nature and development of
Early Historic settlement and the relationship between urban/religious centres and
their surroundings, and so landscape survey has great potential to contribute data
related to spatial and temporal patterns of change. Historical models are limited by
invisible periods in excavations and so landscape archaeology may be employed to
establish more data and comparison material. Within historical archaeology there is
further scope for collaboration between archaeological landscape research, and other
sources of evidence such as art-history, history and epigraphy (Stark, 2006:423).
Many of the traditional issues that initially hindered the development of landscape
archaeology in India are being addressed, although there is still a need for more
research. There are many opportunities to carry out survey as only basic data has
been achieved for large areas of the country, which is insufficient for tackling
important archaeological questions of settlement, economy and subsistence patterns.
144
Chapter 4. Landscape archaeology
145
Chapter 5: Methodology
5.1. Introduction
The following chapter concerns the research design and fieldwork methodology of
the multi-period survey employed at Rāmṭek, as informed by existing landscape
research in India. This research’s methodology was constrained by the scope of the
research but the applicability of previous approaches was kept in mind and
considered in relation to the project scope and the specific landscape. The research
design was organised to maximise the success of identifying sites in the landscape
and the amount of meaningful information gathered within the project logistics. This
study contributes to the large body of landscape research in India and strives to
demonstrate the potential for survey projects with limited resources to contribute
useful data. It has been argued that in a large area such as India, student research is
best directed towards investigation of specific locations to expand our understanding
of the wider landscape (Lahiri, et al., 1996; Shaw, 2007; Suvrathan, 2013). This
study is aimed at investigating the regional development of the Eastern Vākāṭakas to
expand our understanding of the function of Rāmṭek at the centre of the dynasty.
Investigation of a tightly-focused area was aimed at gathering more information to
add to existing evidence from past archaeological explorations to build a picture of
Rāmṭek’s development as a case study for Eastern Vākāṭaka development.
This project uses a number of datasets to achieve the overall aim of the research,
which is to construct a holistic archaeological view of the landscape, testing the
narrative of Eastern Vākāṭaka expansion by establishing the longue durée of the
Rāmṭek area. The primary dataset is the survey evidence collected from three seasons
of fieldwork; a preliminary reconnaissance, a second principal survey season and a
supplementary third season to refine the survey data and carry out data analysis. The
survey data provides specific details of the archaeological evidence around Rāmṭek
and was directed towards documenting evidence related to the religious and
settlement development of the area. It was hoped that details of concepts such as
146
Chapter 5: Methodology
Research design involves a balance between project aims and practical limitations
such as time, availability of resources, funding and team size (Gallant, 1986:405).
The project aims directed the survey in that in order to test the hypothesis regarding
Vākāṭaka investment across the landscape, economic and ritual activity in rural areas
was to be investigated. Certain approaches employed in the fieldwork methodology
were necessitated by project logistics and the research design had to incorporate
limitations such as manpower, time and funding. Furthermore, additional restrictions
encountered within India had to be accounted for, including permissions and
personal transportation. The fieldwork methodology evolved over the course of the
first two field seasons to adapt to increased understanding of the survey area and
limitations encountered in the field.
The survey was intended to work from a tiered sampling strategy with a variety of
intensive and extensive investigations, designed to efficiently use the time available
and explore different types of data:
147
Chapter 5: Methodology
Level 1.
Level 2
Level 3
The ‘informed’ village-to-village survey was the main focus of the fieldwork and
was primarily restricted to the defined survey area. Modern village-to-village surveys
focus on manageable areas of land to enable more intensive coverage. While
Erdosy’s survey (1988) worked within a 10km radius of each village in an entire
148
Chapter 5: Methodology
region, this survey area is contained within a 10km radius from Rāmṭek and includes
numerous, closely-spaced villages used as bases for exploration.
Level 4
A portion of the second field season was dedicated to more intensive methods of
investigation within the survey area and tested basic systematic sampling strategies.
The original decision to extend the systematic investigations out from the survey area
was made due to the aim of seeking pre-existing Early Historic activity in order to
probe the historical narrative of Eastern Vākāṭaka development. It was theorised that
there may have been a topographic influence on the location of earlier sites due to the
proposed shift of the Eastern Vākāṭakas inland to more peripheral areas. Given the
attempt to understand Rāmṭek in its regional perspective the systematic survey was
directed towards understanding the effect of the local topography on site location or
type. However, time did not permit these wider investigations to be carried out and
the primary function of the Level 4 survey became to test the patterns emerging in
the data achieved during Level 3 survey. Hence focus was initially directed to
sampling within the survey region.
In practice, the different types of survey outlined above were found to interact and
the use of Level 1 survey was maintained throughout the period of research. Sites
could be identified through various types of survey. It was decided that if a site was
initially identified on remote sensing imagery then it was recorded as a Level 1
finding before being confirmed through Level 2 field explorations. If a site already
149
Chapter 5: Methodology
identified through remote survey was highlighted during discussions with informants
on the Level 3 survey, then it was recorded as both visible on satellite imagery and
known locally.
Published excavation and survey reports were referred to during this research to
provide a broader context for the Rāmṭek survey data. The published reports were
initially used to inform the Level 2 and 3 surveys and to establish the regional
archaeological and anthropogenic context. Furthermore, published reports were used
throughout the research in interpretation of the field data, and in comparison of the
results with known evidence. Focus was particularly paid to the reports of excavated
Vākāṭaka sites in the region, such as Pauni (Deo & Joshi, 1972; Nath, 1998), Paunar
(Deo & Dhavalikar, 1968), and Mansar (Joshi & Sharma, 2000; 2005). Reports of
relevant explorations recorded in the Indian Archaeological Review and Sali’s
overview of archaeological explorations of Vākāṭaka sites (1998) were useful
sources providing brief descriptions of a number of villages in the wider Rāmṭek
Tehsil with evidence of Early Historic activity.
Local maps and cartographic data are a significant dataset, providing information
relating to known sites and features, and locational data for modern villages.
Topographic maps provide a basic land-use profile to digitize and use as a
background for the survey data. A Survey of India topographic map of Rāmṭek
district (Narain, 1975) was used as a guide to village names and locations, temples
and irrigation facilities. This increased familiarity with the survey area to anticipate
land use, identify villages and compare to exploration data. The maps used here do
not provide the most up to date information (Sheet no. 55 0/7; 1:50,000; First edition,
1975) but more recent maps could not be obtained within the scope of this survey. As
such, the digitised land-use information was kept simple and compared to satellite
150
Chapter 5: Methodology
images from GoogleEarth. The map scan was entered into a GIS and geo-rectified
based on secure latitudes and longitudes so that map features could be digitised by
creating polygons/shapefiles of sites and features of interest.
Satellite imagery can be particularly useful in identifying patterns in site data, which
may emerge from a wider perspective as opposed to observation on the ground. An
additional strength in aerial or satellite imagery is the ability to visualise changes in
the landscape over recent history, and assess degradation of sites. This was a
principal concern in this research as clear evidence of site disturbance was
documented during the survey. Remote sensing data was used as a visual tool and
source of aerial information. The data was incorporated into a GIS to provide a
basemap for the survey data input into the GIS, and to convey information such as
elevation and topography. Satellite imagery was used in the Level 1 survey to detect
sites and for the aerial examination of sites located by field survey. As a
supplementary dataset in this research, remote sensing data was also assessed for its
use in this particular landscape and projects of this scope.
Table 5-1. The types of imagery utilised in this study. Given the financial restrictions of the project,
use was made of free satellite imagery.
Aster and SRTM data are low resolution imagery which were used to provide DEMs
(Digital Elevation Models) for the GIS. A DEM allows observation of topographic
151
Chapter 5: Methodology
variation and the imagery simplifies the landscape so it is more appropriate for
quantitative interpretation. Aster and SRTM data are freely accessible from
http://earthexplorer.usgs.gov/. ASTER (Advanced Spaceborne Thermal Emission
and Reflection Radiometer) data originated from images taken from the Terra
Satellite launched in 1999, and was made available from 2009 as a Global DEM
(Abrams, et al., 2002). ASTER imagery has relatively high spatial resolution, up to
15m (http://asterweb.jpl.nasa.gov/). The SRTM (Shuttle Radar Topography Mission)
project was carried out in 2000 by NASA, the National geospatial-intelligence
agency and the German and Italian space agencies (See Farr, et al. (2007) for an
overview; available at http://srtm.csi.cgiar.org). SRTM created topographic data
which is freely available at a resolution of 90m, which is sufficiently detailed for the
detection of larger features (Galiatsatos, et al., 2009:5-6). Landsat imagery was also
downloaded for use from EarthExplorer via http://landsat.usgs.gov/. (Landsat
programme information, USGS 2003). For coordinated use of the remote sensing
data in a GIS, all imagery must be geo-corrected to the same base co-ordinate
system. The WGS 1984 datum and the Universal Transverse Mercator (UTM)
projection system were used in this project. This handheld GPS was configured to
this system, so that the survey data was compatible and could be integrated into the
GIS. The imagery was re-projected in ArcGIS, which was the software used to
generate the GIS.
Given budget restrictions, GoogleEarth was the main resource used in planning the
fieldwork, both in identification of potential sites, and in assessment of the nature of
the landscape. GoogleEarth can be a valuable source of current land-use information
which can be transferred to the GIS and compared to other data sources. As
GoogleEarth is universally available, it could be accessed in India for aerial
assessment of sites located on the ground, although due to connection limitations it
was not available in the field itself. GoogleEarth was used alongside other remote-
sensing data later in the research as a comparative source for other levels of survey.
In GoogleEarth, the entire planet is covered by medium resolution, simulated, true-
colour imagery, which has been derived from Landsat data (Ur, 2006:35). As in other
satellite imagery one relies on visual interpretation to identify features, which
become apparent through their difference to the surrounding landscape. True-colour
imagery is helpful in the identification of possible sites in the landscape, as variations
152
Chapter 5: Methodology
in vegetation and soil colour may be recognised. If features are identified online in
GoogleEarth, they can be digitised for incorporation into the main GIS by creating
KML (Keyhole Markup Language) files.
Orbview 3 imagery was also used at the fieldwork planning stage and Level 1 survey
to initially scan the landscape for potential sites. Orbview imagery is freely available
online and is at a higher resolution than GoogleEarth, but is in black and white.
When a potential site is located, aerial imagery is useful in assessing its form and
shape, its boundaries, the nature and topography of its landscape surroundings and its
relationship to associated remains such as nearby hydrological features or other sites.
Shading differences in the landscape can indicate the presence of archaeological
remains, as past activity can alter the soil composition. Shadows could indicate
differences in elevation from accumulated deposits. The methodology for
identification of sites using remote sensing data relies largely on recognising changes
in elevation or visual anomalies in factors such as soil colour. It is not immediately
known whether the identified anomaly is archaeological or caused by another
process; soil differences may be geological or caused by factors such as drainage,
agricultural activities and varied land-use. Around Rāmṭek the general appearance of
the ground is affected by the subdivision of fields into areas of different crops.
Changes in elevation may also be connected to natural undulating topography rather
than anthropogenic activity. It is therefore important to verify remote findings by
field visits to confirm the archaeological origin and nature of a site.
Satellite imagery is particularly useful in identifying mound sites and thick scatters
of brick, as well as old tank bunds which can often be seen to correspond to those
recorded on cartographic maps. Mounds can be easily identifiable, although low
elevation can pose a problem if no shadow is evident. Conversely, apparent changes
in elevation visible on satellite imagery may also not be as prominent in reality; the
conditions under which imagery is taken and the methods used in its acquisition will
affect the visibility of sites on the resulting data. Visibility of sites on imagery relies
on a complex combination of factors including the type of imagery, the resolution,
geology, topography, environmental conditions and the nature of remains. The
relationship between ground remains and remote sensing data, and sites and their
visibility on imagery, is generally not well understood and there is limited
information from India. As it is unsure why sites appear on imagery, it is difficult to
153
Chapter 5: Methodology
define sites which will relate to evidence in the field (Galiatsatos, 2004). Remote-
sensing data may also be at too low a resolution for identifying specific types of site.
Field visits enable the collection of artefactual remains, and the recording of
categories of sites which may not be manifested on remote sensing imagery, such as
small religious sites and scatters of artefacts.
The introduction of Geographic Information Systems (GIS) has been one of the most
innovative developments in landscape archaeology since the 1990s, and has
considerably increased the analytical potential of archaeological data due to the
ability to integrate different types of information, such as survey data, environmental
parameters and remote sensing data (Gkiasta, 2008:20). GIS are primarily a
methodological tool used to emphasise the spatial relevance of environmental and
cultural systems. GIS can store, manipulate and analyse data, and can create map
visualisations of different datasets which are spatially related. As an analytical tool a
GIS can answer questions of spatial and temporal patterning and aid in the
reconstruction of past landscapes (Llobera, 2007). However, GIS are subject to the
restrictions of the data they use and the questions asked by researchers (Gkiasta,
2008:23).
During the fieldwork, data was stored in Excel spreadsheets, which recorded the GPS
numbers, feature and sculpture numbers, photograph list and information related to
the ceramic and finds collection. The assignment of site numbers to recorded features
was completed in retrospect and sites were given a unique identifying number.
Following completion of the fieldwork the Excel data was transferred into an Access
2010 Database in order to facilitate analysis and enable querying of the information.
Access databases have the advantage over Excel of permitting the linking of data
through shared attributes. The database was designed to be as simple as possible,
while still providing enough detail to allow comparisons and interpretations of the
data.
154
Chapter 5: Methodology
In this project the GIS was generated through ESRI ArcGIS and connected to the
Access database in order to combine the remote-sensing imagery, site and ceramic
data and topographic maps. ArcGIS allows the mapping of spatial data, while the
Access database stores the attribute data for the GIS. The spatial data from survey
can be displayed on the base maps from remote-sensing imagery, and features on
different imagery can be digitized using points and shapefiles. A GIS is a powerful
tool which allows researchers to query data and produce visual explanations to
display information relating to the site density and type, spatial and chronological
distributions and inter-site relationships. GIS can be used in more sophisticated
analysis, but in this research it was primarily used as a visual aid for interpretation of
the survey data.
The first consideration of the research design was the scale of the project and the
definition of the survey area. Scale refers to the size of the area to be surveyed; large-
scale surveys are appropriate for regional investigations but require relatively high
investment, while small-scale surveys are more feasible for limited projects and may
be justified by research aims, such as the desire to create a nuanced map of a specific
area. The survey region should be appropriate to the fieldwork’s empirical issues but
in practice is also shaped by additional factors, such as accessibility and feasible land
coverage.
The original survey area was designed to provide a tight focus on Rāmṭek, in order to
answer questions about the specific local environment. The boundaries of a survey
are typically delineated by natural, cultural or arbitrary attributes of the area (Plog, et
al., 1978:384). The survey area defined within this project took all of these factors
into consideration: A 10km radius was taken from the central religious complex on
Rāmṭek Hill for the preliminary field season, with the possibility of altering the size
or shape of survey area in future seasons. A 10km radius was considered manageable
for the time and resources available, being small enough to maximise land coverage,
155
Chapter 5: Methodology
but sufficiently large to reveal patterns in the data to relate to regional trends. During
the second season it became apparent that although settlement mounds were being
identified there was an absence of ancient temples, and so it was considered that if a
dispersed pattern of temples was present in the past that this may not be shown under
the restricted survey area. The potential for wider patterns of settlement and ritual
activity partly led to the original design of Level 4 transect survey and positioning of
transects to be discussed below. Archaeologically and ethnographically, there is
some justification for the delineation of a 10km survey zone. The large Vākāṭaka
period water reservoir is situated towards the centre of the survey area and so the
surrounding area would have been part of the cultivatable zone and sites would have
been directly related to the Vākāṭaka economic investment into the landscape.
Furthermore, 10km is thought to constitute two hours walking distance in site
catchment analysis; studies in India have estimated that agricultural Chalcolithic
communities in the Deccan would require an area of around 5km to support their
needs and 10km would be the maximum daily distance for farmers utilising bullock
carts (Pappu, 1988; Pappu & Deo, 1994; Pappu & Shinde, 1990), while in Sri Lanka,
a maximum of 15km a day was estimated as a possible daily travelling distance
(Coningham & Gunawardhana, 2013). Furthermore, this survey area incorporated the
Vākāṭaka capitals at Mansar and Nagardhan, which enabled thorough exploration of
these sites for comparison to the rural landscape. Physically, the size of the radius
was confined by the fringe of densely forested hills demarcating the northern part of
the Rāmṭek area and leading up to Pench national park.
By omitting the larger area of forested hills it is possible that the archaeological
traces of different types of communities may be missed. Areas of forests have been
potential areas of habitations for communities in ancient times and archaeological
research and epigraphic records demonstrate that forest groups would have co-
existed and interacted with settled agricultural communities. Forest people have also
been “integrated” into influential political structures through exchange of products or
gifting and tribute owed to kings (Hecht, et al., 2014; Morrison & Junker, 2002).
While part of the northern hills is included within the survey area it must be noted
that omitting large parts of this environmental zone risks undervaluing the
complexities of human occupation in the region. However, given the lack of
understanding of the immediate hinterland of the Eastern Vākāṭaka centre, it was
156
Chapter 5: Methodology
Figure 5-1. The 10km survey area surrounding Rāmṭek displaying the main, known sites and the
spread of modern villages.
A combination of public transport, private transport and walking was used in the first
reconnaissance season, while independent transport was available for the second and
third field seasons. Initially, the fieldwork timetable in the first season was adjusted
to account for travel limitations so that accessible areas were surveyed during the
week and travelled to by public transport, and more remote areas were surveyed at
the weekend when there was access to a motorbike. The motorbike was suitable for
off-road areas and narrow lanes, and provided a faster method of travelling across the
10km survey area, although field exploration was carried out on foot. This method of
transportation is local and familiar, and therefore cheap and discreet. During the
157
Chapter 5: Methodology
second season a motorbike was available for the entire period, which comparatively
improved the efficiency of the survey and reduced costs. However, the use of a
motorbike does insert additional biases into the survey by raising the chances of the
survey becoming unintentionally limited to more accessible areas or roadways. In
attempts to mediate this, the handheld GPS could be used to monitor daily routes and
show land coverage. The survey area was divided into manageable zones designated
for survey according to a schedule to structure travel across the landscape. It is still
necessary to keep in mind biases of coverage when discussing results and the
patterns of findings.
Over the course of the fieldwork, the team size was restricted to two individuals,
consisting of myself and a partner. During the first season, Dr K. Bhaisare, a local
archaeology research postgraduate from Deccan College, accompanied me in the
field. Mr Mustafa Ahmad assisted in the first three weeks of the second season, and
while he is not local, he is experienced in archaeological exploration and the
identification of sites. Mr Nivesh Sarjare assisted over the remaining field periods,
and was local to the area with fieldwork experience. It was invaluable to be
accompanied by an Indian partner, particularly due to language difficulties in a
Marathi-speaking area and the suspicion generated by being a foreign woman in a
rural area. Archaeological training for the field assistant was not found to
substantially affect the success of the fieldwork but other criteria such as gender, age
and personality were influential, as noted by Shaw in Sanchi (2007:67). When
accompanied by a male, attempts to engage groups of villagers (often groups of men)
in relaxed conversation were more successful and there was less reluctance for them
to divulge information on potential archaeological evidence. In addition to a field
assistant, local informants were often identified in each village through connections
made in Rāmṭek. This method provided a starting point for engaging with the local
inhabitants and removed some of the suspicion associated with our work.
158
Chapter 5: Methodology
The definition of what constitutes a ‘site’, and the extent to which ‘off-site’ features
are recorded, is a long-standing issue in archaeological research and a fundamental
decision in survey design. A project’s definition of ‘site’ relates to the aims and
research questions, and has wider implications including comparability to other
projects (Plog, et al., 1978:386). The concept of ‘sites’ varies between researchers
and projects, but a basic definition by Plog et al. takes ‘site’ to mean a “discrete and
potentially interpretable locus of material”, with sufficient cultural material to enable
inferences to be made about past human behaviour (1978:389). The question of ‘site’
is particularly important when utilising a number of datasets or methods, as each may
enable the identification of different categories of ‘site’. For example, there can be
some disparity between the types of site identifiable on the ground versus what is
visible on satellite imagery. The basic definition of ‘site’ as an area where artefacts
are found is particularly ingrained in archaeological practice; it is used as an
organisational tool for identifying discrete assemblages within a data-set and as the
base of quantitative analysis (Dunnell & Dancey, 1983:271). However, the
distribution of archaeological material across habitation areas and areas of transient
remains can be complex. The space between discrete ‘sites’ may contain extensive
cultural material, and a continuous low density spread of artefacts demonstrates that
human activity was not site-restricted (Dunnell & Dancey, 1983:272). Focus on high-
density areas may therefore exclude evidence related to more ephemeral activity
(Gkiasta, 2008:16). In an attempt to look beyond the traditional discrete ‘site’, ‘non-
site’ and ‘off-site’ methodologies have been developed, which take artefacts or
features as the unit of investigation (Bintliff & Snodgrass, 1988:61; Cherry, et al.,
1988; Dunnell & Dancey, 1983; Foley, 1981; Thomas, 1975). The idea of ‘non-site’
features has particular relevance to India as it brings attention to all parts of the
landscape including significant natural places and ephemeral remains (Shaw,
2007:62).
Some scholars have advocated the removal of the concept of ‘sites’ altogether
(Thomas, 1975), but it can be difficult to operate without this term when discussing
fieldwork results and the temporal and spatial spread of human activity. In practice,
sites may be identifiable by greater density of material remains, representative of
159
Chapter 5: Methodology
This research included three seasons, which varied in length due to project and
budget restrictions and the methodologies planned for each stage of the survey.
The first field season was intended as a general reconnaissance survey to establish
the nature of the landscape and the extent, availability and visibility of archaeological
remains. This season drew from Level 1 research and was a combination of Level 2
unsystematic investigations in the local region followed by Level 3 ‘informed’
village-to-village survey in the defined study area. The results of the first season
were intended to enable focused preparation for the second in-depth field season. The
methodology was preliminary and flexible, with the potential for modification in the
field if adaptation was necessary due to unforeseen difficulties in the previously
unvisited area.
160
Chapter 5: Methodology
The time scale of the first season was three months, as dictated by the season in India
and the demands of the research project in the UK. The first survey period around
Rāmṭek coincided with the middle to end of the dry winter season in India, which
was appropriate due to the cooler weather and relative ease of working in the field.
The general environment presented low vegetation across the hills and areas outside
of agricultural fields, with fairly dense forest on the higher elevation hill range to the
North. The rice harvest prior to the survey resulted in a high proportion of bare,
unploughed or shallow ploughed paddy fields. Standing crops of wheat, planted
following the rice harvest, were present at the beginning of the fieldwork period and
harvested by mid-February. Other crops, such as chickpeas and chillies, were grown
on a smaller scale and harvested in the middle of the field season. The fieldwork
season was not the most vegetation-free period, however the hot summer months
increase the difficulty of conducting fieldwork.
During the first month of reconnaissance fieldwork, regional site visits were made as
part of the Level 2 unsystematic survey. Two weeks were spent familiarising myself
with the region and visiting sites of the Eastern and Western Vākāṭakas and
contemporary dynasties, such as Udayagiri and Ajanta. A further week was spent
attending key Vākāṭaka sites within Nagpur district and viewing artefact collections
from excavations with other members of the project and archaeologists from the
region, including Dr Michael Willis, Prof Hans Bakker, Dr Riza Abbas of the Nashik
Institute and Dr Chandrashekhar Gupta. The sculptural evidence in the Nagpur
University and Central Museum collections, and artefact collections from various
excavations at Vākāṭaka sites, formed a useful reference for the later survey. The
sites visited included the nearby major Vākāṭaka sites of Pauni, Paunar, and
Mandhal, which provided a broader regional context to the survey. The main known
sites in the survey area (Rāmṭek, Nagardhan, Mansar and Hamlapuri) were also
visited to follow on from the short survey carried out in 2006 by Bakker. The
remaining 9 weeks were dedicated to Level 2 and 3 surveys within the defined
survey area.
During the first season of reconnaissance, survey was directed towards areas of high
potential for different types of archaeological remains to maximise returns in a short
period of time and to characterise the potential archaeology of the area. The first
season initially targeted the different environments within the survey area to form a
161
Chapter 5: Methodology
representative view of the landscape and ensure that location-specific activity was
not overlooked. The unsystematic Level 2 survey was directed towards hilltops,
forest, tanks, river banks and agricultural stretches while the initial Level 3 village-
to-village survey was conducted around a selection of village which had been
prioritised. A week was then spent on the each of the environments noted above with
3 weeks of village survey and exploration directed by information. A week was
absorbed by logistics and organisation of the fieldwork once in India.
The different environments were identified through the use of topographic maps and
GoogleEarth prior to the commencement of the fieldwork. Excluding the northern
fringe of large forested hills, there are few areas of high elevation in the survey area
and so all minor hills were investigated for archaeological evidence. In particular this
was directed towards the identification of religious remains due in part to the
historical association of Buddhist sites with hilltops and connected to the project
aims of investigating pre-Vākāṭaka activity. The northern fringe of hills was also
subject to preliminary exploration so as to ascertain whether traces of past less-
settled human activity were visible. These hills have a lack of paths and were walked
into from the roads. Most of the hills, particularly those to the north, were difficult to
access and navigate due to the lack of roads and the forest coverage and had poor
ground visibility due to leaf and shrub cover.
Areas with access to water resources were also given particular attention in the first
field season. The land around the Kanhan River was subject to speculative survey to
ascertain whether types of settlement were drawn to the riverine areas, perhaps to
benefit settled agriculture. While the river location may have shifted slightly over
time, this represents a water source which would have been present in the past. This
area was revisited through Level 3 survey in a later season due to the location of
villages along the river but in the first season targeted unsystematic surveys were
conducted. Additionally, water storage tanks were identified from the topographic
maps and GoogleEarth imagery in order to test the association between tanks and
settlement sites. Past research has also shown the connection between water and
religious activity (Shaw, 2005/2006). The tanks were assessed to see if there were
any visible indication of ancient origins and the surrounding areas were surveyed for
related archaeological material.
162
Chapter 5: Methodology
Following the targeted surveys, speculative survey was undertaken by travelling the
road networks and pathways throughout the area in order to look for prominent
features or mounds. This method is clearly biased towards larger sites, but was
successful in maximising the returns of the survey in a short period.
Key villages were prioritised for Level 3 research in the first season, based on factors
such as appearance, accessibility and location, in the knowledge that all villages
would be surveyed in the second fieldwork period. An apparently mounded shape on
satellite imagery was used to select targeted villages, as it was theorised the mounds
could relate to the continuity of settlements over time. The villages were investigated
by pedestrian survey, paying particular attention to house foundations, areas of
digging and sections for archaeological evidence. This was followed by engagement
with the local community to inquire about archaeological remains or religious sites in
the area, which may otherwise have remained undetected, leading to further
exploration around the settlement and cultivation areas.
The primary methodology of the second field season continued to be the use of Level
3 ‘informed’ survey, but a short systematic element (Level 4 survey) was also
introduced. Following assessment of the data collected during the reconnaissance
survey, the methodology of the ‘informed’ village-to-village technique was adjusted
to counteract some of the limitations first encountered. The 10km radius of the first
season was found to be manageable and provided a specific focus for the fieldwork.
As there was a large area still to cover within that defined radius, and it was still
appropriate to the aims and needs of the project, the boundaries of the survey
universe remained the same during the second ‘informed’ survey. Occasional
explorations were made beyond the 10km radius based on information regarding
areas with high potential for sites, but were kept within Rāmṭek Tehsil.
The beginning of the second field season coincided with the rice harvest and the
ploughing of the fields around Rāmṭek. The second season extended the time in the
field to different months to allow a comparison of visibility and included the period
following the monsoon. The later months of the field season were the same as the
163
Chapter 5: Methodology
previous year, which was previously found to be practical in terms of heat, ease of
working and vegetation cover. Unfortunately, the start of the field season was
affected by unexpectedly high vegetation cover until early December. The high
vegetation was a consequence of late rains, resulting in a late rice harvest and
extended growth of natural flora. Consequently, the ceramic sorting originally
scheduled for the end of the fieldwork period was moved forwards and initial Level 3
investigations began within and around villages while the fields were less accessible.
In total, 4 weeks were spent on pottery sorting and analysis at the Nasik Institute, 14
weeks were spent conducting Level 3 village-to-village survey and 3 weeks were
dedicated to the Level 4 transect-based survey.
The results of the first season indicated that the land coverage by pedestrian survey
had been more limited than expected. The small team size and lack of permanent
164
Chapter 5: Methodology
The ceramic and finds analysis was the primary focus of the third season
(methodology discussed below). The third season also incorporated an additional
short field period to refine aspects of the data collection. This was planned for late
March 2014 to minimise vegetation cover following difficulties with seasonality and
weather during the second season.
Recording of sites in the field was carried out free-hand in a field diary, which
minimised the amount of equipment and was more manageable than loose recording
sheets. However guidelines for data collection were formulated to ensure comparable
results. Sites and features were recorded with brief descriptions and rough
measurements.
All sites were recorded using a handheld GPS (Garmin Etrex 30; GPS accuracy is
15m or less) and the relevant features were photographed. Features were initially
assigned a temporary loci number which could be reorganised at a later date, given
that the nature of the archaeological landscape was unknown. Site boundaries were
visually estimated with their extent being primarily determined on the basis of sherd
density and related features. The size of sites was based on the presence of brick
remains or the density of ceramic material (10 sherds per metre2) and the scatters of
165
Chapter 5: Methodology
archaeological material constituting a site were surveyed to ascertain their nature and
extent. For larger scatters the perimeter was recorded with the GPS, whereas small
discrete features, such as pottery scatters or small shrines, were recorded by a single
GPS reading taken from the rough centre point. Larger features, such as mounds and
earthworks, could be circumnavigated with the GPS to estimate the area and rough
boundaries of sites while the height of mounds and earthworks were judged by sight.
The time spent walking and investigating the extent of sites was variable according
to site size and the necessary recording of features however the collection of ceramic
remains was limited to two hours.
Isolated sculptures and those contained within temples were recorded in situ with
GPS, photographed and described to record their size, material and preservation.
Occasionally, access to temples or close proximity to sculptures was not permitted,
which restricted the ability to measure sculptures. Where possible a 5 or 10cm scale
or a half metre measuring pole was used in images but unfortunately a limitation of
the results is that not all sculptures were measured in the field. The information
relating to each sculpture was kept in an index so that individual pieces could be
identified when assessed for iconographical details and chronology. Most of the
modern shrines offered no obvious archaeological remains and while their location
was recorded, collections of stones and clearly modern sculpture were not. At
temple, notes were made on architectural features for comparison to published
examples. This involved attempts to identify the material used in construction, the
details of architectural decoration, construction and column types and possible
modern adaptations to designs.
Details about the surrounding environment were recorded, such as modern land-use
and site disturbance, soil type and vegetation cover, in order to assess visibility.
Visibility impacts on the recognition of sites and is a central issue in archaeological
survey during both data collection and the subsequent analysis and interpretation
(Terrenato & Ammerman, 1996:92). This may refer to the effect of the survey
environment on site recognition or the obtrusiveness of certain categories of finds,
and it is particularly important in the recovery of low-density surface archaeological
material (Schiffer, 1988:475). The Rāmṭek landscape is generally agricultural or
forested, and so it was necessary to construct a methodology to maximise site
identification and take into account the potential effects of vegetation on site
166
Chapter 5: Methodology
recovery. Visibility can vary over the course of a single survey project due to factors
such as seasonal vegetation growth or changing land-use, and this was apparent
during the Rāmṭek survey (Ammerman, 2004:178). The same month over
consecutive years presented different levels of vegetation cover as a result of annual
climatic variation, and modern activities. Exposure of archaeological deposits may
also be influenced by agricultural practice (Gallant, 1986:416). There have been
numerous studies into the effect of agriculture on surface assemblages, particularly
those which consider the effects of ploughing on the formation, visibility and
recovery of surface assemblages and the subsequent interpretation of such
assemblages (Ammerman, 1981; 2004; Francovich, et al., 2000; Haselgrove, et al.,
1985). While there is the potential for high disturbance from ploughing, there is a
tendency for shallow ploughing in India, as mentioned in the previous chapter. The
majority of farmers observed around Rāmṭek utilised traditional non-mechanical
ploughs and so artificial cuts and sections were deliberately sought out as a means of
testing the differences between buried and surface assemblages.
Portable small finds were recorded by GPS, photographed, assigned a finds number
and often collected if loose on the ground. Few portable finds were identified during
the two survey seasons, including worked stones and ceramic artefacts such as
spindle whorls and figurine fragments. The finds were recorded, photographed and
described.
167
Chapter 5: Methodology
been adopted from statistics in order to enable quantitative evaluation of the evidence
(Ammerman, 1981:78; Nance, 1983). ‘Grab sampling’ involves the collection of
artefacts from a site in a designated period of time. It is subject to selection biases as
larger, prominent artefacts can be overrepresented and is not specific in terms of
collection areas or exact artefact provenance. However, it is kept in use by its ease
and speed of application (Plog, et al., 1978:406). Judgement-based samples are less
quantifiable but can maximise the likelihood of achieving a useful assemblage for
analysis in a limited time (Mattingly, 2000:9). The sherds at sites with a small and
clearly defined area under 2 m² were collected in total, while larger sites were
subjected to a timed collection following recording of the site features (one hour to
two hours dependent on estimated size) by walking the determined extent of the site
and picking up visible ceramics. Although larger diagnostic sherds tended to be more
visible for selection, an effort was made to collect a sample representative of the
wider assemblage (Mattingly, 2000:7).
Over the course of the first two field seasons, the majority of sherds were recovered
from the surface and sections of habitation mounds, with further remains collected
from low density surface ceramic scatter or cuts in modern agricultural and
settlement areas. The ceramics and small finds were stored at the Nashik Institute,
before being subjected to detailed recording in season three.
168
Chapter 5: Methodology
Systematic surveys have been very successfully applied in South Asia and are
generally considered more accountable than unsystematic surveys given the
production of data from which it is easier to glean statistically meaningful results.
The main concern with a village-to-village methodology is that the fundamental
reliance on local information increases the likelihood of omitting certain types of
data and makes it difficult to assess how representative the data is. It becomes almost
impossible to apply any meaningful statistics to the results of unsystematic survey as
one cannot be sure of the absences identified. The decision not to utilise systematic
survey for the entirety of the research was directed by limited resources and attempts
to minimise the risk of collecting significantly less data than would be achieved
through more flexible survey methodologies. However, an element of targeted
systematic survey was included in the second fieldwork season to create an
integrated survey methodology and to enable comparison of the unsystematic and
systematic methodologies. Over the course of the second field season, three weeks
were dedicated to the systematic survey, which accounted for around 14% of the
time.
The systematic survey was not intended to be a full-coverage exercise as it was not
feasible to systematically cover the entire survey environment. Rather, a sample-
based systematic survey was appropriate for the needs and means of this project. The
relative benefits and disadvantages of sample-based versus full coverage surveys
have been widely discussed in literature (Fish & Kowalewski, 1990; Kowalewski,
2008; Terrenato, 2004). A key issue is that of intensity, which refers to the level of
detail in examining the landscape, and may relate to the spacing of archaeologists in
systematic survey or the percentage of land coverage (Plog, et al., 1978:389). Lower
intensity projects require less investment and time while high intensity projects are
typically more expensive. It is generally felt that a higher percentage of sites will be
located with a greater level of intensity and a higher intensity of survey may be
required to identify less prominent sites (Plog, et al., 1978:394). A random selection
of sample areas in systematic survey is intended to remove the tendency to seek sites
in areas of high probability, which is an issue in the ‘informed’ survey.
169
Chapter 5: Methodology
Firstly the systematic survey worked towards moderating any bias from the
‘informed’ survey and was aimed at increasing confidence in the findings from the
informed survey by substantiating or refuting the inferences made from unsystematic
data collection. It was hoped that the results would either support the original
findings and absences, or indicate that the unsystematic techniques were less able to
identify certain categories or patterns of site.
Secondly, due to the research aim of attempting to situate Rāmṭek in its landscape
related to the historical narrative, the decision was made to extend the systematic
field-walking outside of the original survey region and across the topography of
Rāmṭek district. This was theoretically directed towards testing the connection
between past settlement or religious activity and topography. A long transect
measuring 5 by 30 kms was partly laid across the original survey area, extending
from the hills in the north of Rāmṭek through to the Kanhan River to the south. Given
enough time, the intention was to explore outside the original survey area to look for
changes in the distribution, number and size of settlements across different
environmental conditions. Through the placement of the transect across the
topography, it was hoped that patterns would be discernible which could test the
suggestion that Vākāṭaka sites shifted inland in hillier regions, while earlier
settlements tended to be distributed closer to major rivers.
170
Chapter 5: Methodology
Figure 5-2. The randomly generated transects across the 10km survey area.
171
Chapter 5: Methodology
restrictions. Creating random horizontal and vertical strip transects across Rāmṭek
itself would have been more appropriate and should be carried forward into further
research. Regional investigation could then be reserved for a later study with a
broader scope.
Walking long transects was found to be impractical due to the extensive restrictions
on access and the highly agricultural landscape. Moreover, wide spacing would have
been necessary to cover the transects in the time available and would have been
impractical with only one GPS and two researchers. Initially, modern field divisions
were considered for the unit for systematic field-walking (as in Carreté, et al. (1995)
in Tarragona, Spain), but the Rāmṭek field systems were judged to be too complex
and irregular to efficiently carry out this method. Therefore, a regular grid of
quadrats was imposed within the transects to structure the systematic survey.
Alongside regulating the field-walking, this technique has a further benefit in
reducing the work required to digitize field boundaries in the GIS (Mattingly,
2000:8).
Figure 5-3. The randomly selected quadrats within the transects across the original 10km survey area
172
Chapter 5: Methodology
Quadrats have been successfully used in survey projects (See Mueller, 1974; Pryce,
et al., 2011), but transects are frequently argued to reduce distortion in site
distribution, more clearly display relationships between sites, and result in a higher
rate of site discovery (Plog, 1976; Plog, et al., 1978:401). In reality, there may not a
superlative unit size or shape as the effectiveness and applicability of sample units
relies on a multitude of factors, including the nature and distribution of
archaeological remains, practical considerations and the intended project aims
(Schiffer, et al., 1978:11). The decision to use quadrats was based on manageability
and appropriateness; following observations made by Gallant (1986), quadrats can be
easily located and delimited, and are easily covered by a small team, leaving less
room for error. Quadrats are flexible and the selection can be shifted to a
neighbouring or nearby area if there are issues of access or visibility. The selected
transects were divided into 100m² quadrats, of which a random sample of 25% (75
quadrats) were selected for intensive field-walking. The sample of quadrats used in
this study was considered to be a manageable area of terrain given the project
constraints, and it provides a sufficiently large sample for inference. The randomly
sampled quadrats and the background transect grid were loaded onto a handheld GPS
allowing quick and accurate navigation to the sample units. A different GPS model
was used during the second field season, which allowed the upload of custom maps
(Garmin Etrex Legend H; the resolution and error margin is consistent across both
models). The discovery probability of sites has been shown to increase as the
walking interval decreases, and so this interval was kept to a practical minimum
(Schiffer, 1988:475). The quadrats were walked at intervals of 5 metres, which was
easy to navigate and structure, and ensured careful and systematic coverage with
minimum deviation.
Data was recorded per quadrat, relating to ground visibility, land-use and relevant
attributes such as hydrology or settlement development. Stratified sampling is
common in systematic survey which allows the effective assessment of visibility, but
stratifying the survey area on the basis of different land-use and vegetation was not
beneficial as transects mostly crossed areas under cultivation. Therefore, the
methodology used followed that of Gallant (1986), where a system was introduced to
rank the quadrats in terms of percentage of ground visibility. The scale ran from 0 to
10 with 0 representing 0% visibility and 10 being indicative of 100% ground
173
Chapter 5: Methodology
visibility (Gallant, 1986:406). Some quadrats fell across areas of forest or high crop
coverage, but these were walked if possible to avoid skewing the distribution of
random samples. If quadrats fell over inaccessible private land or across bodies of
water the nearest accessible quadrat was walked, with the change marked for later
reference. Surface artefacts were collected per quadrat to indicate patterns in artefact
density and any ‘sites’ were flagged for recording after the quadrat was completed.
This field-walking process was loosely timed in order to structure each day, and
ensure the targeted quadrats were covered in a timely fashion. Given the 5 metre
spacing of field walking within the quadrats, each of the two researchers had to walk
10 strips intensively. Each strip was limited to 10 minutes of walking so that each
quadrat was initially walked within two hours with further time then given if sites of
interest were located. A total collection of visible remains were carried out regardless
of the preservation of surface material and quantities of sherds were recorded per
quadrat. The total quantity for each quadrat could then be compared to visibility to
assess any possible correlation.
The following section outlines the approach to the study of the ceramic assemblage
identified during the village-to-village and systematic survey discussed above. With
no existing typology to work from, the initial ceramic analysis focused on
development of the classification schema to identify fabric wares and vessel shape
types, followed by seriation of the ceramics to construct a feasible relative phasing.
The first stage of sherd sorting was carried out at the end of each fieldwork day in
Rāmṭek, when collected sherds were washed, dried and numbered. The sherds were
assigned ascending numbers with the prefix ‘R’ and marked using black waterproof
ink. As the sherds were numbered they were initially catalogued into an excel
spreadsheet to record the following basic details:
174
Chapter 5: Methodology
Sherd number
GPS location
Site name/number
Date of collection
Type of sherd (rim body, base etc.)
Quantity of sherds (if the sherds were broken from the same piece).
Decoration details
Additional comments
No attempt at outlining the wares was made at this stage, but the material was
subdivided into three major categories; micaceous, red and grey wares.
Following collection of the entire assemblage, sherds were categorised and divided
into ‘wares’ on the basis of fabric, surface treatment, inclusions, colour, and
indications of firing and method of manufacture. A ware may be thought of as a
group of pottery with a consistently similar appearance, composition and method of
production, taken to indicate an origin from the same tradition, historical period,
region or in some cases from the same location – although such precision is rarely
possible. The fabric refers to the physical material of the ceramic (the clay, temper
and other inclusions) and its properties, and observable variations in quality,
inclusions and temper may relate to differences in production or raw materials. An
‘intuitive’ methodology of ware classification was attempted; this involved visual
examination of the material in order to discern groups, which could be later linked to
published wares and classes as possible (Sinopoli, 1991:49). This is a common, quick
and often effective method suitable for basic analysis and for familiarising oneself
with a new assemblage, but it is relatively subjective (Sinopoli, 1991:51). Initial
attempts at ‘intuitive’ sorting encountered difficulties due to the large amount of
minor and apparently irregular variation across the red and micaceous wares and the
apparently continuous scale of size and frequency of mica within the micaceous
175
Chapter 5: Methodology
Firstly, the numbered sherds were recorded individually, which was possible due to
the relatively small yield of pottery. At this stage, simple techniques were used to
assess individual qualitative attributes (the main characteristics recorded can be
found below). When describing fabric and slip colour, use of a Munsell soil colour
chart is ideal to ensure standardisation and comparability to other studies. However,
in this assemblage a large degree of colour variation was observed within and across
wares. Consequently, colour was used as a helpful descriptive tool but was not
treated as a defining feature, as in Smith (2000), and use of a Munsell chart was not
considered worthwhile at this stage. Individual recording was employed to ensure
that each sherd was adequately described so that in the event no distinct ware
divisions emerged during ‘intuitive’ sorting, more defined wares may become
apparent in the data. This method had the additional benefit of familiarising the
author with the assemblage. During this stage all sherds were photographed bar
extremely fragmentary or eroded pieces. Sherds were returned to broad ware groups
prior to the second stage of ‘intuitive’ sorting.
Table 5-2. Outline of the characteristics recorded for each sherd during the first stage of individual
sherd analysis.
Category Explanation
Sherd Sherd number
Final ware This field was entered later following the ‘intuitive’ sorting, and
gives the final ware code for each sherd e.g. ORM4.RS
Preliminary ware Original ware description used in individual recording of sherds,
e.g. Orange Micaceous, Red, Grey
Sub-ware Original sub-ware description used in individual recording of
sherds, related to slip e.g. Red slipped, Black slipped
Sherd type Sherd type code, e.g. body, base, rim
176
Chapter 5: Methodology
Exterior slip/surface Colour of the unslipped ceramic surface or the slip (subjective
colour judgment using simple terms), e.g. Grey; dark brown; orange-red
Interior surface Completed if a specific surface treatment was visible on the vessel
treatment? interior, e.g. black slipped; burnished black slipped; mica wash.
Otherwise left blank.
Core Colour Core colour (subjective judgment using simple terms), e.g. Grey;
black with brown margin
Fabric Medium-fine; medium; coarse; very coarse
Feel Smooth; rough; gritty; powdery/soapy
Inclusions No obvious inclusions; grits; sand; vegetable temper etc.
Fabric Mica Size Size range of the mica included in the ceramic fabric based on
rough size divisions (in mm), e.g. None; trace; and between small
and very large.
Fabric Mica Amount Fabric mica amount on a scale from 0 to 4, based on estimated
percentage
Thickness (mm) Example thickness of sherd body taken using digital callipers
The ‘intuitive’ sorting of specific wares was carried out in the third season of
fieldwork. The decision to use ‘intuitive’ sorting of wares was based on the
overwhelming quantity of fairly utilitarian red and micaceous ceramics, which tend
to be very loosely described or overlooked in publications. Furthermore, there are
few published sequences related to the period and region in question. Within the
ceramic assemblage many of the ‘well known’ types of pottery, which are more
deeply established in archaeological reports, appeared to be absent. It was also clear
that published reports often referred fairly generally to broad groupings which did
not account for the degree of variety presented by this data. Therefore the three main
groups of micaceous, red and grey fabric sherds were subdivided into wares by
arranging them by attribute. Inclusions were taken as a defining feature of wares
alongside core colour and surface treatment, and were identified from visual
observation of the cross section of the clay. Inclusions vary across the wares but
commonly include small grits, mica and vegetable temper. Red or micaceous sherds
were initially arranged in terms of having a uniformly coloured interior, or a
grey/black core, and then by inclusions and fabric appearance or coarseness. The
177
Chapter 5: Methodology
The rim sherds were the initial focus of attention as they tended to be larger and there
was a manageable quantity for detailed analysis. Once the main ware divisions were
established, body sherds were matched to the example rims. If a new ware was
encountered within the body sherds, example pieces were placed on the table for
reference. Occasionally sherds could not be ascribed to a particular ware, leading to a
small proportion being classed as ‘over-fired’ or ‘non-ID’. Once all the sherds were
divided into ware bags, the sherd numbers were recorded against an abbreviated ware
code. Several ware classes contain slipped and unslipped varieties of comparable
fabrics, which were bagged separately, and a suffix was added for slipped sub-
groups. For example, the ware name ‘orange micaceous ware’ (ORM) would refer to
unslipped sherds, with the slipped sub-group being labelled ‘red-slipped orange
micaceous ware’ (ORM.RS). Broader grouping of fabrics, which share key defining
characteristics but were dissimilar enough to be separate wares, were brought
together under ‘ware families’. This created another level of sorting with which to
work within the assemblage, and later aided in assessing broader patterns of ware
association.
Once the fabric categorisation was complete, a ware description was written and
example sherds were photographed. The average colour range of each ceramic ware
established through comparing characteristic sherds to a Munsell colour chart (1994).
During this stage sherds were marked as ‘missing’ if they had been misplaced, or
damaged in transport to the point where their number was not legible. This
constituted 19 sherds of the total assemblage, creating a negligible error margin of 19
out of 5962 (0.32%).
178
Chapter 5: Methodology
The wares identified during the ceramic categorisation were compared to published
ceramics to suggest chronological affiliations and usages. The chronology of
available ceramic sequences has primarily been established using relative dating
methods resulting in comparatively little firmly dated ceramic material for this
region. Early Historic sites have been excavated in Vidarbha and central India, but
the number of fully published sequences relevant to this study is low and they are
often quite dated. The published comparative data in this present study primarily
originates from within the general region, with particular use being made of the
Pauni and Paunar excavation reports. The closest comparison material would be from
the excavations at Mansar, Hamlapuri and Mandhal, but these are not yet published.
Other sequences from across Maharashtra have been used, such as those from
Brahmapuri, Bhokardhan, Kaundinyapura, Nevasa and Nasik, but due to the limited
number of local comparisons, reference has been made to sites in Gujarat and
Madhya Pradesh. The relevance of published examples from sites so far from
Rāmṭek may be questioned, and depends on the parallel development of style and
technology over quite large distances. There is obviously a risk of making false
comparisons. In addition to the use of published sequences, ceramic wares and types
may be identified through reference to ceramic material held in collections. The
ceramic assemblage from the Mansar excavations within the survey area is
stratigraphically related to the periods of interest, but was inaccessible. A few
photographs are available in an online depository (http://mansar.eldoc.ub.rug.nl/),
and the British Museum Wellsted collection contains a small selection of well-
preserved, complete vessels from an in situ 5th century location at Mansar. Limited
Early Historic ceramics were viewed at Deccan College and Nagpur University, but
overall the ability to compare the Rāmṭek ceramics with physical related material
from the region was restricted.
179
Chapter 5: Methodology
developed in the next stage of the ceramic analysis. Broad vessel types are usually
related to function and include bowls, dishes, jars, basins and cooking vessels. As
such, consideration of different vessel shapes and function may indicate the nature of
activities carried out at sites. The divisions between these types are typically based
on the proportions of the vessel, for example jars are ‘restricted’ vessels with an
opening narrower than their maximum diameter reflecting their use in storage, while
bowls have ‘unrestricted’, wide openings for serving (Rice, 1987:212). Shape
typologies tend to follow country-specific conventions, which take into account
typical vessel types in the cultural context. In India a number of characteristic vessel
shapes are defined in published sequences, such as the ‘handi’ which is a type of
deep, restricted cooking vessel. The basis for creating typologies is usually the
assessment of pottery dimensions (Miller, 1985:49). With surface assemblages with
no contextual information, ‘stylistic analysis’ and assessments of how far different
forms fit a “model of variability”, may be all that is possible through comparison to
established sequences (Miller, 1985:199).
Shape type is primarily denoted on the basis of rim form, and so diagnostic rims
were separated from the main assemblage to be intuitively sorted into groups with
similar profiles. The shapes were sorted by first associating them with a general
typology such as that in Miller (1985), which separates out bowl, globular pot and jar
forms. The rims and sherds were then compared to seek out groupings with similar
dimensions of rims, necks and vessel bodies. The dimensions under consideration are
often not directly comparable and show much variety regarding particular
orientations, heights and alignment (Sinopoli, 1991:45). A great degree of variation
is possible with Indian ceramics and there is a large variety of different types. Form
variability is often assumed to reflect functional differences but differences can also
be attributed to factors such as different potters, material, “heterogeneity of the social
context” and interpretation (Miller, 1985:51, 202). Despite great variation, a number
of fairly well-defined types and more common shapes emerged. If the overall form of
the vessel was sufficiently similar to suggest an intended shape, minor variations
were attributed to the imprecision of manual production and the rims assigned to one
type. However, if coherent groups of rim variation were noted across similar shapes,
these were divided into separate but related types. Once types were assigned to the
180
Chapter 5: Methodology
ceramic rims, these could be compared to vessel shapes in published reports in order
to consider chronology and purpose.
Figure 5-4. Rim types 147 and 148, which display a similar overall vessel shape of a double lipped
rim, but were split into two types based on coherent groupings of variation.
Base pieces can carry distinctive attributes to contribute to a shape typology but
those collected during the Rāmṭek survey primarily belonged to one broad type.
Nearly all base sherds appeared to be from a simple bowl form with a string cut base
(Type 11), occurring in coarse red or micaceous wares. Rims considered too
fragmentary or degraded for an exact type classification were recorded as ‘eroded’.
Once all rims were divided by shape, the sherd numbers were allocated a type
number. Numerous unique types were identified and were assigned a type number,
but noted as being ‘unique’. Drawings were made of characteristic examples of each
rim type, unique pieces, example bases/handles, sherds displaying distinctive
decoration, and forms such as ‘sprinklers’ or spouts. The author completed 279
illustrations, with a further 55 drawn by Ben Saunders (Cambridge Archaeology).
The illustrations were digitised by Vicky Herring (Cambridge Archaeology), which
enabled quick comparison of the rim profiles with those in published excavation
reports to find parallels and identify any chronologically distinct types.
181
Chapter 5: Methodology
Figure 5-5. A typical string cut base, commonly seen on type 11 within the ceramic assemblage. The
vast majority of bases identified were of this type.
The ceramic data was recorded in Microsoft Excel and later imported into an Access
database. The tables contain information relating to sherd location, the basic visual
characteristics identified during individual sherd analysis, the final wares following
the ‘intuitive’ sorting and the type numbers and attributes.
The ceramic remains recovered during survey are invaluable for investigating
settlement activity, given the dearth of other archaeological evidence. However, the
study of this data is impaired by significant limitations, which stem from the nature
of ceramics themselves, the context of this project, the state of regional research and
the quality of reference material.
182
Chapter 5: Methodology
remains still lack a secure context. Aside from uncertainties related to site formation
and post-depositional processes, it is not certain that visible surface ceramics reflect
the complete subsurface picture as shallow ploughing may expose only the latest
ceramics present at sites (Millett, 2000:53). Surface ceramics lack associations with
datable stratigraphy or chronological indicators, and so other methods must be
sought to date the ceramics. These may include scientific methods, seriation or
comparison to published sequences (Millett, 2000:53). Absolute dating can be
carried out using techniques such as Thermoluminescence dating, but there was no
permission or funds to attempt scientific dating in the present study. Under such
circumstances ceramic sequences are usually constructed through comparison to
published stratified sequences from excavated contexts. Thus the quality of ceramic
dating will be reliant on the quality of the comparison data, and to an extent the
experience of the researcher (Millett, 2000:54). The accuracy of ceramic
chronologies can be reduced by subjective judgment, particularly if categorisation of
the ceramics is made difficult by sherd condition (Millett, 2000:54). Surface sherds
such as those in this assemblage are more likely to be fragmentary and eroded due to
agricultural processes and general exposure on the ground.
The objective of the ceramic classification was to intuitively define the wares are
types as the quality of reporting of excavated assemblages can limit attempts to
establish parallels between them and surface assemblages. Vessel types and wares
may be only imprecisely described, which complicates comparisons (Khare,
1969:28). Older reports often do not consider residuality of ceramics and the dating
may be questioned if no information is provided on the prevalence of types or
whether all types were reported. Therefore, while ‘period-specific’ types may be
outlined, it may not be clear to what extent these types persisted. Culture-historical
approaches, which treat ceramic change as a proxy for cultural, economic and
political change, tend to focus ‘distinctive’ ceramics assumed to be representative of
certain cultures, often to the exclusion of common, everyday pottery. Issues with
ware classification can even extend to ‘distinct’ ceramics considered to be
characteristic of specific periods, as a range of similar ceramics can be amalgamated
under general terms (Smith, 2002a:142). Well-known ‘pan-Indian’ types such as
NBPW, considered to be highly distinctive Early Historic ceramic, have now been
shown to display internal or local variability and a wide distribution across different
183
Chapter 5: Methodology
Early Historic contexts and the general relationship between remains are poorly
understood, and uncertainty is increased by the reliance on artefact association to
date ceramics, particularly in potentially disturbed contexts. Vākāṭaka period
material remains are generally not well understood, and the contemporary ceramics
have not been fully characterised. Although the Paunar excavation report contains a
useful discussion of Vākāṭaka period ceramics, those from other known Vākāṭaka
sites are not published. Changes in ceramic wares and types over time are dated by
184
Chapter 5: Methodology
their association in stratified contexts with other, more inherently datable artefacts,
such as coins; a general rarity of Vākāṭaka coins has hindered the identification of
this period’s representative ceramics. Coins can only provide a terminus post quem
for deposits, but it seems likely that the overabundance and residuality of the
preceding Sātavāhana coins, alongside a lack of Vākāṭaka coins, may have resulted
in stratified layers being misidentified. Existing published sequences were influenced
by prevalent contemporary theoretical movements, notably that of Early
Historic/Early Medieval deurbanisation which presupposed decline or a lack of
Vākāṭaka activity. Hence, the Vākāṭaka and Post-Vākāṭaka phase is often
disregarded in excavations as being absent or constituting re-used or degenerated
material. At Pauni, the post-Sātavāhana period V (i.e. Vākāṭaka, 3rd to 6th century
AD) is described as disturbed, represented only by buildings of re-used material
rather than ceramic remains, and containing artefacts of mixed chronology (Nath,
1998:9, 14).
5.6. Conclusion
186
Chapter 6. Overall results
6.1. Introduction
During the three field seasons, numerous categories of sites and archaeological
remains were recorded and so the field survey achieved its objective of exploring the
rural hinterland around the major known sites at Mansar, Nagardhan and Rāmṭek in
order to increase our knowledge of the archaeological landscape. A key result has
been the documentation of unrecorded sites of past settlement activity and a number
of smaller or more transient sites. This chapter will summarise and evaluate the
overall findings of the fieldwork, prior to a more in-depth discussion of the
archaeological evidence pertaining to the economic and religious landscape in the
following two chapters.
Firstly, the fieldwork methodology will be appraised, and the applicability and
relative successes of the ‘informed’ versus systematic approaches to archaeological
surveys will be considered. This will lay the foundations for subsequent discussions
of the fieldwork results, which must be interpreted with an awareness of limitations
and biases inherent in their methodologies. Secondly, this chapter will summarise of
the number and type of sites encountered, after which it will present the working
chronological framework for the project. The broad chronological divisions used in
this research relate to the data collected and are comparable to those in previous
archaeological work from the region. This working chronology provides the structural
framework for future discussion of the proposed development of the landscape in this
region.
The initial preparations for fieldwork, conducted through the use of the Level 1, desk-
based research, successfully established the state of the art, contextualised the study
187
Chapter 6. Overall results
and supported preliminary explorations into the nature of the Rāmṭek landscape.
Survey using satellite imagery has not been published for this specific area and so the
data was used to assess the extent of its usefulness in interpreting the archaeological
landscape in this kind of research. Overall, the satellite imagery was useful as a tool
for supporting survey, but it required ongoing comparison with data achieved in the
field, as it was difficult to remotely identify sites. Few potential sites were identified
in advance, perhaps disguised by the complexity of the agricultural landscape,
although several mounds were clearly visible which were later found to have
habitation remains. Mounds and earthworks were unexpectedly challenging to
recognise on satellite imagery, which is perhaps due to the undulating plain and their
relatively low profile, being elevated only a few metres above the plain. The shapes
of field-systems or villages were also not particularly reliable at indicating the
presence of an archaeological site. As noted by Thakuria, et al. (2013), ground-
'truthing' is required to reasonably assess the features of any potential site identified
remotely.
188
Chapter 6. Overall results
may be possible to identify other potential sites on the imagery in areas where no sites
exist now, indicative of site destruction. However, it must be clear that there is no
evidence to support identification of features as archaeological sites without ground
survey.
While Level 1 remote survey may have been more limited as an independent means
of investigating the archaeological landscape, it was found to be useful in assessing
the broad landscape and supporting the Level 2 and 3 surveys. The continued
reference to satellite imagery throughout the field seasons resulted in the
identification of several archaeological signatures on satellite imagery following
recording of these sites on the ground. For example, site 262 is a very low mound
which has been significantly disturbed by farming, but has high density brick scatter.
When the coloured satellite imagery is reviewed this is displayed as an area of ground
with an orange tone, initially not recognised against the variable ground cover of the
surrounding fields. This discovery led to a second review of the imagery, which
resulted in the discovery of one site (276), identifiable by a circle of lighter coloured
soil. This was not visible on GoogleEarth, which instead shows the extent of
ploughing, but panchromatic Orbview 3 data displays the variation clearly. This
demonstrates the benefit of using different satellite imagery in combination as
resolutions vary and a comparison of black and white to colour imagery may more
clearly indicate whether there is a feature or whether the anomaly is more likely to be
an artefact of the photographic/weather conditions at the time. Variations in soil
colour, such as in this example, may be a result of anthropogenic activity but this
feature is not entirely consistent; in some cases, areas of the landscape display a
similar anomaly due to modern differences in land-use. Further survey of the satellite
imagery did not reveal potential areas of habitation outside of those recorded on the
ground through Level 3 survey. This may be interpreted as supporting the patterns
observed during the ‘informed’ survey or it may be related to low visibility or
survival of archaeological remains. It is probable that with higher resolution imagery,
the success of remotely identifying potential sites would be increased.
189
Chapter 6. Overall results
Figure 6-1. GoogleEarth colour imagery of the low mound at site 262. No shadow is cast from the
mound which has been flattened by agriculture but the surface brick scatter is visible as orange
colouration of the ground.
Figure 6-2. Site 276 as seen on GoogleEarth colour imagery. The ploughing of the field is visible and
obscures the shape and tone of the mounded area. See in comparison to Figure 6-3.
190
Chapter 6. Overall results
Figure 6-3. Site 276 identified during survey of the Orbview 3 imagery; the shape of the mound is
more clearly visible than on the colour imagery and the variation is soil colour is recognised.
As noted in the previous chapter, the ‘informed’ survey can result in restricted
coverage due to the reliance on local information, and the inspection of places with
high potential for sites. By using a handheld GPS to track the routes travelled, it was
noticed that the survey tended not to venture into fields further removed from villages
and roads, and the use of a motorbike could have exacerbated this bias towards
accessible areas. Partially, this could be rectified through monitoring progress in the
field and explorations on foot to move into the agricultural or forested areas away
from paths. However, time restrictions and difficulties with access/visibility were
further issues to overcome.
Low surface visibility was common, and the fieldwork was particularly affected by
unpredictable seasonal weather over the three years of the project. In the last two field
seasons, variable rainfall increased the amount of natural vegetation and delayed the
crop growing seasons and harvests to the extent that sites easily visible during
January 2012 were covered by vegetation in April 2014. Both the Level 3 and Level 4
surveys were affected by variables in the field, including visibility, access to private
land and modern activities, but under these conditions, ‘informed’ survey appears to
have been more successful. This is due to the removal of the random element which
can lead researchers directly to areas of high find concentration. There is a clear bias
involved in this method of data collection and so this must be taken into account
when discussing patterns of findings.
192
Chapter 6. Overall results
The effectiveness of both Level 3 and Level 4 survey methods varied according to the
general environment across the research area. On average, the ‘informed’ survey was
easier to implement across the entire survey area, as it is flexible and can be targeted
to maximise returns. The cultivated land south of Rāmṭek was generally easy to
traverse with average visibility, but occasionally there were access issues and areas
heavily disturbed by digging. Within Rāmṭek town, archaeological remains are
potentially obscured by modern infrastructure and the surrounding area consists of
wasteland cleared for building projects. The northern part of the survey area is hilly
and less uniformly cultivated, and thus less manageable in terms of visibility and
access. Areas of high vegetation and forest reduced the effectiveness of ‘informed’
survey and some areas were impassable due to activities such as brick manufacture.
With increased resources and time this area may have been explored more profitably,
but under the constraints of the project the northern hills proved particularly difficult
to traverse. The lack of sites and finds in the north may reflect a true pattern of
evidence, or may be a product of the landscape conditions and methodology. It should
be recognised that certain categories of site, particularly those subject to low visibility
due to their ephemeral nature, could be over-looked.
The systematic field-walking itself was fairly easy to organise limited manpower and
resources, due to the use of random quadrats within transects. The sample areas were
easy to manage and navigate due to their size and shape, and the spacing of the field-
walking could be monitored closely with the GPS. Systematic transect survey tend to
more accurately characterise archaeological landscapes and identify a wide range of
remains than unsystematic reconnaissance survey (Blinkhorn, et al., 2010:8).
However, given the limited time available, the systematic (Level 4) survey did not
result in a higher quality and quantity of findings and was not widespread enough to
facilitate a comprehensive modelling of the local archaeology. The key function of
the limited systematic (Level 4) survey was to act as a control to the results of the
informed (Level 3) survey. Therefore, despite low levels of findings, its inclusion is
still considered important for promoting greater confidence in the findings from the
‘informed’ survey and it was successful in demonstrating patterns of artefact
absences. Although few large sites were identified through quadrat survey, several
areas of higher density ceramic scatter were noted. Ceramic material did not appear to
be widespread over the areas surveyed, but background scatter was visible. The
193
Chapter 6. Overall results
fragmentary and abraded nature of the surface sherds and the difficulty in identifying
background scatter highlights that it is possible this was overlooked in less intensive
explorations. Due to the condition of the remains, the material collected during
quadrat walking is of little use in a temporal study of the landscape. Taking into
account issues of visibility and the small sample size, the returns of the systematic
survey still appear to be low. The general lack of findings supports the indication that
there is little surface evidence outside of discrete mound areas, which are well known
to villagers. The results of the short period of systematic survey indicate that in order
to achieve data comparable to that of the ‘informed’ survey, complete coverage of the
landscape may be required, which was well beyond the scope of the project. It is
likely that with higher investment a fully systematic survey would produce data of
higher resolution, but the practicality of the ‘informed’ survey in this context is clear.
A total of 444 sites were recorded during the Rāmṭek survey; this includes 47 ‘sites’
already known in publications and several outside of the survey area which were
recorded for posterity, such as the location of the Mandhal excavations (recorded as
Level 1 survey findings). The majority of the sites were identified during the Level 3
‘informed’ survey or the subsequent speculative explorations, while 12 sites were
located during the Level 4 survey of quadrat walking.
Table 6-1. The number of sites recorded during the ‘Informed’ survey versus the systematic quadrat
walking, with mention of an additional 47 sites which were previously known to researchers.
The recorded sites may be classified into broad feature groups such as religious sites,
habitation sites, environmental features, scatters and isolated features. ‘Isolated’
sculpture and architectural pieces have been categorised separately from general
194
Chapter 6. Overall results
religious sites as the original find-spots of these remains are not known. Rather, it
often seems likely that they have been re-installed in villages or moved given their
lack of associated remains. On the other hand, religious sites primarily constitute
structural temples and cabūtrās, which are more permanent manifestations of the
religious significance of an area. Environmental features encompass the structural
modifications of the landscape for control of water such as wells and tanks. While
related to habitation activity, these are often found to be isolated features throughout
agricultural areas and so have been categorised independently. A separate site
category has been formed for areas featuring a temple and tank in clear association
with one another, as they are viewed as one complex. Structural remains refer to
discrete brick or stone constructions with a secular or unclear function. A lack of
associated ceramics or remains of habitation debris results in this category being
separated from ‘habitation’ sites for ease of analysis. Habitation sites refer to
extensive remains of settled activity and were always recorded when identified. This
includes clear mounds, disturbed or partially destroyed mounds and abandoned
modern villages. Low-density habitation remains refers to ceramic or brick scatters
with no visible mound and smaller ceramic assemblages or low-density remains in
cuts. Low density habitations remains were distinguished from mounds as while both
relate to habitation activity, they have formed in different ways or have been subject
to significantly different post-depositional factors. The ‘locally known area’ category
relates to locations repeatedly referred to by local inhabitants, but which did not
display any substantial evidence of past remains.
Table 6-2. The broad site categories, to which the 444 recorded sites were assigned.
195
Chapter 6. Overall results
The ‘informed’ village-to-village survey resulted in the discovery of 385 sites related
to occupation of the rural landscape. Temples and shrines (smaller religious
structures) were the most common feature types and will be discussed in chapter
eight, but the majority of these religious sites appear to be modern. Of the sites
containing habitation remains, the most common feature was mounded sites,
associated with high density ceramic remains and large quantities of brick.
Environmental features were frequently encountered, usually in connection with
villages and their neighbouring agricultural land. Ceramic sherds were the most
common category of find during survey, occurring in transient scatters and artificial
cuts as well as in high concentrations at mounds. Brick debris was the second most
common type of archaeological remains encountered.
196
Chapter 6. Overall results
Table 6-3. The number of each feature type identified over the course of the ‘informed’ survey.
197
Chapter 6. Overall results
Table 6-4. The relevant information for each settlement mound site identified during the ‘informed’
survey.
79 4 Destroyed mound at Hamlapuri - the fields are Red and II-III (Up
cut through the original ground surface Mica1 to V on
wares surface)
122 2 Southern outskirts of Rāmṭek - the area has Primarily II-VI
been cut into the original ground surface. Red
Pottery is abundant in the section indicating Burnished
that under the modern occupation there may be ware
further evidence.
154 4 Mounded settlement on the north side of Range of II-III (Up
Rāmṭek (Kawadak). Brick and ceramic Mica wares to V on
remains continue across the fields north of this surface)
area
197 1 Ploughed mound on the west side of Hivra- on Red and I-II (Up to
higher ground than the surrounding village to Mica2 V on
the east wares surface)
202 <1 Village of Hivri – mounded with ceramic and Mica2 to II-V?
brick remains grey/buff
wares
210 1 Mounded fields with brick remains and Mica2 to II-VI?
ceramic scatter. grey/buff
wares
198
Chapter 6. Overall results
236 1 Mound with a section cut away along the Primarily II-V
western side – previous extent unknown Mica2
237 1 A second mound to the west of site 236:: Red and II-IV
Ceramic and brick scatter Mica wares
238 1 Third mound near to site 236: Brick remains Red and II-IV
are scattered across the top of the mound Mica2
wares
242 1 Mound at Beldongri: The village is built Red wares II/III
around the northern and western edge.
Extensive brick remains
248 1 Area of high ground across the river from Range of II-VI?
Banpuri: Extensive brick evidence wares
262 1 Mounded fields with high density brick scatter Red and I-IV?
Mica wares
265 1 Destroyed mound (as at Hamlapuri): pottery Primarily I-IV
and large brick remains are visible on the red and
ground and in sections mica1
wares
276 1 Mounded fields with brick fragments and Mica wares II/III-V?
medium density ceramic scatter to
grey/buff
wares
289 1 Mounded fields with surface ceramics and Primarily II-IV (V
fragmentary bricks. Mica2 on
surface)
318 2 Mounded ploughed field with fragmentary Primarily II-V
brick and medium density pottery scatter Mica2
199
Chapter 6. Overall results
Figure 6-4. The identified mound sites in the survey area featuring brick and ceramic remains.
Aside from pottery and brick remains, the artefacts encountered included tile
fragments, architectural pieces, sculpture and a low number of portable small finds. A
large amount of sculpture was identified and notable pieces will be discussed in
chapter eight. The small finds included fragmentary figurines, agricultural artefacts
such as grinders, and worked stone tools (see appendix three and discussion in
chapter eight). A number of small, rounded ceramic balls and smoothed pottery discs
were identified. Their purpose is inconclusive however similar rounded pottery sherds
have been found at sites such as Nevasa (periods II to VI) and have been interpreted
as play pieces or gaming counters (Sankalia, et al., 1960:384). Spindle whorls formed
200
Chapter 6. Overall results
from ceramics with a hole through the centre were also identified and could relate to
evidence of weaving (Gupta, 1992:134). No other categories of small finds, such as
beads or metal slag were identified. The absence of material related to production
activities, particularly slag, is striking and it is unclear whether this is primarily
related to issues of visibility or an absence in the archaeological record. Metal slag
was found in only low quantities at Kaundinyapura during survey and it was
suggested that iron was not produced in certain areas of the site (Smith, 2001:87). It
could be conjectured that there were centralised areas of specialised production which
then distributed material to the local area (Gogte, 1982). However, given the ubiquity
of metal working slag in excavation reports, the total absence here does suggest that
visibility or inexperience in detection has led to this material not being recognised.
The locally known material ‘Murrum’ was found across the survey area and was
initially mistaken for metal slag given its distinct similarities. It is possible this has
obscured the archaeological material.
During the systematic survey, a random selection of quadrats was walked across the
three transects contained within the study area. Transects one and three largely
covered stretches of agricultural or waste land, while transect two had reduced
visibility as it fell partially over Rāmṭek town and the surrounding area, much of
which is cleared to undergo development. There was a range of environmental
settings in the quadrats surveyed and even when quadrats fell across solely
agricultural areas, conditions ranged from being ploughed, fallow or recently planted.
On the whole, visibility within each quadrat was found to have low correlation with
sherd density; quadrats with no obvious sherds for collection often displayed high
visibility, but equally the quadrats containing the highest number of sherds in each
transect were of average or high visibility. In the agricultural areas, significantly low
201
Chapter 6. Overall results
visibility due to water from canals and the planting of new crops could be theorised to
reduce success in collections. For example, transect three did not yield many ceramic
sherds despite its location in the south of the survey area where a greater number of
sites were recorded during ‘informed’ survey. This area experiences a higher rate of
double cropping and so reduced visibility due to ongoing cultivation may be
responsible. Alternatively, the different distribution of remains may relate to diverse
past activities across the various environments of the survey area.
Figure 6-5. Map giving the visibility across sampled quadrats in the three transects surveyed during
systematic field-walking.
Twelve religious sites or environmental features were discovered during the transect
survey. The small rural shrines, temples and wells which were discovered in the fields
suggest that it is likely further examples of this type of site could be found in
agricultural areas which were not explored in depth. This highlights the bias of the
village-to-village informed survey and its unsystematic coverage as there is the
potential for further sites to exist. However, no mounded areas indicative of ancient
habitation were discovered in proximity to these structures.
202
Chapter 6. Overall results
Table 6-5. The types of site identified during systematic quadrat walking.
The average yield of ceramic sherds per quadrat was 4.36 per 100m2 in transect one,
0.4 sherds per 100m2 in transect two and 1.21 per 100m2 in transect three. The
majority of quadrats yielded no sherds, although a lack of visible finds does not
necessarily equate to an absence of artefacts in the past and low density surface
scatter is found across other quadrats (Gallant, 1986:406). The average yield of
sherds per quadrat is higher for transect one due to the presence of medium density
ceramic scatters in occasional quadrats. These hotspots were typically related to
habitation sites previously identified during the unsystematic investigations, but
systematic survey enabled the collection of ceramics from a wider radius. The
quadrats with the highest quantity of sherds in transect one were located next to site
154 (Kawadak), which was identified during speculative survey around Rāmṭek. The
previously recorded ceramic scatter was found to extend further into the ploughed
fields and settlement area across the road from the main mound. Similarly, the
quadrats around site 169 revealed medium density surface scatter in connection to the
site, which had not been previously recorded. This demonstrates that more complete
information could be achieved from intensive field-walking. Additionally, systematic
walking of quadrats revealed low density scatters of ceramic sherds where no sites
were previously recorded. There were no any other noticeable features, such as a
raised ground profile indicative of a settlement mound, and so such areas may be
classed as ceramic scatters perhaps suggestive of more transient human activity.
Significantly, this implies that less obtrusive sites, or those not known by local
203
Chapter 6. Overall results
inhabitants, could be missed during ‘Informed’ survey and so the systematic survey
has fulfilled its purpose in this respect. It should be emphasised that care should be
taken when suggesting a lack of habitation in certain areas as other forms of human
occupation and rural economic activity could have been present in the past but are
simply hard to determine in the modern landscape.
Table 6-6. The count of ceramic sherds per quadrat for transect one, and the number of quadrats in
which that quantity of sherds was collected.
Transect 1
Count of ceramic Count of quadrats
sherds per quadrat
0 50
1 3
2 3
3 2
4 5
5 1
7 1
8 2
9 1
10 1
20 1
23 1
30 1
41 1
54 1
77 1
Table 6-7. The count of ceramic sherds per quadrat for transect two, and the number of quadrats in
which that quantity of sherds was collected.
Transect 2
Count of ceramic Count of quadrats
sherds per quadrat
0 56
1 1
2 6
3 3
4 1
5 2
6 1
7 2
204
Chapter 6. Overall results
11 1
12 2
Table 6-8. The count of ceramic sherds per quadrat for transect three, and the number of quadrats in
which that quantity of sherds was collected.
Transect 3
Count of ceramic Count of quadrats
sherds per quadrat
0 56
1 1
2 6
3 3
4 1
5 2
6 1
7 2
11 1
12 2
Figure 6-6. Map displaying the sherd density per sampled quadrat with clear low levels of findings
across the majority of quadrats.
205
Chapter 6. Overall results
Theoretically, the sample from the systematic ‘Level 4’ survey could be incorporated
into statistical analyses and be used to extrapolate data regarding the rest of the
survey area. However, within the scope of this study, the results of the systematic
approach were intended to assess the findings of the Level 3 ‘Informed survey’.
Therefore, it is again reiterated that the Level 4 survey did suggest a background level
of human activity across the survey area, perhaps related to low-density rural
agriculture. The survey also failed to identify any unknown large sites indicating past
settled habitation, which strengthens the argument that the Level 3 survey is capable
of recognising such remains. What has been achieved with the systematic survey is
sufficient to enable the prediction that a similar frequency of sherds would be found
across the area, if this methodology were repeated on a wider scale (Barker, 1996).
However, all such findings must only remain hypotheses that need to be tested and
confirmed with future work.
Ongoing degradation and destruction of sites was noted during the course of the
Rāmṭek survey, and reflects observations by multiple scholars that the archaeological
potential of rural India should be a priority for research. Existing damage to sites was
recorded at the start of the survey, and during the three fieldwork seasons there was
progressive disturbance from changing land-use. As discussed in chapter three,
damage to the archaeological remains around Mansar was conspicuous during
Wellsted’s explorations in the 1930s. Wellsted’s reports of habitation remains around
Mansar Hill could not be confirmed as it appears the archaeology already severely
damaged by 1934 has now been completely obliterated. Bakker et al.’s 2006
reconnaissance detected that the site of Hamlapuri (site 79) was the remains of a
destroyed mound; these observations were confirmed in this survey and another site
was identified in the same condition (site 265, Bada). All of the habitation mounds
identified during survey have suffered damage from the encroachment of agriculture
and digging activities. The surfaces of low mounds have been ploughed and
cultivated, and others have been damaged more extensively. At the large site of
Dudhala (site 236), the mound has been cut away along one side, and at Beldongri
(site 242) curious villagers have cut away the corner of the brick mound to reveal the
206
Chapter 6. Overall results
interior. The artificial cuts in both of these examples allowed access to sections,
which can reveal more information about artefact associations and the nature of
buried remains, but the damage also alters the extent of sites and may affect the
remains of different periods unequally.
The habitation site of Naharwani (site 364) presents one of the starkest examples of
the swift destruction at archaeological sites. A complete settlement mound is visible
on 2002 GoogleEarth imagery and 2003 Orbview 3 imagery, but has now been
completely destroyed. On 2009 GoogleEarth imagery it is clear that a small area on
the southwest corner of the mound had experienced limited exploratory digging, but
by 2012 it had been entirely excavated during extraction of Murrum, leaving only
scattered brick and ceramic remains in the spoil.
Figure 6-7. The progression of modern damage at the site (364) of Naharwani from top left as shown
in (a) 2002 GoogleEarth imagery, (b) 2009 GoogleEarth imagery, (c) 2014 GoogleEarth imagery (d) A
photograph of the site during survey (2012).
Even more drastically, an area of cultivated land (site 73) south of the Bhonsle fort in
Nagardhan, which displayed limited ceramic surface scatter in the first field season,
was found to be dug away to a depth of around three metres by the end of season two,
207
Chapter 6. Overall results
revealing copious archaeological remains. With the area’s proximity to the Medieval
fort, and given the known Early Historic potential of Nagardhan, it is unfortunate that
the work was unmonitored. Official development and renovation work at the fort
itself has caused damage to the archaeological remains as digging has cut through
sequences of brick and ceramics, which were discarded in the spoil. A large section of
the western wall of the fort collapsed in 2014, revealing archaeological remains of
seemingly mixed antiquity inside the wall.
Figure 6-8. From left: Modern digging activities south of the fort. Site 73 is shown as (a) cultivated
land on 2012 GoogleEarth imagery and (b) following excavation on 2014 imagery.
Figure 6-9. Photographs show (a) the extent of the disturbance south of the fort at site 73 during the
second field season in 2013, and (b) the collapsed section of wall at the Bhonsle Fort.
208
Chapter 6. Overall results
Large areas of the Rāmṭek landscape are clearly damaged by past and current
quarrying, small-scale mining activities and brick manufacture. It is unknown
whether some were eradicated before this survey commenced. Furthermore, the past
extent of damaged sites, such as the mound at Hamlapuri, is unknown. The situation
witnessed in the Rāmṭek landscape highlights a need for even basic reconnaissance
and recording of sites, such as that attempted in this project. It is important to note
that disturbance of sites may affect the resulting patterns of remains identified during
survey. While destruction of sites can reveal buried ceramics and remains, it is also
unknown how many sites have been destroyed and the land reclaimed for other uses.
Unknown visibility or survival of sites must be taken into account in later discussion
of the patterns evident in the archaeological remains, as it may be that the survey has
not achieved a representative picture of the past landscape.
Several types of archaeological remains identified during the Rāmṭek survey have
relevance to a consideration of chronology and the temporal patterning of sites in the
landscape. The phases of interest in this study are structured by the aims of the
survey, the nature of remains and the published comparison material. The
chronological parameters of this research reflect the questions we have about the
landscape and its development; the fieldwork aimed to elucidate the longue durée
development of the landscape, and questioned whether pre- and post-Vākāṭaka
activity could be identified in order to ascertain the extent of the Vākāṭaka impact on
the landscape. The chronological framework remains broad and the nature of the
archaeological remains did not enable assignment to very precise time periods.
Therefore, the chronology follows the dating set out in publications of excavated sites
and material from the region. Many of the broad phases referred to are quite loosely
defined and not always easily linked to recognised historical periods.
209
Chapter 6. Overall results
material and established typologies, a lack of standardised terminology and the use of
loose and imprecise descriptions in publications which make comparisons difficult.
Architecture, carved architectural fragments and sculpture have the potential to be
dated fairly accurately if they are well preserved and display recognisable stylistic
indicators. However, early architectural remains are rare, and much of the sculpture
was difficult to identify. Past habitation sites in the survey area could only be dated
by reference to ceramic and brick remains, as there was usually no association with
sculpture or architecture. Brick remains are impossible to date accurately in the
absence of scientific techniques, but may be dated in association with other artefacts.
Past research has used typologies of ancient brick dimensions to suggest periods of
manufacture, using the key principle that brick size diminished over time and certain
brick dimensions were common to particular periods. Early Historic bricks are
considered to have been large, and published brick dimensions vary between roughly
30 and 50cm in length by approximately 20-30 cm in width, and 7-10 cm in depth
with a range of variation (Mishra, 1997). Large-sized bricks with similar ‘Early
Historic’ dimensions were common around Rāmṭek, and more complete pieces were
measured for comparison to those in archaeological reports and known structures in
the area.
Various studies have demonstrated that brick dimensions are a poor indicator of
chronology brick size, as the amount of regional variation means that the trend of
diminishing size does not hold up to scrutiny, but this method has been commonly
employed across the subcontinent in the absence of other dating methods (Verardi,
2002). Bricks may be used cautiously as broad indicators of chronology when
associated with ceramics, and can provide support for Early Historic occupation. By
considering brick dimensions it may be possible to demonstrate that there are ancient
bricks across the survey region, which are consistently similar in size to published
‘Gupta-type’ bricks. Bricks and ceramics are also associated with tiles in the survey
area, which are found with Early Historic materials in excavated contexts. Tiles have
been used for dating in other projects (Suvrathan, 2013), but could not be used here as
the pieces were too fragmentary.
Despite there being little scope to refine the working project chronology, it does
nonetheless provide enough of a framework to engage with the periods of interest in
this research. In South Asian archaeology, terms such as ‘Early Historic’, ‘Early
210
Chapter 6. Overall results
Medieval’ and ‘Medieval’ are invariably relied upon, and this project remains in
keeping with past research by referring to these broad phases. In this study, the Early
Historic and Early Medieval phases have been related to the ruling dynasties, based
on the research aim of identifying differences in activity as a consequence of periods
of rule. Chronological discussions in India often rely on dynastic phasing despite the
risk of simplifying the history of development by implying there were distinct
cultural breaks. However, historical divisions can still be useful for structuring broad
developmental periods if the process of change is taken into account. The
chronological framework presented here will be referred to in the later chapters to
structure the dating of the ceramics, sculpture and architecture.
Table 6-9. The working project chronology, which gives both the broad period and the project phasing
from I – VI.
211
Chapter 6. Overall results
The first phase (I) in the project chronology has been termed ‘Prehistoric’ and
‘Protohistoric’, and relates to the general period prior to the beginning of the Early
Historic. Phase I activity was expected to be most visible in finds of stone tools,
characteristic ceramics or megalithic re-working of the landscape. Although
megalithic structures would be expected for Phase I, as shown previously they have
been found to extend into the Sātavāhana period in Vidarbha and could therefore
relate to Phase II in the project chronology. An even later date for megalithic tombs,
up to the 4th century AD, has been reported in north-western Tamil Nadu based on
ceramic, numismatic and C-14 dating (Ramachandran, 1980). Although transitional
phases between the Megalithic and Early Historic periods have been identified at
some sites in Maharashtra such as Khairwada, often terminal dates for the Megalithic
remains are not conclusive and the relationship with the succeeding Early Historic
period is still subject to investigation (Mitra, 1983c:51). According to the majority of
regional publications, they appear to have gone out of use by Phase III at the latest.
As it was, while ceramics and stone tools were encountered during survey and can be
assigned to Phase I activity, there was no evidence of structural remains.
Phases II and III can be broadly referred to as the ‘Early Historic’ period. This period
has been divided into two phases based on the known development of the Rāmṭek
landscape, and the importance of separate phases in relation to the project research
questions. Certain types of remains were expected to be distinct between Phases II
and III, for example the earlier pre-Vākāṭaka phase (II) may have left traces of
Buddhist remains related to the corresponding socio-political and religious context,
while Brahmanical evidence would have emerged in the later Vākāṭaka phase (III).
For other categories of evidence, such as Early Historic ceramics, it was not certain
that the different phases within the overarching period could be determined, as they
are not well established.
Phase IV refers to the Early Medieval period and has been split into IV A and IV B;
Phase IV A relates to the unknown intermediate period in the Rāmṭek landscape prior
to the establishment of Yādava rule around the 11th century AD. Phase IV B refers to
the second phase of the Early Medieval, which witnessed more obvious landscape
development under the Yādava kings. Early Medieval ceramics have not been
adequately defined and so sculpture and architecture becomes particularly important
for identification of this period. Structures originating from these phases may be
212
Chapter 6. Overall results
distinguishable due to their different architectural influences, but the same resolution
was not expected in other remains. As it was not certain that precise developments
could be determined, or that a nuanced picture of Early Medieval settlement could be
established, the two periods A and B were kept under the umbrella of Phase IV.
Phase V relates to the Later Medieval to Mārāṭha period, after the collapse of the
Yādava dynasty and up to the end of the 18th century. Medieval ceramics are better
established that those of the Early Medieval, and the material culture is distinct from
the Early Historic. Ceramics would be expected to be key indicators of settlement.
The earlier part of Phase V correlates with the Vijayanagara period, of which there is
little known evidence in the Rāmṭek landscape. Mārāṭha influence on architectural
remains is known around the 18th century and it was expected that this phase would
be more visible across the survey area. Following the 18th century, from the period of
British rule, developments in the landscape are considered to be Modern, and are
simply designated as such. As the chronological phases may be represented by
different categories of evidence – earlier periods are more likely to be represented by
ceramics and brick remains compared to the Medieval periods which would seem to
have greater quantities of architectural remains - there may be some difficulties in
directly comparing phases of development and evaluating the extent of activity.
6.6. Conclusion
The introduction to the overall fieldwork results provided in this chapter leads into an
in-depth consideration of the different categories of evidence over the next two
chapters. The survey was successful in recording a range of sites relating to multi-
period settlement and religious activity, and has increased our understanding of
Rāmṭek’s rural landscape through the assessment of previously unrecorded sites. Both
the ‘informed’ survey and the quadrat-walking encountered methodological
difficulties but achieved their intended objectives. Given the project limitations, the
‘informed’ survey was found to be easier to implement, and more advantageous than
the trial systematic survey in terms of the amount of information that it yielded. The
‘informed’ survey enabled the recording of hundreds of sites in a strict time frame
with limited resources. The systematic survey did not result in a high number of new
213
Chapter 6. Overall results
sites or ceramic remains, but it has demonstrated that low level ceramic scatter can be
identified. The results of the systematic survey indicate that transient and small-scale
sites may in some cases have been overlooked by the ‘informed’ survey, alongside
ubiquitous sites such as modern shrines and water resources. This must be noted as
transient remains could refer to categories of human activity significant to the rural
landscape and engagement with the Eastern Vākāṭakas.
The fieldwork has revealed the effect of modern activity on archaeological sites and
supports the argument for further landscape surveys in India. Despite having
undergone ASI village-to-village reconnaissance, the Rāmṭek area contains a wealth
of unreported sites, including a number of habitation mounds, which can impart
important information about occupation of the landscape. Landscape survey is
particularly important for this region, given that the nature of settlement and activity
during the Early Historic and Early Medieval periods is barely understood. The
working project chronology presented in this chapter will provide the framework for
discussions of the archaeological remains over the following two sections of the
thesis. The next chapter will focus on ceramic classification, and the resulting relative
seriation, which forms a major part of discussions on the phasing of sites.
214
Chapter 7. Ceramic assemblage
7.1. Introduction
215
Chapter 7. Ceramic assemblage
The broad details of the ceramic assemblage will be presented below, and the
complete tables may be referred to in appendix four. 5,962 sherds were collected
across the 314 km² of the survey area, with a total of 5,943 available for study. Of this
complete assemblage, the majority were fragmentary body sherds, which were
assigned a ware but were undiagnostic for the shape typology.
Table 7-1. Type codes for the ceramic sherds in the Rāmṭek assemblage, and the count of each type.
In all, 74 wares were identified during intuitive sorting; a number of wares are
particularly common in the assemblage but only GW.BLS represents more than 10%
of the overall assemblage. A further 28 wares each constitute over 1% of the overall
assemblage, and the remaining 46 wares each make up less than 1%.
216
Chapter 7. Ceramic assemblage
Table 7-2. Outline of the ware types identified during ceramic analysis.
217
Chapter 7. Ceramic assemblage
218
Chapter 7. Ceramic assemblage
Several general statements can be made about the nature of the Rāmṭek assemblage
based on the ceramic wares. The sherds mostly indicate that ceramic vessels were
wheel thrown, which is in keeping with archaeological reports of Early Historic
material culture (Ray, 1954:552). The fabric quality was described using terms such
as ‘coarse’ and ‘medium’; ‘coarse’ describes less compact material with many
inclusions, while medium fabrics were more compact with fewer inclusions. All of
the ceramics loosely fell within these categories and appear to be fairly utilitarian,
with few appearing medium-fine outside of the varieties of Black and Red (B&R)
ware. A range of firing is demonstrated by core colour, particularly across the red and
micaceous wares, and it is common for sherds to display a black/grey core
representative of unoxidised firing. The closest comparison material from the survey
area – the ceramic assemblage from the 5th century monuments at Mansar - has not
been fully published. However some reports suggest that the majority of ceramics
found in the excavations could be assigned to medium fabric red ware, unoxidized
dull red wares of a similar fabric, coarse and gritty micaceous red wares and fine
polished red wares (Manjhi, et al., 2000:56). Ceramics of the Gupta-Vākāṭaka period
are thought to have had a long duration with a range of minor variations. It has been
traditionally thought that by the Gupta-Vākāṭaka period, Sunga grey wares and
typical ‘Mauryan’ types disappeared from assemblages and were replaced by a range
of red wares (Ray, 1954).
The most common surface treatment for ceramics in the Rāmṭek assemblage was
smoothing and flattening of the surface, and the application of a slip, with burnishing
being common. Burnishing involves rubbing the ceramic surface with a smooth, hard
object to give a shiny finish, but results in a more irregular shine than polishing (Rice,
1987:138; Sinopoli, 1991:25-26). The Rāmṭek red and micaceous wares usually
display various shades of red slip, which is similar to those reported from
contemporary sites, which are also noted to have a polished appearance (Ray, 1954).
The addition of mica to the clay and surface treatments has often been stated to give a
lustrous finish, and thus is typically equated to special vessel types. The Grey/Buff
wares in the Rāmṭek assemblage principally display black slips.
219
Chapter 7. Ceramic assemblage
Pottery with scalloped or thumb impressed applique designs similar to those in the
Rāmṭek assemblage have been found unstratified at Brahmapuri/Kolhapur (T.95)
(Sankalia & Dikshit, 1952). At Pauni, applique ‘roundels’ were described on storage
jars in Period IV (Nath, 1998:46) and at Nevasa decorated sherds of thick storage jars
also featured applique bands with fingertip/rope designs (Sankalia, et al., 1960:273).
The applique flanges and pinched designs familiar to the black slipped wares is
similar to the descriptions of raised bands with fingertip impressions and beaded rims
on burnished black storage wares at Dwarka, Period III (7th to 8th century AD), and
220
Chapter 7. Ceramic assemblage
the ledged carination and applique wavy lines on burnished black ware in period IV
(10th century AD to the Medieval period) (Ansari & Mate, 1966:79).
Table 7-3. The major categories of ceramic decoration encountered in the Rāmṭek assemblage; further
illustrated examples are available in appendix seven.
221
Chapter 7. Ceramic assemblage
Of the 2,549 rim sherds, 1,750 (69%) were successfully categorised into 128 rim
types, with these types represented by two or more examples within the diagnostic
rims. A further 75 rim sherds were identified as unique shapes, and 740 rims (29%)
were recorded as ‘eroded’, and could not be reliably attributed a type. Published
parallels tend to be easier to find for type than ware given the more detailed
descriptions and images, but parallels cannot always be found. Of the 203 total rim
types (including uniques), 94 were found to resemble parallel types closely in
published assemblages (see appendix six). By identifying broad time frames from
published parallels and assessing the occurrence of particular vessel types within
wares, it may be possible to use vessel shapes as chronological indicators and gain a
clearer understanding of the date and life span of some of the wares. This may be
particularly useful when considering the more common and nondescript fabrics, such
as in the red wares, which appear to have a long period of use (Miller, 1985). The
early excavations at Mansar indicated that common types could be identified in the
red and micaceous wares, including vases with flared and out-turned rims, vessels
with globular bodies, carinated handis with short necks and out-curved rims and
sprinklers (Manjhi, et al., 2000:56). If the lengthy Early Historic time frame remained
difficult to sub-divide and the Vākāṭaka ceramics were difficult to define, it was
hoped that types may be identified as uniquely pre- or post-Vākāṭaka in order to
construct a relative chronology.
Several factors support the absence of imported wares to suggest that the Rāmṭek
ceramics are locally produced; a large degree of minor variation is observed in fabric
and vessel shape and there is a high quantity of wares and types in a small
assemblage. The inclusion of mica fragments in the majority of ceramic sherds may
further demonstrate local production as mica is prolific in the area and easily
incorporated into the parent clay. The large range of fabrics and types at individual
222
Chapter 7. Ceramic assemblage
sites, and an occasional association between particular wares/types and certain sites
also supports suggestions of localised production and use. The local production of the
most common wares is supported by the types of vessels present (such as water jars,
and cooking pots), which would have been expensive to transport due to their bulk,
weight and low value. Despite this, many of the vessel shapes are comparable to
published parallels; similar to the ceramics from Smith’s study at Kaundinyapura, the
Rāmṭek assemblage demonstrates the existence of a “wider shared material culture”
at least across the Vidarbha region (Smith, 2001b:75). Numerous parallels were also
found within published Gujarati sequences which may have further implications in
terms of ceramic production and shared material culture.
The proportions of different vessel types are often used in publications to indicate
types of activity and functional differences between and throughout sites. This
assemblage is dominated by functional vessel types intended for cooking and storage,
usually with a general mix of both across sites. This suggests a primarily residential
or domestic situation across the survey area. An in-depth study of functional
differences based on vessel or ware type is not undertaken here; questions of varying
function or status across sites may be addressed following future research when the
ceramics of this region and period are more solidly established and understood.
Currently, the focus of this ceramic study is to attempt to formulate a relative
chronology, in order to elucidate phases of activity at sites. However, it may be
briefly noted that all ceramics appear to be locally produced and utilised. There is an
absence of fine or traded wares and vessel types, which may usually be connected to
implications of trade or social stratification. It is worth noting that the small
proportion of large storage jar rims, occurring in particular storage jar wares, do show
some correlation with certain sites and this pattern gives the clearest indication of
some functional differences. It is noted that the ‘palace’ site at Mansar (site 7) has a
relatively high percentage of storage vessel sherds, which may be related to its elite
role and the receipt of produce from other sites. Similarly, areas around Nagardhan
have higher percentages of storage jars; this is the largest site in the area and may also
be connected to a chain of production in the landscape (to be discussed in chapter
ten). Site 78, a small mound near to Nagardhan with copious brick remains but
relatively few ceramics, also displays a large proportion of storage vessel fragments.
This may support suggestions that the mound is not the result of a build-up of
223
Chapter 7. Ceramic assemblage
habitation activity, but rather may represent a structure or other category of site,
which again was in receipt of surplus from the surrounding population. Mention may
also be made of the correlation between certain vessel types and wares, either due to a
functional or preferential difference. This will be discussed with reference to each
ware type below.
Given the large number of wares overall, ware families were introduced to create
broad groups of wares and add another level of analysis. Thirteen ware families were
outlined with similar characteristics and occurrences. This section will introduce
briefly the characteristics of the ware families and their most common vessel types,
and consider them against published ceramics. In the case of the black and red wares
and the grey/buff wares, the family groupings were fairly cohesive and
straightforward, while the red and micaceous wares were split across different ware
families. A great deal of variation is observed across the red and micaceous wares in
particular and clear divides between wares were hard to define. Moreover, the red and
micaceous wares are largely observed to occur together in varying quantities.
224
Chapter 7. Ceramic assemblage
Table 7-5. The broad ware family groupings in the Rāmṭek assemblage, with their corresponding
specific ware types.
Grey/Buff 1. BAG.BLS
2. BUG
3. BUG.BRSL
4. GW
5. GW.BLS
6. TIG
Grey/Orange 1. GOW1
2. GOW1.RS
3. GOW2
4. GOW2.RS
5. GOWM
Gritty 1. GROMIT1
2. GROMIT1.RS
3. GROMIT2
4. GROMIT2.RS
5. GROMIT3
6. GROMIT3.RS
7. GROMIT4
8. GRW1
9. GRW1.RS
Mica1 1. ORM1
2. ORM2
3. ORM3
4. ORM4
5. ORM4.RS
6. ORM5
7. ORM5.RS
8. ORM6
9. ROM
Mica2 1. BOM1
2. BOM1.RS
3. BOM2
4. BOM2.RS
5. BOM3
6. BOM3.RS
7. BOM4
8. BOM5
225
Chapter 7. Ceramic assemblage
9. BOM5.RS
10. BOM6
11. BOM6.RS
12. ORM1.RS
Red 1. RBC3
2. RBC3.RS
3. RBC4
4. RED2
5. RED2.RS
6. RED3
7. RED3.RS
8. RED4
9. RED4.RS
10. RED5
11. RED5.RS
12. TOR
Storage 1. HRBC
2. HRBC.RS
3. LAST
4. TORM
5. TOW
6. VORBL
7. VORBL.RS
8. VORM
Unknown 1. BLARB
2. NON-ID
3. RED.OF
Table 7-6. The count of sherds falling within each ware family in the Rāmṭek assemblage
226
Chapter 7. Ceramic assemblage
Due to their distinctiveness from the rest of the ceramic assemblage, the black and red
ware family consists of only BAR and F.BAR. This ware family is represented by 41
sherds, and was found at a limited number of sites. Compared to other ceramics
identified, Black and Red (B&R) ware is well-established in the published literature.
It has a characteristic colouring produced by an inverted firing technique, which
results in a reduced black interior and an oxidised red exterior, or an exterior which is
red around the vessel base and black towards the top. B&R ware is described as being
fine and well-levigated with a polished surface, but publications describe a range of
fabric. It appears that regional varieties have been identified, and Subbarao identified
a wide range of coarse and fine wares consistently occurring in shallow bowl forms
when classifying B&R wares from a number of sites (Subbarao, 1958). There is a
question over whether B&R ware can be considered part of a “single cultural
complex irrespective of time and space”, or whether the different ‘varieties’ are
simply the result of similar production methods rather than a chronological and
cultural connection (Srivastava, 1971:375).
B&R ware is usually deemed to be characteristic of the South Indian Megalithic, and
as such is often dated between the 10th and 3rd centuries BC. The ceramic is prolific
across the Vidarbha region alongside extensive Megalithic evidence, but sherds have
been recovered from Early Historic contexts, indicating that it continued in use in
some form throughout the early centuries AD (Subbarao, 1958:173). The occurrence
of B&R ware in the Early Historic period seems particularly common across central
India, including at sites in Maharashtra (Singh, 1969:70; Subbarao, 1958:176). In
Vidarbha, B&R ware was discovered in Period IV (500-150 BC) at Adam (Nath,
1992a:74), and at Paunar B&R ware was found in larger quantities in Period IIA (4th -
227
Chapter 7. Ceramic assemblage
3rd century BC) than in earlier periods (Deo & Dhavalikar, 1968:29). At Nevasa, the
limited quantity of B&R ware in Early Historic Period V (50 BC to 200 AD)
presented a narrower range of vessel shapes, so it has been suggested that it remained
in sparing use perhaps for specific purposes (Sankalia, et al., 1960:276). While B&R
ware evidently had a long history of use, published sequences seem to demonstrate
that it disappears from sites by the end of the 3rd century AD.
The Early Historic variety of B&R ware is often described as being comparatively
rare, degraded and coarser in fabric; the Early Historic B&R ware at Maheshwar and
Navdatoli (400-100 BC) is described as containing impurities with an increasingly
“porous, gritty and dull” core (Sankalia, et al., 1958:134). Similarly, while B&R ware
persisted at Timbarva up to about 300 AD, it is described as “crude” from Period II
(c.50 AD) (Mehta & Patel, 1967:17). Other studies of B&R ware have attempted to
connect increasing thickness to a later date, for example, thin B&R ware was
recorded in Period I at Paunar (c.1000-800 BC) followed by a thick variety in Period
II (up to 3rd century AD) (Deo & Dhavalikar, 1968:6). Ultimately, neither pattern has
been demonstrated in a robust fashion, nor has a means for determining between
Megalithic and Early Historic B&R ware been realised.
This characteristic ware name is used in this study, but it is important to note that the
Rāmṭek Black and Red wares (BAR) are not typical, given their coarseness; it cannot
be determined whether the fabric quality indicates an Early Historic date or is due to
local production, and there are two few examples to discern any potential pattern.
What can be stated is that this is the earliest known ceramic collected during survey,
and relates to the earliest phase of identifiable activity. Based on published parallels,
this ware family corresponds to Phase I or II in the chronological scheme outlined in
chapter 6, and can comfortably be described as pre-Vākāṭaka.
Table 7-7. Count of ceramic sherds within each ware type in the Black and Red ware family.
228
Chapter 7. Ceramic assemblage
All Black and Red ware sherds were rim pieces corresponding to two distinctive
shallow bowl forms, which are shallow bowls with either a featureless or slightly
extended round rim, and do not occur in any other ceramic fabric. The appearance of
the fabric and vessel shapes is mostly consistent with published descriptions. Both
types 1 and 2 have distinct parallels with shallow bowls/dishes reported from Nevasa
Period IV/V (150 BC to 200 AD), pre-Mauryan to Sātavāhana Kaundinyapura, Jorwe
and Malwa phases at Tuljapur Garhi and Paunar Phase I (1000 to 800 BC)
(Bopardikar, 1996; Deo & Dhavalikar, 1968; Sankalia, et al., 1960; Smith, 2000).
The types therefore lend support to a prehistoric to Early Historic time range for the
Black and Red wares.
Table 7-8. Table displaying the two vessel types present in the Black and Red ware family, and the
count of each within the specific Black and Red wares.
Count of Type
Type 1 Type 2 Total
BAR 5 3 8
F.BAR 5 1 6
Total 10 4 14
Figure 7-1. The two types of Black and Red ware rims encountered in the Rāmṭek assemblage.
229
Chapter 7. Ceramic assemblage
Micaceous wares make up 38.7% of the ceramic assemblage (2,299 of 5,943 sherds)
and contribute to the largest overall type of ceramic based on their defining
characteristic; the quantity of fabric mica. However, given the large amount of variety
across the micaceous wares they are split into four families; ‘Mica1’, ‘Mica2’, ‘Dark
Mica’ and ‘Other Mica’. Attempts to establish phasing between the numerous
micaceous wares is difficult as the ware associations at sites frequently indicates a
broad mix of co-occurrences, but the ware family divisions outlined below emerge
from visual characteristics and the most common co-associations.
Table 7-9. Count of ceramic sherds within each of the four different micaceous ware families.
Within these groups are both slipped and unslipped varieties of ceramics, and mica
occurs in the fabric and surface treatment; a coloured slip may contain mica powder
or small flakes, or the exterior and/or interior surface of vessels may be coated with a
thin silvery wash of powdered mica. The micaceous wash may be smoothed or
burnished to give a fine sparkly finish. Inclusion of mica may be related to aesthetic
or practical concerns, or indeed both. It has been suggested that mica is an effective
flux used to reduce the rate of firing (Rice, 1987:94; Shaw, 2007:108). In this study,
fabric mica defined the micaceous wares, but in publications the term ‘micaceous
ware’ can refer to those with little fabric mica but treated with micaceous slips.
Micaceous ceramics are sometimes described as being luxury wares; “special type”
vessels were described as being mica-dusted to give a “metallic” lustre at Ahichhatra
(from 350 AD) (Ghosh & Panigrahi, 1946:41), and at Devnimori, micaceous red
wares are theorised to be “costlier and less used” (Mehta & Chowdary, 1966:77).
Shaw describes the micaceous wares from Sanchi as “deluxe” with a ‘sparkly’
appearance and smooth texture (Shaw, 2007:108). Surface mica is likely to have been
230
Chapter 7. Ceramic assemblage
The useful qualities of mica coupled with its availability, may mean that mica is
present in ceramics across different periods, and its use is more related to
functionality than chronological distinctions. Micaceous wares have been identified
in different chronological contexts including the Megalithic and Early Historic, for
example, mica is recorded as a degraissant in early period ceramics at Kausambi
(Period III: 9th to 6th century BC) (Misra, 1969:206). Dikshit also notes that ceramics
with large amounts of mica used as a degraissant increased in number from the late
Sātavāhana period at Kaundinyapura (Dikshit, 1968:33). Some studies have
attempted to indicate that mica flakes were larger in Megalithic ceramics and
decrease in size in the Early Historic period, but the ware descriptions are vague and
there is no clear justification for this observation. A study of Bronze Age and Early
Historic micaceous red wares in Gujarat refutes this suggestion, as the historic
micaceous wares contain huge quantities of large mica flakes to act as a temper
(Herman & Krishnan, 1994:227). Micaceous wares in Early Historic contexts at
Kaundinyapura resembled Iron Age examples due to their large mica flakes, fragile
fabric, and typical vessel shapes, but were more abundant, which rather suggests their
increased popularity over this later period (Shete, 2009:75-76).
Although published micaceous wares are diverse and invariably loosely defined, there
are similarities between the micaceous fabrics and types described in Early Historic
contexts in this region and those in this assemblage. Both Hamlapuri and Mandhal are
described as ‘single culture’ Vākāṭaka sites with ceramic assemblages dominated by
micaceous red and black wares (Bhaisare, 2010-2011:104). A large number of
micaceous wares were assigned to the Early Historic period in Smith’s survey of
Kaundinyapura, and described as being of coarse, pale red clay with up to 40% fabric
mica. The presence of micaceous wares was used to suggest regional trade and a
preference for this aesthetic, as mica is not local to Kaundinyapura and the closest
231
Chapter 7. Ceramic assemblage
source is around 70km to the east, towards the Nagpur district (Smith, 2002a:145). At
Pauni, coarse micaceous red wares with small mica flakes as a ‘degraissant’, were
most common in the later Sātavāhana period (Deo & Joshi, 1972:65). Similarly at
Paunar, coarse red micaceous wares occur in larger quantities from Periods IIA and
IIB (4th/3rd century BC to 2nd/3rd century AD) (Deo & Dhavalikar, 1968:29). Further
afield, micaceous wares were found from Period II (c.50 AD onwards) at Shamalaji
(Mehta & Patel, 1967:18), and Period VI (200 to 500 AD) at Maheshwar and
Navdatoli (Sankalia, et al., 1958:159).
Table 7-10. Vessel types which are exclusively found in the micaceous ware families and their
proposed phase based on published parallels.
232
Chapter 7. Ceramic assemblage
The most common type (44) identified in the Micaceous wares has parallels with
published examples from Pauni period IV (1st century BC to 3rd century AD), Nevasa
period V (50 BC to 200 AD), Vaisali period IV (200 to 600 AD) and Sanchi (3rd to
2nd century BC) which supports a generally Early Historic chronological range. The
second most common type (47) also related to Pauni period IV and to period III at
Paunar (3rd to 8th century AD).
The Mica1 ware family broadly encompasses the fabrics described as ‘orange
micaceous wares’, which are frequently found together in significant numbers and
tend to be coarse, uniformly orange in colour with minimal surface treatment.
Table 7-11. Individual ware types contained within the Mica1 ware family, and the count of sherds
within each type.
There is a large range of vessel types across this ware family. There are 41 defined
types, 24 of which are only represented by one example in this ware, and there are 5
unique shapes. The types primarily represent a range of simple bowls, lid-cum-bowls,
globular pots, and long necked jars with the most common types being carinated
handis. Types 18 (small bowl with bifacial rim) and 48 (carinated cooking vessel) are
exclusive to this ware family. A number of the Mica1 types show some correlation
with other wares, particularly other micaceous wares as in types 47 and 49, or both
micaceous and red wares (e.g. types 11, 37, 55 and 74).
233
Chapter 7. Ceramic assemblage
Table 7-12. The main vessel types identified within the Mica1 ware family, and the count of examples
per type. The vessel types displayed here have over 3 examples in this ware family.
Figure 7-2. The most common vessel shape present in the Mica1 ware family.
In terms of the proposed chronology, a number of the more utilitarian vessel shapes
are found across different chronological phases in published reports, but the most
commonly occurring types have published parallels dated to the Early Historic
period. Type 55 is the most common rim in the Mica1 ware family and has parallels
to carinated handis in micaceous wares at Bhokardhan (Period IA, 2nd to 3rd century
BC) (Deo & Gupte, 1974). Type 49 is one of the most common vessel shapes in
micaceous ware in the assemblage and has parallels to micaceous carinated vessels at
Maheshwar and Navdatoli (period VI, 200 to 500 AD) but also resembles rims from
7th century AD Sanjan (Nanji, 2011; Sankalia, et al., 1958). Type 48 is another form
of carinated pot with published parallels from Nevasa period IV (150 to 50 BC),
Pauni period II (300 to 100 BC) and Timbarva period I (400 BC to 0 AD). Type 37
has similarities to lids from period III of Baroda (600 to 1000 AD). Type 74 is
recognisable in shapes from Pauni period II (4th to 2nd century BC), Baroda period II
and Devnimori, while type 100 can be compared to sherds from period II at Timbarva
(0 to 300 AD) and Late Sātavāhana Kaundinyapura (references as before). The
234
Chapter 7. Ceramic assemblage
The Mica2 family is the largest group of micaceous wares, and second only to
grey/buff wares in the overall assemblage. This group includes the harder, more
compact micaceous fabrics with a black core. These ceramics are less orange in
colour than those in Mica1, and have been generally subjected to more surface
treatment such as thicker red slips.
Table 7-13. Individual ware types contained within the Mica2 ware family and the count of sherds
within each type.
This ware family has a very large range of vessel shapes, covering 76 types and a
further 18 unique types. Of the 76 defined types, 21 are represented by only one
example. A number of types exclusively occur in Mica2 wares (types 15, 17, 45, 50,
235
Chapter 7. Ceramic assemblage
52, 118, 128 and 186) and these include simple bowls, carinated vessels and globular
pots. As with the Mica1 ware family, the most common type (47) is a carinated handi
and a number of the frequently occurring types are shared across the two micaceous
ware families. There is a particular association of types with other micaceous wares,
and some of the vessel shapes common to Mica2 wares are also present in meaningful
amounts in red and red burnished wares (e.g. types 53 and 58). A number of the types
appearing more commonly in Mica2 wares than Mica1 wares have a range of
comparable published shapes dated to the Early Historic to Early Medieval period.
For example, type 75 is similar to a red ware vase dated to 2nd century BC to 6th
century AD at Sanchi (Shaw, 2007) as well as rims from Early Historic (Sātavāhana)
Kaundinyapura, Baroda period II (100 to 600 AD), Timbarva period I (400 BC to 0
AD), Paunar period III (3rd to 8th century AD) and Bhokardhan period IA (2nd to 3rd
century BC). In terms of decoration, the Mica2 wares generally feature incised
horizontal lines and small notches as in the Mica1 wares. There are also occasional
examples of the stamped floral motif, typically associated with Gupta period ceramics
onwards.
Table 7-14. The main vessel types identified within the MICA2 ware family, and the count of
examples per type. The vessel types displayed here have over 4 examples in this ware family.
236
Chapter 7. Ceramic assemblage
Figure 7-3. The most common vessel shape in the Mica2 ware family
The most common vessel types do appear to have published parallels which usually
correspond to phases II and III in the project chronology. Type 146, a globular pot
with a split rim, is particularly interesting as an example was found at the 5 th century
remains at Mansar, which may strengthen the association of this ware and type with
Vākāṭaka activity. However, the types present in Mica2 often show a wide
distribution across other wares and can have published parallels extending into later
phases (IV). For example, Type 150 has a wide range of parallels in publications
covering Early Historic to Medieval phases, however has only been found to be
comparable to published vessels in very different fabrics (Deo & Dhavalikar, 1968;
Nanji, 2011; Subbarao, 1953) This highlights the potential for certain forms to have
long periods of use for particular functions.
Dark micaceous wares are relatively rare in comparison to other micaceous wares and
tend to appear at the same sites as Mica2 ceramics in particular. The four wares
included in this family are differentiated from other micaceous wares by their fabric
colour, which varies from grey to dark brown. Published reports make little reference
to dark micaceous wares, although ‘micaceous black wares’ were reported from the
excavations at the Vākāṭaka sites of Hamlapuri and Mandhal (Bhaisare, 2010-
2011:104).
237
Chapter 7. Ceramic assemblage
Table 7-15. Individual ware types contained within the Dark Mica ware family and the count of sherds
within each type.
There are 42 vessel types and 8 unique shapes within this ware family, but the vast
majority are only represented by one or two examples. As seen in other micaceous
wares, carinated handis are most common and other broad types include globular pots
and lid-cum-bowls. All of the most common types presented below relate to the most
common shapes in the micaceous wares already discussed. The types present in Dark
Mica ware show a wide distribution across other wares, but primarily match those in
other micaceous fabrics. Published parallel types have a broad chronological range,
relating to phases II-V in this project. Furthermore, the decoration on Dark Mica
wares mostly does not differ to the red micaceous ceramics as it is dominated by
oblique notches or incised horizontal lines with occasional wavy lines.
Table 7-16. The main vessel types identified within the Dark Mica ware family, and the count of
examples per type. The vessel types displayed here have over 2 examples in this ware family.
238
Chapter 7. Ceramic assemblage
Figure 7-4. The most common vessel shape in the Dark Mica ware family.
This ware family only contains the Patchy micaceous fabric, which is considered on
its own due to its distinct appearance and limited distribution. The majority of sherds
were collected from a discrete scatter at one site (49), and all sherds were of a single
type (11, plain bowl), which is ubiquitous across all phases of published sequences.
This bowl type features no decoration and is formed with the commonly identified
string cut base. A specific production and use is implied for this ware, as it was
considered to skew the results if included in the other micaceous families.
Table 7-17. The count of sherds within the ware type ‘PATCHY’, which constitutes the ‘Other Mica’
ware family.
Figure 7-5. Type 11, which is the only vessel shape identified in the Patchy micaceous ware.
239
Chapter 7. Ceramic assemblage
Red wares are the third most common ceramic family in the assemblage, and are
represented by 794 sherds across 76 sites. Two broad sub-groups were originally
defined, consisting of those with a solid red/orange core and those with a black/grey
core. There were also slipped and unslipped varieties, and mica is occasionally found
added into the red slip, similar to the red slipped wares described at Pauni (Deo &
Joshi, 1972:65). Regardless of core colour and surface treatment, all the red wares
grouped here appear to behave in a similar fashion and mostly co-exist.
Table 7-18. Count of ceramic sherds within each ware type in the Red ware family.
The red wares are mostly generic, medium to coarse ceramics with little decoration.
There are difficulties in comparing this group to published reports as local, plain red
wares can be ubiquitous across sites and periods, and are invariably loosely described
(Nanji, 2011:67). Coarse red wares persist throughout all periods at Timbarva (c.400
BC to 1000 AD) (Mehta, 1955:6), and are the most common group of wares present
across all phases at Pauni (Nath, 1998:27-37). Publications do seem to indicate that
the quantity of red wares in excavated assemblages changes over time. Red wares are
often cited as becoming more common in the Early Historic, with a particular
predominance from the Sātavāhana period onwards. At Pauni, red ware is the primary
Sātavāhana period ceramic (Nath, 1998:37), and red wares dominate Period III at
240
Chapter 7. Ceramic assemblage
Paunar (c. 3rd to 8th century AD), occurring in high necked jars and carinated handis,
which were absent in earlier periods (Deo & Dhavalikar, 1968:50, 115). An
abundance of red wares in Early Historic contexts is noted in Period V (150 BC – 250
AD) at Adam (Nath, 1992a:76), and Period II (100 to 600 AD) of Baroda (Subbarao,
1953:36). Red wares with a red wash were abundant in Periods IV and V at Nevasa
(c. 150 BC - 200 AD) in a wide range of utilitarian shapes (Sankalia, et al.,
1960:277). The quantity of red ware is suggested to decrease across the Medieval
period, as suggested at Brahmapuri where red ware is limited in the Bahmani layers
but present in large quantities in the ‘Late Sātavāhana’ period (4th to 9th century AD)
(Sankalia & Dikshit, 1952:62).
Occasional studies have attempted to distinguish between earlier and later red wares
based on decreasing quality of firing and ceramic fabric. This has little support and
such suggestions may have been influenced by prevalent theories concerning urban
decline. A range of firing quality is observed across the red wares in the Rāmṭek
assemblage, but this does not seem to correlate with phasing. There is a long tradition
of red wares in India, which may not have changed substantially, as publications
often note similarities between Early Historic and Medieval red wares (Nanji,
2011:68). For example, coarse red ware is present in all historical phases at Nevasa,
and many shapes and fabrics continue across Periods V and VI despite an 800 year
break in occupation (Sankalia, et al., 1960:277). Vessel types are more likely to
indicate chronological change, but due to the functional nature of many red wares,
generic vessel types can be ubiquitous.
The red ware rims occur in 48 types, with 30 of these represented by one or two
examples. There is an additional 14 unique rim shapes, but no vessel types were
exclusive to red wares. The most common type is 11, a plain bowl which primarily
occurs in red and micaceous wares. Other common vessel shapes include various
high-necked jars/vases, globular pots and lid-cum-bowls. A number of the common
red ware types are regularly found in micaceous wares, while others show a broader
range across other ware families, such as grey/orange or red burnished wares.
241
Chapter 7. Ceramic assemblage
Table 7-19. The main vessel types identified within the Red ware family, and the count of examples
per type. The vessel types displayed here have over 4 examples in this ware family.
Aside from type 11, the most common shape (Type 76) shows distinct similarities
with red ware sherds from Pauni period IV (1st century BC to 3rd century AD),
Baroda period II (100 to 600 AD), Timbarva period 1 (400 BC to 0 AD) and Vaisali
period IV (200 to 600 AD). Type 64 is comparable to a red ware high necked vessel
from period IV at Nevasa (150 to 50 BC) and Mauryan period Kaundinyapura.
However, a range of dates is demonstrated by types such as 37 which is comparable
to a sherd from period III at Baroda (600 to 1000 AD), and type 93 which has
published parallels at Pauni, Timbarva and Paunar which span from 1000 BC to 3rd
century AD. Simple red wares are common, probably locally produced and are likely
to have existed in some form across a range of time. While the most common types
do find parallels in published reports from Phase II/III in the project chronology,
242
Chapter 7. Ceramic assemblage
many of the vessel shapes are similar to published examples common across Phase I
to IV.
There is often no decoration on the Red ware sherds however when it occurs it is
similar in range to that typically found in the micaceous wares. Most common are
incised horizontal lines and oblique notches while limited sherds display the stamped
floral motif or pinched applique bands. The incised or impressed decoration
resembling ‘basketry’ is most common in the Red ware family and as mentioned
appears to be a characteristic style of the Early Historic period (for example, 1st
century BC to 3rd century AD at Pauni (Nath, 1998)).
The ‘sprinkler’ (a specialised type of spout) is one form which is noted in published
reports as being particularly characteristic of Early Historic ceramics; examples of
which were found in plain red ware during this survey at Nagardhan Fort (site 45) and
Kawadak (site 154). Sprinklers are often found in Red Polished ware (RPW), and
consequently are associated with Early Historic activity, particularly during the
Sātavāhana period (Subbarao, 1958:47). Red ware sprinklers have been described at
the Vākāṭaka sites of Hamlapuri and Mandhal; they are considered to have persisted
with little change in form from the preceding Sātavāhana period, albeit in a coarser
fabric, but are not a post-Vākāṭaka ceramic (Bhaisare, 2010-2011:105). Red ware
sprinklers are also noted from the 5th century monuments at Mansar (Manjhi, et al.,
2000:56).
Table 7-20. Sites which have reported ‘sprinklers’ during excavations and the relevant phasing in
which this sherd type was encountered.
243
Chapter 7. Ceramic assemblage
Red Burnished red wares are separated from the plain red ware family, due to their
distinctive material characteristics and a comparatively weak correlation with other
red wares. Red Burnished red wares contain few inclusions and very little mica, and
have a hard, compact fabric with a thick bright or dark red slip with a burnished
lustre. Red Burnished red wares are difficult to distinguish in the consulted published
excavation reports; burnished red wares are often noted alongside other red wares as
in the Shamalaji report (Mehta & Patel, 1967:23), although there is occasional
mention of similar wares, such as the red ware with bright red slip common in Period
III (600 to 1000 AD) at Baroda (Subbarao, 1953:41). Burnished red wares have been
described as one of the two basic types of Indian Medieval ceramic, alongside
burnished black ware (Mehta, 1979:42). However, Medieval ceramics are often
overlooked in publications of Early Historic sites, and so there may be a lack of
information about these wares if they relate to a later period.
Table 7-21. Count of ceramic sherds within each ware type in the Red Burnished Red ware family.
A number of the types represented by red burnished red wares support the theory that
they are distinct to plain red wares. There is a large range of vessel shapes with 60
defined types and 11 unique shapes. 40 of the 60 vessel types are represented by only
one or two examples, and types 31, 160 and 163 exclusively occur in this family. A
number of the vessel shapes are also common in plain red wares and micaceous
wares, but a meaningful similarity to the grey/buff wares may be inferred from the
majority of types. Common types in Red Burnished ware (such as 24, 148, 149, 152,
159 and 184) are most numerous in Grey/Buff wares, and others such as type 43
occur primarily in Red Burnished wares but have limited examples exclusively in
244
Chapter 7. Ceramic assemblage
grey/buff wares. Common forms include varieties of globular pot with a split rim,
shallow bowls and dishes. Based on published parallels, several types relate to Phases
IV and V in the project chronology, while others are common vessel shapes with
earlier parallels from the Early Historic.
Table 7-22. The main vessel types identified within the Red Burnished Red ware family, and the count
of examples per type. The vessel types displayed here have over 4 examples in this ware family.
Figure 7-7. The most common type present in the Red Burnished Red ware family.
The most frequently identified types have already been noted in the Mica2 ware
family. Aside from these vessel shapes, Type 43 bears similarities to a red ware dish
from period VI at Baroda with the late dating of the 19th century AD. Type 149 may
be comparable to grey or burnished black jars found at Pauni period III (1st century
BC to 3rd century AD), Maheshwar and Navdatoli period VI (200 to 500 AD). Such
early data ranges stand in contrast to the chronology of other vessel shapes such as
Type 159, which has been related to a globular jar in a drab black ware at Paunar
(period IV, 10th to 16th century AD). This demonstrates the possibility of a broad
chronology for certain types. In terms of decoration, the Red Burnished wares are
primarily plain but the most commonly found feature are simply horizontal lines
245
Chapter 7. Ceramic assemblage
running around the shoulder of vessels. This type of decoration is more common in
Red Burnished wares than in the red and micaceous wares already described.
Occasional sherds have incised lines or notches and there are a few examples of the
stamped floral motif.
The Grey/Buff ware family is the most common ceramic group in the Rāmṭek
assemblage, and contains the most numerous individual ware (GW.BLS). This family
has a distinct fabric, and displays a different range of types. Grey/Buff wares are
hard-fired with a medium to coarse, fairly compact fabric, which is usually uniformly
light grey. The wares contain small grits as inclusions and mica is very rare in the
ceramic fabric. Occasionally mica is present in the slip, which is usually black or
greyish-brown in colour and burnished.
Table 7-23. Count of ceramic sherds within each ware type in the Grey/Buff ware family.
The black slipped grey wares included in this family are not comparable to early
black slipped wares mentioned in excavation reports as being present in Prehistoric
and Early Historic contexts. The fabrics and shapes here are more akin to much later
grey wares; Medieval plain and burnished black wares are described as being gritty,
fired in a reduced atmosphere, and occurring in a range of utilitarian shapes such as
cooking vessels, bowls, dishes and pots (Mehta, 1979:42). At both Pauni and Paunar
coarse and burnished grey wares are recorded as Medieval ceramics (Deo &
Dhavalikar, 1968:70; Deo & Joshi, 1972:57). Burnished and coarse black slipped
246
Chapter 7. Ceramic assemblage
grey wares were recorded at Medieval Sanjan (Nanji, 2011:69), and are found in
Period VII (Muslim-Maratha) at Maheshwar and Navdatoli (Sankalia, et al., 1958).
At Timbarva and Baroda, black burnished wares are considered to become prevalent
from around 600-1000 AD (Mehta, 1955:6; Subbarao, 1953:41). Further afield,
Indian black slipped and burnished wares with a lack of fabric mica, were reported by
Kennet as present at Ras al-Khaimah around the 7th or 8th century AD (Kennet,
2004a:89). Publications seem to suggest that black slipped wares became common
from this period alongside a reduction in red wares, as at Brahmapuri (Sankalia &
Dikshit, 1952:55).
Many of the most frequently identified shape types in the Grey/Buff ware family are
either common to previously discussed wares, such as Mica2, or currently find no
parallel in the available published literature. However, several of the globular pots are
comparable to vessels from later levels of sites, such as Medieval Sanchi or period IV
at Dwarka (10th century AD to Medieval period). The most distinctive element of the
Grey/Buff is the common decoration of applique flanges or ledged carinations. Other
247
Chapter 7. Ceramic assemblage
forms of decoration include a range of wavy or straight incised horizontal lines and
more common occurrences of the stamped floral motif.
Table 7-24. The main vessel types identified within the Grey/Buff ware family, and the count of
examples per type. The vessel types displayed here have at least 10 examples in this ware family.
Figure 7-8. The most common type present in the Grey/Buff ware family.
Grey/Buff sherds were almost entirely surface finds, most commonly associated with
modern villages and their surroundings. This, alongside the phasing indicated by the
published parallels for ceramic fabric and vessel shape, seems to suggest that they are
late, post-Early Historic ceramics. It is difficult to establish an exact date range given
248
Chapter 7. Ceramic assemblage
that there is often a lack of published information about Early Medieval and Medieval
ceramics, and consequently a number of the vessel shapes do not have a parallel in
the consulted reports. However, it can be stated confidently that this is the latest
group of ceramics in the Rāmṭek assemblage.
Gritty wares are defined by the great quantity and size of grits present in the ceramic
fabric, and their hard-fired, heavy and stony qualities. These ceramics have little
surface treatment beyond occasional slips, and all except two varieties (GRW and
GRW.RS) contain large pieces of mica or micaceous stone in the ceramic fabric.
Gritty wares are found at a limited number of sites, including those with Black and
Red ware, and are generally associated with red and micaceous wares. Gritty wares
very rarely occur at sites dominated by red burnished or grey/buff wares, both of
which are considered to be later in the relative chronology. These associations may
suggest a prehistoric or Early Historic phasing: In reports, gritty wares are mentioned
in Period IV (2nd half of 1st century BC to 3rd century AD) of Pauni (Nath,
1998:37), and Period II (0-300 AD) at Timbarva (Mehta, 1955:6).
Table 7-25. Count of ceramic sherds within each ware type in the Gritty ware family.
249
Chapter 7. Ceramic assemblage
The gritty ware rims can be assigned to 22 types and 5 unique shapes, with most
types only represented by a couple of sherds. The most common type (65), and types
67, 87 137 and 173, are primarily represented by this ware family, with occasional
micaceous or red ware examples. The most common types in this ware family are
heavy storage jars or jars with long flaring rims which seem suited for pouring, and
may imply a functional reason behind the high grit content of these wares. No sherds
display decoration which may support the identification of the sherds as part of
utilitarian vessels. The vessel shapes are quite common but a number are similar to
vases with out-flaring rims said to be common at Megalithic sites (Shete, 2009:74).
Types 65 and 87 find parallels in gritty red ware vessels from Early Historic
Kaundinyapura, with 87 also being comparable to sherds from period I at Dwarka (1st
to 2nd century BC), Pauni period IV (1st century BC to 3rd century AD), Timbarva
period I (400 BC to 0 AD) and Paunar period III (3rd to 8th century AD). Type 67 also
supports a fairly early date range for these ceramics as it bears similarities to sherds
from Pauni period II (3rd to 1st century BC). The remaining, most common shapes in
the gritty ware all correspond to date Early Historic data ranges. This supports a
relatively early phasing for this ware, from perhaps the late prehistoric to Early
Historic periods.
Table 7-26. The main vessel types identified within the Gritty ware family, and the count of examples
per type. The vessel types displayed here have over 2 examples in this ware family.
250
Chapter 7. Ceramic assemblage
Figure 7-9. The most common type found in the Gritty ware family.
The storage jar wares have been separated out due to their low quantity in the
assemblage and their specific intended function. Compared to standard red and
micaceous wares, these ceramic fabrics are thicker and heavier with large rims, and
are particularly coarse with little surface treatment and a higher degree of vegetable
temper. The functionality and limited number of the storage jar sherds reduces their
usefulness in ceramic seriation, as they are utilitarian, lacking in distinctive features
and common across phases in publications. However, their presence at sites may
indicate specific activities and the storage of goods. The most common types (179
and 192) are both large storage jar rims, and other vessels represented are large
globular pots or thick jars. A number of the common types in this ware family are
also represented by red and micaceous wares. Type 172 displays similarities to coarse
red storage jars from period IA at Bhokardhan and type 136 has parallels to storage
jars from around 3rd century BC to 3rd century AD at Pauni and Bhokardhan.
Table 7-27. Count of ceramic sherds within each ware type in the Storage ware family.
Table 7-28. All vessel types identified within the Storage ware family, and the count of examples per
type.
The grey/orange wares are distinct in appearance, with a uniform light grey interior
and orange exterior. Most grey/orange wares do not contain mica and have few
inclusions overall, although a micaceous variety is present, and some are treated with
a thin orange-red slip. They are rare in the assemblage and are difficult to find
parallels for in published literature. They are perhaps considered to be a variety of red
ware in publications but without access to assemblages, I am unable to identify this
more closely.
Table 7-29. Count of ceramic sherds within each ware type in the Grey/Orange ware family.
252
Chapter 7. Ceramic assemblage
Grey/orange ware rims display 22 types and 2 unique shapes, but all are simply
represented by one or two examples. Type 69, which is a variety of long necked jar, is
exclusive to this ware family. One sprinkler was found in this ware, which may
suggest an Early Historic date, as in the red wares. It is further noted that grey/orange
wares are particularly found associated with red wares. This may corroborate the
suggestion that this ware could be absorbed within broader red ware groupings in
publications.
Black slipped red ware is rare within the assemblage as 47 sherds were collected.
While not numerous, as with other less commonly identified sherds, its distribution
may be indicative of aspects of occupation patterns or activity in the landscape. It is
visually distinct, and does not fit the pattern of the red ware family; black slipped red
ware is more similar in occurrence and type to the red burnished red wares and
grey/buff wares. A similar ware is mentioned as occurring across most of the
occupation sequence at Sanjan, with the highest concentration being in layer 5 (from
7th or 8th century AD) (Nanji, 2011:76). At Brahmapuri, black slipped red wares
become common in the late or post-Sātavāhana period through to the Bahmani
period, alongside other black slipped wares (c. 4th to 16th century AD) (Sankalia &
Dikshit, 1952:55).
Table 7-30. All vessel types identified within the Red BLS ware family, and the count of examples per
-type.
253
Chapter 7. Ceramic assemblage
Examples of 7 rim types in RED.BLS were collected; the most common of these
types are primarily represented by grey/buff wares, and type 148 is also numerous in
red burnished wares. Type 100 is the exception as it is a common vessel across all
ware families. Where published parallels have been found for the types present in
RED.BLS, they seem to indicate a Phase IV or V date for this ware.
The ‘unknown’ ware family contains over-fired red wares, ceramic fabrics which
could not be classified and BLARB, which is a rare burnt ceramic fabric not
identifiable to its original ware group (almost certainly a red ware, but charring
precludes an exact ware definition). The ‘unknown’ ware family only incorporates 87
sherds and is not used in the ceramic seriation, given that it is a broad grouping of
largely dissimilar and unclassified fabrics. For example, the NON-ID category
contains unique micaceous wares, including one with a golden micaceous slip, as
well as various red wares.
Table 7-31. Count of ceramic sherds within each ware type in the Unknown ware family.
Figure 7-10. Two sherds included in the ware group NON-ID: a) Sherd 3317 with a golden micaceous
slip, b) Sherd 4221 which is also micaceous but has a light pink fabric and a friable texture.
254
Chapter 7. Ceramic assemblage
In order to demonstrate the ware associations the data was queried on different levels
within Microsoft Access, and manipulated in Microsoft Excel using pivot tables.
When all 74 wares were taken into account, the sheer number was difficult to
organise in order to identify meaningful patterns in site assemblages. Equally the
similarities between some of the ceramic wares can obscure distributions, which is
particularly true of the micaceous and red wares that co-occur at many sites in
different quantities. To simplify the process, attention was focused on the major sites
identified during survey (i.e. those with the largest ceramic assemblages). Discrete
sites which yielded over 0.5% of the total assemblage were considered, as opposed to
255
Chapter 7. Ceramic assemblage
areas of low density ceramic scatter in transects. This comprised 49 sites and included
the major mounded sites with comparatively dense surface scatters.
256
Chapter 7. Ceramic assemblage
Furthermore, only key wares were considered, which were defined as the most
numerously occurring (over 1% of assemblage). This again results in a more
manageable and meaningful dataset, and while there is still a large number of sites
and wares to work with, clearer patterns of association can be discerned. Using a
crosstab query in Access and pivot tables in Excel, sites can be displayed against
ware to show relative concentrations of ceramics across the sites. The cells were
conditionally formatted to be colour coded based on the percentages of wares. This
method may also be carried out using types against wares or sites. At this stage, the
ware family groupings could be tested to ascertain how reliably the ware families
could be defined; the Black and Red, Red burnished and Grey/Buff wares in
particular appear to be coherent. Conversely, the similar behaviour of many of the
micaceous and red wares creates difficulties in breaking down this large
conglomeration of ceramic fabrics but while there is overlap the broad groupings
appear to be supported. In particular the Mica1 wares occur together frequently, while
Mica2 wares appear to display a different pattern of association.
The ware families more clearly indicate the interaction between broad ceramic types
and patterns of association and change, and so the previous method of seriation was
repeated to view ware families against the major sites. The table was conditionally
formatted by colour to relate to the lower, middle and upper percentiles of ware
family occurrences, and the site data was manually rearranged based on comparative
ceramic assemblages. While some sites display a broad range of ware families,
several relative phases appear to emerge. Firstly, the Red and Mica1 ware families
257
Chapter 7. Ceramic assemblage
It may be suggested that there were different periods of dominant wares, and while
these phases are not dated, they can be proposed to be related to historical
developments in chapter nine. Such proposed phasings can be further demonstrated
by re-arranging seriation tables based on the percentages of sherds from the major
ware family (BAR, Red, Mica1, Mica2, Red Burnished and Grey/Buff wares). For
example, by organising the sites based on the highest percentages of red wares, it is
possible to identify a point where the sherds of a different ware family become more
significant in quantity.
Table 7-33. Example table which gives the distribution of red and micaceous wares in sites with over
100 sherds. The data is sorted by quantity of red ware from largest to smallest. This clearly
demonstrates the proposed different phasing of the Red ware family and Mica2, with overlap occurring
between Red and Mica1 wares. See appendix eight for the full seriation table.
258
Chapter 7. Ceramic assemblage
Another distinction is clear between the red or micaceous wares and the Grey/Buff
wares, which are rarely found together, and this is considered to relate to a
chronological separation. The Red Burnished Red wares appear to fit in between
these different phases, given that they have a meaningful correlation with Grey/Buff
wares but also occur at sites with few grey wares but a higher proportion of earlier
red and micaceous wares. It can be noted that Mica2 is more likely than Mica1 to
occur at sites with these later wares, which perhaps lends more support to suggestions
of a different phasing from the Red and Mica1 wares. Based on the associations of
the Red Burnished Red wares, it could be inferred that there is a phase where these
ceramics occurred without grey wares, as well as a period of overlap. There is a
reasonable case for the grey wares to be the latest ceramics in the relative phasing,
given their ubiquity as surface remains and their association with modern villages and
Medieval structures.
Ceramic seriation is of some use for understanding less significant wares, and
generally supports the chronological suggestions from typological comparisons. It is
clear in the limited site assemblages with gritty wares that this family is most strongly
associated with red wares, and often found at sites with Black and Red ware sherds.
These associations place this ware early in the relative ceramic seriation and may
strengthen a proposed Early Historic date or an origin in the Protohistoric period. The
storage wares and grey/orange wares seem to largely correlate with the Red and
Mica1 wares, which shows some diversity in ware production at this point in the
phasing. Dark Mica wares most commonly occur with Mica2 wares, which
demonstrates that they perhaps fit into the third phase of the ceramic progression.
To illustrate the proposed ware family seriation, the example sequences of four
example sites are shown below. The spread of wares at these sites effectively
demonstrates the potential separate phases as well as the proposed overlapping
phases.
259
Chapter 7. Ceramic assemblage
Table 7-34. Ware family distributions at the sites of Hamlapuri (79), Bhijewada (210), Mandri (348)
and Kelapur (353). Counts are shown as a percentage of the assemblage at each site with the total
sherd count given in the final column. Ware families with minimal numbers of sherds have been
excluded and the table has been conditionally formatted to illustrate the spread of wares.
Firstly, Kelapur (site 353) displays an interesting spread of wares, including both the
earliest and latest ware families identified in the assemblage. This site is mentioned in
a Vākāṭaka inscription as having been in use or known about during the Early
Historic period, which may have some significance for the dating of the common red
wares in this site’s assemblage. The assemblage at Kelapur is interesting given the
lack of micaceous wares if they indeed relate to later Early Historic activity (to be
further discussed in chapter nine). Hamlapuri (site 79) is regarded as a purely
Vākāṭaka site based on trial excavations, and field observations by the author seemed
to support this proposition by confirming the uniformity of the archaeological deposit
and identifying a relatively narrow range of wares and types. Hamlapuri is considered
to be one of the better-known points from which to work within the Rāmṭek
assemblage in order to aid the formation of a meaningful seriation; if the site is
Vākāṭaka, Hamlapuri may provide a base for our understanding of the characteristic
ceramic fabrics and types for this phase of the Early Historic. If typical of an Early
Historic site, the Hamlapuri ceramics clearly demonstrate the association between red
and micaceous wares. The less common Mica2 wares at Hamlapuri may indicate a
slightly different, perhaps later, phasing. The mound at Bhijewada (site 210) clearly
demonstrates the proposed different phasing of Mica2, and the range of ceramics is
also demonstrative of the occasional co-occurrence of the Mica2 ware family with
seemingly later wares. In this case the Bhijewada assemblage, and the presence of
Mica2, Red Burnished and Grey/Buff wares, may be indicative of duration of
settlement. Finally, the habitation mound at Mandri (site 348) is an example of a site
dominated by Grey/Buff wares representative of later occupation, with a smaller
quantity of Red Burnished red wares.
260
Chapter 7. Ceramic assemblage
The relative phasing appears to correlate well with the broad dates suggested by
published parallels for the ware types and the diagnostic vessel shapes. When referred
to the project’s working chronology identified in chapter six, the diagnostic pottery
for each phase becomes:
Table 7-35. The working project phasing including information from published reports and the ceramic
seriation.
Ware Key notes from published Notes from relative seriation Proposed
Family literature phase
BAR Characteristically prehistoric Rare in the assemblage but I/II
but extends up to 3rd century occurs at sites with a
AD predominance of red wares
Red Common from the Early Red wares may occur in large Primarily
Historic but red wares have a amounts on their own or are II/III
long duration. They seem to often associated with the Mica1
reduce in quantity by the family. Low levels of red wares
Medieval. Common across are found at most sites with a
phases II to IV wide range of wares
Mica 1 Micaceous wares appear Mica1 is found at sites with red II/III
across general Early Historic wares and is also frequently
phases in excavations. Major present alongside Mica2 wares.
types relate to phases II and Little association with
III Grey/Buff wares
Mica 2 Ceramic fabric relates to Associated with other micaceous II-IV
broadly Early Historic wares. wares, less commonly
The wide range of types find associated with red wares.
published parallels between Appears to be a second phase of
Phases II to IV micaceous ceramics
Red Few published parallels for Most strongly associated with III-V
Burn fabric, but a wide range of Grey/Buff wares, and
types, primarily related to occasionally occurs at sites with
phases III to V Mica2 wares. This possibly
indicates an intermediate
phasing or a long duration of
use.
Grey/ Medieval ceramic wares with Generally occurs at different V
Buff late vessel types sites to red and micaceous
wares, but if present it is
suggested to indicate duration of
settlement, given the proposed
chronological separation. Some
association with Red Burnished
wares, but Grey/Buff wares can
also dominate assemblages
261
Chapter 7. Ceramics
7.5. Conclusion
The results of the ceramic categorisation in the Rāmṭek survey should not be
overstated as there are clear limitations, but a proposed seriation has been
constructed. This provides a floating relative phasing based on ware associations.
Alongside comparison of ware types with published ceramics and the incorporation
of dateable parallels for vessel shape, this has resulted in broad chronological phasing
in line with that presented in chapter six. The chronological information is basic, but
does reliably demonstrate the earliest and latest wares present in the assemblage, and
provides a suggested phasing for the intermediate wares families. A better
understanding of the wares and types present and the relationships between different
types of ceramics has been achieved; given the lack of firm chronological markers,
seriation has produced a reasonable case for ceramic phasing, which can now be
tentatively linked to historical developments in the landscape in the next chapter. This
theoretical overview leaves much scope for further research as a nuanced phasing of
ceramics remains elusive. In particular, it remains challenging to subdivide the
expansive Early Historic and Early Medieval periods using this dataset. The
assemblage is relatively small and composed of numerous, local wares that can
individually be represented by a small quantity of sherds, which makes statistical
work very difficult. Comparison to published types to establish potential dates for the
ceramics was challenging due to the large degree of variation encountered, and the
utilitarian nature of the majority of wares. While the broad phasing can be quite
confidently presented, it is fully expected that it will be challenged by further
research. It is hoped that more data will be gathered in order to test the theories being
proposed in this preliminary study, and continue attempts to establish a reliable
phasing of material remains.
262
Chapter 7. Ceramics
the Rāmṭek landscape and hypothesise about periods of activity at sites. This becomes
particularly important in the next chapter, when the economic development of the
landscape is considered.
263
Chapter 8. The Economic Landscape
8.1. Introduction
Having discussed the overall results of the survey in terms of general findings and the
ceramics collection, this chapter will begin to interpret the data to consider the
Eastern Vākāṭaka economic investment into the Rāmṭek landscape. One of the thesis
aims is to contextualise Rāmṭek in its landscape in order to enhance our
understanding of economic change in the 4th to 5th centuries AD. This contributes to
our understanding of the Eastern Vākāṭakas’ investment in the area and their
organisation of production, agriculture and settlement. It is hoped that this will
contribute to a deeper understanding of the Eastern Vākāṭakas as a political entity and
help to establish the extent of their influence. The evidence presented here primarily
concerns settlement and production in the hinterland of the political and religious
centre, in order to explore the theory that there was a growth of agrarian settlement
under the Eastern Vākāṭakas. The historical narrative suggests that the Eastern
Vākāṭakas asserted their dominance over an underpopulated peripheral area, which
presented an opportunity for the establishment of a new dynastic centre, and so this
chapter seeks to investigate the relationship between this newly established political
and ritual centre and the hinterland
This chapter will address three issues focused around the Eastern Vākāṭakas as a
political entity impacting on the landscape: the nature of settlement in landscape prior
to Eastern Vākāṭaka influence; any visible change under the Eastern Vākāṭakas; and
comparison of Early Historic investment in the landscape to later medieval
occupation. The ceramic seriation from the previous chapter will be assessed spatially
and integrated with other evidence of economic investment into the landscape,
including environmental constructions for water management and other features or
finds related to agricultural production. The assessment of habitation remains helps to
elucidate the rural settlement pattern and indicate changes in the level of past activity
across the landscape over time. The survey was fairly successful in identifying
264
Chapter 8. The Economic Landscape
mounds which are important features in the Indian landscape created through the
persistence of settlements and a build-up of cultural material. They are often seen as
the base of rural archaeology related to settled human life. However, the possibility
exists for transient remains to be related to other subsistence strategies, which still
form an important aspect of the past landscape economy yet are difficult to locate.
This chapter will reflect on what cannot be ascertained about the past landscape from
the identified evidence and the potential absences in the data set.
Irrigation systems and tanks form an important part of the Indian landscape as water
management strategies have been used since ancient times to increase the availability
of water across seasons with variable rainfall (Bauer & Morrison, 2008:2207).
Ancient exploitation of water resources in South Asia, and the spatial distribution of
water management structures, appears to have been essentially directed by natural
patterns of rainfall, runoff and the topography. Given this, it may be suggested that a
modern pattern of environmental structures would not be expected to differ
substantially from the past (Sutcliffe, et al., 2011:781). Extant water reservoirs
around Mansar and Rāmṭek are consistent with archaeological and epigraphical
evidence of Vākāṭaka-period constructions, and further water management systems in
the hinterland may be indicators of widespread economic investment across the
landscape.
265
Chapter 8. The Economic Landscape
As outlined in chapter two, it is long held that the Eastern Vākāṭakas shifted their
capital, and the centre of their sphere of influence, to a relatively underpopulated and
marginal area. Initially it was believed that the Vākāṭakas extended the limits of
agricultural society and transformed the Rāmṭek landscape. However, the spatial
distribution of land-grants and regional archaeological material now suggests that the
Eastern Vākāṭaka dynasty was fairly localised within a region with pre-existing
populations. Regional archaeology demonstrates a long history of development across
Vidarbha, with a degree of influence by the previous centralised authority of the
Sātavāhanas. This is particularly meaningful given suggestions that the Sātavāhanas
developed Mansar and that the Vākāṭakas capitalised on an area with pre-existing
structures and social patterns. It has been theorised that the Vākāṭakas attempted to
engage communities through economic and ritual means but little is known about the
extent of their actual investment into the landscape. As our understanding of the
prehistoric, protohistoric and pre-Vākāṭaka Early Historic landscape of Rāmṭek is
limited, it was hoped that survey would identify either a presence or absence of
archaeological remains to give a preliminary indication of the level of past pre-
Vākāṭaka activity. This was aimed at ascertaining the extent of Vākāṭaka impact on
the landscape. It was hypothesised that having asserted their influence at Rāmṭek, the
Eastern Vākāṭakas and brought this ‘peripheral’ location into networks of regional
production in order to support the prosperity of their kingdom.
The earliest phase of landscape development around Rāmṭek remains elusive after the
field survey. It seems likely that Rāmṭek saw at least limited or transient activity
given that prehistoric occupation is well attested in the region. No stone tools were
identified at Mansar during survey, unlike the findings of the excavations (Joshi &
266
Chapter 8. The Economic Landscape
Sharma, 2005:6). However, a number of worked stone tools were identified from
other sites, two of which may be considered genuinely prehistoric while others are of
less secure origin, which prohibits a definitive chronological attribution. The first
prehistoric stone tool appears to be a Middle Palaeolithic oval flake (Abbas, pers.
comm.) and was collected from a low mound near Umri, to the north of Rāmṭek (site
363). The flake was found on the surface of ploughed agricultural fields, in
association with brick remains and predominantly Early Historic ceramics. It is not
certain that this flake was found in its original location, but it may be taken as
evidence of prehistoric activity in the general area. The second worked stone is a
core, found approximately 2.5 metres below the top of a section in Dudhala mound,
near Nagardhan (site 236). No other material remains were found in the same context,
although the layer did contain ashy deposits, and later ceramic and brick remains
were confined to the top 50 cm to 1 m of the mound. Again, this contributes to the
suggestion of some movement of prehistoric people through the landscape.
Figure 8-1. Section in the mound at Dudhala (site 236) displaying the ashy layer, with the location of
find 20 indicated.
The remaining four microlith/flakes and a core were collected from site 265 (‘Bada’),
in the vicinity of what appears to be the remains of an Early Historic occupation
mound yielding bricks and ceramics. Of these worked stone pieces, two flakes appear
to be Upper Palaeolithic in origin, while the others are of an unknown date. Flake
267
Chapter 8. The Economic Landscape
tools are commonly reported across Vidarbha in early, often aceramic, deposits such
as Period I at Adam (c. 3rd to 2nd millennium BC) (Dikshit, 1986:12; Nath, 1991:93).
However, microliths are known to have continued in use throughout the Early
Historic period (Deo & Dhavalikar, 1968; Nath, 1998). Several of the microliths
identified during survey may originate from a later period as they are associated with
Early Historic material and lack defining prehistoric features. The limited number of
worked stone artefacts recovered may be due to the survey methodology and low
visibility across the agricultural landscape, as well as an untrained eye for this
category of artefacts. However, they are not negligible findings as despite the absence
of the remains of prehistoric production activity or sites displaying prehistoric
modification of the landscape (as seen elsewhere in the region, such as at Junapani),
the presence of stone tools support transient prehistoric activity across the Rāmṭek
area.
A lack of visible Megalithic sites and limited prehistoric pottery in the survey data
may suggest that the settled Iron Age presence at Rāmṭek landscape was low in
comparison to other areas in the region. Iron Age/Megalithic remains would not be
unexpected given the strong tradition of Megalithic activity in Nagpur District and
reports of Megalithic cairns along the bank of Mansar Lake (Joshi & Sharma,
2000:127). However, no standing remains were identified during this survey; this
may be due to site survival or visibility in the agricultural zone or may be a result of
the small survey area, which could potentially miss wider patterns of megalithic
occupation in the region. The more transient prehistoric occupation of Rāmṭek may
have been localised given its more peripheral location and distance from larger rivers.
Black and Red ware is certainly pre-Vākāṭaka, and while it cannot be definitively
assigned to the Iron Age, it does suggest occupation of the landscape prior to the
establishment of the Eastern Vākāṭaka state. The spatial distribution of sites with
Black and Red ware demonstrates that a number are found in the plain south of
Rāmṭek hill, but the two sites with the highest quantity of sherds are found north of
Rāmṭek hill. The sites are all situated in low-lying areas with close proximity to small
rivers or water sources, which may suggest a concern with being located in plain
areas suitable for cultivation. The most northerly site of Kelapur (353) has the highest
number of sherds and is closely related to the tributaries leading into Khindsi
reservoir. The evidence of earlier ceramics at Kelapur is noteworthy as there is
268
Chapter 8. The Economic Landscape
Vākāṭaka literary evidence connecting this village to the construction of Khindsi tank.
It could be suggested that this site was established prior to Vākāṭaka rule in the area,
either in the Prehistoric or towards the beginning of the Early Historic. This supports
the suggestion that the Eastern Vākāṭakas exploited previous investments in the area.
Even more striking is the presence of Black and Red ware sherds at Nagardhan,
which suggest a long history of use at the site, with probable occupation prior to the
establishment of a major Vākāṭaka centre.
Figure 8-2. Map of the survey region displaying the sites with the presence of Black and Red ware
sherds.
269
Chapter 8. The Economic Landscape
Table 8-1. The sites from where Black and Red ware sherds were collected during survey.
Table 8-2. The ceramic assemblages of each site with Black and Red ware sherds. The table displays
the major ware families and the quantities of each appearing at the sites. The ware group ‘Patchy’ has
been omitted as only two sherds were present at Site 213 (Satak).
Black and Red ware is typically found at sites with assemblages containing
significant amounts of red or micaceous wares, which are considered to be broadly
Early Historic (table 8.2). Correlation of the Black and Red sherds with these Early
Historic ware groups may demonstrate the prehistoric foundation of sites followed by
continuation of settlement into the Early Historic. Alternatively, as the associated
archaeological material and site contexts lean towards Early Historic activity, Black
and Red ware may have been in limited contemporary use with Early Historic wares,
as seen in regional Early Historic contexts.
270
Chapter 8. The Economic Landscape
271
Chapter 8. The Economic Landscape
The Vākāṭaka arrival (in Phase III of the project chronology) is often assumed to have
prompted development of a thriving agrarian economy. This section presents the
survey findings related to Early Historic development in order to ascertain whether
economic and agricultural developments predated this proposed state-level influence
and to consider the extent of the Eastern Vākāṭaka impact on the area. If red and
micaceous ceramics are primarily indicative of Early Historic activity then the survey
results indicate a clear expansion in settlement size and number alongside a vast
increase in the quantity and variety of ceramic fabrics and vessel types at some point
in the Early Historic. This implies some socio-economic change resulting in higher
requirement, production and usage of ceramics.
The lack of RPW and other distinct pre-Vākāṭaka Early Historic ceramics leads to
questions about the nature of settled life during this period of Rāmṭek’s development.
However, plain red wares and micaceous wares are prolific across sites in the survey
area, and bear similarities to the local, everyday wares described in published reports
as indicative of the Early Historic period and Sātavāhana activity at regional sites.
The relative phasing established during ceramic seriation suggests a higher rate of
occurrence between Red wares and the early Black and Red ware which may support
hypotheses that the Red ware family contains ceramics representative of pre-
Vākāṭaka Early Historic activity. Unfortunately, these ceramics have ambiguous
phasing and cannot conclusively be identified as pre-Vākāṭaka wares, resulting in an
inability to state definitively at which point in the Early Historic settlements and the
economic landscape developed. Red wares are often considered to have no
“diagnostic value” because of the duration of this ceramic tradition (Shete, 2009:68).
While published parallels are often described as originating from the Sātavāhana
period, it is possible that these vessel types and ceramic fabrics persisted throughout
the Vākāṭaka phase of development.
272
Chapter 8. The Economic Landscape
and several are substantial mounds demonstrating evidence of settled life. The Red
ware sites are primarily spread across the low-lying areas south and just north of
Rāmṭek hill. This may demonstrate a preference for plain areas which can be
theorised to be connected to increasingly settled occupation and perhaps agricultural
expansion. The Nagardhan area shows a conglomeration of sherds and a number of
small mounds in close proximity, while Mansar and Rāmṭek demonstrate high
quantities of Red ware primarily around the monumental centres. Kelapur (site 353)
becomes a medium-sized site which again may be related to the inscriptional
evidence connecting the site to Early Historic activity. Many smaller sites are found
near to small rivers, such as Sonpur (site 386), the area around Old Mandri (Sites 60
and 61), and Bhilewada (site 400). There is an outlying site (Umri, site 363) which
shows that sites were not exclusively situated in the more accessible plain, although
the site is located on the bank of the small river leading to the Khindsi reservoir,
which supports the proposed preference for proximity to water sources.
Figure 8-3. Map of the survey region displaying the sites with the presence of Red ware sherds.
273
Chapter 8. The Economic Landscape
Table 8-3. Examples ceramic sequences from the known Vākāṭaka sites in the survey area
Specific identification of earlier Phase II foundations at Phase III sites may not be
possible through ground survey due to material culture continuity or the quality of the
data and comparison material. However, the ceramic seriation seems to indicate a
phase of sites with an Early Historic Red ware presence but few Mica1 wares - taken
to be representative of Vākāṭaka activity – which highlights the potential for a
number of settlements to have origins in the Phase II Early Historic. Pre-Vākāṭaka
Early Historic (Phase II) occupation represented by Red wares could have
encompassed seasonal populations and more settled groups in an expanding
agricultural area, upon which the Vākāṭakas imposed their new social and religious
order. The association of Red wares with Mica1 wares at other sites may suggest a
continuation of settlement into the Vākāṭaka period, indicating that Phase III activity
was part of an ongoing development with roots in pre-existing local social change.
The arrival of the Vākāṭakas to Rāmṭek may have resulted in the expansion of
settlement, an increased number of sites, and the promotion of agricultural and social
developments, but these transformations could have been built on earlier foundations.
Given the ambiguity of the Red wares, it could be equally valid to argue that the
Phase III saw the highest level of Early Historic investment in the landscape and that
the Eastern Vākāṭakas engaged the Rāmṭek area with wider networks of production,
prompting the political and socio-economic transformations of the 5th century AD.
Red wares are clearly represented at the major sites connected to 5th century Vākāṭaka
activity and an association with Mica1 wares seems to be characteristic of the
assemblages at known Vākāṭaka sites. The ceramic assemblage at Hamlapuri (site 79)
is similar to that around the 5th century monuments on Mansar and Rāmṭek hill and
displays prolific red wares alongside Mica1 wares. Despite the wares and types
having published parallels comparable to pre-Vākāṭaka Early Historic activity, this
274
Chapter 8. The Economic Landscape
Supporting evidence for significant Vākāṭaka economic activity comes in the form of
monumental remains and inscriptional evidence whereas there is a lack of non-
ceramic evidence for significant pre-Vākāṭaka activity. The monumental
constructions would logically require support from the rural population, and
presumably led to an expansion or intensification of settlement and agrarian activity.
If the assemblages at Hamlapuri and Mansar can be used as a point of reference for
Vākāṭaka ceramics, then the similar assemblages found at a number of mounds and
surface scatters across the survey area could be seen as signifiers of economic activity
related to Eastern Vākāṭaka authority. It is interesting to briefly note that the
Hamlapuri ceramics collected during survey differ from those described in previous
archaeological investigations at the site. While the types are broadly similar and
correspond to Early Historic vessel shapes, the majority of ceramic fabrics are plain
red wares or coarse micaceous wares, as opposed to the ‘Imitated’ RPW and fine
micaceous wares recorded by Bhaisare (2010-2011:104). Furthermore, contrary to
Bhaisare’s findings, no black-washed or black-burnished wares corresponding to
Early Historic activity were identified, although micaceous black ware (‘Dark
Micaceous’ here) is recorded in small numbers.
275
Chapter 8. The Economic Landscape
Table 8-4. Sites in the survey area with significant quantities of Red and Mica1 wares, as at Hamlapuri
and Mansar (included within the table).
The distribution of Mica1 sites is very similar to Red ware sites; 52 loci display both
ware groups and the similar site distribution strengthens the argument that both ware
families relate to an Early Historic spread of settlements. The generally lower
numbers of Mica1 sherds might support a suggestion that they represent a tradition of
micaceous utilitarian wares employed alongside the red wares. The more significant
concentrations of Mica1 wares at the three key sites of Vākāṭaka activity may imply a
preference for the production of micaceous wares at larger Vākāṭaka sites, while red
wares may have been more common across rural sites. This could be connected to
Smith’s observation at Kaundinyapura that even utilitarian materials and local
ceramics, such as Mica1, may have some other aesthetic or symbolic value and so
may have been differentiated in the past (Smith, 2002a:146).
276
Chapter 8. The Economic Landscape
Figure 8-4. Map of the survey region displaying the sites with the presence of Mica1 sherds.
Figure 8-5. Map showing the relative proportions of Red and Mica1 wares at sites in the survey area
277
Chapter 8. The Economic Landscape
As discussed in chapter three, Nagardhan’s association with the first Vākāṭaka capital
of Nandivardhana in publications has not been as solidly evidenced as the
identification of Pravarapura at Mansar. This survey confirms that the modern
settlement, the Bhonsle fort and fields surrounding the fort at Nagardhan village are
situated on extensive habitation mounds, and demonstrates that Nagardhan represents
the most comprehensive evidence for a large settlement site in the immediate Rāmṭek
landscape. Abundant red and micaceous wares are associated with Early Historic
bricks and the most significant quantity of pottery in the survey area was collected
from Nagardhan. This provides further archaeological support for its proposed
identification as the ancient Vākāṭaka capital however excavation would be required
to assess the true nature of the archaeological remains.
The total area of mounded land indicating ancient habitation at Nagardhan is roughly
50 ha, which is significantly larger than the other settlement mounds identified during
survey (each averaging around 2 ha). It is difficult to ascertain the past extent of the
site and whether the mounded areas are continuous, as the mounds are irregular in
shape and damaged by modern settlement. The mounded land southwest of the fort
278
Chapter 8. The Economic Landscape
Areas of Nagardhan display a similar range of broad ceramic wares and vessel shapes
to Hamlapuri and Mansar, but it is clear that the assemblages are not entirely
consistent across these known Vākāṭaka sites. Hamlapuri contains a range of largely
homogeneous and coarse fabrics while Nagardhan displays a greater variety of forms
and fabrics. At different locales across Nagardhan itself, assemblages are either
dominated by an assortment of Red and Mica1 wares or a prevalence of Mica2 wares.
There is over twice the amount of Mica2 compared to Mica1 sherds at Nagardhan. It
is possible that the Mica2 ware family represents a different category of
contemporaneous micaceous wares, perhaps related to a functional difference given
that Mica2 wares are less coarse than Mica1 and may have been the product of
localised manufacture at certain sites. The large quantity of Mica2 sherds at the
Vākāṭaka capital at Nagardhan and nearby larger village mounds may be related to
some form of site hierarchy within the rural hinterland; Dudhala mound (site 236) has
an assemblage dominated by Mica2 wares and appears to be the remains of a fairly
substantial settlement area with a small water tank in close proximity. Sites such as
this one may represent a tier of slightly larger settlement sites interspersed across the
plain. Coarser wares at sites such as Hamlapuri may be due to the settlements’ rural
nature or small size.
Nagardhan appears to have been an important site throughout the Early Historic
period and existed alongside a pattern of smaller sites in regular spacing across the
plain. The current investigations do not support the published view that Hamlapuri
and Nagardhan form a continuous ancient site as no archaeological remains were
279
Chapter 8. The Economic Landscape
280
Chapter 8. The Economic Landscape
Table 8-5. Areas of sherd collection around Nagardhan town giving the relative amounts of Red,
Mica1 and Mica2 sherds.
Figure 8-7. Section below the Bhonsle fort displaying layers of brick and ceramic
If there is a chronological division between Mica1 and Mica 2, then while the overall
number of Mica2 sites is similar to the number of Red and Mica1 ware sites, there is
a difference in distribution. This change is not represented spatially as the majority of
larger sites with common Mica2 ceramics remain distributed across the agricultural
plain. However, the change in distribution primarily concerns the quantities of
ceramics which may be related to site size and the expansion of some sites compared
to the contraction of others. Based on the quantity of Mica2 wares, the large sites of
the preceding phase (sites 79, 241, 265, 353 and 363) appear to reduce in size, while
five sites expand (sites 210, 236, 289, 364 and 318). The first four of these are located
281
Chapter 8. The Economic Landscape
in a rough line with Nagardhan, while site 318 (Panchala) has close proximity to the
theorised location of the 5th century Sudarśana reservoir (Khindsi). Mica2 wares
appear particularly focused around Nagardhan and nearby mounds, at the expense of
other sites. Given that there is less Mica2 at several known Vākāṭaka sites including
Mansar and Rāmṭek, it may be theorised that as the power structures collapsed
towards the end of Vākāṭaka rule, these monumental sites became less important and
contracted but the major population centre at Nagardhan persisted as an ongoing
settlement supported by its agricultural base.
Figure 8-8. Map of sites in the survey area displaying Mica2 sherds.
Table 8-6. Sites with the highest proportion of Mica2 wares, demonstrating the common association
with Red and Mica1 wares.
282
Chapter 8. The Economic Landscape
Table 8-7. Sites which are not represented by all three ceramic types, but which have more than 10
sherds of at least one type.
Figure 8-9. Difference in site distribution between those with Mica1 wares and those with Mica2
wares. A number of sites are represented by both types, however the distribution of ceramics changes
across the two ware phases.
283
Chapter 8. The Economic Landscape
There is some ceramic evidence for Early Historic activity on the northern outskirts
of Rāmṭek, at Kawadak mound (site 154). However, the density of modern settlement
makes it challenging to trace archaeological remains and ceramic assemblages tend to
consist of late wares, even in areas of digging. Where Mansar and Nagardhan are
concerned, dissimilar archaeological situations were revealed during survey at these
two ‘capitals’. Mansar displays monumental structures but no ancient ceramic and
brick remains were found outside of the monumental complex. Mansar village is not
situated on a habitation mound and no archaeological remains were located in areas
of digging in the village or the surrounding mining communities. Wellsted reported
the remains of a substantial town-site suggesting dense settlement around Mansar
tank, and theorised that the wider area was likely to have been thinly settled
(Wellsted, 1934:161). Wellsted also stated that while the settlement does not appear
to have been fortified, the surrounding hills may have been occupied by watch-posts
(Wellsted, 1934:162). On the ground, there is no evidence of this reported town-site
and the remains reported to the east of the monumental structures, are completely
untraceable. Fragmentary brick remains were found on the hilltops, but none were
substantial enough to indicate the nature or extent of any structures. At Nagardhan,
there are widespread brick and ceramic remains attesting to settlement at the site.
Furthermore, while it cannot be stated conclusively that monumental constructions
such as a royal ‘residence’ or Vākāṭaka fort existed, brick remains beneath the 18th
century fort, suggest that pre-existing structures have been masked by the Medieval
developments.
If a large settlement existed at Mansar, it would appear the evidence, which was
already greatly damaged by 1928, has since disappeared. Yet, Nagardhan has a
similar level of modern habitation to Mansar but substantial remains can still be
found. It may be hypothesised that the two sites held different functions despite both
being known as state capitals. A ‘capital’ is assumed to be the place from where
copper-plate charters were issued but the survey evidence suggests that a state
‘capital’ could have served a political function without necessarily having extensive
associated settlement for the general population. In this case, the administrative centre
and place of issue for copper-plate grants may have shifted but the population seems
to have remained focused around the agricultural hinterland of Nandivardhana.
Mansar was perhaps intended as a ceremonial and secular centre given its proximity
284
Chapter 8. The Economic Landscape
to Rāmṭek and the line of sight between the two monumental hilltops. These findings
correspond with the observation by Bakker & Isaacson (1993) that there is no
indication that Nagardhan was abandoned following the supposed move of the capital
to Pravarapura given that a charter was issued from Nandivardhana in the last decade
of the 5th century, (Bakker & Isaacson, 1993:68). Furthermore, Mansar itself may not
have been occupied by the general populace as the Mansar inscription states its
maintenance was imposed upon a collection of villages (Meister, et al., 1988:62).
If the term is used in a functional sense then it is proposed that Nagardhan may be
considered an ‘urban’ site in the context of the Rāmṭek landscape. This definition
may be warranted by its comparatively large size and the wide variety of ceramic
wares and vessel types, which implies a broad range of activities and different modes
of production. Smaller mounds tend to have a narrower range of ceramics and
evidence suggests localised production. For example, the ceramic assemblage at
Hamlapuri is dominated by a particular ware (ROM, mostly occurring in type 55) that
does not occur in large numbers at other sites. Despite indications of local
manufacture of ceramics, no ceramic production sites were identified - it may be that
remains are not visible as sites may have been simple pits – and as yet there is no
evidence of commercial activities or other production around Rāmṭek. Production
may have been largely carried out in Nagardhan’s urban centre, which is now under
modern settlement, while rural agricultural communities supported the temples and
production centres. The pattern of contemporary sites around Nagardhan suggests that
the town may have been a focal point for agricultural development in the landscape.
Furthermore, reference may be made to the slightly higher proportion of storage
wares and types in the Nagardhan assemblages, which indicates the provision of
surplus to the main site or production and storage activities.
The main activities around Rāmṭek appear to have been agricultural, as supported by
the artefacts identified during survey, which include quern stones and grinders. Other
evidence of economic investment in the landscape during the Early Historic period is
found in the potential remains of water management resources. Textual evidence
supports that reservoirs, often in the form of stone-faced earthen dams to contain
water from natural sources, were particularly common in South Asia from the Early
Historic period (Bauer & Morrison, 2008:2210). During survey, embankments of
disused tanks were encountered where the water storage area is now used as
285
Chapter 8. The Economic Landscape
agricultural land. These are visible from their earthworks and boundaries through
ground inspection, on satellite imagery and occasionally on topographic maps. Small
tanks identified during the survey often feature curved earthen embankments and
have the potential to have older origins. Dating of reservoirs is difficult as there may
not be any chronologically distinct features and the continued use and renovation of
reservoirs can exacerbate issues with dating (Hegewald, 2002:77). In order to
establish a period of construction, focus tends to be drawn to related epigraphic
evidence or scientific analysis (Bauer & Morrison, 2008:2210). Unfortunately the
majority of the water storage areas near Rāmṭek lack features to prove their antiquity
and there was no association with dateable artefacts. This is unlike the situation
recorded by Shaw at Sanchi, where Naga sculptures associated with reservoirs were
used to propose an Early Historic origin for their construction. Occasional ceramic
sherds and brick remains were found around the embankments at Rāmṭek, but these
do not give a solid indication of date. However, as noted, the use of tanks or water
storage reservoirs in the landscape is proven by the 5th century construction of
Sudarśana, which had the capacity to support the agricultural hinterland. Inscriptional
evidence puts the date of Khindsi reservoir slightly later than the rise of the
Vākāṭakas and so theoretically its construction may have been related to a subsequent
increase in population and intensity of agriculture, which required further water
access.
Early wells can be less easily identified than their more structural Medieval
counterparts, however ring wells are considered to be common from the Early
Historic period. Ring wells were formed from tubes of interlocking stone, terracotta
or burned brick rings (Hegewald, 2002:149; Pande, 1966:218). These structures can
be difficult to date independently; some scholars have linked ring diameter or height
to chronology, but there are also suggestions that variation can be related to
geographical differences (Pande, 1966:211). One ring well was identified in situ
south of Nagardhan fort (site 73), but had been partially destroyed during large-scale
digging. The ring well was constructed of large terracotta rings and most closely
corresponds to the first of four broad categories of wells identified by Pande, where
the bottoms of the upper rings fit into the mouths of the ones below (Pande,
1966:211). The large diameter of the ring well identified at Nagardhan makes it
similar to those identified at sites including Pauni and Paunar (Deo & Dhavalikar,
286
Chapter 8. The Economic Landscape
1968; Nath, 1998:17). Archaeological evidence indicates that ring wells were often
located in settlement areas, for example at Bhir and Besnagar the ring wells were
found within house remains, and at sites such as Nasik, Tripuri, and Nevasa they were
located near habitational structures (Pande, 1966:216). The ring well at Nagardhan is
situated towards the edge of the visible habitation mound in an area of high brick and
pottery remains which suggests the settlement extended across this area. The presence
of this buried ring well suggests that further wells may be present underneath the
modern village. It appears that there is greater potential for Early Historic water
management structures to be present across the Rāmṭek landscape but they may be
harder to identify due to visibility or difficult to securely date to the ancient period.
Figure 8-10. Remains of a ring well in a large area of digging south of Nagardhan fort (site 73).
When viewed together, the archaeological evidence pertaining to the general Early
Historic period indicates an increase in settlement sites and ceramics, which supports
the hypothesis that at some point the area became engaged with wider networks,
pottery production and use of brick. Furthermore, Nagardhan appears to represent the
development of a stand-out urban site in comparison to the agricultural plain. Despite
overall transformations in the nature of the rural landscape, settled life and the
economy and the presence of permanent traces of economic investment across the
landscape, Eastern Vākāṭaka economic investments appear to have been fairly
287
Chapter 8. The Economic Landscape
localised. It may be argued that beyond the redirection of resources there was little
royal investment in the rural communities themselves.
Early Medieval ceramics are not well established and the material is often absent in
excavation reports, but rough red ware, micaceous red ware and black slipped ware
are reported to relate to the Late Vākāṭaka to Mahavihara period at Mansar (Joshi &
Sharma, 2005:25). If these ceramics are a reliable sample of cultural material from
the Mahavihara period then the first two wares demonstrate similarities to the Early
Historic ceramics discussed here. Previous research has stated that there are no
“standard ceramic markers” for the Early Medieval period in the Deccan and there
may be a similar situation in this region. Without more detailed descriptions of the
fabrics and types, it is not possible to identify how the ceramics might differ across
the Early Historic to Early Medieval periods and a fundamental lack of understanding
of ceramic change does limit our ability to establish the length of occupation at sites.
The survey evidence could be interpreted as tentatively suggesting the Early Historic
sites were fairly short-lived and fell out of use sometime following Phase III and prior
to the Later Medieval period (Phase V) when there was generally a shift in site
288
Chapter 8. The Economic Landscape
Phase IVB in the project chronology relates to the period of Yādava influence and
activity in the area towards the end of the Early Medieval period, which is primarily
evidenced by the extensive constructions on and around Rāmṭek hill. Within the
working relative chronology it has been tentatively theorised that the Red Burnished
wares may be representative of Early Medieval or Medieval activity. The Red
Burnished wares were typically obtained from shallow cuts as opposed to deep
sections or surface scatter, placing them stratigraphically in sections above Red and
Micaceous wares but below Grey/Buff wares. Red burnished wares are most typically
found at major sites with later remains, such as Nagardhan and Rāmṭek and the
relatively low quantity of this ware may be due to the remains being buried under
modern occupation. The sites with larger quantities of Red Burnished ware tend to
have a high occurrence of Grey/Buff wares suggesting contemporary usage and a
later phasing for the Red Burnished ceramics, or continuity of settlement from some
point in the Early Medieval phase.
Table 8-8. The sites with the largest quantities of Red Burnished ware in the survey area.
289
Chapter 8. The Economic Landscape
Figure 8-11. Map of sites across the survey area with the presence of Red Burnished wares
The Red Burnished ceramics may tentatively indicate that contrary to the historical
narrative, the landscape was not completely abandoned post-Vākāṭaka. Several sites
such as Bhijewada (210) display a range of wares, which may demonstrate ongoing
occupation from the Early Historic through to the Medieval period. At sites such as
Bhijewada, continuation of settlement may be more likely than repeated periods of
abandonment and re-settlement, particularly as there appears to a change in settlement
pattern before or during the Medieval period. When comparing sites with Red
Burnished ware to those with micaceous wares, there seems to be a reduction in the
number of sites with significant quantities of ceramics. Several of the mounds to the
south of Rāmṭek hill continue to be represented, such as sites 210 and 276, but others
such as 197, 289, 236 and 318 often show a reduction in ceramics. The co-occurrence
of micaceous and Red Burnished wares at several sites may be suggestive of the
290
Chapter 8. The Economic Landscape
During the Early Medieval period, water management strategies advanced and
structures often became more architecturally elaborate with more recognisable
conventions, such as the presence of stepped tanks close to temples (See Japala (site
393) where the temple is connected to a stepped reservoir) (Dandawate, et al.,
2006:3). Bāvlīs and wells are particularly common across villages and agricultural
land in contrast to the limited or undetectable evidence for Early Historic economic
structures outside of the political and religious centres. The nature of water provision
may have changed towards the Medieval period so wells became more common than
tanks, which cannot be easily dated. The wells identified are typically stone
constructions or brick wells with a stone facing, which in some circumstances appear
to feature re-use of large ancient bricks. The structural stone wells correspond to a
relatively late type of simple Medieval wells which feature a “parapet” wall and
pulley apparatus for drawing water (Mehta, 1979:32). The stone wells are primarily
291
Chapter 8. The Economic Landscape
Figure 8-12. Site 307: Stone well in Bori village with octagonal casing and hooped stones on a parapet
for a pulley system.
Figure 8-13. Site 162: Round brick and stone well with a parapet, located in agricultural fields.
292
Chapter 8. The Economic Landscape
Figure 8-14. Site 407: Loose stones near to an old stone well adjacent to a Hanuman temple in
Chorkumari village, which show the hole used for a pulley system to fetch water.
Several other categories of later structural wells were identified in the survey area but
tended to be restricted to the major towns. Bāvlīs are underground wells, which
feature steps leading down into a water storage area and subterranean archways.
Bāvlīs may be constructed of stone or brick and mostly originate from the Medieval
period (Sutcliffe, et al., 2011:784). During survey, bāvlīs were identified in the larger
centres of Rāmṭek and Nagardhan and also to the east of Ambālā Lake, where there is
a group of Medieval structures and several bāvlīs with systems of underground
archways. In Rāmṭek, the temple complex at site 119 featured an excellent example
of a bāvlī but this has been filled in during renovations since the first fieldwork
season. Step wells are another category of well, which are not as common across the
Rāmṭek landscape but can be found in Rāmṭek town. Step wells appear to be
indigenous to India but continued to be developed under Muslim rule around the 14th
and 15th centuries onwards (Hegewald, 2002:156). The recorded distribution of wells
identified during survey does not seem to reveal any noticeable pattern and
topographic maps of the area indicate the presence of further wells not recorded on
survey. The large number of wells pertaining to the later period reflects on greater
levels of economic investment in the rural landscape.
293
Chapter 8. The Economic Landscape
Figure 8-15. View of brick and stone bāvlī to the east of Ambālā tank. In the left hand side of the
picture the long, underground arched bāvlī terminates in a round brick and stone well with a parapet.
Figure 8-16. The bāvlī in a temple complex in Rāmṭek town (site 119), which has since been filled in
and concreted over.
294
Chapter 8. The Economic Landscape
Figure 8-17. Example earthwork in the Rāmṭek landscape. This appears to be an old tank embankment
at Manapur village, with a modern shrine situated on the top (site 227).
295
Chapter 8. The Economic Landscape
Table 8-9. Sites with the largest quantities of Grey/Buff wares in the Rāmṭek assemblage
Figure 8-18. Map of sites across the survey area displaying the presence of Grey/Buff wares.
296
Chapter 8. The Economic Landscape
The sites with the most significant numbers of micaceous and red wares tend not to
display Grey/Buff wares beyond transient surface scatter. These ancient habitation
sites are typically mounded areas located outside of modern villages, whereas the
majority of modern villages with Medieval architectural/ceramic remains are not
situated on a settlement mound. Obvious mounds at Medieval sites aren’t necessarily
common as Medieval settlements frequently expanded horizontally and were not as
likely to build on the remains of earlier settlements (Mate, 2005:85). A change in
settlement pattern is suggested both at abandoned mounds with limited other wares,
such as Devada (site 275) and by the presence of Grey/Buff wares as surface scatter
throughout late Medieval to Modern villages such as Bori, which do not display an
ancient habitation mound but contain Phase V Medieval architecture and wells.
Furthermore, a change in settlement pattern appears to be reflected in the spatial
distribution of sites, as sites with the highest quantities of Grey/Buff wares (outside of
the major towns of Rāmṭek and Nagardhan) are often situated north of Rāmṭek hill,
indicating an expansion of sites. Many of these sites display insignificant quantities of
earlier ceramics, which would suggest a lack of established settlement prior to the
Medieval period. Several smaller sites in the south, which have low numbers of Phase
IV ceramics, remain occupied; sites such as Udapur (276) yield fairly significant
amounts of Red Burnished wares and Grey/Buff wares which may indicate that there
was some continuation or expansion of Early Medieval sites into the Medieval period,
alongside the foundation of new sites.
This distribution of sites may indicate increased agricultural and commercial activity
in areas which were previously more difficult to cultivate and settle. The expansion of
settlement to the north may also be related to the development of Rāmṭek town.
Alongside the numerous late temples, the Grey/Buff wares at Rāmṭek suggest a later
expansion of the settlement prior to its modern development into a large town. These
remains may indicate growing activity around Rāmṭek and increasing importance as
an economic centre. However, it is impossible to ascertain the past extent of the town
and its settlement activity with survey alone, as the earlier remains could have been
concealed or destroyed. Similarly there may be more substantial remains obscured
beneath the current infrastructure at the surrounding modern villages with occasional
Grey/Buff surface sherds.
297
Chapter 8. The Economic Landscape
Prior to this survey, the excavations at Mansar proposed a long history of potential
settlement at the site. However the evidence is not available to be assessed and a
number of scholars have expressed doubts about the interpretations of the excavated
evidence. It was also unsure how this related to Rāmṭek, Nagardhan and the wider
hinterland. The accepted view of landscape development around Rāmṭek was largely
constructed from a historical perspective, which proposed that the Vākāṭakas moved
their state to a newly developing, peripheral region. It was thus proposed that the
period of Vākāṭaka rule resulted in major socio-economic, religious and political
developments in the landscape and transformed rural settlement. Evidence was
severely lacking for the periods outside of Eastern Vākāṭaka rule in the Rāmṭek
landscape and even within the recorded phases the nature of rural settlement was
largely unknown. After the decline of the Vākāṭakas, little was known historically
about the Early Medieval period prior to the establishment of an organised political
authority under the Yādavas.
Therefore, a primary research question of this thesis was to ascertain whether it was
possible to adequately explore Rāmṭek to address the lacunas in our understanding of
the area’s development and assess the extent of Vākāṭaka economic investment in the
landscape. The resulting overview of the landscape development is still broad, but
hypotheses can be made about the nature of rural settlement and the relationship of
the religious centre of Rāmṭek to its landscape. Furthermore, we can hypothesise
about whether royal economic investment was localised or expansive.
The picture that emerges is of an area which was clearly a population centre at several
points in its history, making Rāmṭek more than simply an Early Historic religious
site. Prehistoric activity may have been at a low level resulting in few discernible
archaeological traces, or may have been related to seasonal or transhumant
populations who had limited interaction with wider networks and left only transient
material remains. This scenario is not necessarily at odds with the historical view of
the Vākāṭakas moving into a sparsely inhabited, peripheral zone. There is a
significant difference in the level of activity from the prehistoric to Early Historic
periods, and the survey results go some way towards establishing the nature of Early
298
Chapter 8. The Economic Landscape
Historic rural settlement in the area. The evidence implies that the Early Historic
period saw changes in the landscape and the rural economy, but these can only be
attributed reliably to the overarching period rather than specifically to Vākāṭaka
influence. This narrative highlights the ambiguity of the available evidence and
demonstrates that the Vākāṭakas may not have been the sole driving force behind the
development of the landscape, despite a compelling case for much of the material
evidence to correspond to 5th century activity.
It may be proposed that the religious and urban centres drew on the prosperity of a
local rural landscape and agricultural economy. It appears that alongside the high
level of culture demonstrated by the sites of Mansar and Rāmṭek, Nagardhan
developed into a reasonably sized urban area. Economic development of the rural
hinterland however appears low. Outside of the three political/ritual centres, sites are
small and feature no visible evidence of construction. It seems that the economic
function of the wider landscape was to direct produce into state pursuits and ritual
activity.
Following the Early Historic period, the narrative of the landscape development sees
two lacunas in evidence, from the 6th to 11th centuries and from the 14th to 17th
centuries. Although it was hoped that a nuanced picture of activity in the intermediate
periods would be achieved from settlement evidence, the phasing of sites according to
ceramics and their possible duration of occupation remains equivocal. While it has
been shown that landscape survey can identify areas of past settlement activity,
without reference to reliable excavation data the nature of the evidence results in
provisional or conjectural conclusions. At some point in the Early Medieval or
Medieval period, the wider Rāmṭek landscape saw greater economic investment from
the centralised authority. Beyond the known Yādava temple on Rāmṭek hill, only
tentative evidence of settlement has been identified and the great change in the
landscape seems to have occurred later. The Medieval period witnessed a change in
settlement location and the construction of permanent water management resources
across the landscape.
Discussion of the survey results must be tempered with recognition that the
limitations associated with this type of survey may impact on the types of remains
identified. There is clear potential for further research as a detailed understanding of
299
Chapter 8. The Economic Landscape
the landscape is elusive, but it has been possible to address general trends of
development which may be related to wider historical considerations. It is clearly
necessary to employ a combination of survey and excavation to provide stratified
sequences. Before a deeper sense of the relationships between remains and nature of
Early Historic and Early Medieval diagnostic artefacts can be achieved, this
foundation data must be established.
300
Chapter 9. The Ritual Landscape
9.1. Introduction
Sculptural remains and standing architecture form a large part of the dataset achieved
during the Rāmṭek survey. Religious architecture was the most commonly identified
site type and includes temples and village/field platform shrines (cabūtrā). Study of
sculpture and religious architecture can aid understanding of the development of
religion in the survey area and suggests the religious affiliation of the population,
both past and present. Additionally, the past religious landscape can be linked to the
area’s political and cultural history based on the role of religious practice, patronage
and ideology in political authority. Consideration of the religious evidence may
indicate the effect that ritual and ideology had on the development of settlement and
the landscape. This chapter investigates the relationship between the religious centre
of the Eastern Vākāṭakas with the surrounding landscape and the engagement of the
surrounding population through ritual means, following on from the economic focus
of the previous chapter.
While there are numerous sculptural/architectural remains around Rāmṭek, the vast
majority appear to reflect Late Medieval, Early Modern or Modern activity, and
therefore relate to a known period of occupation in the survey area. Despite this,
study of the local architecture and sculpture may reveal insights into the level of
activity across earlier periods, keeping in mind the potential effect of destruction and
modern development on the visibility of earlier sites. Sculptural and architectural
remains are considered particularly important in attempts to identify activity in the
Early Medieval period following the collapse of Vākāṭaka rule. There are numerous
issues with Early Medieval archaeological remains as noted in the previous chapter;
ceramics and cultural material are generally found to be lacking in archaeological
investigations and the diagnostic cultural material or the amount of change or
continuity in material culture is not well established. However, architecture and
sculpture may give important clues to some level of activity in the Early Medieval
landscape, which otherwise may not be visible in the archaeological record.
301
Chapter 9. The Ritual Landscape
In order to evaluate the relative Eastern Vākāṭaka political and religious investment in
the landscape, an objective of the Rāmṭek survey was to identify religious activity
outside of the Hindu (Brahmanical) mainstream, which was prevalent from the time
of the Vākāṭakas through to the modern period. By investigating different religious
affiliations, it may be possible to discern whether activity in the landscape increased
with the arrival of the Hindu Vākāṭakas, and how closely religion was related to the
political-economic development of the area. Hinduism refers mainly to the two major
groups of Śaivism and Vaiṣṇavism, which have Śiva and Viṣnu as the main focus of
worship respectively, but in reality it may be considered to be a “heterogeneous
amalgam of beliefs and practices” which differs according to geographical and
chronological factors (Lahiri & Bacus, 2004; Sugandhi & Morrison, 2011:921). The
vast majority of temple and sculptural remains around Rāmṭek relate to the dominant
‘Hindu’ or ‘orthodox’ Brahmanical tradition, with evidence of both Śaiva and
Vaiṣṇava affiliation. However, despite the domination of Brahmanical remains, the
religious landscape is complex and there is a variety of evidence relating to the
persistence of local shrines and devotions.
In the past, reference has traditionally been made to the division between ‘Great or
Major’ and ‘Little’ (village/folk-based) traditions (Redfield, 1956; Redfield, 1995)
The ‘Great’ tradition was related to the concept of ‘Sanskritic Hinduism’, after
Srinivas (1952), and is that of the elite and literate population. The ‘Little’ tradition
refers to that of the village communities and is considered to include local forms of
Hinduism (Redfield, 1956:72). The term ‘Sanskritisation’ was coined to describe a
process wherein the ‘Great’ tradition spreads to the ‘Little’ tradition due to increasing
influence (Staal, 1963:263). However, this relationship is shown to be complex as
over time mainstream Brahmanism appears to have appropriated local deities
(Sugandhi & Morrison, 2011:923). Furthermore, scholars have suggested that often
the ‘Great’ tradition has roots in ‘little’ traditions, so rather than a process of
assimilation or displacement, traditions co-exist through the exchange of features
(Eschmann, 1978:82; Singh, 1993:287; Staal, 1963:269). Moving away from this
division recognises that folk traditions were coeval with Brahmanism and it is
important to recognise that the landscape can reveal ritual heterarchies and
temporalities of worship and investment in religious traditions.
302
Chapter 9. The Ritual Landscape
The dating of temples and shrines can be problematic as these structures often have
long histories of use, and may represent different phases of construction and
modification. Nearly all of the temples identified in the survey area are still active
religious sites involved in daily worship and modern superstructures or additions
potentially obscure original features or constructions. Temples are often modified by
the infilling of archways or plastering and whitewashing during renovations. None of
the temples newly recorded on the survey, which are primarily small village
structures, have inscriptions to aid with dating. Without independent evidence for
dating, structures can only be assigned broad phasing based on diagnostic
architectural features, which can be compared to art historical typologies established
by specialist scholars (Hardy, 1995; Hardy, 2007). Broad art historical dating can
assign date ranges which allow the reconstruction of a sequence of architectural
development, which may give an indication of the level of activity across major
periods. However, the evolution of architectural style does not necessarily account for
long periods of use for structures and even their re-affiliation for worship by different
groups of people (Shaw, 2007:75). Temples are usually classified into phases broadly
definable in terms of dynastic associations, based on the general assumption that
architectural styles developed and changed over time resulting in characteristic traits
which relate to specific periods of rule, for example Vākāṭaka (3-5th century AD),
Cālukya (6-8th?), Yādava (12th to 14th century AD) and Mārāṭha or later (18th century
onwards). There are difficulties with such schemes, as common architectural motifs
and styles may have remained popular across different periods and dynastic
periodisation can create gaps in the known history of the landscape. Unfortunately,
the nature of the architectural evidence and the non-specialist approach of this
researcher do not permit more nuanced dating, but the broad overview of religious
practice is still useful.
303
Chapter 9. The Ritual Landscape
pieces are again usually broadly referred to in terms of major dynasties but there can
be distinct local variations and traditions under regional powers (Shaw, 2004:8). The
Gupta dynasty produced a characteristic style of sculptural art and a sophisticated
iconographic tradition (Harle, 1974:8). The Vākāṭaka style on the periphery of the
Gupta Empire shares traits and decorative features with the Gupta tradition (Harle,
1974:24). Sculptures identified at Mandhal, which were locally produced, reflect the
character of Gupta art and display analogous depictions of physical characteristics,
clothing and ornamentation (Jamkhedkar, 1991a:87). Despite mutual aesthetic
principles and iconography, Vākāṭaka art demonstrates regional originalities which
do not have clear Gupta parallels (Bajpai, 1989:103). The excavated Mansar
sculptures embody the spirit of the Eastern Vākāṭaka sculptural tradition and are said
to be defined by their “refined elegance and expressiveness” (Deva & Sharma,
2009:21). A large amount of sculpture was identified during survey, but the vast
majority appeared to be late Medieval or modern (see appendix nine for catalogue).
There are difficulties dating the potentially older sculptures due to their condition as
they are often badly preserved and ongoing worship can result in features being
obscured by layers of sindūr or worn by regular liquid offerings (Blurton, 1992:72).
Here, as in the previous chapter, three overall phases are considered focused around
the 5th century investments by the Eastern Vākāṭaka dynasty. The survey looked for
evidence for pre-existing ritual development in the landscape followed by evidence of
a transformation or formalisation of ritual activity with the arrival of the Vākāṭakas,
in order to infer about the nature of their authority and engagement with the
landscape. Finally, a comparison is made to the post-Vākāṭaka period in order to
establish the relative levels of investment and again reflect on the character of the
Eastern Vākāṭaka state.
304
Chapter 9. The Ritual Landscape
Vākāṭaka Buddhist remains dated to the Mauryan and Sātavāhana periods have been
claimed at Mansar (Joshi & Sharma, 2005:7). The reports of the excavation go as far
as describing Mansar as “virtually the capital or sub-capital” for the Sātavāhanas in
the area (Joshi & Sharma, 2005:8).
305
Chapter 9. The Ritual Landscape
prior to Brahmanical expansion. This is the case in the surveys at Sanchi (Shaw,
2007), Badoh-Patari (Casile, 2009) and Banavasi (Suvrathan, 2013). Similarly, no
trend of Buddhist sites was identified across the plain in the survey zone. Ancient and
abandoned stupas tend to be identifiable by their regular, circular mounds containing
evidence of a solid brick construction. The mound encountered at site 242
(Beldongri) was initially considered to resemble an ancient stupa, but on closer
inspection appears perhaps to be a later secular or religious construction. One further
mound, site 237, possibly resembles a stupa, but insubstantial evidence is detectable
to support or refute this identification. While the site has a high mounded profile, it is
entirely overgrown and it is currently not possible to tell whether this is an ancient
brick structure or whether it served a different purpose.
Figure 9-1. From left a) 2014 GoogleEarth imagery of Beldongri mound (site 242) and b) 2013
GoogleEarth imagery of the second mound at Dudhala to the west of the main mound which is
partially cut away.
area did not come under a significant, unified political authority until the
establishment of Eastern Vākāṭaka rule. An absence of any clear major sites in this
phase is also in keeping with the historical view of the period which dictates that
major centres tended to be situated along rivers rather than in inland, ‘peripheral’
locations (Nath, 1998).
The earliest known ritual structures in the survey area that are still visible today are
the Vākāṭaka constructions on Rāmṭek and Mansar hills, which demonstrate a change
in worship in the landscape from some point around the 5th century AD. This is seen
in the expansion of Śaivism and Vaiṣṇavism alongside more formalised support of
local cults, which attests to the multi-religious investments of the Eastern Vākāṭakas.
Only relevant comments will be made here regarding the structures at Mansar based
on observations made during fieldwork, and there is scope for further evaluation by a
specialist in ancient architecture. The discussion here focuses on the evidence for
307
Chapter 9. The Ritual Landscape
Figure 9-2. The inner and outer phases of the pilaster mouldings on MNS II.
To investigate the chronological situation further, bricks were measured from the
exterior walls of MNS II and III and compared to those used in the interior of the
brick constructions, which was accessible due to heavy rain damage in 2014. Despite
the supposition that the interior of MNS II should represent earlier phases of building,
no significant difference in brick size was found and there does not appear to be any
meaningful difference in quality between the building materials. Bricks identified as
being from different chronological phases overlap in size and, as can be seen in table
9.1, the ‘Vākāṭaka’ bricks are comparable in size to those considered to have
originated from the earlier ‘Maurya-Śuṅga’ period. The bricks measured on survey
(table 9.2) demonstrate that associated bricks in situ can vary substantially in size,
308
Chapter 9. The Ritual Landscape
Figure 9-3. Damage caused by rains to the southern side of MNS II. Inner courses of brick have been
revealed across this structure and at MNS III.
Table 9-1. Brick dimensions referred to in reports of the Mansar excavations (Joshi & Sharma, 2000;
2005).
309
Chapter 9. The Ritual Landscape
Table 9-2. Dimensions of bricks from MNS II (the ‘Palace’ site) at Mansar measured during survey.
All pieces are proposed to be Vākāṭaka bricks and demonstrate the range of sizes in bricks associated
with Vākāṭaka structures. It is noted that some of the complete bricks are comparable to brick sizes
noted in Table 8.1 which have been assigned different dates by the excavators.
The Vākāṭaka period architectural pieces at Mansar and Rāmṭek display a preference
for red sandstone, and have detailed carving, frequently with floral motifs. At
Mansar, the architectural pieces largely lie around MNS III, and it would appear that
they were used to decorate the brick temple construction. The architectural fragments
show striking similarities to the 5th century pieces on the Rāmṭek temples. While
Vākāṭaka period architectural fragments can be found around Rāmṭek and Mansar
hill, no similar pieces were identified in the surrounding village areas during survey.
Therefore, aside from the previously published Vākāṭaka constructions, few early
remains were conclusively identified. The majority of architectural pieces located
across the survey area were carved from yellow sandstone, and can be loosely
categorised as ‘Post-Vākāṭaka’.
310
Chapter 9. The Ritual Landscape
Figure 9-4. Red Sandstone architectural fragments on Mansar hill associated with the stone and brick
monumental constructions.
Outside of the known sites, the absence of extant architecture pertaining to this early
period of Vākāṭaka construction could be due to destruction of remains related to
modern activities or survival of the building materials. Destruction of brick structures
has been witnessed elsewhere; for example the Vākāṭaka period temples at Mandhal
have been completely obliterated by brick re-use and only scatters were visible during
the field visit in 2012. Brick remains were recovered across the survey area, usually
related to ceramic evidence, and may indicate the presence of earlier structures,
however none related to standing monuments. Several mounds were identified which
would appear to represent ancient structures as opposed to ancient habitation sites due
to large amounts of brick and comparatively scarce ceramic evidence (for example,
site 78 to the west of Nagardhan). These fairly small sites may be small subsidiary
temples to those at Rāmṭek and Mansar, or the remains of secular structures such as
outposts, possibly related to the urban site at Nagardhan. As there is no clearly
associated evidence, it is impossible to assign a religious or secular purpose without
excavation.
The mound at Beldongri, site 242, displays substantial remains indicating a buried in
situ brick structure. As mentioned, on first appearances the mound resembles an
ancient stupa, but through ground inspection the mound shape does not appear
consistent with the expected distinctive stupa profile. The mound displays two peaks
with a flat area in between, rather than a single round profile. Furthermore, at around
65m in diameter, any stupa would be quite large. There is little tile evidence, which is
frequently associated with stupa roofing, and a modern cut in the base of the mound
has revealed courses of brick, which appear to form a corner. This may be part of a
311
Chapter 9. The Ritual Landscape
platform, but it seems more likely that it is part of an angular building. The brick
dimensions fall within the range of those typically considered ‘Gupta’-style (Mishra,
1997) and are consistent in size with those from Mansar and Nagardhan, indicating a
contemporary brick building probably unrelated to Buddhism.
Figure 9-5. Beldongri mound (site 242) which is a brick construction with large sized bricks
comparable to those used in other Gupta period constructions.
Figure 9-6. Section on south west side of Beldongri mound. Courses of in situ brick has been reveals
indicated an angular construction.
312
Chapter 9. The Ritual Landscape
Figure 9-7. a) Eroded torso of sculpture number 211, which resembles the shape of the large
Narasiṃha on Rāmṭek hill; b) Kevala-Narasiṃha sculpture from Rāmṭek hill for comparison (photo
credit: Prof Hans Bakker)
An absence of extant religious architecture may suggest that the Rāmṭek religious
landscape was dominated by the monumental sites at Rāmṭek and Mansar; major
temple complexes may have had administrative and religious authority across the
immediate landscape, and so smaller local temples may not have been needed to
provide an administrative function. It may be that the Vākāṭaka ritual presence was
limited to a few key locales and that the wider landscape received lighter investment.
Alternatively the 10km survey area may have been too concentrated on the major
religious site of Rāmṭek to show trends of smaller temples, and a wider survey could
reveal a greater pattern of dispersed Brahmanical religious sites. Smaller village
313
Chapter 9. The Ritual Landscape
traditions would still be expected despite the potential domination of the Rāmṭek and
Mansar temples. However, rural temples may not have survived as brick re-use is
evident in Nagardhan village. Smaller temples related to a village level of religious
activity in the landscape may have been constructed from less permanent materials
such as wood, which may have been related to restricted usage of brick or stone in
particular (Williams, 1983:226).
It would not have been unusual for “religious pluralism” to have been practiced in the
Gupta-Vākāṭaka period, given that the development of regional dynasties had
diversified the religious landscape and an “assimilationist approach” would be
appropriate to encourage allegiance to the kings and state (Narain, 1983:48). The
Guptas demonstrated this in their political system where royal authority was
legitimised through the development of divine kingship, the use of Vedic ritual and
liberality towards religion, despite the dynastic commitment to Brahmanism (Narain,
1983:34). This is reflected in both the Western Vākāṭakas’ popular patronage of
314
Chapter 9. The Ritual Landscape
Buddhism and the Eastern Vākāṭakas’ apparent adoption of locally important deities -
Varāha and Narasiṃha - into the Vaiṣṇava pantheon at Rāmṭek (Bakker, 1992a;
Bakker, 1992b:14).
The Vākāṭaka worship of Vaiṣṇavism has often been connected to Gupta influence,
but it has also been suggested to reflect local development of Vaiṣṇava groups in
Vidarbha; for example Kapur has suggested that the Vākāṭaka rulers required
religious legitimation at the time when sects such as the Sāttvatis wanted royal
patronage (Kapur, 2006:19). Similarly, the shift to Śaivism under Pravarasena II and
the development at Mansar may reflect the growing popularity of the Śiva cult in the
region (Kapur, 2006:28). Due to the sculptural finds at Mansar, Bakker has suggested
that Śiva worship among the Vākāṭaka elite would have been iconic to emphasise
their status and aniconic liṅga worship would appeal more to rural communities
(Kapur, 2006:31).
The most secure iconographical dating of sculptures identified during survey suggests
that the earliest relate to the period of Vākāṭaka control given their similarities to
sculpture from major Vākāṭaka sites such as Mansar, where the sculptures have been
assigned a date from the 5th century (c. 450 AD). Again, it is significant that no
sculptural evidence was conclusively identified from non-Brahmanical traditions in
the survey area, supporting the previous assumption that the influence of
institutionalised religions was only felt from the establishment of Brahmanical
worship under the Eastern Vākāṭakas. The discovery of three Buddhist bronzes near
Nandivardhana led to suggestions that there was a Buddhist presence in the area,
contemporary to the Hindu temples (Meister, et al., 1988:61). It is possible that there
was Buddhist activity, but there is a striking absence of any architectural fragments in
the remains recorded, unlike at Pauni where decorative railing related to an Early
Historic stupa is clearly visible. The lack of Buddhist architectural evidence around
Rāmṭek may indicate that there was not a large established site within the survey
radius.
The aniconic form of Śiva worship in the form of the liṅga is known from the Early
Historic period and remains a common form of worship in modern India (Blurton,
1992:78). The Vākāṭaka period liṅgas at the Mansar temples are carved from
sandstone and have a simple round shaft on a typically square base. They display the
315
Chapter 9. The Ritual Landscape
markings of pārśvasūtra and brahmasūtra which appear around the 5th century AD
(Mitterwallner, 1984). Bakker has compared the Mansar liṅgas to the style of ‘Type
E’ (examples from early 6th century AD at Deogarh) in the typology by
Mitterwallner, which display two vertical lines in the brahmasūtra enclosed by a
pointed pārśvasūtra (Bakker, 2002:13; Mitterwallner, 1984:18). There are also some
similarities with ‘Type D’ (as at Nachna-Kutara, post-Gupta), where the two vertical
incisions are close together, similar to those seen in sculpture 40 (figure. 8.34 below).
The apparently octagonal base of this liṅga is similar to that seen in a 5th century AD
example from Uttar Pradesh in Mitterwallner’s publication, which displays the
incised pārśvasūtra and brahmasūtra and a lower shaft carved into a octagon which
would be embedded in the pīṭha (1984:21). Two 5th century liṅgas were recorded at
Hamlapuri near the Vākāṭaka Ganesh sculpture. These appear to have the pārśvasūtra
and brahmasūtra, and square bases, seen on those at Mansar, and are comparable in
size and shape.
Figure 9-8. Sculpture no. 40 and 43: 5th century liṅgas with pārśvasūtra and brahmasūtra at Mansar.
Across the survey area, other liṅgas were found which may be early icons; a small
sandstone liṅga in Nagardhan appears to be particularly old and reminiscent of the
smaller liṅgas at Mansar. A liṅga found in Nagardhan fort (sculpture no. 10) is
carved from basalt and displays a square base. This may be from the 5 th century or
slightly later and has some similarities to the liṅga at site 202 in Hivri (sculpture no.
88), which is associated with a settlement mound, ceramic and brick remains. Two
316
Chapter 9. The Ritual Landscape
larger liṅgas, also carved from Basalt, were found near to Mansar (sculptures 41 and
42). These both feature quite short round shafts on square bases, but have a slightly
irregular shape which bulges towards the end. There are no discernible markings to
date these liṅgas. All of these examples are bipartite in construction with the abstract
representation of the main shaft and head of the liṅga and a square base which would
be embedded in the ground. Finally, a liṅga was identified in Nagardhan town, which
appears to be exceedingly large (sculpture no. 117); the liṅga is mostly buried in the
ancient mound but has a circumference of 160cm and is reminiscent of the large liṅga
from the top of the Śiva temple at Ahicchatra, dated to the early 5th century AD
(Mitterwallner, 1984:21).
Figure 9-9: Left to right: Sculpture no. 668: Small sandstone liṅga with no discernible markings in
Nagardhan town. Sculpture no. 88: Liṅga in Hivri village associated with brick and ceramic remains
(site 202)
Both Varāha and Narasiṃha seem to have been significant focuses for worship
around Rāmṭek and both have been suggested to have entered later Hinduism through
tribal/local cultic influence (Blurton, 1992:19). Gupta period sculpture in central
India suggests that Varāha was a common god under worship at this time (Blurton,
1992:122). The cult of Narasiṃha has been suggested to have developed in central
India under Vākāṭaka patronage (Welankar, 2009:127). The Narasiṃha cult has been
connected to tribal worship in the Deccan and its current form appears to be the result
of an “amalgamation” of Vaiṣṇava and local religions (Blurton, 1992:125).
317
Chapter 9. The Ritual Landscape
Rāmṭek appears to have been a centre of worship for Narasiṃha given the huge
sculptures on the hill. The Rāmṭek Narasiṃhas have a distinctive “placatory” pose
(Welankar, 2009:113) and the demon Hiranyakasipu, whom Narasiṃha kills in many
sculptures, is absent leading to the modern iconographic assignment of
Kevala, meaning ‘only’ or ‘alone’ (Bakker, pers. comm.). Around 9 replicas of the
Rāmṭek Narasiṃhas have been found at other sites in the area, which supports the
suggestion that the cult was strong during the 5th century AD (Welankar, 2009:123).
A small Narasiṃha image (sculpture no. 25) was found at Koteshwar temple in
Nagardhan during survey, and it closely resembles the large fifth-century Narasiṃha
sculptures at Rāmṭek, with all three being seated in mahārājalīlāsana and holding a
cakra in the right hand (Bakker, 1990:66). The miniature Narasiṃha parallels those
found previously at Rāmṭek, Mansar and Nagardhan in both size and form, which all
reflect the stance and features of the large Rāmṭek images, and have been dated to the
Vākāṭaka period. The miniature measures 25 by 22 cm, and appears to be carved of
sandstone but it has been painted with sindūr. These replica images are theorised to
have been produced regionally for devotees to use in domestic alters (Bakker,
1997:136) As already noted, sculpture number 211 at Beldongri is also theorised to
represent a replica of the large images on Rāmṭek hill. However, in this case the
sculpture is not a miniature depiction (the remaining section of the torso is 43cm in
height) and so may have been the main icon of a subsidiary temple.
Table 9-3. Miniature Narasiṃhas known from the survey area in publications.
318
Chapter 9. The Ritual Landscape
Figure 9-10. Sculpture no. 25: Small Narasiṃha at Koteshwar temple in Nagardhan.
319
Chapter 9. The Ritual Landscape
Figure 9-11. Small Narasiṃha at Nagpur Museum. This sculpture was discovered at Rāmṭek and is
dated to the 5th century AD
320
Chapter 9. The Ritual Landscape
Figure 9-12. (a) Sculpture no. 21: Ganesh sculpture at Hamlapuri, published in Bakker (1997). (b)
Sculpture no. 23: possible 5th century Liṅga found in the modern temple alongside the image of
Ganesh
321
Chapter 9. The Ritual Landscape
Figure 9-13. Sculpture no. 534: Fragmentary sculpture of a boy’s head, possibly originating from the
5th Century AD.
Figure 9-14. Sculpture no. 295: Fragmentary sculpted head with elaborate curled hairstyle.
Figure 9-15. Sculpture no. 294: Sculpture of a head with matted hair at Rāmṭek
322
Chapter 9. The Ritual Landscape
Figure 9-16. (a) Sculpture no. 17: Gaṇa-type figure at Koteshwar temple in Nagardhan. (b) 5 th century
corpulent Gaṇa type figure at Nagpur Museum.
century BC until 11th century AD (Elgood, 2004:335). Lajjā Gaurī images have been
reported from across India but appear to have been particularly significant throughout
the central region (Ray & Ravindranath, 2007). Lajjā Gaurī images have so far been
identified as being particularly concentrated around Maharashtra with approximately
42 recorded plaques and further fragmentary pieces found in this region out of a total
of around 109 known and documented images (Jamkhedkar, 2004:28). Hunter
published the first example from Mahurjhari in 1933 and since there have been finds
at a number of sites including Bhokardhan, Ter, Sanchi and Nagardhan (Janssen,
1993:464). The greatest numbers have been identified around Nagpur where this
survey was focused (Bolon, 1992:5).
Lajjā Gaurī has a wide array of variations but is recognisable by her supine position
revealing her genitals, and she is pictured typically as a headless figure (Bolon,
1992:5). Although scholars in the past have interpreted the goddess to be squatting,
Lajjā Gaurī is customarily thought to be depicted as lying with her legs drawn up and
the feet turned outwards, a pose which is described iconographically as the uttānapad
position (Janssen, 1993:461; Kramrisch, 1956:259). Despite her popularity, there is
little textual or epigraphic evidence to help decipher the role of Lajjā Gaurī beyond
her apparent symbolism of fertility (Willis, 2008:5). Lajjā Gaurī figures appear to be
considered “auspicious”, rather than being connected to a form of eroticism (Bolon,
1992:6). The characteristic uttānapad pose has been connected to concepts of divine
or human creation and variously interpreted as representing “sexual receptivity” or
the act of giving birth (Bolon, 1992:6; Janssen, 1993:459; Mohanty & Deo, 2004:32).
Scholars have attempted to date Lajjā Gaurī images through a typology of “major
regional/period styles” but the simple forms with local features can be difficult to
characterise (Bolon, 1992:12). Jamkhedkar has dated all of the Maharashtra images to
between the 3rd to 6th centuries AD (Jamkhedkar, 2004:28). Bolon identified a four-
fold classification scheme developing from the most simplistic and “nearly aniconic”
representations to a fully human figure and these four forms appear to demonstrate
regional and chronological differences (Bolon, 1992:3):
324
Chapter 9. The Ritual Landscape
Form III has the addition of arms on a complete torso, usually with a lotus
bud in each hand
Form IV is described as fully anthropomorphic (Bolon, 1992:11).
Within the survey area, Lajjā Gaurī plaques were previously known from Hamlapuri
and Mansar (Gupta, 1987:46). Janssen has also reported five Lajjā Gaurī plaques
from Nagardhan, which are kept in Nagpur museum (Janssen, 1993:464). Wellsted’s
collection in the British museum includes a small stone Lajjā Gaurī plaque from
Mansar, measuring 6.6 x 7.6 cm and probably dating to the 4th to 5th century AD
(Willis, 2008:4). During this survey, two Lajjā Gaurī plaques were identified at
Hamlapuri and it is likely that these are the ones already recorded by Gupta. Both
examples are small and simple images most similar to Type II Lajjā Gaurīs, where the
arms are absent and sculpture does not feature a head above the breasts. These images
are similar in size and design to a number of Lajjā Gaurī plaques found in the region
including M46 from the Mansar excavations (Deva & Sharma, 2009:57) and an
image from Ter (Sankalia, 1960:115). Bolon has published similar examples
including a 4th to 6th century AD piece from Alapur, Aurangabad district (Fig. 22, 4 x
4 inches), a figure from Paunar in Wardha district (Fig. 25, 4¼ x 3 inches) and a 4th
to 5th century example from Mansar (Fig. 93, 4 x 3¾ inches) (Bolon, 1992). A Lajjā
Gaurī plaque previously recorded from Nagardhan (15 x 16cm) displays anklets
which are similar to those found here (Jamkhedkar, 2004). The Hamlapuri Lajjā
Gaurīs appear to relate to Phase III, Vākāṭaka period Early Historic, in the survey
chronology.
325
Chapter 9. The Ritual Landscape
Figure 9-17. Lajjā Gaurī figures identified at Hamlapuri: a) Sculpture no. 69. b) Sculpture no. 70, the
second Lajjā Gaurī figure, which has been placed on the cabūtrā upside down.
The origin, development and iconography of Lajjā Gaurī is not well understood,
although the Goddess has been considered to be a folk/village deity related to
agriculturalists (Sonawane, 1988:28). It is often suggested the ‘folk’ deity of Lajjā
Gaurī, who was assimilated into the Brahmanical religion (Bolon, 1992:48). This
would not have been unusual as various forms of mother goddesses may have
originated within local worship before being incorporated into mainstream Hinduism
(Mishra, 2004:56). This is considered to have been a socio-political process, which
attempted to legitimise royal authority and integrate local communities through either
the patronage of ‘tribal’ deities or the absorption or acceptance of local deities into
the Hindu temple (Bolon, 1992:69). There is a disparity between the typically
aniconic focus of ‘tribal’ traditions and the use of mūrtis or anthropomorphic
depictions of the gods in Brahmanical temples and it has been suggested that Lajjā
Gaurī was ‘anthropomorphised’ when she was brought under Brahmanical auspices
(Bolon, 1992:68).
However, it has also been suggested that Lajjā Gaurī was not necessarily a “minor
deity” as she appears to have been widespread, particularly in central India, and
images are fairly numerous at Early Historic sites (Willis, 2008:5). Lajjā Gaurī is
often found to be associated with Śaivite imagery such as the bull and liṅga, and in
situ plaques have been located in Śaivite temples. Evidence suggests that by about the
4th century AD Lajjā Gaurī belonged to the Śaiva Hindu pantheon (Bolon, 1992:48).
Alongside suggestions of Śaivite association, are tentative indications that a lion head
depicted on some of the recorded Lajjā Gaurī plaques may be representative of
326
Chapter 9. The Ritual Landscape
A small, yellow sandstone Viṣṇupada has been published from Nagardhan and has
been related to the possible worship of Viṣṇu’s footprints on Rāmṭek hill. This plaque
displays a pair of footprints in relief with engravings of possible railings, a gada, and
a śaṅkha (conch). In the centre an “enigmatic symbol” is engraved, which resembles
a tree (Bakker, 1997:136). A number of Viṣnupadas were located during survey,
which may have significance as indicators of a local tradition of worship. However,
these do not resemble the recorded plaque at Nagardhan. Although carved in
sandstone, they simply feature a pair of footprints and a carved border. Their phasing
is unknown as all examples are either badly eroded through ongoing libations or are
very crudely carved. Examples were located at Rāmṭek and Kawadak just to the north
of Rāmṭek town; Manapur, Mandri and Khairi, which are located in close proximity
to Rāmṭek and Mansar; and Bori, a village which particularly displays Early
Medieval to Medieval remains.
Figure 9-18. Sculpture no. 271: A small sandstone Viṣnupada, typical of those identified in the Rāmṭek
survey, at site 159 near Kawadak.
commonly found in temples and across villages in rural areas due to their association
with agriculture and fertility (Blurton, 1992:68). Nagas can be found in temples with
different religious affiliations and in the past, appear to have been associated with
both Buddhism and Brahmanism. The “phallic shape” of the Naga creates a clear link
with Śaivism (Blurton, 1992:68). Shaw has discussed the Naga’s incorporation into
Buddhism as a possible method for creating relationships with rural agriculturalists
and the close connection often witnessed between Nagas and water may indicate a
Buddhist connection to water management through the “appropriation” of a popular
ancient deity (Dass, 2006:41; Shaw, 2004:50). Around Sanchi, Shaw recorded
numerous anthropomorphic Naginis and Nagas, usually dated from 1st century BC to
5th century AD, in close association with tank embankments. Shaw used the Nagas as
a proxy indicator of the date of construction of the water resources (Shaw,
2005/2006:47). Anthropomorphic Nagas are totally absent in the Rāmṭek survey
region and the association between Nagas and water is not obvious. Only
occasionally are theriomorphic Nagas situated on the side of tanks and their direct
association with the water management structures is questionable.
Naga/snake worship is strong around Nagpur and early Nagas are already known
from the survey area; a small 5th century Naga, carved from soapstone with a “typical
Vākāṭaka flower cap” was found by Wellsted in the burial shaft at Mansar (Bakker,
2008a:85), and a Naga image is found in front of the Kevala-Narasiṃha temple on
Rāmṭek hill (Bakker, 2007b:38). Both of these Nagas demonstrate that there was a 5th
century tradition of snake worship, but display differences in size, hood number and
body shape. Although a large number of Nagas images were identified on survey,
their dating is inconclusive and there is great variety obscuring any close parallels to
the known 5th century examples. It would not be unexpected if a number of those
recorded were early sculptures.
328
Chapter 9. The Ritual Landscape
Figure 9-19. Sculpture no. 2: Five-headed Naga from the Kevala-Narasiṃha temple at Rāmṭek.
Another regional trend in the survey area is metal snakes. A metal snake was found
during the excavation at Mansar and dated by the excavators to Period I (200 BC to
250 AD) (Joshi & Sharma, 2000:128). However, Bakker disagrees with this dating
and believes the metal snake to be Vākāṭaka (Bakker, 2007a). The iron snake was
roughly 25cm in length and was discovered at the foot of a lime figure of a man as
part of the ritual assemblage for the Puruṣamedha (Ali, 2002). The iron snake at
Mansar has been compared to one from Kauśāmbi (Pl.43, Fig.8, No. 38: Iron model
of a serpent, length 12.1cm) found in layer 1 of a ritual altar (Bakker, 2002:14; 2008-
2009; Sharma, 1960:122). It is difficult to know which of the metal snakes around
Rāmṭek are ancient and which may be modern based on visual inspection as most
have been daubed in sindūr, which further obscures any details. Given their level of
exposure and lack of associated remains, it may be that at least some of these are
modern snakes, but this still supports a continuation of this tradition in the area. The
majority of metal snakes were located in Rāmṭek and on roadside shrines at Mansar,
and a collection were found in the modern mining areas of Mandri and Kandri just
north of Mansar, which could suggest relocation of finds to modern shrines.
Occasional metal snakes were located elsewhere, however these were built into
modern shrines in villages (Nawergaon, Borda and Chargaon) with little to no
evidence of ancient habitation.
329
Chapter 9. The Ritual Landscape
Figure 9-20. (a) Sculpture no.538: 5-hooded metal Naga in a modern shrine in Rāmṭek, (b) Examples
of metal Nagas in a small shrine in Mandri (site 340)
The ceramic evidence found during survey, and connected to the economic landscape
in the previous chapter, may also be related to ritual activity. Miniature pots
discovered during survey may have a votive function, and therefore may be
representative of less established or more private forms of worship. Similar miniature
pots and jars have been identified at regional sites such as Paunar (Gupta, 1992).
Their presence in mound sites may be a reflection on rural worship and an indicator
of small scale shrines in the absence of permanent structural religious sites. The
fragmentary figurines found in habitation mounds alongside Early Historic pottery
may also have been connected to local ritual in the past, such as domestic shrines.
Figurines have been reported from Mansar in large quantities and include images of
divinities, a tortoise and a human skull (Manjhi, et al., 2000:56). Of the three
figurines located during the Rāmṭek survey, two appear to be the heads of bulls and
one is a part of a torso. All are made from micaceous clay and were situated in
mounded areas with extensive remains of Early Historic ceramics. Finds such as these
have significance for considerations of the religious composition of the landscape and
the depth of Vākāṭaka investment. Different levels of worship were present under the
Vākāṭakas with established, permanent constructions at Rāmṭek and Mansar and
indications of more local worship in the surrounding hinterland.
330
Chapter 9. The Ritual Landscape
It is interesting to note that all the conclusive 5th century sculpture identified on
survey was found in the vicinity of the three major sites of political investment, and
Hamlapuri which appears connected to Nagardhan, except for the possible replica of
Narasiṃha at site 242 (Beldongri). The finds of fragmentary 5th century sculptures in
Rāmṭek town may imply the presence of further religious establishments or village
shrines in the area beneath the hill. When characterising Rāmṭek in terms of its
religious function to consider the nature and scope of the Eastern Vākāṭaka polity it
appears there was fairly localised investment, which contrasts the picture achieved
from later remains of religious activity.
Figure 9-21. Distribution of the sculptural finds potentially related to Vākāṭaka activity (phase III in
the project chronology) and sculpture identified as most likely to be post-Vākāṭaka. The 5th century
sculptures are clustered primarily at the three major sites of Rāmṭek, Mansar and Nagardhan, with one
piece located at the site of a brick building in Beldongri, theorised to be a subsidiary temple to the
Narasiṃha temples on Rāmṭek hill.
331
Chapter 9. The Ritual Landscape
In the absence of solid habitation remains pertaining to the Early Medieval period,
other evidence such as sculpture, temple architecture and artefacts such as hero stones
have become more important (Khaladkar, 2007-2008:27). Temples became common
structural features in the Early Medieval period and thus the nature of religious
remains is better known (Dayal, 2005:63). Architecturally, the period immediately
following Vākāṭaka rule is under-represented in the survey data, and as noted in
chapter three, a single structure in Rāmṭek town, the Kālikā Devi temple (site 98), is
thought to date to the 7th or 8th century AD based on its distinctive vaulted roof
similar to the South Indian style (Kramrisch, 1946a:183). Unfortunately, the
purported Mahavihara (Viṣṇukuṇḍin) phase at MNS II contemporary to this
intermediate period cannot be recognised. During the survey, this picture of the
landscape persisted as no previously unrecorded temples displaying similar
architectural features pertaining to around the Cālukyan period could be identified.
The major period of structural development at Rāmṭek following the Vākāṭakas only
appears to have occurred with the spread of the Yādava dynasty around the 12th to
14th centuries AD. Interestingly, no previously unrecorded temples belonging to this
period were securely identified within the survey area, and it may again be that larger
temples requiring higher investment were restricted to in and around Rāmṭek.
However, architectural elements have been found at village shrines which may
indicate a greater presence of past structures and suggest that the picture of temple
architecture across the survey area is not complete.
Particularly across the south of the survey area, several smaller architectural
fragments with common motifs, including floral designs, were identified, but these
are mostly badly-preserved sandstone pieces. Architectural fragments are frequently
found placed at shrines with other miscellaneous sculpted fragments and are most
common at sites such as the fort at Nagardhan, and at villages which have evidence of
pre-Modern occupation, such as Chichala. Dating of these architectural fragments is
not easy as most are isolated, heavily eroded and fragmentary, so that even those with
better preservation can only generally be categorised as ‘Post-Vākāṭaka’ in style. The
spread of remains is largely within modern villages, although it is unknown whether
they have been moved from their original find-spot or whether they exist in sites of
332
Chapter 9. The Ritual Landscape
Figure 9-22. Example of a sandstone architectural fragment located on a small mound near Nandpuri
(site 91).
During survey a number of āmalaka pieces were discovered, which are “cogged”
ring-stones (Dass & Willis, 2002:33). These are normally used to crown the tall
śikharas of North Indian style ‘Nagara’ temples, but are not found on South Indian
‘Dravida’ temples (Kramrisch, 1946b:348). The earliest known āmalakas date to
around the 5th century AD at Gupta period sites (Kramrisch, 1946b:348). At Rāmṭek,
several large āmalaka fragments in red sandstone can be found around the Vākāṭaka
period Varāha shrine, which itself is topped by an āmalaka. Other examples of
fragmentary sandstone āmalakas were found during survey, often re-instated as
objects of worship in small shrines. Sculpture number 142 is located on a platform
shrine north of Rāmṭek and is associated with eroded sculpture. The two other pieces
identified (sculptures 418 and 321) are presented on village shrines as rudimentary
images of the seven mothers (the saptamatrikas). These āmalakas are small with slim
serrated segments, more similar to later examples found on the body of the śikharas
of Early Medieval or Medieval temples than the larger crowning examples of the
earlier period (Kramrisch, 1946b:348).
333
Chapter 9. The Ritual Landscape
Figure 9-23. Āmalaka fragments from across the survey area. Left to right, top row: (a) Red sandstone
āmalaka in the enclosure of the Varāha temple on Rāmṭek hill (b) Sculpture no.142. Left to right,
bottom row: (c) Sculpture no.418: An āmalaka worshipped as the saptamatrikas (d) Sculpture no.321:
An āmalaka worshipped as the saptamatrikas
Two Siṁha-Vyālas were identified at modern temple sites (sculptures 58 and 274),
both of which also display numerous pieces of post-Vākāṭaka sculpture. Vyālas are
zoomorphic architectural features found on the gateways or doorways of temples
(Michell, 2000:30). They are known to feature in Śuṅga, Kushāṇa-Kshatrapa and
Gupta architecture and persisted throughout the Early Medieval to Medieval periods
(Dhaky, 1965). A Siṁha-Vyāla is a lion-headed version of the vyāla, that has been
noted to appear on the door jambs of temples in Maharashtra (described as post-11th
century in (Deglurkar, 1974:164). The two Siṁha-Vyālas are both of yellow
sandstone and appear to be in the same style. The cabūtrā at site 169, north of
Rāmṭek, also includes a piece of post-Vākāṭaka door jamb. The ornate doorjamb
piece (sculpture no. 272) features a male figure or attendant, with a female figure
seated above. There also appears to be another eroded architectural piece at this site
which resembles an elephant. Another doorjamb piece is located at Koteshwar temple
334
Chapter 9. The Ritual Landscape
(sculpture no. 16 at site 50) in Nagardhan. This fragment shows a makara, and a male
figure holding a lotus. These architectural elements all appear to be ‘post-Vākāṭaka’,
and are likely to date from around the Yādava period.
Figure 9-24. (a) Sculpture no. 58: A vyāla fragment found at Manapur temple. (b) Sculpture no. 274: A
vyāla fragment incorporated into a cabūtrā north of Rāmṭek (site 169), which features a number of
architectural pieces.
Figure 9-25. Doorjamb fragments discovered during survey: a) Sculpture no. 272, part of a doorjamb
on a cabūtrā north of Rāmṭek (site 169), which features a number of architectural pieces. b) Sculpture
no.16, architectural door piece featuring a Makara located at Koteshwar temple, Nagardhan (site 50).
335
Chapter 9. The Ritual Landscape
These survey findings are similar to those of the Ballabgarh survey as fragmentary
architectural remains have potentially been discovered which appear to relate to the
7th to 12th centuries AD, but no extant temple structures can be securely dated prior to
the 18th century (Lahiri, et al., 1996:55). The structures themselves may have been
replaced by later constructions, modified to the point that they are less recognisable
as older temples or destroyed, with a traditional explanation being that earlier temples
were ruined following the governance of the Sultanates. It is therefore difficult to
make a statement about the amount of activity across the periods immediately post-
Vākāṭaka collapse up to the later Medieval period. However, it is also important to
note that if the picture of temple architecture is not complete due to preservation or
survival, we may question the comprehensiveness of the picture of the ritual
landscape across other periods.
Mārāṭha and Islamic influence is clearly visible in the design elements of later
temples and structures in the Rāmṭek survey area. In a study of Mārāṭha architecture
in Maharashtra, Sohoni identifies a distinct tradition which appears to be the synthesis
of Hindu (Yādava) and Islamic (Sultanate and Mughal) architectural conventions
(Sohoni, 2011:51). The architectural style around Nagpur is considered to be distinct
from the typical ‘indigenous’ Mārāṭha style as it developed under autonomous
Bhonsle rule and remained slightly removed from the rest of Maharashtra (Sohoni,
1998:33). The Bhonsles appear to have favoured local sandstone for construction and
their temples tended to be more ‘revitalist’ in style indicating an attempt to recover
336
Chapter 9. The Ritual Landscape
Figure 9-26. Chatrī located on Nagardhan mound, west of the fort (site 75), displaying a common
plinth motif.
337
Chapter 9. The Ritual Landscape
Figure 9-27. Architectural details on the platform of a large ornate chatrī at Nagardhan (site 53).
The largest number of later Medieval/ Mārāṭha temples in the survey area are to be
found throughout Rāmṭek town, which is not surprising given the Mārāṭha influence
on the area and the development occurring on the hill and around Ambālā Lake.
There are several older temples outside of the main town, with similar decoration and
architectural features, which appear to be Mārāṭha period constructions. Site 256 in
Nagardhan incorporates several Medieval temples including a relatively large
complex used as part of a school, and to the north of Rāmṭek is another example of a
large later temple that is situated on the bank of a water reservoir (Japala temple, site
393). Bori village has two older temples, which when viewed alongside other
evidence such as old wells, sculptural remains and ceramics, indicate the village was
occupied during the Medieval period. Occasionally, temples in village areas have
foundations or architectural features which appear to be old but are recently painted
and modified, as at site 284 in Lohadongri. Other temples may have early
components or foundations, such as the small temple at Kelapur (site 353), but
renovations hinder identification of chronologically distinct features. Furthermore,
even though temples may appear to relate to a particular period stylistically, the sites
may have long histories of use, and so the visible remains at a site may not
necessarily be representative indicators of all periods of activity. Difficulty also exists
in identifying the ‘tutelary deities’ in individual temples, as it is hard to confirm
whether the sculptures housed there now are original and the names of temples and
their original affiliation is often lost (Shaw, 2007:75).
338
Chapter 9. The Ritual Landscape
Figure 9-28. Examples of Medieval or Maratha period temples in the survey area: a) The temple at
Japala (site 393), which also features an associated stepped reservoir. b) A temple at Lohadongri
village (site 284), whose superstructure has recently been renovated and painted.
Figure 9-29. The distribution of temples and shrines across the survey area. Modern villages have been
indicated on the map and modern temples were most commonly recorded at modern settlement sites.
Temples deemed to be older (Post-Vākāṭaka/Medieval) also demonstrate a correlation with some of the
modern villages suggesting a continuation of settlement. Shrines and temples are also found
throughout the agricultural areas.
339
Chapter 9. The Ritual Landscape
Figure 9-30. South facing Hanuman temple with modern exterior in Nagardhan (site 62).
The majority of sculptures identified during survey reflect the evidence from the
architectural remains and appear to be Medieval or modern. The dominant religion in
the area today is clearly Hinduism and the modern spread of shrines/sculpture is
mostly Śaivite. Around Nagpur, temples usually have the same combination of
sculptures under worship; temples invariably contain a small yoni showing the liṅga
and pīṭha in union, associated with the bull (Śiva-Nandi), Hanuman and a trio of
mūrtis of Ganesh, the snake (Naga) and Annapurna. The triad of Ganesh, Annapurna
and Naga may be depicted on one panel, or may be individual sculptures which are
comparable in size and iconography. The examples of this combination of mūrtis
340
Chapter 9. The Ritual Landscape
appear to be mostly carved from local sandstone or basalt but can be found in metal,
and appear to be mostly Late Medieval or modern. A possible explanation for the
triad may be the association of the deities with three basic human needs: Ganesh
removes obstacles in life; Naga provides protection of the site/house;
Annapurna provides food in the house (Bakker, pers. comm.).
Figure 9-31. Sculpture no. 266: Example of the trimūrti of Ganesh, Annapurna and the Naga which
form part of the Pañchāyatana.
341
Chapter 9. The Ritual Landscape
Figure 9-32. Individual temple niche icon of Annapurna (sculpture 175) located in a Hanuman temple
in Lohara (site 315).
The commonly depicted bull, present in the Pañchāyatana, is the mount of Śiva,
which is widely referred to as ‘Nandi’ (Blurton, 1992:108). The bull sculptures
identified around Rāmṭek fall into several main types; shrines primarily contain a
small Nandi associated with a miniature liṅga and pīṭha but these are usually heavily
eroded, sandstone sculptures, which have lost most of their features. Larger Nandi
sculptures are also identified, usually carved from basalt, and these display clearer
iconography. The larger humped bulls are all seated, with decorative bridles and
garlands, often with bells or flowers, and tend to be found in the porch of Medieval
temples (Dhaky, 1972). This type is also present around Ambālā Lake, which features
extensive Mārāṭha period structures. These iconographic features most closely
resemble later sculptural styles and the associated yonis also appear to be late in style
and manufacture. The yonis mostly appear to be relatively late renderings of the liṅga
and pīṭha, featuring a small oval-shaped pīṭha with carved run-off lines for liquid
libation (Mitterwallner, 1984:26). The Nandis usually display ornamentation over the
body and are quite heavily stylised (Dhaky, 1972).
342
Chapter 9. The Ritual Landscape
Figure 9-33. Left to right: Sculpture no. 275: Painted Nandi at Japala temple (site 393). Sculpture no.
202 Nandi at a Hanuman temple in Bori village (site 311)
Figure 9-34. Sculpture no. 147, ‘Yoni’ with small liṅga and oval, shallow pīṭha.
Hanuman is ubiquitous across the Rāmṭek area and is the most commonly identified
deity (122 out of 696 sculptures identified). Hanuman is often situated in the side
porch of aniconic Śiva temples, or may be unaccompanied in small village shrines
and cabūtrās. Alongside clear depictions of Hanuman are numerous rudimentary
representations, featuring a stone as the central item of worship. This may be related
to Hanuman’s role as a guardian deity and protector of agricultural areas (Lutgendorf,
2007:41). Determining the iconography and date of the Hanuman sculptures is
particularly difficult as the layers of sindūr applied during worship results in
“popular” Hanuman images appearing rounded with only their eyes clearly visible
(Lutgendorf, 2007:59). Although the iconography is relatively late in date, it is
343
Chapter 9. The Ritual Landscape
Figure 9-35. Left to right: Sculpture no. 56: Standing Hanuman with tail curled over his head, hands in
front of the body and a gada under his left arm; Sculpture no. 203: Striding Hanuman in the “heroic”
vīra pose.
344
Chapter 9. The Ritual Landscape
Within the Hindu tradition across the survey area, there are several small temples
dedicated to the Mother Goddess in some form, but the sculptures all appear to be
Medieval or later. The goddesses under worship in the survey area were primarily
stated to be Mata Mai, Durga, Kali or the Saptamatrikas. The sculptures were often
quite rudimentary, as noted with the āmalakas fashioned into the Saptamatrikas, and
the majority of ‘Mata Mai’ shrines contained unhewn stones. The village shrines
encountered across the survey largely consist of rough platforms, often under a tree,
with stones as the object of worship. The platforms are sometimes obviously modern
and formed from concrete, or are occasionally either Medieval themselves or
constructed from re-used carved stone blocks. Stones, often painted with sindūr, are a
common object of worship across rural India, and are considered to be a “permanent”
representation of the “presence of a deity” (Kramrisch, 1983:98). During survey, the
presence of a shrine with unhewn stones as the object of worship was recorded,
however the stones themselves did not receive a unique sculpture number due to a
lack of attributes to aid with chronological interpretation.
Figure 9-36. Site 166: Small modern shrine containing stones, located in an agricultural field.
Ubiquitous in the survey area is the presence of local folk or village practices, co-
existing with mainstream temples. Provincial sculptural remains may be used to
demonstrate the persistence and strength of village traditions and the complexity of
the religious landscape. A great number of sculptural pieces which seem to display a
345
Chapter 9. The Ritual Landscape
folk or village style were identified during survey but interpretation of these pieces is
complex as local iconography is less standardised. Many carved figures display a
similar standing posture and features including round faces, large ears and long
disproportionate limbs. Several small unusual sculptures were recorded across the
survey area, which appear to be local representations of unknown figures.
Figure 9-37. (a) Sculpture no. 282: Unusual folk-style temple alcove piece of unknown figure at Japala
temple (site 393) potentially Medieval (Bakker pers. comm) (b) Sculpture no. 77: Sculpture of
unknown figure on a cabūtrā between Mansar and Rāmṭek, which is said to be Hanuman by villagers
(site 174).
Figure 9-38. Sculptures no. 169 and 170: Folk-style depictions of two figures. Similar sculptures are
very common across the survey area
346
Chapter 9. The Ritual Landscape
Two larger fragments of sculpture at Chichala depicting female figures were noted as
being of interest due to the use of the local red sandstone preferentially utilised for the
5th century Vākāṭaka sculptures. These figures are fairly crude in dimension and
display strange iconography, with broad shoulders and long arms. Both are holding
weapons or attributes but neither is identifiable. As with other figures of a
‘provincial’ style, the date of these sculptures cannot be determined in the scope of
this study.
Figure 9-39. Sculpture no. 449: Red sandstone sculpture of a female torso in a provincial style at
Chichala (site 258)
Another common tradition noted across the survey area was the appearance of village
shrines devoted to ‘Bagh’ with small images of tigers carved either in a seated or
standing position. Worship of the tiger is known amongst the tribes of Maharashtra,
where it may be referred to as ‘Vagh’, ‘Bagh’ or ‘Wagh’ (from the Sanskrit vyāghra
meaning `tiger') (Singh, 2004). Bagh is said to be a part of the ‘lower pantheon’ of
Gods and is viewed as being either responsible for the protection of cattle or as a
force to be placated for general protection (Pillai, 1997:210; Tiwari, 2002:288).
Worship of the tiger is particularly known amongst the Gond tribe which is present in
Maharashtra and across central to north India (Crooke, 1894:322; Tiwari, 2002:26).
Traditional worship of tigers as a totem has also been identified in groups such as the
Mavchi, Bhil, Ganda, Basor and Dhanka in Maharashtra (Gajrani, 2004:257; Singh,
2004). These images again represent the variety of forms of local worship but cannot
be ascribed a phase.
347
Chapter 9. The Ritual Landscape
Figure 9-40. (a) Collection of fragmentary animal sculptures on a village cabūtrā including sculpture
no. 155 which is said to be a tiger. (b) Sculptures no. 496 and 497: Both described as ‘Baghdev’ by
villagers.
Naga worship became more important and widespread under Bhonsle rule in the area
(1730-1816) and these images can be found incorporated into different architecture
such as forts, chatrīs and bāvlīs (Girhe, 2002:546). Numerous Nagas were identified
during survey and these images are typically free-standing, theriomorphic forms of
the snake, with either one or multiple hoods (three or five). Most of these sculptures
appear to be quite late in date, and some are clearly modern, but none have any
particularly distinctive characteristics for dating. As mentioned, Nagas are often
incorporated into the Hindu Temples dedicated to Śiva and Hanuman, as part of the
Pañchāyatana. The Naga images are also located on isolated shrines and in
collections of fragmentary sculpture and stones.
Figure 9-41. Left to right: (a) Sculpture no. 157: Five-headed Naga at a Hanuman temple in Kandri,
north of Mansar (site 426). (b) Sculpture no. 513: Single-headed Naga on village outskirts (site 370)
348
Chapter 9. The Ritual Landscape
Stones with carvings of snakes, called Naga-kals, are common in rural India. They
can depict single snakes with coiled bodies and usually one or five hoods, or “pairs of
intertwined snakes” (Blurton, 1992:108). Naga-kals are particularly common in
eastern, central and southern India, often found in association with water sources such
as rivers or tanks (Hegewald, 2002:36). The ubiquitous Naga-kals across Rāmṭek
vary greatly in size and style but there is no defined iconographical typology to show
the chronological development of either the Naga stones or free-standing images.
None of the Nagas around Rāmṭek are inscribed, which can assist with dating. Naga
iconography is difficult to interpret as the tradition has great antiquity and is very
complex with relationships to both Hinduism and Buddhism (Blurton, 1992:107).
Figure 9-42. (a) Sculpture no. 304: Unusual Nagakal at site 198 in Hivra village. (b) Sculpture no. 54:
Naga stone at the Kālikā Devi temple in Rāmṭek.
Regarding other religious influence on the area, there is frequent Islamic influence on
the later architectural elements as mentioned and Late Medieval or Modern Muslim
tombs were identified in the larger towns of Mansar, Rāmṭek and Nagardhan. Muslim
tombs and platforms often feature re-used bricks, such as at sites 101 and 115, which
demonstrates the ongoing local trend for re-use of building materials. Finally, there is
a large Jain temple complex in Rāmṭek town, Shri Shantinath Jain Mandir, which is
believed to have been constructed under Bhonsle rule (see Sohoni (1998) for a brief
discussion of the architecture of the fifteen stone temples in the Jain complex).
Mirashi reports that, “a huge and beautiful image of Santinatha about 15 ft. in height”
349
Chapter 9. The Ritual Landscape
was recovered around Rāmṭek and installed in the Jain temple (Mirashi, 1966:53;
1968). During survey no sculptural or architectural remains of a clearly Jain origin
were found across the Rāmṭek area and so there is currently no known evidence to
suggest the presence of a Jain community until the later period. One piece of
sculpture was identified in the Bhonsle Jain temple in Rāmṭek town. The original find
spot of this piece is unknown but it is likely to be an alcove piece from one of the
temples within the complex. This sculpture appears fairly late in date, which does not
reveal any earlier trends of Jainism. There is no visible evidence to suggest earlier
foundations at the Jain Mandir but the site has been extensively developed and is
currently undergoing building works to construct a modern temple and hall.
Figure 9-43. A Jain temple in the complex north of Rāmṭek town (site 99).
350
Chapter 9. The Ritual Landscape
Figure 9-44. Sculpture no. 696: Possibly Jain image kept in the Jain Mandir in Rāmṭek.
A relatively small number of carved stones to memorialise the dead were found
around Rāmṭek during survey. Memorial stones are commonly found from the 8th
century AD across Maharashtra, and may be represented by hero-stones (viragals) or
sati stones (mahasatikal or mastikal) (Jamkhedkar, 1985-1986:35). Hero stones
originally memorialised the dead ‘heroes’ from war and cattle raiding, while sati
stones commemorate the immolation of a widow (Khare, 1982:253; Mate, 1982:77).
351
Chapter 9. The Ritual Landscape
Figure 9-45. Left to right: a) Sculpture no. 303: Potential hero stone piece with depiction of the sun and
moon. b) Sculpture no. 152: Sandstone hero stone fragment.
The sati stones around Rāmṭek were found to be separate from shrines and temples,
which supports their identification as memorial images rather than plaques featuring
local deities. Similarly, in Banavasi Suvrathan describes the memorial stones as being
“external to Brahmanical worship” or a subject of their own worship given that they
are usually situated outside of temples or on their own platform (Suvrathan,
2014:225). The sati-stones tend to be rectangular and display similar iconography; the
stones are carved with two standing figures, clearly a man and a woman, who are
holding hands and often bear a sword. The two figures may be representative of the
husband and wife but unlike in some published examples which feature common
symbols of the sun and the moon, no other symbols are carved onto the surrounding
of the Rāmṭek stones (Khare, 1982:252). It is common in Maharashtra to find “crude”
anthropogenic figures carved onto stone tablets (Mate, 1982:77). The Rāmṭek
352
Chapter 9. The Ritual Landscape
memorial stones show some similarities to sati stones identified by Thakuria, which
display a standing sati figure with a raised right arm (Thakuria, 2008-2009:185).
There are also some similarities to 18th century examples (fig.9.2 and 9.3) identified
by Misra, which depict a couple with raised hands (Misra, 2014:164). Sontheimer
identified three primary types of sati stone in Maharashtra and often the couple are
depicted, associated with symbols such as the liṅga, yoni, sun and moon (Sontheimer,
1982:278-281). The Rāmṭek sati stones do not fit within these defined categories and
may be a local expression with its own iconographical tradition. The Rāmṭek
memorial stones are most reminiscent of the Sati stones identified at Banavasi, which
depict a standing couple against a plain background and are considered to represent a
local tradition (Suvrathan, 2014:225). As with the hero-stones, the Sati images would
appear to be post-Vākāṭaka.
Figure 9-46. From top left to right: (a) Sculpture no. 232: Memorial stone (b) Sculpture no. 137: Image
which may be a memorial stone. (c) Sculpture no. 333: Red sandstone memorial stone. (d) Sculpture
no. 342: Memorial stone
353
Chapter 9. The Ritual Landscape
9.5. Conclusion
The evidence presented here goes some way to determining the role of the religious
landscape around Rāmṭek to lead into discussions of how the ritual site was used by
the Eastern Vākāṭakas to engage the population and feed into the state supported
ideology.
There are difficulties in attempts to specifically date the architectural and sculptural
material due to the history of use and current worship. It must also be noted that there
is ambiguity in whether the current spread of remains is representative of activity in
the past. No standing temples beyond those already known in Rāmṭek have been
assigned to periods outside of the phases of Vākāṭaka, Yādava or Mārāṭha
construction. Limited sculptural and architectural evidence may indicate some level
of activity, but is mostly in a poor state of preservation and can be only broadly dated.
Sculptural remains in particular have highlighted the varied modes of worship across
the survey area. However, sculptural fragments and carved images are difficult to
date, and some of the local iconographic traditions are not well understood. A
limitation with the results of the survey is that there is a large quantity of isolated
sculpture or architectural fragments, which cannot be securely connected to temples
or religious institutions. Therefore our understanding is restricted when trying to
consider the context of these fragments and the actual spatial distribution of religious
elements. These categories of archaeological material do however allow some insight
into the religious complexity of the past landscape around the major religious centre,
and the broad chronological phases identified may be used in conjunction with the
ceramic evidence in order to form hypotheses about the occupation of the Rāmṭek
hinterland.
The absence of clear ritual practice in the landscape outside of the major Vākāṭaka,
Yādava or Mārāṭha construction raises questions about whether this is due to survival
or reflects phases of structural building activity related to investment in the landscape
by a centralised authority. Based on the disparity between the absence of old temples
but the presence of fragmentary, eroded architectural pieces, there could be
indications that the currently assessed picture of religious activity in the landscape is
354
Chapter 9. The Ritual Landscape
incomplete. Given the unknown survival rate of earlier temples, it is difficult to come
up with conclusive hypotheses about the religious development of the area.
Small structural temples are not witnessed across the greater landscape until the
Medieval period, and modern villages in the survey area often contain a temple with
Medieval architectural features. This reflects the economic evidence of wider spread
water management structures as noted in chapter eight, and seems to demonstrate
greater investment in rural sites at this time. This may reflect on the different
355
Chapter 9. The Ritual Landscape
character of the later centralised authority and the focused investments of the Eastern
Vākāṭakas at key sites.
356
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
10.1. Introduction
This chapter addresses the original hypothesis of the research by evaluating Eastern
Vākāṭaka political investment in the Rāmṭek landscape, in order to challenge the
historical narrative of their expansion. It aims to consider both the economic and
religious function of the ritual centre to establish the relationship between Rāmṭek
and its hinterland. Although the survey data is not sufficiently nuanced to tie to
specific historical developments, a sequence of broad changes in the spatial and
temporal development of the landscape has been discussed in chapters eight and nine.
The survey data gives some indication of the nature of the rural economy and the
socio-political and religious transformations which occurred in this central area of
Eastern Vākāṭaka control. Early Historic investment in the landscape can be inferred
from the survey data, which may be used to speculate on the nature of kingship and
the political entity of the Eastern Vākāṭakas in the 5th century AD. The survey results
are relevant to a number of topical issues in South Asian archaeology and the key
themes highlighted in this chapter are Eastern Vākāṭaka polity and kingship, the
nature of their political centre, and the socio-economic role of ritual sites through
relationships with the hinterland.
Through postulations concerning the nature of Eastern Vākāṭaka kingship and polity,
this chapter will begin to consider the results of the survey data in the context of early
Indian state development and will highlight the regionality of the formation of early
complex societies. An attempt can be made to consider the coexistence of the Eastern
Vākāṭakas with the Western Vākāṭakas and the Gupta dynasty, including the
possibility of the Vākāṭaka kingdom having been influenced by the Imperial example.
It seems that the co-existing Vākāṭaka branches functioned fairly independently and
responded differently to their local environments through socio-economic and
political adaptation. The Rāmṭek survey data tentatively supports suggestions that the
357
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
It is argued here that although the Eastern Vākāṭaka polity could be classified as a
state development in the perspective of central Indian history, it was fairly localised
and presented a different political character to traditionally extensive regional
kingdoms or empires. The Eastern Vākāṭaka polity appeared to have operated through
the elite establishment of ritual sites like Rāmṭek in order to serve a ritual function as
a focal point of worship in the landscape and to integrate local worship through
religious patronage. The ritual centre may also have acted as a socio-economic
institution, through the re-direction of produce into elite activities in order to
consolidate the immediate territory and legitimate royal authority. The survey data
seems to support the theory that the Vākāṭaka presence at Rāmṭek was active in
comparison to previous periods, and resulted in changes to the nature of settlement
and worship through expansion into a relatively detached and underpopulated area.
However, it becomes apparent that physical investment in the landscape was fairly
sparse and limited to key sites of more intense political and ritual monumentality. The
nascent political centre may have been focused on funnelling production and
resources into the prosperity of the kingdom and its monumental expressions in the
immediate landscape. Rather than adopting an expansionist policy, it could be that the
Eastern Vākāṭakas primarily sought to stabilise their networks of control throughout
Vidarbha. The Eastern Vākāṭaka polity may differ from rigid, traditional views of
states as it could be inferred that the ruling elite had more fluid interactions with the
landscape featuring indirect control over loci situated in a ‘node and network’ system
to exploit resources and focus on a ritual and royal ideology to legitimate their
authority (Smith, 2007:28).
As established from the published literature, the Eastern Vākāṭaka polity was
traditionally considered to have expanded into ‘peripheral’ zones to stimulate the
development of areas less engaged with wider networks. Accordingly, the Eastern
Vākāṭakas are historically viewed as the first dynasty to bring the Rāmṭek area under
358
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
It is not unlikely that the establishment of sites such as Rāmṭek was a method of
extending the Vākāṭaka state presence and strengthening networks of resource
exploitation and production. Rather than being related to economic pressure and the
abandonment of urban sites to pursue an agrarian economy, the shift in focus of
royal/state investment to the ‘periphery’ may have been aimed at integrating wider
areas. Agricultural expansion may have promoted hinterland development in order to
support sites of royal investment and a move to Rāmṭek may have been aimed at
solidifying their territorial hold if existing major sites were considered to be
sufficiently strongly established. This shift may have resulted in an apparent
depreciation of earlier sites, which subsequently would have been located further to
the boundaries of the Vākāṭaka territory. However epigraphic evidence supports the
suggestion that previous capitals remained in use after an official move designated by
the epigraphic places of issue; for example, following the shift of the capital to
Nandivardhana, the Masod Plates still refer to the first capital of Padmapura in the
previous Sātavāhana territory (Bakker & Isaacson, 1993). The ceramic data presented
in this study indicates consistency between the cultural apparatus of the pre-Vākāṭaka
and Vākāṭaka periods, which implies there may have been continuity of settlement
with minimal upheaval or extensive cultural change resulting in Vākāṭaka remains
being overlooked at sites thought to have been abandoned.
Regarding the relocation to Rāmṭek and the more visible trend of land-grants in the
5th century AD, the historical narrative has suggested that the Early Historic
expansion of agricultural villages could be explained as an attempt to bring new areas
under the plough. However the spatial distribution of the inscriptional evidence now
359
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
indicates that they may have been part of an attempt to consolidate existing
agricultural areas through the establishment of temples to act as centres of
engagement. It seems that Rāmṭek would have been roughly the centre of the sphere
of influence of the Eastern Vākāṭakas in the 5th century AD, and acted as a focal point
of economic and ritual activity itself.
The data acquired on survey pertaining to both economic and ritual remains across
the Rāmṭek landscape demonstrates little formalised or settled activity prior to
Eastern Vākāṭaka control. Such a suggestion appears conducive with the historical
narrative of the Eastern Vākāṭakas ‘subjugating’ peripheral areas although the
landscape may have had complex settlement dynamics which are still largely elusive.
The survey has highlighted the possibility of transhumant pastoralism and seasonal
occupation in the landscape, and there is ambiguity in the dating of Early Historic
ceramics which could support pre-Vākāṭaka settlement. It could be inferred that the
Eastern Vākāṭakas’ move to Rāmṭek enabled them to impose a centralised authority
on pre-existing social structures and agrarian systems. Regional excavations have
shown a long history of agriculture in the region and the Vākāṭaka inscriptions
suggest this was the mainstay of the population, which is supported by small finds of
quern stones. As a case study for the economic position of the Eastern Vākāṭakas, the
survey data suggests that the Rāmṭek landscape had a dispersed semi-rural settlement
pattern and an agricultural economy. A dispersed pattern of small villages may reflect
a need to manage cultivation, and it may be significant that the village mounds
identified so far were mostly situated within the cultivatable plain with proximity to
water resources, as opposed to the increasingly hilly and forested areas in the north of
the survey region.
The most convincing evidence for significant change in settlement and the economy
under the Eastern Vākāṭakas is related to the clear monumental investment at Rāmṭek
and Mansar, creation of the Sudarśana reservoir and the establishment of the capital at
Nagardhan. The main activities in the region appear to have been agricultural and a
burgeoning village agrarian economy may have supported a growing population,
indicated by the more extensive habitation remains around Nagardhan. Agricultural
expansion may be connected to local prosperity which allowed the economic surplus
to be directed into channelled into the visible ritual and secular constructions and the
maintenance of royal control and state administration. This seems to correspond well
360
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
with information contained within the Mansar inscription, which states that the
Pravareśvara temple was maintained by a donated group of 26 villages (Meister, et
al., 1988:62). The size of the survey area could feasibly encompass the landscape
and surrounding villages engaged in this localised investment into Mansar.
It has been suggested that the nature of the Eastern Vākāṭaka relationship with the
agricultural area around Rāmṭek could have been an adaptive strategy to engage local
communities and legitimise Vākāṭaka authority across Vidarbha. The expansion of
state society into peripheral areas is suggested to have created a reciprocal
relationship whereby the local area was integrated into wider networks and the state
was also “localised” through the incorporation of rural elements (Sahu, 2012:150).
The significant rise in ceramics during the Early Historic period and the imposition of
the Brahmanical tradition at Rāmṭek supports suggestions of the integration of this
area into wider regional networks. Bakker has theorised that the worship of Varāha
and Narasiṃha in particular at Rāmṭek, suggests conscious “assimilation of popular
cults” in order to incorporate local trends, unify the community in worship, and
legitimise Vākāṭaka rule over the rural population (Bakker, 1992a:14). Eastern
Vākāṭaka society appears to have been grounded in local interactions and flourished
through ritual maintenance of its state system. Both the Western and Eastern
Vākāṭakas appear to have employed religious patronage and ritualism to reinforce
their authority, although their specific approaches manifested differently perhaps due
to adaptation to the local environments.
As the Eastern Vākāṭakas are thought to have consolidated their power through
Brahmanism, it seems the Western Vākāṭakas undertook a programme of patronage
to Buddhism in order to engage with the local communities and legitimise their rule
(Morrison, 1995:210, 214). Whereas the Rāmṭek survey data suggests an absence of
formalised religion prior to Eastern Vākāṭaka rule, the Western Vākāṭakas inherited
territory previously occupied by the Sātavāhanas, and may have encountered a strong
Buddhist presence (Morrison, 1995:210). The proliferation of Buddhist institutions
under the Sātavāhanas was part of an economic and religious system connected to
long-distance and international trade networks (Morrison, 1995:211; Ray, 1985;
Sinopoli, 2001). The Western area is therefore more likely to have developed with
this influence and it is probable that some of the economic structures persisted into
the Vākāṭaka period.
361
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
The development of sites, such as Ajanta would have necessitated a large economic
investment and suggests affluence across the Western Vākāṭaka kingdom (after
Bakker, 1997). It has been suggested that the Western Vākāṭaka exploits were funded
by expansion and control of land or routes, with Buddhist patrons and merchants
playing a larger donative role (Bakker, 1997:44-45). The continuing influence of west
coast trade on the Western Vākāṭaka economy can be inferred from depictions in the
5th to 6th century AD frescos at Ajanta (Kapur, 2006:18). While prosperity is also
demonstrated by the monumental constructions at Mansar and Rāmṭek, the affluence
of the Eastern Vākāṭakas is considered to have been rooted in their local system and a
small-scale agrarian economy, without significant influence from international trade.
This appears connected to the strong tradition of land-grants in the Eastern Vākāṭakas
kingdom compared to the Western Vākāṭaka kingdom, which signifies the adoption
of a different political and economic response to their territory. There is limited
epigraphic evidence and fewer completed excavations to support our understanding
of why land-grants did not play a major role in the Western Vākāṭaka kingdom but it
may be theorised that this branch adjusted to the environment and social background
they encountered in order to successfully engage with existing local communities.
Central to the Eastern Vākāṭaka ‘state’ formation process appears to be the donation
of tax-free land to Brahmins, royal patronage to key ritual/pilgrimage sites and temple
building at the centre of the dynastic territory (Kulke, 1993:10). The land-grants have
been variously explained as revealing political, social and ideological motives to
develop infrastructure, integrate communities, maintain sovereignty and extend the
influence of the Brahmanical religion and social order. Sinopoli has suggested that in
part the land-grant process may have been associated with “structural weaknesses” of
imperial political and economic organisation. Attempts to develop autonomous
religious institutions outside of the state may be indicative that dynasties recognised
their instability and were trying to guarantee longevity (Sinopoli, 2001:170).
362
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
nature of Eastern Vākāṭaka polity and kingship, in light of their deliberate move to
Rāmṭek and establishment of the monumental centres. It appears that the focus of the
Eastern Vākāṭaka investment in the landscape was the ritual sites as opposed to deep
penetration into the rural landscape and creation of urban centralised life, which
reflects on their characterisation as a political entity.
Kennet has suggested that the Vākāṭaka period witnessed a redirection of the
economy and displayed an alternative form of state rule, which may be more difficult
to identify in the archaeological record than the traditional urban response and
indicators of growing socio-economic complexity (Kennet, 2004b:15). Existing
research demonstrates that no incontrovertible evidence for large-scale urban
developments has emerged from the nine excavations within the Vākāṭaka territory
and known sites tend to reveal temples and religious structures (Thakur, 1997-
1998:26; Vajpeyi, 1986:147). Sharma acknowledged that late Gupta or post-Gupta
centres displayed extensive religious remains, but these were not considered to be
evidence for urbanism despite the relationships between monumental constructions,
religious ideology, administration and a site’s economic base (Sharma, 1987b:131,
177). It may be that previous investigations were not sufficiently widespread to
identify urban remains or that settlement shrinkage and the movement of sites based
around shifting centres of power has concealed these habitation remains (Kennet,
2013:348). However, it has also been suggested that Vākāṭaka state rule was not
necessarily based on town and cities but may have been focused on key religious
sites, such as Rāmṭek (Kennet, 2004b:15). The result of this is that areas such as
Rāmṭek may not fit traditional models of urbanism and state formation (Childe, 1950)
which involve “nucleation” of settlements but rather witnessed an alternative
structuring of polities and the economy, with ritual sites acting as foci for activity
alongside a less centralised state administration (Creekmore & Fisher, 2014:3;
Fletcher, 2007:189; Varma, 2008:218).
An alternative organisation of the state apparatus, would see ritual centres such as
Rāmṭek, performing a number of the functions typically associated with urban
centres, which would possibly account for an absence of traditionally defined large
363
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
and densely occupied ‘urban sites’ (Kennet, 2004b:15). The survey data lends support
to this hypothesis, as rather than yielding remains of a single large urban site, the
centre of the Eastern Vākāṭaka kingdom exhibits a trend of small settlement sites
focused around a major religious centre, with several connected capitals with
arguably different functions. It is difficult to ascertain from survey alone whether
Rāmṭek was a flourishing urban settlement site under the Vākāṭakas, or whether the
site was primarily a ritual centre until a later period. Mansar appears to have been a
monumental expression of royal authority alongside Rāmṭek’s dominant religious
significance across the area. Conversely, Nagardhan appears to have drawn in the
population as it seems to have been a relatively large settlement site in the
agricultural plain, surrounded by fairly evenly spaced villages, and further low-
density settlement activity evidenced by ceramic background scatter. These three sites
may have created a local network, each contributing to the different mechanisms used
for integration of the rural communities.
Rāmṭek, Mansar and Nagardhan are at the centre of the Eastern Vākāṭaka kingdom –
the second largest in India at the time – and yet do not constitute a huge urban
conglomeration. Recent research has expanded to consider the context of sites in their
hinterlands to establish their differentiated functions, and has demonstrated that
‘urban’ sites can encompass an array of economic and political formations outside of
compact and developed centres (Coe, 1961:66; Creekmore & Fisher, 2014:7;
Fletcher, 2007:188). Functions typically associated with urbanism have been
identified at dispersed agrarian settlements, and in particular tropical forest
environments sites have been found to demonstrate ‘low-density’ urbanism (Fletcher,
2007:188; Fletcher, 2009:7). This urban form differs to the densely nucleated
settlements expected in the Middle East or the ancient Indian subcontinent
(Wilkinson, 2014:183). Key examples of tropical forest civilisations have been
observed in Mesoamerica (the classic Maya, c.300-900 AD) and mainland South-East
Asia (the Khmer, 802-1431AD) (Fletcher, 2009:2).
Mesoamerican lowland cities tended to be small city states with smaller territories
and dispersed populations (Creekmore & Fisher, 2014:16) (Isendahl & Smith,
2013:132). While there was an apparent lack of traditional ‘cities’, large “ceremonial”
monumental centres have been identified which appear to have been supported by the
produce and labour of the surrounding villages (Coe, 1961:76; Stark & Ossa,
364
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
2007:385). It has been suggested that areas were exploited for resources, and
monumental centres were relied on to maintain societal coherence (Dunning, et al.,
1999:654; Dunning & Beach, 2011:377; Scarborough & Lucero, 2010).
Archaeological surveys at Angkor, Bagan and Anuradhapura have shown that these
sites should also be understood in terms of extensive, low-density settlement patterns
(see Evans, et al. (2007), Hudson, et al. (2001) and Coningham & Gunawardhana
(2013) respectively). The Classic Khmer settlement pattern consisted of a large royal
administrative and cult centre supported by its hinterland of low-density rural villages
so that produce was directed towards cult centres (Coe, 1957:410; Evans, et al.,
2007:14277). A low-density urban landscape model for Anuradhapura suggests that
the dispersed hinterland settlements and population were integrated by focal
monasteries (Coningham, et al., 2013:461). Ceramic scatters are proposed to
represent small peripatetic villages formed around monasteries subsidiary to those in
the monumental sacred centre, which acted as foci to carry out administrative,
political and economic functions (Coningham, et al., 2006:63; Coningham, et al.,
2013:461; Coningham, et al., 2007). The small size of sites around Anuradhapura
means that the settlement pattern is difficult to compare to those of other low-density
cities, and so it has been noted that the settlement pattern could be defined as either
low-density urbanism or dense small rural settlements depending on the criteria
applied (Wilkinson, 2014:193).
365
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
Something akin to low-density agrarian-based urbanism may partly explain the nature
of the Rāmṭek landscape, given that the limited survey appears to have demonstrated
important centres at Rāmṭek-Mansar-Nagardhan with an interconnected hinterland
network of smaller temples and villages. Urbanised places may be considered higher
order settlements in a region of interrelated people and settlements, which provide
centralised functions, such as political administration or economic distribution (Stark
& Ossa, 2007:388). A main point of Coe’s tropical forest civilisation theory is that
agricultural surplus was directed towards non-agrarian elite activity (1961), which
finds parallels in the proposed economic structure of the Eastern Vākāṭaka kingdom.
Furthermore, the centres of low-density urban cities seem to have been able to attract
the population through “political, social, economic and religious means” (Lucero, et
al., 2015:1141).
As the evidence from Nagardhan does not fit the stereotypical model of a compact,
developed urban centre, it may be that the Rāmṭek area is better described as a type of
conurbation (Coe, 1961:66). There are clear comparisons with Casile’s study at
Vadovyapattana during the Badoh-Pathari explorations. Vadovyapattana was an
important royal centre in central India, where the core area was surrounded by rural
settlements. The pattern of archaeological remains did not reflect the traditional
dichotomy between ‘urban’ and ‘rural’ and so the landscape was described as being
“agro-urban” (Casile, 2014:141). Comparison may also be made to Hawke’s survey
at Bharhut, which revealed only five settlement sites appearing as low mounds with
brick and ceramics scatters and no ‘urban’ centre; the largest site at Patarahata
covered 15 ha and had an unknown function (Hawkes, 2014a:359). As at Rāmṭek,
Hawkes did not detect any of the “usual archaeological signatures” of urbanism and
theorised that the local economy and polity functioned on a relatively small scale
(Hawkes, 2014a:360). The larger settlements identified in agricultural areas were
theorised to represent small local centres (Hawkes, 2014a:364). This again supports
the suggestion that ‘local centres’ could have been foci in networks, carrying out a
variety of ‘urban’ functions.
366
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
Rather than denoting the breakdown of state structures, Nandi has suggested that the
increasingly decentralised formations of regional kingdoms apparent during the
Gupta-Vākāṭaka period would not have been an unusual feature of Early Indian state
formation (Nandi, 2000:18). There is evidence to suggest that the ancient Indian
states, such as the Mauryan and Sātavāhana polities, were not highly centralised state
systems as traditionally thought, but were in fact fairly “ephemeral” (Sinopoli,
2001:155). Anthropological models have often been utilised to rationalise this
theorised absence of a “centralised bureaucratic structure” in state formation,
particularly in the late Early Historic to Early Medieval periods (Ali, 2012:9).
Although later than the period of Eastern Vākāṭaka rule at Rāmṭek, we may draw
from such alternative models as archaeological, numismatic and epigraphic evidence
indicates a similar dispersed state structure. Focus is drawn to the establishment and
growth of state systems with different foci and networks, using comparable
mechanisms to the Vākāṭakas as a means of spreading ideological control and
legitimising political authority, such as religious grants (Heitzman, 1991:25;
2009:214-215; Prasad, 2012:34).
Within Stein’s ‘Segmentary’ model for Cola state formation in South India, the state
is considered to have consisted of a core area of centralised power and ‘localities’
(nadus) which fairly autonomously carried out much of the state’s political and
administrative activity, but were united under the “ritual sovereignty” of the royal
authority (Heitzman, 1997:15-16; Stein, 1977; Stein, 1985:394). Heitzman proposed
the idea of ‘temple urbanism’ as a means of state formation, arguing that temples
became part of an integrative, administrative system which supported royal control
(Heitzman, 1997:107). Chattopadhyaya and Kulke’s ‘Integrative’ model of state
formation focuses on the proliferation of rural agrarian settlements and the spread of
Brahmanical ideology, which acted as an overarching mechanism to integrate local
communities and religions (Shrimali, 1993:27). In Orissa, the early centuries AD
appear to have been defined by the development of small kingdoms and Kulke
focused on the role of Brahmins in consolidating these territories with land-grants
used as a mechanism for creating and strengthening local networks (Heitzman,
1997:14; Kulke, 1978:128; Sinha, 2001:159). Chattopadhyaya proposed that the
367
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
It may be proposed that Rāmṭek formed part of an ‘urban’ ritual area, which are
known to have drawn in the population through acting as centres of pilgrimage and
cultural interaction and locations for “economic transactions” (Smith, 2014:311).
During research at Kaundinyapura and Sisupalgarh, Monica Smith stressed the
importance of the hinterland by noting that the economic growth of Early Historic
urban centres may have been based in local exchange and production. Smith has
proposed that alternative models of urban success could have existed which focused
on different levels of trading networks given that the evidence at Sisupalgarh suggests
that Early Historic cities could be “self-sufficient” (Smith, 2005:297). Such research
highlights the potential for a complex system of local, regional and long-distance
exchange to have existed across the Early Historic subcontinent, where emerging
cities were embedded in local networks rather than being connected through regional
exchange like the ancient urban sites of the first urbanisation episode in India (Ali,
2012:9). Smith suggested that while towns such as Kaundinyapura may have relied
more heavily on regional exchange, cities could have become prosperous through
their association with a thriving hinterland (Smith, 2002a:148). The large Early
Historic city of Sisupalgarh displayed minimal regional exchange with a local
ceramic tradition. On the other hand, the material culture at the town-site of Talapada
in its hinterland demonstrates that smaller habitations were active in the regional
economy (Mohanty, et al., 2014:53). Similarly, ceramics and other artefacts at the
town of Kaundinyapura suggest prosperous local and regional exchange with limited
long-distance trade of marine shell alongside a tradition of local rather than traded
fine-wares (Smith, 2002a:145). This evidence suggests that central Indian exchange
networks could have made use of regional materials but largely involved the local
production of domestic artefacts and ceramics (Sinopoli, 2001:174; Smith,
2002a:147). The remains around Nagardhan are reminiscent of those at Sisupalgarh
in that they indicate local manufacture of ceramics and a lack of exchanged goods;
again the site appears to have been “anchored” in the surrounding hinterland with
368
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
If urban sites were more reliant on a strong local hinterland than long-distance
exchange then agriculture is necessary for urban growth (Smith, 2002a:148). If
political organisation was “relatively weak” in the period in which sites such as
Sisupalgarh prospered, and regional dynasties experienced fluctuating territorial
control, then even small urban areas, as potentially seen around Rāmṭek, may have
been the most consistent space in which to focus investment into monumental
constructions (Smith, 2002b:111). Sites considered urban appear to have extended
their influence within a local “urban conglomeration”, which fulfilled a role similar to
the regional exchange of other areas (Smith, 2002a:149). Additionally, local trade has
been suggested to have been a means of engaging the population through creating a
“shared cultural system” (Smith, 2002a:149). The concept of a “local system” is
relevant to this context as the Rāmṭek evidence seems to suggest an interconnected
network of sites secured in a productive hinterland (Smith, 2005:304).
If states are understood as existing within networks then state expansion could be
comparatively low investment as growth relied on the varying control of nodes and
corridors as opposed to delineation of rigid territorial boundaries (Smith, 2007:31).
The ‘node-and-network’ theory allows us to consider a more flexible approach to
landscape undertaken by the ruling dynasties (Smith, 2007:29). Kingdoms could have
prospered by asserting control over nodes through different strategies such as
369
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
“alliance”, “coercion” and the rewarding of areas who adhered to the systems of the
over-arching authority (Smith, 2007:33). This may have significance for the Eastern
Vākāṭaka land-grant tradition and the funding of Brahmana institutions within
established habitation areas, as the dynasty may have been creating or securing nodes
to act as production centres within more extensive networks. Again this could reflect
on a wider climate of potential instability and state dynamics, which kingdoms were
able to navigate more successfully by adjusting to the changing socio-economic and
political landscape (Smith, 2007:31).
If the Eastern Vākāṭaka polity operated within a flexible hinterland using various
networks to obtain resources then this could aid explanation of the apparent mobility
of the dynastic centre. The modes of exploitation in the landscape and the flexibility
of polity shows attributes similar to that of chiefdoms, which are usually considered
to have less permanent and rigid borders than states (Smith, 2007:31). Rāmṭek is the
centre of the Eastern Vākāṭaka polity where the strongest investment of expansionist
power would logically be expected however this is not the case in the data achieved
during the survey. The comparatively low investment of the Eastern Vākāṭakas in the
landscape could suggest their polity resembled a specialised chiefdom, although early
states seem to have had similar strategies for resource acquisition and control of their
territorial limits (Smith, 2007:31). Although seemingly lacked a strongly centralised
administration in a clearly defined territory, the Vākāṭaka polity did feature strong
sovereigns and the unequal distribution of production and surplus given that the kings
controlled the land and channelled production into the centre (Smith, 2003:90-91).
The land-grants also suggest systems used to restrict access to resources, obtain
labour and collect some form of taxation in order to support consumption by royal
and religious spheres, as in a state (Smith, 2003:92). In reality, a strict classification
of states and chiefdoms may not acknowledge the “heterogeneity of socio-cultural
formations” and the concepts of the ‘state’ may simply applied to particular
combinations of “powerful yet elusive methods” of arranging social systems (Smith,
2003:96). The Eastern Vākāṭaka kingdom may be better considered as an Early
Complex Society due to its kingship, social stratification, permanent institutions,
resource acquisition and socio-religious ideology (Smith, 2003:103-104).
Questions regarding state formation are relevant even to the imperial Guptas, as
Virkus has suggested that while their ideology was pervasive, their state may have
370
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
been loosely organised to incorporate diverse polities, site types and local economic
or administrative systems (Virkus, 1992). The Eastern Vākāṭaka polity may also have
incorporated local networks unified by ritual ideology. A key concept in Early
Complex Societies, and indeed traditional states, is that of the creation and
perpetuation of power through the legitimation of authority (Smith, 2003:105). In
more dispersed networks of sites associated with a ritual or political centre,
monumental constructions can be one method of visually demonstrating a particular
social order (Coningham, et al., 2007:18; Dunning, et al., 1999). Religious and royal
monumental expressions can create authority by communicating the king’s power and
administrative systems, which affects the population’s perception and experience of
the landscape (Duncan, 1990; Glatz, 2014:127). Monuments are also able to induce
and convey a communal cultural or religious identity and set of values to integrate the
people (Duncan, 1990:154; Kolb, 2014:154).
At Rāmṭek and Mansar the monumental state sanctuaries and ritual centre could be
viewed as an attempt to legitimise authority and cast a ‘kingly reading’ over the
hinterland. The ‘King’s reading’ refers to the creation of an ‘ideal’ landscape by a
particular model of kingship, which legitimises the king’s claim to power; for
example, the Aśokan model of kingship centres on the creation of religious structures
and public works within the landscape (Duncan, 1990:6). Monuments may be aimed
at communicating ideals of “political and religious life”, so that the king’s reading of
the landscape is conveyed and perceived by different members of the population
(Duncan, 1990:87).
371
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
is juxtaposed by Vidisha and the Sanchi area, and is described as being part of the
“wider suburban sprawl” of the city of Vidisha. Small habitation sites were
documented in Udayagiri’s environs, and the site is suggested to have served a ritual
function in the suburban environment (Shaw, 2007:131). While Nagardhan is not an
‘urban’ site in the same way as Vidisha, but there may be a similar relationship
between its landscape and the ritual centre of Rāmṭek. Agricultural sites in the
surrounding ‘conurbation’ may have created a support network to maintain the ritual
site, which acted as a focal point to legitimise Eastern Vākāṭaka authority.
A connection between the Guptas and Vākāṭakas has been widely theorised in
historical research but the available evidence does not conclusively support any direct
influence (for a discussion of the evidence see Goyal (1989), Goyal (2006) and
Shastri (1992a). The Vākāṭaka inscriptions demonstrate their shared ideals as the
Prakrit of early Vākāṭaka records was replaced by Sanskrit from the mid-4th century
AD (Pollock, 2006:289; Shastri, 1997:180). Mutual epigraphic conventions in the
representation of dynastic power and authority reflects the political ideology and
Sanskrit’s power in legitimising authority through the spread of Brahmanical culture
(Bhattacharya, 2014:25; Bronkhorst, 2011:47; Lubin, 2005:98; Sahu, 2001:14). Willis
372
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
has proposed that a “circle of kings” existed where despite regional autonomous
kingdoms, there was an overall ideologically coherent polity organised by common
values and practices (Willis, 2009:6).
The Guptas and Vākāṭakas seem to have had similar political systems and both
followed a model of kingship like that outlined by the Arthaśāstra, in which areas
with an opportunity for control were settled and the kings practised tolerance towards
local practices to subsequently lead to integration into the Brahmanical system
(Willis, 2009:159). The Brahmana settlements not only strengthened the agricultural
and economic development of an area but also assisted in compelling the people to
accept the new social framework through ritual integration (Willis, 2009:164). Willis
sees religious institutions as having been “dynamic” forces in structuring political
and economic relationships, driven by those with particular intentions, such as kings
priests and the upper echelons of society (Willis, 2009:9).
The Eastern Vākāṭakas and Guptas do seem to have had different economic responses
within their territories as the Guptas do not appear to have been frequent donors of
land unlike their subordinate kingdoms (Singh, 2008:495; 2009:160). Whereas
smaller dynasties may have used land-grants as high-value transactions, the Guptas
were heavily monetised and produced a large amount of precious currency (Raven,
2012). It may be that Gupta territory encompassed mostly established Buddhist
landscapes with an already developed and inherited land tenure system, meaning that
grants of land were unnecessary or impractical. Additionally, the Guptas’ imperial
presence may have reduced their need to establish power and consolidate territory in
their overarching landscape. Conversely, newer or smaller kingdoms may have
required the land-grant system to further organise the land and establish boundaries,
particularly if they were centralised in areas with interrelated but competing
kingdoms.
10.6. Conclusion
This study initially aimed to challenge the historical narrative of Eastern Vākāṭaka
expansion, which describes the colonisation of ‘peripheral’ areas to establish a new
social order, and the creation of an agrarian economy. While the data achieved from
373
Chapter 10. Rāmṭek and the Eastern Vākāṭakas
the Rāmṭek landscape initially seems to provide independent support for the historical
sources, it becomes apparent that the picture of landscape development is more
complex. This chapter has considered the evidence surrounding Rāmṭek’s function as
a socio-economic and religious site in the landscape in order to extrapolate
information about Eastern Vākāṭaka polity, administration and period of rule.
It appears that following their move to Rāmṭek, the Eastern Vākāṭakas reorganised
the landscape to support their state activities. The ritual function of Rāmṭek may be
similar to that at Udayagiri, where the religious centre was involved in the
legitimation of authority. Monumental expressions of power and ritual ideology
seem to have been used to integrate the local communities and legitimise Eastern
Vākāṭaka rule through the creation of shared worship. Furthermore, it appears the site
of Rāmṭek may have served an important economic function through the coordination
of production in the landscape so that the monumental constructions of the Vākāṭaka
kings were supported by a local economic network. The sites of Rāmṭek, Mansar and
Nagardhan may theoretically have acted as ‘nodes’ within local networks of resource
exploitation. The evidence from Nagardhan, which is central to the Eastern Vākāṭaka
territory, suggests that regional kingdoms may not have had a typical ‘urban’
response to state formation. The results of the survey add further weight to
suggestions that Vākāṭaka rule could have centred on key ritual or political sites,
which may not be deemed ‘urban’ in a traditional sense, but which utilised other
mechanisms to integrate the people and consolidate the regime’s power.
374
Chapter 11. Conclusion
This research began with the aim of contributing survey data towards the study of the
Eastern Vākāṭaka dynasty, by attempting to contextualise the ritual centre of Rāmṭek
in its surrounding landscape and characterise the nature of past settlement. Small-
scale survey was undertaken in an attempt to obtain meaningful data in the face of
logistical limitations, in an under-researched area. It was hoped that sufficient data
could be achieved to test the hypothesis of challenging the historical narrative of
regional landscape development under Eastern Vākāṭaka control. This research
therefore aimed to look beyond viewing Rāmṭek as purely a religious site to explore
its relationship to the landscape and establish its role in both economic and ritual
activity. With the aforementioned study having been concluded, this chapter
represents the summary of the main work embodied in chapters two to ten.
This research utilised a variety of approaches and datasets, including the use of
topographic maps and satellite imagery, ‘informed’ unsystematic survey and
systematic quadrat-walking to investigate the nature of the rural landscape at Rāmṭek.
This study was able to draw upon a number of influential projects conducted across
historical sites and their hinterlands, including Hawkes (2010); Lahiri (1996);
Morrison (2009); Shaw (2007) and Smith (2001b). The survey of the Rāmṭek
landscape enabled the recording of 444 archaeological features, including the remains
of ancient habitations and other material related to past settlement and religious
activity. Discussion of the resulting survey data focused on three major categories of
evidence; ceramics, sculpture and architectural remains. A consideration of these
categories of evidence allows the development of the Rāmṭek landscape to be viewed
as a case study for socio-economic trends in the Eastern Vākāṭaka kingdom, in order
to attempt to infer the nature of the Early Historic polity and regional trajectories of
economic or ritual development. Although this thesis outlined a number of problems
375
Chapter 11. Conclusion
which may be faced when working with a complicated dataset, it is hoped that the
potential for small-scale research to contribute meaningful data has been emphasised.
Primarily, this thesis has demonstrated that meaningful investigation into the
archaeology of the Early Historic Indian landscape, and the subsequent acquisition of
data, is both possible and advantageous. A key finding of this project is that data can
be collected on rural sites, despite the impact of modern activity in the swiftly
changing landscape. The discovery of unrecorded archaeological material deepens
our understanding of the nature of settlement and establishes important contextual
information, in order to situate the religious site in its hinterland. The rural landscape
should be a significant focus in the study of the Vākāṭakas, in order to consider the
socio-economic and political development of the region, given the role which rural
communities appear to have played in the transformation of the urban economy and
configuration of state. Future research should be encouraged as survey has the
potential to contribute to our understanding of rural settlement in the Early Historic
landscape, and it is a priority to record these archaeological remains while they are
still detectable. As the nature of the material culture in this period is not well
understood, there is great potential for a variety of research methods to be successful
in exploring the landscape. This study has presented one way of approaching the
questions we have of the archaeological record and thus makes a small contribution to
the expanding body of landscape research conducted in India. Based on the success of
this small-scale project in recovering usable data, it seems that greater resolution and
quality of data could be achieved with intensive systematic survey and the
collaboration of researchers with specialist skills.
This research has met its initial aim of contextualising Rāmṭek, to increase our
understanding of the site’s relationship to the landscape and consider the Vākāṭaka
period in context. This research has provided necessary archaeological data on the
settlement history of the Rāmṭek landscape to extend consideration of the site beyond
the monumental religious constructions and connect it to regional trends. The
importance of regional investigation is clear for elucidating widespread
transformation of politics and the economy. However, the ability to form a
comparative study with evidence from across India is limited by the lack of evidence
for the operation of other dynasties. The Western Vākāṭakas are lacking in studies of
their settlement history and economy, which may be addressed in future research.
376
Chapter 11. Conclusion
Their mention in this thesis was brief, but highlighted the potential complexity of
regional variation and the varied responses of the different dynasties to their
respective environments. Similarly, our socio-economic and political understanding
of the Guptas is remarkably limited despite their importance and proposed Imperial
presence. A route for further research would be to integrate the findings of regional
surveys and to consider other categories of remains, such as the dynastic epigraphic
records, as material evidence related to the political and economic functioning of
Early Complex Societies.
Discussion of chronology has been a major limitation in this project and will not be
resolved without more systematic approaches to archaeological investigation.
Although some limitations were certainly inherent to the research design, the ability
to refine chronologies and create a nuanced picture of development was hindered by
inadequate comparative material and published excavation data. There is great scope
for the application of absolute dating and archaeological science, which could target
confusions surrounding typologies and rate of change in Early Historic and Early
Medieval ceramic assemblages in particular. Ultimately, however, in order to resolve
issues faced in studies of this region and time period well-conducted excavation has
to be a priority. This project has demonstrated that surface studies can only achieve a
limited amount in the absence of reference material. Given the advances in theoretical
standpoints and archaeological practice, and a greater understanding of the potential
complications in research, the logical step is to undertake excavation and begin to
build the foundations of systematic research.
Theories can only be extrapolated from the evidence that was available or visible at
the time of the survey and a nuanced view of regional landscape development in the
Gupta-Vākāṭaka period remains elusive. I welcome further research to confirm or
refute these intimations. Indeed, a key message to be emphasised at the end of this
study is the great potential for future research, not only within this region but also
across India.
377
Chapter 11. Conclusion
This study sought to answer four research questions in order to challenge the
historical narrative of Eastern Vākāṭaka development and it is now possible to
address these, having outlined the results of the landscape survey and discussed the
economic and ritual significance of Rāmṭek in its hinterland related to the socio-
economic and political character of the Eastern Vākāṭaka dynasty.
The first research question sought to investigate the nature of ritual and economic
activity across the landscape prior to the major recorded period of c.4th to 5th centuries
AD. This research has identified material remains which are likely to relate to
prehistoric or emerging Early Historic activity, suggesting that it is possible to
identify evidence for past activity outside of the known periods of Vākāṭaka
expansion and later Medieval development. However, there is an absence of
substantial habitation remains or material related to formalised pre-Brahmanical
religious practice. I have argued that the Rāmṭek area, while subject to potentially
limited or transient settlement, does not appear to have been engaged in regional
networks of production during the prehistoric period. The combination of absences of
pan-Indian ceramics, local ceramics and extant habitation remains have been taken to
suggest the area potentially experienced a rural economy less reliant on production
networks and rooted in local, less-settled activity. A number of factors may be
affecting this picture of pre-Vākāṭaka Rāmṭek and there are certainly questions over
the visibility and survival of archaeological material; an existing settlement pattern
with a developing local economy prior to the Early Historic would not be out of place
in regional history, where ‘urban’ traits are traced back to Chalcolithic and Iron Age
developments. Visibility may be adversely affected by shallow ploughing, limiting
the ability of surface survey to detect features. Furthermore, even when early remains
are discovered, the nature of the evidence itself and of regional comparison material
can limit attempts to assign a reliable chronology. However, it does seem that at some
point in the Early Historic, the production, use and deposition of ceramics vastly
increased and there was a rise in the number and size of settlements, which indicates
economic change in the landscape.
378
Chapter 11. Conclusion
This leads us to our next two research questions: one regarding the nature of the
relationship between the nascent political and ritual centre of Rāmṭek to the
surrounding landscape and the extent of Vākāṭaka impact on settled hinterland; and
the other questioning whether the Eastern Vākāṭakas could be credited with the
economic transformation of the landscape. These were directed towards testing the
traditional narrative that implies the Vākāṭakas greatly invested in the dynastic centre,
as evidenced by the monumental constructions at Rāmṭek and Mansar, in order to
generate the Pan-Indian socio-economic and religious transformations of the 5th
century AD within this region.
Historically, the Vākāṭakas appear to have been the first unified power to directly
control the area but it has been unclear to what extent the establishment of the ritual
centre and state capitals initiated development in the local settlement, the rural
economy and production, as opposed to simply capitalising on existing systems.
There remains little conclusive evidence to suggest a previous political or religious
authority in the area. However, while the ceramic evidence could be interpreted as
indicating some intensification of settlement activity prior to Vākāṭaka rule, it seems
much of the rural landscape development did occur under their influence. The Phase
III Early Historic period under the Vākāṭakas demonstrates a high level of state
investment at the central sites and the first evidence of major institutionalised
religion. Furthermore, Nagardhan appears to have expanded to become a relatively
large habitation site at the centre of a local conurbation.
However, a key finding was that the royal investment was fairly localised so that
production and resource exploitation were subject to the demands of the ritual and
political centres and the agricultural landscape was engaged in local production to
fulfil domestic needs. Rather than the traditionally held, urban-centric model of state
formation, it has been proposed that religious sites could have acted as focal points in
the landscape and that other mechanisms were involved in the consolidation of
control. The ritual site of Rāmṭek may have been established to integrate the rural
landscape into Brahmanical society and legitimate royal authority, through its
position as an important node in local production and exchange networks and through
the communication of a pervasive shared ideology. It could be suggested that the
Eastern Vākāṭaka polity may reflect the proposed situation in the Gupta Empire
379
Chapter 11. Conclusion
where socio-economic and political formations were directed by the ritual ideology of
the kings.
Consideration of the final research question highlights this issue further by comparing
the levels of economic and ritual investment under the Eastern Vākāṭakas to that of
the Post-Vākāṭaka periods. There was clearly a widespread proliferation of permanent
religious and economic structures across the landscape at this time. This forces us to
reflect on the nature of the Eastern Vākāṭaka state. It seems that the ‘state’ focus was
drawn to incorporating rural and agricultural communities into prosperous local
production networks. Such a scenario may be connected to suggestions that the period
witnessed general political instability and so dynasties would have attempted to
secure their territorial control through investing in smaller cities (Sinopoli, 2001:170).
Finally, the seemingly intermittent investment in the dynastic centre of the Rāmṭek
hinterland may reflect the Eastern Vākāṭakas’ flexible consolidative approach to their
landscape. It may be theorised that the efforts of the Eastern Vākāṭaka regime were
more focused than expansionist, and directed towards the strengthening of regional
networks and maintenance of royal and ritual authority.
380
Bibliography
Bibliography
Abrams, M., Hook, S., Ramachandran, B., 2002. Aster User Handbook Version 2,
in: Technology, J.P.L.C.I.o. (Ed.).
Adams, R.M., 1965. Land Behind Baghdad, Chicago: University of Chicago Press.
Adams, R.M., 1981. The Heartland of Cities, Chicago: University of Chicago Press.
Agrawala, V.S., 1953. India as Known to Panini, Lucknow: University of Lucknow.
Alcock, S.E., Cherry, J.F., 2004. Side-by-Side Survey: Comparative Regional
Studies in the Mediterranean World, Oxford: Oxbow Books.
Ali, D., 2004. Courtly Culture and Political Life in Early Medieval India,
Cambridge: Cambridge University Press.
Ali, D., 2012. The Historiography of the Medieval in South Asia, Journal of the
Royal Asiatic Society 22 (1), 7-12.
Ali, Z., 2002. Iron Objects from Mansar Excavation, 1998, in: Margabandhu, C.,
Sharma, A.K., Bisht, R.S. (Eds.), Puraratna: Emerging Trends in Archaeology, Art,
Anthropology, Conservation and History (in Honour of Shri Jagat Pati Joshi),
Delhi: Agam Kala Prakashan, pp. 287-290.
Alizadeh, K., Ur, J., 2007. Formation and Destruction of Pastoral Irrigation
Landscapes on the Mughan Steppe, North-Western Iran, Antiquity 81 (311), 148-
160.
Altekar, A.S., 1960. The History of the Vākāṭakas, in: Yazdani, G. (Ed.), The Early
History of the Deccan: Parts I-Vi, London: Oxford University Press, pp. 149-200.
Ammerman, A.J., 1981. Surveys and Archaeological Research, Annual Review of
Anthropology 10 (1), 63-88.
Ammerman, A.J., 2004. Farewell to the Garden of Eden: Survey Archaeology after
the Loss of Innocence, British School at Athens Studies 11 177-182.
Ansari, Z.D., Dhavalikar, M.K., 1975. Excavations at Kayatha, Pune: Deccan
College, Post-Graduate and Research Institute.
Ansari, Z.D., Mate, M.S., 1966. Excavations at Dwarka, 1963, Pune: Deccan
College Post-Graduate and Research Institute.
Anschuetz, K.F., Wilshusen, R.H., Scheick, C.L., 2001. An Archaeology of
Landscapes: Perspectives and Directions, Journal of Archaeological Research 9 (2),
157-211.
Aryan, K.C., Aryan, S., 1975. Hanuman in Art and Mythology, Delhi: Rekha
Prakashan.
Bajpai, K.D., 1989. Indian Classical Art: The Vākāṭaka Phase, in: Trivedi, S.D.
(Ed.), Essays on Indology, Polity and Administration: In Honour of Padmabhushan
Sh. R.K. Trivedi, Delhi: Agam Kala Prakashan, pp. 103-110.
381
Bibliography
Bajpai, K.D., 1992. Location of Rāmagiri of Kalidasa, in: Shastri, A.M. (Ed.), The
Age of the Vakatakas, New Delhi: Harman Publishing House, pp. 85-91.
Bakker, H.T., 1989a. The Antiquities of Ramtek Hill, Maharashtra South Asian
Studies 5 (1), 79-102.
Bakker, H.T., 1989b. The Ramtek Inscriptions, Bulletin of the School of Oriental
and African Studies, University of London 52 (3), 467-496.
Bakker, H.T., 1990. Ramtek: An Ancient Centre of Viṣṇu Devotion in Maharashtra,
in: Bakker, H.T. (Ed.), The History of Sacred Places in India as Reflected in
Traditional Literature: Panels of the VIIth World Sanskrit Conference, Leiden: E.J.
Brill pp. 62-85.
Bakker, H.T., 1991. The Footprints of the Lord, in: Eck, D.L., Mallison, F. (Eds.),
Devotion Divine. Bhakti Traditions from the Regions of India. Studies in Honour of
Charlotte Vaudeville, Groningen: Egbert Forsten, pp. 19-37.
Bakker, H.T., 1992a. Memorials, Temples, Gods, and Kings. An Attempt to Unravel
the Symbolic Texture of Vākāṭaka Kingship, in: van den Hoek, A.W., Kolff, D.H.A.,
Oort, M.S. (Eds.), Ritual, State and History in South Asia. Essays in Honour of J.C.
Heesterman, Leiden, pp. 7-19.
Bakker, H.T., 1992b. Throne and Temple. Political Power and Religious Prestige in
Vidarbha, in: Bakker, H.T. (Ed.), The Sacred Centre as the Focus of Political
Interest: Proceedings of the Symposium Held on the Occasion of the 375th
Anniversary of the University of Groningen, 5-8 March 1989, Groningen: Egbert
Forsten, pp. 83-100.
Bakker, H.T., 1997. The Vākāṭakas: An Essay in Hindu Iconology, Groningen:
Egbert Forsten
Bakker, H.T., 2002. Religion and Politics in the Eastern Vakataka Kingdom, South
Asian Studies 18 1-24.
Bakker, H.T., 2004. The Vākāṭaka Heritage: Indian Culture at the Crossroads,
Groningen: Egbert Forsten.
Bakker, H.T., 2007a. Human Sacrifice (Puruṣamedha): Construction Sacrifice and
the Origin of the Idea of the 'Man of the Homestead' (Vāstupuruṣa), in: Bremmer,
J.N. (Ed.), The Strange World of Human Sacrifice, Leuven: Peeters Publishers, pp.
179-194.
Bakker, H.T., 2007b. Monuments to the Dead in Ancient North India, Indo-Iranian
Journal 50 11-47.
Bakker, H.T., 2008a. A Funerary Monument to Prabhāvatīguptā?, in: Bautze-Picron,
C. (Ed.), Religion and Art: New Issues in Indian Iconography and Iconology,
London: The British Academy, pp. 81-91.
Bakker, H.T., 2008b. Mansar and Its Eastern Neighbours: Architecture and the
Temples in Nagarā and Dakṣiṇa Kosala, in: Bakker, H. (Ed.), Mansar: The
Discovery of Pravaresvara and Pravarapura, Temple and Residence of the
382
Bibliography
383
Bibliography
384
Bibliography
385
Bibliography
Britt Bousman, C., 2006. Satellite Archaeology for Everyone, The SAA
Archaeological Record: The Society for American Archaeology pp. 32-34.
Bronkhorst, J., 2011. Buddhism in the Shadow of Brahmanism, Leiden: Brill.
Brown, R.L., 2004. Vākāṭaka-Period Hindu Sculpture, in: Bakker, H.T. (Ed.), The
Vākāṭaka Heritage: Indian Culture at the Crossroads, Groningen: Egbert Forsten
pp. 59-69.
Campana, S., Forte, M., Liuzza, C., 2010. Space, Time, Place. Third International
Conference on Remote Sensing in Archaeology, Bar International Series:
Archaeopress.
Caraher, W.R., 2014. Patronage and Reception in the Monumental Architecture of
Early Christian Greece, in: Osborne, J.F. (Ed.), Approaching Monumentality in
Archaeology Albany: State University of New York Press, pp. 39-56.
Carreté, J.-M., Keay, S., Millett, M., 1995. A Roman Provincial Capital and Its
Hinterland: The Survey of the Territory of Tarragona, Spain, 1985-1990, Journal of
Roman Archaeology, Supplementary Series 15.
Casile, A., 2009. Temples Et Expansion D'un Centre Religieux En Inde Centrale:
Lectures Du Paysage Archéologique De Badoh-Paṭhāri Du 5 ͤ an 10 ͤ Siècle De Notre
Ère, Paris: University Sorbonne Nouvelle.
Casile, A., 2014. Changing Religious Landscapes in Gupta Times: Archaeological
Evidence from the Area of Baḍoh-Paṭāri in Central India, South Asian Studies 30
(2), 245-268.
Chakrabarti, D.K., 2001. Archaeological Geography of the Ganga Plain: The Lower
and the Middle Ganga, New Delhi: Orient Blackswan.
Chakrabarti, D.K., Lahiri, N., 1988. A Preliminary Report on the Archaeology of the
Ranchi District with a Note on the Asura Sites, Man and Environment XXII 29-54.
Chakrabarti, D.K., Lahiri, N., 1994. The Iron Age in India: The Beginning and
Consequences, Puratattva 24 12-32.
Chakrabarti, D.K., Tewari, R., Singh, R.N., 2003. From the Ganga Plain to the
Eastern and Western Deccan: A Field Study of the Ancient Routes, South Asian
Studies 19 57-72.
Chakrabarti, D.K., Tewari, R., Singh, R.N., Singh, K.K., Trivedi, R., 2002. Kampil,
Sankisa and Chakranagar: Sites in the Farrukhabad-Etah-Budaun-Mainpuri-Etawah
Region of Uttar Pradesh, South Asian Studies 18 109-119.
Chakravorty, S.L., 2001. Artisanal and Small-Scale Mining in India: International
Institute for Environment and Development, pp. 1-81.
Chari, C.N., Marawar, S.A., Kulkarni, S.G., Bhagwatwar, D.A., Bobde, S.P., 1975.
Geology and Mineral Resources of Maharashtra, Nagpur: Government of
Maharashtra, Directorate of Geology and Mining.
386
Bibliography
387
Bibliography
388
Bibliography
Dass, M.I., Willis, M., 2002. The Lion Capital from Udayagiri and the Antiquity of
Sun Worship in Central India, South Asian Studies 18 25-45.
Datye, V.S., Dhorde, A., 2002. Agriculture in Maharashtra, in: Diddee, J., Jog, S.R.,
Kale, V.S., Datye, V.S. (Eds.), Geography of Maharashtra, Jaipur: Rawat
Publications, pp. 159-179.
Dayal, S., 2005. Towns, Villages and Desertion: Exploring the Early Medieval
Phase in Indian Archaeology, Man and Environment XXX (1), 61-66.
Dayalan, D., 2010. Conservation and Environmental Issues of Taj Mahal, in:
Campana, S., Forte, M., Liuzza, C. (Eds.), Space, Time, Place. Third International
Conference on Remote Sensing in Archaeology: Archaeopress, pp. 221-224.
Deglurkar, G.B., 1974. Temple Architecture and Sculpture of Maharashtra, Nagpur:
Nagpur University.
Deo, S.B., 1973. Mahurjhari Excavation, 1970-1972, Nagpur: Nagpur University.
Deo, S.B., 1985. The Megaliths: Their Culture, Ecology, Economy and Technology,
in: Deo, S.B., Paddayya, K. (Eds.), Recent Advances in Indian Archaeology, Pune:
Deccan College Research Institute, pp. 89-99.
Deo, S.B., Dhavalikar, M.K., 1968. Paunar Excavation, 1967, Nagpur: Nagpur
Vidyapeeth Mudranalaya.
Deo, S.B., Gupte, R.S., 1974. Excavations at Bhokardhan (Bhogavardhana),
Nagpur: Nagpur University
Deo, S.B., Joshi, J.P., 1972. Pauni Excavations (1969-1970), Nagpur.
Déodat, L., Lecocq, P., 2012. Using Google Earth and Gis to Survey in the Peruvian
Andes, in: Kluiving, S., Guttman-Bond, E. (Eds.), Landscape Archaeology between
Art and Science: From a Multi- to an Interdisciplinary Approach, Amsterdam:
Amsterdam University Press, pp. 321-339.
Deol, R.S., 2011. Census of India, Provisional Population Totals, in: Operations,
D.o.C. (Ed.), Mumbai: Director of Census Operations, Maharashtra.
Deotare, B., 2007. Preliminary Report on the Excavations at Bhon and Paturda,
Buldana District, Maharashtra, Bulletin of the Deccan College Post-Graduate and
Research Institute 64-65 87-105.
Deotare, B., Shete, G., Sawant, R., Kathale, V., Naik, S., 2007. Discovery of
Structural Stupa at Bhon, District Buldana, Maharashtra, Puratattva 37 176-185.
Deotare, B.C., Joshi, P.S., Parchure, C.N., 2013. Glimpses of Ancient Maharashtra
through Archaeological Excavations, Pune: Bharatiya Itihas Snkalan Samitee &
Deccan College, Post-Graduate and Research Institute.
Deshpande, C.D., 1971. Geography of Maharashtra, New Delhi: National Book
Trust, India.
389
Bibliography
390
Bibliography
Elgood, H., 2004. Exploring the Roots of Village Hinduism in South Asia, World
Archaeology 36 (3), 326-342.
Erdosy, G., 1985. Settlement Archaeology of the Kausambi Region, Man and
Environment IX 66-79.
Erdosy, G., 1988. Urbanisation in Early Historic India, Oxford: B.A.R.
International Series 430.
Erickson, P.A., Murphy, L.D., 2008. A History of Anthropological Theory, Toronto:
University of Toronto Press.
Eschmann, A., 1978. Hinduization of Tribal Deities in Orissa: The Śākta and Śaiva
Typology, in: Eschmann, A., Kulke, H., Tripathi, G.C. (Eds.), The Cult of Jagannath
and the Regional Tradition of Orissa, New Delhi: Manohar Publications, pp. 79-98.
Evans, D., Fletcher, R., 2015. The Landscape of Angkor Wat Redefined, Antiquity
89 (348), 1402-1419.
Evans, D., Pottier, C., Fletcher, R., Hensley, S., Tapley, I., Milne, A., Barbetti, M.,
2007. A Comprehensive Archaeological Map of the World's Largest Preindustrial
Settlement Complex at Angkor, Cambodia, Proceedings of the National Academy of
Sciences 104 (36), 14277-14282.
Fairservis, W.A., 1956. Excavations in the Quetta Valley, West Pakistan,
Anthropological Papers of the American Museum of Natural History 45 (2), 169-
402.
Fairservis, W.A., 1959. Archaeological Surveys in the Zhob and Loralai District,
West Pakistan, Anthropological Papers of the American Museum of Natural History
47 (2), 277-448.
Falk, H., 2006. Aśokan Sites and Artefacts: A Source-Book with Bibliography,
Mainz am Rhein Verlag Philipp von Zabern.
Falk, H., 2009. Two Dated Satavahana Epigraphs, Indo-Iranian Journal 52 197-206.
Farr, T.G., Rosen, P.A., Caro, E., Crippen, R., Duren, R., Hensley, S., Kobrick, M.,
Paller, M., Rodriguez, E., Roth, L., 2007. The Shuttle Radar Topography Mission,
Reviews of Geophysics 45 (2).
Fergusson, J., Burgess, J., 1880. The Cave Temples of India, London: W.H. Allen &
Co. .
Fish, S.K., Kowalewski, S.A., 1990. The Archaeology of Regions: A Case for Full-
Coverage Survey, Smithsonian Series in Archaeological Inquiry, Washington, DC:
Smithsonian Institution Press.
Fisher, K.D., 2014. The Creation and Experience of Monumentality on Protohistoric
Cyprus, in: Osborne, J.F. (Ed.), Approaching Monumentality in Archaeology,
Albany: State University of New York Press, pp. 355-406.
Flam, L., 1976. Settlement, Subsistence and Population: A Dynamic Approach to the
Development of the Indus Valley Civilisation, in: Kennedy, K.A.R., Possehl, G.L.
391
Bibliography
392
Bibliography
Galiatsatos, N., Wilkinson, T.J., Donoghue, D.N.M., Philip, G., 2009. The Fragile
Crescent Project (Fcp): Analysis of Settlement Landscapes Using Satellite Imagery,
Computer Applications to Archaeology, Williamsburg, Virginia, USA, pp. 1-15.
Gallant, T.W., 1986. “Background Noise” and Site Definition: A Contribution to
Survey Methodology, Journal of Field Archaeology 13 (4), 403-418.
Ghose, B., Kar, A., Husain, Z., 1979. The Lost Courses of the Saraswati River in the
Great Indian Desert: New Evidence from Landsat Imagery, The Geographical
Journal 145 (3), 446-451.
Ghosh, A., 1952. The Rajpatuna Desert: Its Archaeological Aspect, Bulletin of the
National Institute of Sciences in India 1 37-42.
Ghosh, A., 1953. Exploration in Bikaner, East and West 4 (1), 31-34.
Ghosh, A., 1964. Indian Archaeology - a Review, 1961-1962, New Delhi:
Archaeological Survey of India, Government of India
Ghosh, A., 1973. The City in Early Historical India, New Delhi: Indian Institute of
Advanced Study in association with Munshiram Manoharlal Publishers Pvt. Ltd.
Ghosh, A., Panigrahi, K.C., 1946. The Pottery of Ahichchhatra, District Bareilly,
U.P. , Ancient India 1 37-59.
Ghoshal, U.N., 1954. Political Theory and Administrative Organisation, in:
Majumdar, R.C., Pusalker, A.D. (Eds.), The History and Culture of the Indian
People: The Classical Age, Bombay: Bharatiya Vidya Bhavan, pp. 341-355.
Giddens, A., 1979. Central Problems in Social Theory, Berkeley and Los Angeles:
University of California Press.
Girhe, K.M., 2002. Temple Architecture of Nagpur under Bhoslas, in:
Margabandhu, C., Sharma, A.K., Bisht, R.S. (Eds.), Puraratna: Emerging Trends in
Archaeology, Art, Anthropology, Conservation and History (in Honour of Shri Jagat
Pati Joshi), Delhi: Agam Kala Prakashan, pp. 546-554.
Gkiasta, M., 2008. Historiography of Landscape Research on Crete: Leiden
University Press.
Glatz, C., 2014. Monuments and Landscape: Exploring Issues of Place, Distance,
and Scale in Early Political Contest, in: Osborne, J.F. (Ed.), Approaching
Monumentality in Archaeology, Albany: State University of New York Press, pp.
109-134.
Gogte, V.D., 1982. Megalithic Iron Smelting at Naikund (Part II): Efficiency of Iron
Smelting by Chemical Analysis, in: Deo, S.B., Jamkhedkar, A.P. (Eds.), Excavations
at Naikund, 1978-1980, Bombay: Department of Archaeology and Museums, pp. 56-
59.
Gogte, V.D., 2006. Excavations at Chaul, Annual Report of the Deccan College
Post-Graduate and Research Institute, 2004-2005 68-72.
393
Bibliography
394
Bibliography
Haslam, M., Korisettar, R., Petraglia, M., Smith, T., Shipton, C., Ditchfield, P.,
2010. In Foote's Steps: The History, Significance and Recent Archaeological
Investigation of the Billa Surgam Caves in Southern India, South Asian Studies 26
(1), 1-19.
Hawkes, J., 2006. The Buddhist Stupa of Bharhut and Its Sacred and Secular
Geographies, Department of Archaeology, Unpublished: University of Cambridge.
Hawkes, J., 2008. Bharhut: A Reassessment, South Asian Studies 24 1-14.
Hawkes, J., 2009. The Wider Archaeological Contexts of the Buddhist Stūpa Site of
Bharhut, in: Hawkes, J., Shimada, A. (Eds.), Buddhist Stupas in South Asia: Recent
Archaeological, Art-Historical and Historical Perspectives, New Delhi: Oxford
University Press, pp. 146-174.
Hawkes, J., 2010. Bharhut and Its Wider Regional Context, in: Callieri, P., Colliva,
L. (Eds.), South Asian Archaeology, 2007: Proceedings of the 19th Meeting of the
European Association of South Asian Archaeology in Ravenna, Italy, 2-6 July 2007,
Oxford, pp. 133-144.
Hawkes, J., 2014a. Bharhut as an Archaeological Area, in: Chakrabarti, D.K., Lal,
M. (Eds.), History of Ancient India: Political History and Administration (C. 200bc -
Ad750) (from the End of the Mauryan Rule to the Beginning of the Dominance of
Regional Powers), New Delhi: Aryan Books International, pp. 351-366.
Hawkes, J., 2014b. Chronological Sequences and the Problem of Early Medieval
Settlement in India, Puratattva 44 208-228.
Hawkes, J., 2014c. Finding the "Early Medieval" in South Asian Archaeology, Asian
Perspectives 53 (1), 53-96.
Hecht, S.B., Morrison, K.D., Padoch, C., 2014. The Social Lives of Forests: Past,
Present, and Future of Woodland Resurgence, Chicago: University of Chicago Press.
Hegde, K.T.M., Sonawane, V.H., 1986. Landscape and Settlement Pattern of
Harappa Culture Villages in the Rupen Estuary, Man and Environment X 23-32.
Hegewald, J.A.B., 2002. Water Architecture in South Asia: A Study of Types,
Development and Meanings, Leiden: Brill.
Heitzman, J., 1991. Ritual Polity and Economy: The Transactional Network of an
Imperial Temple in Medieval South India, Journal of the Economic and Social
History of the Orient 34 (1/2), 23-54.
Heitzman, J., 1997. Gifts of Power: Lordship in an Early Indian State, New Delhi:
Oxford University Press.
Heitzman, J., 2009. The Urban Context of Early Buddhist Monuments in South
Asia, in: Hawkes, J., Shimada, A. (Eds.), Buddhist Stupas in South Asia: Recent
Archaeological, Art-Historical and Historical Perspectives, New Delhi: Oxford
University Press, pp. 192-215.
Herman, C.F., Krishnan, K., 1994. Micaceous Red Ware: A Gujarat Proto-Historic
Cultural Complex or Just Ceramics?, in: Parpola, A., Koskikallio, P. (Eds.), South
395
Bibliography
396
Bibliography
397
Bibliography
Kervran, M., 1996. Indian Ceramics in Southern Iran and Eastern Arabia: Repertory,
Classification and Chronology, in: Ray, H.P., Salles, J.-F. (Eds.), Tradition and
Archaeology: Early Maritime Contacts in the Indian Ocean. Proceedings of the
International Seminar Techno-Archaeological Perspecives of Seafaring in the
Indian Ocean, 4th Century B.C. To 15th Century A.D. New Delhi, February 28-
March 4 1994, New Delhi: Manohar Publishers and Distributors, pp. 37-58.
Ketkar, D.J., 1970. Subsurface Conditions - Bombay Island, Symposium on
subsurface exploration and foundations of structures in the Bombay region, Indian
National Society for Soil Mechanics and Foundation Engineering 1.
Khaladkar, V., 2007-2008. Archaeological Investigations in the Middle Bhima
Basin, Maharashtra: A Preliminary Report Puratattva 38 24-38.
Khare, D.H., 1982. Memorial Stones in Maharashtra, in: Settar, S., Sontheimer, G.D.
(Eds.), Memorial Stones: A Study of Their Origin, Significance and Variety,
Dharwad: Institute of Indian Art History, Karnatak University, pp. 251-254.
Khare, M.D., 1969. Study of Pottery - a New Approach, Abc of Archaeology or a to
Z of Archaeology, in: Sinha, B.P. (Ed.), Potteries in Ancient India, Patna: The
Department of Ancient Indian History and Archaeology, Patna University, pp. 27-
29.
Knapp, A.B., Ashmore, W., 1999. Archaeological Landscapes: Constructed,
Conceptualised, Ideational, in: Ashmore, W., Knapp, A.B. (Eds.), Archaeologies of
Landscape: Contemporary Perspectives, Oxford: Blackwell Publishers Ltd pp. 1-30.
Kolb, M.J., 2014. Monumentality among the Mediterranean Isles, in: Osborne, J.F.
(Ed.), Approaching Monumentality in Archaeology, Albany: State University of
New York Press, pp. 153-180.
Kowalewski, S.A., 2008. Regional Settlement Pattern Studies, Journal of
Archaeological Research 16 (3), 225-285.
Kramrisch, S., 1946a. The Hindu Temple, Calcutta: University of Calcutta.
Kramrisch, S., 1946b. The Hindu Temple, Calcutta: University of Calcutta.
Kramrisch, S., 1956. Image of Aditi-Uttānapad, Artibus Asiae 19 (3/4), 259-270.
Kramrisch, S., 1983. Unknown India: Ritual Art in Tribe and Village, in: Miller,
B.S. (Ed.), Exploring India's Sacred Art: Selected Writings of Stella Kramrisch,
Philadelphia: University of Pennsylvania Press, pp. 85-120.
Krishnan, M.S., 1982. Geology of India and Burma, 6th ed., Delhi: CBS Publishers
and Distributors.
Kropman, M., 2008. The Seals and Inscriptions from Mansar, in: Bakker, H.T. (Ed.),
Mansar: The Discovery of Pravaresvara and Pravarapura, Temple and Residence of
the Vakataka King Pravarasena II. Proceedings of a Symposium at the British
Museum, 30 June – 1 July 2008., Groningen: Library of the University of
Groningen.
Kulkarni, P., 2003. Coins of Vakatakas, Numismatic Digest 25-26 65-80.
398
Bibliography
Kulke, H., 1978. Royal Temple Policy and the Structure of Medieval Hindu
Kingdoms, in: Eschmann, A., Kulke, H., Tripathi, G.C. (Eds.), The Cult of
Jagannath and the Regional Tradition of Orissa, New Delhi: Manohar Publications
pp. 125-138.
Kulke, H., 1993. Kings and Cults: State Formation and Legitimation in India and
Southeast Asia: Manohar Publishers & Distributors.
Kulke, H., 2004. Some Thoughts on State and State Formation under the Eastern
Vakatakas, in: Bakker, H.T. (Ed.), The Vakataka Heritage: Indian Culture at the
Crossroads, Groningen: Egbert Forsten, pp. 1-10.
Lahiri, N., 1996. Archaeological Landscapes and Textual Images: A Study of the
Sacred Geography of Late Medieval Ballabgarh, World Archaeology 28 (2), 244-
264.
Lahiri, N., Bacus, E., 2004. Exploring the Archaeology of Hinduism, World
Archaeology 36 (3), 313-325.
Lahiri, N., Sethi, V., Purushartha, B., 2002. Historical Archaeology of India: An
Outline of the Work of the Archaeological Survey of India, in: Settar, S., Korisettar,
R. (Eds.), Archaeology and Historiography: History, Theory and Method, New
Delhi: Manohar Publishers & Distributors, pp. 71-116.
Lahiri, N., Singh, U., 1999. In the Shadow of New Delhi: Understanding the
Landscape through Village Eyes, in: Ucko, P.J., Layton, R. (Eds.), The Archaeology
and Anthropology of Landscape: Shaping Your Landscape, London: Routledge, pp.
175-188.
Lahiri, N., Singh, U., Uberoi, T., 1996. Preliminary Field Report on the Archaeology
of Faridabad - the Ballabgarh Tehsil, Man and Environment XXI (1), 32-57.
Lal, B.B., 1954-55. Excavation at Hastinapura and Other Explorations in the Upper
Ganga and Sutlej Basins 1950-52: New Light on the Dark Age between the End of
the Harappa Culture and the Early Historical Period, Ancient India 10 & 11 5-151.
Lal, M., 1984a. Settlement History and Rise of Civilization in Ganga-Yamuna Doab,
from 1500 B.C. To 300 A.D: B.R. Publishing Corp.
Lal, M., 1984b. Summary of Four Seasons of Exploration in Kanpur District, Uttar
Pradesh, Man and Environment VIII 61-80.
Lane, P., 1986. Past Practices in the Ritual Present: Examples from the Welsh
Bronze Age, Archaeological Review from Cambridge 5 (2), 181-192.
Llobera, M., 2007. Reconstructing Visual Landscapes, World Archaeology 39 (1),
51-69.
Lubin, T., 2005. The Transmission, Patronage, and Prestige of Brahmanical Piety
from the Mauryas to the Guptas, in: Squarcini, F. (Ed.), Boundaries, Dynamics and
Construction of Traditions in South Asia, Firenze: Firenze University Press, pp. 77-
103.
399
Bibliography
Lucero, L.J., Fletcher, R., Coningham, R.A.E., 2015. From 'Collapse' to Urban
Diaspora: The Transformation of Low-Density, Dispersed Agrarian Urbanism,
Antiquity 89 (347), 1139-1154.
Lutgendorf, P., 2007. Hanuman's Tale: The Messages of a Divine Monkey, Oxford:
Oxford University Press.
Madella, M., Ajithprasad, P., Lancelotti, C., Rondelli, B., Balbo, A., French, C.,
Rodriguez, D., García-Granero, J.J., Yannitto, V., Rajesh, S.V., Gadekar, C.S., Briz,
I., 2010. Social and Environmental Transitions in Arid Zones: The North Gujarat
Archaeological Project - Nogap, Antiquity Project Gallery 084 (325).
Madry, S., 2007. An Evaluation of Google Earth for Archaeological Aerial
Prospection and Site Survey, in: Clark, J.T., Hagemeister, E.M. (Eds.), Digital
Discovery: Exploring New Frontiers in Human Heritage. Proceedings of the 34th
Conference of Computer Applications and Quantitative Methods in Archaeology,
Fargo, United States, April 2006, Budapest: Archaeolingua.
Mahajan, M., 1992. Administrative Divisions under the Vākāṭakas, in: Shastri, A.M.
(Ed.), The Age of the Vākāṭakas, New Delhi: Harman Publishing House, pp. 65-84.
Mahapatra, S.K., 1994. Indian Archaeology - a Review, 1989-1990, New Delhi:
Archaeological Survey of India, Government of India.
Mahapatra, S.K., 1995. Indian Archaeology - a Review, 1990-1991, New Delhi:
Archaeological Survey of India, Government of India
Maity, S.K., 1970. Economic Life in Northern India in the Gupta Period (Cir. A.D.
300-550), Delhi: Motilal Banarsidass.
Majumdar, R.C., Altekar, A.S., 1967. The Vakataka-Gupta Age (Circa 200-550
A.D.), in: Sarkar, J. (Ed.), A New History of the Indian People, 3 ed., Delhi: Motilal
Banarsidass Publishers Private Limited.
Majumdar, R.C., Pusalker, A.D., 1954. The History and Culture of the Indian
People: The Classical Age, Bombay: Bharatiya Vidya Bhavan.
Malik, S.C., 1973. The Role of Theory in the Study of Archaeology in India,
Purātattva: Bulletin of the Indian Archaeological Society 6 1-11.
Mani, B.R., Banerji, A., Bhatt, P., 2003. Indian Archaeology - a Review, 1997-1998,
New Delhi: Archaeological Survey of India, Government of India.
Manjhi, H., Dorji, C., Banerji, A., 2000. Indian Archaeology - a Review, 1994-1995,
New Delhi: Archaeological Survey of India, Government of India.
Margabandhu, C., 1985. Archaeology of the Satavahana-Kshatrapa Times, Delhi:
Sundeep Prakashan.
Mate, M.S., 1982. Hero-Stones: The 'Folk' and the 'Classic', in: Settar, S.,
Sontheimer, G.D. (Eds.), Memorial Stones: A Study of Their Origin, Significance
and Variety, Dharwad: Institute of Indian Art History, Karnatak University, pp. 77-
135.
400
Bibliography
Mate, M.S., 1990. The Clay Feet?, Bulletin of the Deccan College Post-Graduate
and Research Institute 49 243-249.
Mate, M.S., 2005. Archaeology of Medieval India, Delhi: B.R. Publishing
Corporation.
Mattingly, D., 2000. Methods of Collection, Recording and Quantification, in:
Francovich, R., Patterson, H., Barker, G. (Eds.), Extracting Meaning from
Ploughsoil Assemblages, Oxford: Oxbow Books, pp. 5-15.
Mehta, R.N., 1955. Excavations at Timbarva (Baroda District) April, 1953, Baroda:
Maharaja Sayajirao University of Baroda.
Mehta, R.N., 1979. Mediaeval Archaeology, Delhi Ajanta Publications
Mehta, R.N., Chowdary, S.N., 1966. Excavation at Devnimori (a Report of the
Excavation Conducted from 1960 to 1963), Baroda: The Maharaja Sayajirao
University of Baroda Press.
Mehta, R.N., Patel, S.A.J., 1967. Excavation at Shamalaji, Baroda: Maharaja
Sayajirao University of Baroda Press.
Meister, M.W., Dhaky, M.A., 1986. Encyclopaedia of Indian Temple Architecture:
South India, Upper Drāviḍadéśa, Early Phase, A.D. 500-1075, Philadelphia:
University of Pennsylvania Press.
Meister, M.W., Dhaky, M.A., Deva, K., 1988. Encyclopaedia of Indian Temple
Architecture: North India, Foundations of North Indian Style, Delhi: American
Institute of Indian Studies, Oxford University Press.
Menon, J., 2008. Archaeology of Early Historic South Asia: A Review, in:
Sengupta, G., Chakraborty, S. (Eds.), Archaeology of Early Historic South Asia,
New Delhi: Pragati Publications with Centre for Archaeological Studies and
Training, Eastern India, pp. 15-38.
Menon, K.G., 2002. Indian Archaeology - a Review, 1996-1997, New Delhi:
Archaeological Survey of India, Government of India.
Meshram, P., 2010. Vakataka Coins: Not Myth, a Reality, The Journal of the
Numismatic Society of India LXXII (1), 34-41.
Michell, G., 2000. Hindu Art and Architecture, London: Thames & Hudson
Miller, D., 1985. Artefacts as Categories: A Study of Ceramic Variability in Central
India, Cambridge: Cambridge University Press.
Millett, M., 2000. Dating, Quantifying and Utilising Pottery Assemblages from
Surface Survey, in: Francovich, R., Patterson, H., Barker, G. (Eds.), Extracting
Meaning from Ploughsoil Assemblages, Oxford: Oxbow Books, pp. 53-59.
Mirashi, V.V., 1963. Inscriptions of the Vakatakas, Ootacamund: Government
Epigraphist for India.
Mirashi, V.V., 1964. Ramagiri of Kalidasa, Journal of Indian History 42 131-143.
401
Bibliography
402
Bibliography
Culture. Professor S.B. Deo Felicitation Volume, Delhi: Agam Kala Prakashan, pp.
157-169.
Mohanty, R.K., Smith, M.L., Matney, T., 2014. Excavations and Geophysical
Survey at the Early Historic Town of Talapada, Odisha (2013), Man and
Environment XXXIX (2), 53-63.
Mohanty, R.K., Walimbe, S.R., 1993. A Demographic Approach to the Vidarbha
Megalithic Culture, Man and Environment XVIII (2), 93-103.
Morrison, K.D., 1993. Supplying the City: The Role of Reservoirs in an Indian
Urban Landscape, Asian Perspectives 32 (2), 133-151.
Morrison, K.D., 1995. Trade, Urbanism and Agricultural Expansion: Buddhist
Monastic Institutions and the State in the Early Historic Western Deccan World
Archaeology 27 (2), 203-221.
Morrison, K.D., 2005. Brahmagiri Revisited: A Re-Analysis of the South Indian
Sequence, in: Jarrige, C., LeFèvre, V. (Eds.), South Asian Archaeology 2001, Paris:
Editions Recherche sur les Civilisations - ADPF, pp. 257-261.
Morrison, K.D., 2009. Daroji Valley: Landscape, Place and the Making of a
Dryland Reservoir System, New Delhi: Manohar Publishers & Distributors.
Morrison, K.D., Junker, L.L., 2002. Forager-Traders in South and Southeast Asia:
Long-Term Histories, Cambridge: Cambridge University Press.
Mueller, J.W., 1974. The Use of Sampling in Archaeological Survey, Memoirs of the
Society for American Archaeology 28 i-91.
Mughal, M.R., 1993. Recent Archaeological Research in the Cholistan Desert, in:
Possehl, G.L. (Ed.), Harappan Civilisation: A Recent Perspective, 2nd ed., New
Delhi: Oxford & IBH Publishing Co. Pvt. Ltd. , pp. 85-95.
Mughal, M.R., 1997. A Preliminary Review of Archaeological Surveys in Punjab
and Sindh: 1993-1995, South Asian Studies 13 241-249.
Muley, P., Jain, P.K., Jain, R., 2010. Experimental Studies on Utilisation of Murrum
as Hard Shoulder Material, International Journal of Engineering Science and
Technology 2 (9), 4896-4901.
Nagar, S.L., 1995. Hanumān in Art, Culture, Thought and Literature, New Delhi:
Intellectual Publishing House.
Nagata, K., 2008. The Problems in the Identification of Gaṇa-Like Images from
Mansar: Is It Śiva or Gaṇa?, in: Bakker, H.T. (Ed.), Mansar: The Discovery of
Pravaresvara and Pravarapura, Temple and Residence of the Vakataka King
Pravarasena II. Proceedings of a Symposium at the British Museum, 30 June – 1
July 2008. , Groningen: Library of the University of Groningen.
Nalini, N.S., Rajani, M.B., 2010. The Importance of Precision in Georeferencing
Demonstrated in the Case Study of Archaeological Gis of Chitradurga, Karnataka,
in: Campana, S., Forte, M., Liuzza, C. (Eds.), Space, Time, Place. Third
403
Bibliography
404
Bibliography
405
Bibliography
Pathak, K.B., Dikshit, K.N., 1925. Poona Plates of the Vakataka Queen Prabhavati-
Gupta: The 13th Year, in: Thomas, F.W. (Ed.), Epigraphia Indica and Record of the
Archaeological Survey of India, 1919-1920, Calcutta: Director of the Archaeological
Survey of India. Government of India, Central Publication Branch, pp. 39-44.
Petraglia, M.D., Korisettar, R., Noll, M., Schuldenrein, J., 2003. An Extensive
Middle Palaeolithic Quarry Landscape in the Kaladgi Basin, Southern India,
Antiquity, Project Gallery.
Petrie, C.A., Singh, R.N., 2008. Investigating Cultural and Geographical
Transformations from the Collapse of Harappan Urbanism to the Rise of the Great
Historic Cities: A Note on the Land, Water and Settlement Project, South Asian
Studies 24 37-38.
Philip, G., Donoghue, D., Beck, A., Galiatsatos, N., 2002. Corona Satellite
Photography: An Archaeological Application from the Middle East, Antiquity. 76
(291), 109-118.
Pillai, A., Dikshit, O., Rai, G.K., Fonia, R.S., Prabhakar, V.N., Vikram, B., 2010.
Documentation and Visualisation of Archaeological Data, in: Campana, S., Forte,
M., Liuzza, C. (Eds.), Space, Time, Place. Third International Conference on
Remote Sensing in Archaeology: Archaeopress, pp. 379-392.
Pillai, S.D., 1997. Indian Sociology through Ghurye: A Dictionary, Mumbai:
Popular Prakashan.
Pinto-Orton, N., 2013. Red Polished Ware in Gujarat: Surface Collections from
Inland Sites, in: Abraham, S.A., Gullapalli, P., Raczek, T.P., Rizvi, U.Z. (Eds.),
Connections and Complexity: New Approaches to the Archaeology of South Asia,
California: Left Coast Press Inc. , pp. 195-222.
Plog, S., 1976. Relative Efficiencies of Sampling Techniques for Archaeological
Surveys in: Flannery, K. (Ed.), The Early Mesoamerican Village, New York:
Academic Press, pp. 136-158.
Plog, S., Plog, F., Wait, W., 1978. Decision Making in Modern Surveys, Advances
in Archaeological Method and Theory 1 (ArticleType: research-article / Full
publication date: 1978 / Copyright © 1978 Springer), 383-421.
Pollock, S., 2006. The Language of the Gods in the World of Men, Berkeley:
University of California Press.
Possehl, G.L., 1980. Indus Civilisation in Saurashtra, Delhi: B.R. Publishing
Corporation.
Possehl, G.L., 2002. Fifty Years of Harappan Archaeology: The Study of the Indus
Civilisation since Indian Independence, in: Settar, S., Korisettar, R. (Eds.),
Protohistory: Archaeology of the Harappan Civilisation, New Delhi: Manohar
Publishers & Distributors, pp. 1-46.
Prasad, B.N., 2012. Brahmanical Temples, Maṭhas, Agrahāras and a Buddhist
Establishment in a Marshy and Forested Periphery of Two 'Frontier' States: Early
406
Bibliography
Mediaeval Surma Valley (Sylhet and Cachar), C. 600 Ce - 1100 Ce, Religions of
South Asia 6 (1), 33-60.
Prinsep, J., 1836. Facsimiles of Indian Inscriptions, Journal of the Asiatic Society of
Bengal 5 554-561.
Pryce, T.O., Bevan, H., Ciarla, R., Rispoli, F., Castillo, C., Hassett, B., Malakie,
J.L., 2011. Intensive Archaeological Survey in Southeast Asia: Methodological and
Metallurgical Insights from Khao Sai on, Central Thailand, Asian Perspectives 50 (1
& 2), 53-69.
Rai, S.S., 2010. Application of Integrated Digital Technologies in the Study of
Settlement Archaeology of Kausambi Region, in: Campana, S., Forte, M., Liuzza, C.
(Eds.), Space, Time, Place. Third International Conference on Remote Sensing in
Archaeology: Archaeopress, pp. 325-332.
Rajaguru, S.N., Joshi, P.S., 2013. Maharashtra: Land and People, in: Deotare, B.C.,
Joshi, P.S., Parchure, C.N. (Eds.), Glimpses of Ancient Maharashtra through
Archaeological Excavations, Pune: Bharatiya Itihas Sankalan Samitee & Deccan
College Post-Graduate and Research Institute, pp. 29-42.
Rajan, K., 1994. Archaeology of Tamil Nadu (Kongu Country), Delhi: Book India
Publishing Co. .
Rajan, K., 2003. Archaeology of the Pudukottai Region, Tamil Nadu, Man and
Environment XXVIII (1), 41-56.
Rajani, M.B., Bhattacharya, S., Rajawat, A.S., 2011. Synergistic Application of
Optical and Radar Data for Archaeological Exploration in the Talakadu Region,
Karnataka, Journal of the Indian Society of Remote Sensing 39 (4), 519-527.
Rajeev, C.B., 2005. Indian Archaeology - a Review, 1999-2000, New Delhi:
Archaeological Survey of India, Government of India.
Rajeev, C.B., 2006. Indian Archaeology - a Review, 2000-2001, New Delhi:
Archaeological Survey of India, Government of India.
Ramachandran, K.S., 1980. Archaeology of South India, Tamil Nadu Delhi:
Sundeep.
Rao, M.S.N., 1985. Indian Archaeology - a Review, 1982-1983, New Delhi:
Archaeological Survey of India, Government of India.
Rao, S.R., 1963. Excavation at Rangpur and Other Explorations in Gujarat, Ancient
India 18-19 1-207.
Rao, S.R., 1973. Lothal and the Indus Civilisation, London: Asia Publishing House.
Raven, E.M., 2004. Kings in Copper, in: Bakker, H.T. (Ed.), The Vākāṭaka
Heritage: Indian Culture at the Crossroads, Groningen: Egbert Forsten, pp. 19-31.
Raven, E.M., 2008. Brick Terraces at Ahicchatra and Mansar: A Comparison, in:
Bakker, H. (Ed.), Mansar. The Discovery of Pravareśvara and Pravarapura, Temple
and Residence of the Vākāṭaka King Pravarasena II. Proceedings of a Symposium at
407
Bibliography
the British Museum, London, 30 June - 1 July 2008, Groningen: Library of the
University of Groningen, pp. 1-48.
Raven, E.M., 2012. From Candragupta II to Kumaragupta I: Styles in the Patterning
of Gupta Coin Designs, Journal of Bengal Art 17 51-62.
Ray, H.P., 1985. Trade in the Western Deccan under the Satavahanas, Studies in
History 1 (1), 15-35.
Ray, H.P., 1986. Monastery and Guild: Commerce under the Satavahanas, Delhi:
Oxford University Press.
Ray, H.P., 1987. Early Historical Urbanisation: The Case of the Western Deccan
World Archaeology 19 (1), 94-104.
Ray, H.P., 1999. Early Buddhist Monachism and Its Socio-Economic Implications,
in: Shastri, A.M. (Ed.), The Age of the Sātavāhanas New Delhi: Aryan Books
International pp. 199-204.
Ray, H.P., 2010. The Archaeology of Ritual Spaces: Satellite Images and Early
Chalukyan Temples, in: Campana, S., Forte, M., Liuzza, C. (Eds.), Space, Time,
Place: Third International Conference on Remote Sensing in Archaeology:
Archaeopress, pp. 343-352.
Ray, H.P., Ravindranath, S., 2007. The Archaeology of Ritual Spaces: Satellite
Images and Early Chalukyan Temples, Man and Environment XXXII (1), 89-101.
Ray, N.R., 1954. Painting and Other Arts, in: Majumdar, R.C., Pusalker, A.D.
(Eds.), The History and Culture of the Indian People: The Classical Age, Bombay:
Bharatiya Vidya Bhavan, pp. 537-554.
Raychaudhuri, H., 1960. The Geography of the Deccan, in: Yazdani, G. (Ed.), The
Early History of the Deccan, Parts I-Vi, London Oxford University Press, pp. 1-64.
Redfield, R., 1956. Peasant Society and Culture Chicago: University of Chicago
Press.
Redfield, R., 1995. The Little Community: Viewpoints for the Study of a Human
Whole, Uppsala: Almqvist & Wiksells Boktryckeri Ab
Rees, G., 2011. Buddhism and Donation: Rock-Cut Monasteries of the Western
Ghats, Archaeology: University of Cambridge.
Rice, P.M., 1987. Pottery Analysis: A Sourcebook, Chicago: The University of
Chicago Press.
Rondelli, B., Stride, S., García-Granero, J.J., 2013. Soviet Military Maps and
Archaeological Survey in the Samarkand Region, Journal of Cultural Heritage 14
(3), 270-276.
Rowland, J., Phillipps, R., 2012. Survey Strategy in Disturbed Landscapes: A Case
Study from the Nile Delta, eTopoi. Journal for Ancient Studies Special Volume 3
93-100.
408
Bibliography
Russell, R.V., 1908. Nagpur District Gazetteer, Central Provinces and Berar
District Gazetteers, Calcutta: Bengal Secretariat Book Depot.
Sahu, B.P., 2001. Brahmanical Ideology, Regional Identities and the Construction of
Early India, Social Scientist 29 (7/8), 3-18.
Sahu, B.P., 2012. Recent Perspectives of the State and Debates in Early Indian
History, Indian Historical Review 39 (2), 145-162.
Sali, C., 1998. Vakataka Culture: Archaeological Perspective, Sanshodhak 66? 3-38.
Sankalia, H.D., 1960. The Nude Goddess or "Shameless Woman" in Western Asia,
India, and South-Eastern Asia, Artibus Asiae 23 (2), 111-123.
Sankalia, H.D., Deo, S.B., 1955. Report on the Excavations at Nasik and Jorwe
1950-1951, Poona: Deccan College Post-Graduate and Research Institute.
Sankalia, H.D., Deo, S.B., Ansari, Z.D., Ehrhardt, S., 1960. From History to Pre-
History at Nevasa (1954-56), Poona: Deccan College Post-Graduate and Research
Institute.
Sankalia, H.D., Dikshit, M.G., 1952. Excavations at Brahmapuri (Kolhapur), 1945-
46, Poona: Deccan College Postgraduate and Research Institute.
Sankalia, H.D., Subbarao, B., Deo, S.B., 1958. The Excavations at Maheshwar and
Navdatoli, 1952-1953, Poona & Baroda: Deccan College Post-Graduate and
Research Institute & Maharaja Sayajirao University.
Sawant, R., 2012. Historical Archaeology of Vidarbha, New Delhi: Aryan Books
International
Scarborough, V.I., Lucero, L.J., 2010. The Non-Hierarchical Development of
Complexity in the Semitropics: Water and Cooperation, Water History 2 (2), 185-
205.
Schiffer, M.B., 1988. The Structure of Archaeological Theory, American Antiquity
461-485.
Schiffer, M.B., Sullivan, A.P., Klinger, T.C., 1978. The Design of Archaeological
Surveys, World Archaeology 10 (1), 1-28.
Schopen, G., 1997. Bones, Stones and Buddhist Monks: Collected Papers on the
Archaeology, Epigraphy and Texts of Monastic Buddhism in India, Honolulu:
University of Hawai'i Press.
Schuldenrein, J., Wright, R.P., Mughal, M.R., Khan, M.A., 2004. Landscapes, Soils,
and Mound Histories of the Upper Indus Valley, Pakistan: New Insights on the
Holocene Environments near Ancient Harappa, Journal of Archaeological Science
31 777-797.
Selvakumar, V., 1996. Archaeological Investigations in the Upper Gundar Basin,
Madurai District, Tamil Nadu, Man and Environment XXI (2), 27-42.
Sharma, A.K., 2008. Star Shaped Temples of Mansar, Puratattva 38 195-198.
409
Bibliography
Sharma, G.R., 1960. The Excavations at Kauśambī (1957-1959). The Defences and
the Śyenaciti of the Puruṣamedha, Allahabad: Allahabad University Publications
Sharma, R.S., 1984. How Feudal Was Indian Feudalism?, Social Scientist 12 (2), 16-
41.
Sharma, R.S., 1987a. The Kali Age: A Period of Social Crisis, in: Jha, D.N. (Ed.),
Feudal Social Formation in Early India, Delhi: Chanakya Publications, pp. 45-64.
Sharma, R.S., 1987b. Urban Decay in India (C. Ad 300-1000), New Delhi:
Munshiram Manoharalal Publication.
Sharma, R.S., 1991. Aspects of Political Ideas and Institutions in Ancient India,
Delhi: Motilal Banarsidass Publ. .
Shastri, A.M., 1975-1976. Mandhal Copper-Plate Charter of Pravarasena II, Year 16
Epigraphia India 41 68-76.
Shastri, A.M., 1987. Early History of the Deccan: Problems and Perspectives,
Delhi: Sundeep Prakashan.
Shastri, A.M., 1990. Unique Coin of Vakataka Prithivishena (II?) Indian Coin
Society Newsletter 4 2.
Shastri, A.M., 1992a. The Age of the Vākāṭakas: Harman Publishing House.
Shastri, A.M., 1992b. Vākāṭaka Coins, in: Shastri, A.M. (Ed.), The Age of the
Vākāṭakas, New Delhi: Harman Publishing House, pp. 285-294.
Shastri, A.M., 1992c. The Vākāṭakas: Original Home and Some Other Problems, in:
Shastri, A.M. (Ed.), The Age of the Vākāṭakas New Delhi: Harman Publishing
House, pp. 3-20.
Shastri, A.M., 1997. Vākāṭakas: Sources and History: Aryan Books International.
Shastri, A.M., 1998. The Satavahanas and the Western Kshatrapas, Nagpur:
Dattsons.
Shaw, J., 2004. Nāga Sculptures in Sanchi's Archaeological Landscape: Buddhism,
Vaiṣṇavism, and Local Agricultural Cults in Central India, First Century Bce to Fifth
Century Ce, Artibus Asiae 64 (1), 5-59.
Shaw, J., 2005/2006. Landscape, Water and Religion in Ancient India, Archaeology
International 9 43-48.
Shaw, J., 2007. Buddhist Landscapes in Central India: Sanchi Hill and
Archaeologies of Religious and Social Change, C. Third Century B.C. To Fifth
Century A.D., London: British Association for South Asian Studies, The British
Academy.
Shaw, J., Sutcliffe, J., 2003. Water Management, Patronage Networks and Religious
Change: New Evidence from the Sanchi Dam Complex and Counterparts in Gujarat
and Sri Lanka, South Asian Studies 19 (1), 73-104.
410
Bibliography
Shaw, J., Sutcliffe, J., 2005. Ancient Dams and Buddhist Landscapes in the Sanchi
Area: New Evidence on Irrigation, Land Use and Monasticism in Central India,
South Asian Studies 21 (1), 1-24.
Shaw, J., Sutcliffe, J.V., Lloyd-Smith, L., Schwenninger, J.-L., Chauhan, M., Misra,
O., Harvey, E., 2007. Ancient Irrigation and Buddhist History in Central India:
Optically Stimulated Luminescence Dates and Pollen Sequences from the Sanchi
Dams, Asian Perspectives 46 (1), 166-201.
Shete, G., 2009. The 'Type-Variety' Approach to Utilitarian Pottery of the Early Iron
Age and Early Historical Period in the Western Vidarbha Region: Applications for
Site Chronology, Man and Environment XXXIV (1), 66-76.
Shinde, V., 1989. Land Use Pattern During the Chalcolithic Period in the Tapi
Basin, Purātattva: Bulletin of the Indian Archaeological Society 18 54-60.
Shinde, V., 1991. The Late Harappan Culture in Maharashtra, India: A Study of
Settlement and Subsistence Patterns, South Asian Studies 7 91-96.
Shinde, V., Deshpande, S.S., 2015. Crafts and Technologies of the Chalcolithic
People of South Asia: An Overview, Indian Journal of History of Science 50 (1), 42-
54.
Shipton, C.B.K., Janardhana, B., Koshy, J., Petraglia, M.D., Haslam, M., Korisettar,
R., 2010. Systematic Transect Survey of the Jurreru Valley, Kurnool District,
Andhra Pradesh, Man and Environment XXXV (1), 24-36.
Shokoohy, M., Shokoohy, N.H., 1999. The Dark Gate, the Dungeons, the Royal
Escape Route and More: Survey of Tughluqabad, Second Interim Report, Bulletin of
the School of Oriental and African Studies, University of London 62 (3), 423-461.
Shrimali, K.M., 1987. Agrarian Structure in Central India and the Northern Deccan
(C.Ad300-500). A Study of Vākāṭaka Inscriptions, New Delhi: Munshiram
Manoharlal Publishers Pvt. Ltd. .
Shrimali, K.M., 1992. Some Aspects of the Economy of the Vakatakas, in: Shastri,
A.M. (Ed.), The Age of the Vakatakas, New Delhi: Harman Publishing House, pp.
101-116.
Shrimali, K.M., 1993. Reflections on Recent Perceptions of Early Medieval India,
Social Scientist 21 (12), 25-39.
Singh, B.P., 1991. The Chalcolithic Culture of South Bihar as Revealed by the
Explorations and Excavations in District Rohtas, Purātattva: Bulletin of the Indian
Archaeological Society 20 83-92.
Singh, B.P., 1996. Indian Archaeology - a Review, 1991-1992, New Delhi:
Archaeological Survey of India, Government of India.
Singh, K.S., 2004. People of India: Maharashtra, Part 1, Mumbai: Popular
Prakashan, Pvt. Ltd. .
411
Bibliography
Singh, P., 1969. The Problem of Black and Red Wares in Indian Archaeology in:
Sinha, B.P. (Ed.), Potteries in Ancient India Patna: The Department of Ancient
Indian History and Archaeology, Patna University, pp. 67-74.
Singh, R.N., Petrie, C.A., French, C.A.I., Bates, J., Pandey, A.K., Parikh, D.,
Lancelotti, C., Redhouse, D.I., 2012. Survey and Excavations at Dabli Vas Chugta,
Hanumangarh District, Rajasthan, Puratattva: Bulletin of the Indian Archaeological
Society 42 133-147.
Singh, R.N., Petrie, C.A., French, C.A.I., Goudie, A.S., Gupta, S., Tewari, R., Singh,
A.K., Sinha, R., Srivastava, R., Yadav, S., Singh, V.K., 2008. Settlements in
Context: Reconnaissance in Western Uttar Pradesh and Haryana, Man and
Environment XXXIII (2), 71-87.
Singh, R.N., Petrie, C.A., French, C.A.I., Neogi, S., Pandey, A.L., Parikh, D., Pawar,
V., 2010a. Geoarchaeological Survey and Excavations at Burj-2010, Fatehabad,
Haryana, Purātattva: Bulletin of the Indian Archaeological Society 40 94-101.
Singh, R.N., Petrie, C.A., Pawar, V., Pandey, A.K., Neogi, S., Singh, M., Singh,
A.K., Parikh, D., Lancelotti, C., 2010b. Changing Patterns of Settlement in the Rise
and Fall of Harappan Urbanism and Beyond: A Preliminary Report on the
Rakhigarhi Hinterland Survey 2009, Man and Environment XXXV (1), 37-53.
Singh, U., 1993. Kings, Brāhmaṇas and Temples in Orissa: An Epigraphic Study
A.D. 300-1147, New Delhi: Munshiram Manoharlal Publishers Pvt. Ltd.
Singh, U., 2008. A History of Ancient and Early Medieval India: From the Stone
Age to the 12th Century, New Delhi: Dorling Kindersley (India) Pvt. Ltd. .
Singh, U., 2009. Brāhmaṇa Settlements in Ancient and Early Medieval India, in:
Chattopadhyaya, B.D. (Ed.), History of Science, Philosophy and Culture in Indian
Civilisation, Delhi: Centre for Studies in Civilisations (CSC), pp. 157-175.
Sinha, B.P., Roy, S.R., 1969. Vaisali Excavations, 1958-1962, Patna: The
Directorate of Archaeology and Museums, Government of Bihar.
Sinha, N., 2001. Early Maitrakas, Landgrant Charters and Regional State Formation
in Early Medieval Gujarat, Studies in History 17 (2), 151-163.
Sinopoli, C.M., 1991. Approaches to Archaeological Ceramics, New York: Plenum
Press.
Sinopoli, C.M., 2001. On the Edge of Empire: Form and Substance in the
Satavahana Dynasty, in: Alcock, S.E., D'Altroy, T.N., Morrison, K.D., Sinopoli,
C.M. (Eds.), Empires: Perspectives from Archaeology and History, Cambridge:
Cambridge University Press, pp. 155-178.
Sinopoli, C.M., 2006. Imperial Landscapes of South Asia, in: Stark, M.T. (Ed.),
Archaeology of Asia, Oxford: Blackwell Publishing Ltd, pp. 324-349.
Sircar, D.C., 1969. Landlordism and Tenancy in Ancient and Medieval India as
Revealed by Epigraphical Records, Lucknow: University of Lucknow.
412
Bibliography
Skibo, J.M., 1999. Pottery and People, in: Skibo, J.M., Feinman, G.M. (Eds.),
Pottery and People: A Dynamic Interaction, Salt Lake City: University of Utah
Press, pp. 1-8.
Smith, A.T., 2003. The Political Landscape: Constellations of Authority in Early
Complex Polities California: University of California Press.
Smith, M.L., 1999. "Indianization" from the Indian Point of View: Trade and
Cultural Contacts with Southeast
Asia in the Early First Millennium C.E., Journal of the Economic and Social History
of the Orient 42 (1), 1-26.
Smith, M.L., 2000. Systematic Surface Survey at the Early Historic Site of
Kaundinyapura, India, Man and Environment 25 (1), 75-87.
Smith, M.L., 2001a. The Archaeological Hinterlands of Mahasthangarh:
Observations and Potential for Future Research, in: Alam, M.S., Salles, J.-F. (Eds.),
France-Bangladesh Joint Venture Excavations at Mahasthangarh First Interim
Report, Dhaka, pp. 61-73.
Smith, M.L., 2001b. The Archaeology of an Early Historic Town in Central India,
Oxford: J. and E. Hedges, Archaeopress.
Smith, M.L., 2002a. The Role of Local Trade Networks in the Indian Subcontinent
During the Early Historic Period, Man and Environment XXVII (1), 139-151.
Smith, M.L., 2002b. Systematic Survey at the Early Historic Urban Site of
Sisupalgarh, Orissa, in: Sengupta, G., Panja, S. (Eds.), Archaeology of Eastern
India: New Perspectives, Kolkata: Centre for Archaeological Studies and Training,
East India, pp. 111-125.
Smith, M.L., 2005. Archaeological Research at Sisupalgarh, an Early Historic City
in Eastern India, in: Franke-Vogt, U., Weisshaar, H.-J. (Eds.), South Asian
Archaeology 2003, Aachen: Kommission für Archäologie Außereuropäischer
Kulturen pp. 297-306.
Smith, M.L., 2007. Territories, Corridors, and Networks: A Biological Model for the
Premodern State, Complexity 12 (4), 28-35.
Smith, M.L., 2014. The Archaeology of Urban Landscapes, Annual Review of
Anthropology 43 307-323.
Smith, V.A., 1914. The Vakataka Dynasty of Berar in the Fourth and Fifth Centuries
A.C, Journal of the Royal Asiatic Society of Great Britain and Ireland (ArticleType:
research-article / Full publication date: Apr., 1914 / Copyright © 1914 Royal Asiatic
Society of Great Britain and Ireland), 317-338.
Sohoni, A., 1998. Temple Architecture of the Marathas in Maharashtra, Leicester:
De Montfort University.
Sohoni, A., 2011. Ganesh Temple at Tasgaon: Apotheosis of Maratha Temple
Architecture South Asian Studies 27 (1), 51-73.
413
Bibliography
Sonawane, V.H., 1988. Some Remarkable Sculptures of Lajjā Gaurī from Gujarat,
Lalit Kala 23 27-34.
Sontheimer, G.D., 1982. Hero and Sati Stones of Maharashtra, in: Settar, S.,
Sontheimer, G.D. (Eds.), Memorial Stones: A Study of Their Origin, Significance
and Variety, Dharwad: Institute of Indian Art History, Karnatak University, pp. 261-
281.
Spate, O.H.K., Learmonth, A.T.A., 1967. India and Pakistan: A General and
Regional Geography, New Delhi: Munshiram Manoharlal Publishers Pvt. Ltd. .
Spink, W.M., 2005. Ajanta: History and Development. Volume One: The End of the
Golden Age, Leiden: Koninklijke Brill NV.
Spink, W.M., 2014. Ajanta: History and Development, Volume Six: Defining
Features, Leiden Koninklijke Brill NV.
Srinivas, M.N., 1952. Religion and Society among the Coorgs of South India,
Kodagu: Asia Publishing House.
Srivastava, K.M., 1971. The Problem of the Black-and-Red Ware in Proto-Historic
India, Journal of the Oriental Institute XX (4), 372-417.
Staal, J.F., 1963. Sanskrit and Sanskritization, The Journal of Asian Studies XXII
(3), 261-276.
Stark, B.L., Ossa, A., 2007. Ancient Settlement, Urban Gardening, and Environment
in the Gulf Lowlands of Mexico, Latin American Antiquity 18 (4), 385-406.
Stark, M.T., 2006. Early Mainland Southeast Asian Landscapes in the First
Millennium Ad, Annu. Rev. Anthropol. 35 407-432.
Stein, B., 1977. The Segmentary State in South Indian History, in: Fox, R.G. (Ed.),
Realm and Region in Traditional India, New Delhi: Vikas Publishing House Pvt.
Ltd., pp. 3-51.
Stein, B., 1985. State Formation and Economy Reconsidered, Modern Asian Studies
19 (3), 387-413.
Subbarao, B., 1953. Baroda through the Ages (Being the Report of an Excavation
Conducted in the Baroda Area 1951-1952), Baroda: Maharaja Sayajirao University
of Baroda.
Subbarao, B., 1958. The Personality of India Baroda: Maharaja Sayajirao University
of Baroda.
Sugandhi, N., 2008. Networks of Intention in a Peripheral Landscape, Antiquity
Project Gallery.
Sugandhi, N., Morrison, K.D., 2011. Archaeology of Hinduism, in: Insoll, T. (Ed.),
The Oxford Handbook of the Archaeology of Ritual and Religion, Oxford: Oxford
University Press, pp. 921-933.
Sutcliffe, J., 2007. Hydrology of Ramtek and Mansar (26.02.07).
414
Bibliography
Sutcliffe, J., Shaw, J., Brown, E., 2011. Historical Water Resources in South Asia:
The Hydrological Background, Hydrological Sciences Journal 56 (5), 775-788.
Suvrathan, U., 2013. Complexity on the Periphery: A Study of Regional
Organization at Banavasi, C. 1st - 18th Century A.D.: University of Michigan.
Suvrathan, U., 2014. Spoiled for Choice? The Sacred Landscapes of Ancient and
Early Medieval Banavasi, South Asian Studies 30 (2), 206-229.
Taçon, P.S.C., Boivin, N., Hampson, J., Blinkhorn, J., Korisettar, R., Petraglia,
M.D., 2010. New Rock Art Discoveries in the Kurnool District, Andhra Pradesh,
India, Antiquity 84 335-350.
Tare, A., Dikshit, O., Rai, G.K., Fonia, R.S., Prabhakar, V.N., Vikram, B., 2010.
Geospatial Database of Ahichchatra Using Geoinformatics, in: Campana, S., Forte,
M., Liuzza, C. (Eds.), Space, Time, Place. Third International Conference on
Remote Sensing in Archaeology: Archaeopress, pp. 289-300.
Terrenato, N., 2000. Surface Thoughts: Future Directions in Italian Field Surveys,
in: Bintliff, J., Kuna, M., Venclová, N. (Eds.), The Future of Surface Artefact Survey
in Europe, Sheffield: Sheffield Academic Press, pp. 21-28.
Terrenato, N., 2004. Sample Size Matters! The Paradox of Global Trends and Local
Surveys, in: Alcock, S.E., Cherry, J.F. (Eds.), Side by Side Survey: Comparative
Regional Studies in the Mediterranean World, Oxford: Oxbow Books, pp. 36-48.
Terrenato, N., Ammerman, A.J., 1996. Visibility and Site Recovery in the Cecina
Valley Survey, Italy, Journal of Field Archaeology 23 (1), 91-109.
Thakur, V.K., 1993. The Post-Kushana Urban Decline in India: Implications of
Ecology, Proceedings of the Indian History Congress 53 97-105.
Thakur, V.K., 1997-1998. Decline or Diffusion: Constructing the Urban Tradition of
North India During the Gupta Period, The Indian Historical Review XXIV (1-2), 20-
69.
Thakuria, T., 2008-2009. Memorial Stones from Gujarat: Study of Pāliyas at
Kanmer, Bulletin of the Deccan College Post-Graduate and Research Institute 68-
69 179-190.
Thakuria, T., 2014. Chronology of Vidarbha Megalithic Culture: An Appraisal,
Journal of Multidisciplinary Studies in Archaeology 2 651-657.
Thakuria, T., Padhan, T., Mohanty, R.K., Smith, M.L., 2013. Google Earth as an
Archaeological Tool in the Developing World: An Example from India, The SAA
Archaeological Record: The Society for American Archaeology, pp. 20-24.
Thapar, B.K., 1964-65. Prakash - 1955, Ancient India 20-21 5-167.
Thapar, B.K., 1979. Indian Archaeology - a Review, 1975-1976, New Delhi:
Archaeological Survey of India, Government of India.
Thapar, B.K., 1980a. Indian Archaeology - a Review, 1976-1977, New Delhi:
Archaeological Survey of India, Government of India.
415
Bibliography
416
Bibliography
Varma, S., 2008. The Absence of Urban Centres in Early Historic Kathiawad, in:
Sengupta, G., Chakraborty, S. (Eds.), Archaeology of Early Historic South Asia,
New Delhi Pragati Publications with Centre for Archaeological Studies and
Training, Eastern India, pp. 213-230.
Venkata Subbaiah, P.C., 1989. Protohistoric Investigations in the Central Pennar
Basin, Cuddapah District, Andhra Pradesh, Purātattva: Bulletin of the Indian
Archaeological Society 18 79-84.
Verardi, G., 2002. Excavations at Gotihawa and a Territorial Survey in Kapilavastu
District of Nepal: A Preliminary Report, Rome Lumbini International Research
Institute.
Verardi, G., 2011. Hardships and Downfall of Buddhism in India, New Delhi:
Manohar Publishers & Distributors.
Verghese, A., 2004. Deities, Cults and Kings at Vijayanagara, World Archaeology
36 (3), 416-431.
Virkus, F., 1992. Politische Programme in Gupta-Vākāṭaka-Zeitalter, Dargestellt
Am Beispiel Von Landschenkungsinschriften, Altorientalische Forschungen 19 (1),
121-126.
Walimbe, S.R., 2003. Excavations at the Early Historic Site of Kahali-Brahmapuri,
Annual Report of the Deccan College Post-Graduate and Research Institute, 2001-
2002 39-40.
Wangu, M.B., 2003. Images of Indian Goddesses: Myths, Meanings and Models,
New Delhi: Abhinav Publications
Welankar, V., 2009. The Iconography of Kevala Narasimha: A Reappraisal, South
Asian Studies 25 (1), 113-130.
Wellsted, T.A., 1934. Notes on the Vakatakas of the Central Provinces and Berar,
and Their Country, 4th to 8th Century A.D., Journal of the Asiatic Society of Bengal
XXIX 159-166.
Wilkinson, T.J., 2003. Archaeological Landscapes of the near East, Tucson: The
University of Arizona Press.
Wilkinson, T.J., 2004. Archaeology of Landscape, in: Bintliff, J. (Ed.), Companion
to Archaeology, Oxford: Blackwell Publishing Ltd., pp. 334-356.
Wilkinson, T.J., 2014. Comparative Landscape Analysis: Contrasting the Middle
East and Maya Regions, Archaeological papers of the American Anthropological
Association 24 183-200.
Willey, G.R., 1953. Prehistoric Settlement Patterns in the Viru Valley, Peru
Washington: The Bureau.
Williams, J.G., 1982. The Art of Gupta India. Empire and Province, Princeton.
Williams, J.G., 1983. Vākāṭaka Art and the Gupta Mainstream, in: Smith, B.L. (Ed.),
Essays on Gupta Culture, Delhi, pp. 215-233.
417
Bibliography
Willis, M., 2008. Cosmetics and Goddesses: The Wellsted Collection at the British
Museum, in: Bakker, H. (Ed.), Mansar. The Discovery of Pravareśvara and
Pravarapura, Temple and Residence of the Vākāṭaka King Pravarasena II.
Proceedings of a Symposium at the British Museum, London, 30 June - 1 July 2008,
Groningen: Library of the University of Groningen, pp. 1-8.
Willis, M., van Schaik, S., Hill, N., 2012. European Research Council Synergy
Grant 2012. Beyond Boundaries: Religion, Region, Language and the State, British
Museum, London: European Research Council.
Willis, M.D., 2009. The Archaeology of Hindu Ritual: Temples and the
Establishment of the Gods, New York: Cambridge University Press.
Yashpal, S.B., Sood, R.K., Agrawal, D.P., 1980. Remote Sensing of the Lost
Saraswati River, Proceedings of the Indian Academy of Sciences (Earth and
Planetary Sciences) 89 (3), 317-331.
Yazdani, G., 1930-1955. Ajanta: The Colour and Monochrome Reproductions of the
Ajanta Frescoes Based on Photography, London: Oxford University Press.
418
Appendix Table of Contents
419
Appendix 1: Site List
Appended here is the list of sites encountered during the survey field seasons (2011-2014). These include known sites such as the Rāmṭek temples
and newly identified features. Several sites located just outside of the 10km survey area have been included, as they have featured in the
discussion of the local area. Sites are arranged in numerical order with a simple ascending site code. Sites have been categorised according to their
primary features, as discussed in the main text, and have been assigned a descriptive sub-type.
The information contained here includes the GPS location of sites, a brief description of the nature of the site and a comment on the material
remains present. Religious associations and chronological attributions have been given where possible.
Proposed Potential
Site GPS Site Sub- Brick Ceramic Sculpture Extent (m/km) Type of
Site Type Description Find Phase of phase of
No. No. Type Y/N range Phase or area (ha) survey
structure site
Remains of the
excavation at a 5th
Damaged
Temple century brick temple
1 1 Religious Y N/A N/A N/A III III mound - extent Known
remains (BHK-II). Extensive
unclear
damage due to removal of
brick for re-use.
Remains of the
Temple
2 4 Religious excavation of a 5th Y N/A N/A N/A III III Known
remains
century brick temple
420
Appendix 1: Site List
Remains of the
excavation at MDL-II
Temple
3 6 Religious with associated Nandi Y N/A N/A N/A III III Known
remains
sculpture and
architectural fragments
421
Appendix 1: Site List
Potential tower
Structural Structural
14 17 construction along the Y N/A N/A N/A II/III II/III Known
remains remains
surrounding wall
422
Appendix 1: Site List
423
Appendix 1: Site List
424
Appendix 1: Site List
Siddhanātha Ascetics
Cave next to the
26 22 Religious Cave N N/A N/A N/A III III Known
Guptarāma temple on
Rāmṭek hill
425
Appendix 1: Site List
426
Appendix 1: Site List
427
Appendix 1: Site List
A possibly Medieval
building on the north side
of Ambālā tank,
Structural Structural
41 578 displaying similar motifs N N/A N/A N/A V? V? Known
remains remains
with Islamic influence as
18th century Chatrīs in
the region
428
Appendix 1: Site List
429
Appendix 1: Site List
430
Appendix 1: Site List
431
Appendix 1: Site List
Ancient large-sized
Structural Re-used bricks re-used in modern Informed
57 114 Y N/A N/A N/A N/A III?
remains Brick house construction near (Level 3)
to Nagardhan fort
432
Appendix 1: Site List
433
Appendix 1: Site List
Hanuman temple in
Temple Nagardhan with stones Informed
64 566 Religious N N/A N/A V+ VI V+
(Modern) outside. The temple is (Level 3)
probably modern
434
Appendix 1: Site List
Small modern
temple/shrine in
Shrine Nagardhan. There is a Informed
68 570 Religious N N/A N/A N/A VI VI
(Modern) large stone inside under (Level 3)
worship, and modern
architectural fragments
Temple near to
Nagardhan Fort with
fragmentary sculpture.
Temple Informed
69 571 Religious The temple appears N N/A N/A IV-VI VI IV-VI
(Modern) (Level 3)
modern but there are a
several pieces of Post-
Vākāṭaka sculpture
435
Appendix 1: Site List
Mound southwest of
Nagardhan, past the fort.
The mounded areas are
overgrown and have been
cut back to expand paddy
fields. Brick fragments
and ceramic sherds can Around 3 ha
be collected from around despite being
Settled Informed
71 340 Mound and in the sections. On Y II-IV/V N/A N/A N/A II-V cut back to
habitation (Level 3)
the west side of the expand paddy
mounded area is a large fields
section with ceramic
remains in the bottom of
the cut and sherds
scattered around the
ground surface, which is
cultivated
436
Appendix 1: Site List
437
Appendix 1: Site List
438
Appendix 1: Site List
Area of archaeological
remains at Hamlapuri
where paddy fields have
been cut through the
c. 4 ha walked
original ground surface
with GPS
leaving sections
around area
containing brick and
with visible
pottery. Several 'islands'
Settled Disturbed Unknow pottery and
79 12 remain in the middle of Y II-V+ N/A N/A II-V+ Known
habitation mound n brick remains -
the paddy fields which
mound is
show that the area would
destroyed so
have been mounded. The
true extent
sections appear to be
unknown
uniform and the site was
described as 'single
culture Vākāṭaka' in trial
excavations
Unclear
whether this
Large 5th century Ganesh
area forms part
sculpture and two Liṅgas
of the original
now housed in a modern
80 12 Religious Shrine Y N/A N/A III VI III habitation Known
temple, next to the
mound due to
destroyed mound at
destruction and
Hamlapuri
modern
development.
439
Appendix 1: Site List
440
Appendix 1: Site List
Hanuman temple in
Temple Hamlapuri - modern with Informed
85 491 Religious N N/A N/A V-VI VI V-VI
(Modern) no evidence of an earlier (Level 3)
construction at the site.
Modern temple in
Nandpuri containing a
Temple modern Ganesh sculpture Informed
87 493 Religious N N/A N/A V-VI VI V-VI
(Modern) and a small Nandi (Level 3)
(possibly Late Medieval
to Early Modern)
441
Appendix 1: Site List
Modern Hanuman
Temple in Nandpuri
Temple containing various pieces Informed
89 248 Religious N N/A N/A IV+ VI IV-VI
(Modern) of sculpture, none of (Level 3)
which appear to be
particularly old
442
Appendix 1: Site List
Artificial cut
Low approximately 190 metres
Ceramic
density to the north east of Kapur Informed
93 78 remains in N II-V+ N/A N/A N/A II-V+ 4m cut
habitation Bāvlī. Ceramic remains (Level 3)
cut
remains were collected from this
area
443
Appendix 1: Site List
Low
Low density ceramic Ceramic scatter
density Ceramic Informed
94 81 scatter found in the fields N II-V+ N/A N/A N/A II-V+ within 50m
habitation scatter (Level 3)
north of Rāmṭek hill radius of GPS
remains
444
Appendix 1: Site List
Ceramic remains
Low
Ceramic collected from the
density Informed
102 118 remains in foundations and filling N II-V+ N/A N/A N/A II-V+
habitation (Level 3)
cut material of a collapsed
remains
house in Rāmṭek town.
445
Appendix 1: Site List
Ceramic remains
Low
Ceramic collected from the
density Informed
103 119 remains in foundations and filling N II-V+ N/A N/A N/A II-V+
habitation (Level 3)
cut material of a collapsed
remains
house in Rāmṭek town.
Isolated
Architectur Isolated stone column Phasing Informed
104 120 architectur N N/A N/A N/A N/A
al remains piece in Rāmṭek Unknown (Level 3)
al remains
446
Appendix 1: Site List
A modern shrine in
Shrine Informed
110 127 Religious Rāmṭek with various N N/A N/A V+ VI V-VI
(Modern) (Level 3)
fragments of sculpture
Environme
A large Medieval bāvlī in Informed
111 128 ntal Bāvlī N N/A N/A N/A V V?
Rāmṭek. (Level 3)
Feature
447
Appendix 1: Site List
Low
Ceramic Ceramic remains in spoil Area explored
density Informed
116 206 remains in at a digging site in N II-V+ N/A N/A N/A II-V+ was 25m by
habitation (Level 3)
cut Rāmṭek town 33m
remains
448
Appendix 1: Site List
Temple group on
outskirts of Rāmṭek town.
This area contains several
Temple Informed
119 207 Religious temples and a Medieval N N/A N/A IV-V IV-V IV-V
Group (Level 3)
bāvlī (now filled in), as
well as numerous
sculptural fragments
449
Appendix 1: Site List
450
Appendix 1: Site List
Modern temple in
Temple Rāmṭek, containing a Informed
128 519 Religious N N/A N/A N/A VI VI
(Modern) Hanuman sculpture and (Level 3)
stones under worship
451
Appendix 1: Site List
Modern temple in
Temple Rāmṭek with sculpture Informed
132 523 Religious N N/A N/A V-VI VI V-VI
(Modern) fragments outside (Level 3)
including a head
452
Appendix 1: Site List
453
Appendix 1: Site List
454
Appendix 1: Site List
455
Appendix 1: Site List
456
Appendix 1: Site List
Two sandstone
architectural pieces kept
at a house next to the
ploughed fields
neighbouring Kawadak.
Isolated
Architectur Medium density pottery
156 401 architectur Y II-V+ N/A N/A N/A II-V+ Quadrat
al remains scatter and fragmentary
al remains
brick visible around the
house and is possibly a
further continuation of
the ceramic scatter
related to Kawadak
Environme
Stone bāvlī in sampled
157 393 ntal Bāvlī N N/A N/A N/A V V Quadrat
quadrat
Feature
457
Appendix 1: Site List
Religious
Temple East facing stone
with
162 407 (Modern) / well/bāvlī with modern N N/A N/A N/A VI VI Quadrat
Environme
Well temple facing it
ntal feature
458
Appendix 1: Site List
459
Appendix 1: Site List
460
Appendix 1: Site List
Low
Brick fragments on the
density Brick Phasing Informed
175 154 surface of the low hill Y N/A N/A N/A N/A
habitation Scatter Unknown (Level 3)
around Mansar-Moil
remains
461
Appendix 1: Site List
462
Appendix 1: Site List
Brick remains on a
prominent part of the hill
Low
north of Mansar village,
density Brick Phasing Informed
182 205 uncertain whether these Y N/A N/A N/A N/A
habitation Scatter Unknown (Level 3)
are pieces of old or
remains
modern brick as they are
too fragmentary
463
Appendix 1: Site List
464
Appendix 1: Site List
465
Appendix 1: Site List
466
Appendix 1: Site List
Modern Hanuman
Temple Temple in Hivra village: Informed
195 263 Religious N N/A N/A N/A VI VI
(Modern) The temple contains a (Level 3)
few pieces of sculpture
467
Appendix 1: Site List
468
Appendix 1: Site List
re-used in buildings.
Some pottery present.
The liṅga appears to be
old. It is not under
worship today - it is
viewed simply as a
marker of the village
centre
Collection of small
modern shrines in Hivri
Shrine containing various Informed
203 583 Religious Y N/A N/A IV+ VI VI
(Modern) sculptures, none of which (Level 3)
appear to be particularly
old
469
Appendix 1: Site List
Clearly mounded
cultivated fields.
Throughout the fields are
brick remains and fairly
dense ceramic scatter.
Several stone
architectural pieces have
Settled been uncovered by the c. 1 ha - visibly Informed
210 367 Mound Y II-V+ N/A N/A VI II-VI
habitation farmer. There is a small mounded area (Level 3)
modern shrine at the base
of the mound containing
a badly preserved
sculpture (?) and stones.
This area is likely to be
the site of ‘old
Bhijewada’
470
Appendix 1: Site List
Small embankment
Square area of
forming a roughly square
112m by
Environme tank/water storage area
Embankme Phasing 141m. Wider Informed
213 223 ntal on the north side of Satak Y I-V+ N/A N/A Unknown
nt Unknown scatter (Level 3)
Feature village: pottery sherds
contained
and brick fragments
within 2ha
present
471
Appendix 1: Site List
472
Appendix 1: Site List
473
Appendix 1: Site List
474
Appendix 1: Site List
Embankment surrounding
fields with a modern The
shrine, south of Manapur. embankment
This could be a small old was roughly
Religious Temple disused tank but the area 152 m long but
with (Modern) / within the embankment is the circular Informed
228 234 N N/A N/A N/A VI VI
Environme embankme now grassland. A water storage (Level 3)
ntal feature nt rudimentary area contained
temple/shrine sits on the within was
embankment, and measured at
contains stones dedicated 0.8ha
to Bhiwsen Maharaj
475
Appendix 1: Site List
A seemingly Late
Medieval/Early Modern
Mandir on a platform in
Manapur (Marāṭhā style Informed
231 433 Religious Temple N N/A N/A N/A V V
architectural features, (Level 3)
possibly 18th century).
The temple does not
contain sculpture
Modern Hanuman
Temple in Manapur
containing few seemingly
Temple modern sculptures. Informed
232 434 Religious N N/A N/A VI? VI VI
(Modern) Outside the main temple (Level 3)
is a small platform with a
collection of fragmentary
sculpture
476
Appendix 1: Site List
Environme
Stone well in
234 402 ntal Well N N/A N/A N/A V+ V+ Quadrat
grassland/field area
Feature
477
Appendix 1: Site List
478
Appendix 1: Site List
479
Appendix 1: Site List
Brick remains in
Kawalapur village. Brick
fragments are found
scattered on the ground
Low Extent of
Ceramic / and there is reuse of brick
density surface scatter Informed
239 269 brick in several structures. Y V+ N/A N/A N/A V+
habitation is roughly (Level 3)
scatter Limited pottery was
remains 0.7ha in area
collected but the village
is not mounded and the
archaeological remains
are fairly low density
480
Appendix 1: Site List
Clear mound at
Beldongri: The village is
built around the northern
and western side of the
mound, which has been
cultivated and has large
trees growing on top.
There are clear and c. 1ha of
extensive brick remains, visible
Structural including an area of in material. Informed
242 279 Mound Y II/III N/A N/A III? II/III
remains situ brick wall visible in a Mound (Level 3)
section on the south west approximately
side of the mound. The 65 m diameter
mound is said to be an
old 'large house' by
villagers. There is very
little pottery; instead the
mound appears to
represent a solid brick
construction.
481
Appendix 1: Site List
482
Appendix 1: Site List
483
Appendix 1: Site List
Modern Hanuman
Temple in Nerla,
containing a few pieces
Temple Informed
250 342 Religious of modern sculpture. N N/A N/A VI? VI VI
(Modern) (Level 3)
There are no indications
that the temple has earlier
foundations
484
Appendix 1: Site List
485
Appendix 1: Site List
Brightly painted
relatively old (Late
Medieval, possibly 18th
century with Marāṭhā
influence) Hanuman Informed
255 485 Religious Temple N N/A N/A VI? V V-VI
temple in Chichala with a (Level 3)
smaller adjoining temple
behind, which contains
several sculptures of
female figures
486
Appendix 1: Site List
487
Appendix 1: Site List
488
Appendix 1: Site List
489
Appendix 1: Site List
Hanuman Temple
Informed
270 316 Religious Temple between Nawergaon and N N/A N/A V+ V-VI V-VI
(Level 3)
Sangrampur
490
Appendix 1: Site List
Location of 'Devada
Rithi', which is said to be
an abandoned village,
with an associated
temple. This area is now
ploughed fields. Pottery
Area of
Abandone and brick remains can be
Settled Unknow cultural Informed
275 318 d village collected. The pottery and Y II-V+ N/A V-VI II-VI
habitation n material (Level 3)
(modern) brick may be modern.
roughly 2ha
The village is still marked
on topographic maps. The
temple does not appear to
be old and houses
fragmentary modern
sculpture
491
Appendix 1: Site List
492
Appendix 1: Site List
493
Appendix 1: Site List
Modern temple at
Temple Informed
288 373 Religious Khodgaon which contains N N/A N/A V+ VI VI
(Modern) (Level 3)
some sculpture
494
Appendix 1: Site List
Mound of 'Old
Khodgaon' in agricultural
fields. The mound has
c.1 ha
been cut away to the
Settled mounded field Informed
289 322 Mound south and west side. Y II-IV/V N/A N/A N/A II-IV/V
habitation with surface (Level 3)
Surface ceramics are
ceramics
particularly noticeable
and there are also
fragmentary bricks.
495
Appendix 1: Site List
South-facing modern
Temple Informed
294 354 Religious Hanuman Temple at N N/A N/A N/A VI VI
(Modern) (Level 3)
'Minsi Rithi'
496
Appendix 1: Site List
Small south-facing
Hanuman shrine in
Shrine Masala, which appears to Informed
298 550 Religious N N/A N/A V-VI VI V-VI
(Modern) be modern. There is a (Level 3)
Hanuman sculpture in
front of the shrine
497
Appendix 1: Site List
Small embankment
surrounding a potentially
disused tank north of Extent unclear
Environme
Embankme Bori with low density Phasing but appears to Informed
306 89 ntal N II-IV N/A N/A Unknown
nt ceramic scatter across the Unknown be roughly (Level 3)
Feature
surface of the 0.15ha
surrounding harvested
paddy fields
498
Appendix 1: Site List
Arranged decorated/
worked stone blocks
Informed
310 95 Religious Cabūtrā forming a cabūtrā in Bori. N N/A N/A IV+ Unknown IV+
(Level 3)
On the shrine is a Sri
Pada
499
Appendix 1: Site List
500
Appendix 1: Site List
Modern temple in
Panchala containing
Temple various sculptures Informed
321 499 Religious N N/A N/A VI? VI VI
(Modern) including Ganesh and (Level 3)
Hanuman. Next to the
temple is a round well
501
Appendix 1: Site List
Seemingly modern
temple in the fields near
Temple Informed
324 326 Religious Panchala Khurd. There is N II-V+ N/A Unknown VI II-VI
(Modern) (Level 3)
no obvious evidence of
archaeological remains.
502
Appendix 1: Site List
Temple in Bhandarbodi
Informed
327 251 Religious Temple with small Naga N N/A N/A IV+ VI IV-VI
(Level 3)
sculpture
Large tank at
Bhandarbodi: The tank
has a very large earth
embankment and is
Environme c.3 ha across
possibly old, but there are Phasing Informed
328 ntal Tank N N/A N/A N/A Unknown the water
no visible archaeological Unknown (Level 3)
Feature storage area
remains in the vicinity
and no ceramics or brick
were collected from the
surrounding fields
503
Appendix 1: Site List
504
Appendix 1: Site List
Hanuman Temple
Informed
341 331 Religious Temple between the villages of N N/A N/A V+ V-VI V-VI
(Level 3)
Mandri and Mahadula
505
Appendix 1: Site List
Environme
Informed
347 390 ntal Well Stone built old well N N/A N/A N/A V+ V+
(Level 3)
Feature
506
Appendix 1: Site List
Large earthwork or
embankment to the east
of the mound at Old
Mandri, brick and
ceramic fragments are
found in the surroundings
Low and particularly in the
Ceramic / Ceramics
density disturbed ground by the Informed
350 107 brick Y IV/V N/A N/A Unknown IV/V found along
habitation road where there is a cut (Level 3)
scatter 33m of cut
remains in the embankment. It is
not clear whether this
embankment is a modern
feature built up from
surrounding soil or
whether it has older
origins
507
Appendix 1: Site List
508
Appendix 1: Site List
Small hillock at
Manegaon with an area of
plain at the base where a
British period settlement
apparently used to be.
Abandone
Settled Some pottery was Informed
354 229 d village N N/A N/A N/A VI VI
habitation collected from this area (Level 3)
(modern)
but the fragments of
building material were
modern. British period
coin collected from the
fields
509
Appendix 1: Site List
510
Appendix 1: Site List
511
Appendix 1: Site List
Several small
temples/shrine in the
vicinity of the mounded
area of 'Naharwani: both
are fairly modern, one is
dedicated to
Maruti/Hanuman temple.
Shrine Informed
365 370 Religious Near to the temples is an Y N/A N/A Unknown VI VI
(Modern) (Level 3)
area of architectural and
worked stone blocks.
Scattered brick and
ceramic remains can be
found on the ground
surface, presumably from
the destroyed mound.
512
Appendix 1: Site List
Temple group at
Mukanapur. All temples
appear to be relatively old
Temple (Medieval or Early Informed
371 359 Religious N N/A N/A V-VI V V-VI
Group Modern). Two Hanuman (Level 3)
temples and one smaller
temple with a stone and
bull sculpture
513
Appendix 1: Site List
514
Appendix 1: Site List
515
Appendix 1: Site List
516
Appendix 1: Site List
Modern temples in
Chaugan containing
various fragmentary
Temple sculpture and stones. One Informed
391 475 Religious N N/A N/A VI? VI VI
(Modern) of the small shrines (Level 3)
contains a large stone
painted to look like a
large head with a crown
517
Appendix 1: Site List
518
Appendix 1: Site List
Temple south of
Bhondewada and the
brick manufacturing site:
The temple may be Late
Medieval or Early Informed
395 474 Religious Temple N N/A N/A Unknown V V-VI
Modern and contains (Level 3)
several pieces of
sculpture including a
seated female figure
resembling Annapurna
Hanuman temple at
Religious
Bhondewada. The temple
with Temple / Informed
396 299 is situated on the north of N N/A N/A VI? V V-VI
Environme Tank (Level 3)
the village, on the south
ntal feature
side of a round tank
519
Appendix 1: Site List
Hanuman Temple at
Temple Bhilewada which Informed
399 295 Religious N N/A N/A V-VI VI VI
(Modern) contains a few other (Level 3)
small sculptures
Modern Hanuman
Temple Informed
401 377 Religious Temple near to N N/A N/A V+ VI V-VI
(Modern) (Level 3)
Bhilewada
520
Appendix 1: Site List
Potentially
chiselled/worked large
sandstone blocks lying in
the village of
Isolated
Architectur Chorkumari. Throughout Phasing Informed
405 302 architectur N N/A N/A N/A N/A
al remains the village there is no Unknown (Level 3)
al remains
other evidence of
archaeological remains
and the area is not
mounded
521
Appendix 1: Site List
522
Appendix 1: Site List
523
Appendix 1: Site List
Hanuman temple in
Temple Informed
422 444 Religious Khumari containing two N N/A N/A V+ VI V-VI
(Modern) (Level 3)
Hanuman sculptures
Hanuman Mandir in
Informed
424 314 Religious Temple Bondri, just north of N N/A N/A IV+ V-VI IV-VI
(Level 3)
Mansar
524
Appendix 1: Site List
Small temple/shrine in
Kandri on slightly
undulating natural
mounded area by the
road: The object inside is
unknown, it appears to
Shrine have a face but is shaped Informed
429 458 Religious N N/A N/A Unknown VI VI
(Modern) like a sword/long and (Level 3)
pointed. It could be
something that has been
made to appear like an
object of worship. The
shrine and contents are
modern
525
Appendix 1: Site List
526
Appendix 1: Site List
Temple/shrine containing
Shrine sculpture of two figures, Informed
438 480 Religious N N/A N/A Unknown VI VI
(Modern) in the fields north of (Level 3)
Chargaon
527
Appendix 1: Site List
528
Appendix 2: Temple images
This appendix presents selected images of temples, shrines and cabūtrās recorded
during survey in order to illustrate the typical structures encountered across the
Rāmṭek landscape.
529
Appendix 2: Temple images
530
Appendix 2: Temple images
531
Appendix 2: Temple images
532
Appendix 2: Temple images
533
Appendix 2: Temple images
534
Appendix 2: Temple images
535
Appendix 2: Temple images
536
Appendix 2: Temple images
537
Appendix 2: Temple images
538
Appendix 2: Temple images
539
Appendix 2: Temple images
540
Appendix 2: Temple images
541
Appendix 2: Temple images
542
Appendix 2: Temple images
543
Appendix 2: Temple images
544
Appendix 2: Temple images
545
Appendix 2: Temple images
546
Appendix 3. Small Finds
Find Site GPS Find Type Material Dimensions (cm) Description Suggested Additional
Number Number Number Phase Comments
1 29 53 Figurine Ceramic Height: 8 cm Fragmentary figurine of an animal's II/III?
Width (across head; perhaps a bull given the
head): 7cm Neck visible bridle detail and marks
circumference: where horns may have originally
15cm been. Made from micaceous clay,
with an orange exterior and solid
grey interior. The outside has been
given a micaceous wash. Relatively
heavy.
2 29 60 Ceramic Ceramic Height: 4cm Half of a hollow round ceramic Unknown Associations with
object Length: 7.8cm object with a flattened top and Phase II/III
Width: 4.2cm bottom, and a hole through the ceramics
middle. Wheel-turned (visible
manufacture lines inside). One base
is thicker than the other. Made of
red ware with occasional mica dust
and the odd grit inclusion. Light
and not very sturdy, possibly a
spindle whorl
3 45 86 Figurine Ceramic Height: 3cm Part of ceramic figurine; the upper Unknown
Circumference: torso of a human figure, with what
12cm appears to be a necklace. Made of a
dark micaceous ware, black in
colour with some evidence of a thin
black slip.
547
Appendix 3. Small Finds
4 349 105 Grinding Stone Length: 5.7 cm End piece of a quern-stone or Unknown
stone Width: 7cm grinder. This has a rounded top
Depth: 5cm with a flat bottom possibly from
grinding. It is a fairly heavy stone
with porous appearance, and the
outside has been smoothed.
5 349 105 Stone Stone Length across the Rounded stone piece, which could Unknown
object break: 6.7cm be a body rubber or a gaming disc.
Width: 4.8cm Porous stone, dark grey in colour,
Depth: 3cm with a rough feel. It is broken
roughly in half and is shaped
approximately like a semicircle.
6 353 111 Grinding Stone Length: 5.5cm Small, heavy grey stone block, Unknown
stone Circumference: which could have been a quern-
15.8cm stone or a pestle used for grinding.
It is possible that it could even be
part of a small squared-off liṅga.
7 236 215 Grinding Stone Length: 14cm A gritty sandstone piece which Unknown
stone Circumference: could be a heavy grinding stone or
22cm quern/pestle. It has a rough surface
all over.
8 236 215 Stone Stone Length: 6cm Very smooth, mottled pink and grey Unknown Associated with
object Width: 5.8cm stone with shiny quartz inclusions. Phase II to IV
Depth: 5.7cm The surface has been smoothed and ceramics and
Half polished but it is broken at both Gupta-size bricks
circumference:12 ends. One side of the stone piece
cm has a smooth, flat bottom, which is
slightly concave, so that the cross-
section of the object is a semicircle.
The purpose is unknown; perhaps
part of a grinding stone or part of a
carved sculpture.
548
Appendix 3. Small Finds
549
Appendix 3. Small Finds
16 363 308 Worked Stone Length: 7.6cm A large stone tool; an oval flake I Discussion with Dr
stone Width: 6.1cm made from a hard dark brown R. Abbas
stone. Appears to be a Middle
Palaeolithic flake with prominent
bulb portion and two bidirectional
flake scars. The edges also show
unintentional breaking
17 444 329 Worked Stone Length: 4.2cm Possible microlith/flake: Opaque I/II
stone Width: 1.9cm brown colour, sharp edges, appears
to have been worked.
18 275 318 Ceramic Ceramic Length: 2.7cm Appears to be part of a small Unknown
object Circumference: ceramic pipe, broken at both ends.
5.7cm Made from a hard-fired red ceramic
with few inclusions.
19 236 216 Figurine Ceramic Height: 4cm Part of a ceramic figurine; the head II/III?
Length (nose to of an animal, which appears to be a
neck): 9cm bull. The face is long and seems to
Width: 5cm have had horns. A bridle is visible,
and there is a small raised lump on
top of the head. Made from a
micaceous ceramic, with mica on
the exterior and traces of a dark red
slip.
550
Appendix 3. Small Finds
20 236 216 Worked Stone Length: 5cm Worked flint-like stone/ Bifacial I Discussion with Dr
stone Width: 3.5cm unidirectional core. Seems to be a R. Abbas
Circumference: core on a nodule (river pebble) with
11cm clear evidence where small flakes
have been taken off. Two
generations of flaking, the tool has
been reused. The flaking on the
smoother side of the core is glossy,
while the second side shows a later
episode of flaking (or it could be
that one surface was exposed and
became polished). The core seems
to have been sharpened to a point
so could have been used as a tool
itself.
21 79 124 Ceramic Ceramic Width: 1.3cm Small clay ball, which has no Unknown
object Circumference: obvious hole and so doesn't appear
3.2 to be a bead. Perhaps a piece for
gaming.
22 45 Ceramic Ceramic Length: 6.5cm Part of a large decorated pipe or the II/III
object Widest spout of a vessel. Made from a red
circumference: ware with few inclusions but
14cm treated with a dark red slip (similar
to RED2.RS).
23 73 361 Stone Stone Height: 1.4cm Just over half of a sandstone ball, Unknown
object Width: 2.1 no perforations and so it doesn't
Circumference: 7 appear to be a bead. Could be a
gaming piece.
551
Appendix 3. Small Finds
24 Quadrat T1.62 Ceramic Ceramic Length: 10.5cm Large ceramic spout from a vessel, II/III Similar to the
object Widest Orange-red exterior fabric and spouts of the
circumference:17. black interior; the ware is similar to 'Roman imitation'
5 RBC2. Wheel-turned (visible vessels found at
manufacture lines), surfaces have Kolhapur
been smoothed but it still has a (Dhavalikar,
fairly rough feel. No evidence of 1999:139)
surface treatment or decoration.
25 29 60 Ceramic Ceramic Radius: 2.6cm Broken, small ceramic weight or Unknown
object Width: 1cm whorl. Rounded ceramic object
with a hold through the middle.
Made from a red ware, rough with
no surface treatment.
26 236 216 Ceramic Ceramic Length: 9cm Small cup or vessel made from a III
object Rim diameter: heavy gritty micaceous ware
7cm (Similar material to GROMIT4 and
found with large pieces of another
vessel). Irregular in shape, appears
hand formed, narrow interior which
would not hold much liquid.
552
Appendix 3. Small Finds
Find number: 1 2 3 4
5 6 7 8
553
Appendix 3. Small Finds
9 10 11 12
Find number: 13 14 15 16
554
Appendix 3. Small Finds
17 18 19 20
21 22 23 24
555
Appendix 3. Small Finds
Find number: 25 26
556
Appendix 4: Sherd count
This appendix contains the sherd count of the ceramics assemblage with the final
ware and type designation of each rim sherd.
Sherd Site
GPS. No. Ware Sherd Type Rim Type
Number No.
3 21 24 RBC2.RS R 33
4 21 24 BAG.BLS R 148
9 20 54 ORM4 R 187
10 20 54 ORM4.RS R 142
15 92 13 RBC2.RS R 148
18 92 13 RED.BLS R 159
24 79 12 ROM R 55
25 79 12 RED2 R 76
26 79 12 RED3.RS R 55
27 79 12 ROM R 49
28 79 12 ROM R 55
30 49 11 ORM5 R 49
31 49 11 BOM5 R 49
32 49 11 BOM1 R 11
51 49 11 ORM6 R 11
56 45 RBC1.RS R 150
59 45 RBC1.RS R 151
60 45 RBC2.RS R 43
62 45 BOM2.RS R 111
67 7 129 RBC1.RS R 143
70 7 129 RED2.RS R 11
73 7 129 RED2 R 150
81 7 129 ORM5 R 49
134 45 GW.BLS R 184
135 45 BOM2.RS R 47
138 45 BOM2.RS R 151
156 45 ORM2 R 47
158 45 GW.BLS R 148
161 45 ORM3 R 85
190 49 11 PATCHY R 11
199 49 11 ORM6 R 75
557
Appendix 4: Sherd count
205 49 11 PATCHY R 11
207 49 11 PATCHY R 11
217 49 11 PATCHY R 11
218 49 11 PATCHY R 11
230 49 11 PATCHY R 11
234 49 11 PATCHY R 11
258 49 11 BOM2 R 11
286 29 53 RBC2.RS R 148
291 29 53 GW.BLS R 158
296 29 53 RBC2.RS R 112
297 24 37 RBC2.RS R 109
301 24 37 RED5.RS R 152
302 24 37 RED2 R 11
305 24 37 RBC2.RS R 180
317 24 37 BOM2.RS R 187
325 28 47 RBC2.RS R 148
331 28 47 RBC3 R 24
333 28 47 RBC2.RS R 24
334 28 47 RBC2.RS R 148
340 23 33 RBC3 R 26
341 23 33 BOM5.RS R 44
353 23 33 RBC2.RS R 184
357 20 55 FIDAM R 47
375 20 55 BOM6 R 75
381 20 55 ORM4 R 187
398 29 59 RED2 Lamp 202
400 29 59 GW.BLS R 148
405 29 59 RBC3.RS R 24
416 29 59 RBC1.RS R 148
420 29 59 RBC2.RS R 24
424 29 59 RBC1.RS R 148
429 29 59 RBC2.RS R 24
442 29 59 RBC2.RS R 40
447 29 59 RBC2.RS R 148
452 29 59 RED1.RS R 58
453 29 59 RED2 R 100
459 29 59 RBC2.RS R 148
464 29 59 RBC2.RS R 38
477 29 59 RED.BLS R 148
558
Appendix 4: Sherd count
482 29 59 RBC3 R 8
493 29 59 RBC2.RS R 148
513 29 59 RED1.RS R 100
518 29 59 RBC3.RS R 148
539 92 13 RBC1.RS R 160
541 92 13 ORM4 R 143
542 92 13 ORM4 R 143
547 92 13 BAG.BLS R 167
550 92 13 RBC2.RS R 113
583 45 GW.BLS R 112
585 45 GW.BLS R 159
587 45 GW.BLS R 150
588 45 BUG.BRSL R 24
590 45 GW.BLS R 111
592 45 BAG.BLS R 184
598 45 BAG.BLS R 27
599 45 BUG.BRSL R 159
612 45 DARM R 148
613 45 DARM R 104
617 45 RBC2.RS R 24
618 45 BUG.BRSL R 112
620 45 GW.BLS R 149
622 45 GW.BLS R 159
623 45 GW.BLS R 24
624 45 BAG.BLS R 40
625 45 BAG.BLS R 149
628 45 GW.BLS R 24
630 45 BAG.BLS R 148
631 45 GW.BLS R 148
632 45 GW R 159
633 45 GW R 159
634 45 GW R 184
635 45 DARM R 79
636 45 BAG.BLS R 184
637 45 BAG.BLS R 152
638 45 BAG.BLS R 153
640 45 BAG.BLS R 153
642 45 GW.BLS R 184
644 45 BAG.BLS R 152
559
Appendix 4: Sherd count
560
Appendix 4: Sherd count
561
Appendix 4: Sherd count
764 45 BOM2.RS R 24
765 45 GOW1.RS R 69
766 45 BOM2.RS R 24
767 45 RBC2.RS R 148
770 45 RED2 R 11
771 45 BOM3.RS R 100
772 45 BOM2.RS R 143
773 45 ORM1.RS R 47
775 45 BOM2.RS R 143
777 45 RBC2.RS R 143
782 45 BOM2.RS R 100
785 45 BOM2.RS R 58
787 45 ORM5.RS R 100
788 45 RBC1.RS R 151
789 45 RBC3.RS R 99
790 45 RBC2.RS R 143
791 45 RBC2.RS R 24
792 45 BOM2.RS R 26
793 45 BOM2.RS R 143
794 45 RBC3.RS R 111
797 45 GW.BLS R 40
802 45 DARM R 59
807 45 BOM3.RS R 24
811 45 GW.BLS R 156
815 45 DARM R 11
816 45 BOM2.RS R 143
830 45 GW.BLS R 159
832 45 RED2 R 100
839 45 DARM R 5
840 45 DARM.RS R 78
841 45 ORM4 R 139
844 45 BUG.BRSL R 100
848 45 RBC2.RS R 148
850 45 RED2 R 150
851 45 BOM2.RS R 155
864 45 BOM3.RS R 100
870 45 DARM.RS R 75
881 45 ORM4.RS R 100
894 45 BOM2.RS R 24
562
Appendix 4: Sherd count
563
Appendix 4: Sherd count
1029 94 81 BOM6 R 4
1030 94 81 ORM5 R 173
1036 94 81 ORM5 R 150
1059 46 86 BOM1.RS R 100
1060 46 86 ORM1.RS R 47
1061 46 86 DARM R 44
1062 46 86 ORM1.RS R 47
1063 46 86 BOM1.RS R 47
1065 46 86 BUG.BRSL R 142
1066 46 86 BOM2.RS R 29
1069 46 86 BOM2.RS R 155
1070 46 86 RBC2.RS R 143
1071 46 86 ORM1 Lamp 204
1074 46 86 BOM1.RS R 58
1075 46 86 BAG.BLS R 142
1077 46 86 BOM1.RS R 47
1080 46 86 BOM3.RS R 100
1081 46 86 BOM2.RS R 151
1082 46 86 BOM2.RS R 143
1085 46 86 ORM1.RS R 47
1086 46 86 BOM2.RS R 100
1087 46 86 BOM2.RS R 113
1100 95 83 GW.BLS R 5
1144 305 88 RED2 R 11
1153 49 11 PATCHY R 11
1154 49 11 PATCHY R 11
1155 49 11 ORM3 R 11
1160 49 11 PATCHY R 11
1161 49 11 PATCHY R 11
1162 49 11 PATCHY R 11
1167 49 11 BOM2 R 11
1168 49 11 PATCHY R 11
1174 49 11 PATCHY R 11
1176 49 11 PATCHY R 11
1179 49 11 PATCHY R 11
1182 49 11 PATCHY R 11
1185 49 11 PATCHY R 11
1188 49 11 PATCHY R 11
1190 49 11 PATCHY R 11
564
Appendix 4: Sherd count
1191 49 11 PATCHY R 11
1192 49 11 PATCHY R 11
1193 49 11 PATCHY R 11
1194 48 98 ORM2 R 48
1196 48 98 ORM2 R 48
1198 48 98 ORM2 R 48
1199 48 98 ORM2 R 48
1202 48 98 ORM5 R 48
1203 48 98 ORM2 R 48
1215 48 98 ORM3 Lamp 201
1219 48 98 RED2 R 11
1242 52 104 RBC3 R 11
1244 75 100 RBC2.RS R 99
1245 75 100 BOM1.RS R 146
1246 75 100 BOM1.RS R 146
1248 75 100 GW R 183
1258 349 106 GW.BLS R 110
1259 349 106 GW.BLS R 24
1264 349 106 GW R 180
1271 75 100 GW.BLS R 183
1294 76 101 BOM2.RS Lid 39
1295 76 101 BOM1.RS R 196
1296 76 101 ORM2 R 120
1299 76 101 BOM4 R 57
1301 76 101 RBC2.RS R 151
1303 76 101 RBC3 R 73
1308 76 101 BOM2.RS R 116
1309 76 101 BOM2.RS R 99
1310 76 101 ORM1.RS R 146
1312 76 101 BOM2 R 52
1314 76 101 RBC3.RS R 73
1316 76 101 BAR R 2
1317 76 101 BOM1.RS R 178
1318 76 101 RED2 R 11
1322 76 101 NON-ID R 14
1324 76 101 BOM2.RS R 150
1325 76 101 BOM1.RS R 151
1327 76 101 BOM1.RS R 24
1336 76 101 BOM1.RS R 11
565
Appendix 4: Sherd count
566
Appendix 4: Sherd count
567
Appendix 4: Sherd count
568
Appendix 4: Sherd count
569
Appendix 4: Sherd count
570
Appendix 4: Sherd count
571
Appendix 4: Sherd count
572
Appendix 4: Sherd count
573
Appendix 4: Sherd count
574
Appendix 4: Sherd count
575
Appendix 4: Sherd count
576
Appendix 4: Sherd count
577
Appendix 4: Sherd count
578
Appendix 4: Sherd count
579
Appendix 4: Sherd count
580
Appendix 4: Sherd count
581
Appendix 4: Sherd count
582
Appendix 4: Sherd count
583
Appendix 4: Sherd count
584
Appendix 4: Sherd count
585
Appendix 4: Sherd count
586
Appendix 4: Sherd count
587
Appendix 4: Sherd count
588
Appendix 4: Sherd count
589
Appendix 4: Sherd count
590
Appendix 4: Sherd count
591
Appendix 4: Sherd count
592
Appendix 4: Sherd count
593
Appendix 4: Sherd count
594
Appendix 4: Sherd count
595
Appendix 4: Sherd count
596
Appendix 4: Sherd count
597
Appendix 4: Sherd count
598
Appendix 4: Sherd count
599
Appendix 4: Sherd count
600
Appendix 4: Sherd count
601
Appendix 4: Sherd count
602
Appendix 4: Sherd count
603
Appendix 5: Ware descriptions
Code BAG.BLS
Distinguishing A grey inner fabric sandwiched between outer brown fabrics with
features a black slip, generally burnished on one or two surfaces.
Surface Surfaces flattened and smoothed, and then the external surface is
treatment treated with a thin black slip, which is usually burnished. The
interior surface is also occasionally black slipped and burnished, or
treated with a thin mica wash.
Decoration Commonly, the shoulder is decorated with incised lines running
horizontally around the vessel, with occasional additional
decoration such as oblique notches or wavy lines. Stamped
floral/sun designs are common on body sherds, usually at the
shoulder. Some rims feature a pinched applique design
Manufacture Wheel-turned
Firing Hard
Colour Black shiny outer slip. 7.5YR 5/6 strong brown outer fabrics with
(Munsell) a 7.5YR 6/1 grey core.
Fracture Rough but regular and generally perpendicular to surfaces.
Occasionally the fracture is inward facing
Inclusions Occasional sub-angular mica fragments (<2mm) and mica dust,
more on surfaces. Addition of mica usually does not seem
deliberate but perhaps due to the parent clay. Roughly sorted sub-
angular sand, occasional long, thin vegetal voids in fabric and rare
small sub-angular clear grits (2-3mm)
Thickness Generally between 4mm and 7mm body thickness, but some
sherds are up to 10mm thick.
General A hard-fired, medium-thick brown and grey ware usually with a
description thin shiny black slip, and occasionally a thick slip. Surfaces appear
to have been burnished. The fabric is brown with grey core, and
similar in feel to the grey ware (GW), but often not as lightweight.
This ware appears to be found mainly in restricted jar forms.
604
Appendix 5: Ware descriptions
Code BAR
Distinguishing Black and red ware with distinctive colouring indicative of inverse
features firing. Coarser fabric that usually associated with this ware, but
still highly burnished on both surfaces.
Surface Surfaces smoothed and flattened. Both surfaces slipped and then
treatment burnished.
Decoration None – One sherd shows two incised lines running horizontally
beneath the rim
Manufacture Wheel-turned
Firing Hard
Colour Both surfaces are black burnished. Bottoms of exterior surfaces are
(Munsell) 5YR 6/6 reddish yellow
Fracture Slightly rough but fairly regular and perpendicular to the surfaces
Inclusions Roughly sorted sub-rounded sand grains (1mm), rare mica dust,
very rare sub-angular clear grits (<2mm)
Thickness Body ranges from 4mm up to 8mm. Usually 6 or 7mm in thickness
General A hard-fired, relatively fine fabric Black and Red Ware with a
description smooth hard feel to the surfaces, which have been slipped and
highly burnished. Characteristic surface colouration from the
inverse firing technique, and a uniform black or dark grey core.
Occurs in limited types of shallow carinated bowl forms. The
fabric is less fine than is typically associated with the characteristic
Black and Red Ware (See F.BAR) – BAR sherds are usually
slightly thicker and less smooth than F.BAR with some grit
inclusions.
Similar wares: Shaw 2007 ‘Crude Black and Red Ware’ –
Medium thickness fabric, poorly levigated.
Possible date range: 2nd century BC to 2nd century AD (Ansari &
Dhavalikar, 1975; Mehta & Chowdary, 1966; Shaw, 2007)
605
Appendix 5: Ware descriptions
Code BLARB
Surface Surfaces appear smoothed and flattened with slip on one or both
treatment sides. On one side is burnished, usually the exterior but it is the
interior surface on two occasions
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Mostly blackish grey, with occasional patches of 7.5YR 6/4 light
(Munsell) brown.
General A hard fired, medium ware with rare/occasional mica dust which is
description mostly on surfaces, and well sorted sand with no other evident
inclusions. The sherd surfaces feel smooth and hard and have been
smoothed and slipped, with one surface burnished. Burnt deposits
are apparent on one side of the sherds.
606
Appendix 5: Ware descriptions
Code BOM1
Manufacture Wheel-turned
Colour Patchy with areas approximating to 2.5YR 5/4 weak red, 5YR 6/4
(Munsell) light reddish brown or 5YR 6/6 Reddish Yellow on surfaces with a
dark grey/blackish core.
607
Appendix 5: Ware descriptions
Code BOM1.RS
Distinguishing A red/brown/orange outer fabric sandwiching a dark grey/black
features inner fabric. Large and small mica fragments throughout, common
on surfaces and within fabric. Red slipped.
Surface Surfaces smoothed and flattened. Exterior surface treated with a
treatment red slip, which is generally thin. Slip usually has mica mixed in, or
can be coupled with a thin micaceous wash.
Decoration Generally none, occasional incised horizontal lines and one sherd
which features incised triangular notches
Manufacture Wheel-Turned
Firing Medium - Hard
Colour Patchy with areas approximating to 2.5YR 5/4 weak red, 5YR 6/4
(Munsell) light reddish brown or 5YR 6/6 Reddish Yellow on surfaces with a
dark grey/blackish core. The slip is usually 2.5YR 4/6 dark red.
Fracture Rough irregular fracture but generally perpendicular to the
surfaces, Occasionally forms laminations due to high mica content
Inclusions Dense small to medium sized mica fragments (1-4mm), roughly
sorted occasional sand grains (1mm), occasional poorly sorted,
angular clear grits (2-3mm), occasional poorly sorted rounded
brown grits (2-3mm)
Thickness Generally 4mm to 7mm – a few sherds in the range of 8mm to
10mm thickness.
General Fabric similar to BOM1 but distinguished by a red slip. Rough,
description medium sandy-brown to orange-brown fabric with a dark
grey/blackish core. Densely micaceous which causes a fairly
irregular fracture, but the fabric is still relatively hard-fired and
compact. Mica appears on the surfaces and throughout; other
inclusions are grits and sand but the mica is the dominant and
characteristic inclusions. The surfaces have been smoothed and
flattened, and then slipped. The red slip is usually quite thin and
dark red in colour, and often has the addition of mica fragments.
Carinated vessels dominate in this ware, with evidence of
cooking/smoke.
608
Appendix 5: Ware descriptions
Code BOM2
Manufacture Wheel-turned
Firing Hard
Colour Patchy with areas approximating to 5YR 5/1 grey, 2.5YR 5/4 weak
(Munsell) red, 5YR 6/4 light reddish brown on surfaces with a dark
grey/blackish core.
Thickness Generally 4-7mm but some sherds are in the range of 8-13mm
thickness
609
Appendix 5: Ware descriptions
Code BOM2.RS
Distinguishing A red/brown/orange outer fabric sandwiching a dark grey/black
features inner fabric. Large and small mica fragments throughout fabric and
on surfaces but less common than in BOM1. Red slipped.
Surface Surfaces smoothed and flattened and exterior surface has a red slip.
treatment Interior occasionally has a red slip or is treated with a micaceous
wash
Decoration Usually none but occasional incised lines and notches
Manufacture Wheel-turned
Firing Hard
Colour Patchy with areas approximating to 5YR 5/1 grey, 2.5YR 5/4 weak
(Munsell) red, 5YR 6/4 light reddish brown on surfaces with a dark
grey/blackish core. Slip is 2.5YR 4/6 dark red
Fracture Rough fracture, perpendicular to the edge
Inclusions Frequent to moderate small to medium (1 – 4mm) sub-angular and
angular mica fragments. Poorly sorted sub-angular clear and
rounded brown grits (2mm), occasional long vegetal voids, fairly
well sorted sand grains.
Thickness Body sherds usually between 3 to 7mm but some sherds are in the
range of 8 to 13mm thickness.
General Rough, medium sandy-brown to reddish-brown fabric with grey or
description black core. Frequent to moderate mica fragments throughout fabric
and on surfaces. Mica content is less dense than in BOM1 and less
visible on surfaces. This ware is hard-fired and compact despite
the mica content. Fracture is rough due to inclusions but generally
perpendicular to the surfaces and does not delaminate. Use of a
thin red slip distinguishes this ware from BOM2: Slip is usually
dark red with some appearing more brownish-red or orange-red.
Mica is often added to the slip, and occasionally there is evidence
of treatment with a thin micaceous wash.
610
Appendix 5: Ware descriptions
Code BOM3
Decoration Usually none, one sherd with impressed decoration, and one with
stamped floral motif
Manufacture Wheel-turned
Firing Hard
Colour Patchy with areas approximating to 5YR 5/1 grey, 2.5YR 5/4 weak
(Munsell) red, 5YR 6/4 light reddish brown on surfaces with a dark
grey/blackish core
Thickness Generally between 4 and 7mm, a few sherds are between 8 and
11mm in thickness
611
Appendix 5: Ware descriptions
Code BOM3.RS
Distinguishing Similar to BOM2 - A red/brown/orange outer fabric sandwiching a
features dark grey/black inner fabric. Small mica fragments occasionally
throughout, less common that BOM2, more on surfaces. Red
slipped.
Surface Smoothed and then flattened. Exterior surface treated with a thin
treatment dark red slip. Occasionally a further micaceous wash is added to
the slip.
Decoration Usually none, occasional incised lines
Manufacture Wheel-turned
Firing Hard
Colour Patchy with areas approximating to 5YR 5/1 grey, 2.5YR 5/4 weak
(Munsell) red, 5YR 6/4 light reddish brown on surfaces with a dark
grey/blackish core. Slip is 2.5YR 4/6 dark red
Fracture Rough fracture, perpendicular to the edges
Inclusions Moderate to occasional sub-angular fragments, mostly small
(<2mm) but also medium in size (up to 4mm). Roughly sorted
small sand grains, occasional small angular clear grits (2-3mm)
Thickness Sherds usually between 4 and 7 mm, occasional sherds between 8
and 11mm with one at 25mm thickness
General Red slipped version of BOM3: Rough feel medium fabric with
description sandy-brown or reddish-brown exterior and a dark grey or black
core. Mica is less frequent and smaller than in BOM2, but there are
visible mica inclusions in fabric and on the surface. This ware is
compact, hard-fired and wheel thrown. The surfaces have been
smoothed and flattened and a red slip has been applied, usually on
the exterior surface only. The thin slip is usually a dark red, but
occasionally more orange-red in colour, and mica fragments have
been mixed into the slip. No other surface treatment is visible
except occasional evidence of a micaceous wash.
612
Appendix 5: Ware descriptions
Code BOM4
Distinguishing An orange outer fabric sandwiching a dark grey/black inner fabric
features with very common small mica dust/fragments throughout and on
surfaces. Smooth burnished appearance and patchy surface colour
Surface Surface smoothed, flattened and treated with a micaceous wash
treatment before being possibly burnished. Interior surface has also been
smoothed and treated with a micaceous wash. Surface treatment
gives a silvery/shiny appearance
Decoration Usually none, two sherds with incised oblique lines/notches on
carination
Manufacture Wheel-turned
Firing Hard
Colour Patchy with areas approximating to 5YR 5/1 grey, 2.5YR 5/4 weak
(Munsell) red, 5YR 6/4 light reddish brown on surfaces: Core 5YR 3/1 very
dark grey to 10YR 4/1 Dark grey
Fracture Rough irregular fracture with some laminations from mica
inclusions. Generally perpendicular to the surfaces but sometimes
outwards or inwards facing fracture
Inclusions Dense small to medium (1 - 4mm) angular and sub-angular mica
fragments, and dense mica dust. Poorly sorted occasional sub-
angular and angular small clear grits (2-3mm)
Thickness Sherds generally between 4 and 8mm in thickness but occasional
sherds in the range of 8 to 14mm thickness
General This ware is a hard fired, wheel thrown ware with a densely
description micaceous fabric. The outer surface appears in a range of colours.
This ware’s appearance and mica content is similar to ORM2, but
the orange fabric has a grey core. The fabric mica occurs as small
to medium sized fragments or dust, with smaller fragments being
dominant. The fabric is fairly hard and compact despite mica
inclusions. Other inclusions are not obvious but occasional. The
surface treatment appears to be smoothing and flattening with a
micaceous wash. This wash seems to have been smoothed again or
burnished lending a smooth finish with a silvery sheen from the
fine mica. This ware often occurs in carinated vessel forms.
613
Appendix 5: Ware descriptions
Code BOM5
Distinguishing An orange outer fabric sandwiching a dark grey/black inner fabric
features with frequent small mica dust/fragments throughout but more
common on surfaces.
Surface Surfaces smoothed and flattened, and no other visible surface
treatment treatment on most sherds. Occasional treatment with a micaceous
wash, usually on exterior but sometimes on interior.
Decoration Usually none, two sherds feature incised lines
Manufacture Wheel-turned
Firing Hard
Colour Patchy with areas approximating to 5YR 5/1 grey, 2.5YR 5/4 weak
(Munsell) red, 5YR 6/4 light reddish brown on surfaces with a dark
grey/blackish core.
Fracture Rough and irregular fracture but generally perpendicular to the
surfaces
Inclusions Frequent small (<2mm) sub-angular mica fragments and frequent
to dense mica dust throughout. Occasional vegetal voids on
surfaces, well sorted sub-angular sand grains and rare sub-angular
small clear grits (<2mm).
Thickness Sherds usually within the range of 4 to 7mm thickness. A few
sherds are between 8 and 14mm
General A hard, compact fabric with an orange exterior and a grey core,
description finer than BOM1 and BOM2. The mica content is frequent to
dense but the fragments and dust are small throughout unlike in
previous BOM wares. Mica is more apparent on surfaces than in
fabric due to small size. The surfaces have been smoothed and
flattened but there is no other visible surface treatment and sherds
still feel rough
614
Appendix 5: Ware descriptions
Code BOM5.RS
Distinguishing An orange outer fabric sandwiching a dark grey/black inner fabric
features with moderately common small mica dust/fragments throughout,
but more common on surfaces. Red slipped
Surface Surface smoothed and flattened and then red slipped. Occasional
treatment evidence of additional mica wash on exterior. Interior surface very
occasionally has evidence of a micaceous wash.
Decoration Usually none, but some sherds display incised lines, notches and
notched applique (beading impression)
Manufacture Wheel-turned
Firing Hard
Colour Patchy with areas approximating to 5YR 5/1 grey, 2.5YR 5/4 weak
(Munsell) red, 5YR 6/4 light reddish brown on surfaces with a dark
grey/blackish core. Slip is 2.5YR 4/6 dark red
Fracture Rough and irregular fracture but generally perpendicular to the
surfaces
Inclusions Frequent small (<2mm) sub-angular mica fragments and frequent
to dense mica dust throughout. Occasional vegetal voids on
surfaces, well sorted sub-angular sand grains and rare sub angular
small clear grits (<2mm). Very rare sub-rounded brown grits
(<2mm)
Thickness Usually between 4 and 7 mm, with occasional sherds in the range
of 8 to 10mm
General Same fabric as BOM5 but treated with a red slip - A hard, compact
description fabric with an orange exterior and a grey core, finer than BOM1
and BOM2. The mica content is frequent to dense but the
fragments and dust are small throughout unlike in previous BOM
wares. Mica is more apparent on surfaces than in fabric due to
small size. The surfaces have been smoothed and flattened and
treated with a thin red slip. The slip is usually a dark red but can be
more orange-red in colour. Mica is often added to the red slip
615
Appendix 5: Ware descriptions
Code BOM6
Distinguishing An orange outer fabric sandwiching a dark grey/black inner fabric
features with occasional small mica dust/fragments throughout but more
common on surfaces. Can resemble a red ware due to small
amount of mica
Surface Surfaces flattened, no other visible surface treatment, except
treatment occasional traces of a micaceous wash, usually on exterior
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Patchy with areas approximating to 5YR 5/1 grey, 2.5YR 5/4 weak
(Munsell) red, 5YR 6/4 light reddish brown on surfaces with a dark
grey/blackish core.
Fracture Rough fracture but generally perpendicular to the edges
Inclusions Moderate small (<2mm) sub-angular mica fragments and frequent
mica dust. Common roughly sorted sand grains and occasional thin
long vegetal voids on surfaces. Occasional poorly sorted sub-
rounded and sub-angular clear grits (2-3mm)
Thickness Most sherds are between 4 and 7mm in thickness. Occasional
sherds in a larger size range of 8 to 12mm
General A general fabric category with variation across the ware but
description generally can be described as a hard fired and wheel-turned fabric
with a dull brownish-red surface and grey/black core. Sherds are
lacking in surface treatment and often eroded. Moderate mica
content, usually dust and very small fragments, with mica more
common on surfaces. Similar to a medium red ware but mica
inclusions appear intentional and too frequent for this fabric to be
classed as a red ware. Fabric feels sandy and rough, but still hard
and compact so no material comes away if the surface is rubbed
616
Appendix 5: Ware descriptions
Code BOM6.RS
Distinguishing An orange outer fabric sandwiching a dark grey/black inner fabric
features with occasional small mica dust/fragments throughout but more
common on surfaces. Can resemble a red ware due to small
amount of mica. Red slipped and less common than BOM6
Surface Surfaces flattened and smoothed and treated with a thin red slip.
treatment Occasionally an additional mica wash is added to the exterior
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Patchy with areas approximating to 5YR 5/1 grey, 2.5YR 5/4
(Munsell) weak red, 5YR 6/4 light reddish brown on surfaces with a dark
grey/blackish core. Slip is 2.5YR 4/6 dark red
Fracture Rough and irregular but generally perpendicular to the surfaces
Inclusions Moderate small (<2mm) sub-angular mica fragments and frequent
mica dust. Common roughly sorted sand grains and occasional
thin long vegetal voids on surfaces. Occasional poorly sorted sub-
rounded and sub-angular clear grits (2-3mm), and occasional
rounded brown grits (<3mm)
Thickness Sherds all between 4 and 6.5mm in thickness, except one outlier at
9.4mm
General A general fabric category similar to BOM6 with variation across
description the ware but generally can be described as a hard fired and wheel-
turned fabric with a dull brownish-red surface, grey/black core and
a red slip. Moderate mica content, usually dust and very small
fragments, with mica more common on surfaces. Similar to a
medium red ware but mica inclusions appear intentional and too
frequent for this fabric to be classed as a red ware. Fabric feels
sandy and rough, but still hard and compact so no material comes
away if the surface is rubbed. Slips mostly dark red but
occasionally more orange-red. Small mica fragments visible in slip
617
Appendix 5: Ware descriptions
Code BOM7
Distinguishing A rough orange outer fabric sandwiching a dark grey/black inner
features fabric with occasional small mica fragments throughout but more
common on surfaces. Grit and vegetable inclusions
Surface Surfaces flattened, no other visible surface treatment
treatment
Decoration Usually none, one sherd with incised oblique lines
Manufacture Wheel-turned
Firing Medium to Hard
Colour Patchy with areas approximating to 5YR 5/1 grey, 2.5YR 5/4 weak
(Munsell) red, 5YR 6/4 light reddish brown on surfaces with a dark
grey/blackish core
Fracture Rough fracture but perpendicular to the edges
Inclusions Moderate to frequent small (<2mm) sub-angular mica fragments
and mica dust. Common sand grains and occasional poorly sorted
sub-angular clear grits or rounded brown grits (2-4mm). Rare large
sub-angular clear grits (>4mm), long thin vegetal voids
particularly within fabric but occasionally on surfaces
Thickness Most sherds are between 4 and 7mm in thickness. Occasional
sherds in a larger size range of 8 to 10mm
General Wheel thrown, medium to hard fired, coarse fabric, with an orange
description exterior and a grey/black core. Rough, gritty feel and fabric is
lacking in surface treatment - Utilitarian in appearance. Numerous
types of inclusion with mica and sand being the most prominent
618
Appendix 5: Ware descriptions
Code BUG.BRSL
Distinguishing A buff/grey ware with brown slip, usually burnished on one or two
features surfaces. Occasional small sand grains in fabric
Surface Surfaces flattened and smoothed then treated with a thin or thick
treatment brown slip on exterior. Slip is often burnished and does not usually
contain mica (if mica is visible it is only trace amounts). Interior is
occasionally slipped in brown or black and burnished, or may have
a micaceous wash
Manufacture Wheel-turned
Firing Hard
Colour 2.5YR 5/1 dark reddish grey fabric with 7.5YR 6/3 light brown
(Munsell) surfaces and 7.5YR 3/2 dark brown slip
Inclusions Very rare small (<2mm) sub-angular mica fragments and rare mica
dust in fabric. Occasional roughly sorted sub-angular sand grains
and very rare sub-angular clear or opaque white grits (2-3mm)
619
Appendix 5: Ware descriptions
Code BUG
Distinguishing A buff/grey ware with occasional small sub-angular sand grains in
features fabric.
620
Appendix 5: Ware descriptions
Code COM
Surface Surfaces flattened and smoothed and then treated with a micaceous
treatment wash or thin slip. This micaceous wash appears to have been
smoothed or burnished. Micaceous wash also visible on interior of
some sherds
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Surfaces are 7.5YR 6/4 light brown. Fabric is 2.5YR 6/1 grey.
(Munsell)
Thickness Sherds range between 2.3 and 6.5mm with one slightly thicker
sherd measuring roughly 8mm
General A hard fired wheel-turned micaceous ware with a smooth feel and
description silvery sheen. Sherds are usually thin and treated with a micaceous
wash or slip to lend a silvery appearance. Fabric is fairly densely
micaceous throughout but compact and hard despite mica content.
The fabric is similar to BOM4 but a grey-buff colour throughout.
621
Appendix 5: Ware descriptions
Code DARM
Distinguishing A dark grey/buff micaceous ware with frequent mica fragments on
features surfaces and within the fabric. Occasionally mica-slipped on both
surfaces or treated with a thin black slip. Large roughly sorted sub-
angular sand grains apparent in fabric.
Surface Surfaces flattened and smoothed then occasionally treated with a
treatment micaceous wash on either surface. Exterior surface of sherds
sometimes treated with a thin black slip
Decoration Usually none, occasional sherds with incised horizontal lines on
shoulder portion. A couple of sherds display lines coupled with
other decoration such as a wavy line or incised oblique notches.
Decorated sherds are those with a slip.
Manufacture Wheel-turned
Firing Medium to Hard
Colour Patchy colouration. Some areas black/dark brown 7.5YR 5/3
(Munsell) brown through to 7.5YR 6/4 light brown
Fracture Rough, irregular fracture but generally perpendicular to the
surfaces
Inclusions Frequent to dense small to medium (1-4mm) angular and sub-
angular mica fragments throughout. Roughly sorted large sub-
angular sand grains and occasional poorly sorted small angular and
sub-angular clear/opaque white grits (<2mm)
Thickness 4 to 7mm in thickness with a few sherds in the range of 8 to 14mm
General A medium fabric dark brown or black micaceous ware, medium to
description hard fired and wheel-turned. Distinguished from BOM as the
fabric is always dark brown or black, occasionally with buff/grey
areas, but not orange or red. A fairly general ware category with a
range of mica size and frequency: Usually moderately to densely
micaceous and always with mica visible in the core and on the
surfaces. Some sherds have been treated with a thin black slip
which may be burnished. Slip contains fragments of mica, and
those without a black slip often appear to have a micaceous wash
on the surfaces.
622
Appendix 5: Ware descriptions
Code DARM.RS
Distinguishing A dark grey/buff micaceous ware with large mica frequent mica
features fragments on surfaces and within the fabric. Occasionally mica-
slipped on both surfaces but always treated with a thin dark red
slip. Large roughly sorted sub-angular sand grains apparent in
fabric
Surface Surfaces flattened and then smoothed and treated with a thin red or
treatment brownish-red slip mixed with small mica fragments. Sheen on
some sherds indicates an additional micaceous wash. Interior
surface occasionally treated with a micaceous wash or slipped.
Decoration Usually none, occasional sherds with incised horizontal lines
Manufacture Wheel-turned
Firing Medium to Hard
Colour Patchy colouration. Some areas black/dark brown 7.5YR 5/3
(Munsell) brown through to 7.5YR 6/4 light brown. Areas of 2.5YR 6/6 red
slip
Fracture Rough, irregular fracture but generally perpendicular to the
surfaces
Inclusions Frequent to dense small to medium (1-4mm) sub-angular or
angular mica fragments, and common mica dust. Roughly sorted
large sub-angular sand grains and occasional poorly sorted small
angular and sub-angular clear/opaque white grits (<2mm)
Thickness 4 to 8mm in thickness, four sherds between 9 and 13mm thick
General A medium to hard fired wheel-turned ware with a medium buff-
description grey, dark brown or black micaceous fabric. A fairly general ware
category with a range of mica size and frequency: Fabric is
moderately to highly micaceous and treated with a red or
brownish-red slip with visible small mica fragments mixed in.
Inclusions of grits and sand, and a level of mica which is similar to
BOM2 orBOM3. Distinguished from BOM as the fabric is always
dark brown or black, occasionally with buff/grey areas, but not
orange or red. Dark red or brownish-red slip, often very worn.
623
Appendix 5: Ware descriptions
Code F.BAR
Distinguishing A fine, thin Black and Red Ware - highly burnished on both sides.
features
Surface Surfaces smoothed and flattened. Both surfaces slipped and then
treatment burnished or brought to a polish.
Decoration None (One sherd with a horizontal line to delineate the rim)
Manufacture Wheel-Turned
Firing Hard
Colour Both surfaces black burnished. Bottoms of exterior surfaces are
(Munsell) 5YR 6/6 reddish yellow.
Fracture Smooth, even and perpendicular to the surfaces
Inclusions Well-sorted rare sub-rounded sand grains (1mm), very rare mica
dust
Thickness Body ranges from 2mm to 4mm in thickness
General Smooth, hard-fired, thin fabric with the distinctive black and red
description colouring resulting from inverse firing. Uniformly dark grey or
black throughout the fabric and core. Both surfaces are highly
burnished or polished, and have a hard smooth feel. The fabric has
a metallic ring. This ware occurs in characteristic shallow
carinated bowl forms.
Similar wares: Shaw 2007 ‘Black and Red Ware’ – Medium
thickness fabric, well levigated, hard smooth fabric
Possible date range: 10th century BC to 3rd century BC (Shaw,
2007). Most common in excavations c. 1000BC but continues up
to Early Historic in association with NBPW (Ansari & Dhavalikar,
1975)
624
Appendix 5: Ware descriptions
Code FIDAM
Distinguishing : A thin, finer version of DARM with only fine mica dust on the
features surfaces and in the fabric.
Decoration None
Manufacture Wheel-turned
Firing Hard
General A fairly fine dark micaceous ware, wheel-turned and hard fired.
description Surfaces treated with small/dust mica and there is also small mica
throughout fabric. A thin hard ware which is quite compact as the
mica is fine. Sand inclusions are visible but grits or vegetable
temper are not obvious. Fabric is mostly a dark grey/black or
buff/grey. Surfaces have been smoothed, occasionally slipped and
then washed with mica.
625
Appendix 5: Ware descriptions
Code GOW1
Distinguishing Orange surfaces with a thick grey core, relatively fine fabric.
features
Decoration None
Manufacture Wheel-turned
Firing Medium-Hard
Colour Some 5YR 6/6 (Reddish Yellow) or 5YR 5/6, others 10YR 5/6
(Munsell) Red and 2.5YR 6/8 Red. Core is grey GLEY2 6/5PB
Inclusions Very rare mica dust, rare roughly sorted sand grains, occasional
vegetal voids on surfaces and rare small sub-angular clear grits
(2mm)
General A fairly fine, medium to hard fired ware with distinctive orange or
description pink-orange exterior around a thick uniform grey core. The fabric
is reasonably well levigated, compact, and medium to hard in feel.
It can be quite lightweight. Few evident inclusions in the fabric
and the mica appears to be a natural inclusion.
626
Appendix 5: Ware descriptions
Code GOW1.RS
Distinguishing Orange surfaces with a thick grey core, relatively fine fabric. Thin
features orange-red slip.
Surface Surfaces flattened and smoothed, thin red slip applied, usually to
treatment exterior but occasionally both sides. Slip is close in colour to the
surface of the ware
Decoration None
Manufacture Wheel-turned
Firing Medium-Hard
Colour Some 5YR 6/6 (Reddish Yellow) or 5YR 5/6, others 10YR 5/6
(Munsell) Red and 2.5YR 6/8 Red. Slip is close to the original surface
colour, and usually 5YR 6/6 (Reddish Yellow) and core is grey
GLEY2 6/5PB
Inclusions Very rare mica dust, rare roughly sorted sand grains, very rare
poorly sorted sub-angular opaque white grits (<3mm), occasional
thin vegetal voids particular on surfaces.
General A fairly fine, medium to hard fired ware with distinctive orange or
description pink-orange exterior around a thick uniform grey core. The fabric
is reasonably well levigated, compact, and medium to hard in feel.
It can be quite lightweight. Few evident inclusions in the fabric
and the mica appears to be a natural inclusion. The thin slip is an
orange-red in colour and often very close in colour to the fabric
surfaces.
627
Appendix 5: Ware descriptions
Code GOW2
Distinguishing Orange surfaces with a thick grey core, coarser fabric than GOW1
features with more evident inclusions
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Some 5YR 6/6 (Reddish Yellow) or 5YR 5/6, others 2.5YR 5/6
(Munsell) Red. Core is grey throughout: GLEY2 6/5PB
628
Appendix 5: Ware descriptions
Code GOW2.RS
Surface Surfaces flattened and smoothed, thin red slip applied, usually to
treatment exterior but occasionally both sides, and some evidence of a
micaceous wash on a few sherds
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Some 5YR 6/6 (Reddish Yellow) or 5YR 5/6, others 2.5YR 5/6
(Munsell) Red. Slip is 2.5YR 6/6 Red or 5YR 6/6 (Reddish Yellow). Core is
grey throughout: GLEY2 6/5PB
Thickness Sherds between 6.5 and 10 mm with one sherd measuring around
13mm thick
629
Appendix 5: Ware descriptions
Code GOWM
Surface Surfaces flattened and smoothed, a thin red slip is often applied to
treatment the exterior and there is evidence of a micaceous wash on a few
sherds
Decoration None
Manufacture Wheel-turned
Firing Medium-Hard
Colour Some 5YR 6/6 (Reddish Yellow) or 5YR 5/6, others 2.5YR 5/6
(Munsell) (Red). Slip can be 2.5YR 6/6 (Red) or 5YR 6/6 (Reddish Yellow).
Core is grey: GLEY2 6/5PB
General A micaceous version of Grey and Orange wares (GOW), and most
description similar to GOW2. A medium to hard fired ware with distinctive
orange or pink-orange exterior around a thick uniform grey core.
Rough, medium fabric, with a hard feel and generally heavier than
GOW1 or GOW2. Mica fragments appear as a deliberate inclusion
in this ware. This fabric has been treated with a thin red slip,
similar in colour to the fabric surface
630
Appendix 5: Ware descriptions
Code GROMIT1
Distinguishing : An orange fabric with frequent sub-angular large grits and
features medium to large mica fragments. Thick bodied, hard and very
gritty
Surface Occasional traces of a thin micaceous wash. Otherwise surfaces
treatment rough and uneven despite some evidence of flattening
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Variable colours across ware: typically surface is 10YR 6/4 Light
(Munsell) yellowish brown or 5YR 6/6 Reddish yellow. Core 2.5YR 6/1
Grey or 10 YR 4/2 Dark greyish brown
Fracture Rough, irregular fracture which sometimes delaminates in the
particularly micaceous sherds
Inclusions Frequent/dense roughly sorted sub-angular medium to large clear
and opaque white grits (2-5mm), occasionally grits are larger than
5mm. Moderate poorly sorted round brown grits (2-4mm),
Frequent to dense small to large (1-6mm) angular mica fragments
Thickness Sherds usually within the range of 12 to 21mm, some occasionally
thinner and in the range of 7 to 10mm thickness
General A hard-fired rough ware with a densely micaceous fabric and a
description large quantity of medium to large mixed grits. Patchy exterior
surface colour but essentially an orange fabric with a grey core.
Fabric feels hard and compact, and is difficult to break with a
rough fracture with some laminations from the large amount of
mica. The fabric is heavy and the large size and quantity of grits is
the defining feature. Surface feels rough despite some evidence it
has been flattened or smoothed on some sherds, as the large
quantity of grits forms an irregular surface. Commonly appears in
long, thick flaring rim forms, and some site correlation is apparent.
631
Appendix 5: Ware descriptions
Code GROMIT1.RS
Distinguishing An orange fabric with frequent sub-angular large grits and medium
features to large mica fragments. Thick bodied, hard and very gritty.
Evidence of a thin red slip
Surface Surfaces rough and uneven despite some evidence of flattening,
treatment but treated with a thin red slip and occasional micaceous wash
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Fabric surface variable: 7.5YR 6/4 Light brown to 5YR 6/6
(Munsell) Reddish Yellow. Core Colour 5YR 6/1 Grey or 5YR 5/3 Reddish
Brown. Slip ranges from 2.5YR 6/8 Red to 2.5YR 5/8 Red
Fracture Rough, irregular fracture which sometimes delaminates in the
particularly micaceous sherds
Inclusions Frequent/dense sub-angular and angular large clear and opaque
white grits (2-5mm), occasionally grits are larger than 5mm.
Moderate poorly sorted round brown grits (2-4mm), Frequent to
dense small to large (1-6mm) angular mica fragments
Thickness Sherds usually within the range of 10 to 15mm, some occasionally
thinner and in the range of 7 to 9mm thickness
General A hard-fired rough ware with a densely micaceous fabric and a
description large quantity of medium to large mixed grits, as in GROMIT1,
but with a thin weak red slip. Fabric has a grey core and
brown/orange surface. Fabric feels hard and compact, and is
difficult to break with a rough fracture with some laminations from
the large amount of mica. The fabric is heavy and the large size
and quantity of grits is the defining feature. Surface feels rough
despite some evidence it has been flattened or smoothed on some
sherds, as the large quantity of grits forms an irregular surface.
Commonly appears in long, thick flaring rim forms, and some site
correlation is apparent.
632
Appendix 5: Ware descriptions
Code GROMIT2
Distinguishing An orange fabric with frequent to dense sub-angular medium to
features large grits and common mica
Surface Surfaces rough but smoother than in GROMIT1, seem to have
treatment been flattened or smoothed. Occasional micaceous wash
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Variable colours across ware: typically surface is 10YR 6/4 Light
(Munsell) yellowish brown, 5YR 4/3 Reddish Brown or 5YR 6/6 Reddish
yellow. Core 2.5YR 6/1 Grey or 10 YR 4/2 Dark greyish brown
Fracture Rough, irregular fracture but generally perpendicular to the
surfaces
Inclusions Frequent to dense roughly sorted medium to large (1-4mm) sub-
angular and angular clear grits, and occasional sub-rounded and
rounded brown grits (2-4mm). Occasional grits larger than 4mm.
Moderate small to medium (1-4mm) angular and sub-angular mica
fragments throughout
Thickness Sherds usually between 7 and 10mm thickness, occasional sherds
are slightly thinner (5-6mm) and some sherds up to 14mm
thickness
General A hard-fired, medium, wheel-turned ware with fabric similar to
description GROMIT1 but with smaller better sorted grits and less common
and smaller mica fragments. Fabric is thick but thinner than in
GROMIT1. Appears in a range of colours but typically reddish
brown or light brown with grey or dark brown core. Fabric has a
rough feel, but the surfaces have been flattened or smoothed and
the smaller size and quantity of grits means that the surface is more
regular than in GROMIT1. The ware has a heavy feel, and is hard
and compact
633
Appendix 5: Ware descriptions
Code GROMIT2.RS
Distinguishing An orange fabric with frequent to dense sub-angular medium to
features large grits and common mica
Surface Surfaces rough but smoother than in GROMIT1, seems to have
treatment been flattened or smoothed. Treated with a thin red slip and
occasional micaceous wash.
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Variable colours across ware: typically surface is 10YR 6/4 Light
(Munsell) yellowish brown, 5YR 4/3 Reddish Brown or 5YR 6/6 Reddish
yellow. Core 2.5YR 6/1 Grey or 10 YR 4/2 Dark greyish brown.
Slip: 2.5YR 6/8 Red or 2.5YR 4/6 Dark Red
Fracture Rough, irregular fracture but generally perpendicular to the surfaces
Inclusions Frequent to dense roughly sorted medium to large (1-4mm) sub-
angular and angular clear grits and occasional sub-rounded and
rounded brown grits (2-4mm). Occasional grits larger than 4mm.
Moderate small to medium (1-4mm) angular and sub-angular mica
fragments throughout.
Thickness Sherds usually between 6 and 10mm thickness, occasional sherds
are slightly thinner (5-6mm) and some sherds up to 15mm thickness
General A red slipped version of GROMIT2 which is a hard fired wheel
description thrown ware with fabric similar to GROMIT1 but with smaller
better sorted grits and less common and smaller mica fragments.
Fabric is thick but thinner than in GROMIT1. Fabric appears in a
range of colours but typically reddish brown or light brown with
grey or dark brown core. Fabric has a rough feel, but the surfaces
have been flattened or smoothed and the smaller size and quantity
of grits means that the surface is more regular than in GROMIT1.
The ware has a heavy feel, and is hard and compact. The slip is thin
and typically dark red.
634
Appendix 5: Ware descriptions
Code GROMIT3
Manufacture
Firing Hard
Thickness Most sherds within the range of 13 to 20mm with some larger
sherds between 22 to 33mm thick
General A thick hard-fired ware with an orange exterior fabric and black
description core. Fabric is compact, thick and heavy with multiple types of
inclusions, including common grits and mica throughout and
vegetable temper which is particularly apparent on surfaces. This
ware is similar in appearance to VORBL but is harder, heavier and
with more obvious grits as opposed to vegetable voids.
635
Appendix 5: Ware descriptions
Code GROMIT3.RS
Surface Surfaces flattened and treated with a thin red slip. On some sherds a
treatment thin micaceous wash also appears to have been added
Decoration None
Manufacture
Firing Hard
General A thick hard-fired ware with an orange exterior fabric and black
description core. Fabric is compact, thick and heavy with multiple types of
inclusions, including common grits and mica throughout and
vegetable temper which is particularly apparent on surfaces. This
ware is similar in appearance to VORBL but is harder, heavier and
with more obvious grits as opposed to vegetable voids. Surfaces
have been treated with a thin pale red slip.
636
Appendix 5: Ware descriptions
Code GROMIT4
Decoration None
Manufacture
Firing Hard
637
Appendix 5: Ware descriptions
Code GRW1
Distinguishing A grit tempered red ware, similar to GROMIT1 and GROMIT2 but
features without the addition of mica
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Surface colour typically 5YR 4/3 Reddish Brown or 10YR 6/4
(Munsell) Light Yellowish brown. Core colour 10YR 5/1 Grey or 10YR 4/1
Dark greyish Brown
Thickness Sherds generally between 6 and 12mm with some larger sherds in
the range of 15 to 28mm thick
638
Appendix 5: Ware descriptions
Code GRW1.RS
Distinguishing A red slipped grit tempered red ware, similar to GROMIT1 and
features GROMIT2 but without the addition of mica
Surface Surfaces appear to have been flattened, and have been treated with
treatment a thin red slip
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Surface colour typically 5YR 4/3 Reddish Brown, 10 YR 4/2 Dark
(Munsell) greyish Brown or 10YR 6/4 Light Yellowish brown. Core colour
10YR 5/1 Grey or 10YR 4/1 Dark greyish Brown. Slip: ranges
from 2.5YR 5/8 Red to 2.5YR 6/6 Red
Thickness Generally between 7 and 9mm thick but occasional thinner sherds
between 4 and 7mm thick
639
Appendix 5: Ware descriptions
Code GW.BLS
Distinguishing A common ware with uniform grey fabric containing roughly
features sorted sub-angular sand grains. This ware has a black slip (often
burnished) on one or two surfaces
Surface Surfaces have been smoothed and flattened and then treated with a
treatment thin or thick black slip on one or two surfaces, but most commonly
the exterior. The slip is usually burnished. The interior surface
frequently displays a thin micaceous wash
Decoration Range of decoration, most commonly incised horizontal lines,
often coupled with oblique notches or lines, notched applique,
wavy lines or a stamped floral/sun design. The stamped decoration
is most commonly found in this ware. Rims often have thick
applique designs or a thin notched applique edge (beading)
Manufacture Wheel-turned
Firing Hard
Colour Exterior surface is black slipped, and often burnished. Fabric is
(Munsell) 10YR 6/1 grey.
Fracture Regular, fairly even and perpendicular to the surfaces
Inclusions Very rare small mica fragments (<1mm) and rare mica dust,
roughly sorted occasional sub-angular sand grains, occasional
poorly sorted angular and sub-angular clear or opaque white/grey
grits (1-3mm). Occasional long thin vegetal voids in fabric
Thickness Sherds generally between 2.5 and 10mm thick with some falling in
a large range of 10-17mm thick
General Similar to GW in being a hard-fired medium grey ware with a
description uniform mid-grey to light bluish-grey fabric but with a black slip.
This ware feels hard and can be relatively lightweight but is
generally finer and harder than GW. Surfaces feel fairly even as
they have been smoothed and flattened and then slipped with a thin
or thick black slip, but the interior surfaces can be slightly rough.
The slip is often burnished and occasionally contains traces of
mica, although mica is rare in the fabric and probably
unintentional. Some distinctive decoration including thick applique
rims and stamped designs.
640
Appendix 5: Ware descriptions
Code GW
Distinguishing A mid grey fabric with rare mica fragments and roughly sorted
features sub-angular sand grains and grits. Some sherds have a pinkish-
orange discolouration from the surrounding soil.
Surface Surfaces have been smoothed and flattened. Sometimes no other
treatment surface treatment is apparent and surfaces are dull and matt, but
frequently sherds have a thin buff/light grey slip usually on the
exterior surface with occasional traces of burnishing
Decoration Range of decoration, most commonly incised horizontal lines,
often coupled with oblique notches, or a wavy line. Rims often
have thick applique designs
Manufacture Wheel-turned
Firing Hard
Colour Fabric is 10YR 6/1 grey. Occasionally has discolouration from soil
(Munsell) giving sherds patches of 10R 6/3 pale red or 10YR 6/4 light
yellowish brown. If slip is present it ranges between 10YR 6/3 pale
brown and 10YR 6/1 grey
Fracture Rough, but fairly even and perpendicular to the surfaces
Inclusions Very rare small mica fragments (<1mm) and rare mica dust,
roughly sorted occasional sub-angular sand grains, occasional
poorly sorted angular and sub-angular clear or opaque white/grey
grits (<2mm)
Thickness Usually within the range of 4 to 10mm thickness with two sherds
measuring slightly less (smallest = 2.5mm) and two slightly more
(thickest = 15mm).
General A hard-fired medium grey ware with a uniform mid-grey to light
description bluish-grey fabric. This ware feels relatively lightweight compared
to other wares and the surfaces feel rough, with thicker sherds
occasionally having a slightly porous appearance. This ware is
often treated with a thin buff or light grey slip which sometimes
appears to have been burnished. Unslipped sherds may have lost
their slip as they tend to be badly eroded.
641
Appendix 5: Ware descriptions
Code HRBC
Distinguishing A thick, heavy red ware with a black core and orange outer fabric.
features
Surface Surfaces flattened but no other surface treatment visible
treatment
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Outer fabric 5YR 6/6 (Reddish yellow) or 5 YR 5/6 (Yellowish
(Munsell) Red), with a dark grey or black core
Fracture Rough and irregular with some laminations, usually an inwards or
outwards facing fracture
Inclusions Occasional long thin or rounded vegetal voids in fabric, occasional
poorly sorted small to medium sub-angular and angular clear or
opaque grey grits (2-6mm). Roughly sorted sand grains and rare
poorly sorted small mica fragments (<2mm)
Thickness All sherds are thick usually ranging between around 16 to 22mm.
Some sherds are slightly thicker with the maximum thickness
observed being 42mm
General A heavy hard-fired and thick-bodied red ware with an orange
description fabric and a black core. This ware has a very irregular fracture and
is quite coarse, with no decoration or surface treatment - perhaps a
coarse storage ware. Clear vegetable temper leaving voids in fabric
and on surfaces, usually long voids but some very clear grain and
husk impressions on surfaces, including obvious rice grain
impressions. Other inclusions are occasional small to medium
grits, which are poorly sorted and in various shapes and sizes, and
rare mica fragments.
642
Appendix 5: Ware descriptions
Code HRBC.RS
Distinguishing A thick, heavy red ware with a black core and orange outer fabric.
features Similar to HRBC but red slipped and not as common
Decoration None
Manufacture Wheel-turned
Firing Hard
643
Appendix 5: Ware descriptions
Code LAST
Distinguishing An abrasive hard fired fabric with orange fabric and black core,
features visible vegetable and sand temper
Surface Surfaces have been flattened and usually treated with a thin red
treatment slip.
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour 2.5YR 6/8 Red, 2.5YR 6/4 weak red with patches of 7.5YR 6/4
(Munsell) Light brown
Inclusions Rare mica dust, common vegetal voids on surfaces and within the
fabric ranging from round voids to long and thin. Roughly sorted
common sand grains, occasional poorly sorted angular small to
medium clear grits (2-4mm)
General A thick, hard heavy ware, with brown to orange fabric and a black
description core. This ware feels rough and abrasive. Surfaces have been
somewhat flattened and are sometimes clearly slipped with a thin
red slip but no other surface treatment is visible and there is no
decoration. Vegetable and sand temper is dominant and the ware
also contains some grits and rare mica dust (natural inclusion from
clay?).
644
Appendix 5: Ware descriptions
Code ORM1
Surface Surfaces smoothed and flattened, no slip but occasionally there are
treatment traces of a thin micaceous wash, particularly on exterior surfaces
but sometimes on interior
Decoration Usually none, but one sherd has incised horizontal lines
Manufacture Wheel-turned
Firing Hard
Colour Surfaces mainly 2.5YR 5/4 weak red, 5YR 6/4 light reddish brown
(Munsell) or 5 YR 6/6 Reddish yellow
Inclusions Dense small to large sub-angular and angular mica fragments (1-
5mm), rare poorly sorted sub-angular and angular clear or opaque
white grits (2-3mm), moderate roughly sorted sand grains, and
occasional small vegetal voids particularly on surfaces
Thickness Sherds generally between 3 and 8.5mm thick, a few sherds in the
range of 9 to 16mm
645
Appendix 5: Ware descriptions
Code ORM1.RS
Distinguishing A red/brown/orange fabric throughout. Large and small mica
features fragments throughout, particularly common on surfaces. Red
Slipped
Surface Surfaces have been smoothed and flattened and then treated with a
treatment thin red slip usually on the exterior but occasionally on both
surfaces. There is also sometimes evidence of an additional
micaceous wash on the exterior or on the interior
Decoration Usually none but a couple of sherds display incised horizontal
lines or small incised notches
Manufacture Wheel-turned
Firing Hard
Colour Surfaces mainly 2.5YR 5/4 weak red, 5YR 6/4 light reddish brown
(Munsell) or 5 YR 6/6 Reddish yellow. Slip is 2.5YR 5/6 Red
Fracture Rough and irregular, occasionally delaminating but generally
perpendicular to the surfaces
Inclusions Dense small to large sub-angular and angular mica fragments (1-
5mm), rare poorly sorted sub-angular and angular clear or opaque
white grits (2-3mm), moderate roughly sorted sand grains, and
occasional small vegetal voids particularly on surfaces
Thickness Most sherds fall between 3 to 8mm in thickness, however 3 sherds
are between 10 and 12mm thick
General A medium to coarse hard-fired and densely micaceous fabric,
description orange/red throughout. Same as ORM1 but with a red slip. This
fabric is more compact and harder fired than ROM, and not as
orange in colour – it is more similar to BOM1.RS but without the
black or grey core. This fabric has a rough feel, although surfaces
have been smoothed and flattened. The thin slip is usually a dark
red colour with small mica fragments or dust mixed in –
occasionally the slip is more orange-red in colour. This ware
typically appears in carinated vessel forms.
646
Appendix 5: Ware descriptions
Code ORM2
Distinguishing A distinctive pale orange fabric with very common small to
features medium mica fragments throughout, more on surfaces. Similar to a
fine version of ROM.
Surface Surfaces have been smoothed and flattened and then most probably
treatment treated with a thin micaceous wash or slip and burnished to
achieve a smooth finish
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour 2.5YR 6/6 Red (Appears a pinkish-orange) fabric and surface
(Munsell)
Fracture Rough fracture, generally perpendicular to the surfaces but
forming small laminations from the mica content
Inclusions Dense small to medium (<4mm) mica sub angular and angular
fragments. Frequent mica dust occasional angular clear grits 2-
3mm poorly sorted, roughly sorted sand grains
Thickness Sherds usually 3 to 10mm thick, a few thinner sherds around 2mm,
and some sherds in the range of 11 to 20mm
General A hard-fired, medium ware with a pale orange/red fabric colour
description throughout. There is dense mica in the fabric in small to medium
sized fragments and an additional micaceous slip/wash. This ware
has a smooth silvery appearance from the surface treatment, and
has a smooth hard feel. Mica content is similar to ROM, but
perhaps slightly less and better sorted so that combined with the
hard firing this is a more compact fabric compared to ROM. Very
distinctive due to the smooth micaceous burnished surface and the
uniform pale orange/red colour. Common in rounded rims and
carinated vessels.
647
Appendix 5: Ware descriptions
Code ORM3
Distinguishing A fairly coarse orange fabric with common medium and small
features mica fragments and moderate angular grits throughout surfaces
and fabric
Surface Surfaces have been smoothed and flattened but there is no other
treatment visible surface treatment
Decoration None
Manufacture Wheel-turned
Firing Medium to hard
Colour Surface is around 5YR 6/6 (reddish yellow) or 7.5YR 6/6 (Reddish
(Munsell) yellow)
Fracture Rough and fairly irregular but generally perpendicular to the
surfaces
Inclusions Frequent small to medium angular mica fragments (1-4mm) and
mica dust, moderate sub-rounded sand grains, common poorly
sorted sub-angular clear grits (2-3mm) and occasional sub-rounded
brown grits (3-4mm). Rare larger sub-angular grits, usually clear
or translucent grey. Occasional vegetal voids, particularly on
surfaces
Thickness Most sherds fall within the range of 3 to 9mm thick, some are
slightly thicker between 9 and 16mm
General This is a relatively coarse, medium-fired fabric similar to BOM7
description but with a uniform orange colour throughout. This fabric has a
rough, gritty feel where mica and dust can rub away and stain the
fingers orange, as in ROM. Unlike ROM however, the mica
content is average and there are a variety of inclusions including
grits, sand grains and some vegetable temper. Although surfaces
appear to have been smoothed and flattened there is no other
visible surface treatment and surfaces can be quite eroded. Appears
to be a utilitarian ware, and often occurs in simple cord-made
bases.
648
Appendix 5: Ware descriptions
Code ORM4
Distinguishing Orange fabric with occasional small mica, mica dust which is more
features common on surfaces, and occasional medium grits
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Surfaces are 5YR 6/6 Reddish yellow or 7.5YR 6/6 Reddish
(Munsell) yellow
Thickness Most sherds are between 3 to 8mm thick, occasional sherds in the
range of 8.5 to 14.5mm
Code ORM4.RS
Distinguishing Orange fabric with occasional small mica, mica dust which is more
features common on surfaces and occasional medium grits. Red Slipped
Surface Surfaces flattened and smoothed and then usually exterior surface
treatment is treated with a thin red slip. Occasionally there are traces of an
additional thin micaceous wash on the exterior surface
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Surfaces are 5YR 6/6 Reddish yellow or 7.5YR 6/6 Reddish
(Munsell) yellow. Range of red shades in slip but generally 2.5YR 5/6 Red to
10YR 5/8 Red
Fracture Rough and irregular but generally roughly perpendicular to the
surfaces
Inclusions Occasional to frequent small to medium sub-angular mica
fragments (1-4mm) and mica dust, occasional poorly sorted small
to medium sub-angular/angular clear grits (2-5mm), and rare
rounded small brown grits (<2mm). Common roughly sorted sand
grains and occasional long, thin vegetal voids particularly on
surfaces.
Thickness Sherds all fall within the range of 3 to 9mm thickness except one
which measures 10.6mm
General A medium, hard-fired ware with a uniform orange-brown fabric
description throughout. Ware is rough to the feel with common grit and sand
inclusions. This fabric also contains frequent mica dust and small
mica fragments, but has lower mica content than in wares ORM1-
ORM3. The fabric is hard and compact, and surfaces have been
flattened and slipped. The thin red slip appears in a range of
shades, usually a dark red or orange-red, and small mica fragments
have been mixed into the slip.
650
Appendix 5: Ware descriptions
Code ORM5
Distinguishing Relatively fine orange fabric with only occasional mica dust and
features small fragments in fabric and on surfaces. Few other inclusions
Surface Surfaces have been flattened and smoothed, and usually no other
treatment surface treatment is visible although occasionally there are traces
of a thin micaceous wash on the exterior or interior surface
Decoration Usually none, although occasional sherds display incised lines and
one appears to be incised or ‘basket’-impressed
Manufacture Wheel-turned
Firing Hard
Colour Surfaces are 5YR 6/6 Reddish yellow or 5YR 5/6 Yellowish red
(Munsell)
Inclusions Common to dense mica dust and common small sub-angular mica
fragments (<1mm), occasional roughly sorted fine sand grains, and
rare poorly sorted small angular clear or opaque white grits
(<2mm)
Thickness Most sherds fall between 3 and 8mm thick, a couple are slightly
thinner than 3mm (c. 2.5mm) and a few sherds are in the range of
9 to 11mm thick
General A medium thickness, hard-fired micaceous ware, with frequent to
description dense mica dust and small fragments. This ware has a rough sandy
feel although surfaces appear to have been smoothed and flattened,
and fine sand grains do appear as an inclusion. This ware is a
uniform orange/sandy-brown colour throughout, and resembles a
red ware except in the high quantity of mica dust which indicates it
is an intentional inclusion. The fabric has a hard, compact feel, and
also has a fairly regular fracture as the inclusions are small and the
fabric medium to fine in quality.
651
Appendix 5: Ware descriptions
Code ORM5.RS
Distinguishing Relatively fine orange fabric with only occasional mica dust and
features small fragments in fabric and on surfaces. Few other inclusions.
Red slipped.
Surface Surfaces have been flattened and smoothed, and then treated with a
treatment thin red slip. Occasionally there are traces of a thin micaceous
wash on the exterior or interior surface
Decoration Usually none but one sherd displays incised lines
Manufacture Wheel-turned
Firing Hard
Colour Surfaces are 5YR 6/6 Reddish yellow or 5YR 5/6 Yellowish red.
(Munsell) Slip is usually around 2.5YR 5/6 Red
Fracture Rough but regular and perpendicular to the surfaces
Inclusions Common to dense mica dust and common small sub-angular mica
fragments (<1mm), occasional roughly sorted fine sand grains, and
rare poorly sorted small angular clear or opaque white grits
(<2mm).
Thickness Most sherds are in the range of 3 to 7, but a couple are slightly
thicker between 8 and 10.7mm
General A medium thickness, hard-fired micaceous ware, with frequent to
description dense mica dust and small fragments. This ware has a rough sandy
feel although surfaces appear to have been smoothed and flattened,
and fine sand grains do appear as an inclusion. This ware is a
uniform orange/sandy-brown colour throughout, and resembles a
red ware except in the high quantity of mica dust which indicates it
is an intentional inclusion. The fabric has a hard, compact feel, and
also has a fairly regular fracture as the inclusions are small and the
fabric medium to fine in quality. The thin red slip ranges in colour
from an orange-red to dark red, and has small mica fragments
mixed in.
652
Appendix 5: Ware descriptions
Code PATCHY
Distinguishing A rough surfaced medium to thin-bodied orange and black
features micaceous fabric, patchy and pitted on the surface. Mica and
vegetable temper.
Surface Surfaces appear to have been flattened but no other surface
treatment treatment is visible
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Surfaces mainly 10YR 5/4 (Yellowish brown), mottled with 10YR
(Munsell) 4/2 (Dark Greyish Brown). Core 10YR 3/1 (Very dark Grey)
Fracture Rough and irregular fracture with uneven break, sometimes
perpendicular to the surfaces and sometimes inwards or outwards
facing
Inclusions Frequent small to medium angular mica fragments (1-4mm),
poorly sorted sub-angular and angular clear or translucent grey
grits (1-3mm) and rounded brown grits (<2mm), very rare large
clear or translucent grits (>5mm). Common rounded and long, thin
vegetal voids visible on surfaces and within fabric.
Thickness Sherds are mostly 3.5 to 8mm with the majority between 4 and
6mm thick. Some sherds are in a larger size range of 8 to 12.3mm
General A hard-fired, relatively coarse micaceous ware with an orange
description surface fabric and dark grey or black core. Distinctive patchy
appearance to exterior, with areas of darker brown on orange –
unclear whether this is part of manufacture or something that has
happened after deposition as most appear to be from the same site.
This ware has a coarse, rough feel and an irregular fracture, with a
high number of inclusions, including mica fragments and grits. The
surfaces have been slightly smoothed or flattened on the exterior
but there is no other surface treatment, and no sign of decoration.
This ware is almost exclusively restricted to small straight sided
bowls with a cord pulled bottom.
653
Appendix 5: Ware descriptions
Code RBC1.RS
Distinguishing Hard, relatively fine red/orange ware with occasional mica. Similar
features to RED1 but with a black core. Red slipped.
Surface Surfaces have been smoothed and flattened and then treated with a
treatment red slip. Slip can be thin or thick and is often burnished.
Occasional traces of an additional thin mica wash on the exterior
and sherds may also have a micaceous wash on the interior surface
Decoration Usually none; most common decoration is incised horizontal lines.
Occasional sherds have incised notches or a small incised applique
line (beaded appearance)
Manufacture Wheel-turned
Firing Hard
Colour Surface fabric is 5 YR 6/6 reddish yellow or 5 YR 5/6 yellowish
(Munsell) red with a dark grey or black core. Slip is usually 2.5YR 5/6 Red
Fracture Rough but regular and perpendicular to the surfaces
Inclusions Rare mica dust, more on surfaces. Very occasional clear vegetal
voids are visible on surfaces and in fabric, occasional roughly
sorted sub-angular small clear grits (2-3mm). Roughly sorted sub-
angular sand grains
Thickness Most sherds fall within the range of 2.5 to 8mm thickness (with the
majority measuring between 4 and 6mm), with a few slightly
thicker sherds measuring between 8 and 11.3mm.
General A hard-fired orange fabric with a black/dark grey core. This ware
description feels hard and fairly smooth, and is relatively fine in comparison to
other fabrics. There are few obvious inclusions, mainly fairly well
sorted sand grains and occasional small grits. The mica dust that is
apparent in some sherds is at a low level and appears to be a
natural inclusion from the parent clay. Mica may be more apparent
on surfaces due to a thin micaceous wash as part of the surface
treatment. Surfaces feel quite smooth and are well slipped. The
slip, which is usually a dark red but can appear more orange-red in
some sherds, may be thick or thin and is often shiny from
burnishing. The slip sometimes contains small mica fragments.
This ware appears mainly in restricted jar forms.
654
Appendix 5: Ware descriptions
Code RBC2.RS
655
Appendix 5: Ware descriptions
Code RBC3
Distinguishing A hard, sandy red/orange ware with a black core. This ware has
features fairly rough surfaces.
Surface Surfaces have been smoothed and flattened but no other surface
treatment treatment is generally obvious. Occasional sherds have traces of a
thin micaceous wash on the exterior or interior surface
Decoration Generally none but occasional sherds have incised horizontal lines.
One has both horizontal lines and incised oblique lines, and one
sherd has a stamped floral/sun motif.
Manufacture Wheel-turned
Firing Hard
Colour Surface fabric is 5 YR 6/6 reddish yellow or 5 YR 5/6 yellowish
(Munsell) red with a dark grey or black core
Fracture Rough, but fairly regular and generally perpendicular to the
surfaces
Inclusions Moderate mica dust, mostly on surfaces and rare small sub-angular
mica fragments (<2mm), occasional thin long vegetal voids, more
on surfaces. Occasional poorly sorted sub-angular small clear grits
(2-3mm) and rare rounded small brown grits (1-2mm). Moderate
roughly sorted small sand grains
Thickness Most sherds measure between 3 and 9mm thick, some sherds are
slightly thicker and measure from 10 up to 23mm thick
656
Appendix 5: Ware descriptions
Code RBC3.RS
Distinguishing A hard, sandy red/orange ware with a black core. This ware has
features fairly rough surfaces but is red slipped
Surface Surfaces have been smoothed and flattened, and then treated with a
treatment usually thin red slip (slip is occasionally thick). Occasional sherds
show evidence of burnishing and others have traces of a thin
micaceous wash on the exterior or interior surface.
Decoration Usually none; one sherd displays incised horizontal lines and one
rim has pinched applique decoration
Manufacture Wheel-turned
Firing Hard
Colour Surface fabric is 5 YR 6/6 reddish yellow or 5 YR 5/6 yellowish
(Munsell) red with a dark grey or black core. Slip is usually 2.5YR 5/6 Red.
Fracture Rough, and fairly irregular but generally perpendicular to the
surfaces
Inclusions Occasional small sub-angular mica fragments (<2mm) and
moderate mica dust. Occasional long thin vegetal voids, more
common on surfaces. Occasional poorly sorted sub-
rounded/rounded brown grits (2-3mm, occasionally >3mm) and
common poorly sorted sub-angular and angular small clear or
opaque white grits (2-3mm but with occasional larger grits).
Moderate roughly sorted sand grains
Thickness Most sherds measure between 4 and 10mm thick, occasional
sherds measure from 12 up to 28mm thick
General A hard-fired, medium fabric with a sandy-brown/orange exterior
description fabric and dark grey to black core. This ware is similar to
RBC2.RS, but has a coarser fabric with more common and larger
inclusions - The fabric contains occasional large grits, and a higher
sand content. The mica fragments and dust are more common on
surfaces and still appear to be a natural inclusion. The fabric has a
hard and rough feel, despite the surfaces having been smoothed
and flattened, before being slipped. The red slip is usually a dark
red but sometimes appears more orange-red, and often has small
mica fragments mixed in.
657
Appendix 5: Ware descriptions
Code RBC4
Distinguishing Rough orange ware with a black core. Occasional mica but still
features classed as a red ware. Similar to RED4 but with black core
Surface Surfaces are often eroded but have been smoothed and flattened.
treatment There may be traces of red slip on a couple of sherds but this is
unclear due to erosion. Most appear to have had no other surface
treatment
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour Surface fabric is 5YR 6/6 reddish yellow, or 5YR 5/6 yellowish
(Munsell) red
Fracture Rough and irregular but generally perpendicular to the surfaces
Inclusions Rare small sub-angular mica fragments (<2mm) and mica dust.
Frequent poorly sorted sub-angular and angular clear or opaque
white grits (2-4mm) and occasional sub-rounded brown grits (3-
5mm). Roughly sorted moderate sand grains and occasional to
frequent vegetal voids throughout fabric and on surfaces
Thickness Most sherds fall within the range of 4 to 9.5mm thickness, with
some larger sherds measuring between 11.5 to 20mm.
General A relatively coarse but hard-fired red ware, similar to RED4 but
description with an orange outer fabric and a dark grey/black core. This ware
has fairly frequent inclusions, mainly grits and sand grains but also
occasional mica. The mica is still infrequent enough to appear to
be a natural inclusion. This fabric has a rough, sandy feel and often
eroded surfaces, although most have clearly been flattened. There
is no clear surface treatment beyond smoothing and flattening
although some sherds may have very small traces of a red slip.
This ware commonly appears as bases/small bowls, and has no
decoration indicating it is fairly utilitarian.
658
Appendix 5: Ware descriptions
Code RED.BLS
Distinguishing A red/orange fabric with a black or dark brown slip on one or two
features surfaces.
Surface Surfaces have been smoothed and flattened and then treated with a
treatment thin black or dark brownish-grey slip on one or both surfaces. The
slip is occasionally burnished and there are rare traces of an
additional thin micaceous wash.
Decoration Usually none but some sherds have incised horizontal lines, and
there are occasional examples of incised notches
Manufacture Wheel-turned
Firing Hard
Colour Outer surface is black or 2.5YR 3/1 reddish brown, and inner
(Munsell) surface is approximately 5YR light reddish brown to 5YR 5/3
reddish brown
Inclusions Rare to occasional mica dust and occasional roughly sorted sub-
angular sand grains. Occasional poorly sorted sub-angular small
clear grits (2-3mm) and rare rounded brown grits (3-4mm).
Occasional long, thin vegetal voids
Thickness Sherds are within the range of 3 to 8.5mm thick with one sherd
measuring 11mm thick
659
Appendix 5: Ware descriptions
Code RED.OF
Manufacture Wheel-turned
Firing Hard
Colour Patchy, generally around 7.5YR 5/3 Brown, 7.5YR 6/4 Light brown
(Munsell)
Inclusions Rare small (<1mm) sub-angular mica fragments and mica dust,
occasional sand grains and occasional poorly sorted small clear grits
(<2mm) and sub-rounded brown grits (<2mm)
Thickness General range between 3 and 10mm with a few slightly thicker
sherds up to 18mm
General This is a general ware category used to describe red wares which
description have been over fired and appear discoloured and difficult to identify
as a result. These red-wares are generally hard fired and
occasionally treated with a thin red slip.
660
Appendix 5: Ware descriptions
Code RED1.RS
Distinguishing A red/orange fabric with hard firing and fine inclusions. Red
features slipped. Very similar to RBC1.RS but orange fabric throughout
Surface Surfaces have been smoothed and flattened and then treated with a
treatment red slip. The slip is usually thick, but may be relatively thin, and is
often burnished. There are occasional traces of a micaceous wash,
particularly on the interior surface.
Decoration Usually none but occasional sherds display incised horizontal lines
Manufacture Wheel-turned
Firing Hard
661
Appendix 5: Ware descriptions
Code RED2
Distinguishing A red/orange fabric, rougher than RED1.RS but with generally fine
features inclusions. Occasional mica.
Decoration Occasional sherds with incised horizontal lines. One sherd with an
applique band with thumb impressions and another with incised or
impressed basket decoration
Manufacture Wheel-turned
Firing Hard
Colour 5YR 6/6 (Reddish yellow) or 5YR 5/6 (Yellowish red).
(Munsell)
Fracture Rough but generally regular and perpendicular to the surfaces
Inclusions Very rare small sub-angular mica fragments (<2mm) and rare mica
dust, more on surfaces. Poorly sorted occasional sub-angular and
angular clear grits (2-3mm) and rare rounded brown grits (2-3mm).
Roughly sorted sub-angular sand and occasional thin and long
vegetal voids.
662
Appendix 5: Ware descriptions
Code RED2.RS
Distinguishing A red/orange fabric, rougher than RED1.RS but with generally fine
features inclusions. Occasional mica. Red slipped.
Surface Surfaces smoothed and flattened. Treated with a red slip, usually
treatment thin and occasionally burnished.
Manufacture Wheel-turned
Firing Hard
Colour 5YR 6/6 (Reddish yellow) or 5YR 5/6 (Yellowish red). Slip is
(Munsell) usually 2.5YR 5/6 (Red)
Inclusions Very rare small sub-angular mica fragments (<2mm) and occasional
mica dust. Poorly sorted, occasional sub-angular and angular clear
grits (2-3mm), rare poorly sorted larger angular grits and rounded
brown grits (2-3mm). Roughly sorted sub-angular sand and
occasional thin vegetal voids.
General A hard-fired medium red ware, with orange fabric throughout. The
description fabric has a hard, relatively rough feel and quite compact surfaces,
which have been smoothed and flattened. This fabric is not as hard
or fine as RED1.RS. This ware is treated with a red slip which is
usually quite thin and can be eroded. The slip is often burnished and
usually dark red or orange-red in colour. Although the inclusions
are usually relatively fine, occasional sherds contain rare large grits.
663
Appendix 5: Ware descriptions
Code RED3
Manufacture Wheel-turned
Firing Medium
Colour Fabric is 2.5YR 6/8 (Red), varying to 5YR 6/8 (Reddish yellow) in
(Munsell) some sherds
Fracture Smooth, but often worn. Generally regular and perpendicular to the
surfaces
Thickness Usually between 3 and 10mm, with occasional thicker sherds in the
range of 10 to 16mm
664
Appendix 5: Ware descriptions
Code RED3.RS
Surface Surfaces flattened and treated with a thin red slip on the exterior
treatment surface
Decoration Usually none; one sherd has incised or impressed basket decoration
and another has applique decoration.
Manufacture Wheel-turned
Firing Medium
Colour Fabric is 2.5YR 6/8 (Red), varying to 5YR 6/8 (Reddish yellow) in
(Munsell) some sherds. Slip is usually 10YR 5/8 (Red)
Fracture Smooth, but often worn. Generally regular and perpendicular to the
surfaces
General A medium-fired, red ware with a red slip and orange fabric
description throughout. Red slipped version of RED3 but less common. It may
be that RED3 originally had a red slip as often the red slip on
RED3.RS is worn away or fragmentary, and the surfaces of this
ware are powdery. The slip is thin and usually orange-red in colour.
Soapy or powdery feel, and the surface comes away when rubbed.
Usually relatively fine to medium with fine inclusions but some
rougher sherds have a sandier or gritty feel.
665
Appendix 5: Ware descriptions
Code RED4
Manufacture Wheel-turned
Inclusions Very rare mica dust. Frequent round and straight, thin vegetal voids
on surfaces and in the fabric. Frequent to common poorly-sorted
sub-angular and angular small clear grits (2-4mm) and opaque white
grits (often >4mm), occasional poorly-sorted rounded/sub-rounded
brown grits (4mm). Common, roughly sorted sub-angular sand
grains.
Thickness Sherds vary between 4 and 10mm in thickness. One sherd measures
3mm and there are also occasional sherds up to 13.7mm thick
666
Appendix 5: Ware descriptions
Code RED4.RS
Distinguishing A relatively coarse and gritty red/orange gritty fabric, with a red
features slip. Not commonly found
Decoration None
Manufacture Wheel-turned
Inclusions Very rare mica dust. Frequent round and straight, thin vegetal voids
on surfaces and in the fabric. Frequent to common poorly-sorted
sub-angular and angular small clear grits (2-4mm) and opaque white
grits (often >4mm), occasional poorly-sorted rounded/sub-rounded
brown grits (4mm). Common, roughly sorted sub-angular sand
grains.
667
Appendix 5: Ware descriptions
Code RED5
Distinguishing A fine pale orange fabric, which is softer fired than RED1. Rarely
features found
Decoration None
Manufacture Wheel-turned
Firing Medium
Inclusions Very rare mica dust. Rare, poorly sorted rounded brown grits (2-
3mm) and occasional fairly well-sorted sand grains
Thickness Most sherds are in the range of 5mm thickness, one sherd is 10mm
thick
668
Appendix 5: Ware descriptions
Code RED5.RS
Distinguishing A fine pale orange fabric, which is softer fired than RED1. Red
features slipped.
Surface Surfaces smoothed and flattened with a thin red slip, which may be
treatment burnished or treated with a micaceous wash
Decoration None, one sherd shows traces of incised horizontal lines
Manufacture Wheel-turned
Firing Medium
Colour 5 YR 6/6 (Reddish yellow) or 5/6 (Yellowish red) with slip 2.5YR
(Munsell) 5/6 (Red)
Fracture Fairly smooth, regular and perpendicular to the surfaces
Inclusions Very rare mica dust. Rare poorly sorted rounded brown grits (2-
3mm) and occasional fairly well-sorted sand grains. Very rare
vegetal voids on surfaces and white calcitic inclusions
Thickness Sherds are usually between 4 and 7mm thick. A couple of sherds are
<4mm and a couple of sherds are thicker, up to 8.3mm
General A relatively fine medium-fired pale orange fabric, which is similar
description to RED1 but is softer fired. This ware has few inclusions, a smooth
and often soapy/powdery feel with smoothed surfaces. There has
been no other surface treatment. This fabric is rare in the
assemblage but more common than the unslipped RED5, which
may represent sherds where the red slip has worn away. The
exterior surface has been treated with a thin red slip, usually orange-
red in colour. The slip may also have a thin micaceous wash or
some evidence of burnishing.
669
Appendix 5: Ware descriptions
Code ROM
Distinguishing Very rough orange fabric throughout, with very common large
features angular mica fragments
Surface Surfaces appear to have been flattened and perhaps smoothed, but
treatment no other surface treatment is visible and sherds are now heavily
eroded and rough
Decoration None
Manufacture Wheel-turned
Firing Medium
Colour 5YR 6/6 reddish yellow fabric
(Munsell)
Fracture Rough, irregular fracture, often delaminating slightly into sheets
due to high mica content. Frequently an outwards facing fracture
Inclusions Very dense large angular mica fragments, size ranging from 1 –
6mm. Occasional poorly sorted sub-angular clear and translucent
grey grits (2-3mm) and poorly sorted sub-angular/rounded sand
grains
Thickness Most sherds are in the range of 3 to 8mm in thickness but some
sherds also measure between 9 and 15mm with one measuring
30mm thickness.
670
Appendix 5: Ware descriptions
Code TIG
Distinguishing A thin brittle grey ware with a range of slip colours, usually brown
features or black.
Surface Surfaces have been smoothed and flattened, and the exterior surface
treatment has been treated with a thin brown/grey/black slip, as well as
occasionally the interior surface. Occasional sherds also show
burnishing of the slip or a micaceous wash.
Decoration Usually none, one sherd shows incised horizontal lines
Manufacture Wheel-turned
Firing Hard
Colour 5 YR 6/1 Grey with 5 YR 4/3 or 4/2. Occasional Black slip, and
(Munsell) some reddish brown surfaces from soil.
Fracture Fairly smooth, regular and perpendicular to the surfaces.
Inclusions Occasional mica dust, and occasional thin, long vegetal voids on
surfaces. Fairly well-sorted small sub-angular clear grits (2mm) and
rare larger sub-angular clear grits, occasional fairly well-sorted sand
grains
Thickness Sherds are between 2 and 6mm thick with most sherds in the range
of 3-4mm thick
General A hard-fired, thin brittle grey ware. This is a medium to relatively
description fine fabric which is grey throughout. This ware contains occasional
mica dust throughout the fabric and on surfaces, and moderate sand
grains and small grit inclusions spread evenly throughout the fabric.
The exterior surface, and occasionally the interior, has been treated
with a thin slip. The slip appears in a range of brown/grey/black
shades and is occasionally burnished or treated with a micaceous
wash. The slip is most commonly brownish-grey. It is unclear
whether this fabric constitutes a separate ware or represents thin
body sherds of another ware, as it only occurs as body sherds and
has a range of slips.
671
Appendix 5: Ware descriptions
Code TOR
Manufacture Wheel-turned
Firing Hard
Thickness Most sherds fall between 2 and 5mm thick, a couple are just over
5mm thick
General A hard-fired thin brittle orange ware, with orange fabric and
description occasionally a thin grey or brown core. The fabric contains frequent
mica dust but this occurs more on surfaces and does not qualify it as
a micaceous fabric. Very similar to TORRS but lacking the slip and
less common; however with some sherds it could be that the slip has
worn away. It is unclear whether this constitutes a separate ware
type or whether this represents the thin body sherds of another ware
such as RED2/RBC2. It only occurs as body sherds and is not
common.
672
Appendix 5: Ware descriptions
Code TORM
Distinguishing A rare thick-bodied orange fabric with grit and mica temper, similar
features to ORM3.
Decoration One rim has no decoration, the other has an applique fringe
Manufacture Wheel-turned
Firing Hard
673
Appendix 5: Ware descriptions
Code TORRS
Surface Surfaces are smoothed and flattened and the exterior has been
treatment treated with a thin red/brown slip and then frequently burnished.
Occasional sherds show clear burnishing marks, and some have thin
micaceous washes.
Decoration None
Manufacture Wheel-turned
Firing Hard
Colour 5 YR 6/6 (Reddish yellow) or 5/6 (Yellowish red). Slip can be
(Munsell) between 2.5YR 5/6 (Red) and 2.5YR 4/4 (Dusky red)
Thickness Sherds generally fall between 2 and 5mm thick, with occasional
sherds around 5 and 6mm thick.
General A hard-fired thin brittle orange ware, with orange fabric and
description occasionally a thin grey or brown core. The fabric contains frequent
mica dust but this occurs more on surfaces and does not qualify it as
a micaceous fabric. The ware has been treated with a thin dark red
or brownish-red slip, which is often clearly burnished. It is unclear
whether this constitutes a separate ware type or whether this
represents the thin body sherds of another ware such as
RED2/RBC2. It only occurs as body sherds and is not particularly
common.
674
Appendix 5: Ware descriptions
Code TOW
Distinguishing An abrasive hard fired fabric with vegetable and sand temper.
features
Decoration None
Manufacture Wheel-turned
Firing Hard
Inclusions Rare small angular mica fragments. Common poorly sorted angular
clear grits (2-4mm) and occasional larger clear grits (>4mm),
occasional poorly sorted small rounded brown grits (2-3mm).
Roughly sorted common sand grains. Common long, thin vegetal
voids on surfaces and within the fabric
Thickness Most sherds are within the range of 7-13mm thickness with several
thicker sherds measuring between 16 and 22mm
675
Appendix 5: Ware descriptions
Code VORBL
Thickness Sherds are usually between 8 and 16mm thick but some larger
sherds range between 17-27mm
676
Appendix 5: Ware descriptions
Code VORBL.RS
Surface Surfaces flattened and then treated with a thin red slip. No other
treatment surface treatment visible
Decoration Usually none, but one sherd has incised oblique lines
Manufacture Wheel-turned
Firing Hard
Colour Surface fabric is 5YR 6/6 (Reddish yellow) or 5YR 5/6 (Yellowish
(Munsell) red) with a dark grey/black core. Slip is 2.5YR 5/6 Red
Fracture Rough and irregular, generally perpendicular to the surfaces but
occasional laminations from inclusions
Inclusions Frequent small to medium angular mica fragments (1-4mm),
common voids from vegetable temper, particularly on surfaces,
leaving long thin voids and rounded impressions. Occasional,
roughly sorted sand grains, and occasional grog. Occasional poorly
sorted small angular clear grits (2-3mm), occasional small rounded
brown grits (<2mm) and rare larger clear grits (>4mm).
Thickness Two sherds measure around 14mm and two measure 18 and 22mm
thickness
677
Appendix 5: Ware descriptions
Code VORM
Thickness Sherds usually fall within the range of 8 to 14mm thick, a couple
range from 16 to 22mm thick.
678
Appendix 6: Ceramic Typology
This appendix concerns the types of vessels identified during the ceramic analysis.
Relatively little comparative data is available; the majority of the Rāmṭek ceramic
types do not sufficiently resemble the published examples in the excavation reports
consulted. Where a parallel type was identified in published research, this has been
included along with the type’s subsequent broad date range within the project
chronology.
Following the overview of the ceramic types is a reference table outlining the main
chronological phasing referred to in excavation reports.
679
Appendix 6: Ceramic Typology
Type General Type Description Published Parallel Period Project Notes Unique
Phase Sherd
No.
1 Shallow Characteristic shallow bowl/dish of Nevasa: T.28d & T.28g, dish Nevasa P.IV & I/II Only in
bowl/dish Black and Red ware. The vessel has in burnished B&R ware, P.V; Black and
an in-turned featureless rim with a T.59a, dish in burnished Kaundinyapur Red Ware
rounded terminus. The body of the B&R ware; Kaundinyapur: (2001), Pre-
vessel is convex and curves in to a T.39-3 in coarse Red and Mauryan;
flat bottom. Exterior of the vessel Black ware; Tuljapur Garhi, Tuljapur Garhi,
sides may be decorated with incised T.13; Paunar: T.11A in B&R Phase B (Jorwe);
lines ware Paunar P.1
2 Shallow Characteristic shallow bowl/dish of Nevasa: T.28h, wide dish in Nevasa P.IV & I/II Only in
bowl/dish Black and Red ware with everted B&R ware and T.59, dish in P.V; Black and
rim. Vessel has convex, curved burnished B&R; Kaundinyapur Red Ware
sides and a flat bottom. The rim Kaundinyapur: T.53-10 in (2001), Late
turns slightly outwards with a B&R ware; Pauni T.17, Sātavāhana;
rounded terminus and may be bowl in medium fabric B&R Pauni (1998),
grooved on exterior. ware; Tuljapur Garhi, T.17 P.III; Tuljapur
Garhi, Phase A
(Malwa)
3 Bowl Small shallow bowl with curved Maheshwar & Navdatoli, Maheshwar P.V; II 5249
sides, and small rounded rim. The similar to T.99, burnished Pauni (1998)
rim is grooved on the outside red slipped ware; Pauni, P.IV; Timbarva,
T.11, red slipped red ware P.1;
bowl; Timbarva, T.25, plain Kaundinyapura
red ware bowl; A, P.IV; Paunar
Kaundinyapura T.171, red P.IIA
ware bowl; Paunar T.44E
tan ware bowl
4 Bowl Rounded rimless bowl with convex Maheshwar & Navdatoli: Maheshwar & II
sides and slightly narrower mouth Similar to T.84b, rimless Navdatoli, P.IV;
than the widest point of the vessel bowl in burnished red ware; Pauni (1968), P.II
body. Pauni, T.33, red slipped
ware
680
Appendix 6: Ceramic Typology
5 Bowl Small to medium sized bowl with a Timbarva, similar to T.67 , P.II II
wide mouth and small rim. The rim in plain red ware
is grooved or undercut on the
exterior, and usually has a flattened
top. The rim turns slightly inwards.
The sides of the vessel are convex
and curve downwards into a
shallow to medium-depth bowl
6 Bowl Wide mouthed medium-sized bowl
with relatively straight sides. The
rim has a flattened top but a
featureless exterior. The top of the
rim may feature small grooves
7 Bowl Wide mouthed, shallow small bowl 3179
with convex curved sides. The
flattened rim turns slightly inwards.
681
Appendix 6: Ceramic Typology
11 Bowl Simple small bowl with straight Devnimori, T.1; Baroda, Devnimori, c. II/III/IV
sides, a small base which has been T.8d; Pauni, T.60, dull red 2nd-4th century
string cut, and a featureless rim ware bowl; Shamalaji, T.55, to 7th/8th century
which varies from rounded slightly bowl with thread marks on AD; Baroda, P.II;
pointed at the terminus. Very base; Timbarva, T.61, bowl Pauni (1968) P.II;
common type in crude red ware (Pankti); Shamalaji P.II-
Bhokardhan, T.1 in coarse III; Timbarva P.I;
red ware Bhokardhan P.IA
12 Bowl Medium-sized bowl with straight 998
sides and an up-turned rounded rim.
The sides of the vessel have rows of
grooves for decoration.
13 Bowl Medium-sized, wide-mouthed bowl 4870
with straight sides sloping inward
to the base. The sides are grooved.
The rim is undercut on the outside
while the interior is grooved.
14 Bowl Small bowl with narrow base and
relatively straight sides leading to a
wide mouth. The rim is small and
slightly undercut on the exterior,
with a triangular section. The walls
of the vessel are relatively thin.
682
Appendix 6: Ceramic Typology
16 Bowl Small wide-mouthed shallow bowl Pauni, T.51 in red slipped Pauni (1968), P.II II
with slightly convex sides. The rim ware
is undercut on the exterior and
triangular in section with a rounded
terminus. The interior of the rim
may be slightly grooved.
17 Shallow bowl Wide-mouthed shallow bowl with a Vaisali, similar to T.6, in red P.III II
bulbous rounded rim. The rim is ware
undercut on the exterior and
grooved on the interior. The sides
of the bowl are convex and turn
quite sharply inwards. The sides of
the bowl are relatively thick and
towards the bottom of the body, the
vessel may bear incised decoration
683
Appendix 6: Ceramic Typology
684
Appendix 6: Ceramic Typology
24 Shallow Wide-mouthed, shallow bowl with Dwarka, T.170, basin in Dwarka, P.IV; V
bowl/dish thick up-turned rim. The rim either burnished black ware; Nevasa, P.VI;
has a rounded or pointed terminus Nevasa: T.117c, shallow Kaundinyapura
which turns slightly inwards. The 'dough' in coarse red ware; (1968) P.VI
rim is often undercut on the exterior Kaundinyapura similar to
before the body curves inwards. T.110, red ware basin
The sides of the bowl are either
straight of slightly convex.
25 Shallow bowl Small shallow bowl with an in- Sanchi: Similar to T.6, Sanchi 1st to 4th II/III
turned rim, with a pointed terminus. shallow bowl in micaceous century AD;
The body of the bowl and the rim ware; Devnimori: T.66, Devnimori c.
turn sharply inwards from the point broad mouthed bowl in 2nd-4th century
on the shoulder of the bowl. The micaceous ware; Shamalaji, to 7th/8th
angle of the widest point of the T.134, dish in micaceous century;
bowl is less than 90 degrees. ware Shamalaji from
P.IIA
26 Shallow Wide-mouthed bowl with an Sanchi: Similar to T.27, 6th century B.C. I
bowl/dish incurved and externally collared Basin in various fabrics to 3rd/2nd
rim with small undercut. century B.C.
Sometimes the rim features an
internal groove, which creates a
slight overhang.
27 Bowl Large, fairly steep-sided bowl with Brahmapuri: T.5a, Dish with Brahmapuri, V
in-turned and pointed rim. The rim chocolate slip; Hastinapura, Bahmani;
has an interior and exterior No. Xia Hastinapura, late
projection, giving a triangular 11th to early 15th
profile. There is a small ridge century AD
beneath the rim and occasionally
the body of the bowl is decorated
with incised lines or notches.
685
Appendix 6: Ceramic Typology
28 Bowl Wide-mouthed bowl with an in- Maheshwar & Navdatoli: Maheshwar & II
turned rim, with a rounded terminus T.100 & 100b, Basin in Navdatoli, P.V;
The shoulder of the vessel at the burnished red slipped ware; Nevasa, P.IV
base of the rim has a small ridge, Nevasa: Similar to T.49,
from which the rim and sides of the Basin in red washed ware;
bowl turn inwards. The interior Nasik, T.19b, dish in red
angle at the point is just over 90 ware; Shamalaji, T.48 in
degrees. The shoulder of the bowl plain/burnished black ware
may be decorated with horizontal
incised lines. Due to the in-turned
rim, the mouth of the bowl is
slightly narrower than the shoulder.
29 Flanged bowl Wide-mouthed bowl with an in- Maheshwar & Navdatoli: Maheshwar & ?
turned rim with a rounded terminus. Similar to T.79, Kunda in Navdatoli, P.IV;
The widest part of the bowl tat the coarse red ware; Dwarka: Dwarka P.IV;
end of the rim has an undercut Similar to T.171, basin in Nasik, P.IIA
projection which droops over. The burnished black ware
outer part of the rim and projection
is grooved, and the shoulder has a
slight ridge. The interior of the rim
is also slightly undercut and droops
over. The sides of the bowl are
fairly straight and it appears to be
relatively deep
686
Appendix 6: Ceramic Typology
32 Carinated bowl Wide-mouthed carinated vessel Maheshwar & Navdatoli: Maheshwar & III/IV
with short neck, sloping shoulder Similar in external profile to Navdatoli, P.VI;
and a small ridged carination. The T.135c carinated vessel in Dwarka, P.III
short rim varies between a rounded red slipped ware; Dwarka:
form and a slightly more pointed Sherd 3359 shares
form. The rim extends internally to similarities with T.99, Pot
marginally restrict the opening of with ridged carination in
the vessel so that it is narrower than coarse black ware
the widest point of the bowl. The
body of the bowl turns inwards
from the carination
687
Appendix 6: Ceramic Typology
34 Carinated bowl Wide-mouthed, medium-sized Maheshwar & Navdatoli, Maheshwar, P.IV II 4255
carinated bowl with a rounded, T.76d, carinated bowl in ; Kaundinyapura
mostly featureless rim. Below the coarse red ware; (1968), P.IV
rim is a pointed ridge from which Kaundinyapura, similar to
the body extends inwards. T157, red micaceous ware
vessel
36 Lid-cum-Bowl Wide-mouthed lid-cum-bowl with a Pauni: T.25, lid-cum-bowl in Pauni (1998) II/III
small featureless rim and long red slipped red ware; Pauni P.IV; Pauni
flange around the widest part. The (1972), T.58, dull red ware; (1972) P. II,
flange is angled slightly upwards Kaundinyapur: T.19-1; Kaundinyapur,
and the neck and body both extend Maheshwar & Navdatoli: Early Historic;
inwards from this point. The T.98 in slipped red ware; Maheshwar &
interior angle where the rim and Nasik, T.23a lid in red ware; Navdatoli, P.V;
body meet is slightly over 90 Timbarva T.71 in plain Nasik, P.II;
degrees. The sides of the vessel black ware; Paunar T.95B Timbarva, P.I;
curve inwards to a shallow bowl red slipped lid Paunar P.III
37 Lid-cum-Bowl Small lid-cum-bowl with a vertical Baroda: T.55, bowl/dish P.III IV
featureless rim. Below the rim is a with incurved rim and
small ridge or projection to enable carination
the lid-cum-bowl to fit another
vessel. From this small flanged
waist, the body slopes inwards to a
shallow bowl.
688
Appendix 6: Ceramic Typology
38 Lid-cum-Bowl Small lid-cum-bowl with a straight Baroda: Similar to T.23 & Baroda P.II; II/III
neck and vertical or slightly out- 23a, Carinated dish; Paunar: Paunar P.III
turned rim with a flattened top. The T.164A, carinated cover
vessel has a small ridge at the
shoulder from which the body
extends inwards to create a shallow
vessel. The sides of the bowl are
initially fairly straight
40 Shallow dish Large wide-mouthed, shallow dish Sanchi: Similar to T.41a, Sanchi, phasing
with thin sides and a slightly thicker bowl in Late Medieval grey unknown
rim. The rim is either rounded or ware & T.43, shallow bowl
slightly squared off with a slight in Medieval unslipped red or
groove underneath. The sides of the black ware
dish curve inwards slightly then
appear to continue into a wide flat
base
689
Appendix 6: Ceramic Typology
42 Shallow dish Wide mouthed shallow dish with Brahmapuri: T.19, dish
quite vertical sides and a carinated
shoulder from which the bottom of
the bowl turns sharply inwards. The
sides of the bowl have grooves
above the carination and below the
rounded rim. The rim is small and
slightly out-turned.
43 Shallow dish Large dish with a long extended rim Baroda: T.117, shallow Baroda, P.VI V
with rounded terminus. The upper bowl in burnished red or
portion of the rim is usually black ware E47
concave but may be slightly convex
after the groove before the rim
terminus. There is an interior ridge
just before the main body of the
dish curves down. The shoulder of
the dish may have a slight ridge
690
Appendix 6: Ceramic Typology
44 Dish Large dish, slightly deeper than Sanchi: Decorated examples Sanchi, c.3rd-2nd II/III
T.43, with a long flaring rim. The similar to T.12, in B&R or centuries BC;
rim is curved and slightly grey slipped wares; Pauni, Pauni (1998)
overhanging, with a rounded T.1 dish in red polished P.IV; Nevasa,
terminus, and has a small interior ware; Nevasa, T.86b, P.V; Vaisali P.IV
groove before the main bowl of the shallow plate in red washed
dish. The exterior of the body has a ware; Vaisali, T.8, red ware
small carinated shoulder, from plate
which the body curves inwards.
Some examples have decoration on
the upper face of the rim
45 Shallow dish Shallow carinated bowl or dish with Maheshwar & Navdatoli: Maheshwar P.VI; II/III
wide mouth. The vessel has a Similar to T.150 in Pauni (1968), P.
sloping rim with a pointed terminus Micaceous ware; Pauni, II; Shamalaji
and often grooves on the upper T.54, red slipped dish; from P.IIA
face. Immediately below the rim is Shamalaji, T.136, dish in
a carinated shoulder from which the micaceous ware
body of the vessel curves inwards.
The carination varies between being
a small ridge and a more defined
point. The bowl appears to have a
fairly shallow, rounded base. Some
examples have incised notches on
the shoulder and others simply have
incised horizontal bands
691
Appendix 6: Ceramic Typology
692
Appendix 6: Ceramic Typology
48 Carinated Pot A medium to large carinated pot Nevasa, similar to T.48, in Nevasa P.IV; II
with a closed mouth. The body red slipped ware; Pauni, Pauni (1968) P.
below the rim slopes to a carinated similar to T.68 in upper II; Timbarva P.I
shoulder at the widest point of the profile, mica slipped ware;
vessel. The upper portion of the Timbarva, similar overall to
body above the carination is longer T.57 in red burnished ware
than in the carinated handis. The
bottom portion of the body curves
inwards into a seemingly globular
body. The relatively small rim
flares outwards, with a pointed
terminus.
49 Carinated handi Carinated handi rim form found on Maheshwar & Navdatoli, Maheshwar & III, IV Exclusively
both medium and small examples. similar to T.152a, carinated Navdatoli, P.VI; occurs in
The small rim forms a simple out- vessel in ware with quartz Sanjan occupied Micaceous
turned curve, with a flattish top and grains; Sanjan, T.77, in mica from 7th century wares
rounded terminus. The shoulder of red ware AD
the vessel is carinated and often
defined by a groove. The body of
the vessel curves inwards from the
carination to a medium depth to
shallow globular body. This is a
fairly common rim type, however
while all examples have the same
rim profile and angle to the body,
not all have the carination.
693
Appendix 6: Ceramic Typology
50 Carinated Pot Medium-sized carinated pot Maheshwar & Navdatoli: Maheshwar & II/III Exclusively
possibly for cooking, with a wide, T.119a, carinated vessel in Navdatoli, P.VI; occurs in
slightly closed mouth. The small black ware; Nevasa: similar Nevasa, P.V Micaceous
rim extends out at 90 degrees to the to T.69a in tan-slipped ware wares
neck, with a flattened top and a
small external rounded terminus.
The rim is often undercut and
drooping. The neck has small raised
horizontal bands on the external
surface before a ridged carination
with an upward pointed external
extrusion. The body then slopes
into a wide rounded body.
51 Globular pot Globular pot with a ridged
carination at the shoulder and a
concave neck. The body from the
small upturned ridge is rounded
rather than sloping directly inwards.
The neck of the vessel has slopes
up to a small out0turned rim. The
neck may be decorated with a series
of grooves. The rim has a rounded
terminus with a grooved top and is
undercut on the exterior
52 Globular pot Globular pot with a concave neck
and curved shoulder portion
featuring a small pointed ridge.
From the neck the body appears to
be wide and round but this portion
is missing on all examples. The rim
is uncut and droops over the neck,
and features a small groove on the
interior
694
Appendix 6: Ceramic Typology
695
Appendix 6: Ceramic Typology
696
Appendix 6: Ceramic Typology
57 Globular pot Globular pot with a constricted Pauni: T.19, vase in mica Pauni (1998), II/III
neck and long flaring rim. The rim slipped red ware; P.IV; Maheshwar
is featureless with a simple rounded Maheshwar & Navdatoli: & Navdatoli
or pointed terminus and usually similar to T.151, in ware P.VI;
extends directly from the narrowest with quartz grains; Kaundinyapura
point of the neck in an upwards and Kaundinyapura T.8b, wide (1968),
outwards flaring manner. In such mouthed cooking vessel in Sātavāhana
cases, The rim meets the body at an micaceous ware
angle of just less than 90 degrees.
In some examples the rim appears
part of a concave neck flaring out
from the body giving a more
rounded appearance. The body
portion appears to extend into a
round globular pot
58 Globular pot Simple small pot with a short Sanchi: T.91a, cooking Sanchi: 2nd-4th II/III
concave neck and small splayed out vessel in coarse grey century AD;
rim. The rim terminus is rounded or ware/burnished red ware; Devnimori c.
slightly pointed, and extends out at Devnimori: T.23 & T.24, 2nd-4th century
a roughly 90 angle from the body. pear shaped pots in plain red to 7th/8th century
The body extends down and out and ware; Pauni: Similar to T.3, AD; Pauni
appears to become rounded or oval globular pot in red ware; (1998), P.III;
in shape. This may be a small to Pauni (1972), T. 108, Pauni (1972)
medium serving or storage vessel. micaceous red ware; Paunar P.III; Paunar P.II
T.24D coarse red jar
59 Pot Pot with concave, restricted neck Dwarka: T.13, in slipped red Dwarka, P.I; II
and a medium-long flaring rim ware; Hastinapura, similar to Hastinapura,
which meets the body at around a No.XXXVIII; Pauni, T.66, P.IV; Pauni
90 degree angle. The rim may in micaceous red ware; (1978), Phase II
feature a slight bulge near to the Shamalaji, T.100,
neck and has a rounded terminus. plain/burnished red ware
The body of the vessel appears to handi
be globular or oval in shape
697
Appendix 6: Ceramic Typology
60 Jar/Globular jar Large storage jar rim type: The rim Devnimori: Similar to T.60, Devnimori c. III Primarily
turns out from the body at just over pot with flared rim in 2nd-4th century Gritty
a 90° angle, and flares straight micaceous red ware; Paunar to 7th/8th century Micaceous
outwards to a length of 35 to T.61 course micaceous jar AD; Paunar, P.II Wares
70mm. Some examples display a
slight downward curve to the rim,
and all have a simple rounded or
pointed terminus. The body does
not remain on most examples but
appears to continue into a globular
or oval jar form
61 Globular jar Jar with very rounded body 3978
(majority of vessel body is
missing). The jar has a short
straight, vertical neck with a small
rounded rim. The rim is slightly
out-turned
62 Jar Storage jar rim form with a
medium-length concave neck and
small out-turned rim. The neck
features a small ridge which may be
decorated with an incised pattern.
The rim has a flattened top and
often a slight undercut projection
inwards. At the narrowest part of
the neck is an interior groove
698
Appendix 6: Ceramic Typology
64 Jar Large storage jar rim type: The Nevasa, similar to T.47b, in Nevasa, P.IV; II
vessel has a concave neck which red slipped ware; Kaundinyapura
smoothly transitions into a long up Kaundinyapura T. 13 high (1968) Mauryan
and outwards flaring rim. The rim necked red ware ghada (300-100BC)
has a slightly thicker terminus shaped vessel
which is more rounded on the
interior and make by defined with a
small groove. The body of the
vessel is missing but appears to be
rounded at the top. This may later
extend into an oval shaped jar
65 Jar/Vase Jar with long, flaring rim, which Kaundinyapura: T.53-8 in Kaundinyapura III
may be designed for pouring. The micaceous fabric (2001), Early
rim extends out from the neck and Historic
is usually quite straight. The end of
the rim is slightly bulbous with a
rounded terminus. The body portion
is missing but from the profile
where it meets the rim, it would
appear to extend into an oval jar
form. This rim type usually occurs
in a heavy, gritty ceramic fabric.
699
Appendix 6: Ceramic Typology
67 Jar A jar with a long flaring neck and Pauni, T.75, in red slipped Phase II (3rd to II
relatively wide mouth. The neck of ware 1st century BC)
the vessel is more out-flared and
curved than in T.65, with a more
pronounced rounded rim terminus.
The end of the rim may be slightly
undercut. Though most of the neck
and body is usually missing, the
neck of the vessel appears to be
constricted.
68 Jar Jar rim form with a constricted neck Timbarva, similar to T.74, Timbarva, P.II II
and long flaring neck/rim. The rim ware with micaceous slip
has a rounded terminus and is
undercut on the exterior, with a
small groove on the interior top.
The mouth of the vessel is quite
wide in comparison to the neck.
While the body portion is missing it
would appear to be a rounded or
oval jar form
69 Jar Jar with relatively long concave Timbarva, similar to T.51 & Timbarva, P.I II
neck portion leading up to an out- 52, in red burnished ware
turned rim. The rim is quite deeply
undercut and droops over on the
exterior with a rounded appearance.
The body of the vessel is missing
but it appears the neck is narrow
and constricted, indicating a jar for
storage or pouring
700
Appendix 6: Ceramic Typology
71 Jar/Vase Jar/Vase with medium length Brahmapuri, T.79, Pot with Brahmapuri, Late II/III
concave, flaring neck and small flaring rim; Dwarka, T.44, in Sātavāhana;
out-turned rim. The rounded rim is slipped red ware; Dwarka, P.II;
undercut and droops slightly. There Maheshwar & Navdatoli, Maheshwar &
is a groove around the middle of the T.132, in red slipped ware; Navdatoli, P.VI;
rim exterior. The neck of the vessel Paunar T.92I red slipped Paunar, P.III
is very narrow and constricted, high neck jar
indicating a jar/vase for pouring or
serving. Based on published
parallels the body of the vessel may
be quite wide
72 Jar/Vase Very similar jar/vase form to T.71. Dwarka, T.48, in burnished Dwarka, P.II; II
The vessel has a long concave red ware; Maheshwar & Maheshwar &
flaring neck and a small out-turned Navdatoli, T.86f, high- Navdatoli P.IV
rim. The rim has a pointed terminus necked vessel in burnished
and flattened exterior face, and it is red ware
undercut to droop slightly. There is
a groove towards the top of the rim.
The neck is narrow and constricted
and the vessel body turns out at the
base of the neck. The body appears
to continue into a globular or oval
jar/vase
701
Appendix 6: Ceramic Typology
73 Jar/Vase Another similar jar/vase with a high Devnimori, T.38, water Devnimori c. II/III
concave neck and small flaring out- pot/pitcher in plain red ware; 2nd-4th century
turned rim. The rim has a groove Pauni, similar to T.28, vase to 7th/8th century
before its terminus which creates a in dull red slipped red ware; AD; Pauni
squared off end. The rim is pointed Hastinapura, No.XXXIV ; (1998), P.IV;
at the top and slightly undercut. The Bhokardhan, similar to Hastinapura P.IV;
neck of the vessel is very narrow T.7A, coarse red ware pot Bhokardhan, P1A
and constricted, ideal for pouring,
and it appears the body is globular
or oval in shape
74 Jar/Vase Jar or vase with a long flaring neck Timbarva, similar to T.73, Timbarva, P.II; II
and small rim with more prominent red ware pot; Kaundinyapura
groove on top creating a beaked Kaundinyapura, Similar to (1968) P.V Late
appearance. The high neck is T35b Sātavāhana
concave and may extend fairly
vertically or be flared outwards.
The groove on the top of the rim
may be for receiving a covering,
and creates a flat exterior surface to
the rim. The vessel has a narrow,
restricted neck.
75 Jar/vase Jar with a long flaring neck and Sanchi: T.148, jar with Sanchi, 2nd II/III
small out-turned rim: the shoulder clubbed rim in red slipped century BC to
and body portion is always missing ware; Kaundinyapur: Similar 5th-6th century
in this type. The remaining neck to T.47-3, in red ware; AD;
appears to be quite straight and Baroda: Similar to T.31; Kaundinyapur
would indicate a narrow, restricted Timbarva, T.50, in red (2001), Early
opening. The rim is undercut and burnished ware; Paunar Historic
tapers. It has a triangular section T.92K red slipped jar with (Sātavāhana);
and squared off terminus. high concave neck; Baroda, P.II;
Bhokardhan T.17A globular Timbarva, P.I;
pot with red wash Paunar, P.III;
Bhokardhan, P1A
702
Appendix 6: Ceramic Typology
76 Jar/Vase Jar/Vase with a high concave neck, Pauni: T.10, vase in Pauni (1998), II/III
constricted at the base of the neck. medium, dull red ware; P.IV; Baroda,
The rim flares outwards and droops Baroda, T.26; Timbarva, P.II; Timbarva
over a small undercut. The rim T.28, high necked pot in P.I; Vaisali, P.IV
terminus may be rounded or plain red ware; Vaisali, T.13
slightly squared off. The body of red ware
the vessel appears to continue into
an oval jar form.
77 Jar/Vase Jar/vase with a long straight neck Bhokardhan: T.81, pot with Bhokardhan, II
which flares outwards in a funnel high concave neck in red P.1B
shape from the body. The rim is slipped ware
undercut and droops over. The rim
also has a groove on the outside
giving it a split appearance. The
neck of the vessel is restricted, but
the body portion is missing
703
Appendix 6: Ceramic Typology
83 Jar Jar with long, concave neck and Dwarka, similar to T. 46, Dwarka, P.II; II 5653
out-turned rim. The rim has an with more pronounced Pauni (1968), P.II
interior groove and a groove on the groove on top, burnished red
outer face ware; Pauni, T.73a, dull red
ware
84 Jar/Vase Jar/Vase with long concave neck 4075
and a rounded out-turned rim. The
rim has a groove on the exterior
704
Appendix 6: Ceramic Typology
88 Jar Jar with out-turned flaring rim and Pauni, Similar to T.30, vase Pauni (1998), II 706
a relatively straight neck. The rim is in mica slipped dull red ware P.IV
squared off and has a groove on the
interior
705
Appendix 6: Ceramic Typology
89 Jar Jar with long straight neck. The rim Baroda, similar to T.31 Baroda, P.II II/III 3009
is undercut and droops over to a
slightly pointed terminus
706
Appendix 6: Ceramic Typology
94 Jar Jar with globular body and Nasik, T.7b, straight necked Nasik, P.IIA I
restricted opening. The vessel has a vessel in red ware
short concave neck with a long,
vertical rim. The rim has a small
rounded terminus and a slightly
ridged exterior
95 Jar Small jar with a narrow neck which 1622
flares outwards slightly. The rim is
turned inwards and is undercut on
the interior with a flattened top.
707
Appendix 6: Ceramic Typology
99 Lidded Jar Necked globular vessel/jar with a Maheshwar & Navdatoli: Maheshwar & II/III All Red
constricted opening. The rim may T.132a, vessel with ringed Navdatoli, P.VI; slipped red
be flattened or tapering, and partly rim in red slipped ware; Dwarka P.II; or
protrudes inwards. The neck also Dwarka, similar to T.51, Sanchi, c. 5th-6th micaceous
bears a small flange either with an rims with a projecting band to 11th-12th c wares
oblique square undercut ring or a in Red ware; Sanchi, T.105c, AD
triangular profile, from which the Lidded vase/jar in Red
neck slopes inwards. The fairly slipped/burnished ware
straight angle of the neck suggests
these are not small lids (as in Pauni,
T.26 (Fig. 15), P.IV) but the top of
a vessel designed to hold a lid or
covering
100 Pot Small globular pot rim type. The Pauni: T.7, vase in coarse Pauni (1998), I/II/III Wide range
rim curves outward from the body red slipped red ware; P.II; Baroda P.II; of wares
at around 90 degrees and generally Baroda, T.7, pear shaped jar; Devnimori c. across Red,
thickens slightly into a simple Devnimori, T.25/26; 2nd-4th century Micaceous
rounded rim terminus. Some Shamalaji, T.79, to 7th/8th century and a
examples have a slightly triangular plain/burnished red ware AD couple of
terminus. The body slopes down Buff/Grey
from the rim into a rounded
globular shape
101 Pot Pot with a short neck and flaring
rim. The neck slopes inwards from
a seemingly globular body. The rim
is undercut and has a simple
rounded terminus. The top of the
rim is curved and incised with thin
grooves.
708
Appendix 6: Ceramic Typology
709
Appendix 6: Ceramic Typology
710
Appendix 6: Ceramic Typology
111 Pot Globular pot with sides that slope Dwarka, similar to T.106 in Dwarka, P.IV IV
inwards with a straight line. The slipped red ware
neck turns outwards at an angle of
around 100 degrees below the
rounded bulbous rim. On the
interior of the rim is a groove at the
narrowest part of the vessel
opening. The pots sometimes
feature incised horizontal lines on
the exterior as decoration
112 Pot Medium-sized globular pot with a
vertical neck leading to a small
bulbous rim. Just below the rim the
neck curves out slightly. The rim
itself faces inwards and has a
groove on the interior
113 Pot Similar to type 112: A globular pot
with a short slightly concave neck
ending in an oval-shaped bulbous
rim. The rim extends outwards and
inwards, being undercut on both
sides. The interior lip slightly
narrows the opening of the vessel
114 Jar Rim of a large jar? The neck has 3545
straight sides and turns slightly
outwards. The rim droops over the
neck on the exterior and has a
pointed terminus
711
Appendix 6: Ceramic Typology
712
Appendix 6: Ceramic Typology
713
Appendix 6: Ceramic Typology
714
Appendix 6: Ceramic Typology
715
Appendix 6: Ceramic Typology
133 Pot Wide-mouthed globular pot with no Sanjan, similar to T.147, in Sanjan, late 7th to IV
neck. The sides of the vessel slope black slipped grey ware early/mid-9th
inwards and the rim flares out century AD
directly from the body. The angle
forms is roughly 90 degrees. The
interior side of the rim is flattened
and the rim has a rounded terminus
134 Pot Similar to T.134: Globular pot with
no neck. The rim meets the body at
around 90 degrees and is more
curved and less angular than in
T.133
135 Pot Small-sized globular pot similar to 5244
T.133 where the rim meets the body
at around 90 degrees and the rim
has a flattened edge on the interior.
Below the rim is a small ridge
136 Pot Similar to T.134: Globular pot with Pauni, T.18, vase in mica Pauni (1998), II
a very short concave neck and a rim slipped red ware; Pauni P.IV; Pauni
that turns outwards and is undercut. (1972), T.106, dull red (1972) P.III;
The rim has an oval section and storage jar; Bhokardhan, Bhokardhan, P1A
slightly pointed terminus similar to T.8, coarse red
ware pot
716
Appendix 6: Ceramic Typology
138 Jar Appears to be the rim and neck of a Pauni, T.99, dull red ware Pauni (1968), II 1510
large, wide-mouthed jar/basin. The basin P.III
sides of the vessel are thin and
extend inwards from the rim. The
rim curves outwards and is slightly
undercut with a round terminus.
The rim is also externally grooved
and has a flat top leading to the
interior of the vessel. Below the rim
of the interior is another groove.
139 Pot Similar to T132. A thin walled
globular pot with no neck. The rim
on this type extends out from the
body at roughly 75 degrees and is
therefore angled upwards. The rim
has a flattened edge on the interior
and a rounded terminus On the top
of the rim is a small groove. Some
varieties feature incised lines on the
topside of the rim
140 Pot/Jar Rim of a wide-mouthed jar or 2706
globular pot with very thin walls.
Only the rim piece remains. The
rim is thin, flaring outwards and
externally undercut. The exterior of
the rim has a groove to create a split
lip. On the interior there is another
groove just about the narrowest part
of the opening.
717
Appendix 6: Ceramic Typology
143 Jar Globular jar similar to T45. This Pauni, T.25 mica slipped red
vessel type has a short restricted ware
neck, fairly wide mouth and slightly
undercut rim. The rim flares out
horizontally and has two grooves
on the exterior giving it a triple
lipped appearance. The middle
section of the rim is less
pronounced than the upper and
lower section, but this panel bears
no decoration as in the other type.
718
Appendix 6: Ceramic Typology
719
Appendix 6: Ceramic Typology
148 Jar Medium sized closed storage jar Baroda, T.43. jar in grey Baroda, P.III IV Primarily
rim type. Rim has external ware; Shamalaji, T.22, Grey wares
split/double lip on a horizontal plain/burnished black ware and Red
external overhang. Top is flattish, burnished
occasionally more rounded and red wares
most examples have a small
internal groove just down from the
rim. Some examples have
horizontal grooves below the short
neck on the shoulder exterior. Body
appears to be rounded and globular
149 Jar Medium sized closed storage jar Maheshwar & Navdatoli, Maheshwar & II/III
rim type. The globular jar has a similar to T.120, globular Navdatoli, P.VI;
medium-length sloping neck which vessel in burnished black; Pauni (1968),
leads up to the rim. The rim has an Pauni, T.113, in grey ware P.III
external split/double lip on a
horizontal external overhang. The
top of the rim is flattened and there
is a small groove on the interior.
Similar to T7 but with a less
distinct neck and more pronounced
overhanging rim
150 Jar Medium sized closed storage jar Baroda, similar to T.40, jar Baroda P.III & IV/V
rim type. Rim has external in grey ware but also similar Period IV/V;
split/double lip and is undercut. to T.70 in red slipped ware; Sanjan, from 7th
Some examples have horizontal Sanjan, T.197, in black century AD;
grooves below the short neck on the slipped grey ware; Paunar Paunar P.IV
shoulder exterior. Body appears to T.142A drab black pot
be rounded and globular.
720
Appendix 6: Ceramic Typology
152 Jar Medium-sized wide-mouthed jar Dwarka, T.162, in burnished Dwarka, P.IV IV/V
with short neck and small triangular black ware
rim. The rim is positioned with a
point at the top and straight edge
either vertical or at a 45 degree
angle. Short neck leads into a
globular pot/jar form
721
Appendix 6: Ceramic Typology
722
Appendix 6: Ceramic Typology
723
Appendix 6: Ceramic Typology
724
Appendix 6: Ceramic Typology
725
Appendix 6: Ceramic Typology
170 Globular pot Globular pot with a short neck Kaundinyapura, similar to Kaundinyapur II
curving upwards to a thick rim. The T.37, large storage jar in (1968), P.V Late
rim is inverted and square in coarse red ware Sātavāhana
profile. On the interior the rim is
undercut due to a small groove.
171 Globular pot Small globular pot with no neck 4665
and a small rounded rim. The rim
faces outwards and is defined by a
slight undercut
172 Pot Large close-mouthed jar or pot with Bhokardhan, similar to Bhokardhan, II
bulbous body. Small rim with thick T.16C, coarse red storage jar P.1A
inner bulb or projection. Rim is
either rounded or slightly squared
off
173 Globular pot Small-sized globular pot with no Pauni, T.74, mica slipped Pauni (1968), P.II II
neck and an inverted rim. The ware
rounded body leads straight into the
rim which features a small groove
on the exterior. On the interior the
rim has a rounded terminus and
restricts the opening of the vessel
174 Globular pot/jar Appears to be a globular pot or jar Sanjan, similar to T.90, in Sanjan, c. III/IV
with no neck. The rim is inverted Black slipped grey ware; early/mid 9th to
and features a slightly pointed end Paunar similar to T.115A, mid/late 12th
on the exterior. The interior red ware neckless storage century AD;
terminus of the rim is rounded and jar+E180 Paunar P.III
makes the opening the narrowest
part of the vessel. The upper
portion of the body may have
incised decorations
726
Appendix 6: Ceramic Typology
179 Heavy Pot Large storage vessel with relatively Pauni, T.104, dull red Pauni (1968), II
narrow closed rim and thick fabric. storage jar P.III
The rim extends upwards from the
short neck and forms a rounded
terminus. The body of the vessel
extends down and out from the
neck in a curve giving it a globular
appearance
727
Appendix 6: Ceramic Typology
728
Appendix 6: Ceramic Typology
186 Basin Deep steep-sided bowl with in- Dwarka, similar to T.78, in Dwarka, P.III; IV
turned rim. The rim has a rounded slipped red ware; Paunar Paunar P.IV
terminus either turned up or down T.160 drab black neckless
with ridges on the top. The external jar
base of the rim has a more
pronounced ridge. The sides of the
vessel either extends down and then
curves in or is convex all round into
a globular body. The vessel is thin-
sided and some examples have an
internal ridge halfway down the
body.
729
Appendix 6: Ceramic Typology
187 Pot Globular pot with a short, concave Pauni, T.30, mica slipped Pauni (1968), P.I I/II
neck and a large flanged rim. The red ware & T.112, in & P.III
body of the vessel features incised micaceous red ware
decoration of incised lines and
notches. The rounded rim faces the
interior of the vessel and is slightly
undercut. On the exterior is an
upwards facing flange with a
groove and further incised
decoration on the top
188 Jar Medium-sized globular pot or jar Paunar, similar to T.142 Paunar, P.III III 4392
with a short concave neck and an neckless storage jar
outwards flaring rim with a rounded
terminus. The rim is undercut and
features a pinched applique fringe.
The top of the rim features several
grooves with incised decoration.
189 Jar Thick sided jar with constricted 2790
concave neck. The body portion is
missing but it would appear to be
globular or vase shaped. The rim is
deeply undercut and overhangs.
The rim is ridged and has an
applique fringed decoration. The
interior of the rim has a small
groove
190 Jar Thick sided jar with large bulbous Paunar, T.114C, storage jar Paunar P.III 4668
rim. Most of the neck is missing but with beaded rim in red
begins to slope inwards from the slipped ware
rim with straight sides indicating a
restricted neck. The rim is rounded
and undercut with an applique
fringe. The interior of the rim is
rounded and slightly undercut too.
730
Appendix 6: Ceramic Typology
191 Storage jar Large storage jar rim type with a 3157
short concave neck. The underside
of the neck/rim features several
ridges and grooves. The rim has a
triangular profile and is faces
outwards. The exterior of the rim
has a small groove, below which is
a ridged fringe
192 Heavy Pot Very large storage vessel with thick Primarily
fabric and closed rim. Rim curves in Storage
inwards to form a rounded internal Jar wares
lip. The rim also has an external
extruded rounded lip and a flat rim
top, giving a triangular section. The
body curves downwards,
presumably extending into a long
oval shape.
193 Basin Large vessel with wide mouth and Sanchi, similar to T.72, in Sanchi, 1st-4th II/III
outwards splayed collared rim. The grey ware; Shamalaji, century AD;
body extends almost vertically from similar to T.126, handi in Shamalaji from
the neck, with slight rounding to micaceous ware; Vaisali, P.IIA; Vaisali
produce a deep wide basin. On T.1, red ware basin; Paunar, P.III; Paunar,
some examples there is a small T.73A coarse micaceous P.IIB
raised rounded band below the globular pot
neck. The rim displays a small
rounded internal lip and a larger
external lip with a rounded
terminus. The rim top is decorated
with small incised horizontal lines.
731
Appendix 6: Ceramic Typology
194 Basin Deep, steep sided basin or bowl Maheshwar & Navdatoli, Maheshwar & III
(with flattened base on one T.141, Kunda in coarse red Navdatoli, P.VI
example). Thick sided vessel made ware
from coarse material. No definition
leading to the rim, just thinning of
sides and rounded terminus
195 Basin Vessel with straight sides and flat Nevasa,T.106, Kunda in Nevasa, P.V; II 4946
top, with slightly rounded exterior coarse black ware; Sanchi, Sanchi 1st-4th
to the rim. The top of the rim is similar to T.50, heavy century AD
decorated with an incised beaded lidded basin in red slipped or
pattern. The exterior of the vessel micaceous ware
has small ridges
732
Appendix 6: Ceramic Typology
733
Appendix 6: Ceramic Typology
734
Appendix 6: Ceramic Typology
735
Appendix 7: Ceramic Illustrations
736
Appendix 7: Ceramic Illustrations
737
Appendix 7: Ceramic Illustrations
738
Appendix 7: Ceramic Illustrations
739
Appendix 7: Ceramic Illustrations
740
Appendix 7: Ceramic Illustrations
741
Appendix 7: Ceramic Illustrations
742
Appendix 7: Ceramic Illustrations
743
Appendix 7: Ceramic Illustrations
744
Appendix 7: Ceramic Illustrations
745
Appendix 7: Ceramic Illustrations
746
Appendix 7: Ceramic Illustrations
747
Appendix 7: Ceramic Illustrations
748
Appendix 7: Ceramic Illustrations
749
Appendix 7: Ceramic Illustrations
750
Appendix 7: Ceramic Illustrations
751
Appendix 7: Ceramic Illustrations
752
Appendix 7: Ceramic Illustrations
753
Appendix 7: Ceramic Illustrations
754
Appendix 7: Ceramic Illustrations
755
Appendix 7: Ceramic Illustrations
756
Appendix 7: Ceramic Illustrations
757
Appendix 7: Ceramic Illustrations
758
Appendix 7: Ceramic Illustrations
759
Appendix 7: Ceramic Illustrations
760
Appendix 7: Ceramic Illustrations
761
Appendix 7: Ceramic Illustrations
762
Appendix 7: Ceramic Illustrations
763
Appendix 7: Ceramic Illustrations
764
Appendix 7: Ceramic Illustrations
765
Appendix 7: Ceramic Illustrations
766
Appendix 7: Ceramic Illustrations
767
Appendix 7: Ceramic Illustrations
768
Appendix 7: Ceramic Illustrations
769
Appendix 7: Ceramic Illustrations
770
Appendix 7: Ceramic Illustrations
771
Appendix 7: Ceramic Illustrations
772
Appendix 7: Ceramic Illustrations
773
Appendix 7: Ceramic Illustrations
774
Appendix 7: Ceramic Illustrations
775
Appendix 7: Ceramic Illustrations
776
Appendix 7: Ceramic Illustrations
777
Appendix 7: Ceramic Illustrations
778
Appendix 7: Ceramic Illustrations
779
Appendix 7: Ceramic Illustrations
780
Appendix 7: Ceramic Illustrations
781
Appendix 7: Ceramic Illustrations
782
Appendix 7: Ceramic Illustrations
783
Appendix 7: Ceramic Illustrations
784
Appendix 7: Ceramic Illustrations
785
Appendix 7: Ceramic Illustrations
786
Appendix 7: Ceramic Illustrations
787
Appendix 7: Ceramic Illustrations
788
Appendix 7: Ceramic Illustrations
789
Appendix 7: Ceramic Illustrations
790
Appendix 7: Ceramic Illustrations
791
Appendix 7: Ceramic Illustrations
792
Appendix 7: Ceramic Illustrations
793
Appendix 7: Ceramic Illustrations
794
Appendix 7: Ceramic Illustrations
795
Appendix 7: Ceramic Illustrations
796
Appendix 8: Ceramic Seriation
The following table presents the relative seriation of the ceramic assemblage. As noted in the main text, this table is based on sites with over 0.5%
of the total ceramic assemblage (29 sherds) and key wares which form over 1% of the total assemblage. The table has been conditionally
formatted based on percentage of ceramic ware groups and the sites have been manually rearranged to determine patterns of ware association.
Site BAR Gritty Grey/ Storage Red Mica1 Mica2 Dark Other Red Red Grey/ Unknown Sherd
No. Orange mica Mica BLS Burn Buff Count
353 16.4% 10.0% 10.0% 0.9% 27.3% 0.9% 3.6% 4.5% 0.0% 0.9% 5.5% 18.2% 1.8% 110
325 5.7% 47.5% 3.8% 1.3% 12.0% 1.3% 8.9% 0.6% 0.0% 0.0% 14.6% 3.2% 1.3% 158
262 2.8% 16.8% 8.4% 1.9% 32.7% 6.5% 16.8% 3.7% 0.0% 0.0% 2.8% 2.8% 4.7% 107
197 2.2% 22.2% 9.4% 5.6% 21.1% 5.0% 20.6% 5.0% 0.0% 0.0% 6.1% 0.6% 2.2% 180
196 0.0% 0.0% 12.1% 12.1% 48.5% 3.0% 6.1% 0.0% 0.0% 0.0% 6.1% 12.1% 0.0% 33
363 0.0% 2.2% 9.6% 8.1% 51.1% 5.2% 8.9% 0.7% 0.0% 0.0% 4.4% 2.2% 7.4% 135
316 0.0% 2.3% 18.6% 9.3% 34.9% 18.6% 7.0% 0.0% 0.0% 0.0% 0.0% 4.7% 4.7% 43
76 2.7% 2.7% 0.0% 0.0% 20.3% 17.6% 36.5% 4.1% 0.0% 0.0% 6.8% 5.4% 4.1% 74
72 0.0% 7.8% 7.0% 11.7% 31.3% 23.4% 7.8% 2.3% 0.0% 0.0% 6.3% 0.8% 1.6% 128
265 0.0% 2.4% 0.0% 9.8% 58.5% 25.6% 1.2% 1.2% 0.0% 0.0% 0.0% 0.0% 1.2% 82
59 0.0% 0.0% 2.8% 2.8% 47.2% 30.6% 0.0% 13.9% 0.0% 0.0% 2.8% 0.0% 0.0% 36
241 0.0% 6.2% 1.2% 8.6% 34.6% 29.6% 9.9% 2.5% 0.0% 0.0% 1.2% 4.9% 1.2% 81
7 0.0% 4.5% 0.9% 18.9% 28.8% 21.6% 9.0% 0.0% 0.9% 0.0% 12.6% 1.8% 0.9% 111
79 0.0% 1.3% 1.8% 3.1% 26.7% 53.8% 7.6% 1.3% 0.9% 0.4% 0.0% 0.9% 2.2% 225
213 1.1% 16.9% 9.0% 3.4% 11.2% 22.5% 19.1% 6.7% 2.2% 0.0% 3.4% 1.1% 3.4% 89
71 0.0% 11.3% 1.6% 14.5% 11.3% 51.6% 3.2% 0.0% 0.0% 0.0% 4.8% 0.0% 1.6% 62
797
Appendix 8: Ceramic Seriation
71 0.0% 11.3% 1.6% 14.5% 11.3% 51.6% 3.2% 0.0% 0.0% 0.0% 4.8% 0.0% 1.6% 62
400 0.0% 5.7% 3.4% 3.4% 18.4% 11.5% 24.1% 8.0% 0.0% 1.1% 1.1% 19.5% 3.4% 87
29 0.0% 0.0% 0.0% 0.4% 27.4% 3.9% 6.5% 1.7% 0.0% 3.5% 48.3% 8.3% 0.0% 230
21 0.0% 0.0% 0.0% 0.0% 22.2% 5.6% 22.2% 0.0% 0.0% 0.0% 33.3% 16.7% 0.0% 36
92 0.0% 0.0% 0.0% 0.0% 14.3% 25.7% 5.7% 5.7% 0.0% 2.9% 28.6% 14.3% 2.9% 35
48 0.0% 0.0% 0.0% 0.0% 10.3% 64.1% 10.3% 2.6% 2.6% 0.0% 7.7% 0.0% 2.6% 39
20 0.0% 0.0% 2.3% 0.0% 15.9% 40.9% 29.5% 2.3% 0.0% 0.0% 9.1% 0.0% 0.0% 44
94 0.0% 0.0% 0.0% 0.0% 14.9% 55.3% 19.1% 2.1% 0.0% 0.0% 4.3% 4.3% 0.0% 47
48 0.0% 0.0% 0.0% 0.0% 10.3% 64.1% 10.3% 2.6% 2.6% 0.0% 7.7% 0.0% 2.6% 39
364 0.0% 0.6% 1.3% 0.0% 7.1% 23.1% 51.3% 9.6% 0.0% 0.0% 3.2% 2.6% 1.3% 156
49 0.0% 1.2% 0.6% 0.6% 3.1% 16.7% 15.4% 1.9% 59.3% 0.0% 1.2% 0.0% 0.0% 162
351 0.0% 0.0% 1.8% 0.0% 0.0% 14.3% 73.2% 7.1% 0.0% 0.0% 0.0% 3.6% 0.0% 56
46 1.2% 0.0% 0.0% 0.0% 2.4% 4.9% 70.7% 13.4% 0.0% 0.0% 2.4% 3.7% 1.2% 82
318 0.0% 0.7% 2.0% 0.0% 2.7% 1.4% 66.7% 18.4% 0.0% 0.0% 5.4% 2.7% 0.0% 147
236 0.0% 4.7% 0.6% 0.0% 2.3% 4.7% 61.4% 19.3% 0.0% 0.0% 4.7% 1.8% 0.6% 171
289 0.0% 9.0% 3.2% 1.3% 6.4% 9.0% 55.8% 8.3% 0.0% 0.0% 3.8% 0.0% 3.2% 156
47 0.0% 1.5% 0.0% 0.0% 0.0% 7.6% 48.5% 24.2% 0.0% 0.0% 9.1% 9.1% 0.0% 66
75 0.0% 10.3% 0.0% 0.0% 10.3% 10.3% 27.6% 17.2% 0.0% 0.0% 3.4% 20.7% 0.0% 29
210 0.0% 2.5% 2.5% 1.3% 1.9% 3.8% 43.4% 9.4% 0.0% 1.3% 20.8% 12.6% 0.6% 159
45 0.4% 0.4% 0.7% 0.0% 3.8% 2.9% 22.5% 8.0% 0.0% 1.1% 10.7% 47.9% 1.6% 449
276 0.0% 0.0% 1.1% 0.0% 4.5% 3.4% 16.9% 4.5% 0.0% 2.2% 22.5% 43.8% 1.1% 89
202 0.0% 2.7% 2.7% 2.7% 2.7% 0.0% 29.7% 8.1% 0.0% 0.0% 21.6% 27.0% 2.7% 37
155 0.7% 0.0% 0.0% 0.0% 2.0% 2.7% 11.5% 5.4% 0.0% 2.0% 17.6% 58.1% 0.0% 148
312 0.0% 0.0% 5.3% 0.0% 2.6% 7.9% 23.7% 7.9% 0.0% 0.0% 7.9% 44.7% 0.0% 38
352 0.0% 1.0% 3.1% 0.0% 2.1% 2.1% 12.5% 2.1% 0.0% 0.0% 8.3% 68.8% 0.0% 96
107 0.0% 0.0% 0.0% 1.3% 3.8% 2.5% 10.0% 3.8% 0.0% 1.3% 21.3% 56.3% 0.0% 80
798
Appendix 8: Ceramic Seriation
122 0.0% 0.0% 0.0% 0.0% 5.5% 1.1% 2.2% 0.0% 0.0% 4.4% 70.3% 13.2% 3.3% 91
106 0.0% 0.0% 0.0% 0.0% 3.8% 0.0% 0.0% 1.9% 0.0% 1.9% 17.3% 73.1% 1.9% 52
116 0.0% 0.0% 0.0% 0.0% 2.5% 2.5% 5.0% 0.0% 0.0% 0.0% 35.0% 55.0% 0.0% 40
348 0.0% 0.0% 1.0% 0.0% 2.5% 0.5% 5.4% 2.5% 0.0% 1.5% 17.3% 68.8% 0.5% 202
386 0.0% 0.0% 1.0% 0.0% 5.2% 0.0% 4.1% 0.0% 0.0% 0.0% 13.4% 76.3% 0.0% 97
356 0.0% 4.7% 0.0% 0.0% 7.0% 0.0% 2.3% 0.0% 0.0% 0.0% 9.3% 74.4% 2.3% 43
275 0.0% 0.0% 0.0% 0.0% 6.5% 0.0% 0.0% 0.0% 0.0% 0.0% 6.5% 82.3% 4.8% 62
349 0.0% 0.0% 0.0% 1.5% 7.4% 0.0% 0.0% 0.0% 0.0% 2.9% 2.9% 83.8% 1.5% 68
350 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 2.7% 0.0% 1.4% 5.4% 90.5% 0.0% 74
Grand
Total 41 253 134 136 761 623 1111 289 103 41 654 1247 82 5475
799
Appendix 9: List of Sculpture
This appendix contains general information regarding the sculptural pieces identified
and recorded during survey. Each sculpture was assigned a number when discovered.
A number of sculptures are less useful for a chronological study of the landscape as
they belong to an unknown phase of activity or are broadly identified as ‘Post-
Vākāṭaka. Many are highly fragmentary or eroded, and features are often obscured by
the layers of sindūr applied in worship.
800
Appendix 9: List of Sculpture
6 61 31 Ganesh Very eroded sandstone Ganesh sculpture incorporated into modern III
shrine at Ambālā
7 63 33 Nandi Nandi Sculpture in modern shrine at Ambālā: Left leg bent up, bell IV+
around neck, back left leg tucked under body
8 70 40 Nandi Nandi sculpture pulled from Ambālā Tank. The left front leg is bent 45x24x65 IV+
up; both back legs are visible on right side. Decorated with bells
and garlands
801
Appendix 9: List of Sculpture
13 63 33 Naga Large Naga in modern shrine at Ambālā (with Nandi): Heavily IV+
covered in Sindūr but appears to have one or two heads
15 71 30 Nandi Seemingly late period Nandi at Ambālā: Left leg raised and both 80x34x65 IV+
back legs visible on right side. Adorned with bells and garlands.
Pronounced, slightly pointed hump and horns
16 103 50 Architectural Comparatively intricate sculpture at Koteshwar Temple, 37x20x12 IV+
Nagardhan: architectural fragment. Smoothed/polished sandstone.
At the bottom is a makara, with a kneeling male figure above,
holding onto what might be a lotus?
17 103 50 Gaṇa-type figure Red sandstone incomplete sculpture at Koteshwar Temple: possibly 35x29x11 III
Vākāṭaka - similar to one seen at Nagpur Museum. Appears to be a
Gaṇa-type figure (although may be Kubera): Two armed, rotund-
bellied, seated resting with left leg up. Bracelets visible, particularly
on right wrist. Necklace also visible (with amulet type design),
possibly anklets as well
18 95 310 Nandi Eroded sandstone Nandi on small platform with other sculpture 20x28x16 IV+
under worship in Bori, left front leg tucked underneath body, bridle
visible on face
20 96 48 Liṅga/Yoni Small sandstone Liṅga in modern shrine by tank next to Nagardhan 35x26 III
Fort: Round pīṭha. Back of liṅga is damaged/cut away (pīṭha:
100x84cm)
802
Appendix 9: List of Sculpture
27 103 50 Ganesh Very badly eroded sandstone Ganesh at Koteshwar Temple: Seated 64x43x15 Unknown
cross legged, appears 4 armed
28 103 50 Ganesh Badly eroded and incomplete Ganesh sculpture at Koteshwar 36x37.5x19 Unknown
Temple: 4 arms
29 103 50 ? Two figures: very eroded. Architectural piece? One is seated cross- 32x31x9 IV+
legged the other standing, standing figure may be female? Post-
Vākāṭaka door post?
30 103 50 Hero Stone Part of Hero Stone at Koteshwar Temple: Striding figure with 28x25x13 IV+
sword and shield
31 103 50 Hanuman Hanuman sculpture at Koteshwar Temple in heroic pose standing V+
on the demon, with tail curled overhead. Obscured by sindūr
803
Appendix 9: List of Sculpture
33 115 91 Ganesh Very eroded sandstone sculpture at Shrine on route to Nandpuri: 4 59x43x23 Unknown
armed Ganesh sat with left leg tucked under body
34 115 91 Mahishamardini Sandstone sculpture of female figure at shrine on route to 67x41x7 Unknown
Nandpuri: standing with right foot on an ass. Necklaces and
bracelets visible. Short arms, provincial style
35 115 91 ? Badly eroded sandstone sculpture of 4-armed male figure at shrine 85x39x16 IV+
on route to Nandpuri
36 115 91 Naga Bottom part of small sandstone Naga showing curled tail in figure 27x22x10 Unknown
of 8 similar to that seen in early depictions: Head missing
37 121 105 Naga Small sandstone Naga: as part of a sculpture pile at shrine in Unknown
Rāmṭek town with numerous fragmentary and eroded pieces of
sculpture. Single hood
39 127 110 Hanuman Large Hanuman in modern shrine, standing with hands in front of V+
the body
40 150 6 Liṅga Sandstone Liṅga in excavated temple West of Mansar Palace: 56 (+base III
Brahmasūtra and pārśvasūtra. Base is Octagonal 14)x35
41 157 177 Liṅga Liṅga on top of Hill east of Mansar-Moil by Modern shrine. Square (49 liṅga) Unknown
base, appears there is some evidence of a line similar to the 103x37
Brahmasūtra and pārśvasūtra but it is unclear whether this is an
intentional feature. End of shaft is slightly bulging
804
Appendix 9: List of Sculpture
42 160 179 Liṅga Isolated liṅga on its side by village at base of hill with similar 83.5 (51 liṅga) Unknown
profile to type.42. Square base, damaged and no visible 43
Brahmasūtra and pārśvasūtra.
43 164 10 Liṅga Sandstone liṅga in the star-shaped temple at Mansar: clear 60 (+base 12) III
Brahmasūtra and pārśvasūtra and square base x43
44 199 113 Hanuman Hanuman sculpture in small temple next to tank at Rāmṭek. V+
Standing Hanuman with Gada under left arm, hands held in front of
body and tail over head
45 207 119 Liṅga/Yoni Liṅga in Temple in main Rāmṭek town: small liṅga, round pīṭha V+
with deep run off lines for water libation
46 207 119 Naga-Kal Double Naga stone: Straight necks with single loops in tails, oval IV+
shaped hoods, very basic
47 207 119 Naga Naga with coiled body and hood with 5 heads Unknown
48 207 119 Hanuman Small Hanuman sculpture, standing straight with hands in front of V+
body, gada and tail not so clear, covered in sindūr
49 207 119 Naga-Kal Two Naga figures carved on a stone in collection of sculpture/rocks V+
being worshipped outside temple: bodies have a single loop in
centre and hoods face outwards next to each other. Single heads,
stone has rounded shape to top
50 211 152 Naga-Kal Small Naga-Kal at Nāgarjun Temple on the hill near Rāmṭek: Unknown
modern Ganesh Temple with Naga figure, 5 heads and double
curled body
51 217 43 Ganesh Ganesh sculpture in Medieval shrine at Ambālā: East end of 44x27x12 IV+
Ambālā tank at base of hill. Seated with 4 arms, Snake around
belly. Appears to be late depiction for use in temple alcove
52 217 43 ? Sculpture of a female figure with hands on hips. Jewellery on upper IV+
arms, necklaces, large earrings/part of headdress, holding
something in left hand
805
Appendix 9: List of Sculpture
53 14 98 Ganesh 4 armed seated Ganesh at Devi Temple in Rāmṭek. May be an Early 45x40 III
Historic depiction - sindūr obscures details. Sitting to side with
corpulent body, appears to be carved from sandstone
54 14 98 Naga-Kal Naga stone at Devi Temple in Rāmṭek: single head: Snake has one 58x43x15 Unknown
wide hood and a body that forms one loop
55 111 353 Naga Naga shrine outside modern temple near Kelapur. 5 hoods, appears Unknown
to be sandstone although covered in Sindūr
56 233 227 Hanuman Hanuman Sculpture next to Manapur temple and tank. Incomplete 62x55 V+
and painted in Sindūr: Only top half of sculpture remains: tail
curled overhead, late design
57 233 227 Nandi Nandi, similar to S015 at Ambālā. Heavily decorated with V+
prominent hump and horns. Left leg raised and both back legs
tucked under to the right. Elaborately decorated with bell designs
on back
58 233 227 Vyāla/Yali Siṁha-Vyāla at Manapur Temple next to Tank area. Fragment of 41x23x18 V+
Lion head carved from sandstone, possibly from column piece
59 233 227 Nandi Fragment of small sculpture: animal head that appears to be Nandi? 13x8 IV+
Highly decorated with elaborate headdress, although no ears/horns
visible
60 233 227 Hanuman Hanuman sculpture with tail overhead and hands held in front of V-VI
body
61 241 404 Naga Small 5 headed Naga in shrine by village of Bhilewada. Other Unknown
stones in worship
62 242 410 Naga Chota Mahadev Temple: 5 headed Naga Unknown
63 242 410 Ganesh Chota Mahadev Temple: Seated Ganesh, with 4 arms? Unknown
806
Appendix 9: List of Sculpture
64 242 410 Nandi Nandi at Chota Mahadev, pronounced hump and horns, right front IV+
leg tucked under, left front leg up and back legs visible on both
sides
68 351 84 Bhairava Sandstone carving of Bhairava, being worshipped as a female. 112x62 IV+
Probably later folk art
69 351 84 Lajjā Gaurī Next to Bhairava: small Lajjā Gaurī. No head, type II on Bolon's 8.5x6.5 III
typology
70 351 84 Lajjā Gaurī Upside down Lajjā Gaurī, more detailed than S069 but overall 8x6 III
similar depiction
71 351 84 ? Small sandstone sculpture of two figures: May be Shiva/Parvati? 12x10 Unknown
73 351 84 ? Smaller standing figure with four arms - provincial style V-VI
74 247 83 ? Said to be a seated figure of a naked woman holding a child in her 40 Unknown
lap. Owners said it was carved of red sandstone and that the hair is
visible going down the back. Also stated the figure has three eyes. 3
eyes are visible, but figure appears as an old man with no hair -
possibly modern, or too obscured by paint to be identified
807
Appendix 9: List of Sculpture
75 248 89 Nandi Large Nandi with right leg bent and left tucked under body, neck 50x60x60 IV+
decoration
76 493 87 Ganesh Seated 4 armed Ganesh in Nandpuri, looks to be late/modern V-VI
77 250 174 ? Small sculpture outside Hanuman temple on Mansar Road: seated 40x23 Unknown
figure, large features and squat body. No clear parallel and covered
in sindūr so hard to identify. Could be folk art
78 251 327 Naga-Kal 5 headed Naga carved on a stone with 4 coiled body Unknown
80 251 327 Annapurna Carving of seated Annapurna with ladle across lap, necklaces and IV+
headdress but not clearly female
81 253 330 Hanuman Standing Hanuman figure with hands in front of body V+
82 262 260 ? Small carved plaque with two figures, male and female holding 17x10 IV+
spears. Very small image
83 262 260 Sati stone Pair of matching standing figures in a local folk style 19x49 IV+
84 265 253 Hanuman Striding Hanuman in heroic pose with thick coating of sindūr V+
86 266 261 Naga Small metal Naga, hood visible, body has been concreted into floor 9 Unknown
of shrine
87 266 261 Naga Naga which appears to have 5 heads 16x12 Unknown
808
Appendix 9: List of Sculpture
88 271 202 Liṅga Square base Liṅga: could be old but hard to tell without obvious 50 (+base 40) III
markings: centre point of Hivri village 35x35
89 585 205 Ganesh Seated 4 armed Ganesh with right leg tucked under body. Covered 32x20 Unknown
in sindūr
90 585 205 Naga-Kal 5 headed Naga with double coiled body carved onto stone. Covered 32x20 Unknown
in sindūr
91 585 205 Annapurna Small temple icon of Annapurna with ladle across lap 32x20 Unknown
93 585 205 Nandi Small Nandi associated with the yoni. Fairly crude carving and 27x9x23 Unknown
obscured by sindūr
94 585 205 Hanuman Standing Hanuman with Gada 74x55 V+
95 584 204 Naga-Kal Stone with two one headed Nagas - similar to Sculpture 49 overall 34x28 Unknown
96 583 203 Annapurna Seated female with large ear plugs and headdress, hands on knees 27x21 IV+
but no visible object across lap
97 583 203 ? Standing male figure with ear plugs in provincial style. Probably 32x24 Unknown
quite late but phasing unknown
98 583 203 ? Striding figure - in provincial style 32x19 Unknown
99 272 191 Naga 5 headed Naga with double coiled body 38x22 Unknown
100 274 193 Ganesh Seated Ganesh with two arms - no other features identifiable 40x22 IV+
101 274 193 Ganesh Seated Ganesh, appears to be only two armed 37x20 IV+
809
Appendix 9: List of Sculpture
102 274 193 Naga Perhaps a 5 headed Naga, with 4 coiled body, end of tail visible at 38x20 Unknown
bottom
103 274 193 Ganesh 4 armed seated Ganesh, crude appearance Unknown
104 274 193 ? Standing figure crudely carved onto a stone - folk style image with Unknown
disproportionate limbs
105 275 194 Naga 5 headed Naga with heavily coiled body - included in trimurti 34x20 IV+
106 275 194 Ganesh Seated 2 armed Ganesh - temple icon as part of trimurti 35x19 IV+
107 275 194 Annapurna Image of Annapurna with the ladle - part of trimurti 31x21 IV+
108 275 194 Nandi Small crude image of Nandi 36x26x6 IV+
109 277 215 Naga Small very eroded one headed Naga Unknown
110 348 216 ? Standing figure carved on stone with hands held in front of body 50x27 Unknown
111 348 216 ? Standing figure holding a long object - folk style 38x26 Unknown
112 348 216 Ganesh Seated Ganesh with tilted head? Other features not visible 41x25 Unknown
113 348 216 Nandi Very small badly preserved Nandi 28x11x25 Unknown
114 348 216 Liṅga/Yoni Small Liṅga and pīṭha 7x30x40 IV+
115 348 216 ? Fragmentary and eroded carved stone, could be the remains of a Unknown
female figure but impossible to clearly identify
810
Appendix 9: List of Sculpture
116 283 248 Ganesh 4 armed seated Ganesh at Rithi Bhagi IV+
117 288 58 Liṅga Very large liṅga in Nagardhan town, but mostly buried. 38 (visible) x45 III
Circumference is very large: 160cm
118 290 297 Naga One headed snake with hood - part of trimurti IV+
121 290 297 Naga 5 headed Naga with heavily coiled body Unknown
122 290 297 Nandi Nandi with left leg bent and right tucked under body, quite IV+
decorated but also covered in Sindūr
123 290 297 Hanuman Standing Hanuman in heroic with foot on demon, holding gada in V+
right hand and with tail curved overhead. Late style
124 292 220 Annapurna Image of Annapurna with the ladle - part of trimurti IV+
125 292 220 Ganesh Seated Ganesh, probably with 2 arms - part of trimurti IV+
126 292 220 Naga 5 headed Naga with heavily coiled body - part of trimurti IV+
127 293 379 Naga One headed Naga with single coil in body carved onto stone (stone H 50) Unknown
33x20
128 295 399 Nandi Headless Nandi outside Hanuman temple: very eroded, small in Unknown
size
129 295 399 Liṅga/Yoni Liṅga and pīṭha - the liṅga does not match the base of the yoni Unknown
811
Appendix 9: List of Sculpture
130 295 399 ? Apparently modern carving of a standing figure with hands on hips VI?
and large headdress
131 295 399 ? Quite modern looking standing figure with hands on hips, female VI?
132 295 399 Hanuman Central Hanuman figure of temple: standing in heroic pose holding V+
up the Mountain in the left hand
133 296 440 Hanuman Standing Hanuman in Chaugaon: features are indistinguishable due IV+
to sindūr
134 299 396 Hanuman Large and apparently modern standing Hanuman, with hands held VI?
in front of body and Gada by side
135 299 396 ? Small unknown figure placed on outside of temple - folk style Unknown
136 299 396 ? Small figure placed on outside of temple, almost like Lajjā Gaurī 11x12 Unknown
but too eroded and painted for accurate identification
137 300 397 Memorial stone Carved stone with two standing figures standing side by side. Right IV+
figure has earrings and bangles. Folk art from the second
millennium?
138 303 407 Hanuman Standing Hanuman with Gada on right side of body: cannot V+
distinguish details due to Sindūr
139 305 386 Hanuman Very small standing Hanuman in Sonpur V+
140 307 171 ? Fragmentary seated figure, headless as damaged but it could be old. 18x22 III
Body is corpulent - perhaps a gaṇa or image of Ganesh?
141 307 171 Ganesh Very eroded sandstone sculpture - hard to distinguish but most Unknown
probably Ganesh with the left leg drawn up
812
Appendix 9: List of Sculpture
143 309 361 Hanuman Standing Hanuman sculpture in Gudhegaon: Gada on right side V+
144 309 361 ? Squatting figure, very hard to distinguish as eroded Unknown
145 309 361 Naga Naga with coiled body, hood looks large enough for 5 heads Unknown
148 311 326 Naga 5 headed Naga with knotted body Unknown
149 312 410 Naga 5 headed, modern looking, Naga 27x23 VI?
150 312 410 Ganesh Seated Ganesh, could be modern but entirely covered in thick 35x27 VI?
Sindūr
151 312 410 Nandi Nandi with decorations around neck and back, left leg raised. Small 35x40 IV+
pointed hump
152 314 424 Hero stone Carved stone with two striding figures (in battle?), one with sword 45x34 IV+
and shield, the other with a sword
153 314 424 Narasiṃha Small Narasiṃha tearing the demon across lap - only example of 17x9 IV+
this iconography in survey area
154 314 424 ? Standing animal, possibly back of a monkey? Short round tail, Unknown
missing its head. Folk art
155 314 424 Bagh Animal that appears to be lying down - Hard to distinguish features IV+
but can see paw? Identified locally as Bagh
813
Appendix 9: List of Sculpture
156 315 426 Naga 5 headed Naga with double coiled body 36x22 Unknown
157 315 426 Naga 5 headed Naga with unequal sized heads, middle is most prominent, 32x17 Unknown
double coiled body
158 315 426 Hanuman Standing Hanuman with Gada on left side, fairly modern VI?
appearance
159 315 426 Naga Very eroded 5 headed Naga with knotted body Unknown
160 315 426 Annapurna Seated image of Annapurna with the ladle. Part of temple trimurti Unknown
161 316 270 Hanuman Standing Hanuman with gada on left side 67x32x19 V+
162 316 270 Hanuman Male figure, standing or striding, with strange face and headdress V+
(could be a local depiction of Hanuman?)
163 317 271 Hanuman Striding Hanuman holding Gada in right hand V+
164 317 271 Naga Very covered in Sindūr but appears it could be 5 headed with Unknown
double coiled body - part of trimurti
165 317 271 Annapurna Image of Annapurna with the ladle - part of trimurti Unknown
166 317 271 Ganesh Seated Ganesh, very covered in Sindūr so cannot tell if 2 or 4 Unknown
armed - part of trimurti
167 320 272 Hanuman Standing Hanuman sculpture at Sangrampur Ki Toli, hard to V+
distinguish
168 320 272 Annapurna Seated image of Annapurna carved onto stone: object across lap but IV+
held up on left side in this depiction.
814
Appendix 9: List of Sculpture
169 321 283 ? Standing figure with sword and object in left hand: appears to be Unknown
female - folk art
170 321 283 ? Larger standing figure also with sword and object in left hand, 33x14 Unknown
could be male - folk art
171 323 295 Hanuman Standing Hanuman with gada on left side of body V+
172 324 315 Hanuman Standing Hanuman, badly eroded and covered in sindūr V+
173 324 315 Naga 5 headed Naga with knotted body - part of trimurti 23x16 Unknown
174 324 315 Ganesh Seated Ganesh with left knee up, appears only 2 armed - part of 28x14 Unknown
trimurti
175 324 315 Annapurna Seated figure, Annapurna with the ladle across lap, definitely ear 25x14 Unknown
plugs/earrings but top of head covered in Sindūr - part of trimurti
176 326 324 Ganesh Unknown sculpture, very covered in sindūr, possibly seated figure: 40x30 Unknown
Ganesh? Also looks like it could be a Naga from the side but it is
quite large
177 328 340 Naga Metal snakes: one very clear hood with head on the left. Various Unknown
shapes and sizes
178 328 340 ? Standing figure, male, hard to distinguish, possibly quite modern VI?
179 331 341 Hanuman Standing Hanuman with gada on the left side, very covered in V+
Sindūr
180 233 227 Ganesh Small possibly 4 armed seated Ganesh, very covered in Sindūr, Unknown
can't see details
181 233 227 Naga 5 headed Naga with coiled body 26x23 Unknown
815
Appendix 9: List of Sculpture
182 233 227 Śiva Shiva sitting on a lion with right hand raised and snake around neck VI?
183 233 227 Viṣṇu Pada Possibly modern Viṣṇu Pada ?x25 VI?
184 233 227 Annapurna Bottom half of seated figure, highly decorated, with object across 18x16 IV+
lap. Legs would appear to belong to a female: draped materials and
anklets. The object looks like the end of a gada? Sitting on a lotus?
Back shows material fold covering down from waist. Could be
Yādava period
185 238 278 Hanuman Standing Hanuman in Charkurda temple. Hands held in front of V+
body, Gada on left side and tail curled overhead, late (fairly
modern) design
186 238 278 Nandi Highly decorated Nandi in Charkurda temple, large head in IV+
proportion to body
187 238 278 Ganesh 4 armed seated Ganesh, probably quite modern, holding an axe and VI?
elephant goad?
188 238 278 Naga 5 headed Naga with knotted body 34x20 Unknown
189 238 278 ? Seated figure, female? Broken sculpture: 4 arms, knotted hair or 36x22 Unknown
headdress
190 342 250 Naga-Kal Carved stone with one headed Naga with wavy but not coiled body, Unknown
Nerla
191 342 250 Naga-Kal Carved stone with Naga. Possibly 5 heads given size of hood, Unknown
coiled body and long tail hanging out from the knot of the body
192 342 250 Hanuman Standing Hanuman with foot resting on demon?) and tail curled VI?
overhead: appears fairly modern
193 342 250 ? Two figures sitting on a chariot/cart Unknown
816
Appendix 9: List of Sculpture
194 343 299 Hanuman Standing Hanuman with gada by left side in Sirpur V+
195 343 299 Annapurna Annapurna with ladle across lap, very eroded and covered in Sindūr 22x15 Unknown
- part of trimurti
196 343 299 Naga Presumably a Naga with perhaps 5 heads due to hood size and body 16x24 Unknown
shape - part of trimurti
197 343 299 Ganesh Seated figure, fairly indistinguishable due to erosion and sindūr but Unknown
presumably Ganesh- part of trimurti
198 343 299 Nandi Simple small Nandi figure, crudely depicted 18x20 Unknown
199 344 304 Hanuman Modern Hanuman sculpture: standing with tail straight up VI?
200 345 311 Naga Free standing 5 headed Naga on coiled body 23x15 Unknown
201 345 311 Garuda Standing figure of Garuda with sword: bird wings clearly visible 105x58 IV+
and large pointed nose. Seems to be wearing a decorated cap,
earrings, necklace and bracelets
202 345 311 Nandi Decorated Nandi with left leg raised and back legs to the right side. 50x26x43 IV+
Extensive bridle and bell designs, not very prominent hump but
long pointed horns
203 345 311 Hanuman Hanuman standing on a demon with gada raised in right hand and 108x59 V+
tail curled over head
204 345 311 Ganesh Ganesh plaque in wall of temple, 4 armed, seated, basic carving - 47x28 Unknown
part of trimurti
205 345 311 Annapurna Annapurna with ladle across lap, headdress and large earrings - part 51x30 Unknown
of trimurti
206 345 311 Naga 5 headed snake with knotted body carved onto stone plaque - part 45x33 Unknown
of trimurti
817
Appendix 9: List of Sculpture
207 345 311 Liṅga/Yoni Proportionally large liṅga on pīṭha with deep run off lines 23x40x60 Unknown
209 347 243 Nandi Basic Nandi, not particularly highly decorated, with small hump 30x39x23 IV+
210 347 243 Hanuman Modern looking standing Hanuman with gada by left side and tail VI?
straight up to the right
211 347 243 Narasiṃha Eroded top half of sculpture: body position and shape of the 43x34x18 III
shoulders is reminiscent of the Narasiṃhas in the Rāmṭek temples
212 347 243 Naga-Kal Naga plaque with single head and figure of 8 body with straight tail 33x20x12 Unknown
coming down
213 349 190 Naga Single headed Naga, quite free standing, with coiled body 22x12 Unknown
214 353 279 Nandi Nandi with legs tucked under and fairly prominent hump 32x29x14 Unknown
215 353 279 Liṅga/Yoni Quite large liṅga and pīṭha 12x48x34 Unknown
216 353 279 Hanuman Hanuman standing on head of demon with club in right hand and 80x42 VI?
something in left hand. Tail curled over head
217 354 293 Hanuman Standing Hanuman with Gada on left side and tail curled over head V+
218 354 293 Hanuman Hanuman standing on demon, holding gada up in right hand and V+
something to chest in left hand. Tail curled behind
219 355 252 Nandi Eroded and headless Nandi Unknown
220 355 252 Liṅga/Yoni Simple yoni with small liṅga Unknown
818
Appendix 9: List of Sculpture
221 360 355 Hanuman Large, possibly modern, standing/striding Hanuman with right hand VI?
raised
222 360 355 Liṅga/Yoni Small round pīṭha with liṅga Unknown
223 359 371 Hanuman Small standing Hanuman, mostly obscured by sindūr V+
224 359 371 Hanuman Striding Hanuman holding gada in right hand and left hand holding VI?
up another object, standing on demon, looks modern
225 364 317 ? Largest figure in temple at base of dam. Standing male figure, long 64x24 Unknown
body and arms, crudely carved - folk art?
226 364 317 ? Striding 4 armed figure with pleated skirt and necklace. Quite badly 36x20 Unknown
eroded - unclear who is depicted
227 364 317 ? Main sculpture under worship: standing, possibly 4 armed, said to 57x32 Unknown
be female, measured without head projection from sindūr build-up
229 364 317 ? Top half of standing male (?) with right hand raised 15x8 Unknown
231 364 317 Hanuman Standing Hanuman with right hand raised and tail straight up on V+
right side
232 364 317 Memorial Stone Small stone with two figures carved inside a border, hands held in IV+
front of body: A type of local memorial stone?
233 364 317 ? Top half of small stone plaque with carved figure, male, standing 12x8 Unknown
819
Appendix 9: List of Sculpture
234 364 317 ? Standing figure with large ears, hands clasped low in front of body, ?x10 Unknown
crude carving - folk style
235 364 317 Nandi Eroded Nandi, very damaged 37x45 Unknown
236 364 317 Naga Single headed Naga with coiled body in a figure of 8, similar to 40x22 Unknown
others seen across survey area
237 364 317 Naga Tall single headed Naga, broken below hood, with coiled body 67x41 Unknown
238 366 219 Naga Naga looks modern, 4 coils in body and large hood VI?
239 366 219 Liṅga/Yoni Larger of the two liṅga and yoni pieces on the shrine in Satak. Unknown
Eroded yoni and short liṅga
240 366 219 Liṅga/Yoni Small yoni with liṅga which looks to be added on at a later date Unknown
241 366 219 Naga-Kal Naga carved on a stone, singular head and single curl in body, Unknown
circular appearance
242 368 276 Hanuman Broken Hanuman sculpture (headless) on mound area near to 80x47 V+
Udapur, standing with foot on the demon and with clothing/object
hanging between legs
243 370 365 ? Mata Mai temple: Small standing figure, carved on a stone backing. Unknown
Arms raised at sides, appears to have a necklace, and right hand
perhaps holding something
244 373 288 Hanuman Striding Hanuman, very ornate, holding gada up by right shoulder, V+
foot raised, looks like it is curled around or resting on
something/someone holding up something in left hand (the
Mountain?) drapes of material and jewellery visible
245 373 288 Ganesh Small seated 4 armed Ganesh, right leg bent up with hand resting Unknown
on knee. Left lower arm holding something out (sweetmeats?)
Upper 2 arms raised and holding attributes
820
Appendix 9: List of Sculpture
246 373 288 Hanuman Part of a Hanuman sculpture? Although facing left unlike others, V+
striding position, right leg bent but top half missing
247 373 288 Annapurna Image of Annapurna with the ladle - part of trimurti Unknown
248 377 401 Hanuman Standing Hanuman near Bhilewada, covered in sindūr. Hands in V+
front of body
249 378 387 Hanuman Old Soneghat: Modern Hanuman, standing with hands in front of VI?
body, tail curled around head and gada under left arm
250 378 387 Nandi Nandi with raised left leg, both back legs tucked under and to the Unknown
right, fairly prominent hump.
251 378 387 Ganesh Small seated Ganesh, obscured by Sindūr, presumably 4 armed - Unknown
part of trimurti
252 378 387 Annapurna Annapurna with the ladle across lap - part of trimurti Unknown
253 378 387 Naga Small Naga statue, highly coiled body with probable 5 headed hood Unknown
- part of trimurti
254 379 389 Hanuman Probably modern standing Hanuman, hands in front of body and VI?
tail overhead, gada under left arm
255 380 383 ? Small carved standing figure in shrine by Pindkapar, crude figure Unknown
and badly preserved. Two other stones more difficult to make out -
all in folk style
256 383 375 ? Top of sculpted sandstone figure in small roadside shrine, shoulders 21x25 Unknown
and head. Rounded head, large ears and what appears to be a bun
on top
257 283 248 Ganesh Quite large Ganesh sculpture at Bhagi Rithi. 4 armed and seated 70x40 Unknown
with right leg bent
821
Appendix 9: List of Sculpture
258 385 377 ? Outside Dongri: small figure in wall of shrine, hard to distinguish, Unknown
seems to have 4 arms, quite squat, either sitting or moving/dancing
259 386 373 Hanuman Standing Hanuman, nearly completely obscured in sindūr V+
260 386 373 Nandi Small Nandi, quite squarely carved, but mostly obscured, appears Unknown
quite modern
261 386 373 Ganesh? Presumably a seated Ganesh but completely obscured Unknown
262 386 373 Annapurna Annapurna with a necklace around neck and hair piled on head. Unknown
Crossed legs, and ladle resting in lap and hands - part of trimurti
263 386 373 Naga Coiled body, obscured by sindūr but appears to be a 5 headed hood- Unknown
part of trimurti
264 387 376 ? Seated figure, male, wearing robes on lower half and topless. Both 48x23 Unknown
knees are drawn up with hands resting on knees. Some form of
headdress or decoration - folk style?
265 388 357 Naga 5 headed Naga with double coiled body, probably quite modern Unknown
266 389 358 ? Seated figure, probably female, right knee raised and left leg tucked 50x20 IV+
in front. Holding an object up in both hands, long straight object
like sword in right hand and club/gada shaped object in left hand.
Appears to have long hair hanging down to shoulders as well as
piled on head
267 389 358 ? Second seated figure in the Musewadi temple, left knee bent up 50x37 Unknown
with right leg in front of body. Crude carving, small legs compared
to torso and head. Hands resting on knees, perhaps hair behind the
head, no objects/attributes visible - folk style
822
Appendix 9: List of Sculpture
268 391 335 ? Small standing figure with legs together and arms out to side, ?x5 Unknown
wearing robes, arms and head broken off
269 394 158 Liṅga/Yoni Small modern liṅga and pīṭha, black in colour VI?
271 395 159 Viṣṇu Pada Small sandstone Vishnu Pada, plain, undecorated 14x14 IV+
272 397 169 Architectural Architectural fragment, with several figures. A larger figure, torso 46x16 IV+
only, right hand held up and left by hip. Above is a seated female,
side on, leaning on left hand with right hand raised. Appears to be
part of railing or door jamb?
273 397 169 Elephant? Sandstone animal figure: appears it could be an elephant with a 27x15 IV+
raised trunk, can also see quite large ears and decoration around the
back of the neck and over the back. From position, it could be
rearing up on back legs? Architectural piece?
274 397 169 Vyāla/Yali Sandstone Lion headed Vyāla, badly eroded, mane is more visible 20x11 IV+
but face is difficult to work out. Architectural decorative piece
275 194 393 Nandi Nandi sculpture, painted so difficult to see the carved stone, 55x30x65 Unknown
prominent small hump and decorated body, quite square carving
with thin legs. Both back legs out to the right and left front leg bent
up
276 194 393 Hanuman Covered in Sindūr but fairly modern Hanuman with hands in front 80x41 VI?
of body, standing straight up, presumably a gada under left arm
277 194 393 Naga Coiled body, knotted close to hood, single head but defined features 50x20 VI?
and tongue visible - part of trimurti
278 194 Ganesh Presumably a 4 armed seated Ganesh, right leg bent up, very 42x29 Unknown
obscured by sindūr - part of trimurti
823
Appendix 9: List of Sculpture
279 194 393 Naga Free standing 5 headed Naga, black in colour and on a base with 34x33 VI?
same designs as later architecture. Body coiled around with hood
rising up, like a real snake
280 194 393 Annapurna Seated figure, body quite feminine given thin waist and potential 49x26 VI?
breasts but face seems male with earrings, crown/headdress, and
large ears. Decorated with jewellery. Something behind or on left
shoulder (possible snake connecting the figure to Śiva?). Sitting on
lotus style platform, left leg bent in front and right knee up and foot
turned out. Ladle held in hands across lap
281 194 393 Liṅga/Yoni Liṅga and pīṭha with snake coiled around liṅga 20x56x84 VI?
282 194 393 "Kālidāsa" Seated slim-waisted figure in yogic position, said to be Kālidāsa by 61x27 IV+
locals. Figure is sat on a lotus style platform with an animal hand
under his legs - there is also another small seated figure in the front
of the platform. Hands held on lap, left holding out an object but
right broken, remains of some object with beaded shape visible.
Arm bands, bracelets and necklaces visible. Distinct beard and
curled moustache visible and hair piled on head. Quite long face
with big eyes and well defined features despite sindūr covering.
Folk art with Śaiva ascetic matted hair. Below figure appears to be
a buffalo head and a devotee. Probably a temple niche and part of a
larger structure
283 194 393 Annapurna Unknown figure resembling Annapurna, seated with crossed legs, 46x26 IV+
quite portly, topless but wearing robe with hanging bit on bottom.
Round face with large decorated ears but hair difficult to work out.
Holding a ladle in hands across lap as seen in other sculptures
824
Appendix 9: List of Sculpture
284 194 393 ? Another unknown figure at Japala temple: Also seated, definitely 37x27 IV+
male with similar folk characteristics like the figure said to be
Kālidāsa. Legs are crosses and left hand is resting on lap holding
out an object. Right hand holds an object with interesting shape
similar to the broken piece on the Kālidāsa figure. Topless and less
jewellery, slim waisted, face quite obscured but still appears to have
a moustache
285 194 393 Pīṭha Pīṭha with a flower instead of a liṅga at the centre. Long and quite Unknown
eroded
286 194 393 Naga Probably modern Naga, body curled around itself with visible end VI?
of tail, tall slim hood with 5 heads
287 194 393 Hero Stone Carved stone piece with quite geometric crude figure carved on. Unknown
Standing figure in triangular skirted dress, left arm raised and
holding a spear? Right hand down and holding an object. Round
face with prominent ears - possible hero stone in folk style
288 405 380 ? Modern sculpture of a goddess, feet turned to right and arms by VI?
side. Looking face on, wearing short sari like outfit and necklace.
Hair down to shoulders
289 405 380 ? Carved stone piece with crude standing figure, arms by side - folk Unknown
art
290 406 161 Naga-Kal Broken Naga stone with single hooded Naga, uncoiled tail with Unknown
slight curve in it - could have had a second Naga to the right side,
based on design
291 410 165 Naga Metal Naga? Partially concreted into shrine, straight body, curling Unknown
at bottom and hood potentially represents 5 heads
292 200 114 Nandi Damaged Nandi, missing part of face. Back legs curled under to the Unknown
left, right front leg bent up. Simple carved decorations on back,
fairly crude depiction
825
Appendix 9: List of Sculpture
293 200 114 Annapurna Sitting figure, possibly with ladle across lap. Large ears visible, Unknown
appears male, with crossed legs?
294 200 114 Śiva Head of a sculpture: Hair piled on head or tall headdress - from the III
side you can see the matted hair. Eyes appear to be closed, wide
nose, lips are full. Reminiscent of 5th century design
295 200 114 ? Partial head, eyes upwards visible. Large almond eyes. III
Hair/headdress reflects those on some Mansar sculptures, see
Bautze 2008. There is a bun behind and appears to have pleated
hair in the front. On top it may be possible to see a rosette or some
other headdress feature
296 199 113 Viṣṇu Pada Small plain Viṣṇu Pada, now sealed in concrete but possible older, Unknown
appears to be of sandstone
297 199 113 Nandi Small, squat Nandi with legs tucked under to left side and front Unknown
right leg bent up. Decorated around neck, small upraised head and
little hump
298 199 113 Liṅga/Yoni Large modern pīṭha sculpted around a stone liṅga, small liṅga, VI?
fairly square
299 199 113 Naga Small free standing Naga, coiled upright body and rounded hood, Unknown
appears to be single headed
300 199 113 Ganesh Small seated Ganesh, obscured by sindūr - part of trimurti Unknown
301 199 113 Annapurna Small seated figure, left leg in front and right knee bent up. Quite Unknown
squat, ladle across lap - part of trimurti
302 419 207 ? Small seated sculpture, right leg bent up, hands resting on knees. Unknown
Large material piece draped between legs. Round face, prominent
ears and hair on head in small bun? Anklets, necklace and earrings
visible, not particularly fine carving - folk art
826
Appendix 9: List of Sculpture
303 421 198 Hero Stone Part of a Hero stone? Only sun and crescent moon visible Unknown
304 421 200 Naga-Kal Naga stone: small Naga image in square carved onto stone, figure 37x21 Unknown
of 8 body and clear hood
305 421 198 Hanuman Fairly modern standing Hanuman , hands in front of body, gada 80x40 VI?
under left arm and tail curled over head with a twist in it
306 421 198 Nandi Large Nandi, quite heavily decorated and square in shape. Right leg 52x22x45 IV+
bent up and back legs tucked under to the left. Bridle and bells clear
307 421 198 Ganesh 4 armed seated Ganesh, right leg bent up, very obscured by Sindūr- 30x18 Unknown
part of trimurti
308 421 198 ? Figure with crossed legs and robes on, holding hands in prayer in 31x20 Unknown
front of body. Two arms also visible behind. Crown or headdress
visible and quite prominent ears - part of trimurti
309 421 198 Naga 5 headed Naga with highly coiled body - part of trimurti 32x20 Unknown
310 422 201 Naga Platform shrine in Hivra with large modern Naga, figure of 8 body VI?
and tall 5 headed hood
311 422 201 Nandi, liṅga & Small Nandi, liṅga and yoni set, quite decorated and seems late in Unknown
Yoni date. Bridle and bells visible
312 423 212 Hanuman Small standing Hanuman, hands in front of body and tail just V+
visible curled overhead. It and other associated sculpture, is entirely
obscured by sindūr
313 424 222 Architectural Sandstone architectural piece, partially obscured by sindūr and also IV+
upside down. A face is visible with full lips and a fairly large nose.
The Headdress/crown appears to be part of a wider structural
element
827
Appendix 9: List of Sculpture
314 424 222 Naga Naga with coiled body and tail end coming out of the centre. Large Unknown
head
315 426 224 Śiva Sculpture of Śiva, painted blue. Śiva is sitting and holds his right VI?
hand up while the left rests on his knee. A snake is coiled around
his shoulders with the hood on the left side
316 427 225 Architectural Sandstone architectural Fragment (?) in a 'Kali' shrine, rectangular IV+
block with some carvings
317 427 225 Architectural Smaller architectural piece IV+
318 428 226 ? Standing figure with completely round and flat face, no hair or ears 54x20 Unknown
but facial features picked out quite geometrically. Arms by sides,
crudely carved - folk art
319 429 229 ? Tulja Bai'(?) Covered sandstone figure, with umbrella and 15x25 Unknown
backing/halo to head. Small bun on top of head
320 429 229 Naga-Kal Naga stone with coiled body and tail pointing down. One head 50x26 Unknown
321 430 230 Amalaka Amalaka architectural fragment from top of temple. IV+
322 430 230 ? Very eroded standing figure, hard to distinguish Unknown
323 434 232 Naga-Kal Naga stone with circular curled body and thin head/hood pointing 46x34 Unknown
straight up
324 434 232 Nandi Small Nandi with quite prominent hump, covered in sindūr 29x20 Unknown
325 434 232 Liṅga/Yoni Large pīṭha, with liṅga, but may be later addition, very circular with 21x14 VI?
long projection. Yoni measures 85L by 52W
326 434 232 Ganesh Headless sitting figure, probably Ganesh, pot-bellied, hands resting 35x50 Unknown
on lap, left leg in front and right knee bent up
828
Appendix 9: List of Sculpture
327 434 232 Hanuman Very basic/crude standing Hanuman, probably modern. VI?
328 435 233 Hanuman Small standing Hanuman, holding gada over right shoulder? 30x13 V+
329 435 233 Naga-Kal Naga stone with loosely coiled body and apparently single head and Unknown
hood
330 435 233 Ganesh Seated Ganesh, apparently only with two arms. Sindūr covered so Unknown
details not visible
331 435 233 Naga Small Naga with rounded coiled body and probably a 5 headed Unknown
hood - obscured by sindūr
332 436 273 Naga Modern Naga stone with figure of 8 body and 5 headed triangular VI?
hood
333 438 266 Memorial Stone At Mata Mai temple: Two standing figures holding hands, male on 32x34 IV+
left and female on right. The male is holding a sword (?) and
skirted/robed from waist down, and the female is holding a water
pot (?) and is also robed with material at the front and beaded
necklaces, upper arm torque and bangles are visible. Both have
prominent ears, particular the female, and larger round earrings are
also visible on her. Face is eroded away on the female, the male has
some features visible, flat nose and almond eyes, smiling? Hair or
band around forehead. Possibly also has beaded necklace. Pieces
such as this could be memorial stones in provincial style
334 439 267 Hanuman Large standing Hanuman with foot on demon, tail curled overhead V+
and gada held up in right hand, while left hand is in front of body
335 439 267 Hanuman Smaller heroic Hanuman with Gada in right hand, tail up by right V+
side and left hand holding up the mountain
829
Appendix 9: List of Sculpture
336 439 267 Ganesh Seated Ganesh, appears to be 4 armed. Left leg bent up with right Unknown
leg in front. Long head, seems to have a headdress with a central
ornament - part of trimurti
337 439 267 Annapurna Sitting figure, looks female, wearing sari like robes and necklaces. Unknown
Right leg up, same style as the Ganesh. Object across lap held in
hands. Prominent ears and looks to have a headdress also - part of
trimurti
338 439 267 Naga Tightly coiled body with clearly 5 headed hood - part of trimurti Unknown
339 439 267 Liṅga/Yoni Small pīṭha with liṅga (might not be original) VI?
340 439 267 Nandi Small Nandi, back legs tucked underneath to right hand side, right Unknown
front leg raised, lots of decoration including bridle and bells, small
pointed hump
341 439 267 Annapurna Another seated Annapurna probably quite modern, wearing sari VI?
style robes, prominent ears with big round earrings and
triangular/diamond shaped necklace and ladle in hands across lap.
Right knee bent up and left leg in front. - part of trimurti
342 440 268 Memorial Stone Stone with standing figures holding holds, man on left and woman Unknown
on right. Male's right arm is broken so cannot see if holding object.
Both wearing robes and decorated with beaded necklaces. Female
also has torques and bangles. The Male's upper half seems clothed
in a braided material? Both have large ears and the female has large
round earrings in. Male has hair carved into stripes with a band
across, female has hair drawn back and an ornament in the middle
of forehead. Both had smiling faces, eyebrows and quite simplistic
faces. Similar in style to GPS 438. Behind them the stone is carved,
perhaps with sun? Said to be Rukhmani by local children - folk
depiction
830
Appendix 9: List of Sculpture
343 441 269 Naga Metal snake, body straight up and tail curled around at bottom. Unknown
Representation of 5 heads/forked tongues coming out of hood
344 441 269 Naga Hood of metal snake visible, 5 heads Unknown
345 442 420 Naga Coiled knotted body with the body arching around the back of the Unknown
hood and the tail coming out the front, stone Naga with 5 clear
heads - part of trimurti
346 442 420 Ganesh Crudely carved sitting Ganesh, hard to determine features, appears Unknown
to have only 2 arms, quite elongated - part of trimurti
347 442 420 Annapurna Annapurna with ladle held in hands across lap. Seems to have a Unknown
headdress or hair piled up. Face is slightly obscured, but is quite
full and round with large almond eyes - part of trimurti
348 442 420 Nandi Small quite eroded Nandi with left leg tucked under and right front Unknown
leg raised.
349 442 420 Liṅga/Yoni small rounded pīṭha with liṅga, in proportion with the Nandi Unknown
350 443 421 Naga Small free standing Naga with single coil in body and one hood 29x10 VI?
351 444 422 Hanuman Standing Hanuman with Gada on left side and hands held in front V+
of body.
352 444 422 Hanuman Slim standing Hanuman with hands in front of body and gada on V+
left side. Tail can be seen curling over head
353 444 422 Nandi Small square Nandi with collar and decoration on back, front legs Unknown
curved in strange way - crude depiction
354 444 422 ? Top of figure, incomplete and eroded, hard to distinguish Unknown
831
Appendix 9: List of Sculpture
355 444 422 Ganesh Sitting Ganesh, with legs crossed in front, 4 armed, holding Unknown
attributes, left lower arm holding out something, right is resting on
knee - crude depiction - part of trimurti
356 444 422 Annapurna Annapurna with ladle held in hands across lap. Wearing anklets, Unknown
large ears with round rosette style earrings visible - part of trimurti
357 444 422 Naga Coiled Naga with body curled around hood and tail coming out the Unknown
front, seems to have only one head - part of trimurti
358 445 423 ? Sitting animal, headless. Slim, straight front, legs. Sitting on 27x10x31 Unknown
haunches with tail visible up the back, shaped like a lion's tail (?)
359 445 423 Bagh Said to be Bagh but badly eroded: Tiger. Long animal, lying down. 18x12x42 Unknown
4 legs visible and front of snout. Folk-Art, very hard to date
360 446 416 Naga Free standing Naga with 5 heads: simple coiled body Unknown
361 446 416 Hanuman Standing Hanuman, hands in front of body and gada by left side, V+
tail just visible over head
362 446 416 Nandi Nandi, very obscured by sindūr, but back legs tucked under to right Unknown
363 446 416 Liṅga/Yoni Quite deep pīṭha, probably with replacement liṅga Unknown
364 446 416 Naga Naga obscured by Sindūr but can see coiled body, also appears to Unknown
have 5 heads- part of trimurti
365 446 416 Annapurna Annapurna - seated figure with right leg bent up and ladle in hands Unknown
across lap - part of trimurti
832
Appendix 9: List of Sculpture
366 446 416 Ganesh Sitting Ganesh, can only clearly make out two arms and hands Unknown
resting on knees. Trunk hangs to left of body, quite pot-bellied -
part of trimurti
367 447 417 ? Small carved piece, unsure what is being depicted: Three 8x8 Unknown
projections and two holes between them - could be part of small
Naga or architectural piece
368 447 417 Hero Stone Figure carved onto stone, with left hand in front of body, prominent 16x12 Unknown
ears and earrings. Broken and eroded - possibly part of later folk
hero stone
369 447 417 Annapurna Main figure under worship, sitting on pedestal, right leg down left 48x24 V-VI
crossed on lap: Annapurna with ladle in hands across lap. Wearing
sari type robes and bangles. Appears could be female.
Crown/Headdress and hair to neck. Halo behind head. Probably
quite modern
370 448 418 Naga Small free standing Naga with simple curled bottom of body and VI?
large 5 headed hood
371 448 418 Hanuman Basic standing Hanuman with left foot slightly raised with bent V+
knee and gada on right side
372 448 418 Nandi Very square Nandi with legs tucked under body, some decoration Unknown
visible but mostly obscured
373 448 418 Liṅga/Yoni Broken sandstone pīṭha with separate liṅga Unknown
374 448 418 Ganesh Crude carving built into wall of temple and mostly obscured. Can Unknown
see the Ganesh has 4 arms, a pot belly and legs underneath the body
- part of trimurti
375 448 418 Ganesh? Another obscured sculpture with legs tucked under the body, Unknown
possibly with 4 arms but they are raised and not resting on the legs,
could be Ganesh but face is obscured and there is no trunk hanging
downwards
833
Appendix 9: List of Sculpture
376 448 418 ? Standing figure, obscured by sindūr and built into temple wall, Unknown
appears potentially to be female with hands by hips
377 450 419 Hanuman Standing Hanuman, Hands in front of body and gada down by left V+
side. Tail is clearly curled behind the head and Hanuman has a
headdress
378 451 413 ? Small standing figure, left arm by side and right raised. Eroded so Unknown
features missing - folk depiction
379 451 413 Hanuman Very basic standing Hanuman with left leg bent and raised, resting VI?
on demon. Tail is curled overhead and right hand raised while left
rests on left knee. Probably modern
380 451 413 Ganesh Part of a sitting figure, probably Ganesh, with right leg bent and Unknown
hand resting on knee, left leg in front of body - part of trimurti
381 451 413 Annapurna Sitting figure of Annapurna, right leg bent and knee raised, ladle Unknown
held in hands across lap- part of trimurti
382 451 413 Naga Small Naga, quite obscured but coiled body and hood visible. Unknown
Appears to have 5 heads - part of trimurti
383 451 413 Naga Small sandstone coiled Naga with 5 heads in own shrine: knotted Unknown
body
384 451 413 Pīṭha Large round pīṭha with short projection, no liṅga Unknown
385 451 413 Nandi Small Nandi with back decorations, bridle and belled collar, long Unknown
flat ears and a small hump
386 451 413 ? Sitting figure, looks male, left leg bent in front of body and right Unknown
knee up. Two arms, hard to distinguish features - folk depiction
834
Appendix 9: List of Sculpture
387 452 414 Naga Small metal snake with single hood next to trident VI?
388 454 441 Hanuman Modern looking Hanuman sculpture, striding with left leg raised. VI?
Gada in right hand and left hand holding up the Mountain (?)
389 455 442 Naga Probably modern Naga stone, figure of 8 body and large 5 headed VI?
hood
390 457 428 Naga Probably modern Naga stone, figure of 8 body and large 5 headed VI?
hood
391 458 429 ? Rounded 'head' area and long pointed section, very angular with Unknown
spike at end. Rounded portion almost appears to have animal like
facial features - local item of significance or folk depiction
392 459 430 Naga Metal Naga with curled body as a base and a single head in an open Unknown
hood
393 460 431 ? Sitting/reclining figure, covered in sindūr so mostly obscured. Unknown
Appears to have right hand on raised knee and resting back on left.
Quite pot-bellied, face completely obscured
394 460 431 Nandi Top portion of a quite large Nandi, front legs and head, covered in Unknown
sindūr
395 462 433 Hero Stone Top of a carved stone, rounded top, possibly hero stone, very Unknown
eroded so can't make out detail
396 463 434 Śiva & Consort? Sitting Śiva with crossed legs and snake around neck with hood to VI?
left side. Consort (?) also has snake on right shoulder and is holding
a trident. Potentially a tiger as well on right shoulder Painted, hard
to distinguish features, material or age, probably modern
835
Appendix 9: List of Sculpture
397 463 434 Ganesh Seems to be a small 4 armed seated Ganesh, with left leg across lap, VI?
probably modern
398 464 435 Trimurti Frieze with 3 figures: Seems to be Shiva with a snake coiled around Unknown
neck, small Ganesh, Annapurna
399 464 435 Naga Tall metal Naga on small coiled base Unknown
400 464 435 Liṅga/Yoni Small pīṭha and liṅga, which look fairly modern VI?
401 464 435 Nandi Fairly modern looking Nandi with decorated back, prominent Unknown
pointy hump, same stone as the liṅga
402 466 437 Durga? Female figure with 8 arms sitting on a tiger. Lowest hands, left is Unknown
resting on leg and right is broken but appears to have been held up.
Holding various attributes in back hands including a knife, a trident
403 466 437 Naga Thick metal Naga with one head Unknown
404 467 187 Naga Small metal Naga with one head, mostly buried in the shrine. Head Unknown
has pointed end
405 468 188 Hanuman Possibly modern Hanuman, striding with Gada in right hand and VI?
holding up the Mountain in the left
406 471 403 Bagh Very eroded sculpture with rounded head Unknown
407 471 403 Bagh Larger eroded figure with rounded head, no real discernible Unknown
features. From position, could be Nandi
408 471 403 Bagh Another round headed figure, appears to be lying down from body Unknown
position, cannot distinguish any features
409 472 394 Hanuman Modern Hanuman sculpture, striding with gada in right hand and VI?
holding up the Mountain in the left
836
Appendix 9: List of Sculpture
410 473 398 Hanuman Standing Hanuman, probably modern, Gada under left arm and tail VI?
overhead but to the left
411 474 395 Annapurna Annapurna with ladle across lap. Wearing sari like robes and a Unknown
headdress
412 475 391 ? Large head, with trident on forehead and crown placed on top: Unknown
modern rudimentary object of worship?
413 475 391 Naga Probably modern Naga with figure of 8 body and large 5 headed VI?
hood
414 475 391 ? Standing robed figure, male, holding right hand up - folk depiction Unknown
of unknown figure
415 477 384 Hanuman Very obscured by sindūr, standing hanuman with hands in front of VI?
body and tail over head
416 477 384 Nandi Small Nandi, quite eroded, with prominent hump and belled collar Unknown
visible. Left front leg is curved and raised
417 477 384 Liṅga/Yoni Tall stone pīṭha with oval shaped liṅga attached Unknown
418 478 97 Amalaka Piece of an amalaka from a temple in a roadside shrine being 37x34 IV+
worshipped as the seven sisters (associated with the farming land)
419 479 381 Hanuman Probably modern, quite crude sculpture of Hanuman, kneeling on VI?
left leg with gada clasped in hands in front of body
420 479 381 Hanuman Standing Hanuman with hands in front of body, gada under left arm V+
and tail curled over head
421 479 381 Nandi Small stone Nandi with fringed collar, bridle and back decoration, IV+
legs tucked under body front and back
422 479 381 Liṅga/Yoni Appears to be one piece, rounded bottomed pīṭha with side Unknown
projections and small liṅga
837
Appendix 9: List of Sculpture
423 479 381 Ganesh Sitting Ganesh with left knee bent up. Badly preserved and covered Unknown
in sindūr - part of trimurti
424 479 381 Annapurna Sitting Annapurna with right knee bent up, and ladle held in hands Unknown
across lap. Face obscured but large ears and round earrings visible -
part of trimurti
425 479 381 Naga Naga with coiled body, seems to have a single head. Scales on body Unknown
are visible - part of trimurti
426 480 438 ? Carved stone with two sitting figures, mirror image of each other: 24x25 Unknown
each has an outside hand raised, the inside leg bent up with hand
resting on knee. Faces quite obscured, sculpture is eroded - folk
image
427 481 439 Naga Shrine with metal trident and snake winding round (modern?) but VI?
hood of another metal Naga also present
428 482 172 Hanuman Architectural fragment or carved stone with striding figure, V+
probably Hanuman, only torso visible
429 482 172 ? Architectural fragment or piece of carved sculpture, design hard to Unknown
identify as fragmentary
430 482 172 Hanuman Standing Hanuman, very obscured but Gada is visible held up in V+
right hand
431 482 172 Naga-Kal Part of a Naga stone with body twisted into backwards S shape and Unknown
round single head
432 482 172 Ganesh Sitting 4 armed Ganesh. Legs are underneath the body with feet Unknown
touching, trunk is raised on left side of body and seems to have
headdress, quite crude carving and obscured with sindūr
433 482 172 Hanuman Although badly preserved image of a heroic Hanuman V+
838
Appendix 9: List of Sculpture
434 482 172 Hanuman Badly preserved and sindūr covered but could be a striding V+
Hanuman with Gada in right hand. Although could also be another
figure holding a sword as face seems to have ears
435 483 173 ? Possibly modern sculpture of goddess with multiple arms: 4? Unknown
Holding various attributes but covered in clothing
436 484 254 Hanuman Standing Hanuman sculpture, hands in front of body and gada V+
under left arm
437 484 254 Trimurti Trio of figures: L-R Ganesh, Annapurna and Naga Unknown
438 484 254 Naga Naga stone with double figure of 8 coiled body and 5 headed hood Unknown
439 485 255 ? Sitting figure carved on a stone with left leg hanging down and 17x31 Unknown
right leg drawn up to chest. Left hand is between legs and right
holds and object to ear/pours something over body. F
440 485 255 ? Standing female figure, hands in front of body, long ears, robes 29x11 VI?
clear on bottom with material hanging in centre - folk art
441 485 255 ? Smaller standing female figure, hands in front of body, but not 27x11 VI?
touching. Resting on hips? Long ears, slightly different robes like a
sari - folk art
442 486 256 Hanuman Standing/Striding Hanuman with left foot raised on demon, let hand 105x48 V+
in front of body, right hand raised and tail curled over head
443 486 256 Ganesh 4 armed seated Ganesh in wall of temple. Left leg bent and right in 21x16 Unknown
front of body, definite headdress - part of trimurti
444 486 256 Naga Naga in wall of temple with coiled body and 5 heads - part of 22x19 Unknown
trimurti
445 486 256 Annapurna Sitting figure in wall of temple, Annapurna with ladle held across 26x17 Unknown
lap, main piece obscured by sindūr - part of trimurti
839
Appendix 9: List of Sculpture
446 486 256 Nandi Eroded smaller Nandi, hardly any features visible, lying in typical 38x10 Unknown
position
447 486 256 Nandi Larger eroded Nandi, also hardly any features but clear posture and 23x10 Unknown
shape of body
448 488 258 ? Red sandstone carved piece, seems to be upper torso and arms but 22x20 Unknown
the stone has been placed upside down on the platform shrine -
provincial style
449 488 258 ? Red sandstone torso of a female figure with 4 arms. The upper left 38x32 Unknown
hand seems to be holding a small trident and the lower right hand a
sword, the other two are damaged. The chest area is bare except for
a necklace. This sculpture has a strange shape with the broad square
shoulders. Not Vākāṭaka, difficult to date - provincial style
450 489 263 Hanuman Modern looking Hanuman sculpture, striding with left foot up, gada VI?
in right hand and mountain in left
451 489 263 Naga Small 5 headed Naga stone with coiled body - part of trimurti VI?
452 489 263 ? Small female figure, cross legged with hands in front of chest - part Unknown
of trimurti
453 489 263 Ganesh Small cross legged figure with hands held up in prayer position in Unknown
front of body. Statue covered in sindūr - part of trimurti
454 489 263 Nandi Small square Nandi with legs tucked under body, prominent hump Unknown
and large horns. Collar decoration and bridle visible but crudely
carved
455 489 263 Liṅga/Yoni Small round pīṭha and liṅga Unknown
840
Appendix 9: List of Sculpture
456 490 264 ? Standing figure with robed lower half and large beaded necklace Unknown
hanging around neck. Hands are on hips, in swathes of material?
Figure appears to be male, has long ears and something on forehead
- folk depiction
457 490 264 ? Larger standing figure in same style, robed bottom half, large Unknown
beaded necklace and hands on hips in the material. Rounded face,
long ears and something on forehead - folk depiction
458 490 264 ? Top part of a sculpted piece with upper torso and head of figure: Unknown
Holding hands in prayer position in front of body. Hair on top of
head in a bun - folk depiction
459 491 85 Hanuman Standing Hanuman, seems to be modern. Obscured with sindūr. 52x31 VI?
Gada under left arm
460 491 85 Ganesh Small seated Ganesh, obscured by Sindūr 23x12 Unknown
461 491 85 Annapurna Fairly heavily decorated sitting figure on a lotus style pedestal, with 28x18 Unknown
left leg across lap and ladle held in hands across lap. Wearing sari
like robes and a diamond/triangular pendant. Has a headdress on
and seems to have a third eye
462 491 85 Naga Totally obscured by sindūr but can just see coiled of body and hood 30x15 Unknown
(presumably 5 heads given width)
463 492 86 Naga Appears to be a modern Naga stone, figure of 8 body and large VI?
hood but only one head
464 493 87 Ganesh Sitting Ganesh, with 4 arms and elongated body and head. Right leg 71x44 Unknown
bent up and hand resting on knee - part of trimurti
465 493 87 Nandi Small Nandi, quite plain but could be because of sindūr, back legs Unknown
tucked under to left and right front leg bent up. Prominent hump on
back and small horns
841
Appendix 9: List of Sculpture
466 495 337 Hanuman Standing Hanuman, Hands in front of body and Gada under left V+
arm
467 495 337 Ganesh Sitting 4 armed Ganesh holding attributes, right leg drawn up - part Unknown
of trimurti
468 495 337 Naga Coiled body and single pointed head in hood Unknown
469 495 337 Annapurna Annapurna with left leg drawn up and ladle held across lap. Unknown
Wearing a headdress and has prominent ears, could be male or
female- part of trimurti
470 496 338 ? Standing figure, obscured by sindūr and badly eroded. Cannot make 21x16 Unknown
out features
471 496 338 ? Second standing figure even more eroded and damaged Unknown
472 498 320 Memorial Stone? Carved stone with two figures. Very hard to distinguish, large 39x30 VI?
faces, seem to be sitting, quite crudely carved - folk depiction
473 499 321 Naga-Kal Tall Naga stone with triple curled body and straight up hood with Unknown
single head
474 499 321 Hanuman Standing Hanuman, gada under left arm, quite short and rounded in 67x31 V+
shape
475 499 321 Nandi Small Nandi with large hump and horns, bridle and decoration just 25x16 Unknown
about visible
476 499 321 Liṅga/Yoni Small pīṭha with liṅga, in proportion to the Nandi 11x21 Unknown
477 499 321 Naga Probably modern Naga with figure of 8 body and large 5 headed 32x45 VI?
hood
842
Appendix 9: List of Sculpture
478 499 321 Ganesh Also possibly modern 4 armed sitting Ganesh. Quite heavily 38x59 VI?
decorated with small curling trunk, large ears and headdress and
robed lower half. Left leg curled into lap and lower left arm holding
sweetmeats? - part of trimurti
479 499 321 ? Small standing figure: Hands appear to be by hips, ears visible and 37x23 Unknown
possibly hair or some sort of halo behind head - folk depiction
480 500 342 Hanuman Large standing/striding Hanuman, completely obscured 66x33 V+
481 500 342 Annapurna Small Annapurna with ladle across lap, large head but most features 39x23 Unknown
obscured
482 501 343 Nandi Small Nandi with front legs bent up, belled collar and back 40x20x45 Unknown
decoration, small pointed hump and pointed horns
483 501 343 Liṅga/Yoni Liṅga is very big in comparison to pīṭha and carved from a different 44x55x80 Unknown
stone? Pīṭha is quite eroded - possibly pieces of mixed date
484 501 343 Ganesh Part of 4 armed Ganesh built into recess in wall of temple. Features 26x24 Unknown
obscured - part of trimurti
485 501 343 Annapurna Sitting Annapurna with ladle across lap, probably female but again 25x18 Unknown
totally obscured by sindūr - part of trimurti
486 501 343 Naga Small Naga with coiled body and 5 heads - part of trimurti 23x18 Unknown
487 501 343 Hanuman Large standing Hanuman, with hands in front of body and tail V+
probably overhead
488 503 345 Naga Small metal Naga with rounded model and flat hood with single Unknown
head
843
Appendix 9: List of Sculpture
489 504 346 ? Sitting figure with left leg across lap. Left hand resting on lap 44x27 IV+
holding out an object and right hand holding a sword. Figure is
wearing robes and beaded necklaces/decoration over shoulders as
well as a diamond shaped pendant, and has a beaded
headband/headdress on. Large prominent ears, female. Appears to
be same figure as sculpture 266 due to the position, and attributes
490 504 346 ? Standing figure, with broken off arms. Robed in lower half but with 40x12 Unknown
bare legs and wearing some decoration on chest. Male but face
quite obscured - folk depiction
491 504 346 Viṣṇu Pada Very eroded Viṣṇu Pada Unknown
492 505 359 Hero stone Striding figure holding something in right hand and possibly a 52x31 Unknown
shield in left, carved on a stone - local style
493 505 359 ? Standing figure with 4 arms, two held down by sides and other two 37x17 Unknown
raised and holding attributes. Robed with some jewellery but
appears to be male. Headdress and prominent ears visible -
provincial style
494 505 359 ? Bottom half of a standing sculpture, waist and tops of legs. 8x10x14 Unknown
495 505 359 ? Head of a sculpture, possibly related to the bottom half. Tall Unknown
headdress but facial features obscured
496 505 359 Bagh Large headed animal sculpture, possibly tiger or lion. Able to make 24x16x34 Unknown
out mouth and snout, placement of eyes and perhaps ears (holes on
side/top of head). Legs are not defined but joined together.
Standing but short legs. Long tail hanging down on back
497 505 359 Bagh Standing animal sculpture with short legs and relatively large, 20x10x28 Unknown
slightly raised head: can make out snout and perhaps nostrils
844
Appendix 9: List of Sculpture
498 506 360 Hanuman Large standing Hanuman with hands in front of body, gada by left V+
side and tail over head
499 506 360 Nandi Nandi with raised head, legs tucked under body, with clear hooves, 44x18x30 IV+
small hump and bridle/collar visible. Design of cloth over back?
Large ears and pointed horns
500 506 360 Liṅga/Yoni Medium-sized pīṭha with cuts for water run-off and liṅga 40x25 Unknown
501 506 360 Ganesh Seated Ganesh, presumably with 4 arms. Right leg drawn up and 28x20 Unknown
left in front of body, features obscured by sindūr - part of trimurti
502 506 360 Naga Tightly coiled body wrapped behind hood, and large hood with 5 28x20 Unknown
heads - part of trimurti
503 506 360 Annapurna Sitting Annapurna with right leg drawn up and ladle in hands across 28x19 Unknown
lap, face completely obscured by sindūr- part of trimurti
504 507 362 Naga Small possible modern Naga with figure of 8 body and large hood VI?
but only one head
505 508 378 Naga-Kal Stone with small carved Naga, small wavy body and head. Corner Unknown
of stone has modern inscription
506 508 378 Liṅga/Yoni Large flat yoni with long projection, on square base with small 100 Unknown
liṅga in centre. Base has chisel marks
507 508 378 Hanuman Tall standing Hanuman, hands held in front of body, tall headdress V+
508 508 378 Naga Naga with figure of 8 coiled body and wide triangular hood with 5 VI?
heads
509 509 408 ? Standing figure with sari like robes, probably female. Right hand Unknown
raised and left by hip (holding a sword?) Large rounded features
and hair or headdress - provincial style
845
Appendix 9: List of Sculpture
510 510 409 Naga Free standing Naga with figure of 8 body, large hood and 5 heads VI?
511 511 368 Bagh Small standing animal sculpture with rounded head and pointed 58x10 Unknown
body, on short joined together legs. Eye (right) and ear (? Left) just
visible on head but otherwise featureless?
512 512 369 ? Standing figure, probably female with hands held in front of body, Unknown
features eroded - folk depiction
513 513 370 Naga Naga with figure of 8 body and large hood. Single head with visible Unknown
neck area and markings like ridges or scales, mouth open
514 514 124 Naga-Kal Naga stone with figure of 8 body, end of tail up and round hood Unknown
with one head
515 514 124 Naga-Kal Naga stone with single loop in body, and tail up, large oval shaped Unknown
hood with one head, head is visible and quite triangular
516 514 124 Naga-Kal Naga stone with long Naga, S shaped body with loop at the end, Unknown
round hood for one head
517 514 124 Naga Free standing Naga sat on coiled body with large triangular hood - Unknown
seems like one head
518 514 124 Naga Free standing Naga on coiled body piled up 2 or 3 times and small VI?
hood for what seems like one head. Head has been broken off in
past
519 515 125 ? Standing figure, quite crude, possibly modern? Wearing tunic, short Unknown
legs and long arms by side - folk depiction
520 515 125 ? Upper part of small carved figure, appears to be sitting, eroded and Unknown
obscured - cannot identify
521 515 125 ? Standing figure carved on stone, very crude, few features - folk Unknown
depiction
846
Appendix 9: List of Sculpture
522 516 126 Naga Modern Naga with a single head and hood, and tail curling out onto VI?
floor. Unlike any others and also quite obscured by sindūr
523 517 105 Hanuman Large standing Hanuman with hands in front of body and clear V+
gada under left arm
524 517 105 ? Very eroded 4 armed figure, sitting/feet together under body. Two Unknown
hands in front of body and two behind presumably holding objects
525 517 105 Naga Naga with very twisted knotted body and fan-like hood with 5 Unknown
heads. Body has a lot of detail with carved scales and ridges - part
of trimurti
526 517 105 Ganesh Eroded Ganesh, sitting with left leg in front of body and right knee Unknown
drawn up. 4 armed, wearing beaded necklace, with large ears and
relatively small head and clear headdress - part of trimurti
527 517 105 Annapurna Sitting figure of Annapurna with left leg in front of body and right Unknown
knee drawn up with hand resting on knee. Left hand holds an
object. Body is decorated with necklaces including a
triangular/diamond shaped pendant and anklets. Clearly large ears
with round earrings but face is obscured - folk depiction
528 517 105 Ganesh Sitting figure of Ganesh? Left leg in front of body and right knee 34x24 Unknown
drawn up with leg hanging down. Pot-bellied, 4 clear arms. Lower
arms are broken off, left upper arm clearly holds a lotus flower.
Body is decorated with necklaces and bangles. Some form of
headdress of hair decoration and prominent ears with earrings
529 517 105 ? Standing figure, fairly eroded, with hands in front of body. Head is 37x22 Unknown
obscured and it is difficult to tell if there are more arms - folk art
847
Appendix 9: List of Sculpture
530 517 105 Memorial Stone Stone with two carved figures, holding hands, probably male and 42x32 Unknown
female. Outside hands are holding something. The right figure is
slightly smaller and has prominent ears and earrings. Rest of stone
and details are very eroded - provincial style
531 517 105 ? Headless figure, standing with hands in front of body, other 30x16 Unknown
features indistinguishable - folk depiction
532 517 105 Naga-Kal Naga carved on a stone with round coiled body and upright hood 41x26 Unknown
with single head
533 517 105 Naga-Kal Naga carved on a stone with a single knot in body and one head 61x29 Unknown
534 517 105 ? Head of a sculpture with very round face and full lips, hair around 17x12 III
head is textured, and appears to be curly. Could be fairly old
535 517 105 Nandi Part of a Nandi, head and front left leg. Belled collar very clear and 44x18x32 Unknown
pointed horns
536 517 105 Ganesh Very eroded and crude figure of Ganesh with apparently four arms. 41x27 Unknown
Features completely gone, could have been sitting or could be
broken
537 518 127 Naga-Kal Naga stone with coiled body (2 figure of 8s) and a triangular hood Unknown
with 5 heads
538 518 127 Naga Metal Naga with curled body as a base and a triangular hood with 5 Unknown
clear separate heads
539 521 130 Ganesh Small seated Ganesh with left knee drawn up and arm/trunk resting Unknown
on knee. Seems to just be 2 armed, covered in sindūr - part of
trimurti
540 521 130 Naga Naga with very coiled body and wide hood, hard to distinguish but Unknown
could be 5 heads due to hood shape - part of trimurti
848
Appendix 9: List of Sculpture
541 521 130 Annapurna Small seated Annapurna with left leg in front of body and ladle Unknown
held in hands across lap. Wearing sari like robes, large ears with
rounded earrings - part of trimurti
542 521 130 Hanuman Striding/walking Hanuman, turned to his right side. Right hand in VI?
front of body and left raised, tail over head
543 522 131 Nandi Nandi with heavy decoration including belled collars and back 40x21x48 Unknown
decoration. Small hump and square head. Legs tucked under body,
back legs are awkwardly carved
544 522 131 Liṅga/Yoni Round deep pīṭha with short projection and side additions, small 12x37x57 Unknown
square liṅga
545 522 131 Hanuman Standing Hanuman with left foot on demon, left hand in front of V+
body, right raised and tail over head
546 522 131 Naga? Naga? Totally covered in sindūr, but appears to be a coiled body Unknown
with one hood in the centre
547 522 131 Ganesh Large modern Ganesh. Sitting with left leg in front of body, pot- 150x70 VI?
bellied
548 523 132 Hanuman Head of Hanuman, badly damaged V+
549 523 132 ? Head of a sculpture, very round with clear ears. Features mostly Unknown
obscured by sindūr
550 523 132 Hanuman Large standing Hanuman with hands in front of body and gada VI?
underneath left arm
551 524 133 Hanuman Hanuman standing on demon, with left arm in front of body and VI?
right hand raised
552 524 133 Liṅga/Yoni Stone pīṭha and liṅga, seem to be one piece Unknown
553 524 133 Ganesh Sitting Ganesh, obscured by sindūr - part of trimurti 18x28 Unknown
849
Appendix 9: List of Sculpture
554 524 133 Naga 5 headed Naga with triangular hood and coiled body - part of 18x28 Unknown
trimurti
555 524 133 Annapurna Sitting figure of Annapurna with right knee drawn up and ladle held 18x28 Unknown
in hands across lap. Necklaces clear- part of trimurti
556 524 133 Nandi Nandi with small hump and horns. Coiled decoration around back - 24x18x30 IV+
bridle and bell decoration
557 525 134 Hanuman Short standing Hanuman with hands in front of face, and Gada by V+
left side. Completely covered in sindūr
558 525 134 Ganesh Sitting Ganesh, completely obscured by Sindūr but appears to be 4 27x29 Unknown
armed, and has left knee drawn up - part of trimurti
559 525 134 Annapurna Sitting Annapurna with left knee drawn up and ladle in hands 23x16 Unknown
across lap. Decorated with necklaces and bangles. Large ears with
rounded flower style earrings. Hair seems to be in bun on head -
part of trimurti
560 525 134 Naga Naga with tightly coiled body and 5 heads- part of trimurti 23x17 Unknown
561 526 135 Bagh Animal sculpture possibly tiger or lion, sat on haunches, but front 50x20x30 Unknown
legs missing. Rounded head, though seems snout/mouth is still
visible. Tail going up back
562 526 135 Naga-Kal Naga stone with tightly coiled body and short wide hood for 5 Unknown
heads
563 207 119 Hanuman Fairly large standing Hanuman with left foot raised up and right V+
hand raised up. Tail overhead. Covered in Sindūr
564 207 119 Hanuman Standing Hanuman, seems to have tail overhead to right side of V+
body not left. Covered in Sindūr
565 207 119 Viṣṇu Pada Eroded sandstone Viṣṇu Pada with run off area for libations Unknown
850
Appendix 9: List of Sculpture
566 207 119 Annapurna Sitting figure of Annapurna with right leg in front of body and left Unknown
one drawn up, ladle horizontal in hands across laps. Appears to
have a headdress
567 207 119 Nandi Possibly quite modern Nandi, legs tucked under body, belled VI?
twisted bridle and decorated cloth on back, small hump and pointed
horns
568 207 119 Ganesh Modern looking Ganesh, sitting with left knee drawn up, 4 armed VI?
and lower hands resting on knees
569 554 150 Hanuman Standing Hanuman with hands in front of body, features completely V+
covered by sindūr
570 554 150 Nandi Small Nandi with raised curved front leg, small hump and crossed Unknown
bridle over back
571 554 150 Ganesh Sitting Ganesh with left leg drawn up, 4 arms and a headdress - part Unknown
of trimurti
572 554 150 Naga Heavily twisted body with 5 heads very well defined. Body has lots Unknown
of detail with scales, definition and ridges - part of trimurti
573 554 150 Annapurna Annapurna with left knee drawn up, ladle in hands across lap. Not Unknown
very decorated but large long ears with earrings - part of trimurti
574 531 139 Hanuman Standing Hanuman with hands in front of body, gada under left arm V+
and tail curled over head
575 531 139 Nandi Small Nandi, crudely carved with little decoration Unknown
576 531 139 Liṅga/Yoni Round pīṭha with pointed projection and small liṅga which may not Unknown
be original, wider at top and slimmer at base
851
Appendix 9: List of Sculpture
577 532 140 Hanuman Standing Hanuman, left foot raised, right hand raised, left hand in VI?
front of body. Appears modern and also covered in Sindūr
578 533 141 ? Standing figure with round flat face, covered in material Unknown
579 533 141 ? Small standing figure with round flat face, covered in material - Unknown
folk depiction
580 533 141 ? Small sitting figure with archway design behind, large earrings and Unknown
headdress visible but rest covered up - folk depiction?
581 534 142 Naga Possibly modern Naga stone with figure of 8 body and tall slim VI?
hood with 5 heads
582 535 184 Naga Top portion of a metal snake with large triangular hood Unknown
583 535 184 Naga Small hood of a metal Naga with one head Unknown
584 536 185 ? Very eroded and obscured sculpture hard to distinguish or identify Unknown
585 538 143 Hanuman Large, probably modern sculpture of Hanuman striding with left VI?
foot resting on demon. Left hand in front of body with right hand
raised and tail over head
586 539 144 Hanuman Standing Hanuman with hands in front of body and Gada by left V+
side, tail over head
587 540 145 Hanuman Striding Hanuman with left foot raised, left hand in front of body 110x50 V+
and right hand raised, tail possibly over head
588 540 145 Hanuman Small kneeling Hanuman, probably modern with large Gada over 20x10 VI?
right shoulder
852
Appendix 9: List of Sculpture
589 540 145 Hanuman Small kneeling Hanuman, probably modern with large Gada over VI?
right shoulder
590 540 145 Hanuman Striding small Hanuman, heavily decorated but also quite obscured V+
by sindūr
591 540 145 Ganesh Sitting Ganesh with right leg drawn up, 4 armed - part of trimurti 23x17 Unknown
592 540 145 Annapurna Sitting Annapurna with right leg drawn up and ladle in hands across 23x17 Unknown
lap - part of trimurti
593 540 145 Naga Naga with coiled body and single head - part of trimurti 23x16 Unknown
594 540 145 Nandi Small simple Nandi, covered in sindūr so hard to define 25x14x32 Unknown
595 541 146 Nandi Small Nandi with raised head, no decorations, and small hump Unknown
596 541 146 Naga Naga with coiled body and 5 heads Unknown
597 541 146 Annapurna Sitting figure of Annapurna with right leg drawn up and ladle in Unknown
hands across lap. The sculpture is very obscured by sindūr
598 541 146 Ganesh Small sitting Ganesh with presumably 4 arms but very obscured Unknown
with sindūr
599 542 147 Nandi Nandi with feet tucked underneath body and belled decoration all Unknown
over body and around neck. Small pointy hump and small pointy
horns
600 542 147 Liṅga/Yoni Round pīṭha with quite long projection and side bits. Liṅga may be Unknown
original, Liṅga is on round base
853
Appendix 9: List of Sculpture
601 542 147 Ganesh Sitting Ganesh with left leg in front of body and right bent, lower Unknown
hands resting on knees, back arms holding attributes. Headdress
and jewellery clear - part of trimurti
602 542 147 Annapurna Late or possibly modern sitting figure of Annapurna on a lotus Unknown
pedestal. Quite heavily decorated with necklaces, anklets and
bracelets. Headdress and earrings visible. Ladle in hands across lap
- part of trimurti
603 542 147 Naga Naga on a pedestal with twisted body looped around neck and small Unknown
5 headed hood - part of trimurti
604 542 147 Hanuman Standing Hanuman with hands in front of body and gada by left V+
side, tail is probably over head
605 544 149 Hanuman Standing Hanuman with hands in front of body, large gada on left V+
side and presumably end of tail hanging down to the right
606 544 149 Nandi Quite eroded Nandi, with legs tucked under body and bells visible Unknown
around neck and back. Small pointed hump and horns
607 544 149 Annapurna Sitting figure of Annapurna with right leg drawn up and ladle Unknown
across lap
608 544 149 Ganesh Sitting Ganesh with 4 arms, covered in Sindūr Unknown
609 544 149 Naga-Kal Naga carved with knotted body that curves around neck, straight up Unknown
hood with 5 heads
610 546 301 Naga-Kal Probably modern Naga with figure of 8 body and large 5 headed Unknown
hood
611 545 300 Architectural Architectural piece with two figures carved on: right figure has Unknown
raised right arm. Left has a strange bottom portion of torso, almost
looks curved, perhaps kneeling
612 545 300 Nandi Very eroded Nandi, quite rectangular in shape, prominent hump Unknown
and some clear decorative features. Underneath head there is still a
portion of stone - crude depiction
854
Appendix 9: List of Sculpture
613 549 314 Naga-Kal Naga carved on stone, with body curled into spiral and one oval 44x33 Unknown
hood and head
614 549 314 Hanuman Striding figure of Hanuman - not complete, only torso: right hand 38x37 V+
raised and left holding gada
615 549 314 ? Standing figure, very crude design with arms on waist - folk 38x33 Unknown
depiction
616 549 314 Nandi Small Nandi with slightly raised head, prominent small hump, not 17x11x29 Unknown
well preserved, bells and bridle visible
617 549 314 Nandi Larger Nandi with very square face and pointed horns, very small 22x14x33 Unknown
hump
618 549 314 Naga Very eroded and obscured but Naga with coiled body and 22x14 Unknown
presumably 5 heads based on shape of hood
619 549 314 Hanuman Standing Hanuman with large head, hands in front of body and 49x28 VI?
gada under arm. Tail just visible behind head
620 550 298 Hanuman Modern Hanuman sculpture, striding with foot resting on demon. VI?
Gada in right hand over shoulder and left hand holding up the
mountain, heavily decorated
621 550 298 Hanuman Small badly preserved striding Hanuman with left foot resting on V+
demon, Gada on right side and left hand holding up the mountain
622 551 291 Memorial stone Stone with two figures carved on, seems to be male on the left and VI?
female on right. Standing side by side with arms down. Features
fairly obscured but generally crudely carved - modern folk
depiction?
623 552 296 Liṅga/Yoni Possibly modern pīṭha with large liṅga and Nandi combined. Naga VI?
wrapped around the liṅga
855
Appendix 9: List of Sculpture
624 552 296 Hanuman Completely obscured by Sindūr but still appears to be Hanuman V+
striding to left side
625 552 296 Hanuman Completely obscured by Sindūr but still appears to be Hanuman V+
standing with hands in front of body
626 553 303 Bagh Animal sculpture resembling tiger of lion as seen elsewhere, head 24x20x12 Unknown
raised slightly and snout, mouth and ears are recognisable. Sitting
back on haunches
627 555 281 Hanuman Badly preserved Hanuman, covered in Sindūr. Standing straight V+
with hands in front of body, gada appears to hang on left side of
body
628 555 281 pīṭha Round, damaged pīṭha with pointed projection but no liṅga Unknown
629 556 282 Parvati? Probably modern sculpture in Mata Mai temple. Goddess sitting 42x28 VI?
cross legged wearing a sari, jewellery and headdress with hair
hanging behind head. 4 arms holding attributes including trident
630 556 282 ? Very small stone piece with carved figure on but badly eroded and Unknown
difficult to distinguish
631 557 284 Hanuman Standing Hanuman with hands in front of body, gada under right V+
arm and tail over head
632 557 284 Ganesh Ganesh sitting cross legged on a platform, appears to only have two Unknown
arms but hard to tell with sindūr covering - part of trimurti
633 557 284 ? Figure sitting on a platform with right leg drawn up and left in front Unknown
of body with female figure sitting on knee. 4 arms, lower right is
resting on knee, back two hold attributes and lower left behind the
female figure
856
Appendix 9: List of Sculpture
634 557 284 ? Female figure sitting cross legged on platform with hands held in Unknown
prayer position in front of body
635 557 284 ? Male figure sitting cross legged on a horse drawn chariot? 4 arms, Unknown
lower two holding objects and resting on knees, upper two raised
and also holding attributes
636 558 285 ? Interesting squatting figure with left hand in between legs and right 18x26 Unknown
hand raised and touching shoulder. The figure is leaning to the right
side of body. Some details are visible such as perhaps material
around legs and torques on top of arms. Something in between legs
and behind left hand (phallus?) Face raised, rounded - folk art
637 558 285 ? Broken piece of sculpture, definitely a figure sitting on a platform, 8x10x22 IV+
right leg is visible, left is presumably bent in front of body
638 559 286 Naga Naga with coiled body and single head, could be modern 45x34 VI?
639 559 286 Ganesh Sitting Ganesh with four arms. Right lower hand raised and left 19x28 Unknown
holding out sweetmeats. Upper two arms holding attributes
640 559 286 Annapurna Seated Annapurna with ladle across lap in hands. Figure has 20x12 Unknown
potentially got a headdress on and long ears
641 559 286 Liṅga/Yoni Small Liṅga and pīṭha liṅga is comparatively big. Yoni has pointed 11x23x36 Unknown
projection and small additions to side
642 559 286 Nandi Small Nandi with a large head, collar and basic carvings on back 26x10x21 Unknown
for decoration, and prominent hump. Legs are tucked under the
body. Pointed horns and head slightly raised
643 560 287 Nandi Small Nandi with large collar and simple carved decoration on Unknown
back. Prominent hump which is quite square in shape. Square head
with bridle and pointed horns
857
Appendix 9: List of Sculpture
644 560 287 Hanuman Possibly modern Hanuman, striding with left leg raised, foot resting VI?
on demon, Left hand in front of body and right hand raised. Tail
curled overhead
645 560 287 Naga Naga with figure of 8 body and large hood with 5 heads Unknown
646 560 287 Liṅga/Yoni Small pīṭha with projections on side and with larger liṅga - shape of Unknown
the liṅga bulges towards top
647 561 246 Hanuman Large striding Hanuman with left leg raised quite high. Left hand in VI?
front of body and right on hip. Tail curled overhead to right side of
body with curl in the end of tail
648 561 246 Liṅga/Yoni Very eroded sandstone pīṭha and liṅga, broken so only circular part Unknown
is intact
649 561 246 Naga Small sandstone Naga with figure of 8 body and tail coming out at Unknown
the bottom. Head is close to body, small hood or broken
650 561 246 Ganesh Small seated Ganesh, not well preserved, seems to be 4 armed, Unknown
right knee drawn up
651 561 246 ? Standing figure, badly eroded, right arm raised and left down but Unknown
broken. No features discernible - folk art
652 561 246 Architectural Unknown carving: two items, same shape like columns. Unknown
Architectural fragment?
653 561 246 Nandi Very eroded Nandi with no head, right front leg out and others Unknown
presumably tucked under body. Some decoration visible, but badly
preserved
654 562 247 Hanuman Fairly modern looking Hanuman, striding to left with left leg raised VI?
on demon. Left hand in front of body and right raised, tail over
head
655 562 247 ? Presumably male sitting figure covered with cloth, possibly modern VI?
and made of metal. Tall headdress
858
Appendix 9: List of Sculpture
656 562 247 ? Possibly female sitting figure with halo around back of head, Unknown
similar to male figure, same material. Also covered by cloth
657 562 247 Nandi Nandi with large body and quite small head. Decorated over back Unknown
and around neck. Pointed horns and rectangular hump. Right front
leg drawn up and others tucked under body
658 562 247 Liṅga/Yoni Large round pīṭha with short projection and small liṅga Unknown
659 562 247 Liṅga/Yoni Small pīṭha with wide bulging liṅga, quite a deep/tall pīṭha for its Unknown
size
660 563 61 Bagh Sitting animal sculpture, back on haunches, but face completely Unknown
obscured by sindūr
661 563 61 Hanuman Standing Hanuman, obscured by Sindūr, with headdress and hands V+
raised holding objects
662 564 62 Hanuman Small striding Hanuman with gada held up to right shoulder and V+
left hand raised presumably holding up the mountain. Covered in
Sindūr
663 564 62 Ganesh Small seated 4 armed Ganesh with legs crossed underneath him and Unknown
lower arms on lap
664 564 62 Nandi Small Nandi with head raised. Covered in Sindūr and badly Unknown
preserved.
665 564 62 Liṅga/Yoni Small pīṭha and short liṅga, covered in sindūr Unknown
666 565 63 Hanuman Tall modern looking Hanuman, striding with left leg raised, VI?
standing on demon, gada over right shoulder and left arm holding
up the mountain. Tail up behind body to right
667 566 64 Hanuman Standing Hanuman completely obscured by sindūr V+
859
Appendix 9: List of Sculpture
668 567 65 Liṅga Sandstone liṅga, small shaft on tall square base. No discernible 100 III
markings. Looks to be old and could be from the period of interest,
although without markings it is hard to date
669 568 66 Hanuman Striding Hanuman with left foot raised, left hand in front of body, V+
and right hand raised. Tail clearly overhead. Seems to have a
headdress
670 568 66 Viṣṇu? Standing figure in decorated border. Figure has 4 arms: the lower 41x28 IV+
two are clear with left hand raised and right palm facing front and
pointed down. The back two blend into the border, and appear to be
holding flowers. Figure is highly decorated with necklaces and
jewellery, and a tall headdress. Robed on lower half. Vishnu?
Unusual depiction of hands as they cross the frame. Flowers may
indicate Surya
671 569 67 Hanuman Standing Hanuman with hands held in front of body, gada under V+
left arm. Very covered in sindūr
672 569 67 Hanuman Small standing figure, possibly Hanuman with hands in front of V+
body
673 569 67 Ganesh Sitting Ganesh with left leg drawn up, 4 arms, lower 2 holding Unknown
attributes and upper two seem to be part of border of sculpture
674 569 67 ? Two standing figures flanking the Ganesh, mirror images of each Unknown
other, standing females in oval doorway shaped border, each with
an arm raised (inner arm to Ganesh), other hand held in front of
body. Seem to have long ears but very obscured by sindūr
675 571 69 Ganesh Small sitting Ganesh with 4 arms but may be modern? Covered in 30x20 Unknown
sindūr
860
Appendix 9: List of Sculpture
676 571 69 Śiva Sculpture painted blue, appears to be Śiva. Two armed, sitting with 46x26 Unknown
left leg drawn across body. Right hand held up. Snake over left
shoulder. Hair piled on head, appears matted as ascetic. Face and
figure looks like the sitting figure at GPS 558
677 571 69 Devotee Sitting figure, cross legged with hands held in front as in prayer. 15x26 IV+
Halo behind head, and hair in bun on top of head. Appears to be a
divine devotee or saint but there is no attached iconography
678 571 69 Liṅga/Yoni Round pīṭha with small projection and small liṅga ?x49x65 Unknown
679 571 69 Nandi Small eroded Nandi, long body but no head, quite prominent hump 22x12x19 Unknown
but badly preserved
680 571 69 Nandi Small Nandi with clear bridle and decoration, and hump but badly Unknown
preserved
681 571 69 Hanuman Striding Hanuman? Standing with foot raised. Right hand up and 120x60 V+
left in front of body. Seems to have headdress, could be modern but
very covered in sindūr
682 571 69 Ganesh Small Ganesh, trunk in front of body, arms and legs visible but the 40x21 Unknown
whole figure is obscured in sindūr
683 571 69 ? 2 standing figures, very crude and probably modern VI?
684 571 69 Nandi Large Nandi, painted blue, very crude with legs carved on, large 45x22x33 VI?
collar and hump. Looks like a modern imitation of a Nandi
861
Appendix 9: List of Sculpture
685 571 69 Maitraka? Lower portion of a seated female figure. Clearly female with 30x18 IV+
necklace hanging between breasts. Right leg visible but left either
broken away or quite far behind to the left. Left hand very clear and
resting on or holding a stone between legs. Decoration around
waist/top of clothing. A dancing maitraka that would have been part
of a group. The hand may belong to another figure. Fairly old but
post-Vākāṭaka
686 194 393 Hanuman Second Hanuman figure at Japala temple. Standing with hands in V+
front of body
687 199 113 Hanuman Standing Hanuman with hands in front of body, tail overhead and V+
gada underneath left arm
688 417 115 Hanuman Striding Hanuman figure with foot on demon V+
689 575 37 Hanuman Striding Hanuman with foot on demon and tail curved overhead V+
with curl in the end. Holding gada in right hand, face obscured
690 575 37 Hanuman Broken image of Hanuman in crude folk style: Standing with left V+
foot on demon, and right hand raised. Large ears with earrings
691 574 34 Nandi Ornate medium-sized Nandi with decorated bridle and pointed 80x34x65 Unknown
hump. Sitting facing the temple with left front leg up and back legs
tucked to right side
692 580 117 Nandi Large Nandi, quite short in length but tall with straight neck. Small 56x23x68 Unknown
hump on back and legs tucked under body. Belled bridle visible
693 580 117 Ganesh Sitting Ganesh with trunk curled to left side. Left knee raised with 36x23 Unknown
hand resting on it. Two armed
862
Appendix 9: List of Sculpture
694 580 117 Annapurna Small image of figure sat with ladle across lap, Annapurna although 34x18 Unknown
not noticeably female. Large ears, quite crude carving - folk
depiction
695 580 117 Naga 5 headed fairly modern looking Naga with body twisted around 28x18 VI?
neck
696 99 Jina Sculpture kept in the Jain complex at Rāmṭek, possibly with Jain IV+
affiliation? Figure sitting in meditative pose flanked by other
smaller figures. Looks to be part of a scene or architectural detail
rather than standalone piece
863
Appendix 10: Sculpture images
This appendix presents selected images of sculptures recorded during survey. Those
too eroded or obscured by sindūr have been omitted.
9 10 11
12 15 16
864
Appendix 10: Sculpture images
21 22 23
24 25 26
27 28 29
865
Appendix 10: Sculpture images
33 34 35
36 39 40
41 42 43
866
Appendix 10: Sculpture images
51 52 53
54 56 57
58 59 60
867
Appendix 10: Sculpture images
70 71 77
78 81 82
83 84 86
868
Appendix 10: Sculpture images
91 92 95
96 98 104
869
Appendix 10: Sculpture images
870
Appendix 10: Sculpture images
871
Appendix 10: Sculpture images
872
Appendix 10: Sculpture images
873
Appendix 10: Sculpture images
874
Appendix 10: Sculpture images
875
Appendix 10: Sculpture images
876
Appendix 10: Sculpture images
877
Appendix 10: Sculpture images
878
Appendix 10: Sculpture images
879
Appendix 10: Sculpture images
880
Appendix 10: Sculpture images
881
Appendix 10: Sculpture images
882
Appendix 10: Sculpture images
883
Appendix 10: Sculpture images
884
Appendix 10: Sculpture images
885
Appendix 10: Sculpture images
886
Appendix 10: Sculpture images
887
Appendix 10: Sculpture images
888
Appendix 10: Sculpture images
889
Appendix 10: Sculpture images
890
Appendix 10: Sculpture images
891
Appendix 10: Sculpture images
892
Appendix 10: Sculpture images
893
Appendix 10: Sculpture images
894
Appendix 10: Sculpture images
895
Appendix 10: Sculpture images
896
Appendix 10: Sculpture images
897
Appendix 10: Sculpture images
898