Proceedings

44 International Congress of Americanists·

Congreso Internacional de Americanistas
Manchester 1982

General Editor: Nortnan Hamtnond

Calendar s in Mesoam erica and Peru

Native American computa tions
of time
edited by
Anthony F. Aveni
and

Gordon Brothersto n

BAR International Series 174

1983
 B.A.R.
122 Banbury Road, Oxford OX2 7BP, England

GENERAL EDITORS

A. R . Hands , B.Sc., M.A., D.Phil.

D.R. Walker. M.A.

B.A.R. - S 174, 1983: 'Calenda rs in Mesoamerica and Peru: Na ti ve American

computations of time' .

© The Individual Authors, 1983.

Price £ 15. oo post free throughout the world. Payments made in

currency other than sterling must be calculated at the current rate
of exchange. Cheques should be made payable to B .A .R. and sent
to the above address.

ISBN 0 86054 223 8

For details of all BAR publications in print please write to the

above address. Information on new titles is sent regularly on
request, with no obligation to purchase.

Volumes are distributed direct from the publisher. All BAR prices
are inclusive of postage by surface mail anywhere in the world.

Printed in Great Britain

Contents

List of contributors V

Introductory note vii

1 The base of the Venus tables of the Dresden Codex, and

its significance for the calendar-correlation problem

Floyd G. Lounsbury 1

2 The Grolier Codex: a preliminary report on the content

and authenticity of a 13th-century Maya Venus almanac

John B. Carlson 27

3 Quichean Time Philosophy

Barbara Tedlock 59

4 Fechamiento arqueoastron6mico en el altiplano de Mexico

Arturo Ponce de Leon H. 73

5 The structure of the Zapotec calendar

Gordon Whittaker 101

6 0bservaciones del sol y calendario agricola en Mesoamerica

Franz Tichy 135

7 Ciclos agricolas en el culto: un problema de la

correlacion del calendario mexica

Johanna Broda 145

8 The year 3113 BC and the fifth Sun of Mesoamerica: an

orthodox reading of the Tepexic Annals

Gordon Brotherston 167

9 Archeoastronomical fieldwork on the coast of Peru

Gary Urton and Anthony Aveni 221

10 Towards a General Andean star calendar in ancient Peru

R.T. Zuidema 235

Appendix 263

iii
 ACKNOWLEDGEMENTS

Thanks are due to Dorothy Gibson, Denise Jackson, Alison Bower,

Jean Poynter and above all Geoffrey West for their help in prepa ring
the typescript of this volume, and to Tony Young for help with the
figures. Financial and material assistance was provided by the
Publications Fund of the 44th International Congress of
Americanists, and by the Department of Literature, University of
Essex.

iv
 LIST OF CONTRIBUTORS

Anthony F. Aveni, Charles A. Dana Professor of Astronomy and

Anthropology, Colgate University, Hamilton, New York, USA.

Dr Johanna Broda de Casas, Instituto de Investigaciones Hist6r icas,

Universidad Nacional Autonoma de Mexico, Mexico 20 D.F.

Gordon Brotherston, Professor of Literature, University of Es sex,

Colchester, England.

John B. Carlson, Director, Center for Archeoastronomy, Unive rsity

of Maryland, College Park, Maryland, USA.

Floyd G. Lounsbury, Professor Emeritus, Department of Anthropo logy,

Yale University, New Haven, Connecticut, USA.

Arturo Ponce de Le6n, Arquitecto, Col. Country Club, Mexico 21 D.F.

Dr Barbara Tedlock, Director of American Studies, Asso ciate

Prof es so r of Anthropology, Tufts Uni ve rs i ty, Medfo rd,
Massachusetts, USA.

Professor Dr. Franz Tichy, Institut fiir Geographie, Univers itat

Erlangen-Nilrnberg, Kochstr. 4, D 8520 Erlangen.

Dr Gary Urton, Assistant Professor of Anthropology, Colgate

University, Hamilton, New York, USA.

Dr Gordon Whittaker, Volkerkiindliches Institut, Universi tat

Tilbingen, D 74 Tlibingen.

R.T. Zuidema, Department of Anthropology, University of Il linois,

Urbana, Illinois, USA.

V
 INTRODUCTORY NOTE

The papers in this volume have all arisen from the Sympos ium
"Calendars and Chronology" which the editors chaired at the 44th
International Congress of Americanists, in Manchester 1982.
Between them, these contributions represent or refer to the main
current developments in the study of native chronology both
Mesoamerican and Peruvian (or Inca). Their approaches include the
archeological considerations of site and structure alignments (Urton
and Aveni, Ponce de Leon); the elucidation of native texts both in
their original script (Lounsbury and Carlson on the hierogly phic,
and Whittaker and Brotherston on the iconographic writin g of
Mesoamerica) and as transcribed into the alphabet and related
colonial sources (Broda, Zuidema); and enquiry into the practices
and science of American Indians living today (Tedlock, Tichy).

The problems addressed are chiefly those of ascertainin g the

sources of perceived periodicity in astronomical and other natural
phenomena, as well as the more strictly calendrical issue of how
such phases and cycles were integrated into the numerically
consistent patterns of economics, as in tribute collection, and of
shamanist religion, as in the 26O-unit arithmogram of Mesoamerica
known as the tonalamatl (whose ultimate origins can be found in the
nine moons that elapse between the first missed menses and birth).
In the Mesoamerican area particular attention is paid to the long-
standing debate over whether the calendrical year was the seasonal
(tropical) year of agricultural tribute or the unvarying 365 -day
year intercalated into the tun calendar of the Maya lowlands, most
papers favouring the former; and to the possibility that the
Thompson Era date of 3113 BC is valid for the year and tun calendars
alike, the inaugural "4 Movement" sign invoked by the former being
the spring equinox of that year. As regards comparisons betwee n
the otherwise quite distinct societies of Mesoamerica and the Inca
Tahuantinsuyu , parallels emerge in customs associated with the
equinoxes and solstices and in the imp6rtance of both the sidereal
and the synodic moon for the administration of tribute.

Two further papers related to the Symposium are not reproduced

in full here. The first, by the late Rafael Girard, is very
substantially quoted and drawn upon in Tichy's paper; the othe r,
by Manuel Alvarado, is summarized in the Abstracts and Sympos ia
reports published by the Congress. In order to tackle the thorny
question of leap-day intercalation in Mesoamerica, Al var ado turns
to the pre-Positivist tradition of Sahagun and the cleric Fab rega
and proposes multiple readings of such native sources as the
opening chapter of Borgia (pp. 1-8) and Madrid (pp. 77-78); h is
transcription of the latter is given in the appendix. In all, the
papers presented at the Symposium and the discussion that attended

vii
them may be hoped to reveal a little more of the rich intellige nce
responsible for the native American computation of time.

AA
GB

1 IX 1983

viii
FLOYD G. LOUNSBURY

The Base of the Venus Table of the Dresden Codex, and its
Significance for the Calendar-Correlation Problem

The longcount date that appears in the preface to the Venus table of
the Dresden Codex poses a problem that up to the present has resisted
satisfactory solution. So also does one of the intervals tabulated in the
fourth tier of numbers on that same page. These problems, which are
related, bear on the astronomical circumstances and the chronology of the ~
Venus table, and indeed on Mayan chronology in general; for the
interpretation accorded to them affects crucially the solution of the
Maya-to-European calendar-correlation problem. A resolution of the Venus
problems is now at hand; and this opportunity is taken to present it.
A brief review of the context precedes.

Background Information

A perpetual Venus calendar is presented on five consecutive pages of

the Dresden Codex (pages 46-50 in traditional pagination, 25-29 in
corrected order). A sixth page that is prior to these (page 24) contains
prefatory material, including the problematic items referred to above.
These two parts ' of the Venus table will be referred to respectively as the
'main tab le ' and the 'preface'. (See Tables 1 and 2.)

In the main table the successive periods between expected first and
last appearances of the morning star and first and last of the evening star
are schematized as 236 days (morning star), 90 days (invisibility before
and after superior conjunction), 250 days (evening star), and 8 days
(invisibility at inferior conjunction), accounting thus for the 584 days of
the mean synodic period of the planet. A sequence of five synodic periods,
presented on the five pages of the table, constitutes the cycle that harmo-
nizes the 584-day Venus calendar with the 365-day annual calendar or haab
[5 x 584 = 8 x 365, = 2920]. Thirteen repetitions of the scheme, on as
many lines of these five pages, harmonize it in turn with the running of
the 260-day sacred almanac, or tzolkin, uniting all of these into a great
cycle of length equal to two of the 52-year calendar rounds [13 x 2920 =
146 x 260, = 104 x 365, = 65 x 584, = 37960, = 2 CR]. The thirteen lines
of almanac days are matched by three alternative lines of days from the
annual calendar, the latter in effect turning the entire scheme into three
separate tables of calendar-round days. It was shown by Teeple (1926,
1930) that the three tables were designed as replacements one for another,
and that yet a fourth (the first in the series of replacements) is
specified in the preface to the table. Moreover, the rule for providing
additional lines of days from the annual calendar is clear, so that the
useful life of the table is extendible indefinitely into the future. A
transcription of the calendrical and numerical content of this table is

1
Table 1. The calendrical content of the Main Table of the Dresden Codex Venus
pages. Horizontal blocks correspond to pages of the Codex (pp. 46-50).
Reading order is preserved, with horizontal lines substituting for vertical
columns. Noncalendrical portions are omitted. The 'base' positions of the
table (last line) are in boldface.

46:
3 11 6 1 9 4 12 7 2 10 5 13 8 CIB 4Yaxkin 236 9 Zac 19 Kayab 236
2 10 5 13 8 3 11 6 1 9 4 12 7 CIMI 14 Zac 327 19 M.ian 4 Zotz 90
5 13 8 3 11 6 1 9 4 12 7 2 10 CIB 19 Tzec 576 4 Yax 14 Pax 250
13 8 3 11 6 1 9 4 12 7 2 10 5 KAN 7 Xul 584 12 Yax 2 Kayab 8
47:
2 10 5 13 8 3 11 6 1 9 4 12 7 AHAU 3 Omtru 820 3 Zotz 13 Yax 236
1 9 4 12 7 2 10 5 13 8 3 11 6 oc 8 Zotz 910 13 M::>l 3 M.Jan 90
4 12 7 2 10 5 13 8 3 11 6 1 9 AHAU 18 Pax 1160 18 Uo 8 Olen 250
12 7 2 10 5 13 8 3 11 6 1 9 4 I.AMAT 6 Kayab 1168 6 Zip 16 Olen 8
48:
1 9 4 12 7 2 10 5 13 8 3 11 6 KAN 17 Yax 1404 2 M.ian 7 Zip 236
13 8 3 11 6 1 9 4 12 7 2 10 5 IX 7 M.ian 1494 7 Pop 17 Yaxkin 90
3 11 6 1 9 4 12 7 2 10 5 13 8 KAN 12 Chen 1744 17 Mac 2 Uo 250
11 6 1 9 4 12 7 2 10 5 13 8 3 EB 0 Yax 1752 5 Kankin 10 Uo 8
49:
13 8 3 11 6 1 9 4 12 7 2 10 5 I.AMAT 11 Zip 1988 16 Yaxkin 6 Kankin 236
12 7 2 10 5 13 8 3 11 6 1 9 4 E1ZNAB 1 M::>1 2078 6 C,eh 16 Clmnu 90

2 10 5 13 8 3 11 6 1 9 4 12 7 I.AMAT 6 Uo 2328 11 Xul 1 Mac 250

10 5 13 8 3 11 6 1 9 4 12 7 2 CIB 14 Uo 2336 19 Xul 9 Mac 8

50:
12 7 2 10 5 13 8 3 11 6 1 9 4 EB 10 Kankin 2572 15 Qmnu 0 Yaxkin 236

11 6 1 9 4 12 7 2 10 5 13 8 3 IK 0 Uayeb 2662 0 Tzec 10 Zac 90

1 9 4 12 7 2 10 5 13 8 3 11 6 EB 5 Mac 2912 10 Kayab 15 Tzec 250

9 4 12 7 2 10 5 13 8 3 11 6 1 ABAD 13 Mac 2920 18 ltayab 3 Iol 8

2
Table 2. Numerical and calendrical content of the Preface to the Venus
Table of the Dresden Codex (Dr. 24).

[l Ahau] [18 Kayab] [4 Ahau] 1.1.1.14.0 15.16.6.0 10.10.16.0 5.5.8.0

8 o.mtru l Ahau l Ahau 1 Ahau 1 Ahau

1.5.14.4.0 9.11.7.0 4.12.8.0 1.5.5.0

l Ahau l Ahau 1 Ahau l Ahau

4.17.6.0 4.9.4.0 4.1. 2.0 3.13.0.0

9 9 6 Ahau 11 Ahau 3 Ahau 8 Ahau
9 9
6 16 9 3.4. 16.0 2.16.14.0 2.8.12.0 2.0.10.0
2 0 16 13 Ahau 5 Ahau 10 Ahau 2 Ahau
-0- 0 0
4 Ahau 1 Ahau 1 Ahau 1.12.[8].0 1.4.6.0 16.4.0 8.2.0
8 Clmhu 18 Kayab (18 Uo) 7 Ahau 12 Ahau 4 Ahau 9 Ahau

NOIE: Bracketed items are restorations (top left) and an emendation (the uinals digit
in 1.12.8.0). The parenthesized item (18 Uo) is arithmetically out of place, due
apparently to lack of alternative space. Hyphens flanking the zero of 6.2.0 (lower
left) are for the ring of the 'ring rumber', marking it as negative. The colunms of
three dots are in lieu of noncalendrical hieroglyphs.

Table 3. Chronology of the bases of the Dresden Codex Venus Table.

(a) Mayan chronology, as determined by the intervals recorded in the
Preface to the table. (b) Hypothetic Julian-Christian equivalents
according to the Thompson correlation, JDN = MDN + 584285.

-6. 2.0, 1 Ahau 18 Kayab (0). -3119 Aug 30.

-6. 2.0 + 9. 9.16. 0.0 = 9. 9. 9.16.0, 1 Ahau 18 Kayab (A). 623 Feb 6.
9. 9. 9.16.0 + 5. 5. 8.0 = 9.14.15. 6.0, 1 Ahau 18 Kayab (B). 727 Jan 11.
" " " " " + 10.10.16.0 = 10. o. 0.14.0, 1 Ahau 18 Kayab (C). 830 ~c 16.

" " " " " + 15.16. 6.0 = 10. 5. 6. 4.0, 1 Ahau 18 Kayab (D). 934 Nov 20.

" " " " " + l. 1. 1.14.0 = 10.10.11.12. o, 1 Ahau 18 Kayab (E). 1038 Oct 25.
" " " " " + l. 5.14. 4.0 = 10.15. 4. 2.0, 1 Ahau 18 Uo (F). 1129 ~c 6.
10.10.11.12. 0 + 4.12. 8.0 = " " " "" " " " " ("). " " "
" " " "" + 9.11. 7.0 = 11. o. 3. 1.0, 1 Ahau 13 Mac (G). 1227 Jtm 15.
" " " "" + [14.10. 6.0] = 11. 5. 2. o.o, 1 Ahau 3 Xul (H). 1324 ~c 22.

3
given here in Table 1.

The day from which any one such version of the table proceeds is
conventionally known as its 'base'. This is its 'Day Zero' (not 'Day
One'). The table is so constructed that its base is the day of an expected
heliacal rising of Venus as morning star, four days after inferior conjunc-
tion, coinciding with an almanac day 1 Ahau. Its concluding day, when it
is allowed to be run through to the end, is necessarily on the same
calendar-round day as is its base; but it is two calendar rounds later in
the longcount. It too should schedule a heliacal rising of the morning
star on a day 1 Ahau; but because of a small discrepancy between the true
mean synodic period of Venus (583.92 days) and the whole-number approxima-
tion employed by the Maya (584 days), the planet must necessarily tend to
appear a few days ahead of time (about five days on the average) toward the
end of the great cycle of the table. After initial experiments, and
apparently only after this fact had become fully appreciated, the indicated
practice was to stop short of a complete run of the table, at an earlier
date where a 1 Ahau could be found that was closer to a heliacal rising of
the morning star than would have been the one at the end of the table.
This better alternative would then serve as a base (a new epoch) for the
next run of the table. Locating the better alternative could be done by
rule (once the rule was discovered): stop six and a half tuns short of a
complete run in order to locate a 1 Ahau that will offer a four-day
correction, or thirteen tuns short for one that will offer an eight-day
correction. The almanac day 1 Ahau appears to have been inviolate for
epochs of the morning star (it had a mythological charter); and with this
constraint, the corrections can come only in four-day modules and at six-
and-a-half-tun intervals. Successive bases are thus fixed in the almanac
but are movable in the year. The primary base of the table is a day 1 Ahau
18 Kayab. This is the calendar-round day with which the Dresden Codex
system began. Replacement bases are indicated at 1 Ahau 18 Uo, 1 Ahau 13
Mac, and 1 Ahau 3 Xul, in that order of succession, respectively effecting
an eight-day and two four-day corrections.

The Problem

The chronological pos1.t1.on of the 1 Ahau 18 Kayab base of the main

table is not directly indicated in that table. Neither are those of the
replacement bases, although they can be located relative to the primary
base, and to each other, at successive intervals of 4.12.8.0, 4.18.17.0,
and 4.18.17.0 (a doubly and two singly foreshortened runs of the table).
The preface, however, highlights a day 1 Ahau 18 Kayab which it places .at
9.9.9.16.0. The most natural assumption would be--and has usually been--
that this must designate the primary base of the main table. The trouble
with that, as Eric Thompson expressed it (1950: 226), is that "in no
correlation so far suggested, which is not derived solely from astronomical
data, does 9.9.9.16.0 coincide with a heliacal rising of Venus after
inferior conjunction." Correlations can of course be coined ad hoc to
accomplish this end when other pertinent evidence is ignored; but those
that have taken historical evidence into account are not able to achieve
it. Thompson noted that by his correlation (which is derived from post-
Conquest historical data) the prediction is about sixteen days too early
for the event. He therefore opted for an alternative suggested by Teeple
(1930: 97-98), which was to identify the base of the main table with a
1 Ahau 18 Kayab that was eight calendar rounds later, at 10.10.11.12.0.
4
Justification for entertaining this as a possibility was found in the
presence of the number 1.5.14.4.0 which is recorded in the fourth tier of
the preface (see Table 2). Teeple noted that "this number consists of
eight calendar rounds plus 4.12.8.0, the length of the 1 Ahau 18 Kayab
table" (i.e., as foreshortened by thirteen tuns).

It is this number that poses the other part of the problem. Its
interval can be seen as consisting of eight calendar rounds (1.1.1.14.0)
plus 4.12.8.0, i.e., of four complete runs of the main table plus one
doubly foreshortened one; or it can be seen as consisting of six calendar
rounds (15.16.6.0) plus twice 4.18.17.0, i.e., of three complete runs of
the main table plus two singly foreshortened ones. In either case it is an
interval that combines periods of uncorrected error accumulation with
another, or others, providing partial but insufficient correction in such
proportion as to leave a net residual error of about seventeen days
[l.5.14.4.0 = 185120, = 317(583.92) + 17.36].

This number, moreover, when applied to a day 1 Ahau 18 Kayab, leads to

a day 1 Ahau 18 Uo (the first of the replacement bases). But it is not the
only number specified there that does so; for so also does 4.12.8.0, which
is recorded in the same tier. Either these two numbers were intended to
apply to two different dates 1 Ahau 18 Kayab that are eight calendar rounds
apart (as are 9.9.9.16.0 and 10.10.11.12.0) and to lead .to the same 1 Ahau
18 Uo; or else they were to apply to the same 1 Ahau 18 Kayab (which one?)
and to lead to two different 1 Ahau 18 Uo dates eight calendar rounds
apart. There is nothing in the context to warrant, or even to suggest,
entertaining the second of these notions; for there is no indication of
more than one 1 Ahau 18 Uo date. Other 1 Ahau 18 Kayab dates, however, are
indicated or implied. The ring-number base in the preface records one at
minus 6.2.0, a mythical charter date seventy-two calendar rounds before the
one of 9.9.9.16.0, and thus prior to the epoch of the longcount. And the
top tier of numbers in the preface records the first four multiples of the
grand-cycle length of the main table, together with specifications that
they lead to days 1 Ahau, which in a context implying application to 1 Ahau
18 Kayab, would refer to four different resultant dates on this day at
successive two-calendar-round intervals. There are thus five 1 Ahau
18 Kayab dates in historical time indicated in the preface, in addition to
one in mythic time. Their chronology, together with that of the indicated
replacement bases, can be outlined as in Table 3, employing the intervals
recorded in the preface.

But the problem remains. What can have been the rational~ for a
number like 1.5.14.4.0, with such a seemingly inappropriate property? And
why was the date 9.9.9.16.0 given such prominence and made the focus of the
preface if it was not the intended base of the main table or even a
historical heliacal-rising date? Its discrepancy by the Thompson corre-
lation has posed for many a serious obstacle in the way of their acceptance
of that correlation. I hope now to remove that obstacle.

An Astronomical Test

In the following pages there are presented some data drawn from the
Tuckerman planetary tables, together with conclusions that appear to be
inferrible from the data. The Thompson correlation will be assumed (in its
original value: Julian day number= Maya day number+ 584285), since I am
5
convinced that it represents the truth; but it can as well be understood
merely as a working hypothesis, about to be put to a test.

The Venus table of the Codex assigns a mean value of eight days for
the duration of the planet's invisibility at inferior conjunction, without
representing the deviations from that mean for particular periods (which
would have been infeasible because of the inconstancy of their distribution
over the span of time covered by the table). The moment of conjunction
must therefore be ascribed to the middle of this eight-day period, though
it too represents a mean with deviations because of variability in the
planet's synodic period. Thus, corresponding to any one of the prescribed
days for an anticipated heliacal rising of the Morning Star, there is
assumed a mean position for inferior conjunction four days earlier.

It is of interest to see how good may be the match between predictions

from the Dresden Codex table and the astronomical computations for the same
phenomena as obtainable from planetary tables. For example, consider the
date to which Thompson ascribed the 1 Ahau 18 Kayab base of the Venus
table. In his interpretation this was eight calendar rounds after
9.9.9.16.0, which is 10.10.11.12.0, or by his correlation, A.D. 1038,
October 25 (Julian). According to the codex table, this should be a mean
position for a heliacal rising of Venus. The predicted position for the
inferior conjunction is therefore four days earlier, on 10.10.11.11.16,
10 Cib 14 Kayab, A.D. 1038, October 21. Consulting the Tuckerman tables,
however, we see that the inferior conjunction actually occurred seven days
earlier than this, on October 14. Thus, a positive error of seven days
must be ascribed to the prediction for this event, requiring a minus seven-
day correction.

The error in a single datum such as this is not by itself, however,

any reliable indication of the validity of the espoused interpretation of
the table, for the reason that the durations of individual synodic periods
of the planet deviate from the approximately 584-day mean by varying
amounts, up to three or four days either side of that mean. These
deviations fluctuate back and forth within a five-period cycle for short
terms (the 2920 days of the five-page span of any one line of the Dresden
Codex table), with a further slow circulation of their distribution over
longer periods of time. What is needed then is not a single datum such as
that just illustrated, but rather a set of five consecutive ones whose
separate errors can be averaged. When the preceding ones in this case are
determined, to make the required set of five, it is seen that their errors
were respectively +5, +4, +7, +3, and +7, in that order, to the nearest
whole day in each case. That is to say, the predictions overshoot their
respective targets by those amounts, and they require corrections of
-5, -4, -7, -3, and -7 days to bring them into accord with the real
phenomena. The mean of this set of errors is 5.2 days--a number that has
a familiar ring! [It is equal to 65 x 0.08, the accumulation of mean error
rn one complete run of the table of the codex.]

When the same test is made for the set of five inferior conjunctions
just prior to the 1 Ahau 18 Kayab of 9.9.9.16.0, which is A.D. 623 February
6 (Julian) by the Thompson correlation, it is found that the individual
errors at this time were -14, -16, -16, -14, and -17 days respectively,
again to the nearest whole day in each case. These predictions-if that is
what they were--fall short of their targets by these amounts, and would
require corrections of those magnitudes, with opposite sign, to bring them

6
into line with the phenomena. The mean of this set of errors 1.s -15.4
days.

The same test can be carried out for the inferior conjunctions prior
to each of the intermediate 1 Ahau 18 Kayab dates that are indicated in the
top tier of the prefatory table in the codex. The one next prior to
Thompson's base date is of particular interest. This is six calendar
rounds after 9.9.9.16.0, viz., 10.5.6.4.0, which by the assumed correlation
is A.D. 934, November 20 (Julian). The errors in this set of predictions
are +l, -2, +2, -1, and O, to the nearest whole day. The mean of the set
is 0.0 days.

At the two remaining indicated 1 Ahau 18 Kayab pos1.t1.ons-two and four

calendar rounds after 9.9.9.16.0--the corresponding mean errors for the
sets of five are -5.0 days (for the set just prior to 10.0.0.14.0) and
-10. 2 days (for that prior to 9.14.15.6.0).

The errors and mean errors may be determined also for the predicted
inferior-conjunction dates associated with the 1 Ahau 18 Uo, 1 Ahau 13 Mac,
and 1 Ahau 3 Xul bases , as these are distributied chronologically in the
Teeple-Thompson interpretation. These were termination dates of cycles
that were foreshortened for the purpose of error reduction, a 4680-day
foreshortening (13.0.0) accomplishing an eight-day correction in the first
of these, and two 2340-day foreshortenings (6. 9.0) for two separate
four-day corrections in the other two_ At each of these dates there are
two values of the error function: one for the prior location of the date
in the Venus table, and one for its relocation in the base position. The
two error values differ by amounts of 8 days, 4 days, and 4 days respec-
tively in the three cases, for the reasons just noted.

The data for all of these error determinations are presented in

Table 4. The table contains eight sets, with five hypothetic inferior-
conjunction dates in each, these being in every case four days prior to a
presumptive heliacal-rising date scheduled by the Dresden Codex table. The
first five sets are those which are just prior to the five 1 Ahau 18 Kayab
bases (A-E) indicated in the preface to the codex table; the final three
sets are those just prior to the 1 Ahau 18 Uo, 1 Ahau 13 Mac, and 1 Ahau
3 Xul replacement bases (F-H). The last inferior-conjunction date of each
set is designated by the same letter (from A to H) as is used in Table 3 to
label the base that follows it four days later. For example, the last date
of the first set of Table 4 is designated as inferior conjunction A
(at 9.9.9.15.16); it is four days prior to the base that is designated as
base A (9.9.9.16 .0) in Table 3. The others of the set are at 584-day
intervals (1.11.4) successively prior to that. [NOTE: The use of the
capital letters A-H to designate successive bases follows the practice
introduced by Thompson (1950: 226), except that his B, which was spurious,
is here omitted; his C therefore here becomes B, and so on to the end of
the series.]

The columns of Table 4 give the following information:

(1) Maya day numbers in longcount notation for predicted dates of

inferior conjunction, according to the chronological interpretation of the
Dresden Codex Venus table and its preface that was outlined in Table 3.
(2) The corresponding Julian day numbers, according to the Thompson
correlation: Julian day number= Maya day number+ 584285.

7
Table 4. Predicted dates of inferior conjunction of Venus, in sets of five,
preceding each of the eight successive bases indicated in the Dresden Codex
tables; together with corresponding actual dates, errors of prediction, and
mean errors per set. (Maya-to-Julian by the 584285 correlation.)
Predicted dates of inferior conjunction Actual Errors Mean
dates of of errors
Maya longcount Julian Chr. year, & inferior predic- per
day number day number Jul. cal. day conjunction tion set

9. 9. 3. 7. 0 1946305 616 Sep 10 616 Sep 24 -14

9. 9. 5. 0. 4 6889 618 Apr 17 618 May 3 -16
9. 9. 6.11. 8 7473 619 Nov 22 619 Dec 8 -16
9. 9. 8. 4.12 8057 621 Jun 28 621 Jul 12 -14
(A) 9. 9. 9.15.16 8641 623 Feb 2 623 Feb 19 -17 -15.4
9.14. 8.15. 0 1984265 720 Aug 15 720 Aug 23 -8
9.14.10. 8. 4 4849 722 Mar 22 722 Apr 3 -12
9.14.12. 1. 8 5433 723 Oct 27 723 Nov 6 -10
9.14.13.12.12 6017 725 Jun 2 725 Jun 11 -9
(B) 9.14.15. 5.16 6601 727 Jan 7 727 Jan 19 -12 -10.2
9.19.14. 5. 0 2022225 824 Jul 20 824 Jul 23 -3 ,
9.19.15.16. 4 2809 826 Feb 24 826 Mar 3 -7
9.19.17. 9. 8 3393 827 Oct 1 827 Oct 5 -4
9.19.19. 2.12 3977 829 May 7 829 May 12 -5
(C) 10. 0. 0.13.16 4561 830 Dec 12 830 Dec 18 -6 -5.0

10. 4.19.13. 0 2060185 928 Jun 24 928 Jun 23 +l

10. 5. 1. 6. 4 0769 930 Jan 29 930 Jan 31 -2
10. 5. 2.17. 8 1353 931 Sep 5 931 Sep 3 +2
10. 5. 4.10.12 1937 933 Apr 11 933 Apr 12 -1
(D) 10. 5. 6. 3.16 2521 934 Nov 16 934 Nov 16 0 0.0

10.10. .5. 3. 0 2098145 1032 May 29 1032 May 24 +5

10.10. 6.14. 4 8729 1034 Jan 3 1033 Dec 30 +4
10.10. 8. 7. 8 9313 1035 Aug 10 1035 Aug 3 +7
10.10.10. 0.12 9897 1037 Mar 16 1037 Mar 13 +3
(E) 10.10.11. 11. 16 2100481 1038 Oct 21 1038 Oct 14 +7 +5.2

10.14.17.11. 0 2131425 1123 Jul 11 1123 Jul 8 +3

10.14.19. 4. 4 2009 1125 Feb 14 1125 Feb 15 -1
10.15. 0.15. 8 2593 1126 Sep 21 1126 Sep 18 +3
10.15. 2. 8.12 3177 1128 Apr 27 1128 Apr 26 +l
(F) 10.15. 4. 1.16 3761 1129 Dec 2 1129 Dec 1 +2 +1.4

10.19.16.10. 0 2167045 1221 Jan 17 1221 Jan 17 0

10.19.18. 3. 4 7629 1222 Aug 24 1222 Aug 20 +4
10.19.19.14. 8 8213 1224 Mar 30 1224 Mar 29 +l
IL 0. 1. 7.12 8797 1225 Nov 4 1225 Nov 1 +3
(G) 11. o. 3. 0.16 9381 1227 Jun 11 1227 Jun 8 +3 +2.2

4.15. 9. 0
11. 2202665 1318 Jul 27 1318 Jul 22 +5
4.17. 2. 4
11. 3249 1320 Mar 2 1320 Mar 1 +l
4.18.13. 8
11. 3833 1321 Oct 7 1321 Oct 2 +5
5. 0. 6.12
11. 4417 1323 May 14 1323 May 11 +3
(H) 11. 5. 1.17.16 5001 1324 Dec 18 1324 Dec 16 +2 +3.2

8
 (3) The corresponding Christian year and Julian calendar dates.
(4) The actual dates of inferior conjunction, as determined from the
Tuckerman planetary tables.
(5) The errors of prediction: negative when the table predicts too
early, positive when too late.
(6) The mean errors, per set of five, corresponding to the five-period
cycle.

The error magnitudes from the last two columns are the object of our
interest. They are as follows:

Set A. -14, -16, -16, -14, -17; mean error: -15.4 days.
Set B. -8, -12, -10, -9, -12; mean error: -10.2 days.
Set c. -3, -7, --4, -5, -6; mean error: -5.0 days.
Set D. +l, -2, +2, -1, O·, mean error: 0.0 days.
Set E. +5, +4, +7, +3, +7; mean error: +5.2 days.
Set F. +3, -1, +3, +l, +1; mean reduced error: +1.4 days.
Set G. 0, +4, +l, +3, +3; rrean reduced error: +2.2 days.
Set H. +5, +l, +5, +3, +2; mean reduced error: +3.2 days.

Those listed for Sets F, G, and Hare the reduced errors, after relocation
of their respective dates to positions just prior to bases of the table.
In their original locations they are eight days greater in Set F, and four
days greater in sets G and H, their means being respectively 9.4, 6.2, and
7.2 days.

The sequence of mean errors from this tabulation, for the indicated
runs of the Venus table from 9.9.9.16.0 (the first 1 Ahau 18 Kayab) to
11.5.2.0.0 (1 Ahau 3 Xul), is graphed in Figure 1. The abscissas of the
points represent the chronology of the bases; the ordinates are the magni-
tudes of the accumulated mean errors.

That the progression of mean-error magnitudes for the first five sets
should be linear, and that the slope of their graph should be approximately
5.2 days per two calendar rounds, could of course be known in advance. It
was not necessary to extract data from the planetary tables in order to
demonstrate that. (The slight deviations from the 5.2-day module in the
cited values are merely an artifact of the use of whole-day approximations
in computing dates and individual errors.) What could not have been known
in advance was their actual values, or that their graph would cross the
zero line on exactly one of the 1 Ahau 18 Kayab dates. It was to obtain
the actual values, not their differences, that the data in Table 4 were
assembled. The results turn out to be more interesting than anticipated.
The following points deserve note:

(1) At inferior conjunction D, four days prior to base D, the error in

the prediction from the Dresden Codex table is zero. Also equal to zero at
this time is the mean error for the set of five inferior conjunctions prior
to the base. Base D, moreover, is the only one in the series of 1 Ahau 18
Kayab bases for which this is true. It is therefore the best qualified
base, and hence the most probable one, for the historical institution of
the 18 Kayab line of the Dresden Codex table.

(2) This was a unique event in historical time. For a given calendar-
round day (say 1 Ahau 18 Kayab) to recur at a given point in the astronomi-
cal Venus cycle (say four days after a mean inferior-conjunction position),

9
 Fig. 1 Mean-error graph for predicted times of inferior conjunction
of Venus:
(a) SOLID LINES, according to the prescriptions of the
Dresden Codex tables;
(b) BROKEN LINES, Thompsons's hypothetic early base
sequence.
Abscissas represent time, marked at two-calendar-round
intervals (104 years). (Abscissas of the bases are
tabulated inside the frame of the graph.)
(Julian-Christian equivalents are by the 584285 correlation)
Ordinates represent mean errors, in days, averaged for the
five successive inferior conjunctions immediately prior to
each of the specified bases.

10
9 /
I
I
8
7 "
I I /
,/
I

I
I
/ //
6 I I
I
I
I / I I
I
5 I / I
1/
,
I

4 I /
1/
II
3
2
-
e (Th)
F
0
-1
-2
-3
-4
-5
-6 Dresden Codex bases:
-7 A 9. 9. 9.16.0 (1 Ahau 18 Kayab) 623 Feb 6
B 9.14.15. 6.0 (1 Ahau 18 Kayab) 727 Jan 11
-8 C 10. o. 0.14.0 (1 Ahau 18 Kayab) 830 Dec 16
-9 D 10. 5. 6. 4.0 (1 Ahau 18 Kayab) 934 Nov 20
E 10.10.11.12.0 (1 Ahau 18 Kayab) 1038 Oct 25
-10 F 10.15. 4. 2~0 (1 Ahau 18 Uo) 1129 Dec 6
-11 G 11. o. 3. 1. 0 ( 1 Ahau 13 Mac) 1227 Jun 15
H 11. 5. 2. o.o (1 Ahau 3 Xul) 1324 Dec 22
-12 Thompson's hypothetic early bases:
-13 A2 9.10.15.16.0 (1 Ahau 8 Zac) 648 Sep 22
B 9.15.14.15.0 (1 Ahau 18 Zip) 746 Apr 1
-14 C 10. 0.13.14.0 (1 Ahau 13 Kankin) 843 Oct 9
-15 D 10. 5.12.13.0 (1 Ahau 3 Yaxkin) 941 Apr 17
.. ··- - - - -----
-16

10
 or for such a position in the Venus cycle to return again to a given day in
the calendar round, is not possible within the dimensions of historical
time. It can be approximated to within about five days (give or take a day
or two) after just two calendar rounds; but this is hardly close enough to
count as 'same point' or 'same day'. The next such approximation is not
attainable until after 111 or 113 calendar rounds (ca. 5768 or 5872 years),
at which times the deviation is once more reduced to within the plus-or-
minus five-day range. After that, it cannot happen again until after a
total of 224 calendar rounds have passed (ca. 11,640 years). Thus, if
.there was ever a heliacal rising of Venus on a day 1 Ahau 18 Kayab that
motivated the ascription that is made in the Dresden Codex, and if the
Thompson correlation is correct (or any correlation that respects Landa's
equation), then the one of base D was it. Its date was A.D. 934 November
20 (Julian), equal to Maya 10.5.6.4.0 by the Thompson correlation. There
can have been no other.

(3) The first historical run on the 18 Kayab line of the table must
have proceeded from that date. It continued through the entire length of
the table, two full cal ender rounds, reaching the next 1 Ahau 18 Kayab base
at 10.10.11.12.0, A.D. 1038 October 25; whereupon a second run, on the same
18 Kayab line, went forward from this latter base. It is apparent that the
mean error had been accumulating and that a correction was already due
before this second run began; but none was made. The most obvious and
probable reason for the failure to correct at this time is simply that the
ingenious method for accomplishing it had not yet been discovered. It
might also have been that the Maya astronomers were not yet fully convinced
that this was to be an irreversible accumulating error, rather than an
oscillating one. The invention of the correction device is thus best
ascribed to the period between bases E and F, i.e., after the last of the
1 Ahau 18 Kayab bases (10.10.11.12.0), before which it should have been
applied but wasn't, and before the 1 Ahau 18 Uo base (10.15.4.2.0), when a
double correction was applied. This places it in the interval between A.D.
1038 and 1129. Its institution is one of the matters commemorated in the
prefatory page of the Dresden Codex table.

(4) The chronology of the succeeding segments of the graph is

familiar, as it follows the Teeple-Thompson arrangement of bases E through
H (Thompson's F through I). An 18 Uo line, though absent from the main
table, is implied by recorded intervals and is explicitly noted in the
preface. Its absence from the main table is explainable if it is assumed
that it was already obsolete when the codex was composed, the current line
then being the one of 13 Mac. This would be the appropriate recension,
then, for a period beginning in A.D. 1227 (11.0.3.1.0). [In an earlier
paper (1978: 788) I conjectured that the current base was the one of 18 Uo.
That, I am now sure, was a mistake.] The retention of the 18 Kayab line,
thought it also belonged to the past, is explainable as being due to its
special status as the original of this formulation, and to its adoption as
the ultimate reference to which to relate succeeding replacements.
(Another possible function will be noted later.) Looking to the future,
the next replacement--the 3 Xul line (to take effect 11.5.2.0.0,
A.D. 1324)--was already reckoned and included in the table. It would have
been the most recent contribution to the table as it stands in the codex,
occupying a line that may have been held by 18 Uo in a previous version.

(5) The error graph over the periods prior to base D clearly reflects
an uncorrected backward extrapolation of the Venus calendar, employing the

11
whole-number approximation for its reckoning (584 days per synodic period,
or two calendar rounds per grand cycle). It cannot possibly reflect any
series of past observations; else its slope would be zero. It is thus a
computational figment, as is also the 9.9.9.16.0 hypothetic base which it
has as its initial terminus. Neither can this stretch be dispensed with in
interpretation of the table; for not only is it implied by the stated
increments of two, four, six, and eight calendar rounds in the top tier of
the prefatory table, but it is made inescapable by two of the increments
specified in the fourth tier, viz., 4.12.8.0 and 1.5.14.4.0, which derive
the 1 Ahau 18 Uo base from two different 1 Ahau 18 Kayab bases eight
calendar rounds apart. A question is thereby posed: Why should there be
an extrapolation of this particular length--six calendar rounds--rather
than of some other length? Or why one leading to 9. 9. 9.16.0, rather than
to some other equally artificial prior base? If a plausible motive for
this particular extrapolation were to be identified, the solution to the
most vexing problem in the interpretation of the Dresden Codex Venus table
would be at hand.

It is possible now to offer a reason for the extrapolation and for its
particular length, as well as for the choice of the day 1 Ahau 18 Kayab as
an epoch for Venus reckonings. The two matters are related. As for the
latter, that choice can hardly have been motivated by considerations of
frequency; for the date of base D was the only time in all of Maya history
when there could have been a heliacal rising of Venus on the morning of a
day 1 Ahau 18 Kayab, or an inferior conjunction of Venus on a day that is
four days prior to that position in the calendar round. Therefores either
it was fortuitous (i.e., for a nonastronomica l reason) that they chose
1 Ahau 18 Kayab as the primary base of the Venus table of the codex, or
else there was something special about this one.

As already noted, the date of base D by the 584285 correlation was

November 20 (Julian) of the year 934. On that date Venus and Mars came
into conjunction with each other; and that must also have been the date of
their heliacal rising, i.e., of · first sighting as morning star at the
horizon, just ahead of the sun, after their respective periods of invisi-
bility (short for Venus, considerably longer for Mars). Though the moment
of their conjunction was toward the end of the day (ca. 11:20 p.m.,
90th-meridian time), they were close to each other as they rose that
morning, with about nine-tenths of a degree between them, Mars ahead of
Venus; and by the next morning they had changed places, with about a third
of a degree between them, Venus ahead of Mars. On November 20 their
western elongations at rising time were about 6.6 degrees for Venus and
about 7.5 degrees for Mars; by the next morning, at the same time, they
were about 8.2 degrees for Venus and 7.8 degrees for Mars. Their
latitudes (from the ecliptic) were close to zero: -0.16 of a degree for
Venus, and -0.14 for Mars. Their situation with respect to each other can
be seen in the graphs of Figure 2, in which the celestial longitudes of
Venus, Mars, and Sun are plotted for this date and for some intervals
before and after. [The plotted points and the abscissas on the graph are
for 10:00 a.m. at 90 degrees west, because these are the values that can be
taken most easily from the Tuckerman tables (which are for 7:00 p.m. at 45
degrees east).] As for the question of their possible visibility just
before dawn on these dates, the following considerations are pertinent:

12
(1) The minimum necessary elongation of Venus from the sun in order to be
visible at the horizon ranges from two or three degrees to nine or ten,
depending on the time of the year (and other less regular factors). For
the third week in November it is about five and a quarter degrees.
On November 20 of the year with which we are concerned the actual elonga-
tion was about 6.6 degrees; on November 19, the morning before, it was
about 5.0 degrees; Thus, so far as this condition is concerned, it is
unlikely that it could have been sighted on the 19th, but it is both
possible and probable that it was seen on the 20th. (2) The corresponding
requirement for Mar s is greater than that for Venus, because of Mars'
lesser brilliance. It is thus not possible for it to have been seen as
a separate body on the morning of the 20th; but in the days that followed,
as the two planets separated from each other, it would sooner or later have
become apparent that what had initially been perceived as one celestial
body was in fact two, and that Mars had been in conjunction with Venus
-a fact which the Maya astronomers might have anticipated and whose
confirmation they may have been awaiting. [NOTE: Attempts to calculate
and to quantify empirically the conditions for first and last morning or
evening visibil ity of a planet or a star at the horizon have been in terms
of what is known as its arcus visionis, the necessary depression (negative
altitude) of the sun below the horizon at that moment. The concept and the
procedure go back at least to Ptolemy. See Schoch (1924), Langdon,
Fotheringham , and Schoch (1928: 49-52, 94 ff.), van der Waerden (1943),
Huber (1982: 11-14, 84-87), and Weir (1982: 40 ff.), as well as Ptolemy's
Almagest (Manitius 1963, II: 393-4). Here I have expressed it simply in
terms of the planet's elongation from the sun, which in the tropics, for
present purposes, is a sufficiently close approximation. The two values
coincide on dates of the ecliptic's zenith passage, and in Mayan latitudes
the maximum. difference between them (at the winter solstice) amounts to
something between four-tenths and eight-tenths of a degree, which
difference I have attempted to take into consideration.]

If we assume that the almanac day 1 Ahau was already mythologically

chartered as a day special to the Morning Star Venus (and there are reasons
for suspecting that this goes well back into the Classic period), then its
heliacal rising on this date, a "l Ahau", would by virtue of this fact
alone have been a marked event; for this is something that can be
approximated no oftener than once every 97 and a half years or so. And its
conjunction with Mars on this occasion would have made it doubly marked.
The Maya astronomers (their 'god watchers') could not have failed to take
note of it. The zero value of the Venus error graph on this date
implicates it as the starting date for the Dresden Codex table. The
coincidence of these facts can hardly be fortuitous.

Having Mars in the picture along with Venus has an interesting

consequence. The mean synodic period of Mars is 779.94 days. That of
Venus is 593.92 days. The error graph shows that during this time, and for
well over a century thereafter, the length of the Venus period was being
reckoned as a whole number, 584 days, without correction. So it can be
assumed that the period of Mars was also, viz., as 780 days, or three
lengths of the almanac. The lowest common multiple of these two numbers is
exactly six calendar rounds [113880 = 6 x 18980, = 146 x 780, = 195 x 584].
The error graph shows that the problematic 9.9.9.16.0 base was a six-
calendar-round extrapolation back from the indicated starting date of the
table. We have now, in the Mars-Venus relationship, a likely reason for
the magnitude of that extrapolation.
13
Fig. 2 Celestial longitudes of Venus, Sun, and Mars on November 20,
A. D. 934, and during pe riods immediately preceding and
following.

250
9
8
7
6
245
4
3
2

240
9
8
7
6
235
4
3
2

230
9
8
7
6
225
21 26 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 4 9 14
Sept. October November December January
A.D. 934 935

14
 Fig. 3 Celestial longitudes of Venus, Sun, and Jupiter on December
6, A.D. 1129, and during periods immediately preceding and
following.

7
6
265
4
3
2

260
9
8
7
6
255
4
3
2
1
250
9
8
"'s..
7 I)

6
iI)
0
&>
245 A

4
3
2
1
240
-
2 7 12 17 22 27 6 11 16 21 26 6 11 16 21 26 31 5 10 15 20 25
October November December January
A.D. 1129 1130

15
 If a separate reason be required for the extrapolation (apart from
that for its particular magnitude), and for making the fictitious base the
focal piece of the preface, it could perhaps be said that this was a good
'Maya' thing to do; in fact they didn't stop with just that, but extended
it another twelve times t hat length, setting up yet another fictitious base
on another day 1 Ahau 18 Kayab, this one before the beginning of the
current longcount era. The one at 9.9.9.16.0 is a first step in that long
projection into the past, establishing a Venus-Mars compound cycle of six
calendar rounds, after which its twelfth multiple reaches into mythic time
prior to the chronological epoch. That is the whole point of a ring-number
base; it is to find a pre-zero precedent, viz., a date that is similar in
its properties to an important historical date, and that can be reached by
a long-reckoning number that is an integral multiple of all of the
pertinent cycles. (Cf. Lounsbury 1976.)

There could also have been a practi c al use for the 9.9.9.16.0 base.
The Maya astronomers must have been in possession of a sizable collection
of records of planetary observations prior to the 1 Ahau Venus-Mars event
of 10.5.6.4.0, and prior to the codification of the Venus calendar with the
18 Kayab of that date as its base. The hypothetic base six calendar rounds
earlier would have been a useful construct. It would have permitted
forward rather than backward reckonings to the observational data. During
the period covered by perhaps the majority of the records its error would
have compensated fairly well for the other error that is contained in the
use of a table premised on whole-number reckoning.

The extrapolations may thus be understood as 'historical reconstruc-

tions' on the part of the Maya astronomers, for the purpose of fixing bench
marks: the proximate one (9.9.9.16.0) for reckonings in recent history,
and the ultimate one (-6.2.0) for determining the supposed start of it all
in the current order of the universe. And they are yet another manifesta-
tion of the Maya interest in compound cycles.

1 Ahau 18 Uo

In an earlier section, reasons were suggested as to why a corrective

foreshortening of the grand cycle and a resulting base shift were not
instituted prior to 10.10.11.12.0, when they were clearly overdue, and ·why
a second historical run of the table-carrying along an already accumulated
mean error of five days--was allowed to proceed from the 1 Ahau 18 Kayab of
that date. The proposed reasons were (1) that the ingenious method for
accomplishing the correction had not yet been discovered, and (2) that the
Maya astronomers may not yet have been fully convinced that this was to be
an irreversible and steadily accumulating error rather than an oscillating
and eventually self-correcting one. There is now another reason that
appears plausible: (3) it might well have been, because of the extraordi-
nary Mars-Venus coincidence of 10.5.6.4.0, and because of the investment in
theoretical constructs derived therefrom, that the commitment to 1 Ahau
18 Kayab bases was too strong for these easily to be set aside; it may have
been not only 1 Ahau that was sacrosanct, but 18 Kayab as well. If this
were the case, it would have taken some other extraordinary event to break
the precedent and set a new one. By one of those strokes of cosmic luck
with which the Maya astronomers seem to have been blessed (and contrary to
our sense of reasonable probabilities) such an event did take place on
1 Ahau 18 Uo, 10.15.4.2.0, thirteen tuns before the end of the second run

16
of the table on the 18 Kayab line. By the 584285 correlation, that was
December 6 (Julian) of the year 1129; and on that date there was a
reenactment of the event of 1 Ahau 18 Kayab, 10.5.6.4.0 (A.D. 934, November
20), but with Jupiter in the earlier role of Mars. Their situation is
shown in Figure 3. Western elongations at rising time were about 8.28
degrees for Venus and 7.55 for Jupiter, with latitudes of L48 degrees
north of the ecliptic for Venus and 0.38 north for Jupiter. Their moment
of conjunction in longitude was about two hours after sunrise of the
previous day, and it cannot be ruled out that they might actually have been
sighted that morning. It was a time of the year when the magnitude of the
elongation required for a first visibility of Venus was decreasing and was
between six and five degrees. Jupiter, however, like Mars, requires a
somewhat greater elongation for separate visibility; and it may again have
been some days before their separate identities were resolved (and the
anticipations of the astronomers verified). In any case, when the nature
of the situation had become clear, it would have been seen that the 1 Ahau
rising of that date was an extraordinary one. If a sign were needed to
persuade holdouts that the time had come to let go of 18 Kayab, and that it
was propitious for another 1 Ahau (which would also eliminate most of the
by now accumulated nine-day error), this could well have been it. Once
instituted, the device of base shifting could be applied thereafter as
needed, as indeed it was.

Introduction of this present datum into the picture for 1 Ahau 18 Uo

might seem to weaken the case for the Mars-Venus conjunction that was taken
to have endowed 1 Ahau 18 Kayab with special significance; for with two
such cases at hand it might be supposed that it is not too hard to find
convenient conjunctions and simultaneous heliacal risings of the planets on
suitable occasions. That impression, however, would be false. The last
time prior to A.D. 934 when Venus was in conjunction with Mars at or within
a day of heliacal rising was in A.D. 6. And the last time prior to
A.D. 1129 when Venus came close to being in conjunction with Jupiter in
approximately similar circumstance was in A.D. 244. And neither of these
was even close to a day 1 Ahau. These give an indication of the frequency
of such events. They are, for the problem at hand, 'once only'.

Some Other Chronological Interpretations

1. Thompson (1950). The chronology posited here (as in Table 3 and in

Figure 1) includes only the dates and intervals for which explicit
authority is found in the codex. It agrees only in part with Thompson's
(1950: 226); for he made some speculative backward extrapolations of his
own. The two schemes agree with each other in their placement of base A at
the beginning of the sequence, and of E, F, G, and Hat the end of it
(Thompson's F, G, H, and I); but between A and E (Thompson's F) he posited
four bases for which there is neither calendrical nor numerical evidence in
the codex, reconstructing these on the basis of an assumption that the
Venus calendar was in working order by 9.9.9.16.0 and that the device of
corrective six-and-a-half-·tun foreshortenings was then known and was being
applied , and not supposing that the series of two-calendar-round multiples
in the top tier of the preface were intended as base determiners. His
hypothetic sequence, starting with base A, was the following (relettering
his B to I as A2 to H to facilitate comparison with ours):

17
 9. 9. 9.16.0 (A) 1 Ahau 18 Kayab [ 623 Feb 6]
+ 1. 6. 0.0 = 9.10.15.16.0 (A2) 1 Ahau 8 Zac [ 648 Sep 22]
+ 4.18.17. 0 = 9.15.14.15.0 (B) 1 Ahau 18 Zip [ 746 Apr l]
+ 4.18.17 .0 = 10. 0.13.14.0 (C) 1 Ahau 13 Kankin [ 843 Oct 9]
+ 4.18.17.0 = 10. 5.12.13.0 (D) 1 Ahau 3 Yaxkin [ 941 Apr 17]
+ 4.18.17.0 = 10.10.11 . 12.0 (E) 1 Ahau 18 Kayab [1038 Oct 25]
+ 4.12. 8.0 10.15. 4. 2.0 (F) 1 Ahau 18 Uo [1129 Dec 6]
+ 4.18.17.0 = 11. o. 3. 1.0 (G) 1 Ahau 13 Mac [1227 Jun 15]
+ 4.18.17.0 = 11. 5. 2. 0.0 (H) 1 Ahau 3 Xul [1324 Dec 22]

His reconstructed bases A2 to Dare derived by backward reckonings from

base E, employing singly foreshortened runs of the table, until left with a
remainder of 1.6.0.0 between his fourth reconstruction (A2) and the 1 Ahau
18 Kayab of 9.9.9.16.0 (base A) which is highlighted in the preface. This
remainder he supposed to have been the number that was intended · for the
place actually occupied by 1.5.5.0 in the fourth ·tier of the preface, and
he ventured it as a correction of that number (Thompson 1935: 63-64; 1950:
225, 226 ).

A mean-error graph for Thompson's sequence of hypothetic bases prior

to E has been superimposed (in dotted lines) on the graph of Figure 1.
It depic ts data drawn from the Tuckerman tables in the same manner as that
of Table 3 (though not included there). The sequence misses the only
chance that there ever was for a heliacal rising of Venus on a day 1 Ahau
18 Kayab, and it provides no explanation for the large error at the initial
terminus or for the precipitous correction (overcorrection) implied by the
supposed 1.6.0.0 interval. This entire reconstructed portion of Thompson's
sequence must therefore be discarded.

Thompson noted that the final number in the table of the preface
(equal to eight calendar rounds) is the distance between the two positions
of 1 Ahau 18 Kayab in his chronological sequence; but he considered that
"it may be pure coincidence" (1950: 226), and he attached no particular
significance to this or to the preceding multiples of the two-calendar-
round interval. In the interpretation proposed here, however, these latter
have been taken to determine prior Maya-hypothetic bases, in regular
progression from the observational base, backward to the one calculated to
initiate a supposedly previous Venus-Mars cycle.

~ Teeple (1930). Teeple, in his final treatise (1930: 94-98), made

no attempt to link the two 1 Ahau 18 Kayab dates of this series. Rather,
he took them to be mutually exclusive alternatives for the primary base of
the main table. He posited therefore either the series from base E onward
as we have it here, and as Thompson had it, with the base of the table at
10.10.11.12.0, or an alternative analogous series placed eight calendar
rounds earlier, with the base at 9.9.9.16.0. One or the other of these he
thought (then, 1930) was probably correct. The error graph corresponding
to the later placement is the same as that of Figure 1 from base E to the
right, with everything to the left of E cut off; that is, provided that the
Maya chronology is interpreted by the 584285 correlation, which was the one
for which Teeple gave the best arguments and about which he seemed the
least pessimistic at that time. Like Thompson's adaptation and extension
of it, it misses completely its only chance for a heliacal rising on a day
1 Ahau 18 Kayab and its only chance for a zero error. And it offers no
explanation for the 9.9.9.16.0 date or for the problematic interval
1.5.14.4.0. The error graph for the earlier hypothetic placement, assuming
18
again the same correlation, would be the same graph as from E to H,
transposed eight calendar rounds to the left (so as to begin on the
abscissa of A) and 20.8 ordinates downward (so as to begin at -15.6). This
would place the entire graph between the ordinates -20 and -11, implying
errors in that range throughout. Errors of such magnitude are hardly to be
expected in a functioning table of this sort, so Teeple (p. 105) proposed
interpreting the Landa equation with 20 or 30 days leeway to compensate for
it, allowing for that much variation in the correlation in case this should
be the required placement of the base. [But a calendrical error of that
magnitude in Landa' s information is hardly more likely, and the need to
impute it is a cause for doubt concerning the interpretation.]

~ Teeple' s first interpretation (1926). In Teeple' s previous

treatment of this subject (1926) he had posited neither of these two
series, but instead, one that is placed two calendar rounds later than the
earlier of them, thus locating the 1 Ahau 18 Kayab base at 9.14.15.6.0 and
the others accordingly; and he reconstructed two hypothetic prior bases,
analogous to Thompson's but six calendar rounds earlier, on the assumption
that the correction procedure was already in operation. Then he proposed
two of what seemed to him at that time to be reasonable possibilities for
an astronomically derived calendar correlation (they are equivalent to the
constants 492622 and 427803, though he did not express them in those
terms). Both of these yield just about an optimum placement of the error
graph, with heliacal rising on or within a day of the 1 Ahau 18 Kayab of
9.14.15.6.0 (abscissa B of Figure 1, this representing A.D. 476 January 25
by the first correlation, or A.D. 298 August 9 by the second) and with the
range of error magnitudes nicely balanced either side of zero, from -3.8 to
+4.2 days. The hypothetic correlations, of course, were chosen expressly
to achieve that effect (along also with that of yielding correct moon
ages). But they were premised on some other considerations shown later to
be wrong, and they took no account of either of the two prime considera-
tions for a historically valid correlation, namely, the Landa equation and
the Katun 13 Ahau condition; instead, they yield results that are grossly
at variance with these. By the time he wrote his 1930 treatise he had
discarded both the chronology and the correlations of this earlier paper.
They are of interest, however, for what they illustrate about the nature of
the problem and the history of attempts to understand it.

~ · Spinden (1928), Schulz (1935), Ludendorff (1937), Dittrich (1937).

The heliacal rising of Venus on November 20, A.D. 934, four days after the
inferior conjunction of November 16, is the necessary and crucial datum for
any interpretation of the Dresden Codex table that is premised on an
ascription of its 1 Ahau 18 Kayab base to an observable astronomical event
of the prescribed kind, and on a calendar correlation that respects Landa's
equation--as do both the Thompson and the Spinden correlations. It is
ironic that this date was given no notice by either Teeple or Thompson; for
it is the only time in all of Maya history that the cycle of Venus and that
of the Maya calendar came together in the required combination. The
magnitude of the error at 9.9.9.16.0 by the Thompson correlation should
have given the clue to look exactly six calendar rounds later. Spinden, by
a fortuitous coincidence, happened onto the date and recognized its
property; and it (or rather the day before) became the anchor for his Venus
chronology. It was noted also by Schulz (1935), Ludendorff (1937), and
Dittrich (1937), and it figured similarly in their interpretations. Both
Ludendorf f (p. 89) and Dittrich (p. 349) observed that with neither the
Thompson nor the Spinden correlation did the problematic 1 Ahau 18 Kayab of

19
9.9.9.16.0 correspond to a heliacal rising of the morning star as had
generally been supposed and as was seemingly implied by its prominent
display in the preface and by the place of that calendar-round day in the
main table. And Schulz (pp. 58, 66) made a point of the fact that it would
be necessary to add six calendar rounds to it when employing the Thompson
correlation, or eleven when using the Spinden, in order to reach the only
possibility that there ever was for a 1 Ahau 18 Kayab heliacal rising,
which was that of November 20, A.D. 934, and which would be Maya 10.5.6.4.0
by the Thompson correlation, or 10.18.9.15.0 by the Spinden. This date he
took to be the starting point of the 18 Kayab line of the main table of the
Codex. Dittrich (pp. 337-9) gave an elegant derivation of the date and
proof of its uniqueness; and he showed that if 9.9.9.16.0 were considered
to be the indicated base of the 18 Kayab table, it would require a
correlation constant of 698164 (six calendar rounds greater than
Thompson's, eleven greater than Spinden's). This, however, he immediately
rejected, on the grounds that researches on the correlation problem had
already shown that there could be only two acceptable alternatives, five
calendar rounds apart, viz. 489384 (Spinden's) and 584284 (an accepted
variant of Thompson's). The value 698164 which he derived from equating
9.9.9.16.0 with A.O. 934 November 19 would agree with Landa' s equation as
well as did the Spinden and the Thompson values, but it would not satisfy
the Katun 13 Ahau condition, and it was therefore to be rejected. He
concluded that it was necessary to abandon the notion that the table of the
codex might have rested on a 1 Ahau 18 Kayab observation of the morning
star at horizon prior to A.O. 934. This meant that the 9.9.9.16.0 date
that was highlighted in the preface to the Venus table (which he took to be
the base of the main table) would require some explanation other than the
supposed observational one.

The chronology of the historical bases (D to H) as posited by these

writers was thus essentially in agreement with the one that is proposed in
this paper when interpreted in terms of the European calendar. When
interpreted in terms of Mayan chronology however, it is something quite
different. Spinden 's placements of the bases D through H (as labeled here)
are just one day earlier in Christian chronology; but they are five
calendar rounds later in the Maya longcount. A mean-error graph for his
scheme would be like ours from D through H, but with all error values (the
ordinates) reduced by a day, and with the abscissas differently interpreted
in Mayan chronology, though not in European. The Maya 9.9.9.16.0 abscissa
would be five calendar rounds further removed from base D (two and a half
blocks further to the left), and the ordinate of its 'error' at that point
would have to be found several feet off the lower edge of the page. (The
difference is an odd number of calendar rounds; which removes the date far
from any inferior conjunction or heliacal rising. It is instead about 28
days before a superior conjunction.) Spinden did not comment specifically
on base E, but its position is implied by those of D and F. For base H
he mentioned three alternatives, at intervals of two calendar rounds, the
first approximating a final heliacal setting of the evening star, the
second an inferior conjunction, and the third a heliacal rising of the
morning star. It is the last of these that is in agreement with ours, and
that conforms to the specification of a base. [The one-day differences are
due to his interpretation of the Puuc-style Maya date in the Landa equation
(12 Kan 1 Pop= A.D. 1553 July 16, O.S.) as representing Classic 12 Kan
2 Pop rather than 11 Akbal 1 Pop.]

Schulz outlined two alternative chronologies for the same sequence of

20
bases, both of them--but for a single item--agreeing with ours in their
assignments of bases to European chronology, and one of them agreeing also
in its Mayan placements; while the other agrees, again with that exception,
with Spinden's. The first of Schulz's alternatives thus represents a
historical precedent for the chronological interpretation offered in this
present paper, though the Mars conjunction with Venus on the crucial date
was not noted by him and did not figure in his determination. An unfavor-
able judgment on his part as to the likelihood of survival of knowledge of
the longcount into the period covered by the Venus table (A.D. 934-1324)
led him (p. 59) specifically to exclude the possibility that the date
9.9.9.16.0 might be explainable as a calculated backward projection from
the date of the historical base of the table, namely, A.D. 934 November 20
(10.5.6.4.0 by the Thompson correlation, 10.18.9.15.0 by the Spinden). He
took it rather to have been a forward projection from Old Empire times, one
that was preserved--for a reason unstated--for utilization in a later
period. [It would thus have to be seen as an early Classic Maya prediction
that came true only six, or eleven, calendar rounds late.] But for this
judgment and its consequence, and the absence of the Mars datum, and the
exception referred to above (namely, his placement of the 13 Mac base two
calendar rounds too early, 6.9.0 before 18 Uo!), Schulz's first alternative
would have been the same as that proposed here. As it was, however, it
lacked satisfact ory explanation of either the date 9.9.9.16.0 or the
interval 1. 5.14.4. 0.

5. Makemson 0943, 1946). In her earlier treatment of this subject

Makemson accepted the validity of the Landa year-bearer equation; took note
of Dittrich's proof of the uniqueness of the 934 November 20 identification
of a Venus heliacal rising with a Maya day 1 Ahau 18 Kayab (assuming the
Landa equation to be correct); acknowledged the four successive bases
determined by the top tier of numbers in the Venus table preface as
possibilities for its assignment; and on the basis of an assumption of
contemporanei ty of the 13 Mac line of the main Venus table (which she
showed to be the current line) with the particular rerun of the eclipse
table that would have its nodes agreeing with their locations in the table,
she determined that the 934 November 20 event must be identified with the
1 Ahau 18 Kayab of 10.5.6.4.0. This result, she noted, "may be added to
the mounting sum of confirmations of the Goodman-Thompson correlation."
She noted also that "If ••• 10.5.6.4.0 represents a heliacal rising of
Venus, it follows that 9.9.9.16.0 must place the planet 16 days before
heliacal rising, and the reverse is also true." Further, she took note of
the Mars conjunction with Venus on the 10.5.6.4.0 heliacal-rising date, and
the role that this must have had in the backward numerical projections to
9.9.9.16.0 and -6.2.0. Her results were thus in agreement with those to
which I have been led, by a somewhat different route, starting with
10.5.6.4.0 (base D) and proceeding backwards; but proceeding forwards they
disagree. The final 18 Kayab base (E) is missing in her scheme, the 18 Uo
base (F) is unaccounted for, and those that follow--13 Mac, and 3 Xul
(G and H)-fall two calendar rounds earlier. This discrepancy results from
her assumption of close contemporaneity of the Venus table and the eclipse
table, her somewhat early placement of the latter, and a different manner
of judging the currency of any given version of the table. It has the
consequence of providing no explanation for the problematic number
1.5.14.4.0 in the fourth tier of the table of the preface, which she did
not take into consideration. This, as noted earlier, is equal to the sum
of eight calendar rounds and 4.12.8.0 (i.e., four complete runs of the main
table plus one doubly foreshortened run), and it derives the 18 Uo base

21
from the 18 Kayab base of 9.9.9.16.0. In ignoring this, her placement of
the remaining bases fails to conform to the prescriptions of the preface.

That was 1943. In 1946 she abandoned all that, or at least the most
essential parts of it, and went on to propose a new correlation. A part of
the reason had to do with that very number, 1.5.14.4.0, that had been
bypassed in her earlier presentation. Though it did not figure in that
publication, it had not been absent from her thinking about the matter; for
she later wrote (1946: 58-59) that at the time she had been struck by the
curious fact that the excess of that number over 317 mean Venus periods was
exactly right for removing the discrepancy of -17.6 days between 9.9.9.16.0
and the time of heliacal rising as given by the Thompson correlation, and
that it had seemed to her then that the number must have been computed for
that very purpose. This, she noted, was equivalent to saying that
9.9.9.16.0 was not intended to represent a heliacal rising at all, but was
chosen for some other purpose. She submitted her theory to Thompson before
publication, but according to her testimony he was not convinced by her
argument either for this notion or for the Mars connection, and: ''He
maintained that the Maya date was clearly intended as the epoch of the
ephemeris and a heliacal rising of Venus." Concerning this she then
commented: "But if 9.9.9.16.0 represents a heliacal rising of Venus, a
fresh problem arises with regard to the Goodman-Thompson correlation,
[namely] how it came about that an error of -17.6 days is found, when the
cumulative error of the ephemeris is always in a positive direction." This
is yet another aspect of the problem to which Thompson made reference (as
quoted earlier in this paper) but for which he offered no solution. He
only bypassed it by adding the invented figure of 1.6.0.0 to the date in
order to find a new base from which to proceed.

Makemson was very close to an astronomically as well as logically

satisfactory resolution of the entire problem within a chronological
framework compatible with the Thompson correlation. Had she insisted on
her suspicion about the fictitious nature of the 9.9.9.16.0 date, and had
she been willing to accept also the remaining implications of the
1.5.14.4.0 interval, her conclusions and her chronology would have been the
same as those that have been proposed in this paper. Instead, by the time
she wrote her 1946 monograph she had embarked on a survey of possible
correlations such as might be deducible from astronomical evidence alone,
deciding not to be bound by the constraints of the Landa equation, about
which she had developed doubts. And she followed up her deduction that
"a correlation which represents the Maya date [9.9.9.16.0] with a positive
error should be given more weight than one which places [it] an equal
number of days before heliacal rising," as does Thompson's. In this she
was abandoning her former hypothesis, to which Thompson had objected, that
9.9.9.16.0 was a computed projection into the past, and that this accounted
for the negative error at that date. In its place she was now substituting
a hypothesis that 9.9.9.16.0 was a forward projection from a table already
in use before that date, which would then make positive errors expectable.
The correlation that she found--the only one from her survey to fulfill
that condition, along with meeting other astronomical and historical
requireme nts (not including the Landa testimony however)--ended up
differing from the Spinden correlation by just 246 days (with correlation
constant equal to 489138, Spinden's being 489384). Her further arguments
in support of this interpretation (speculations about seasonal relevance of
Maya rituals) will not be considered here, except to note that they are
wide of the mark.

22
Conclusion

After having completed the main part of this paper, in which I pre-
sented the results of my own inquiry, and turning then to a review of
previous interpretations of the problem--and after finishing with those of
Teeple and Thompson and going on to those with which I had been only
slightly if at all acquainted--I found myself in the not unfamiliar
circumstance of having discovered things that others had discovered long
before; in this case, the date A.D. 934 November 20, the uniqueness of the
combination of calendrical and astronomical events of that date, and their
signal importance. Under most such circumstances that would have left me
without a paper; but not quite this time. Although crucial pieces of the
puzzle had been located and identified earlier, their proper assembly has
remained incomplete to the present time. So also has their full documen-
tation.

Spinden was the first to discover the crucial date. His interpre-
tation for the period covered by the historical dates was essentially
correct on the astronomical and European-calendrical side, but it was five
calendar rounds off on the Mayan side of the equation. Since this is an
odd number of calendar rounds while the multiples that are compatible with
the Venus period are even, this left the important date of 9.9.9.16.0 off
in limbo before superior conjunction and without an explanation. Spinden
made only the briefest of reference to it, noting that by his correlation
it would fall (in A.D. 363) on April 12 in retroactive Gregorian dating, a
seasonal position to which he attached importance (he associated it with
the 19 Xul heliacal risings, those next prior to 18 Kayab).

Schulz, . when considering the possibilities under the Spinden correla-

tion, observed (1935: 66): ''By the Spinden correlation the date 9.9.9.16.0
relates to the superior conjunction of Venus (besides which its position in
the tropical year is of course to be taken into consideration)." The
codex, however, clearly sets it up for a first visibility of the morning
star, four days after inferior conjunction. [The Spinden equivalent of
this date, A.D. 363 April 11 (Julian) or 12 (Gregorian), was actually a
good possibility for a last visibility of the morning star. But that is
not what the context in the codex calls for.]

It was left to Dittrich to contrive a rationale for that date under

the Spinden correlation. Noting first that it could not represent an
actual heliacal rising, he took it to be a cipher of sorts, as if it were a
secret code. He considered the 19 Xul and . the 18 Kayab heliacal-rising
dates (April and November in the mid-tenth century) to be the more reliable
ones, and he concluded (p. 349): "If one wanted to camouflage the date of
the epoch, the natural thing to do would be to declare, instead of the
November base, an April one. The initiate would of course know that the
date should be moved forward into the future by the amount that it would
take to cause the day 1 Ahau 18 Kayab to fall in November." How much would
that be? The drift of the haab within the tropical year amounts to about
12.6 days per calendar ~ d [(365.2422 - 365) x 52 = 12.5944.]
The interval from November 24 ( the Gregorian date of 1 Ahau 18 Kay ab in
A.D. 934) to April 12 (its proposed mask) is 139 days. 139 divided by 12.6
is equal to 11 and a small fraction. The hidden interval must therefore be
eleven calendar rounds! And so the initiate could derive the true epoch
from the publicl y declared false one: 9.9.9.16.0 (1 Ahau 18 Kayab) +
1.8.19.17.0 (11 CR) = 10.18.9.15.0 (1 Ahau 18 Kayab), this latter being the

23
longcount equivalent of A.D. 934 November 19 (Julian), or November 24
(Gregorian), by the Spinden correlation [5 CR further along than its
10.5.6.4.0 counterpart by the Thompson].

This remarkable hypothesis is the best that could be done by way of

the Spinden correlation for the date that is the frontispiece of the
Dresden Codex Venus table and that had to be allowed to stand (one way or
another!) for a heliacal rising of the morning star.

Makemson, in her second attack on the problem, and going over to a

Spinden-l ike correlation (i.e., with a 12.9.0.0.0 rather than an
11.16.0.0.0 Ka tun "13 Ahau"), attempted to remedy this inherent discrepancy
by cutting 246 days off from the Spinden correlation constant; but in so
doing, she had to discard the historical underpinning of that correlation
(which i t shared with Thompson's) in the Landa year-bearer equation.

The discrepancy in 9.9.9.16.0 was of lesser numerical magnitude under

the term s of the Thompson correlation, but the logical dimensions of the
problem were not appreciably less. Schulz and Makemson, when they were
entertaining the possibility of that correlation, came the closest to a
solution; but for reasons already reviewed each fell short of attaining it.
Teeple and Thompson were still farther from it, having failed even to
recognize the historical date that offered the only possibility ever for an
articulation of the 1 Ahau 18 Kayab position of the Maya calendrical cycle
with the heliacal-rising position in the astronomical Venus cycle.

The refractory pieces of this puzzle have been the 9.9.9.16.0

longcount date, which is given a very special treatment in the Venus table
preface but which has been difficult to relate to the apparently intended
phenomenon of Venus, and the number 1.5.14.4.0, which on the face of it
seems incongruously composed and whose function in the table's chronology
and correction procedure has been obscure. As noted earlier, these two
items have posed for many scholars an obstacle in the way of their
acceptance of the Thompson correlation. The solution that has been
described here provides the necessary explanations for both of them. It
differs only slightly from what Schulz outlined as possibilities under the
Thompson correlation, or from what Makemson proposed under that correlation
in her first study; but those differences are crucial. It offers support,
moreover, for the Thompson correlation. Strong support for that
correlation, it may be noted, has been found also in Venus-related dates
from Classic Maya inscriptions at several sites. The apparent pertinence
of some of those dates to Venus was first shown by Kelley (see Kelley and
Kerr 1973, and Kelley 1977), and their compatibility with astronomical
chronology when interpreted by the Thompson correlation I have shown in
another paper (1982).

The validity of Thompson's correlation derives from the historical

document on which it was founded (see Thompson 1927, 1935), a document
whose evidence has been unwelcome to some, and which they have tried to
explain away. The Venus chronology now is corroborative, as has been lunar
chronology from the beginning. The problem of the Dresden Codex Venus
table can no longer be held against the correlation. Rather, it may be
seen now as a type case illustrating· that reputed failures of this
correlation . may derive not from wrongness of the correlation, but from
faulty analyses of some of the problems to which it has been applied.

24
 REFERENCES

Dittrich, Arnost
1937 Der Planet Venus und seine Behandlung im Dresdener Maya-Kodex.
Sitzungsberichte der Preussischen Akademie der Wissenschaften,
1937: 326-355. Berlin.

Huber, Peter J.
1982 Astronomical Dating of Babylon I and Ur III. Monographic Journals
of the Near East, Occasional Papers 1/4. Malibu, Calif.: Undena
Publications.

Kelley, Dav id H.
1977 Maya astronomical tables and inscriptions. In: Native American
Astronomy, ed. A. F. Aveni, pp. 57-73. Austin, Texas: University
of Texas Press.

Kelley, David H., & Kerr, K. Ann

1973 Mayan astronomy and astronomical glyphs. In: Mesoamerican Writing
Systems, ed. E. P. Benson, pp. 179-215. Washington, D.C.:
Dumbarton Oaks.

Langdon, S., Fotheringham, J. K., & Schoch, Carl

1928 The Venus Tablets of Ammizaduga. London: Oxford University Press.

Lounsbury, Floyd G.
1976 A rationale for the initial date of the Temple of the Cross at
Palenque. In: Palenque Round Table Series, vol. 3, ed. Merle
Greene Robertson, pp. 211-224. Pebble Beach, Calif.: Robert Louis
Stevenson School; Austin Texas: University of Texas Press.

1978 Maya numeration, computation, and calendrical astronomy. In:

Dictionary of Scientific Biography, ed. Charles C. Gillispie, vol.
15, pp. 759-818. New York: Charles Scribner's Sons.

1982 Astronomical knowledge and its uses at Bonampak, Mexico. In:

Archaeoastronomy in the New World, ed. A. F. Aveni, pp. 143-168.
Cambridge: Cambridge University Press.

Ludendorff, H.
1937 Zur Deutung des Dresdener Maya-Kodex (Untersuchungen zur
Astronomie der Maya, Nr. 11). Sitzungsberichte der Preussischen
Akademie der Wissenschaften, 1937: 75-98. Berlin.

Makemson, Maud Worcester

1943 The Astronomical Tables of the Maya. Contributions to American
Archaeology 42; Carnegie Institution of Washington, Publ 546, pp.
183-221. Washington, D.C.

1946 The Maya Correlation Problem. Vassar College Observatory, Publ.

no. 5. Poughkeepsie, N. Y.

Manitius, Karl (tr. and annot.)

1963 Claudius Ptolemaeus, Handbuch der Astronomie. 2nd ed., 2 vols.,
Vorwort und Berichtigungen von 0. Neugebauer. Leipzig: B. G.
Teubner.

25
Schoch, Carl
1924 The 'arc us visionis' of the planets 1n the Babylonian
observations. Monthly Notices of the Royal Astronomical Society
84 (9): 731-734. London.

Schulz, R. P. C.
1935 Zur Chronologie der Maya. Zeitschrift fur Ethnologie 67: 49-68.
Berlin.

Spinden, Herbert J.
1928 Maya inscriptions dealing with Venus and the Moon. Buffalo
Society of Natural Sciences, Bulletin, vol. 14 (no.I), pp. 1-59.
Buffalo, N.Y.

Teeple, John E.
1926 Maya inscriptions [III]: The Venus calendar and another
correlation. American Anthropologist 28: 402-408.

1930 Maya Astronomy. Contributions to American Archaeology, vol. 1,

no. 2; Carnegie Institution of Washington, Publ. 403, pp. 29-116.
Washington, D.C.

Thompson, J. Eric S.
1927 A Correlation of the Mayan and European Calendars. Field Museum
of Natural History, Publ. 241 (Anthropological Series 17.1), pp.
1-21. Chicago.

1935 Maya Chronology: The Correlation Question. Contributions to

American Archaeology 14; Carnegie Institution of Washington, Publ.
456, pp. 51-104. Washington, D.C.

1950 Maya Hieroglyphic Writing: An Introduction. Carnegie Institution

of Washington, Publ. 589. Washington, D.C. [2nd Ed., 1960, 3rd
Ed., 1971, Norman, Okla.: University of Oklahoma Press.]

Tuckerman, Bryant
1964 Planetary, Lunar, and Solar Positions, A.D. 2 to A.D. 1649, at
Five-day and Ten-day Intervals. American Philosophical Society,
Memoirs, vol. 59. Philadelphia.

van der Waerden, B. L.

1943 Die Berechnung der ersten und letzten Sichtbarkeit von Mond und
Planeten und die Venus tafeln des Ammisaduqa. Berichte der
mathematisch-physikalischen Klasse der Sachsischen Akademie der
Wissenshcaften zu Leipzig 94: 23-56.

Weir, John D.
1982 The Venus tablets: A fresh approach. Journal for the History of
Astronomy 13: 23-49.

26
JOHN B. CARLSON

The Grolier Codex: A Preliminary Report on the Content and

Authenticity of a Thirteenth-Century Maya Venus Almanac

The Maya civilization of eastern Mesoamerica attained the

highest levels of literacy in the pre-Hispanic New World. It is
known that the Maya of the Classic Period (ca. A.D. 250-900)
possessed screenfold "codex" books painted on stuccoed bark paper
because there are depictions of them on Classic Maya pottery, and
decomposed examples have been found in burials. Regrettably, none
of these Classic codices is known to have survived. The major
source of Classic texts is the monumental inscriptions carved on
stone and wood and painted on stucco murals.

We are most fort unate, however, that four Maya codex books from
the Pont-Classic Period (ca. A.D. 900-1520) did survive the
destructive forces of time and the Spanish conquest. Three of
these, the Dresden, Madrid and Paris Codices (named for the European
cities whose libraries house them) have long been recognised as Maya
manuscripts. However, their exact proveniences and how and when
they came to Europe are not known.

A fourth screenfold codex containing Maya day glyphs appeared

in 1966 in the hands of Dr. Josu~ Sa~nz, a collector of
pre-Columbian art in Mexico, along with several other extraordinary
Maya artifacts also composed a perishable materials (Meyer
1973:ch.l). As are the other three Maya books, this codex was
painted on bark paper coa ted with white stucco. Unlike the others,
however, only the obverse is painted and it possesses a number of
non-Maya "peculiarities" in content and style. The unusual style
and uncertainties of provenience have caused at least a few
well-known Maya scholars, such as Sir Eric Thompson (1975) to
suggest, if · not insist, that the fourth Maya codex is a modern
fabrication.

According to the most reliable accounts (M.D. Coe, personal

communication; see also Meyer 1973:ch.1), Josu~ Sa~nz obtained the
11-page codex fragment along with three pieces of unstuccoed paper,
a turquoise inlaid mask, a beautifully carved small wooden box, a
sacrificial knife, and several other items from looters or their
representatives at a jungle airstrip, probably 1n the state of
Chiapas. The cache was allegedly found in a large wooden box 1n a
dry cave. The mask, now generally accepted as authentic, 1s on
display in the Dumbarton Oaks Collection in Washington, D.C. (see
von Winning 1968:pl.333). The carved box with Maya inscriptions,
also authenticated, has been described in detail by Coe (1974). It
is an important consideration that these two artifacts of perishable
materials, both authentic and of exceptional quality, allegedly
appeared together with the equally perishable codex fragment. This
would tend to support the truth of the story of the cache in the dry
cave. The alternative would be to suggest that the huaqueros found

27
Figure 1 A black-and-white reproduction of the color facsimile of
the 11-page Grolier Codex, published by Coe (1973).

2 3
4 5 6
 Figure 1 The Grolier Codex

7 8 9
 11
10

t ...
.... ~ - ~: .
.

1- ~c....-"

w
,-,.
two (or more) genuine, rare pre-Columbian pieces (perhaps from
different sources) and then included a faked codex with them.

The Grolier Codex - A Maya Venus Almanac

The codex first gained public and professional attention on

April 20th, 1971 when its existence was announced by Michael D. Coe
at the opening of an exhibition entitled "Ancient Maya Calligraphy"
sponsored by the Grolier Club of New York. The next day an article
by George Gent (1971:49) appeared 1n the New York Times showing
pages 4 through 7 of the "Grolier Codex". Although it has also been
known as Codex Sa~nz and Codex Coe, it was given the name "Grolier
Codex" by Coe (1973) in the published catalogue of the Grolier Club
Exhibition. This handsome oversized volume, The Maya Scribe and His
World, contains the only published facsimile along with Coe's
analysis and page by page commentary. A black and white copy of
this color facsimile is reproduced here as Figure 1. The original
codex has approximate dimensions of 18.0 cm for the greatest height
of a page; the average width is 12.5 cm. Coe's facsimile reproduces
the eleven fragmentary pages slightly greater than half size,
followed by one of three blank sheets of bark paper found with the
codex. As reported in the Maya Scribe (Coe 1973:150), a piece of
this unpainted bark paper yielded a radiocarbon date of A.D. 1230
+ 130. It is now generally conceded that the paper itself 1s
genuine thirteenth century Mesoamerican paper. However, a question
has remained in some minds as to whether the Grolier Codex 1s a
genuine Maya book or a modern forgery on ancient paper.

From the first revelation of its existence, several Mayanists

independently reached the conclusion that the Grolier Codex is a
fragment of a Venus almanac with the same basic calendrical
structure as that found in the twelfth-century Maya Dresden Codex
(see Figure 2 and Table 2). Venus, the second plane t from the Sun
with an orbit interior to that of the Earth, exhibits a Morning
Star/Evening Star manifestation. The Maya astronomers knew that
Morning and Evening Star were both aspects of the same entity and
were able to observe that the full cycle of Venus was 584 days~
This 1s quite close to the true average value of the synodic period
of Venus, 583.92 days. From the Dresden Codex Venus tables, we know
that the Maya divided the full Venus cycle of 584 days into four
stations with the following canonical durations: (1) Morning Star
(235 days); (2) disappearance and invisibility at Superior
Conjunction (90 days); (3) Evening St_ar (250 days); and (4)
disappearance and invisibility at Inferior Conjunction (8 days).
Also from the Dresden Codex, as well ~s from ethnohistoric and
ethnographic accounts throughout Mesoamerica, we are led to believe
that Venus was seen as a fearsome, death-dealing entity associated
with dire prognostications. In particular, the heliacal rising of
the Morning Star is singled out 1n these accounts for special
perils. The first appearance in the east of the Venus God
Tlahuizcalpantecuhtli, "Lord of the House of Dawn" (an aspect of
Quetzalcoatl, the "plumed serpent") was occasioned by deadly
spearing rays. Mesoamerican Venus imagery shows malevolent deities
wielding spears and launching atlatl darts into their unfortunate
victims (see Seler 1904).

32
 The Maya calendrical astronomers were always keen to
commensurate and coordinate the cycles of time and numbers that they
found significant in order to create even greater periods. The
Mesoamerican "Calendar Round" of 18,980 days is composed of 52 of
their 365-days "Vague Years" and 73 of their all-important 260-day
"Sacred Almanacs". If we further require that the 584-day Venus
period commensurate with the 365-day year and the 260-day almanac, a
grand Venus Calendar of 37,960 days is created which is exactly two
Calendar Rounds, a period composed of 65 Venus cycles, 104 vague
years, and 146 Sacred Almanacs. This is the calendrical basis for
the Venus calendars found in both the Dresden and Grolier Codices.

The Dresden Codex Venus Almanac

Figure 2 contains black and white drawing of the six Venus

pages of the Dresden Codex with the calendrical notations
transcribed in Table 2. Page 24, which provides a scheme to
maintain long-term agreement between the almanac and astronomical
reality and place it in the time frame of the Maya Long Count,will
not be discussed here as no such table was found with the Grolier
Codex. The Dresden Venus Almanac itself is five pages, numbered 46
through 50. Each of these five pages is one Venus cycle of 548
days. The left side of each page is divided into four columns
giving the calendrical dates, reading from left to right, for
disappearance before Superior Conjunction, first appearance as
Evening Star in the West, disappearance before Inferior Conjunction
and heliacal rise as Morning Star in the east, respectively. The
right side of each page shows events related to the heliacal rise of
some sort. The central image is a Venus deity in the act of spearing
a victim depicted below. The associated glyphic texts relate to
these images and contain dire prognostications associated with the
first appearance of Venus as Morning star.

The Dresden Venus Almanac contains five pages-- as do similar

Venus almanacs in the pre-columbian "Borgia Group":Cospi,Vaticanus B
and Borgia codices (Seler 1904) -- because of a fortuitous
astronomical coincidence. As it happens, five Venus periods of 584
days exactly equal eight Maya Vague Years of 365 days. Thus,
following this 2,920-day period, Venus will return to the same
calendrical position relative to the solar year. This is the basis
of the 104-year grand cycle described above, because if we repeat
these five pages thirteen times we achieve the additional
commensuration with the 260-day Almanac. At the upper left of each
Venus Almanac page are groups of Maya day signs with coefficients in
rows of four and columns of 13. These are the dates of days reached
by Venus in the 260-day Almanac at each of the four stations. To
read the tables, we proceed to the top line from left to right
through each of the five pages (eight years total) and then back to
the first page and begin on the second line and repeat until all
thirteen passagesare complete. This is the 104-year great Venus
century of the Maya. More detailed discussions of these pages in the
Dresden Codex may be found in the commentaries of Thompson (1972)
and Lounsbury (1978).

33
 •
24

Figure 2 The Venus Almanac pages of the Dresden Codex (1975).

These are pen-and-ink drawings originally made by
Carlos A. Villacorta and corrected by Ferdinand Anders.

34
 46

35
 - ........
e • •• • • • • • •
.:.:.::

0
c=:::o C 00 <l o o ~.;i,A~
c:=a ~~
~ c::=:::, --- . ":,'-I C1 '.'l 0
;t=::::::)
___ ~ ''"''"'"=
--··· c:::::::=,

47

Figure 2 The Venus Almanac of the Dresden Codex

•••••••••••••••

... --
- ... ==
~-.
.... :...:..
~

0 0000 o o ~
E§3 c'--3~

=
c::== ~ =
c::====zc:==OOO
c:::::::::J

48

37
 49

Figure 2 The Venus Almanac of the Dresden Codex

38
 50

J9
The Dresden and Grolier Venus Almanacs -- A Comparison

When we look at the Grolier Codex (Figure 1) we see on the left

of each page (with the exception of page 11) a column of wha~ were
originally 13 Maya 260-day Almanac dates that match pre cisely the
four columns of dates on Dresden pages 48 and 49 and the first two
columns on page 50. Thus, where each Dresden page represen ts one
Venus cycle of 584 da ys, each Grolier page is one Venus station.
Since four Grolier pages correspond to one Dresden page, the Grolier
Codex must have originally contained a Venus almanac of 20 pages of
which seemingly 11 pages have su rviv ed. Table 1, taken from Coe
(1973:160) shows the Maya great Venus cycle from the Dresden Codex
with a shaded area indicating the overlapping calendrics of the
Grolier. Grolier page 11 was not included becaus e, being
fragmentary and lacking the calendrical dates, Coe was unable to
place it in the sequence. If by chance the Codex had been complete
when found, we might yet hope to find the additional pages or
fragments at a later date.

The pictorial content of the Grolier shows what few would

dispute as character istic Mesoamerican Venus iconography. Standing
personages are seen to be armed with spears, atlatls, sacrificial
blades, and in one case (p.9) a stone. In these aggressive poses,
they dominate bound captives, decapitate a victim (p.6) or sling
atlatl darts into temples (pp.5,8) a Mesoamerican symbol for
conquest. Coe (1973) gives a page -by-page commentary with tentative
identifications of several of these standing figures, e.g. The Death
God (God A) on page 6 and Maize God (God E) on page 9. He also
notes that 20 different deities are named in the Dresden Codex, one
for each Venus station, listed on row 17 across the five pages of
the Almanac (See Table 2). Thompson (1972) labelled these Deities A
through T, and this sequence is also given 1n Table 1. Each
personage named also has a directional association (on line 16),
namely A (north), B (west), C (south), D (east), etc. Oddly enough,
this same pattern of god names is listed again on lines 21 (for
pages D46 and D47) continuing on line 22 (for pages 048 through
D50). The directional associations maintain the same relationship
on line 24 as before, however, the whole pattern is shifted for this
second sequence to one station ahead, i.e. starting with Deity T, A,
B, C, etc. up to Deity S for the last column of page DSO. The
reason for this shift seems clear. For example, on Dresden page 49
column a (equivalent to Grol ier page 5), the column of Lamat dates
indicate a deity (Dresden Deity M) presiding over the event of
disappearance of the Morning Star before Superior Conjunction. Also
in column 049a, line 19, the cummulative number of 1988 elapsed days
from the beginning of the almanac on page D46 is listed. If this
deity's influence is not limited to the Venus event itself on the
given day, but extends over the subsequent period up until the next
station is reached, this shift would be a method of recording the
duration of the influence over that following period. In our
example here, the duration of invisibility at Superior Conjunction
is canonically 90 days. This 90-day figure is listed at the bottom
of Dresden column 49b on line 26 close to the Deity M glyph of the
second sequence on line 22. The 90-days when added to the 1988-day
cummulative total in column D49a, line 19, gives a new total, 2078
days, given in D49b line 19. The period of invisibility ended, Venus

40
reappears as Evening Star in the west reaching the Etz'nab days with
the new Dresden Deity N presiding over the event. The west/Deity N
event is duely recorded at column position 16, 17 in the first deity
sequence of D49b where it belongs.

The preceeding lengthy exposition was given because it helps us

to understand the unique though parallel system in the Grolier
almanac. As indicated above, Dresden column D49a corresponds to
Grolier page 5 and D49b to G6. In both cases, the first Venus
station is the disappearance of the Morning Star before the period
of invisibility at Superior Conjunction on a Lamat day. Ninety days
must be added to th is day to reach reappearance as Evening Star on
an Etz'nab day. On Grolier page 5, these 90 days are written in an
unusual form above the two-headed serpent headdress of the standing
figure, which is probably the Grolier version of Dresden Deity M.
Two black bars for the Maya number 10 are contained in a
red-painted, tied bundle, creating a so-ca 1 led "ring number". To
the right of this are four red dots. In the vigesimal system, each
of the four dots counts for one group of twenty and, with the two
bars representing the number 10, a total of 90 days is indicated.
There are several known Maya uses for ring numbers, but they usually
indicate a step back in time. In this case, the ring number of 90
days, when added to the Lamat date to the left, wi 11 reach the
Etz' nab date on the next page. This was the day for the seating of
the Death God depicted -- the potential Grolier version of Dresden
Deity N. Just as in the Dresden Codex scheme, these Grolier figures
preside over a given Venus event and also relate to the subsequent
duration of time until Venus reaches the next station. The Grolier
ring numbers function the same way on each page, but with an
additional peculiarity. Within the ring bundle, the numbers are
always formed in standard Maya bar-and-dot fashion with a maximum of
4 dots. However, for the red numbers to the right that count groups
of twenty days, we see strings with as many as 12 dots and no bars
appear. The creation of numbers greater than 4 using strings of
dots is not a Maya characteristic; it comes essentially from the
Mixtec tradition of counting.

Though style will not be discussed in this paper, the Grolier

is not a "purely" Maya book when compared to the other three
codices. It is clearly a hybrid showing the influence, broadly
speaking, of what has been cal led the Mixteca-Pueb la style. The
influence of this style was extensive throughout much of Mesoamerica
in the Post-Classic Period. Though Eric Thompson (1975) rejected
the authenticity of the Grolier Codex largely on stylistic grounds,
I would suggest that the Grolier is just the sort of codex one would
expect to find in thirteenth-century Mesoamerica from the
Post-Classic Maya boundary areas such as Chiapas, the Gulf Coast,
and across the lower Yucatan peninsular to Belize and the east
coast of Quintana Roo. The iconography and style of the Grolier
relate remarkably closely to that found along the east coast of the
Yucatan peninsular. A recent study by Arthur Miller (1982) of the
murals of Tancah and Tulum contains images strikingly similar in
style to those found in the Grolier. It is also significant that
most of this material was discovered recently and was thus largely
unavailable to any potential faker in 1966. The details of these

41
 .e:,
t

.. . 1
,,i
f
;
;1

> ~it~~~f~-7~'.
~~
. ~,_f;- ~' ,:· -'", (.-f
- • - , • .1, ...
1 • ~/ i::,~~..,f.~;
,;._-<·-~-~> T~
::t~
Mi "
. ._. ,. _ li
I

"'---•><=----~-~ - ~ - -- - ·--"···- .. ___j

Groli er page 9

Figure 3 A bla ck- and -white reproduction of the Grolier Codex p.9
with the pre l i minary reconstruction by J.B. Carlson of p. 10/11.

42
 .. .. ... ,.
· ..

\, '·' -3--~~·ii:. . .·, •. '· .

~~},l~~;( ~
---"-~-- -\ti~ t;J: . :....... •.;.<f!' ·, -- .f:~- -~~ •. , . -~.,, ,.... .. ,_ -

Grolier page lO/ll

43
arguments, together with a careful critique of Thompson's (1975)
points of refutation will be presented in an expanded study to be
published by Dumbarton Oaks.

To return briefly to the question of the possible relationships

between the deities named in the Dresden sequence and those figures
shown in the Grolier, in only two cases is there a clear-cut
correspondence. Dresden Deity N has a recognisable glyphic name.
It is Schellhas' God A, the Death God, which is clearly depicted
with its collar of "death eyes" on the appropriate Grolier page 6.
Dresden Deity Q is the Maize God, Schellhas' God E, and a good case
can be made that this is the figure correctly shown on Grolier page
9. It is difficult to know if we should expect more detailed
correspondences. The Dresden and Grolier Codices are stylistically
very different and probably come from opposite corners of the Maya
region, e.g. the Dresden Codex probably derives from northern
Yucatan, perhaps even Chichen Itza. On the other hand, it is
equally difficult to imagine why a faker sophisticated enough to
create the Grolier Codex in the 1960s would not put in more detailed
correspondences, such as, for example, Dresden Deity I from D48a,
which is the Maya Sun God, (Schellhas' God G).

The Authenticity of the Grolier Codex -- New Evidence

The essential clinching argument for the authenticity of the

Grolier Codex is that it contains information that could not
possibly have been known to a potential faker in the early 1960s.
This evidence comes from the recent work of Floyd Lounsbury (1982)
who has discovered Venus-related and timed events in the Classic
Maya inscriptions -- events that are battles or raids apparently
initiated for the purposes of obtaining captives for subjugation or
sacrifice. However, before pursuing the significance of this
discovery to the content of the Grolier, we must first examine the
sequence of macabre figures found on Grolier pages 2, 6 and -- as we
will show -- on page 10.

While working on the iconography of the Grolier Codex and other

Venus tables during the summer of 1982 as a member of the Dumbarton
Oaks "Borgia Codex Group" Summer Study Seminar, I came to the
realisation that Grolier page 11 must actually be a fragment of page
10. In principle, page 11 could have been from any of the other 10
missing pages. However, from a detailed look at its contents, it is
clear that the macabre figure of page 11, like the personage on page
5, must be holding up an atlatl as indicated by the dart that he has
already launched toward the body of water below.

Measurements of the pages were made and the pieces were shifted
horizontally with the vertical displacement fixed by a red baseline
beneath the feet of the standing figures on pages 9 and 11. A
preliminary, tentative reconstruction is offered in Figure 3, which
shows Grolier page 9 joined to a hypothetical atlatl-wielding
macabre figure on the reconstructed Grolier page 10/11. This
reconstruction is further corroborated by the water-staining pattern
on page 11. Though this is not clearly visible in the black and
white version in Figure 3, the colour facsimile shows an approximate

44
match of the unstained area below the chin of the death figure with
the underlying fibre remnant of page 10 from which page 11 must have
been torn. It will be most important to examine the codex firsthand
to verify or refute this hypothetical reconstruction.

If indeed the original Grolier page 10 contained a skeletal

figure, it falls naturally into a pattern with the other macabre
figures on pages 2 and 6. They correspond to deities who would
preside over the first appearance in the west of Venus as Evening
Star following invisibility at Superior Conjunction, and who would
likely retain their influence over the Evening Star period.

It had been one of the Thompson's (1975) arguments that the

Mesoamericans really were only interested in Venus at heliacal rise
as Morning Star. He used the fact that the Dresden table
illustrates only this Morning Star event to argue that the other
stations were much less important. He implied that the alleged
modern fabricator of the Grolier had not grasped that the Maya would
never have been interested in these other Venus stations, even
though the Maya listed them all with their associated deities in the
Dresden tables. Thompson further chose to ignore that even in the
Dresden almanac, the same Venus glyph is used for all of the
stations with nothing to single out the Morning Star manifestations
as special.

Beginning with events depicted in the famous Bonampak murals

and chronicled in the texts painted on the walls, Lounsbury (1982)
has been able to build a most convincing case that the Classic Maya
were interested in astrologically timing raids or battle events to
significant stations in the Venus cycle, be they calculated or
observed. The glyphs for the events in question are the so-called
"star-over-earth", "star-over-shell", and "star-over-embl em-glyph
-main-sign" glyphs. These glyphs have for some time been associated
with raids and warfare. David Kelley (1973, 1977) had also begun to
explore the possibility that these "star" events were associated
with 'manifestations of Venus. Using Thompson's original 584,285
Maya-Christian calendar correlation, Lounsbury (1982) has been able
to show that many of the "star" events are associated with Venus.
Significantly, the most numerous category of event is the appearance
of Evening Star, for which Lounsbury cites eight examples.

I had heard Lounsbury's presentation of these results at the

September 1981 Oxford Archaeoastronomy Symposium but had never seen
a draft of the paper until July 21, 1982 when I visited him at Yale
to discuss, among other things, the Grolier Codex. The preprint of
the paper was a revelation because Lounsbury had discovered another
Venus event glyph, a "toothy skull with distinctive markings" (see
Figure 4). This toothy skull glyph is also used as an iconographic
element in examples such as the headdresses worn by figures on Stela
16 and on Temple IV Lintel 3 at Tikal (also shown in Figure 4). In
either glyphic or iconographic context, the association is with a
Venus event, and with the Evening Star in particular.

This exciting discovery provides new insight into Maya

astrology; specifically, it connects the various Venus events and
 3
A B
4
G
,.. H
/i.
I
i
>
I
:

5
4

History of Katun 12
History of Katun 11
Clause B
Clause B

Figure 4(a) -- Two clauses from the historical texts from the middle
panel of the Palenque Temple of the Inscriptions tablets (from
Schele 1980). In the first example (Clause B of the History of
Kat un 11) the "toothy skull" Venus glyph is found at posit ion AS and
is associated with an appea rance of Evening Star. In the second
example (Clause B of the History of Katun 12), the Venus skull glyph
is found at position H6. The "star-over-shell" event glyph is at
position G7. The date of this passage marks the maximum eastern
e 1 on g a t i on o f Venus as Even i n g St a r , the ti me when it reverses i ts
apparent motion relative to the sun.
 -- · .

- ··

Al

A2

AJ

A4

81
Cl

82
C2

83
CJ

84 C4

Figure 4(b) Drawing of Tikal Stela 16 by W.R. Coe (from Jones 1977:
38). The "toothy skull" Venus serves as the headdress of Ruler
A, and the "star" Venus symbol may be seen at the back of the
skull. The date, 9.14.0.0.0 6 Ahau 13 Muan, coincides with
the first appearance of the Evening Star.
 A 8 C D G H

2 2

3 3

4 4

5 5

6 6

7 7

8
~

-, 9
,I

:(
,,?
-r'

(,. '
llll). \

- - 1?

'-~ '. _
:~L --------- ~~I .
J '

'-"'--£...0....CU_,_,___ _ _ ;_ -~~~t~-;t'.;:~.nf ,?;~:_------_- ~ ..:_'-~-'u-~_____.' .- ~

Figure 4 (c) Drawing of Tikal Lintel 3 from Temple IV by W.R. Coe

(from Jones 1977: 46). The "toothy_ skull" with the "star"
Venus symbol at the back forms part of the headdress of the
standing figure. In the glyphic text to the left, a "star-
ove r-shel l" Venus event is indicated at posit ion B4a. This
event occurred on the date (9.15.12.2.2) 11 Ik 15 Chen
indicated at position B3, A4. Lounsbury (1982: 157) notes that
on this day, July 28 A.D. 743, "Venus as Morning Star was in
precise conjunction with Mercury."
Figure 5 A pen-and-in k drawing from Seler (1961: 70) of the Venus
god Tlahuizcalp antecuhtli~ from Trecena IX of Codex Telleriano-
Reme ns is (18 9 9) •

49
military ventures. It also directly correlates the death-head Venus
glyph with the sequence of three macabre figures on Grolier pages 2,
6, and 10/11 which correspond to the first ·appearance of Evening
Star. No hypothetical faker in the 1960s, no matter how clever,
could possibly have known to put such skeletal figures on those
specific pages.

One further piece of confirming evidence comes from the

post-conquest Codex Telleriano-Rernensis (1899) from Central Mexico.
The main figure depicted on the page describing the ninth "Trecena"
(the ninth of a series of 13-day divisions of the 260-day Almanac),
is Tlahuizcalpantecuhtli, the god of the Morning Star (see Figure
5) • This rendition of the "Lord of the House of Dawn" port rays the
god with his characteristic face-painting, headdress, and striped
body. A unique feature however is the toothy skull with flint blade
protruding from the nasal cavity seen attached to the back of the
head -- a death head wearing the headdress of Tlahuizcalpantecuhtli.
A clue to the meaning of this unusual portrayal is given by the
Spanish gloss below the picture:

Este tlauiz calpantecutli quiere dezir la maflana quando

amece~e y lo mesmo es sefior de aquella claridad quando quiere
anoche~er. Este es seftor destros treze dias, ayunavan los quatro
prosteros. (Codex Telleriano-Remensis 1988:26).

Translated, this passage may be rendered:

This Tlahuizcalpantecuhtli means lord of the morning when it

dawns and equally he is lord of that bright star when evening comes.
He 1s the lord of these 13 days, they fasted during the last four.

I int ~rpret this gloss as a statement that, at least in Central

Mexico around the time of the conquest, Tlahuizcalpantecuhtli was
Lord both of the Morning Star and of the Evening Star. Furthermore,
it seems that the artist of the Codex Telleriano-Remensis intended
to show both of these manifestations of the Venus God with the death
head clearly being the Evening Star. When we compare this figure
with Evening Star sacrificer of Grolier page 6, the coincidence is .
particularly striking. Other ske~etal forms of Tlahuizcalpantecu
-htli may be seen in the Borgia and Cospi Codices (Seler 1904) as
well as on the Tizatl4n murals (Noguera 1927). In these cases
also, I would suggest that we may be viewing the Evening Star
manifestations of the deity, a macabre Venus character that we now
know the Maya associated with astrologically-timed raiding events to
obtain captives. These Central Mexjcan examples further strengthen
the case both for the pan-Mesoamerican nature of this Venus
tradition and for the genuineness of the Grolier Codex.

Conclusion

The Grolier Codex, despite the unfortunate circumstances of its

discovery, provides a most important contribution to our
understanding of the Post-Classic Mesoamerican world. Not only does
it give us a new look at the process of cultural amalgamation taking
place in the Maya area at that time, but it also corroborates
and contributes to the new evidence for the importance of the
Mesoamericans of Venus events other than the heliacal rise of the
Morning Star.

Because of the lingering questions about the authenticity of

the Grolier Codex, very little if any research has been done since
it was first publicly displayed in the early 1970s. Furthermore,
the manuscript has apparently been returned to Mexico where it has
been essentially unavailable for general study. Though technically
the Codex is owned by Josu~ Sa~nz, I understand that it is now
kept in the "bodega" of the National Anthropological Museum in
Mex i co C i t y • A c a r e f u 1 re - e x am i n a t i on sh o u 1 d now b e made in the
light of the new evidence presented in this paper.

Radiocarbon dates should be obtained, using the recently

developed small sample technique, for fibers of the bark paper from
one of the actual codex pages. The paint and stucco must be
analysed chemically for any traces of modern inks or pigments. The
patterns of water staining should be examined to see how the
sections of the codex may have fitted together when found. In this
context, I believe that the evidence suggests that the codex may
have been damaged or tampered with at some time in antiquity after
its placement in the cache -- long before its modern discovery. It
may be possible to tell if the ten missing pages of the almanac were
attached to the present manuscript when found by the huaqueros.

Finally, it should be possible to verify the hypothesis that

Grolier page 11 is actually the lower portion of page 10. It is
clear from the published facsimile (Coe 1973) that the lower portion
of page 10 had been ripped off leaving plies of the underlying
fibrous paper. Since the page 11 fragment would have overlapped a
portion of this area, it should be possible to compare the water
staining and fiber patterns to ascertain the precise relationship of
the two pieces. Furthermore, the published facsimile seems to be in
error in attaching a piece of red-painted border to the right edge
of page 11. This must certainly be incorrect and an inspection of
the original should verify this.

In the Handbook of Middle American Indians, John Glass

(1975:12) lists the only sixteen known preconquest pictorial
manuscripts of Mesoamerica -- and two of these may actually date
from the early colonial period. The Grolier Codex, as a genuine
hybrid-style thirteenth-century Maya Venus almanac, adds
substantially to the corpus of pre-Columbian books as the future
seventeenth entry in Glass' census. It is the fourth Maya codex,
the only one currently residing in Mexico. As such, we hope that
this important Mexican national treasure will soon be acknowledged
and made available for scholarly research.

51
 Paga (Grolier)
10
Column A B C 0 0 H K 1,4 N 0 a A 8 T
Poallloo' 6C ES IC MS SC ES IC MS SC ES IC MS SC ES IC SC ES IC MS
Ring number
90 2~ 90 250 1k) 2~
Day 1lgn Clb Clml Clb Ahau 0c Ahau Lamal Kan Ix Kan Eb Lamat Etz 'nab Lamat Clb Eb lk Eb
3 2 5 13 2 1 ◄ 12 1 13
I 3
l 11 13 12 2 10 12 11 1 II
11 10 13 8 10 II 12 7 II 8 11 e 8 7 10 6 7 e II ◄
e 5 8 3 5 ◄ 7 2 ◄ 3 e 1 3 2 6 13 2 1 ◄ 12
1 13

II 8
'
11
11 13 12 2 10 12 11 1 II 11 10 13 8 10 II 12 7
8 8 7 10 5 7 e II ◄ e 6 8 3 6 4 7 2
Vl
I.\) ◄ 3 e 1 3 2 5 13 2 1 ◄ 12 1 13 3 11 13 12 2 10
12 11 1 II 11 10 13 8 10 II 12 7 II 8 11 e 8 7 10 5
7 e II ◄ e 6 8 3 5 ◄ 7 2 e u
◄
2 1 ◄ 12 1 13 '3 11 13 12 2 10 12
'
11 1
1

II
3

11
2

10
6

13 8
10 II 12 7 II 8 11 a~ 8 7 10
f 6 7 8 II e
◄ 6 8 3
!I ◄ 7 2 ◄ 3 8 1 3 2 !I j 13 2 I 4 12 I 13 3 11
13 12 2 10 12 11 1 II 11 10 13 8 10 II 12 7 II 8 11 8
e 7 10 !I 7 e II ◄ e 6 8 3 6 1 e
◄ 2 ◄ 3 1

• SC, 6upeno, Conjuncik>I\.

ES, [-,Ing Siar .
IC, lnle<k>t Coniunc:tlOn.
MS, Morning Siar .

Table 1 The Scheme of the Maya Venus Cycle from the Dresden Codex
with the surviving portion of the same almanac in the
Grolier Codex enclosed (from Coe 1973: 160, Table 3).
 Table 2 The "Scheme of the Venus Cycle" on the left halves of
Dresden Codex pp. 46-50, restored and corrected by Thompson
1

(1972: 66).

SCHEME OF THE VENUS CYCLE ON DRESDEN 46-50 (restored and corrected)

Page46 Page47 Page 48 Page49 Page 50

Line Cib Cimi Cib Kan Ahau Oc Ahau Lam.at I Kan Ix Kan Eb I Lamat Etz'nab Lamat Cib I Eb Ile Eb Ahau
1 3 2 5 13 2 1 4 12 1 13 3 11 13 12 2 10 12 11 1 9
2 11 10 13 8 10 9 12 7 9 8 11 6 8 7 10 5 7 6 9 4
3 6 5 8 3 5 4 7 2 4 3 6 1 3 2 5 13 2 1 4 12
4 1 13 3 11 13 12 2 10 12 11 1 9 11 10 13 8 10 9 12 7
5 9 8 11 6 8 7 10 5 7 6 9 4 6 5 8 3 5 4 7 2
6 4 3 6 1 3 2 5 13 2 1 4 12 1 13 3 11 13 12 2 10
7 12 11 1 9 11 10 13 8 10 9 12 7 9 8 11 6 8 7 10 5
8 7 6 9 4 6 5 8 3 5 4 7 2 4 3 6 1 3 2 5 13
9 2 1 4 12 1 13 3 11 13 12 2 10 12 11 1 9 11 10 13 8
Vl
w 10 10 9 12 7 9 8 11 6 8 7 10 5 7 6 9 4 6 5 8 3
11 5 4 7 2 4 3 6 1 3 2 5 13 2 1 4 12 1 13 3 11
12 13 12 2 10 12 11 1 9 11 10 13 8 10 9 12 7 9 8 11 6
13 8 7 10 5 7 6 9 4 6 5 8 3 5 4 7 2 4 3 6 1
14 4 14 19 7 3 8 18 6 17 7 12 0 11 1 6 14 10 0 5 13
Yaxkin Zac Zee Xul Cwnku Zotz' Pax Kayab Yax Muan Ch'en Yax Zip Mol Uo Uo Kankin Uayeb Mac Mac
16 N. w. s. E. N. w. s. E. N. W. s. E. N. W. s. E. N. W. s. E.
17 A B C D E F G H I J K L M N 0 p Q R s T
18 Red 1/ 2 Red½ Red½ Red½ Red Red Red ½ Red Red Red Red Red Red Red Red Red Red Red Red
Vemu Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus
19 236 326 576 584 820 910 1160 1168 1404 1494 1744 1752 1988 2078 2328 2336 2572 2662 2912 2920
20 9 19 4 12 3 13 18 6 2 7 17 5 16 6 11 19 15 0 10 18
Zac Muan Yax Yax Zotz' Mol Uo Zip Muan Pop Mac Kankin Yaxkin Ceh Xul Xul Cumku Zee Kayab Kayab
21 T A B C D E F G Winged Winged Winged Winged Winged Winged Winged Winged Winged Winged w;nged
Chuen Chuen Chuen Chuen Chuen Chuen Chuen Chuen Chuen Chuen Chuen Chuen
22 Winged Winged Winged Winged .. .. .. .. H I J K L M N 0 p Q R s
Chuen Chuen Chuen Chuen
23 Red Red Red Red Red Red Red Red .. .. .. .. Red Red Red Red Red Red Red Red
Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus Venus
24 E. N. W. s. E. N. W. s. E. N. w. s. E. N. w. s. E. N. W. s.
25 19 4 14 2 13 3 8 16 7 17 2 10 6 16 1 9 0 10 15 3
Kayab Zotz' Pax Kayab Yax Muan Ch'en Ch'en Zip Yaxkin Uo Uo Kankin Cumku Mac Mac Yaxlan Zac Zee Xul
26 236 90 250 8 236 90 250 8 236 90 250 8 236 90 250 ;~ 236 90 250 8
REFERENCES

Codex Borgia
1976 Codex Borgia, Biblioteca Apostolica Vatican, (cod.
Borg. Messicano 1), Commentary by Karl Anton
Nowotny. Akademische Druck-u. Verlagsanstalt,
Graz-Austria
Codex Cospi
1968 Codex Cospi, Calendario Messicano 4093,
Biblioteca Universiteria Bologna. Introduction
and summary by K.A. Nowotny. Akademische Druck-u
Verlagsanstalt, Graz-Austria.

Codex Dresden
1975 Codex Dresdensis, Sachsische Landesbibliothek
Dresden, (Mscr. Dresd. R.310). Commentary by
Helmut Deckert and Ferdinand Anders.
Akademische Druck-u. Verlagsanstalt, Graz-
Austria.

Codex Grolier
1973 The Grolier Codex, Grolier Club of New York
Exhibition on Ancient Maya Calligraphy catalog
no. 87. In The Maya Scribe and His World, by
Michael D. Coe, The Grolier Club, New York.

Codex Madrid
1967 Codex Tro-Cortesianus (Codex Madrid), Museo De
Am~rica Madrid. Introduction and summary by
F. Anders. Akademische Druck-u. Verlagsanstalt,
Graz-Austria.

Codex Paris
1968 Codex Peresianus (Codex Paris), Bibliotheque
Nationale Paris. Introduction and summary by
F. Anders. Akademische Druck-u. Verlagsanstalt,
Graz-Austria.

Codex Telleriano-Rernensis
1899 Codex Telleriano-Rernensis, Manuscrit Mexicain no.
385 l la Bibliotheque Nationale. Transcription
and commentary by E.-T. Harny, Lithographic
production by the Due de Loubat, Paris.

Codex Vaticanus B
1972 Codex Vaticanus 3773 (Codex Vaticanus B),
Biblioteca Apostolica Vaticana. Introduction
and summary by Ferdinand Anders. Akademische
Druck-u. Verlagsanstalt, Graz-Austria.

Coe, Michael D.
197 3 "A Carved Wooden Box from the Classic Maya
Civilisation". In Primera Mesa Redonda de
Palenque Part II: A Conference on the Art,
Iconography and Dynastic History of Palenque.
December 14-22, 1973, pp.51-57. Merle Greene

55
 Robertson, ed., The Robert Louis Stevenson School,
Pre-Columbian Art Research, Pebble Beach :
California.

Gent, George
1971 "Manuscript Could Change Views on Mayas' Religion".
The New York Times, April 21, p.49 New York.

Glass, John B.
1975 "A Survey of Native Middle American Pictorial
Manuscripts". In Handbook of Middle American
Indians, Vol.14, Robert Wauchope, gen. ed.,
Guide to Ethnohistorical Sources, Part III,
Howard F. Cline, Vol. Ed. pp.3-80. University
of Texas Press, Austin.

Jones, Christopher
1977 "Inauguration Dates of Three Late Classic Rulers
of Tikal, Guatemala". American Antiquity, Volume
42, No.I, January, pp. 28-60. Society for American
Archaeology, Washington, D.C.

Kelley, David H.
1977 "Maya Astronomical Tables and Inscriptions". In
Native American Astronomy, Anthony F. Aveni, ed.
pp.57-74. University of Texas Press, Austin.

Kelley, David H. and K. Ann Kerr

1973 "Mayan Astronomy and Astronomical Glyphs". In
Mesoamerican Writing Systems, a Conference at
Dumbarton Oaks, October 30-31, 1971, Elizabeth P.
Benson, ed., pp. 176-216. Dumbarton Oaks Research
Library and Collections, Washington D.C.

Lounsbury, Floyd G.
1978 "Maya Numeration, Computation, and Calendrical
Astronomy". In Dictionary of Scientific Biography
Volume XV, Supplement I, Charles Couston Gillispie,
ed., pp. 759-818. Charles Scribner's Sons, New York.

1982 "Astronomical Knowledge and Its Uses at Bonampak,

Mexico." In Archaeoastronomy in the New World ,
A. F. Aveni, ed., pp. 143-168. Cambridge University
Press, Cambridge.

Meyer, Karl E.
1973 The Plundered Past. Atheneum, New York.

Miller, Arthur G.
1982 On the Edge of the Sea: Mural Painting at Tancah-
Tulum2 Quintana Roo, Mexico. Dumbarton Oaks,
Washington, D.C.

Noguera, Eduardo
1927 Ruinas de Tizatlan, Tlaxcala: Los Altares de

56
 Sacrificio de Tizatlan, Tlaxcala. Publicaciones
de la Secretaria de Educaci6n PGblica, Mexico City.

Schele, Linda
1980 Notebook for the Maya Hieroglypic Writing Workshop
at Texas. Institute of Latin American Studies, The
University of Texas at Austin, Austin.

Sele r , Eduard
1904 "Venus Period in the Picture Writings of the Borgian
CodexGroup". In Mexican and Central American
Antiquities, Calendar Systems and History,
translated and edited by Charles P. Bowditch,
Smithsonian Institution Bureau of American Ethnology
Bulletin 28, pp.355-391, Washington D.C. (Originally
published in Verhandlungen der Berliner Gesellschaft
fur Anthropologie, Ethnologie, und Urgeschichte,
1898, pp. 346-383).

1961 Gesammelte Abhandlungen zur Amerikanischen Sprach-

und Altertumskunde, Volume IV, Akademische Druck-u.
Verlagsanstalt, Graz-Austria.

Thomps on, J. Eric S.

1972 A Commentary on the Dresden Codex. American
Philosophical Society, Philadelphia.

197 5 "The Grolier Codex". In Contributions of the

University of California Archaeological Research
Facility, no. 27, Studies in Ancient Mesoamerica,
II, John A. Graham, ed. pp.1-9~ Department of
Anthropology, University of California, Berkeley.

von Winning, Hasso

1968 Pre-Columbian Art of Mexico and Central America.
Harry N. Abrams. lrtc. Publishers, New York.

57
BARBARA TEDLOCK

Quichean Time Philosophy

Beginning with the important work of Heinrich Berlin (1958) and

Tatiana Proskouriakoff (1960), there has been a revolution in our
thinking about the supposed Mayan glorification of or obsession with
time. Previously it had been assumed that all Mayan hieroglyphic
texts dealt solely with recurrent cyclical astronomical matters
(Thompson 1954). Then, with the sudden breakthrough in
hieroglyphic textual analysis it became clear that many of the texts
on Classic Maya stelae recorded births, dynastic successions,
alliances, marriages and deaths of individuals. Nonetheless, not
all Mayan texts, even on stone monuments, concern such historical
matters. Thus far, the extant Precolumbian codices (Madrid,
Paris, Dresden and Grolier) are believed to contain primarily
cyclical, ritual and astronomical matters. We have a difficult
theoretical problem here because in Western thought, beginning with
the Classical Greeks, temporal modes are described as either
cyclical and recurrent or as linear and non-repetitive. This
dichotomy ultimately restson the contradiction between Plato's view
of time as the measure of the motion of a body and Aristotle's view
of time as the motion of bodies. Since Classic Mayan cyclical and
lineal time models are clearly present archeologically in the same
sites at the same time horizon, it would appear that the Mayas had
somehow resolved this apparent contradiction, if indeed they even
took these models to be in conflict or contradiction. Our problem,
then, is to examine this supposed contradiction in order to
construct a new model of Mayan time philosophy.

As a first step, let me assume that the Maya may have had
different times for separate provinces of their reality--
biological, astronomical, mechanical, psychological, historical,
religious, social-- and that these various time perspectives
underwent a process of "totalisation" (as Sartre would call it)
recorded and rationalized within the intermeshing cycles of their
calendars. I come to this intuition from the study of modern Mayan
multimetrical temporal concepts and rituals involving dialectical
thought patterns which go far beyond the dialectics of polarization
(thesis, antithesis, synthesis), as historically exemplified in
Hegelian and Marxist circles, to include the dialectics of
complementarity, overlapping or mutual involvement, and
reciprocity. 1 These insights concerning the dialectics of Mayan
temporal concepts come from reflect ions upon data gathered during
twenty months of anthropological fieldwork, including a formal
apprenticeship to a calendar priest in Momostenango, Guatemala. 2

The community of Momostenango is located in the Midwestern

Highlands of Guatemala at 15.04° north of the Equator. Although

59
this community is located in the tropics, the climate is quite
temperate because it lies from one to three thousand meters above
sea level. The population of this municipality is approximately
45,000 persons, of whom the overwhelming majority (98%) are Quiche-
speakers. Momostenango is perhaps best known to travelers for its
fine woolen blankets and to ethnographers for its celebration of 8
Batz', the largest ongoing calendrical ritual anywhere in
Mesoamerica today that is scheduled according to the 260-day
(rajilabal k'ij 3 "counting of suns and days") sacred almanac. Here
reside nearly 10,000 initiated female and male "burners" (poronel)
and "daykeepers" (ajk'ij), who are formally trained and initiated
in calendrical rituals (in accordance with both the solar 365-day
cycle and the sacred 260-day cycle) by a formal three-tiered
hierarchy of male priest-shamans known as "mother-fathers"
(chuchkajawib).

Unlike many other Guatemalan communities, where calendar

experts have been persecuted and peripheralized by Roman Catholic
priests and indigenous lay catequistas, the Momostecan hierarchy of
calendar experts has prevented such religious domination by
carefully controlling access to the coveted positions of eldership
within the community. As elsewhere in Mesoamerica, one may speak
of a "civil-religious hierarchy" in Momostenango, but here the
bridge between religious and governmental duties is provided by this
group of calendricists rather than by the leaders of the local
Catholic confraternities. Thus, men who are actively working
toward the respected position of elder within the community tend to
avoid serious involvement in the confraternities and instead undergo
further training in calendrical ritual, so as to ascend in the
local hierarchy.

The first level of religious training is that of a simple

"burner", a man or woman who may approach the outdoor community
altars in order to burn incense and make offerings to the ancestors
and deities. Novices undergoing training for this office are known
as ''burdens" (e'lrnmal) during the 65-day period known as "washing for
the work service" (ch'ajbal chac patan), which begins shortly after
noon on 1 Cawuk. On this date the teacher arrives at Paja', "In
the Water," a shrine which is located in a barranca on the edge of
a stream just west of the town center. There the teacher begins
the "back part of the path" (rij ubinibal, literally "rear or back
of the instrument-for-walking or traveling"), which is the current
expression in Quiche for indicating the beginning of a set of
calendrical rituals. This spatio-temporal concept places the
speaker in the position of starting off behind the final or "big"
day of the initiation rather than as standing before, as we express
it in phrases such as "the night before Christmas".

Beginning the 65-day ritual series slightly after high noon,

when the sun is directly overhead, raises a question as to when the
Quiche begin and end their days. This issue has been much debated
in the literature on other Mayan groups, with noon and sunset being
the favourite candidates for the boundary line between two
successive days (La Farge 1930: 657; Redfield and Villa Rojas 1934:
184; Lincoln 1942: llO; Villa Rojas 1945: 143-44; Thompson 1950:
102). At first glance the case of the 1 Cawuk would seem to

60
support the noon theor y , an d so would th e way in which the
Momostecans sometimes take not ice of the fa ct that the bells of the
parish church are st r iking tw el v e . Having st o pped what they are
doing long e no ugh to glance up at th e sun, t hey may then say a
brief prayer with their fa ces s till u plifted slightly, addressing
the sun by i nvo ki ng the day-- for exampl e , Sa' j la Ajaw Wajxakib
K'anil, "Come here, Lord 8 K' anil." Bu t t h e ~ if several people
happen to be to gether a t th is tim e, th ey wi ll greet one another (as
if they had just met) b y using th e e xpression t h at is appropriate
for the rest of t he afternoon: x be k'ij, "the sun (day) went," in
which x- i ndic at es co mpl eted action - - t h e i mplication being that
the sun has just a l ready d one so me thing, ra t h e r t han that something
has just begun . This is qui t e a disor ien t at ing experience to t he
outsider, es pecially if t hat outsider already kn ows that high noon
is ordinarily referred to co nversationally in Quiche as nic'aj k'ij
"middle of the s u n (or da y o r time)." How can the "middle of the
day" also be the ti me when the day is greeted and asked to enter or
come here, and s i multaneously be the time when t h e day "went"? In
what sense is n oon the beginning , middle, and end of the day ,
sun, and time? This case is partly resolved when one learns that
the path of the sun is referr e d to in Qu i c he as o x ib u x ucut " t h re e
c o r n ers , " consist i ng of sunrise (rele b a l k ' i j , ~ h e comi ng -o ut
place o f the sun"), noon (nic ' aj k' i j, "mi ddle of the sun, day o r
time") , a n d sunset (uk a jibal k ' i ~ "the go i ng-down place of th e
sun"). These thr ee "co rners"are considered the three main
turning points or t r ansitions in time, in which the influences of
day (l_' ij) an d night ( a k'a b ) overl ap in a dialectics of mutual
involvement.

At the 15° latitude where Momos t enango is l ocated, night comes

on rapidly at s unset, which occ urs around 6 p.m., with only very
slight seas onal fluctuations. At this time in the evening one
greets people on t h e p ath i n Momostenango wi th xoc ak'ab "night
entered." Ni gh t c on tinue s on into early dawn. At this time there
are black stre aks low on t h e eastern horizon, with a yellow
background that s lowly int ensifies and deepens into orangeness (ya'
k'ak' chuwi xe kaj , literally "fire is given there at the bottom
edge of the sky") . Then all of the colour slowly disappears and
the rising of the sun itself brings about a process called sakir
uwach ulew (" the f ac e o f th e earth whitens or lightens"). If one
meets a person on t he pa th at this time the pr oper greeting is
sakiric "it is g ett in g wh i te or light," and this continues to be
the proper greet i ng until noon. However, even though the sun has
now appeared, this time of day is known as nimak'ab "big night,"
which indicates that the night has grown very large or long and is
nearing its en d . Th is se em s stran g e until one realizes that the
period of time between dawn and noon is considered an especially
delicate, cool time when the sun is slowly climbing up the sky.
Then when it reach e s n oon it is quite strong for a time, but it is
already spoken of as if it had completed itself.

In the attempt to decide begi n ning and ending points for a

day , one enco u n t ers the same sort of problems I have discussed
el s ewhere in connect i on with t h e misdirected search for a beginning
or ending point for the 260-day cycle (Tedlock 1982: 93-97). Just
as I discovered that Momo stecans were more interested in offering a

61
particular day as the mirldle of the cycle rather than in bpeculating
on the question of its "beginning," so I must suggest here that the
important issue at the scale of a single day is the midpoint. At
noon, the day is maximally 8 K' anil (or whatever the Day Lord is),
and the influence of 8 K'anil rises before that moment and declines
after it. A dream occurring on the previous night, at whatever
hour of the darkness, will be spoken of as having been handed from
7 Can to 8 K'anil; a dream of the following night is handed over
from 8 K'anil to 9 Toj, and so forth. From the fact that a day
has a sharply marked middle it does not follow that-- at least not
in Quichean dialectics-- that is has sharply marked beginning and
ending points. The moments of sunrise and sunset are certainly
used to mark time, but they do not provide absolute boundaries for
the influence of successive Day Lords. As for "midnight," that is
a moment that is reckoned in accordance with events that belong to
the night itself rather than a boundary between two "days," as we
shall see later on.

Now, returning to the rites performed on the afternoon of 1

Cawuk, we can see more clearly why this time of day is chosen.
Once noon has past, the influence of 1 Cawuk is strong but is
already declining, pointing forward, as it were, to the rest of
the series of rituals which it begins. Starting with the second
day in the series (1 E) and from then on, a teacher may well prefer
the early morning hours, before the capacity of a Day Lord to
listen to prayers has been worn out. He returns on 1 E to the
same shrine he visited the first time (Paja', In the Water), where
he again asks permission of the Day Lord-- and of his own direct
ancestors-- to train the novice. After this second ritual he
begins to give intensive instruction in time reckoning, cosmology,
observation of the sun (ki'j), moon (ic'), and stars (ch'umil)t
herbal and shamanic curing and properri tual behaviour to be
followed at the outdoor shrines. Now, the rhythm of his visits to
the shrines speeds up to 7- and 6-day intervals in order to
i nte rcala t e or insert the 8-day series ( wajxaki bal) into the 1-day
series (junabal) that was started first. With the first 8-day, 8
Cawuk, there is also a shift of locale to Ch'uti Sabal, "Little
Declaration Place," which is a shrine located on a hilltop half a
kilometer due west of Paja'. Six days later on 1 Can the teacher
returns to the low shrine at Paja'.

This series of 1-days and 8-days (see fig. 1) rotates back and
forth at 7- and 6-day intervals to produce the following meter: 1
Cawuk + 13 = 1 E + 7 = 8 Cawuk + 6 days = 1 Can + 7 days = 8 E + 6
days = 1 Batz'+ 7 days = 8 Tijax + 6 = 1 C'at. Expressed in the
language of Western music, this 65-day "back of the path" time
period opens with a 13/65 time signature, in which the right-hand
figure (65) indicates the unit of measurement (1/65 of the total
time period under consideration), and the left-hand figure (13)
indicates the number of such units in each measure. In this ritual
series, however, after only one measure the meter speeds up and
alternates back and forth four times from 7/65 to 6/65 through eight
measures, producing an irregular multi meter. Th is mult ime te r
resolves itself and achieves an exciting asymmetrical balance
through the principle of dialectical complementarity, in which the
distinctions (7 and 6) are simultaneously in an alternating

62
relationship to each other-- 7, 6, 7, 6, 7, 6, 7, 6-- and in
dialectical completion of each other-- the original 13 is matched
with 7 + 6 = 13, repeated four times. A third type of dialectical
complementarity, known as direct opposition, is present in the
spatial dimension of the rituals, in the shift from low to high
place (Paja' /Ch'uti Sabal); low to high number (1/8); east to
west; and wet to dry.

After a person has been initiated as a "burner," she or he may

decide to go on to be a "daykeeper," by completing another 65-day
"back of the path" permission period known as "washing for the
mixing pointing" (ch' ajbal baraj pun to). This period begins at the
Paja' shrine with 1 Quej (see fig. 1), which is then followed 13
days later by 1 Junajpu, also at Paja'. At this point the cycle
begins the alternating 7- and 6-day multimeter pattern: 1 Quej + 13
= I Junajpu + 7 = 8 Quej + 6 = 1 Aj + 7 = 8 Junajpu + 6 = 1 Came +
7 = 8 Aj + 6 = 1 Cawuk + 7 = 8 Carne+ 6 = lE. The first thing to
notice about these two 65-day time periods is that the second one,
for the more advanced initiation as an ajk'ij "daykeeper," actually
falls earlier within the 260-day cycle than the first 65-day period
for the training as a "burner" (see fig. 1). Secondly, these two
65-day time periods are overlapping, so that the first and second
days of the "washing for the work service" are also, respectively,
the third-to-last and final days of the "washing for mixing and
pointing." Thus, the teacher completes the chronologically later
65-day period (work service) first and then waits nearly 250 days to
begin the earlier (mixing and pointing) 65-day period.

The initiation as either a "burner" or a "daykeeper" is

celebrated in a two-step procedure which takes place outside of the
65-day (rij ubinal "back part of the path") ritual series. In the
case of a''burner" these rituals are performed on 7 Tz'i' and 8
Batz', while in the case of a "daykeeper" the days are 8 Batz' and
9 E. The rituals on 7 Tz' i' take place at either the home of the
novice or of the teacher beginning at sunset (ukajibal k'ij, "the
going down of the sun," which is simultaneously the term for the
western direction), culminating with fireworks late in the evening.
Then, on 8 Batz', the rituals take place at the 8-day shrine,
Ch'uti Sabal, beginning very early in the day, after Venus as
morning star (junajpu) has risen, or if Venus happens to be an
eveing star (rask'ab) or simply is not visible at all, when the
earliest rays of light have appeared among the black clouds at the
horizon (xe kaj "end of the sky"). If the person initiated will
become a burner·, then this is the final day of his/her initiation.
Later, if the person is being initiated as a daykeeper, s/he will
celebrate 8 Batz' once more and then go, on the next day (9 E), to
a higher shrine known as Nima Sabal, "Large Declaration Place."
The first day of these two sets of initiatory days (7 Tz'i' and 8
Batz'-, respectively) is known as the "broom" (rnesebal), while the
second day (8 Batz' and 9 E, respectively) is known as the "big
day" ( nirna k' i j ) •

The ideal pattern in Momostenango is to speed up the initiatory

process by completing these two overlapping 65-day cycles during
just one 260-day period, so that the novice is simultaneously
initiated as both a burner and a daykeeper (see fig. 1). In order

63
to "double-time" these two initiations the "permissions" must begin
with the second, or more advanced "mixing and pointing" series and
then, before the completion of this cycle, the first day (1 Cawuk)
of the first set of "work service" must be i nte rcala ted. Th is
means that the next day (8 Came) of the "mixing and pointing" set
must be picked up after the first day (1 Cawuk) of the "work
service" set has occurred. This is followed by 1 E, which occurs
as the last day in the "mixing and pointing" and the second day in
"work service." The double-time series then continues straight on
with the remainder of the days of the work service series.
Initiation in this double-timed system is celebrated during the
three-day consecutive period of 7 Tz'i', 8 Batz' and 9 E. When
the two levels of initiation are combined (through a dialectics of
mutual involvement) within a single 260-day cycle, 7 Tz'i' is known
as the "broom" ( m ese bal), 8 Batz' is the "eve" ( m ixpr ix) and 9 E is
the "big day" ( nima k' i j).

The pattern whereby rites of passage are marked by ·three

consecutive days of ritual activities at the end of a double-timed
cycle is replicated on the higher levels of the religious hierarchy
of Momostenango, ranging upward through the mother-fa the rs of the
pat r ili neage s, the can tons and the town as a who le. Thus, for
example, in the case of either of the two mother-fathers of the
town, the "back part of the days" before the initiation into this
office can theoretically be done in two separate time periods. The
first one, the "backpack" or ~ka'bal (literally "pack frame" or
"yoke"), stretching from 1 Quej to 1 Toj (see fig. 2), is divided
into 65 + 65 + 52 days = 182 days (or one half of the solar year).
This involves the burning of great piles of copal incense, praying
and making of other offerings for the benefit of Momostenango, and
even for the entire world. The second is the "washing of the
shrine" or ch'ajbal rech awas (literally "washing of the taboo"),
which stretches f r o ~ o j to 1 Batz' and is likewise divided into
65 + 65 + 52 = 182 days. The two parts can be done in two separate
solar years, bu t in 1976 when the new mother-father for Pueblo
Viejo was initiated they were double-timed. The "back part of the
days" began with the "later" series, the "washing of the shrine,"
and ran as follows: 1 Toj + 13 = 1 Ik' + 7 = 8 Toj + 6 = 1 Tz'iquin
+ 7 = 8 Ik' + 6 = 1 K'anil + 7 = 8 Tz'iquin + 6 = 1 Imox + 7 = 8
K'anil + 6 = 1 Ix= 65 days. What would have been the "earlier"
series; the "backpack," began 13 days after 1 Ix on 1 Quej, .and
was thereafter intercalated with the "washing of the shrine" series.
The combined backpack and washing rituals of 1 Quej were followed by
a return, 7 days later, to the washing rituals on 8 Ix, a day
which is skipped in the backpack series when it is done alone.
With the exception of 8 Ix, the two sets of rituals were done
together from 1 Quej until 1 Batz', combining the washing of the
town shrines (i.e. removal of previously burned copal) and the
backpacking of the town. Next the backpacking continued on alone,
but only until 1 C'at. Thus, in actual practice what could have
been 2 x (65 + 65 + 52 = 182) = 364 days were overlapped (through
the dialectics of mutual involvement) and reduced to 65 + 13 + 65 +
52 = 195.

As in the case of the combined burner and daykeeper initiation

discussed earlier, the initiation of mother-fathers of the town

64
took place on three consecutive days. The first or "broom" day was
9 E, the second or "eve" was 10 Aj and the third or "big" day was
11 Ix. If the backpack and washing series had been done over two
separate years, they would have each been culminated by a two-day
ritual. I could detail many more examples of contrasting two-day
and three-day rituals, but the general rule is this: wherever such
rituals are the culmination of two overlapping counts of days, the
observances will span three days. Otherwise, they will span only
two.

Reflecting on the so-called "tzolkin triad" found in the

eclipse tables of the Dresden codex (pp. 53a-58b), one wonders if
this might also have marked off overlapped cycles. Various
hypotheses concerning the possible function of this triad have been
advanced. For example, Satterthwaite (1965: 623) thought that it
might have been used as +/- 1-day allowance (e.g. a three-day error
range) in lunar correction or variation. An alternative
interpretation is that it was used to shift from one line to the
next above it at certain periodic intervals (Andrews 1940: 156;
Lounsbury 1978: 796). Most recently, it has been suggested that
it functioned as a one-day recession of the window-defining base
dates necessitated by the recycling of the table (Bricker and
Bricker 1983: 12). Alternatively, I would suggest that the triads
in this table indicate that double or even multiple cycles are
overlapped, quite aside from the question of variability or
recession in the astronomical phenomena that may correlate with
these c y c 1 e s. The Brick e rs f o cu s on the so 1 a r asp e ct of th is
table, but following my sense of the Mayan (or at least Quichean)
preference for •overlapping dialectics and for multimetrical time
reckoning, I follow Kelley in calling these table "lunar-solar"
(1976: 42-43). The ethnographic research needed to pursue these
issues further has barely begun (see Remington 1977; Neuenswander
1981). What is specifically needed is more information on
contemporary highland Guatemalan conceptualization and observation
of the moAn, and of the nocturnal sky in general. The richness of
what remains to be learned is only hinted at in what I will be able
to sketch out here.

There is a general and widespread interest in the night sky in

Momostenango. This is particularly true during the cold, dry
season months of November, December, January and February, when
people rise aft~r midnight and go visiting, consult diviners and
attend seances. At this season of the year, on nights without a
bright moon, the winter Milky Way or ube tew, "ice road," clearly
reveals its rift; that part of the Milky Way is called xibalba be,
"road to the underworld." Night rituals which take place during
this season are marked in terms of the rising of certain stars and
constellations. These time markers include the rise of Regulus
(jun ch'umil, "one star"), Orion's Belt (oxib ch'umil, "three
stars") and the Big Dipper (pac', "cupped hands"). During this
same season, on the night of the f ul 1 moon ( r ij a ic '), when the
Milky Way is barely visible, people rise when the moon is at its
peak near midnight and walk to the various hot mineral springs in
the community in order to bathe. These trips to the springs
involve courtship and sexual liaisons, which are timed in
accordance with the phases of the moon.

65
 When the moon is "small" (alaj) it is a time when all of the
world is considered tender-- animals, plants, trees and peopl~
During the fifteen days (7 + 8) of the waxing moon, butchering,
harvesting, woodcutting and sexual relations are avoided. Then,
on the night of the full moon and for the following fifteen days
until the dark of the moon, when she is "buried" (mukulic), all ot
these activities become propitious, since the moon, and all
humanity with her, are now hard or mature (ri j). These idealized
fifteen-day intervals recall the similar intervals that appear seven
times in the lunar-solar table of the Dresden codex.

The night of the full moon (jun ak'ab ube "one night her road")
is a particularly important night each month in Momostenango. Only
on this night does the moon's path-- called (like the sun's path
oxib uxucut, "three corners"-- cross the sky from east to west in
a single night. The three corners consist of moonrise in the east
(relebal ic', "the coming-out place of the moon"), midnight of the
full moon(pa nic'aj "at the halfway" or "middle") and moonset in
the west (ukajibal ic' "the going-down place of the moon"). In
order to contrast them, the solar and lunar triangles are referred
to respectively as chupam sakil "in the light" and chupam k'ekum "in
the darkness." It was not until I understood these triangles that
I could understand a remark that a Momostecan layman once made while
we were walking down a path just after sunset. On seeing the full
moon that had just come up, he said: "The sun has risen." Here I
would suggest that at least some of the kin or sun glyphs in the
Dresden lunar-solar tables (see 56a, s2--i;-:- 54b, 56b, 57b and
58b), superimposed on boundaries between light and dark
backgrounds, .might be metaphors for the full moon rather than
literal indications of the sun.

In the more esoteric world of priest-shamans, there is an even

greater and more serious interest in the paths of the heavenly
bodies. Those who have been initiated as mother-fathers at any
level of the three-tiered hierarchy visit the highest hills within
the community on regular schedules, and the two who are at the
highest level visit the sacred four-directional mountains as well.
Some of these sacred places have good views of the sky and of large
stretches of the horizon. The following information on ritual
cycles was gathered without knowledge of the possible importance of
82-day cycles (3 x 27 1/3 = 82) in charting the motion of the moon
among the constellations (see Aveni 1980: 67-82).

Momostecan patrilineage leaders visit Nima Sabal, one of the

highest shrines, on the following sequence of days: 9 Quej + 13 =
9 Junajpu + 13 = 9 Aj + 13 = 9 Came + 13 = 9 Cawuk + 13 = 9 E + 13 =
9 Can+ 4 = 13 Toj = 82 days. On the first day of this series they
open their particular patrilineage's shrine at Nima Sabal at sunset
and remain for some time, burning_ incense, praying to their dead
predecessors in office by name and observing the night sky. In
these prayers they mention the phases of the moon and its position
in the night sky. They are particularly interested in the seasonal
variations in the moon's path through the summer Milky Way (saki be,
"white road"), which they compare with its passage through the
bifurcated winter Milky Way. They discuss these variations among
themselves and occasionally make notes on their Gregorian calendars

66
at home. Certain men are known as experts in predicting rain
according to the phases of the moon in its seasonal voyage through
the Milky Way. These men carefully observe the night sky on all
seven days of this series of 9-days (belejebal), but this is not
considered necessary by the others, who observe the night sky with
any seriousness only on the opening day (9 Quej) and 82 days later
on the closing (13 Toj), when they once again arrive at sunset to
pray, burn incense and observe the night sky before they close the
shrine for a 22-day period. They will repeat this pattern once
again on 9 Batz'+ 13 = 9 C'at + 13 = 9 No'j + 13 = 9 Tz'i' + 13 = 9
Ak'abal + 13 = 9 Ajmac + 13 = 9 Toj + 4 = 13 Ajmac, at which point
another 82-day cycle has passed. Now the shrines remain closed for
74 days before the first cycle, from 9 Quej to 13 Toj, begins
again. These 82-day periods are referred to in Quiche as chac'alic
"to be staked, suspended, stabilized or set," which was explained
to us as referring both to the firm placement of a table on four
legs and to the forked poles which are planted to support the roof
beams of a new house.

The point of particular astronomical interest here is that

wherever the moon (if visible) might have been located among the
constellations on a given 9 Quej or 9 Batz', it would be in the
same position 82 days later on the following 13 Toj or 13 Ajmac,
respectively. It requires further fieldwork to confirm the
present evidence that this cycle is precisely what the mother-
fathers are observing when they make their nocturnal visits to Nima
Sabal. I should add that I also have evidence of four different
82-day periods in the rituals performed by one of the two mother-
fathers who s~rve the entire town of Momostenango, but those
rituals, which involve overlapping 82-day cycles with others of 40
and 65 days, require a separate paper to themselves.

The reckoning of 82-day periods in Momostenango, and their

connection with ritual mountaintop visits that are known to include
observation and discussion of the night sky, make it necessary to
r~-open the question as to whether Mayan astronomy included not only
synodic moon-reckoning, but sidereal reckoning as well. On the
synodic side, the conceptual identity of the moon with the sun, on
just the one night of the full moon, may give us new ways of
reading and interpreting the lunar-solar pages in the Dresden
codex. And the "tzolkin triads" in those same tables may signal
the overlapping of two or more cycles, as do three-day rituals at
Momostenango, rather than providing for errors of measurement.

The conception of the oxib uxucut or "three corners" of the

paths of the sun and the full moon may call for a re-opening of the
question as to whether ancient Mayan astronomers were interested in
angles, though of course the corners in question here are angles
attributed to the celestial movements themselves rather than to the
geometry of observation.

On the issue of Mayan time philosophy, it is apparent that

contemporary Quiche thinking about time frequently follows a
dialectics of overlapping or mutual involvement. There are moments
of dialectal complementarity, as when time intervals both alternate
(as between 7 and 6 days) and complete one another (to add up to the

67
meaningful pattern number of 13), or when two separate shrines (one
for 1-days and the other for 8-days) are in direct and unmediated
opposition in a particular ritual cycle. On the other hand, a
dialectics of mutual involvement is seen when burner (or work
service) rituals overlap with daykeeper rituals, when washing
rituals overlap with backpacking rituals,- and when night overlaps
with day.

Notes

1 For excellent discussions of various forms of dialectics see

Gurvitch (1964) and Sartre (1976).

2 This fieldwork was made possible by a Research Fellowship from

the State University of New York at Albany and by a summer Faculty
Fellowship from Tufts University. Tony Aveni, Dennis Tedlock and
Mary Jane Cramer have all made useful suggestions on the first draft
of this paper, which the author has incorporated. Any remaining
ambiguities are her own. I am most grateful for the support
provided by these individuals and institutions.

3 The orthography used for Quiche in this paper is the practical

one suggested by the Instituto Indigenista Nacional de Guatemala
(see David G. F.o x, Lecciones elementales en Quiche, pp. 15-18).
Vowels are pronounced as in Spanish, e is like the vowel in English
"met," and o is like the vowel in English "foot." Consonants are
also as in Spanish, with an equivalence between c (used before a,
o and u) and qu (before e, i), except fork, which is articulated
with the tongue farther back than for c or qu; tz, which is like
the German Zeit; w, which is like English w; x, which is like
the English sh; and b, which is a glottalized p. Other
glottalizations are indicated by '

68
References

Andrews, E. Wyllys

"Chronology and Astronomy in the Maya Area," in The Maya and

Their Neighbors, ed. L. Hay~~- New York: Appleton-
Century, pp .. 150-61. (1940)

Aveni, Anthony F.

Skywatchers of Ancient Mexico. Austin: University of Texas

Press. (1980)

Berlin, Heinrich

"El glifo 'emblema' en las insc ripciones may as." Journal de la

Societe des Americanistes, 47, 111-19. (1958)

Bricker, Harvey M. and Victoria R. Bricker

"C 1 a s s i c Ma y a P r e d i c t i on of S o 1 ar E c 1 i p s e s • " Current

Anthropology, 24, 1-24. (1983)

Craine, Eugene . R. and Reginald Reindorp

The Codex Perez and the Book of Chilam Balam of Mani. Norman:
------- ---
University of Oklahoma Press. (1979)

Fox, David G.

Lecciones elementales en Quiche. Guatemala: Instituto

Linguistico de Verano. 0973)

Gurvitch, Georges

The Spectrum of Social Time. Dordrech t, Holland: D. Reidal.

(1964)

Kelley, David Humiston

Deciphering the Maya Script. Austin: University of Texas

Press. (1976) - -

La Farge, Oliver

"The Ceremonial Year at Jacaltenango." Twenty-third

International Congress of Americanists in New York, 1928.
New York, pp. 656-60. (1930)

69
Lincoln, J. Steward

"The Maya Calendar of the Ixil of Guatemala." Contributions

~ American Anthropology and History, 38, 99-128. (1942)

Lounsbury, Floyd G.

"Maya Numeration, Computation, and Calendrical Astronomy,"

in Dictionary~ Scientific Biography (New York: Scribner),
15, suppl. 1, ed. C.C. Gillispie, pp. 759-818. (1978)

Neuenswander, Helen

"Glyphic Implications of Current Time Concepts of the Cubulco

Achi (Maya)." Manuscript prepared for Centro de Estudios
Mayas, Universidad Nacional Autonoma de Mexico. (1981)

Prouskouriakoff, Tatiana

"Historical Implications of a Pattern of Dates at Piedras

Negras, Guatemala." American Antiquity, 25, 454-75.
(1960)

Redfield, R. and A. Villa Rojas

Chan Korn, A Maya Village. Carnegie Institution of Washington

Publication-4~ Washington, D.C. (1934)

Remington, Judith A.

"Current Astronomical Practices among the Maya," in Native

American Astronomy, ed. Anthony F. Aveni. Austin:
University of Texas Press, pp. 75-88. (1977)

Sartre, Jean-Paul

Critique of Dialectical Reason. Translated by Alan Sheridan-

Smith. London: NLB. (1976)

Satterthwaite, Linton

"Calendrics of the Maya Lowlands," in Handbook of Middle

American Indians, 3. Archaeology of Southern Mesoamerica,
ed. Robert Wauchope and Gordon R. Willey. Austin:
University of Texas Press, pp. 603-31. (1965)

Tedlock, Barbara

Time and the Highland Maya. Albuquerque: University of New

Mexico Press. (1982)

Thompson, J. Eric S.

Maya Hieroglyphic Writing: an Introduction. Norman:

University of Oklahoma Press. (1950)
 The Rise and Fall of Maya Civilization. Norman: University
of Oklahoma Press. (1954)

Villa Rojas, Alfonso

The Maya of East Central Quintana Roo. Carnegie Institution

of Washington Publication 559. Washington: D.C. (1945)

Fig. 1

WORK SERVICE MIXING POINTING

(Burner) (Daykeeper)

1 Quej

1 Junajpu
8 Quej
1 Aj
8 Junajpu
1 Came
8 Aj
1 Cawuk 1 Cawuk
8 Came
1 E 1 E
8 Cawuk
1 Can
8 E
1 Tijax
8 Can
1 Batz'
1 C'at

71
 Fig 2

BACKPACK WASHING THE SHRINE

1 Toj

1 Ik'
8 Toj
1 Tz' iquin
8 Ik'
1 K'anil
8 Tz'iquin
1 Imox
8 K'anil
1 Ix
8 Imox
1 Quej 1 Quej
8 Ix
1 Junajpu 1 Junajpu
8 Quej 8 Quej
1 Aj 1 Aj
8 Junajpu 8 Junajpu
1 Came 1 Came
8 Aj 8 Aj
1 Cawuk 1 Cawuk
8 Came 8 Came
1 E 1 E
8 Cawuk 8 Cawuk
1 Can 1 Can
8 E 8 E
1 Tijax 1 Tijax
8 Can 8 Can
l Batz' 1 Batz'
8 Tijax
1 C'at
8 Batz'
1 No'j
8 C'at
1 Tz'i'
8 No' j
1 Ak'abal
8 Tz'i'
1 Ajmac
8 Ak' abal
1 Toj

72
ARTURO PONCE DE LEON H.

Fechamiento arqueoastronómico en el altiplano de México

El siguiente estudio* pretende como objetivo general aportar

algunas ideas que sirvan en la investigaci6n y análisis de los
elementos arquitect6nicos y espacios urbanos prehispánicos
proponiendo estos como instrumentos para el conteo del tiempo;
mediante la posible observancia de las posiciones solares en el
horizonte, señaladas por las orientaciones de dichos elementos.

En razón de la mecánica que se establece para el conteo del

tiempo mediante las estructuras piramidales, es posible aproximar o
comprobar la época de construcción de estos monumentos; es decir
fecharlos arqueoastron6micamente. Dicho sea de paso también
· comprueba el "aparente" desfasamiento del calendario nahuatl, con
respecto al movimiento solar medio y propone, a diferencia del
criterio "europeo" a base de intercalación de días, el ajuste
"físico-calendárico" a base de la medici6n del número de días
desfasados, mediante el cambio de orientación de las pirámides,
por ende el cambio de la posición solar en el horizonte. Semejante
esto a la corrección o serie secundaria del calendario maya: ya que
si ésta medía los días calendáricos desfasados respecto a un origen,
las diferentes orientaciones de los sitios prehispánicos medían
físicamente este mismo desfasamiento, con lo cual era posible
predecir las estaciones y ciclos agrícolas.

En el altiplano de México existen muchos lugares cuyas

características de trazo, ubicación y realización arquitectónica,
están relacionadas con ciertas condiciones de tipo orográfico y
astronómico, esto es, en su construcción indudablemente algo tiene
que ver que el eje de una construcción apunte a un cerro y que éste
sea referencia a la ubicación de un cuerpo celeste, generalmente el
sol, en una fecha de terminada. 1 En general la arq ui tect ura
ceremonial en la América antigua tiene un carácter esencialmente
solar si bien estos edificios tienen algunas relaciones con la
culminación, puntos de ocultación o apariciones helíacas de algunas
estrellas (Marquina 1934; Avení 1977, 1980), el carácter más
importante de estos edificios es el solar; como lo sugiere la regla
general en su orientación oriente-poniente, la cual no obstante su
variedad de unos edificios a otros, en muy raros casos, se~ala más
allá de los puntos en que se observa el sol en el horizonte.

Los ejes de ~us estructuras indican exactamente los puntos de

su registro solar. Llamamos días de registro sotar a cuatro días
del año, dos al amanecer cuando el sol en su movimiento aparente

* Las fotos citadas más abajo pertenecen a la colección del autor.

73
recorre el hemisferio sur y dos al atardecer cuando el sol recorre
el hemisferio norte. Dicho de otra forma, el eje de la estructura
señala cuatro fechas, en las que el sol aparece o desaparece en el
horizonte excatamente en el punto indicado por el eje. Por
ejemplo: una estructura que tenga 17° de orientaci6n al sur del
oriente, registrará al oriente la salida del sol, los primeros
días de febrero y noviembre, cuando el sol recorre el hemisferio
sur y al poniente el ocultamiento del sol los primeros días de mayo
y agosto, cuando el sol recorre el hemisferio norte.

Existe una relación entre los "días de registro solar" de estas

estructuras y la posici6n de la tierra en su eje de traslaci6n
alrededor del sol; si arbitrariamente dividimos en cuatro los
posible movimientos de la tierra alrededor del sol, serán cuatro
los posibles días de registro solar, así el sol deberá registrarse
en el horizonte dos días al amanecer y dos días al atardecer, como
en el ejemplo anterior.

La orientaci6n de cada centro ceremonial, entonces, nos

indica cuál es su registro solar; pero si comparamos los registros
solares de cada centro, esto es la dirección de sus ejes, podremos
notar sensibles diferencias. ¿Tiene esto alguna explicaci6n?
¿Cuál fue la razón de estas diferencias? Existen varios estudios
sobre las relaciones astronómicas de diferentes sitios o estructuras
en particular, pero considero principalmente éuatro los autores que
han analizado el tema desde el punto de vista global y genérico,
proponiendo las razones en la variedad de las orientaciones.
Escalona Ramos ( 1940) menciona que las orientaciones varían con el
tiempo a razón ·de un grado cada 260 años y fija el origen de estas
orientaciones al corriente-poniente astronómico, por el año 2914
a.c., con lo cual dice el investigador se puede de terminar la época
aproximada de la fundación de una ciudad o de la construcción de un
edificio, sabiendo la orientaci6n de su eje principal. F. Tichy
(1976), al investigar la región de Puebla, forma tres familias de
distintas alineaciones, con lo cual propone que las variaciones en
las orientaciones, responden a un módulo angular mesoamericano de
4.5° como veinteava parte de 90°, submúltiplo ideal de un sistema
vigesimal. Anthony Aveni ( 1977) estudia un grupo de orientaciones
de 17º entre otras y el ajuste casi perfecto en Teot ihuacan con la
posición de la puesta en el horizonte de las pléyades, así como la
salida helíaca de este mismo grupo de estrellas el día en que
ocurría el primero de los dos tránsitos del sol por el cenit de
Teotihuacan, la ciudad más grande e influyente de las ciudades
contemporáneas de la antigua América. También propone la relación
de fechas agrícolas, con la posición de salida o puesta del sol en
los sitios con orientaciones de mayor desviación. Fujisyoshi
Yoshio (1981) relaciona diferentes sitios arqueológicos de todo el
mundo y establece en base a los períodos agrícolas, diferentes
según zonas geográficas, que las orientaciones en general señalan
la posición solar en el horizonte, los meses de mayo y junio.

Las ideas hasta ahora más significativas para el enfoque del

presente estudio) o sea la señalización solar en el horizonte,
mediante la orientación de las estructuras, pienso que puedan ser
las enunciadas por estos investigadores. Si para Escalona las
orientaciones van cambiando sis temáticamente, un grado cada 260
años, a partir del año 2914 a.c., entonces el grupo de sitios con
orientaciones de mayor desviaci6n (20° a 25º) por ejemplo, tendrán
que construirse entre el año 2286 d.C. al 3686 d.C., cosa que sale
~ -,¡¿, ;~ ,~, ,-~.~ .. ,_ • .:-,·_ ,_ . . :_.... . En cambio Aveni, Tichy y Yoshio, suponen
que ias orientaciones de los centros ceremoniales, se han definido
con cierta variedad preestablecida, siendo la idea en común para
estos investigadores el considerarle a la diversidad de
orientaciones un medio con el cual fijar un calendario agrícola.

Ahora bien, basándonos en la idea de que una estructura señala

cuatro fechas mediante una misma orientación, vamos a desarrollar
varias alternativas, que relacionen el tiempo probable de
fabricación del monumento con sus días de registro solar. Para lo
cual vamos a utilizar una gráfica circular (véase lámina 1) que
señala en el sentido angular la ascensi6n recta del sol y, en el
sentido concéntrico, las épocas culturales. Así de esta forma
podremos graficar la posición de la tierra en la eclíptica los días
en que las estructuras piramidales, según su época de fabricación,
señalan en el horizonte la salida u ocultamiento del sol. Para la
primera alternativa vamos a suponer, de los cuatro registros de
cada estructura, los dos que corresponden al poniente, 4 es decir
las orientaciones que señalan el ocultamiento del sol en el
horizonte dos veces al afio, una cuando el sol va de primavera a
verano (véase Graf. lA en lámina 1) y otra cuando el sol va de
verano a otoño (véase Graf. lB en lámina 1). Para la segunda
alternativa vamos a suponer los otros registros solares de cada
edificio que señalan al oriente la aparici6n del sol en el
horizonte, también dos veces al año, cuando el sol va de invierno
a primavera (véase Graf. 2A en lámina 1) y cuando va de otoño a
invierno (véase Graf. 2B en lámina 1).

Las siguientes dos alternativas suponen uno de los cuatro

registros de cada estructura y les consideran una secuencia temporal
respecto a las estaciones, así en la tercera alternativa (véase
Graf. 3A en lámina 1) que hemos denominado "secuencial di recta" los
edificios registran primeramente en verano, posteriormente en
otoño, invierno, primavera, verano nuevamente y por último otra
vez en otoño, siguiendo el sentido de las estaciones. En la
cuarta alternativa (véase Graf. 3B en lámina 1) "secuencial
inversa", los registros solares son en primavera, invierno,
otofio, verano, primavera nuevamente y por último otra vez en
invierno siguiendo el sentido inverso a las estaciones. El
análisis de las posibles alternativas de registro solar, centrando
nuestra atenci6n en las dos últimas que presentan una secuencia
16gica, nos permite establecer que estas diferencias se deben a que
cada centro ceremonial fija la orientaci6n de su eje en distintas
fechas, esto es, observando el sol en diversos momentos de su
movimiento aparente.

Las fechas de registro solar guardan, entonces, una

correspondencia con el órden de las estaciones del año conforme
pasan los siglos. Esto nos lleva a relacionar los monumentos
arqueol6gicos con el cómputo del tiempo. Los centros ceremoniales
y el calendario se conjuntan en una importante celebraci6n: el
Fuego Nuevo (Palacios 1934, Saenz 1967, García 1975). Si como
hemos visto, el "Fuego Nuevo" se registraba por la orientaci6n

75
axial de las estructuras y si éstas se iban orientando de diferente
forma, a manera que los registros solares iban sucediéndose en
diferentes meses del año, según como iba pasando el tiempo. ¿Será
posible pensar que estas celebraciones cíclicas de cada 52 años
también se iban llevando a cabo en meses diferentes y no siempre en
el mismo mes? El diferente registro solar de cada centro
ceremonial nos indicaría en consecuencia que la ceremonia del Fuego
Nuevo se efectuaba en diferentes fechas. 5 A esta misma conclusión
llega el Dr. Alfonso Caso (1968), proponiendo el desfasamiento de
las celebraciones del "Fuego Nuevo", conforme pasan los siglos.
Para una mejor comprensi6n de esto vamos a analizar de qué forma
sucede.

Son dos los aspectos más importantes en la cronología de

Alfonso Caso, la colocaci6n del día portador del año, que lo
considera no al principio del año, sino al final de la XVIII
veintena y la no inclusi6n del día correctivo equivalente al día
bisiesto europeo. 6 El año indígena de 365 días, según Alfonso
Caso, no se ajustaba con días agregados a lo largo del siglo, por
eso cada fiesta de Fuego Nuevo se adelantaba 13 días
aproximadamente, es decir se desfasaba con respecto al año solar
medio, también en el sentido inverso a las estaciones. Si los
centros ceremoniales y el calendario cambiaban sus fechas de
celebraci6n cada 52 años de manera inversa al sentido temporal de
las estaciones del año, por ejemplo: si el Fuego Nuevo y la
orientación del templo se fijaban a fines de mayo, conforme pasaban
los siglos iban registrándose ambos en abril, marzo, febrero etc.

Alfonso Caso pudo establecer que el último Fuego Nuevo fue

7 a partir de esta
celebrado el 31 de enero gregoriano de 1508,
fecha se puede graficar retrospectivamente cada ciclo de 52 años.
El siguiente paso consiste en compasar las fechas de registro solar
de cada sitio, según la alternativa "secuencia l inversa" (véase
lámina 1) con las fechas de cada Fuego Nuevo. En esta gráfica se
puede constatar la relaci6n que existe entre la variaci6n de los
registros solares y las celebraciones de ciclos calendáricos de cada
52 años, los dos se van desfasando en el sentido inverso de las
estaciones, conforma pasan los siglos. Este desfasamiento se
puede precisar, estableciendo año, mes y día en que cada ciclo se
cumpli6, elaborando una tabla retrospectiva de Fuegos Nuevos,
según cronología de Caso (véase lámina 2). Si a esta tabla se la
coteja con las fechas de registro solar y los años en que la
historia y/o la arqueología sitúan a cada lugar, se verá la
correspondencia que existe entre cada monumento y cada ciclo de 52
anos.

Comencemos por el Huizachtecatl en Iztapalapa (véase plano 1),

el sitio en que según las fuentes se celebraba la fiesta de Fuego
Nuevo, 8 todavía unos años antes de la conquista. El eje de esta
estructura con 12° 15' al sur del oriente señala a la cumbre del
volcán, Huitlaxochiotl, que se encuentra al norte del poblado de
Tláhuac, por lo que el registro solar diurno de esta pirámide es el
día 20 de febrero. Todos los historiadores coinciden que la
primera celebración de Fuego Nuevo (García 1975) que se llevó a cabo
en este lugar fue la que dio nombre al período de 1351 a 1403 o,
según cronología de Alfonso Caso, 1352 a 1404 y como el siglo era
 r·;.- ---- - .. .r.--- ,. .:,:~. . _, -- i .,- ~
'\ \_../'·,./ fechamiento
\ arqueoast ronomico
1

- 1, 1 p 1 • 11 o

+••· + SIMBOLOGIA.._

--- ----l- O

- - •i•
o
reatos a,_lbQlco,
mojonara natural
do la oatructura arq .
( .
... _ ... ,¡, contonltndo coord1nada1

tx\> geo,¡rclfica1 (latitud o longitud)

l
eje que une vart01 1itio1 arq,
·, - - - tjt 1ol1llclal
porte agua I de lo cutnca dtl
valla dt m·•lco (SARH)

1
1+ ,o· -T~,-~~r~- ;

_\ 1
+
( '- - -'"'
_
/
+ ,o· \
r ' ·-
\
,._,./
_J
•
•
rivera dtl IGIIO en 1521
ragi1tro 101 or diu mo
reolltro 1olar v11p1rtino

i ·--=--,::: _!
·---r '- .... \ ·l~ai==s= .,-±-c-,------.---=-~ ---------,,-----
.('.:,
; tloloc
..,' C.CHAPt.L TEPEC ~.9,,;
l. 1~ , ' 1\ - - ---. Z.ARQ .
1
; _/ - ·-::- ..,--_e:-~ - :,
CALIXT AHUACA 0. 1
1· 'lx;c.la palma(,,.,."....... ¡
1
A c.telapón

J \ CUICUILC~ )
1
- j. / ...-.:~ ...
+
1-
~-'</ __ __ + +
-----1
-----1 ~ lÍEQUIPA

\ i

t --

,,,- - 7± 1t•oo·____ j ___ _ +

¡

•·e, !
1
1 ;
1
¡ 1

-·-- -l------ --

c:......._ _......:J:.._ __ •L
1

_ ._¡.___
+ 1
...l
+_ J.I
i

-+ + "- .)
Plano 1 Región central de México
designado con el nombre del último año de ese siglo, la gráfica
ase lámina 2) señala el 25 de febrero de 1404 como día de la
~elebración del Fuego Nuevo. Esta diferencia de cinco días entre
la fecha calendárica y el registro de las efemérides, se presenta
en casi todas las estructuras analizadas., decimos casi, pues en
ocasiones la lectura corresponde exactamente a la fecha prevista.
El número de cinco días nos hace pensar en los Nemontemi, los cinco
días aciagos con que finalizaba el año indígena. 9 Asimismo otra de
las constantes que nos vamos a encontrar, es que la celebración del
ciclo calendárico correspondiente al registro solar en los sitios,
es precedido en 50 años aproximadamente por el inicio del
asentamiento ya concocido, esto pudiera ser debido a que primero se
realizó el poblamiento, se estudió y se afinó geográficamente la
ubicaci6n y orientación del sitio 10 y tiempo después se registró
solarmente el ciclo calendárico.

El siguiente lugar que analizaremos es el Templo Mayor de

Tenochtitlan; las excavaciones han revelado once etapas de
construcción para la fachada principal y siete perimetrales (Matos
1981), el eje de orientaci6n de las más recientes puede ser
considerado actualmente con un valor de 7° 25', 11 prácticamente el
mismo eje 12 promedio de 7° 30' que presentan las calles del centro
de la Ciudad de México (véase plano 1 ). Siguiendo la orientación
de 7.5° al sur del oriente, registramos la efeméride solar el 4 de
marzo. 13 Sabemos históricamente que la ciudad de Tenochtitlan qyi
conoció el Conquistador fue fundada en el período de 1300 a 1352,
y que en este año debió celebrarse el Fuego Nuevo, la fecha según
la gráfica (véase lámina 2) el 9 de marzo. Una vez más la
diferencia de 5 días entre ambas fechas.

En Calixtlahuaca al norte de la ciudad de Toluca, los ejes de

las pirámides de Quetzalcoatl y de Tlaloc, 15 son difíciles de
precisar por los paños irregulares con que cuentan; pero se pueden
considerar un valor promedio de 1.5° al norte del poniente, por lo
que sus registros serán a finales de marzo. Calixtlahuaca tiene un
período de asentamiento que va de 1200 a 1500 de nuestra era (Piña
1963) y su registro debió efectuarse según la gráfica (véase lámina
2) por el año de 1248 el día 3 de abril, unos días después de su
registro solar.

Teopanzolco situado al oriente de la ciudad de Cuernavaca

(véase plano 1) tiene el eje de la estructura principal 16 orientado
2.5° al norte del poniente, señalando el pico sur del Cerro del
Aire. De los dos templos que rema tan la estructura, el del lado
norte, de planta irregular, cuenta en cada una de sus esquinas con
un pilar exterior, ligeramente separado (2 cm. aprox.) del paño del
muro. El paso del sol es registrado en forma especial a través de
esta separación el 29 de marzo, cuando éste se oculta atrás del
Cerro del Aire. Teopanzolco está ubicado arqueológicamente hacia
el afio 1200 de nuestra era (Piña 1963). Según análisis
retrospectivo de las fechas conmemorativas (véase lámina 2)
encontramos que el 3 de abril de 1248 debió celebrarse el ciclo
calendárico. Los cinco días de diferencia se vuelven a registrar.

Los lugares que veremos a continuación, igual que los dos

últimos, tienen un fechamiento arqueológico que, sumado al

78
análisis retrospectivo de los ciclos de 52 años, servirá para
corroroborar las efem€rides solares. Veamos el caso de Tlapacoya
(véase plano 1) ubicado a 20 km. al oriente de la Ciudad de México;
orientadas al norte del oriente, 1
una de las pocas estructuras
como La Venta¡ Cuicuilco, el eje de esta estructura señala al
cerro Telap6n.
1 Por la orientación 20.5° al norte del poniente su
registro solar es el 24 de mayo; argueol6gicamen te se sitúa hacia
19 El análisis retrospectivo
los afios 500 o 400 antes de Cristo.
(véase lámina 2) señala el 23 de mayo de 468 antes de nuestra era.
La diferencia es de un día.

Cuicuilco en el sur de la ciudad de México: su orientaci6n

20 que se encuentra en el parteaguas
señala el Cerro del Papayo
oriente del Valle de M€xico, con medio grado de variación al nort~
del oriente, 21 registra el 23 de marzo, casi los equinoccios,2
orient aci6n confirmada fºr las estructuras prehispánicas colindantes
en la Villa Olímpica; 2 la tabla retrospectiva (véase lámina 2)
señala el 21 de marzo del año 208 antes de Cristo, fecha que
difiere en dos días al registro de la estructura.

La Pirámide del Sol en Teotihuacan presenta algunos problemas

24 por la
para la medici6n de su eje oriente-ponie nte
"reconstrucci ón" hecha de ella pero se puede establecer
25 al sur del oriente que difiere
aproximadamente en un órden de 17°
a los 15° 28' de la perpendicular a la Calzada de los Muertos.
Entre otras referencias, pues no mencionaremos los marcados que
definen ejes de registro solsticiales y equinocciales o de paso
solar cenital en la Cuenca, voy a hacer mención de la relad.ón que
guarda con la estructura que René Mill6n (1968) denomina Zona 31,
A, B, C y Den el plano 43 y que se encuentra a 1125 metros al
poniente de la Pirámide del Sol (véase plano 1) desde donde se
26 Arqueol6gicame nte,
registra la · efem€ride solar el 7 de febreto.
la Pirámide del Sol se sitúa en el año 100 a.c., según la gráfica
(véase lámina 2) corresponde al 13 de febrero del ano 52 a.c., con
cinco días de diferencia.

La Pirámide de Cholula, con su eje solsticial de 25.5°

27 indica en su registro solar el 21 de
aprox. al sur del oriente,
diciembre (Tichy 1978). Fechada arqueológicame nte en su primera
etapa por el año 150 de nuestra era, según la gráfica (v€ase lámina
2) corresponde al 24 de diciembre del año 155 d.C.: la diferencia
es de 3 días.

La Pirámide de Quetzalcoatl en Teotihuacan, orientada 16° 30'

al sur del oriente, la misma que la Avenida Este-Oeste (Avení
1977), registra la salida del sol el 30 de octubre. Fechada
arqueológicamen te por el 300 d.C., según tabla retrospectiva (v€ase
lámina 2) el 3 de noviembre del año 363 d.C.; que difiere en cuatro
días.

Analicemos por último la Pirámide de la Serpiente Emplumada en

Xochicalco (véase plano 1) y ~ue tiene una orientaci6n de 16°
2 que indica su registro solar el .5
aprox. 28 al norte del poniente
de agosto, situada arqueológicam ente por el año 700 d.C., según
análisis retrospectivo (véase lámina 2) el 7 de agosto de 727 d.C.:
dos días de difencia.

79
 En este sistema se puede observar que en un período de 1508
años se repite la fecha inicial de registro (véase lámina 2), esto
es después de 29 ciclos, 30 de 52 años se debe a que cada 52 años
esta fecha se va adelantando 12.594 días
3 1 por no ir de acuerdo con
el año solar medio (véase lámina 3). Así el desfasamiento total en
1508 años será de 365.2376 días, con uná diferencia al año solar
medio de 6 minutos 30 segundos (tiempo) por lo que prácticamente las
fechas de registro se repiten en cada uno de los cuatro centros
ceremoniales cada 1508 años. ¿Será por eso que los centros
ceremoniales construídos en las cercanías de Teotihuacan quince
siglos más tarde tienen una orientaci6n similar a la de la Pirámide
del Sol? (Avení 1980) La definici6n de un sitio para el
establecimiento humano debi6 de reunir una serie de condicionantes
agrícolas, ecol6gicas, de seguridad, de pesca, de caza entre
otras más, pero evidentemente también geográficas y calendáricas.
Si alguno de los centros ceremoniales analizados anteriormente se
hubiese situado un centenar de metros al norte o al sur, ya no se
cumplirían ciertas condiciones calendáricas de su registro solar.
Por ejemplo, si la Pirámide del Fuego Nuevo en Iztapalapa se
hubiese construído en la base del Cerro (véase plano 1), el
registro solar mediante el Cerro Huitlaxochiotl no sería en la fecha
establecida; o la Pirámide de Teopanzolco, si estuviese más al sur
o al norte (véase plano 1) no registraría su eje sobre el Cerro del
Aire la fecha 29 de marzo. El mismo caso para Cuicuilco o la
Pirámide de la Serpiente Emplumada en Xochicalco.

Todo lo anterior nos hace pensar que los conocimientos

astronómicos y calendáricos debieron aparecer en el Altiplano antes
de la erecci6n de los centros ceremoniales, ya que son numerosos
los sitios arqueol6gicos con evidentes implicaciones geográficas,
astron6micas y calendáricas, que muestran sertales de ocupaci6n
anteriores a la época de construcci6n de los sitios de culto.
Sabemos que el sistema calendárico en Mesoamérica era muy semejante
en su estructura, aunque la nomenclatura variara. ¿Acaso la forma
de registrar el tiempo mediante la orientación de los edificios fue
la misma en toda Meso américa ? 32 ¿Se ordenarían por las mismas
reglas los imponentes edificios de la Zona Maya, los de Oaxaca,
los de la vertiente del Golfo de México, los que se ubican en los
Altos de Guatemala o los de la Zona Andina? Al menos algunos de
esos sitios apuntan esta probabilidad.

A más de los sitios analizados anteriormente, existen otros en

el México antiguo que por su ocupaci6n, orientaci6n y época de
construcci6n, parecen estar en el mismo contexto que hemos
establecido (véase lámina 4), aunque algunos por su fechamiento
hasta ahora conocido llegan a no coincidir por un margen de 50 a 152
3
años con los ciclos calendáricos, como Tenayuca, 33 Teotenango,
y el Juego de Pelota en Xochicalco (véanse fotos 16, 17 y 18). En
cambio sí parecen coincidir Tepoztlan,
35 _ Tlateloco, Tula, 36
Culhuacan, 37 y Chalcatzingo, todos estos en el Altiplano. En el
Area Maya, tenemos de Chichen Itzá el edificio de las Monjas y la
Casa Colorada, con registros entre los días primeros de septiembre
y los últimos de agosto, según la gráfica (véase lámina 4) por el
año 600 de nuestra era, que coincide con el fechamiento conocido,
también en Chichen Itzá el Castillo de Kukulcán, el Templo de los
Gu€rreros y en Tulum el Castillo, todos estos con registros a

Bo
mediados de mayo por el ano 1100 D.C., el Juego de Pelota en
Chichen Itzá de estilo Maya-Yucateco (Piña 1980) situado entre el
año 1000 y el 1200 D.C., cuya orientaci6n de 17º (Avení 1980)
registra el 4 de mayo, en forma interesante a través de sus arcos o
altares superiores en los muros laterales, 3 8 según la tabla
retrospectiva (véase lámina 4) el 11 de mayo del año 1091 D.C., lo
cual concuerda con el contexto establecido.

La Venta de Tabasco con sus 8° al norte del oriente, bien

podrá sefialar al amanecer de los días filtimos de agosto o primeros
de septiembre, su fechamiento conocido de 1000 a 800 A.C., el
ciclo calendárico (véase lámina 4) que según la tabla retrospectiva
señala, sería el año (600 D.C. 1500) 900 A.C.
aproximadamente. 39

Es criticable el hecho de presentar un contexto tan extenso,

temporal y geográficamente. ¿Cómo es posible que se pueda explicar
de la misma forma la razón de la orientación de los centros
ceremoniales en el Preclásico, en el Clásico y en el Pos tclás ico,
cuya diferencia en el tiempo llega a más de mil años? ¿Acaso
tuvieron las mismas razones, los constructores de Cuicuilco,
Teotihuacan, Xochicalco y Tenochtitlan para ubicar y orientar sus
templos y ciudades?

El presente estudio propone un esquema general del

funcionamiento, posiblemente el más importante de los centros
ceremoniales. Si estos "corregían" su orientación con respecto a
los anteriores cada ciclo calendárico de 52 años prehispánicos,
observando el movimiento aparente del sol en el horizonte, se puede
decir entonces que el calendario prehispánico sí contenía ajustes
para ir acorde al movimiento solar. Pero esta corrección no era al
estilo "europeo", pues no lo admitía su sistema combinatorio
(Palacios 1934; Caso 1968; Carrasco 1980). Su elemento
correctivo era la medición del número de días desfasados, entre el
calendario de 365 días y el año solar medio (mediante el cambio de
orientaci6n de las Pirámides) de esta forma se sabía del
desplazamiento con respecto a las estaciones, con lo cual sí se
podían prever los ciclos agrícolas.

81
 +15 v---~ ENSAYO
arqueoastronómico
altiplano
+1 de
méxico
-'<

/ y . - ·500
SIMBOLOGIA
-
•
o REGISTRO SOLAR
VARIACIONES DEL R.S. EN EL
1

~TM . T.~ . o TIEMPO

SENTIDO DE LAS ESTACIONES

ALTERNATIVAS DE REG . SOLAR

~CAL -··-·•-IA (cuando el sol va de primave-
ra a verano)
-··-•--1B (cuando el sol va de verano a
otoño) ·

_. -· -2A ( cuando el sol va de invierno

a primavera)
-·---2B (cuando el sol va de otoño
a invierno)
________ 3A (siguiendo el sentido de los
estaciones, verano, otoño,
invierno, prima ve ro etc. )

TLAP .Tlapacoya
CUl Cuicuilco
PST P. del Sol Teotihuacan r
CHO Cholula
PQT P.Quetzo. Teotihuacan
PSE P.Serp. Emp.Xochicalco
TEOP Teopanzolco l ALTERNATIVAS
ldmina n.o.
DE REGISTRO SOLAR.
_____________
CAL Calixfl.ahuaca
TM TemploMayorTenochtitla1
PFN P.FuegoNuewlztopalapc \..ª·
......_ ponce de leon h. ~

Lámina 1 Alternativas de registro solar

 ENSAYO
arqueoastronómico
altiplano
de
méxico

SIMBOLOGIA
O REGISTRO SOLAR

- --•·· - VARIACIONES DEL R.S. EN EL

TIEMPO
◊ SENTIDO DE LAS éSTACIONES

-------r\.__CICL.OS CALENDARICOS C/25años

.--i.;-w(sec;¡un cronologla de A. Caso)

ALTERNATIVAS DE REG . SOLAR

_______ 3B (inverso al sentido de las

de las estaciones, primavera
invierno, otono, verano,
primavera otro vez¡ etc.)

TLAP Tlopocoyo
CUI Cuicuilco
P. del Sol Teotihuocon
Cholulo
PQuetzo . Teotihuocon
P.Serp. Emp.Xochicalco
TEOP Teoponzolco
I'. LOS REGISTROS
lamina"º·
CAL Colixtlohuoco SOLARES Y LA CRONOLO
T M Templo MoyorTenochtitla GIA DE ALFONSO CASO.
P FN P Fuego Nuevo lztopolop O. leon h.
''481 ponce de
ij 8

Lámina 1 (cont.) Los registros solares y la cronología de Caso

~- s o o - - 23 movo 468
____________ ......_.,"., .

.·
•
TLAPACOYA
TIEMPO

rENSAVO
11 movo 416
•
400,,-
• ... 29 obr 11 364

.I
16 obrl 1 312
300,•
CUICUILCO
------- 21
3 i obr 11
marzo
260
208 arqueoastronómico
200-- 9 marzo 156 altiplano
100. ~ •••
25 febrero 104
de
TEOTIHUACAN P. S.
13 febrero 52
méxico ...,j
t
31 1
- - . - - . 1---+-----+"---+
enero
+ o - - - - · - - · - -· - - · - - · - - · - - - · - - · - - · --- --·-• - · - - . - - . 18 enero 52
• 6 enero 104 r ~
100' ~

CHOLULA
••• 24 diciembre 155 SIMBOLOGIA
11 diciembre 2.01
200, ..
• 28 noviembre 259

300••
• 16 noviembre 311
• FUEGO NUEVO (CICLO DE 52 AÑOS),
1 •• noviembre 363
3 SEGUN CRONOLOGIA
TEOTIHUACAN P.0 .
400, .. 21 octubre 415
• 8 octubre 467
i 500-- • 26 septiembre 519 ◊· REGISTRO SOLAR DE LA
• 13 septiembre ESTRUCTURA
600, ... • :1 ogJsto

100, ...
XOCHICALCO PSI:.
•
• IJ

25
7
aoosto
OQu~tO
J·1llo
·,27
·,79
I
sv
PERIODO DE ASENTAMIENTO

•
1. ■ ..
12
·~ ':I
Ju li o
juniv
831
883
{gj, S1
POSICION TERRESTRE C/OCTAVO
RESPECTO A LAS ESTACIONES :
E P Equinoccio de Primavera
1 ·: ·¡1,nio 93.:;
• 4 junio 987
SV Solstlclo de Verano

• 23 mayo 1 039
EO Equinoccio de Otollo

• 11 may o 1 091
SI Solstlclo de Invierno
10 0 •- • 29 abril 1 143
• lñ o Dr il
::::::::.. ORIENTACION DE LA ESTRUCTURA
20(, -
TE OPANZOLCL... -
• 248 3 atri 1 1
1 7° SEGUN C/ OCTAVO

CALIXL.AHUACA@l------------------------------------------t--+----+--l-3C-1;--t
21 marzv
3C C• • e \...
9 marzv 1 3
TENC'-CHr1wnM1N•'o•._-1....- - - - - - - - - - - - -- - - - - - -- - - - - - - - - - - - - - ------+--=-+..:.:.:.::=-='---+'----=_·.,=-4¿
T 25 febr .. ro 1 4C4
4 •-.,.:.k.JIZACHlEP E l ~ ' - - - - - - - - - a l - - - - - - - - - - - - - - - -- - - - - - - - - - - - - - - - - - - - - - - - - + - - - + - - - - + -- - - i r
13 febrer.:: 1 4 ~•-

+ 1 500- - . e
l 1 111
• - - - - -- •- • _ __ _ _ _ __ _ _ _ _ __ _ _ _ __ _ _ _ __ ___ __ _ _ _ _ :.:
31 ener v 1 50 8
-1_5_6_L-1
. . . R.::.EG-"'l-"-S-'-T'-'-R-'-O_S,=..Oc..cl~A-"-R-'---+---'----~♦
18 e•,er0
1omina. 2. DIAGRAMA TIEMPO
DE ruEGOS NUEVOS .
\/S. REGISTRO SOLAR.
SEGUN CRONOLOGIA

DE ALFONSO CASO
~--------__,
o
~
\.a. p o n c e d e I e n h . ~

Lámina 2 Diag rama tiempo vs . reg istro solar

 varia clones angulare & de las posiciones
desfosomiento
cado 52 años
preh i spdnicos
mo,imiento. de
tronslac,ón
solares en el horizonte c/52 años prehlap. ENSAYO
posiciones
terrestres
V O r i a ei o n e s a n g u 1 a r e s
6c
arqueoastronómico
3
1

posicione, terrestres 26 Iº 't!' 3º 4° 5º altiplano

ti erro . 2.9.__g_ .1 - . - . so ·,stt-~ITo ·----¡¡j--¡¡ • Tan" ,;r-. ~- ·- ·-
4 en su orbltaJ c/52 años 25
/
1 2 ~ -. .._......... ■ de
calendáricos prehipanicos 2
3
4
s
4
5
3

6
1

1
. ~-~■■••···•··
- - -
~ ..... - 1 ...
1
111 ......
méxico

6 ! 23; 6
7
7
. 8 -- --
1

· --. - ,· -.,
1

· - , - · - ,~-
1
·~
.. -~ SIMBOLOGIA
~
- 'fu • v V "' u

_.,~- -
8
9
10 11
9
10 1

.~- ............variaciones angulares.

mayores en primavera y
~
2
EQUINOCCIO _ _ . ___ ~ - _ . __ . _ _ . EQUINOCcJ - -
111 12
-■ en otoño, mlnimas en verooo
1

DE PRIMAVERA DE OTOÑO 12 13
.......... --- --·····••· - -····•--
1 1
y en invierno.
-· - -··
13 14

! :)'_:)
14
15
15
16
-
--- ............ ..........
-
- •-·- ·- · - · _-:
1
- - -variacion anoular (4.5°)
segun F. Ti chy.(ver 1am. 15)

1 20 - -
16
17
17
18
"'·
--... -.
.
~
...,.~--+Posicibn solar en el horizonte
poniente.
18 19 ~
19 1 ••4-
o 1 19' 2J:)
21
21
20

22 ~--· ~ -- · - ·-
1
. .......11 -
_, ___
,:;
~..,.....,..,Qiclon solar en el horlzon
- - -
, 11
¡ . ,aJ
17?
22 23
23 24
24 25
- 'f u \.o'- 11 IV
- 1
•11•--- -·-......$ - orien1e .

~
\ 25 26
26 27
27 28 ---·••111 ........-._, __
---~-~,~,-;; .
para los calculos azimutales
se considero:
28a 29 . ¡ 1
send=COSO·CO8 lat:
norte
lat . = 18.93° ( lat teoponzolco)
altura del horizonte= O
azimut no se representa graflcamente la
excentricidad de i°a orblta terrestre.

3.ACIMUT OC LA POSICOJ
1amina.
SQAR .EN EL t-mlZONTE
C..-52 ANOS PREHIS~NICOS.
a. p o n c e d e I e b n h.

Lamina 3 Acimut de la posición solar en el h,o rizonte c/52 años prehispánicos

 ENSAYO
arqueoastronómico
altiplano
de
méxico

SIMBOLOGIA
e t •CICLOS CALENDARICOS C/52AÑO
(segun cronologlo de A. Coso)
SENTIDO DE LAS ESTACIONES
PERIODOS DE ASENTAMIENTO'.
TLAP Tlopocoyo
CUI Culcuilco
PST P. del Sol Teotlhuacon
CHO Cholulo
PQT P. Quetzolcootl Teotlhuocon
<X PS E P. Serp. Emp .Xochicolco
'----+------,----+ Ña, TEOP Teoponzolco
?':U ...J CAL Colixtlohuoco
FTI TM Tempo Mayor Tenochtitlon
Q. "·« P F N P. del Fuego Nuevo lztopolopo
• CHAL Cholcotz ingo
- ■ J PX Juego de Peloto Xochlcolco
i lTENG Tenongo
1-11cu L Culhuocon
iffüJ TULA Tula
...Jt .TENA Tenoyuco
.:xl l.TLA T Tlotelolco

~= !]MTEPO P.del Tepozteco

MON CH I los Monjas Chiche~ ltza
.:x: C CO!...CHI Casa Colorado Ch1chen ltza
::e: CAS CH I Castillo Chichen I tzo
~ : CAS TUL Castillo Tulum
~ T G CH I T.de los Guerreros Chicheo ltzo
~- J P CHI J. de Peloto Chichen ltzo

"º·4.ESQUEMA DE FE-
IJmino
CHAMIENTQ ARQUEOASTR
NOMICO EN EL MEXICO
ANTIGUO. ,
ij 8 a. once de leon h

Lamina 4 Esquema de fecha m iento arqueoastronómico en el México a nt i g uo

NOTAS

1 Como el Observatorio Astronómico (Grupo E) en Uaxact ún Petén,

para determinar las fechas de los solsticios y los equinoccios
(Morley 1980). El Observatorio o Caracol en Chichen Itzá, para
determinar las fechas de equinoccios y la puesta lunar con máxima
declinaci6n el 22 de marzo (Ricketson 1928, en Aveni 1980), así
como los eventos de Venus y el sol que ocurren según 20 alineaciones
ar qui t e c t 6 n i ca s (Av en i , Gi b b s , Ha r tu n g 1 9 7 5 ) • Ta mb i é n la s
estructuras C y Den Xochicalco, señalando el ocultamiento solar en
los días de posición cenital con la línea visual desde la "Estela de
los glifos" hasta el canto de la base derecha de la estructura D.
(Aveni 1975). La orientación de estas estructuras también señala
el punto en el horizonte (Ponce de Le6n 1982) en el que el sol se
oculta 3 días después del equinoccio, señalando el cuarto
submúltiplo calendárico del año. Como las estaciones no son
iguales en número de días, el equinoccio no es la cuarta parte del
año, sino los días señalados por estas estructuras. También es
conocida históricamente la festividad de Tlacaxipehualiztli que se
hacía cuando el sol estaba en medio de Huichilobos, cuyo fen6meno
visual pudiera ser semejante al que se observa en Teopanzolco (véase
foto 4) y que en la Ciudad de México actualmente se registra
únicamente por sus calles (véase foto 7). Estelas 10 y 12 en Copan
Honduras (Avení 1980) cuyo eje señala el ocultamiento del sol los
días 12 de abril y 1 de septiembre.

2 Los e'jes oriente-poniente, en la' gráfica que resume las

orientaciones en la América antigua (Avení 1980).

3 Se le ha llamado también "posici6n solar en el horizonte"

(Tichy 1978).

4 Algunas estructuras (principalmente en el Preclásico) están

orientadas hacia el norte del oriente, es decir registran en
primavera y verano, pero al amanecer.

5 Esto difiere del concepto tradicional que desde los primeros

cronistas se ha conocido (Carrasco 1980 y Caso 1968).

6 Juan Palacios en 1934 en su estudio sobre el principio del

siglo indígena t ·a mpoco incluye el día bisiesto, pero como el
portador lo considera el principio del año calendárico, la fecha
del Fuego Nuevo último en Mesoamérica le resulta el 2 Acatl, 7 de
mayo (juliano) de 1507, después al transportar la fiesta del 2
Acatl al 4 Acatl (según Chimalpain) obtiene el 26 de julio del mismo
año, (80 días después) día del segundo paso cenital en Tenayuca.

87
7 Difiere del dato (1507) que tradicionalme n te se ha conocido,
porque el portador del afio 3 Calli (afio en q u e es conquistada
definitivamente Tenochtitlan) resulta en 1522 y no en 1521, ya que
al considerar este portador al final y no al principio del año
prehispánico, resulta que al año cr i stiano de 1521, termina 18
días antes del día 3 Calli por lo que éste cae al 18 de enero de
1522 (juliano), equivalente al 27 de enero del año gregoriano de
1522, continuando con esta correlaci6n (sin la inclusión de días
bisiestos) se deduce que la fecha cristiana en que se celebr6 el
ú 1 timo Fuego Nuevo, que dio nombre al anterior ciclo de 5 2 anos,
fue el día 31 de enero del año 1508.

8 Tezozomoc (García 197 5) señala los lugares donde descendió el

Fuego Nuevo o se "ataron los años": 1090 en Acahualitzingo; 1142
en Coatepec; 1194 en Tecpayoacan-- hasta aquí en años Cetochtli--
1247 en Chapultepec; 1299 en Culhuacan y de la sexta en adelante,
ya no las precisa. Chimalpain dice (García 197 5): "nuestros años
se atan en la cumbre del Huizachtecatl, cae-- el fuego-- en el
encendedor de barrena".

9 Si el portador del año 2 Acatl, fuese el día último de los

Nemontemi (25 de febrero de 1404) entonces el registro solar de la
estructura cinco días antes 10 Tochtli (20 de febrero de 1404),
tendría el lugar último del mes XVIII que en la cronología de Caso
lo ocupa el portador 2 Acatl.

10 ¿Podría ser acaso ésta una de las razones de la pe regr inaci6n

azteca~

11 Los valores dados hasta antes de las excavaciones han sido: 7°

30' (Aveni 1977) y 7° 00' (Tichy 1978). Los resultados obtenidos
en marzo de 1981 por quien esto escribe, en la realización de las
mediciones acimutales de los ejes principales y de las estructuras
visibles (en esa etapa de las investigaciones) del Templo Mayor y
las diferencias que entre sí guardaban entonces, son los
siguientes. Para la segunda fase constructiva (misma que cuenta,
aún con los dos templetes) se midieron dos ejes. El eje superior
que pasa en medio de los dos templos, y el eje virtual inferior que
va de la parte media inferior (entre las dos escaleras en el nivel
de desplante de éstas) hacia la parte media de la entre calle
(ranura o junta) opuesta a las escalera~ en el nivel de desplante
actual del monumento. Para el primero el valor resultante fue de
97° 25' (medido a partir del norte astronómico) y para el segundo de
98° 48'. Es de considerar que el desplome que presentan las
fachadas oriente y poniente de dicha fase constructiva fue motivado
por el "bufamiento" del terreno cuyo comportamiento no ha sido
homogéneo en toda la superficie de las excavaciones, encontrándose
que la mayor elevación o bufamiento se ha registrado en la parte
sureste del monumento, razón por la cual se ha propiciado una
diferencia acimutal entre los ejes mencionados de la estructura.
Para las últimas etapas constructivas de la fachada principal,
consideradas a partir de la tercera fase, el eje que se ha logrado

88
medir hasta la fecha es el que va sobre la parte media superior
(entre las dos escaleras) de cada una de las estructuras. Este eje
presenta una medici6n de 96° 02' con una diferencia aproximada de
- 1° 23' con respecto al eje superior de la segunda fase y de - 2°
46' con relaci6n al eje interior de la misma. Si consideramos que
estas últimas fases fueron afectadas también por el levantamiento
del terreno al faltar la carga con la que estaba trabajando antes de
iniciar las excavaciones, es de suponer que exista un esviaje
(semejante al existente en la segunda fase) entre el eje superior,
con respecto al inferior o sea que si a la medici.6n realizada de 96º
02' le adicionamos la diferencia obtenida (1° 23') que existe con
respecto a la segunda fase, nos dará un valor para este eje
superior de 97° 25'. Por otro lado es congruente que los ejes
principales de la traza colonial de la Ciudad de México tuviesen la
misma orientaci6n que la última fase constructiva prehispánica o sea
97° 3,0' (promedio) que es práct:'i.elá:mehte el mismo de 97° 25'
supuesto por comparaci6n deductiva con respecto a la segunda fase.

Conclusiones:

·I · El eje del pasillo de la segunda fase (perimetral) se puede

considerar con un valor original cercano a los 98° 48'.

II Los muros oriente y poniente tienen 7° 30', los muros norte y

sur tienen 97° 30'; por lo que el pasillo central queda descuadrado
con respecto a la estructura.

III El eje, desde la tercera fase hasta la última fase, se puede

considerar con un valor original a los 97° 25'.

IV El Templo Mayor presenta un esviaje en los ejes oriente-

poniente de la segunda fase (pasillo central) con respecto al eje
que comprende las fases gosteriores (de la tercera a la última) cuyo
valor aproximado es de 1 23'.

V Pensando en que la orientaci6n del pasillo central de la

segunda fase señalará la posici6n solar en el horizonte, serían
cuatro los días señalados: al oriente 28-29 de febrero y 10 de
octubre y al poniente el 24 de septiembre y el 10 de abril. Visto
esto bajo el contexto establecido (véase lámina 4) la primera fecha
nos indicaría que la segunda fase se habría construído por el año
1400 d.C., después de la 3a. fase, lo cual es incongruente. La
segunda fecha nos indicaría que esta fase se hubiese construído por
el 500 d.C., cosa que también es improbable, pero la cuarta fecha,
por el año 1200 d.C. se aproxima a lo que mencionan algunos
investigadores (Vega 1979) sobre el asentamiento anterior a la
fundaci6n por los aztecas.

12 Además existen varios ejes en la Ciudad de México, cuyo trazo

urbano es herencia en gran medida de la antigua ciudad de
Tenochtitlan, que presenta como muchos sitios en Mesoamérica una
relaci6n de ubicaci6n con respecto a otros lugares. Así, el
lugar del Templo Mayor se encuentra definido posiblemente por tres
ejes. El de la original Calzada de Tenayuca (González 1973), hoy
Calzada de Vallejo, que unía la á
Pit. mide de Tenayuca, Tla teloco,

89
el Templo Mayor de Teno c. htitlan, el Capulli, Zoquiapan, la que
fuera isleta de Acolco, y remataba en el centro de Huixachti tlan,
uniendo estos sitios en forma simb6lica y física. Otro eje que
posiblemente define la ubicación de esta antigua ciudad (véase plano
1) es el que contiene la traza del centro ceremonial y ciudad de
Cholula en Puebla (Ponce de Le6n 1982), cuyo rumbo astron6mico
señala el solsticio de invierno, este eje después de señalar en el
porteaguas de la Cuenca, el cerro Tehuicocone como mojonera
natural, cruza por el Templo Mayor. El tercer eje propuesto para
esta ciudad (véase plano 1), es aquel que va del Templo Mayor al
cerro de Chiquihuite (Ponce de León 1982), que definiría el norte
astronómico, es decir , las estrellas situadas arriba de la cumbre
del cerro girarían alrededor de un punto situado sobre este mismo,
visto desde el Templo Mayor.

13 Esta fecha ya había sido mencionada, así como las

correspondientes, 10 de octubre, 6 de septiembre y 6 de abril
(Tichy 1978).

14 Según las fuentes el año fue 1325, pero, aplicando la

cronología de Alfonso Caso, el año 2 Calli es el 1326 y esto es por
la posición del portador del año (véase nota 7).

15 La orientaci6 n de los d i ferentes paños por ejemplo en la

Pirámide de Quetzalco atl varían de oº a 2° 17' al sur del oriente,
por lo que tomamos el promedio entre 1° 12' y 1° 50' (Avení 1980).

16 El otro eje, norte-sur, de esta estructura cruza por los

siguientes sitios (Ponce de Le6n 1982): principia al sur en el
cerro Acatipa (véase plano 1) que contiene restos arqueológicos,
sigue al norte por la Pirámide de Teopanzolco, continúa más al
norte por el volcán Tres Cruces (por el que también cruza el eje
norte-sur de la Serpiente Emplumada de Xochicalco) (véase nota 28) y
termina en la Pirámide de Téquipa al sureste del pueblo del Ajusco.
De esta forma, este eje une tres sitios arqueol6gicos, dos en el
Valle de Cuauhnahuac y uno en el Valle de México. Esta unión de
tres sitios arqueológicos, por el mismo eje, no es el único caso
en el Altiplano, ya que Teotenango, Cuicuilco y la Pirámide del
Fuego Nuevo en Iztapalapa (Ponce de León 1982) son unidos por un eje
sols t icial (véase nota 2 O). El mismo caso es el eje que une
Malinalco, La Pirámide de Palpan de Baranda y Xochicalco (Ponce de
León 1982). El significado de esta unión, . quizá simb6lica, de
sitios arqueológicos de diferentes épocas culturales, no la sabemos
pero dudamos sea circunstancial.

17 En general la orientación de las estructuras piramidales en

Mesoamérica se presenta de la siguiente forma. Cuando es al
oriente, señala hacia el su.r-o'riente y cuando es al poniente señala
nor-poniente, pero Tlapacoya, como algunas otras, mira hacia el
nor-oriente y al sur-poniente. Este grupo de orientaciones tienen
en común el pertenecer al Preclásico (véase nota 4).

90
18 Como la Pirámide de Tlapacoya está en la ladera nor-oriente del
cerro no es posible visar el horizonte poniente, ya que lo obstruye
el cerro, siendo visible únicamente el horizonte oriente.

19 Si bien el período de asentamiento en Tlapacoya se prolonga

hasta el período azteca, el inicio de éste es en el Preclásico
(Pifia 1963).

20 Se puede decir que en la Pirámide redonda de Cuicuilco se

manifiestan tres tipos de ejes: el que señalan las rampas oriente-
poniente, oº 30', el de la tumba superior 18° aprox. y el de la
tumba inferior con 26° aprox. al norte del oriente los tres. Por
la diferencia de ejes en la misma estructura y bajo el contexto de
la variaci6n de orientaciones conforme pasa el tiempo, parece
congruente el que la tumba inferior, construída primeramente,
registre solisticialmente en junio (como el eje Iztapalapa-
Cuicuilco-Teot enango) (véase nota 16), la tumba superior
(construída después) registre en mayo y la estructura que presenta
las rampas (construídas en los últimos años de su ocupaci6n)
registre en marzo, siguiendo el sentido inverso a las estaciones.

21 La rampa poniente presenta tres escalones cuyas orientaciones

varían de 2° a 8° al oriente del norte, por lo que las
perpendiculares a estas terrazas serían de 92° a 98° aprox., que
coinciden con los valores asignados a esta pirámide: 97° 38', 95°
42' y 91° 43' (Avení 1980).

22 En el Altiplano no se han identificado estructuras que

registren la posición solar en el horizonte exactamente el día del
equinoccio; las estructuras con registro más cercano pueden ser:
el Juego de Pelota en Xochicalco (véanse fotos 16, 17 y 18), las
estructuras C y D también en Xochicalco, la Pirámide que se
encuentra junto a la Alberca de Villa Olímpica, que aún cuenta con
el desplante de muros estucados y cuya orientaci6n dif~ere en 0.5
grado al oriente poniente astronómico. Esto nos hace pensar en la
divisi6n del "año numérico" o "días de mitad del año" (Tichy 1978;
Carrasco 1980) correspondientes al 24 de marzo y 20 de septiembre,
intervalo más sencillo entre el solsticio estival e invernal. El
Templo Mayor de Tenochtitlan pudo haber registrado el equinoccio
mediante la posici6n solar con 22° de altura sobre el horizonte
(Avení 1978; véase también foto 4). En la zona maya sí parecen
existir edificios que registran este día, y son bastante conocidos
en Chichen Itzá, Uaxactún etc.

23 Actualmente la efeméride solar oriente de la estructura más

importante en Villa Olímpica, que aún cuenta con paños de muros
confiables, desgraciadamente no se puede observar, pues la
obstruyen los edificios colindantes a la Pirámide; pero en base al
cálculo hecho mediante la resolución del triángulo astronómico, por
lecturas solares, se puede decir que este registro solar se lleva a

91
cabo también tres días después del equinoccio de primavera (véase
nota 22.).

24 Mediante análisis de fotos aéreas estereosc6pic as y

cartografías (SDM. SPP. CAVM) se puede apreciar, en conjunto, una
desviaci6n en la orientaci6n de la Pirámide del Sol con respecto a
la Calzada de los Muertos o inclusive con la Avenida Este-Oeste.
Pero la comprobaci6n física, mediante la medición topográfica que
hemos llevado a cabo, refleja variaciones en los diferentes paños.

25 Parece coincidir con la orientación de la Ciudadela, 16º 55'

al sur del oriente (Aveni 1980). Millón (1968) le atribuye la
misma orientación a la Pirámide del Sol y a la Calzada de los
Muertos. Marquina (1964) le asigna aproximadamen te 17º; Aveni
(1980) no le asigna valor y Gaitán (1974) le asigna 16º 30'.

26 Para este tipo de orientaciones (17°) se le asigna como fechas

de registro (Tichy 1978) el 4 de febrero, 5 de mayo, 8 de agosto y
7 de noviembre, que definen ocho períodos del calendario numérico
(Carrasco 1980).

27 Tichy le asigna 25° (Tichy 1978) y Avení seg6n varias faces:

26º 16', 24° 15', 23º 47', 25° 32' y 25º 38'.

28 El eje oriente-ponient e de esta pirámide (véase plano 1) queda

señalado al oriente por el cerro Jumiltepec (Ponce de León 1982)
cuya cumbre· contiene restos de edificios prehispánicos rodeados por
un foso perimetral que circunda el sitio. La cornisa norte mide
105° 39', la sur 106° 09', lo que promedia 105° 54' ( 106 aprox.),
lo cual concuerda con el registro solar al oriente, constatado
fotográficame nte (véanse fotos 1, 2 y 3). Por fases Aveni mide
276° 45' y 277° 13' (Aveni 1980). De forma semejante a la Pirámide
de Teopanzolco el eje norte de esta pirámide senala al Cerro de Tres
Cumbres, en la parte más alta de la sierra que separa el Valle de
Cuauhnahuac del Valle de México (véase nota 16).

29 Al poniente de la Pirámide de la Serpiente Emplumada, se

encuentra una estructura más alta cuya orientación varía 0.5 grado
hacia el norte del poniente con respecto a la Pirámide de la
Serpiente Emplumada. La razón de esto podría ser por la
diferencia de altura entre las dos estructuras, ya que el sol
cuando se oculta el 5 de agosto, visto desde la Pirámide Serpiente
Emplumada se alinea al eje de ésta antes de ocultarse, pues desde
ésta no se observa el horizonte; en cambio, desde la estructura
poniente, sí se observa el horizonte; pero cuando el sol ya se
oculta con 0.5 grado más de orientaci6n al norte del poniente.

30 En 1508 años solares medios (de 365.2422 días) el n6mero

total de días serán 550,785.2376 y en 1508 años calendáricos

92
prehispánicos (de 365 días) serán 550,420 días siendo la diferencia
365.2376 días. Es decir, si cada 52 años se hacía una correcci6n
en los "registros solares" (véase lámina 3), cambiando su
orientaci6n con respecto a la anterior, se puede considerar que el
calendario prehispánico sí contenía co r recciones para ir acorde con
el movimiento solar; pero este ajuste no era el estilo "europeo",
ya que no 1~ admitía su sistema combinatorio (Palacios 1934; Caso
1968; Carrasco 1980). Su elemento correctivo era la medici6n del
número de días del desfasamiento entre el calendario de 365 días y
el año solar medio de esta forma si se tenía conocimiento de los
días que se desfasaba el calendario con respecto a las estaciones
por lo que sí se podía n prever los "ciclos agrícolas".

31 La diferencia de 12.594 días en la posición solar en el

horizonte produce una variación angular en las orientaciones de la
Pirámide, que no es homogénea (véase lámina 3), pues el
movimiento aparente del sol en el horizonte es mayor en los
equinoccios que en los solsticios. Esto es al considerar como
unidad base el "tiempo", diferente a la unidad "ángulo" propuesta
por Tichy (1978).

32 Al menos la calendárica mesoamericana puede ser considerada

como evidencia de la actividad astronómica (Gibbs 1977).

33 El registro solar en Tenayuca es el 7 de mayo, que difiere en

80 días al supuesto (26 de julio) por Marquina (1934). Este autor
le asigna una orientaci6n de 20° 47' al norte del poniente, la
estructura en realidad tiene 17° 42' (Aveni 1980), es decir 3°
menos; por lo que el día de paso solar cenital, el sol no es
señalado en el horizonte por el eje de la Pirámide.

34 El registro del sitio sería al poniente, con la posición solar

senalada minutos antes de su ocultación, semejante a la Pirámide de
la Serpiente Emplumada en Xochicalco (véase nota 28).

35 El eje poniente con 17° 42' al norte del poniente (Aveni 1980)
registra casi a mediados de mayo. Pero al oriente, por la altura
del cerro (lo mismo que en Teotenango y en la Pirámide de la
Serpiente Emplumada en Xochicalco; véanse notas 28 y 34), el sol
incide sobre el eje a mediados de febrero. El eje norte-sur de
esta estructura señala al cerro Chichiltepetl.

36 Aquí en Tula, lo mismo que en Xochicalco, Teotenango y

Tepoztlan (véanse notas 28, 34 y 35), por estar al horizonte-
poniente más alto de lo común el registro es en los últimos días de
julio y no a principios de agosto, según los 17° aprox. al norte
del horizonte.

93
37 El convento de Culhuacan está edificado sobre restos
arqueol6gicos, cuya cerámica abarca desde Teotihuacan IV hasta el
Azteca I (Sejourné 1970). Desde la colonia, iglesia y convento
figuran como parte del pueblo de Culhuacan. Puesto que los
conquistadores elevaban siempre las iglesias en lugares ya
consagrados, con la intención de transferir sobre ellas un antiguo
prestigio, es improbable que monumentos tan costosos hayan sido
edificados en un lugar abandonado, desprovisto por consecuencia de
fuerza social. Es seguro que el eje de las ruinas de la primera
iglesia sea el mismo que el de la estructura prehispánica, como en
otros muchos sitios, ya que los ejes de trazo de los sitios
coloniales vienen siendo los mismos que los prehispánicos, como en
la Ciudad de México. La orientaci6n de las ruinas de la primera
iglesia, 18° 50' aprox. al norte del poniente, registra los
primeros días de agosto. La fecha que de la fundación de Culhuacan
transmite Chimalpain (Sejourné 1970) 670-10 Tochtli conviene
perfectamente al Teotihuacan IV (Sejourné 1970) y debe anteceder en
50 años aprox. a la celebración del ciclo calendárico en ese sitio,
según gráfica (véase lámina 2) el 7 de agosto del año 727 d.C., lo
cual concuerda con lo planteado (véase lámina 4).

38 La funci6n asignada a estos elementos ha sido la de ser

recintos para sacerdotes o personajes principales (Pina 1980).

39 Dado que nuestra gráfica (véase lámina 4) no reporta años

anteriores al 500 a.c., para encontrar a qué año corresponde el
registro solar, el cálculo lo hacemos de la siguiente forma: los
primeros días de septiembre en nuestra gráfica (véase lámina 4)
serían registradas por el año 600 d.C.; como los registros se
repiten cad·a 1508 años (véase lámina 3) el registro de la estructura
debía de ser 1500 años aprox. antes del 600 d.C., es decir, el 900
a .c.
BIBLIOGRAFIA

Avení, A.F.

Archaeoastron omy in pre-Columbian America. Austin:

University of Texas Press. (1975)

Astronomía en la América Antigua. México: S. XXI. (1980)

Skywatchers of Ancient Mexico. Austin y Londres: University

of Texas Press. (1980)

Broda, J.

"Cosmovisión y estructuras de poder en el México prehispánico".

Comunicacione s (Proyecto Puebla-Tlaxca la, México), 15.
(1978)

Carrasco, P.

"La sociedad mexicana antes de la Conquista", en Historia

general de México, I. México: Colegio de México. (1980)

Caso, A.

Los calendarios prehispánicos. México: UNAM. (1968)

El pueblo del sol. México: F CE. (1978)

Closs, M.P.

"El mecanismo para la determinaci6n de fechas en la tabla de

Venus del C6dice de Dresden", en Avení ( 1980).

Everett, R.B.

La gran inundación. Vida x._ sociedad en la Ciudad de México

(16z"9=i638). México: SE P. (1975)

Escalona, R.A.

Cronología x._ astronomía maya-mexica. México. (1940)

Federico, T.

"Simbolismo del Juego de Pelota en el Popolvuh", en Estudios de

cultura maya. México: UNAM. (1973)

Fujiyoshi, Y.

El enigma de la cultura de megalitos x._ el calendario solar.

Japón. ( 1981_)_

95
Gaitán, M.

"La triple cruz astronómica de Teot ihuacan", en XLI Congreso

de Americanistas. México. (1974)

Gibbs, S.

"La calendárica mesoamericana como evidencia de actividad

astron6mica", en Avení (1980).

González, A. L.

Plano reconstructivo de~ región de Tenochtitlan. México:

INAH. (1973)

Haberland, W.

Cultura de la América indígena, Mesoamérica central. México:

F CE.

Hartung, H.

"El ordenamiento espacial en los conjuntos arquitectónicos

mesoamericanos", en Comunicaciones (Proyecto Puebla-Tlaxcala,
México), 15. (1978)

"Arquitectura y planificación entre los antiguos mayas.

Posibilidades y limitaciones para los estudios astron6micos",
en Avení (1980).

Kurt, B.J.

Estudios estratigráficos en Tlapacoya, Estado de México.

México: SE P. INAH. (1978)

León Portilla, M.

De Teotihuacan a los aztecas. México: UNAM. (1977)

-- ---- - -- -----
México-Tenochtitlan: su espacio x_ tiempo sagrados. México:
INAH. (1978)

López, A.

Enterramientos humanos de la zona arqueológica de Cholula,

Puebla. México: S E P. -INAH.--(1976)

Lorenzo, A.

Uso~ interpretación del calendario azteca. México: Porrúa,

S.A.. (1979)

Martínez, H.J.

Significación cronológica de los ciclos mayas. Mérida,

 Yucatán. (1928)

Mar quina, I.

Arquitectura prehispánica. México: INAH. (1964)

Matos, M.E.

"El proyecto Templo Mayor", en Ciencia x._ desarrollo (México),

34. (1979)

Una visita al Templo Mayor. México: INAH. (19 81)

Menzel, D.H.

Guía de campo de las estrellas y los planetas. España: Ed.

Omega-:-s.A. (1967) ---

Millón, R.

"Urbaniza t ion a t Teo ti huaca n. The Teot ihuacan Mapping

Pro ject", en Proceedings ~ the E_ International Congress of
Americanists. (1968)

Morley, S.G.

La civilización maya. México: F CE. (1980)

Noriega, R.

La Piedra del Sol x._ _!_§_ ciclografías astron6micas del México

antiguo. México.

Nuttall, Z.

"La observaci6 n del paso por el zenit por los antiguos

habitantes de la América tropical", en Pub. S E P. 17, 2O
(México). (1928)

Palacios, E.J.

"La orientación de la Pirámide de Tenayuca y el principio · del

año del siglo indígena", en 25° Congreso Internacional de
Americanistas. (1934)

"Las fechas de Xochicalco de la Piedra del Sol y del Códice

vaticano A ". México. ( 1947)

Palerm, A.

Obras hidráulicas. prehispánicas en el sistema lacustre del

Valle de México. SE P. INAH. (1973)
Piña Chan, R.

Ciudades arqueológicas de México. México: INAH. (1963)

97
 Chichen Itzá, la ciudad de los brujos del agua. México: FCE.
(1980)

Ponce de León, A.

"Arqueoastronomía en el Altiplano de México". Gaceta. México:

UNAM. (1980)

Popenhoe, H.

"An Astronomical Calendar in a Portien of the Madrid Codex",

en Avení (1975).

Reyes, V.M.

"Trazado de los nxmumentos mayas", en Escalona ( 1940).

Rojas, R.T.

Nuevas noticias sobre las obras hidráulicas prehispánicas y

el Valle de México.
en -
coloniales - México: SE P. INAIL
(1974)

Ruz, L .A.

"Más datos históricos en las inscripciones de Palenque", en

Estudios de cultura maya, 9. (1973)

"Obras de divulgación sobre la cultura maya", en Antropología

e Historia (INAH), 21. (1978)

Saenz, C.A.

El fuego nuevo. México: INAH. (1967)

Sahagún, F.B.

Historia general de las cosas de la Nueva España. México:

Porrúa. (1975) - - - ----

Sejourné, L.

"Arqueología del Valle de México, . Culhuacan". V. I. México:

INAH. ( 1970)

Smiley, C. H.

•~eroglíficos mayas asociados al sol, la luna y los planetas",

en Estudios de cultura maya (México: UNAM), V. IX. (1973)

Soustelle, J.

La vida cotidiana de los aztecas. México: F C E. ( 1956)

- - - - - - - - -- - -

98
Tichy F.

"El calendario solar como princ1p10 de organización del espacio

para poblaciones y lugares sagrados", en Comunicaciones
(Proyecto Puebla-Tlaxcala, México), 15. (1978)

Toussaint, M.

Planos de la Ciudad de México. Siglos XVI y XVII. Estudio

hist6rico ,-urbanístico y_ bi bliográf ic0:- - México: INAH.
(1938)

Vega, S.C.

El recinto sagrado de México-Tenochtitlan. Excavaciones 1968-

69 y_ 1975-76. México: INAH. (1979)

Wolf, E.

Pueblos y_ culturas de Mesoamérica. México: Ed. Era. (19 6 7)

99
GORDON WHITTAKER

The Structure of the Zapotec Calendar

Zapotec civilization, radiating from Monte Alban in the Valley

of Oaxaca, spans virtually the entire stretch of Mesoamerican
history from just after the end of the Olmec era to the rise of the
Aztecs, a space of about 2000 years. Situated between the Mixe-
Zoquean and Mayan cultures of Southern Mesoamerica and the di verse
cultures of the North, it could well provide the key to an
understanding of the development of such a significant pan-
Mesoamerican trait as the 260-day divinatory calendar, for which
there is abundant evidence in the Preclassic and Classic monuments
of Monte Alban.

The question of whether or not the Olmecs were the inventors of

the divinatory calendar will be left aside here. Suffice it to say
that the earliest evidence (to date) for its use comes from
Oxtotitlan in Guerrero in a clear Olmec context dated between 900
and 700 B.C. (Grove 1970: 18-20, 32). The item concerned is a
painting on a cave wall depicting a cipactli-like head around which
are six circles or dots (Fig. l)~ It must be emphasized that an
interpretation of this as a day-date or calendrical narre 6 Dragon
(Cipactli) . is by no means firm, but the likelihood that it is
indeed such is strong, and there are no compelling alternatives to
this interpretatiog.

At this point it would be appropriate to review the basic

features of the Mesoamerican calendar. The latter is composed of a
sequence of numbers 1 to 13 revolving endlessly in concert with a
set of 20 day-names, the resulting 260 combinations of which are
grouped in 20 sets of 13 days called trecenas (Fig. 2). The first
day in the calendar is traditionally the equivalent of the Aztec 1
Cipactli, although other beginning points, notably Acatl, -are
attested in some areas. The numeral coefficient is usually placed
before the day-name, but this is reversed in the Zapotec system and
occasionally elsewhere.

The naming of years comes from the application of the repeating

260 day-sequence to a conventionalised, and invariable, solar year
of 365 days. In the solar calendar days are grouped in 18 sets
(360 days in all). Following these 18 "months", as they are
called, are 5 'empty' days, the nemontemi of the Aztecs and the
uayeb of the Maya. The solar year is named after the day in the
divinatory calendar falling either:

(a) on the day following the nemontemi,

(b) on the last day before the nemontemi,

1Q1
 0

Fig 1 Painting 3, Oxtitlan (after Grove 1970:

Figs. 14 and 15)

Fig. 3 Monument 3, San Jose Mogote (Marcus 1976: Fig.2)

W2
 Fig. 2 The divinatory calendar

I Cipactli 10 8 2 9 3 10 4 11 5 12 6 13 7

II Ehecatl 2 9 3 10 4 11 5 12 6 13 7 ®a
III calli 3 10 4 11 5 12 6 13 7 {iJ>a 2 9

r.J Cuetzpalin 4 11 5 12 6 13 7 {iv a 2 9 3 10

V Coatl 5 12 6 13 7 10a 2 9 3 10 4 11

VI Miquiztli 6 13 7 l:iJa 2 9 3 10 4 11 5 12

VII Mazatl 7 lJ)a 2 9 3 10 4 11 5 12 6 13

VIII Tochtli 8 2 9 3 10 4 11 5 12 6 13 7 §0
IX Atl 9 3 10 4 11 5 12 6 13 7 10a 2

X Itzcuintli 10 4 11 5 12 6 13 7 Ga 2 9 3

XI Ozana.tli 11 5 12 6 13 7 ®a 2 9 3 10 4

XII Malinal.li 12 6 13 7 {})a 2 9 3 10 4 11 5

XIII Acatl 13 7 Ga 2 9 3 10 4 11 5 12 6

XIV Ocelotl 10a 2 9 3 10 4 11 5 12 6 13 7

~ Cuauhtli 2 9 3 10 4 11 5 12 6 13 7 liVa
XVI Cozcacuauhtli 3 10 4 11 5 12 6 13 7 ®a 2 9

XVII Olin 4 11 5 12 6 13 7 1®a 2 9 3 io

XVIII Tecpatl 5 12 6 13 7 @a 2 9 3 10 4 11

XIX Quiahuitl 6 13 7 1Ga 2 9 3 10 4 11 5 12

XX Xochitl 7 1©a 2 9 3 10 4 11 5 12 6 13

The m.nbers in colums a:::rtbine with eadl day--naoo to label the days.
!lle numbers within circles indicate the nurrber of eadl trecena in the
2~ay cycle.

103
 or (c) on the last day of the nemontemi.

Because of the five extra days the day-name labelling a given

year will always be situated 5 positions further than the one naming
the previous year. This results in four possible year-names
combined with the entire sequence of numbers 1 to 13-- a set of 52
combinations. all told. Day-names in their function as year-names
are referred to (with and without accompanying numerals) as year-
bearers. In the majority of Mesoamerican calendars the year-
bearers fall in positions 3 (Calli), 8 (Tochtli), 13 (Acatl) and
18 (Tecpatl) of the day-name series, alternatives one position up
or down occurring in certain areas or in certain periods.

The Calendar of the Preclassic

For the first conclusive examples of the use of the 260-day

calendar we must turn to the Valley of Oaxaca, to the sites of San
Jose Mogote and Monte Alban. A carved stone slab excavated at San
Jose a few years ago has been proclaimed as the first documented use
of the calendar, dating to the Middle Preclassic, c. 600 B.C.
(Marcus 1976: 44-45). Between the feet of a figure resembling the
so-called danzantes of preclassic Monte Alban appears a date read by
Marcus as "l Earthquake" (Fig. 3).

If taken alone, the style of the calendar hieroglyphs below

the San Jose danzante would suggest a date no earlier than the
P r o t o - C1 a s s i c , beg inn in g around 2 0 0 A. D. w i th the Firs t
Intermediate Period (Transici6n II-IIIA) and Monte Alban Period IIIA
(c. 250-450 A.D.). The sign for the numeral 1 is embellished with
a curved base and the U-bracket, a hallmark of the Classic
unattested in earlier periods. Furthermore, the day-sign
identified by Marcus as the 17th Zapotec day-name "Earthquake"
takes a form here which is not attested before the Monte Alban
Classic. In fact, a virtual duplicate is found on the stone
marker excavated in the Zapotec quarter of Teotihuacan, which is
some thousand years later than the date Marcus assigns to the San
Jose slab, coming as it does from Monte Alban IIIB (c. 450-700
A. D.).

Caution is, moreover, advisable in accepting Marcus'

translation of the 16th-century Zapotec day-name Xoo "Fury,
Violence, Turbulence" as "Earthquake", which is only one of the
extended meanings of the word (it is, for example, used also in
connection with storms). Marcus' translation, influenced
understandably by the frequent seismological association of the
equivalent day, and related sign, in other Mesoamerican calendars,
is brought into question by the early iconographic use of the day-
sign to represent falling water, a torrential downpour, or the
like. The essentially aqueous nature of this sign is further
attested on a Classic Veracruz figurine and in a mural on House Eat
Palenque, where it is rendered in blue with attendant drop elements
(cf. Seler 1915: 428, 482-86).

Since officials of the Mexican Institute Nacional de

Antropologia e Historia have informed me of some uncertainties

'1 04
regarding the exact chronological context of the stone slab as
excavated by Flannery and Marcus, it would seem best to withhold
final judgement on the matter until such questions are resolved.
Nevertheless, given the fact that we are dealing with glyphs on a
danzante-style monument, I would argue, albeit tentatively, for a
terminal Late Preclassic date.

With Monte Alban I (c. 500-200 B.C.) the first indisputable

evidence for the Zapotec, and Mesoamerican, calendar comes to
light in the form of five inscribed stelae at Monte Alban.
Situated alongside the earliest level of the Danzante wall (Caso
1947), these monuments are assigned to the middle of the period
(Scott 1978: 15, 42, 68-71). With the exception of stelae 14 all
record dates in the divinatory calendar, Stelae 12, 13, and 17
(Fig. 4) providing at the same time evidence for the solar calendar.

Alfonso Caso, whose incisive studies of the hieroglyphics of

Monte Alban (1928; 1947; 1965) laid the groundwork for research on
the devlopment of the Zapotec calendrical sys tern, was confronted
here with the major problem of differentiating between the various
kinds of calendrical sign one might expect to be represented in a
Mesoamerican writing system. Foremost was the task of separating
day-signs from signs for larger units such as years, months and
four- or five-part divisions of the divinatory calendar.

The Year-bearers

With the archaic texts of Stelae 12 and 13, plus the much
later stelae of Classic Monte Alban at his disposal, Caso set about
determining the value of an important glyph which he called the
"head of Cocijo" (the god of rain and lightning; see Fig. 5) or,
more accurately, ''headdress of Cocijo" (Caso 1928: 45-49; 1947:
28-29). Noting that one set of glyphs always occurs with numerals,
and that the number is never higher than 13, Caso concluded that
the signs in question must be day-names. Since a bare handful of
such signs also appeared below the headdress element of Caso's ''head
of Cocijo", usually at the beginning of a column, he had little
hesitation in identifying these as year-signs (1928: 45-59).

Unable to match up the Zapotec year-bearers with the Mixtec and

Aztec series (Calli/House, Tochtli/Rabbit, Acati/Reed,
Tecpatl/Flint), Caso was drawn to an alignment one series higher
(Ehecati/Wind, Mazatl/Deer, Malinaili/Twi.sted Grass,
Olin/Movement) by his recognition of an apparent deer's head
occurring with the year-sign on Classic Period monuments (1928: 54),
and by the existence of such a series among the Postclassic Cuicatec
and, alledgedly, Classic Maya.

This deer glyph (Caso's Glyph G) is, however, two signs,

consisting of the heads of a deer and of a rabbit or hare. Both
share a lolling tongue, which Caso took to be the hallmark of the
deer alone, whereas in fact the latter's antlers might be taken as
a better diagnostic feature-- at least as far as Zapotec
inscriptions are concerned.

105
 13

:~{[f.:·.-., -~:

:;:;' "' ~
ti ... ~11,:lt!t~J
··•••· ,;c··se"'G' : .' !'(j; ~;'.~
. · ·:•f i·
• ""
.;t'
..,
;} F.~
·"'"· '
~tf
,~. /<1$;

., t~. ·.:. ,fNJ~

,, - ::>'-i;if)
:-, ,, . i41
~ 171 ;;..KF-;;;.c!lf-' -~~....._=.....,. ·.: ~
,~
i~,.
1-i
-----
~~~..a!¼
I·1
..
-~- 4-
~
·~
~, •. . ~
, '-

Stela 15
~ ~-':....:~ ' . ;-. •

Fig. 4 The Stela e o f Mo n te Al ban I (Caso 1947: Figs. 10, 11, 14, 15)

106
Fig. 5 Year-glyph (upturned), Stela 3 (after Caso 1928: Fig. 31)

Pericxl Acatl Tecpatl Calli Tochtli

I unattested

II

III
!£J
IV

Fig. 6 The Year bearers

107
 The Monte Alban year-bearers were identified by Caso as
Turquoise (or Jade), Deer, Serpent (or Serpent Mask), and Bat
(1947: 29). Of these the first and last are unknown on other
Mesoamerican calendars. The second and third, on the other hand,
are found on almost all day-name lists since they are situated a
mere two positions apart. Even reinterpreting Deer as Rabbit will
not solve the problem, since these are still only three positions
apart, instead of the required five.

In 1976 at the Annual Meeting of the Society for American

Archaeology I presented a means of bringing the Monte Alban
evidence into an acceptable Mesoamerican pattern. As I have
discussed elsewhere (1980: 26-36; 1981; 1983), a comparison of
Zapotec inscriptions from different time-depths will show that the
symbols employed to name the days in their function as year-bearers
were actually subject to change from period to period (Fig. 6).
Not four, but eight, different signs are recorded attached to
Caso's year-glyph, and among these are the three labelled
Turquoise, Deer and Serpent. I am at some odds to explain the
absence of a year Bat, but a careful study of all published and
unpublished inscriptions at Monte Alban has forced me to the
conclusion that no sign depicting this or anything similar is
attested from any period at the site.

From Period 1 Stelae 12, 13 and 17 we may extract three of the

four Monte Alban · year-bearers, which take the form of a dragon's
head, a crosspiece device and a jaguar's head (the latter regarded
by Marcus, 1976: 46, as a month name). The first is Caso's ''head
of Cocijo" and year Serpent or Serpent mask combined. If we
compare the sign in Figure 5 with the year-name on Stela 12, we
find that in both cases the dragon-like head should be considered
not a part of the year glyph, but rather a quite separate year-
name, and that the headdress alone functions as a sign for year.
Neither a year Serpent nor a year Dragon will, however, match up
in compatible positions with a year Jaguar.

A comparison of the Period I year-bearers with those of Period

III, as found on the stone cover of Tomb 104, provides the
solution: of the years named in sequence, the second and third are
depicted by a flint knife and house-- matching the Classic Maya
(cf. Thompson 1960: 124), Mixtec and Aztec year-bearers Tecpatl -and
Calli-- and, therefore, the first and fourth, in the form of
Caso's Serpent Mask and Turquoise, must be equivalent to Acatl and
Tecpatl. The "turquoise" glyph, which I prefer to call "swirling
bands" on formal grounds, may be compared with the Maya Lamat
(=Tochtli), which is similarly divided into four chambers. By
further comparison with later inscriptions, the crosspiece and
jaguar of Period I may be lined up with the Flint and House of
Period I I I ( Wh it take r 198 0 : 3 3- 3 4). In these ear 1 y cases we are
dealing with symbolical associations of the day-names in their
special role as year-bearers.

The Trecena Cycles

A calendrical sign which appears only once at Monte Alban, on

108
Stela 17 of Period I (see Fig. 4), supplies a vital clue in
ascertaining the exact position of the year-bearer in the Zapotec
calendar. In the upper right-hand column below the date 2 0zomatli
or Monkey (with fingers standing for the digits) is a long cigar-
like sign followed by the number 18. Caso has suggested (1947: 10-
11) that this may be the name of a month in which 2 0zomatli is the
eighteenth day, a hypothesis which can be tested now that the year
dates are identifiable.

If the year-bearer 12 Tecpatl fell on the first day, all

months of that year would have begun with a day Tecpatl, and 2
0zomatli would have been the 14th day of Month 7. If the year-
bearer fell on the 360th day, the date would have been the 13th of
Month 12. A year which had 12 Tecpatl as its 365th day would,
however, yield for 2 0zomatli a position 18 in Month 12.

The matter would seem to be resolved by this close match, but

there are reasons for finding this solution less than satisfactory.
Firstly, there is one other time division which meets the
necessary requirements: the 13-day period. If the year-bearer
fell on the 360th day, instead of the 365th, 2 0zomatli would then
have occurred in position 12 of the 18th 13-·day period counting from
4 Quiahuitl, the first day of the year (Fig. 7).

Locating the year-bearer on the 360th day of the year would

accord best with the evidence available to us for the Mixtec and
Aztec calendars, which stand in a direct line of descent from the
calendar of Monte Alban. It is also reasonable to expect that a
Mesoamerican culture would avoid placing the year-bearer within the
five ill-fated days at the end of the year, and that the first day,
as under the Classic Maya, and last day of the solar month cycle,
as under the Aztecs, would have been considered rather more
appropriate.

Further light may be shed on this problem by another period

glyph, Case's Glyph W (catalogued W(hit taker) 542-545); see Fig.
8), which in contrast to the first glyph (W864) is of frequent
occurrence in periods I and II. This sign has been variously
analysed as a day-name (Caso 1928: 44, 95), a month-name (Caso
1947: 29; Marcus 1976: 46-47), due to the discovery of texts where
it is followed by numbers above 13, and as a count of elapsed time
(Prem 1971: 120-21). An examination of the context of this s ·ign
has led me, however, to the conclusion that it stands for the
ritual semilunation known as the trecena (Whittaker 1980: 36-37;
1981; 1983), and takes in fact the form of a conventionalised
half-moon. The number following the sign places a named day in an
elapsing trecena of that number in the divinatory calendar. Unlike
the Maya who recorded month dates as elapsed time, the Zapotecs
recorded dates in elapsing time, a distinction of some importance
to the decipherer (cf. C6rdova 1886: 15Sf f. for the 16th-century
Zapotec modes of counting time). By means of the trecena device it
is possible to anchor a number of day-signs to their corresponding
positions on the day-name list. Thus, Stela 15 places a day 1
Xochitl (Zapotec Loo "Face") in Trecena 4, while Stela 15 records
2 0zomatli in Trecena 14.

10<)
 I Cipactli 10e 2 9 3 10 4 11 5 12- ,6 13 7 i-100 2 9 3 10
',
,4, 11 5 12 6 l3 7 :10e 2 9 J 10

®e f0e
I

II Ehecatl 2 9 3 10 4 11 5 12 6 13 7 l2 9 l 10 '4'
, .. 11 5 12 6 13 7 2 9 3 10 4 11
..,
III Calli 3 10 4 11 s 12 6 13 7 15119 2 9 3 10 ,4, ll s 12 6 13 7 ®a 2 9 3 10 4 11 5 12

IV CUetzpalin 4 11 5 12 6 13 7 ®a 2 9 3 10 '4'
, '\
11 5 12 6 13 7 fj) a 2 9 l 10 '4'
, .. 11 5 12 6 13

V Coatl 5 12 6 13 7 {i)a 2 9 l 0 '4'

, .. 11 ' 5 12 6 13 7 {i)a 2 9 3 10 ;.: 11 5 12 6 13 7 {i)
VI Miquiztli 6 13 7 fve 2 9 3 10 4 l 5 12 6 13 7 {iJa 2 9 3 10 :4: 11 5 12 6 13 7 {lJa 2

VII Mazatl 7 l1Ja 2 9 l 10 4 11 5 2 6 13 7 10e 2 9 3 10 '4'

, .. 11 s 12 6 13 I 7 l\.;.(8 2 9 l

VI II Toc:htli 8 2 9 3 10 4 11 5 12 6 l 7 ®a . 2 9 J 10 .. 4'
, '
11 5 12 6 13 7 @a 2 9 3 10 4

IX Atl 9 3 10 4 11 5 12 6 13 7 fi)a 2 9 3
.,4,,
10 '4'
, '
ll 5 12 6 13 7 lDa 2 9 l
,, 0 4 11 s
X It:zcutntli 10 4 11 5 12 6 13 7 ®a 2 9 l ilO 11 5 12 6 13 7 @a 2 9 l 10
...
l 5 12 6

~
~
0
XI O.Z:anatli

XI I Halina.111
11

12
s
6
u
13
6

7
13
{f)a
7 ®a
2 9
2

J
9

10
l

(
10

11
" 4,
, '
5
.I

12
11 5

6
12
13
6

7
13

{!Ja
7 ®a
2 9
093 10
l
' ,
,4.
10

11
' 4,
,'
5
11

12
, '
5

6
2

l
6

7
13
{I) 8
7

XIII Acatl 13 7 Ga 2 9 . 3 10 4 11 5 12 6
I
•l ) 7 Ga 2 9 3 10 '4'
, .. 11 5 12 6 13 7 Ga 2 9

XIV Ocelot.l {!) 8 2 9 l 10 4 11 5 12 6 13 7 {iJa 2 9 l 10

.,4,, 11 5 12 6 13 7 {J) 8 2 9 l 10
..4,
'>N Cuauhtli 2 9 3 10 4 11 5 12 6
1317 ®a 2 9
.. ,
3 10 , . 11 5 12 6 13 7 ®a 2 9 10 4 11

XVI Co'Lcacuauhtli 3 10 4 11 5 12 6 13 7 Ga 2 9 J 10 , 4,. 11 5 12 6 13 7 [ii) 8 2 9

.. ,
l 10 4 11 5 12
XVII Olin 4 11 5 12 6 13 7 ®a 2 9 J 10
' ,
,4.. 11 s 12 7 ®a 10 Is 12 6 13

's'0 • fj)a
6 13 2 9 l , 11
®a . 4, (i§)
XVIII Teq:,atl 5 12 6 13 7 2 9 3 0
...
'4' 11 5 12 6 13 7 (i§) 8 2 9 3
. 4,
10 , . 11 13 7

XIX Ouiahuitl 6 13 7 {j) 8 2 9 3 10 ,4: 11 5 12 6 13 7 fva 2 9 3 10

, '
11 5 12 6 13 7 2

XX Xochitl 7 10e 2 9 3 10 4 11 5 12 6 ll 7 10e 2 9 J 10 ""

,4, 11 5 12 6 13 7 1~8 2 · 9 3

Fig. 7 Solar trecenas of the year 12 Tecpatl. Divinatory trecenas

are numbered in small circles. The beginnings of solar trecenas are
marked with an X. Dates recorded on Stela 17 are within large
circles.
W542 W864

c:::J

Fig. 8 The trecena glyphs

 A comparison of the two period signs of Monte Alban I reveals a
certain formal resemblance between them, which strongly suggests
the derivation of the first from the trecena glyph. Since the
latter depicts a half-moon, a related meaning must be sought for
the derivative. "Month" would again seem a logical choice, but
this presents difficulties. A reading 2 Ozomatli, Month 18 will
not fit any placement of the year-bearer, the only acceptable
readings in this case being Month 7 or Month 12.

The possibility that the sign in question is the name of one

particular month, and that 18 is the position of the day within
this (12th) month, may be ruled out by virtue of the fact that
such a sign would not simply be a derivative of the sign for half-
moon, but would have to have graphic characteristics distinguishing
it unmistakenly from seventeen other signs in this category. An
examination of the glyphs for the Classic Maya month names will
illustrate this point.

The Novena Cycle

Indirect evidence for the year-bearer on the 360th day, and

therefore support for the identification of the mystery glyph as
that for "solar trecena", comes from Tablet 14 of Period II (c. 200
B.C. - 250 A.D.). This inscription (Fig. 9), the longest known to
us from Monte Alban, records dates anchored in four trecenas. The
central column-- and main text-- beginning with the name of the
year, 6 Tochtli, refers to the subjugation of a town on the day 11
Calli (Whittaker 1980: 37-39, 134-36; 1982). The first column
(from the right) is a preface which reads: "Trecena 5 (named) Reed
descends to 7 Quiahuitl". The primary date 11 Calli occurs four
days later, still within the 5th trecena, which began with 1 Acatl
(Reed). In the upper left of the tablet is a colophon of reduced-
size hieroglyphs. This two-column text contains four dates: on
the right a probable day 8 Olin and below it a damaged 5 Mazatl or 6
Tochtli (the year-bearer) in Trecena 15; and on the left, in a
most unusual mirror-image sequence: Trecena 4, 2 Cipactli; 10
Calli, Trecena 2.

The dates 11 Calli and 6 Tochtli require no further comment,

but the other dates present a curious pattern. In the divinatory
calendar 10 Calli, 2 Cipactli, and 7 Quiahuitl are separated from
each other by 9-day intervals, and 8 Olin falls on day 9 of the
last month of the year. The preface and colophon, therefore, seem
to be concerned with placing the subjugation of the town not simply
in the ritual context of the 13-day trecenas, but rather in that of
the very important Mesoamerican series known as the Lords of the
Night.

Of relevance here is the coupling of dates in the far left

column with deity heads identifiable from comparative Mesoamerican
iconography. Above the date Trecena 4, 2 Cipactli appears the
head of a god known conveniently by the Aztec label Xipe, who is
lord over the 4th trecena, which begins with 1 Xochitl, the
Zapotec sign for which is the face of Xipe.

112
 ~
C t

tgJ
~
I ~0
0
• •
Fig. 9 Tablet 14 (after Caso 1947: Fig. 44, with
corrected Trecena 15)

1 13
 'lJ,
~-- s s
I Clpact.11 10-s
,' 2 9 J ,4, 11 12 6 13 7 ,
10e@ 9 .. J 10 4 11 12 6 7 10e 2 9 J 10
. 7,,
II Ehecatl 2
.. J ,
9 J 10 4 S 12 6 13 ®e 2 9 J 10 4 11 's'
, ' 12 6 13 7
, /0e
2 9 :( 10 11
III C4111
,'
10 4 ll
I
-s 6 13 7 fiif e 2 9 J
0·0 s 12
. 13,
6 13 7
15
8
,
2 9 10
..11,
4 11 5 12

, ''
IV CUetzpalin 4 11 s 12 '6' 13 7
' 2, s
'
f0e , ... 9 J 10 4
..
11 12 6 7 ®e 2 9 J 10 4
, '
5 12 6 13
V Coatl 5 12 6 13 7 fiJe 2 9 J 10 4 11 ,s,
,
12 6 13 7
1

lJ)e 2 9
' J, 0 11 5 12 6 13 7
,0
,~
, , '
VI Mic,.tlztll 6 13 7 '1G e 2 9
\,
10
, ' 11 s 12 6 lJ 7 {l>v
, ' 2 9 J 10 5 12 '6"'
, . 13 7 fiJe 2
{i)e I'
,
VII Mazatl 7
.. 9,
2 9 J 10 11 s 12 6 ' 13 7 l3Je 2 9 3 10 '11' s 2 6 13 7 l:va 2 'g'
,' 3
VI II Tochtli

IX Atl
e

9
2

J
... ,
, ...
10 4
10

11
4

s
11

12
's'
,
6
.. 12

13
6

7
lJ

fi>v
7

, ...
,~

2
e

9
2

J
9
10
'1'
,"
4
10

11
4

s
11
12
s
,"'6'.
"$ lii>a,. ,
12
13
6 J 7 'ft})
2
8

9
2

J
9

10
3
'4'
,'
10

11
4

s
X ltzculntll 10 4
,' 11 s 12 6 'p' 7 RVe 2 9 J 10 )( s 12 6 13 7 RVs 2 ,9 . 10 11 5 12 6
XI Ozanatli 11
..12,
s 12 6 13 7 t0 8 2 9 >: 10 4 11 s 12 6 13 7 ,@a
, ... 2 9
,
10 4 11 s ', 12., 6 13 7
~
~
.;-
XI I Hal inal 11
, ... 6 13 7 ©a'
, " 2 9 J 10 4 11 5 :12 '6'
, .. 13 7 0e 2 9 J 10 4
"
11 5 12 6
. 7, Ga 13 7 0e
XIII Acatl 13 7 Ge 2 9 J 10 4 '1{ s 12 6 13 7 Ge 2 ', (... 3 10 4 11 5 12 6 lJ 2 9
' 3, , " . 10
,. . ~
XIV Ocelot! 0e 2 9 10 4 11 s '10 8 ~ 9 J 10 6 13 7 1@a' 2 9
, 12 6 13 7
,. , ,
~ Cua\l\tll 2 9 3 ',10. . . 4' 11 5 12 '6' 13 7 ®a 2 9 J 10 ,V.. 11 5 (08 2 9 10 11
,' , ' . s , . 12
XVI Cozcacuauhtl1 J 10
. s, 4 11 s 12 6 13 7 10 8 2 '9'
,'
10 4 11 s 12 6 2 9
,
3 10 11 , ..
,fiJe ..12,
XVII Olin

XVIII Teq>atl
4

5
11

12
,
I
6
'
12

13
6

7
13
@
7
8
. 9,
,'
2 9
2
3 10
9
,,,
3

'
10

11
4

5
11

12
s
6
I '
13
6

7
13
@e
7

9 3
9

10
J

4
',10..
11
4

s
11

12
s
6
12
' 13,
6

7
13
{0
XIX Quiahuitl

XX Xochitl
6

7
' 13
'10a
7

2
r 7 8

9
2

3
, ...
10 4
3 10

11
4

5
11
... ,
,12.
5
6
12

13
6

7
"c§i) G,
r
4
e 2 9
2

J
9

"10'
J 10

11
4

s
11

12
' 5,
, "
6
12

13
6

. 7
13

0"'
r,a..
7 'lJ) a
9
, '

Fig. 10 Novenas of the year 6 Tochtli. Divinatory trecenas are

numbered in small circles. The beginnings of novenas are marked
with an X, ·and year dates 5 Calli and 6 Tochtli with a+. Dates
recorded on Tablets 10 and 14 are within large circles.
Fig. 11 Tablet 16 (Caso 1947: Fig. 45)

115
 Similarly, below the date 10 Calli, Trecena 2 we find the
head of a god who has affinities with the Aztec Tezcatlipoca
Tepeyollotl, a jaguar deity dominating the 2nd trecena, which
begins with 1 Ocelotl (Jaguar). Over Xipe's head is a vessel of
water, his symbol as god of the dew known to the Aztecs as the
Night Drinker. Under the Tezcatlipoca aspect is a flaming torch,
symbol of this god known as the Night, the Wind.

If the solar year were divided in early Zapotec times into

novenas, 9-night periods matching the 13-day periods, I felt it
might be possible to map out the placement of these novenas within
any given year. A count of novenas tied to the solar year would
result in an "empty" (uncounted) day at the end of every second year
(or every 729 days, = 9x9x9). This pattern was applied to the year
6 Tochtli of Tablet 14, and the result can be seen in Figure 10.
It is significant that the three dates preceding 11 Calli fall on
initial nights of three consecutive novenas, and that this is
possible only if the year-bearer is situated on the 360th day. The
preponderance of night symbolism in this inscription may indicate
that the event commemorated was a night action.

A final piece of evidence in support of a Zapotec year-bearer

preceding the five empty days at the end of the solar year comes
from Tablet 16 (Fig. 11). The left-hand column has a curious
characteristic shared by Tablets 12 and 15: an upside-down trecena
glyph superfixed with an element resembling feet. Above the
compound is the head of a feline, probably a puma, and below it
that of a monkey, cartouched and followed by the number 18.

Because the number is too high, the last two glyphs cannot be
read together as a day-name and coefficient. Caso (1947: 29) has
suggested that the down-turned trecena compound stands for day
zero-- the first day of an elapsed-time count-- of his Month W,
but this clashes with his interpretation of the number as the
position of the day in the month, and of the monkey head as the
name of a month (1947: Fig. 66).

If glyph W542 stands for divinatory trecena, as I have argued,

two other items must be present: a named day and the number of the
trecena to which it belongs. Since 18 cannot be the coefficient of
a day-name, it must be the number of the trecena, in which case
the intervening monkey head is unlikely to be a day-name. On
Tablet 14, as we have seen, one trecena glyph is followed first by
its rumber (5) and then by its name ( [1] Acatl).

On Tablet 16 this practice seems to be reversed, but Trecena

18 should commence with and be named Ehecatl (general Mesoamerican
"Wind", but Zapotec Quij "Fire"), not Ozomatli, or Monkey, as
found here. The analogy of the year-bearers, where alternate
symbols may be substituted for the standard signs, may have some
bearing on the problem. Since the numerical context is sufficient
to prevent confusion, the expected flaming torch for Fire/Wind had
been replaced by an animal associated with this day-name position--
in Aztec times the monkey is a primary aspect of the wind god, who
is also god of the day Ehecatl.

116
 The position of the day within Trecena 18 remains to be
determined. A day zero as initial day must be rejected on the
weight of the evidence from Stela 13, where the first day of
Trecena 4 is written with a finger for the number 1 and the trecena
glyph is upright. The alternative is to take the down-turned
trecena glyph as signifying the end of a trecena or, together with
the super£ ix, "at the foot of the (descending) trecena". The
numeral 13, which can occur only with the final day-name of a
trecena, is attested only once in this context on existing
monuments of Preclassical Monte Alban, and thus there is no barrier
to such an interpretation on the grounds of graphic convention.
Interestingly, the only other numeral virtually absent from the
corpus is 9, a reflection perhaps of its portentous nature.

If, then, the day recorded is the final day of Trecena 18,
the day-name alone would suffice to complete the required format.
This should be, as a glance at Figure 2 will demonstrate, a day 13
Ocelotl/Jaguar in the general Mesoamerican calendar, and above the
trecena information on Tablet 16 one indeed finds the head of a
feline. This day 13 Ocelot! is, moreover, the first day of the
year 8 Acatl named on the tablet, but again only if the year-bearer
falls on the 360th day.

The Day-Names

We are now faced with felines occupying two positions in the

Zapotec day-name series: Calli (III) and Ocelotl (XIV). While a
probable puma head appears alone for the day Ocelot!, the jaguar
head alternates with a house glyph as both year- and day-sign for
the position Calli in the Monte Alban Preclassic, a fact evident
from Tablet 10 (Fig. 12), where the context will permit only a
reading 12 Calli, Trecena 8. In the year 6 Tochtli, as recorded
on Tablets 10 and 14, the day 12 Calli is the initial day of a
no, ena (see Fig. 10) in the first month, and is also within the
first full divinatory trecena of the year.

In one instance (Fig. 13) trecena information seems to conflict

with the days with which it is associated. 5 Xochitl and 8 Calli
are recorded on Tablet 13 in connection with an apparent Trecena 15
(Caso 1947: Fig. 43). A check of the positions of these two days
reveals, however, that the trecena in question should be 16 ih an
elapsing-time system. I have personally examined this inscription
and have found that, although this section has undergone
considerable erosion, in the space above the three bars of the
numeral 15 traces of the missing dot may still be discerned.

Largely as a result of the frequent application of the day-and-

trecena formula, it has been possible to identify more than half of
the signs representing day-names in the Preclassic, and a few
others may be tentatively classified owing to their close
resemblance to day-name glyphs of later cultures. The approximate
evolution of day-names over the centuries in the Valley of Oaxaca is
summarised in Fig. 14. It should be kept in mind that the pre-
Conquest 'names" provided here are merely convenient labels derived
from the form of the hieroglyphs, and are not necessarily identical

117
 0 0
c::::,
c::=:,

Fig. 12 Tablet 10 (Caso 1947: Fig. 42)

Fig. 13 Tablet 13 (Caso 1947: Fig. 43)

118
 lieneraJ.
Mesoarnerican Preclassic Classic Postclassic onwards
Name Meanings
I Dragon Dragon? Dragon Chi l ta Dragon
II Wind Torch Three-part Scroll Quij Fire
III House House/Jaguar House Quela Night
IV Lizard Queche Frog?, Lizard?
V Serpent Serpent Zij Misfortune
VI Death Death's-head Death's-head Lana Gloomy
VII Deer Deer China Deer
VIII Rabbit Bands Bands Lapa Fragmented
(Rabbit as year)
IX Water Water Water Niza Water
X Dog Fire-~erpent Tail? Tella Conflict?
f,-1,
f,-1, XI Monkey Monkey Monkey Loo Monkey
\.{)
XII Twisted (Grass) Pija Twisted
XIII Reed Reed Cocijo Dragon Quij Reed
(Dragon as year)
XIV Jaguar Puma
xv
Jaguar/Puma ·
,Queche
...,
Feline
Eagle Naa Milpa ?
XVI Vulture Owl Owl/Vulture Loo Crow
XVII Movement Rushing/ Falling Water Xoo Fury
Falling Water
XVIII Flint (Knife) Crosspiece Flint Copa Cold
XIX Rain Rain Rain Cappe Lightning
xx Flower (Xipe) Face (Xipe) Face Lao Face

Fig.14 Evolution of Zapotec day-''names" (cf. Whittaker 1980: Fig.7)

to the meanings of the actual names used in the periods concerned.
There is, of course, no guarantee that day-names were always in
the order they take in the Mixtec, Aztec and Yucatec calendars.
The original sequence at Monte Alban can be ascertained only by
means of a careful search for anchored dates and sequences of days
in the inscriptions available to us.

Since the 16th-century Valley Zapotec day-names are neither

static in form nor in every case transparent in meaning and
unambiguous in spelling, it will continue to be necessary to employ
standard Nahuatl terminology, the referents of which are well
understood, for Zapotec dates on monuments. Sole use of Spanish
or English translations of the Nahuatl terms, while frequent in
Mixtec studies, is inadvisable here becauseof the lack of
concordance between Zapotec glyph and Nahuatl meaning, which would
result, for example, in labelling day-name XX (glyphically a Xipe-
like face) Flower.

Developments in the Classic

The inscriptions of Period I constitute related texts in style,

format and content, and this holds no less true for the Mound J
tablets of earlier Period II. At the beginning of the Classic
period, Monte Alban IIIA (c. 250 A.D. - 450 A.D.), a new grouping
appears in the form of two related sets of South Platform stelae
(Caso 1928; Whittaker 1977; Marcus 1980). Following Preclassic
convention, each inscription has a reading order of . columns
downwards and from right to left when the glyphs face left
(standard), and left to right when the glyphs face right. When
the columns oppose each other, the right-hand column takes
precedence.

A study of the dates on the stelae reveals a remarkable

co··relation between their sequence in the 52-year cycle and their
placement around the South Platform (Fig. 15). The first set,
located at both ends of the structure's important north face,
follows the sequence (from right to left): Stelae 6 (years 10
Tochtli and 1 Tochtli), 5 (5? Tochtli), 4 (no year), 3 (10
Acatl) and 2 (13 Tochtli). It is no coincidence that the only
years found are Tochtli and Acatl years. The vast majority of the
post-Period I dated inscriptions at Monte Alban, and for that
matter elsewhere in the Zapotec area, record a year Tochtli, and
apart from the occasional Acatl, other year positions are virtually
unknown. This suggests the possibility that the Zapotecs ordered
events to fit the close or beginning of a recurring four-year
subcycle, in a kind of ritual systematization of history.
Alternatively, it may simply be that they preferred to set down
events at four-year intervals.

Be that as it may, the inscriptions of the early Classic have

not yet yielded many of their secrets. The familiar trecena glyph
of the Preclassic disappears from view, although the signs of the
days beginning trecenas are now and then used in its place, an
extension of the practice found on Period II Tablets 14 and 16.

120
Fig. 15 The chronological and locational order
of the South Platform stelae

121
 Stela 2 (front and right side)

E
~3
000

~
Stela 6 (left side and front)

Fig. 16 Corner stelae of the North Face Group (after Caso 1928)

122
 Ste la 7
Ste la 8 .

Ste la 1
Plain Stela
year-name
13 Olin 12 Ehecatl? 5 Tochtli 7 Itzcuintli? 3 Quiahuitl 9 Ozornatli 1 Cozcacuauhtli 13? Tochtli
day 117 day 1_42 day 148 day 150 day 159 day 191 day 196 day 208

+260 9 days 9 days 4 X 9 days

Fig. 17 The date series of the Four Gates Group

 A day 13 Cozcacuauhtli (with an owl-lLke glyph) crops up twice
on monuments of the North Face Group, both times in connection with
apparent period signs (Fig. 16). At the extreme northwest corner,
on Stela 6, it occurs below a cartouched element which rests in a
basin over the lower half of a human body in the cross-legged
seating position. In the northeast corner, on Stela 2, the date
precedes a cartouched glyph surmounted by a hand and subfixed by the
same seating element as on Stela 6. Since the cartouched elements
do not appear as day-signs elsewhere at Monte Alban, there is a
strong possibility that we are dealing with a larger unit - a month
or one of the 52- or 65-day periods within the divinatory cycle. A
month interpretation, however, will not work for the years 1
Tochtli and 13 Tochtli, nor will a 65-day period fit.

Only in the 52-day subdivision of the calendar does 13

Cozcacuauhtli carry significance, as the final day of the third
period of the cycle. The context of Stela 6 suggests that the
period-sign recorded is that of the third period which is just
coming to an end. On Stela 2 a different period-sign follows 13
Cozcacuauhtli, and the hand element is setting it down. The
fourth period, which begins here, is ruling when the event of the
main text takes place on 3 Quiahuitl, a mere three days after 13
Cozcacuauhtli. On the Mound II Tablet (Caso 1965b: Fig. 17),
which bears the same year date as Stela 2, not to mention related
iconography, the seated element supports a variant of the Olin sign
below a cartouched element which elsewhere seems to mark the
commencement of a trecena. The first day of the Olin trecena is,
interestingly enough, the first day of the fourth 52-day period.

The second set of Period IIIA stelae (1, 7, 8, and the Plain
Stela), which I call the Four Gates Group after one of the
traditional names for the South Platform (Whittaker 1980: 166-69),
is cast in a style resulting from intensifying foreign contact with
the Teotihuacan-dominated North and Southeast. A series of days in
th! year (13?) Tochtli are named in the order given in Fig. 17.
That this was a time of great change is seen in the lack of a single
convention for the representation of numerals. On Stela 8 five-
bars were placed under the dots as in the Preclassic, on the Plain
Stela bars occur both above and below dots, and on Stelae 1 and 7
they appear only above the dots. Numerals may be found without
internal detail as on the fronts of Stelae 7 and 8, or embellished
as on the sides of these same monuments, where, moreover~ a
filler element seems to be found in connection with the number 7
(Whittaker 1977; cf. Marcus 1980: 60).

The elements comprising numeral coefficients are as a rule

arranged in horizontal layers, these being in all but one instance
below the day-names associated. On Stela 1, nonetheless, several
coefficients are found standing vertically beside or below the day-
or year-name. No difference in fun ct ion seems to be invo 1 ved in
such cases.

A semblance of order is restored in the 2nd Intermediate Period

(or Transici6n IIIA-IIIB, 5th century A.D.), as exemplified by the
reverse of the cover of Tomb 104 (Fig. 18). Again numerals are
found both beside and below day-names, but there is a significant

124
Fig. 18 The reverse of the Tomb 104 entrance slab (after Caso 1938:
Fig. 94, with correction of year 9 Acatl on upper edge.

~ia-

Embellished numeral 1,
cappe Tomb 104 slab (reverse)

Calendrical name
Quiaaappe/1 Quiahuitl
as Cocijo pendant

Fig. 19 The prefix Quia- as a Classic Period glyph

125
change in function heralded by the vertically-aligned numerals. In
each case where the latter occurs a coefficient is already present
in standard horizontal arrangement below the neighbouring day-name,
so vertical numerals cannot be day-name coefficients.

In one instance (at lower left) the vertical numeral 7 is

followed by a sign which I have identified as the Zapotec equivalent
of the Maya kin "sun, day" (Whittaker 1976; 1980: 28, 40-41),
which it closely resembles. To their right is the day 7 Mazatl.
Since this is seven days after the commencement of the divinatory
cycle on 1 Cipactli, the statement "7 days" refers to the position
of the primary date in elapsing time.

Below this information is a row of three cartouched glyphs

which are followed by a vertical 15 before a day-name with
coefficient 7. On its own underneath the first sign is one further
cartouched glyph. These four glyphs are the first evidence we have
for the Zapotec cocijo or 65-day period, described in C6rdova's
Arte del idioma zapoteco (1576, 1886). Similar to the signs for
52-day divisions in Monte Alban IIIA-- and possibly adapted from
them-- these glyphs stand each for a cocijo of five trecenas,
counting forward from a base of 1 Cipactli. The third sign,
therefore, should represent Trecenas 11 to 15 inclusive, and the
vertical 15 next to it would thus seem to indicate Trecena 15.

The day-sign following takes the form of a cartouched Greek

cross above an odd basin-like element, and in the Arrowhead Series
at Mound J the emblem of District H, the Mixteca Alta (Mixtec
Nudzahui ''Land of Rain"), is a combination of quasi-Greek cross and
water/rain elements (Whittaker 1980: 114-15, 133-34). In Trecena
15 the day in question is 7 Atl/Water, the first of the five empty
days at the end of the year 6 Tochtli (recorded at the right).

The wide scope of innovation in the artistry of the Early

Cl tssic gives way in Period IIIB (c. 450-700 A.D.) to a sharp
decline in the sculptural quality and textual coherence of monuments
at Monte Alban. Cluttered with crude, ill-proportioned
iconography and awkward hieroglyphs, the latter become increasingly
difficult to interpret. Stela 9 at the entrance to the North
Platform is no exception (see Batres 1902; Caso 1928: Figs. 49-
52), although the calendrical information it displays is
nonetheless of great importance.

On the north side of the stela two individuals stand above a

row of glyphs carrying the message: "3 Cozcacuauhtli, day of
capture, Cocijo 1 (?)". At the foot of the monument the
inscription concludes: "11 Quiahuitl, day of sacrifice, year 6
Tochtli", below which the glyphs for Cocijos 2, 3 and 4 are given.
3 Cozcacuauhuitli falls in the second trecena of Cocijo 1, 11
Quiahuitl in the 2nd of Cocijo 4. The period glyphs differ
somewhat from those found in Monte Alban IIIA and the 2nd
Intermediate Period, but a relationship between them seems evident.
The glyphs for the 52-day periods apparently were adopted for the
later cocijos, the 2nd and 3rd doubling up as alternates for the
3rd cocijo.

126
The Postclassic and after

With the end of the Classic comes also an end to the last
traces of the tradition of writing at Monte Alban. The growth of
iconographic and notational systems at _the expense of writing is
virtually unparalleled in the history of civilisations, but this
transformation, all but complete by the mid-1st millennium A.D.,
gave rise to a versatile new system mixing elements of writing
(phonetic renditions of names, creation of verbal glyphs, etc.) in
a dominant iconographic structure. Durable stone declined in use
as a primary medium of communication, the emphasis passing to a
highly perishable medium: paper. Largely for this reason our
knowledge of Postclassic calendrics has to rely, not on the sparse
and sculpturally poor stone tablets and stelae available from sites
around Monte Alban, nor on codices of paper and hide, none of
which seems to have survived the Conquest, but on the writings of a
Spanish priest of the late 16th century, Fray Juan de Cordova.

In his Arte of 1578 Cordova set down some tantalizingly meager

data on the nature of the Valley Zapotec divinatory calendar,
including a nonetheless complete transcribed day-count, beginning
with 1 Cipactli and divided into trecenas which are in turn grouped
into 65-day periods. Each trecena (cocij) and 65-day period
(cocijo) bears the name of its first day, continuing a practice
going back at least to the Late Preclassic.

A further tradition reflected in hieroglyphic practice is that

of placing numerical coefficients after day-names. The Zapotec
culture is the only one in Mesoamerica known to have this both
hieroglyphically and linguistically as the standard convention.
Although the numerals are self-explanatory, the day-names are not.
Since C6rdova did not think it necessary in his summary to delve
into the meaning of the day-names, we are left in a few cases
ha ·ing to hazard a guess where the standard Mesoamerican day-name
series or the hieroglyphic record does not provide clues. Tonal
and other phonetic features are inadequately and often
inconsistently registered in Cordova's work, which only adds to the
complexity of the problem. No fully satisfactory interpretations
have been given to date for the day-names in positions X
(Itzcuintli) and XV (Cuauhtli), and several others have been rather
questionably translated by Caso (1947; 1965), where he departed
from analyses by Seler (1901; 1904) and Cruz (1935).

As if this were not enough, these· difficulties are compounded

by the so-called particles prefixed to day-names. Completely
lacking parallels in the Aztec and Maya calendars, they evade
translation even more than the day-names, for which we at least
have equivalents in other parts of Mesoamerica, and documentation
in the hieroglyphic script. Interestingly, all Zapotec
calendrical names recorded after the Conquest in the Spanish script
have these prefixes, but to my knowledge none has the expected
numeral coefficient, which is always given on the stone monuments.

127
 This immediately brings to mind the possibility that the
prefixes, as I prefer to call them, are remnants of an old number
system, but counting against such a hypothesis is their lack of
resemblance to any Otomanguean set of numerals, and even if one
looks at languages unrelated to Zapotec the prospect seems no
brighter. There are, furthermore, not thirteen prefixes as Caso
(1965: 944-45) maintains, but nine, of which four repeat at
regular intervals.

Leaving aside the occasional irregularity in the day-count

sequences, which may result either from slips in transcription or
deliberate ritual exceptions, the following pattern emerges:

+3 +3

1 Quia-

2 Pe(la)- 5 Pe(la)-

3 Peo(la)- 9 Peo(la)-

4 Ca(la)-

6 Qua(la)-

7 Pil (la)- 10 Pil (la)-

8 Ne(la)-, La- 11 Ne(la)-, La-

12 Pino-

13 Pece-

Of these one is clearly present in the hieroglyphic record as

far back as the Protoclassic: the prefix Quia-, which occurs with
great frequency on urns of the rain god Cocijo ("Thunderbolt").
Effigies of this deity are normally adorned with a pendant bearing
his calendrical name 1 Quiahuitl (Zapotec Quiacappe, C6rdova's
Quegappe, "l Flash of Lightning"). The upper element is decorated
at its corners with the phonetic symbol quia/ quie, found among
other things in the conventionalized mountain (quia) used with
place-glyphs, but also used symbolically for rain (quia) in the
glyph for the day Quiahuitl (see Figure 19).

Wilfrido Cruz, an Isthmus Zapotec, attempted to explain the

prefixes as normal Zapotec words in a ritual context (e.g. peo as
"moon"), and while many of his interpretations (1935) are somewhat
less than compelling, his hypothesis does seem on the right track.
Particularly astute is his observation that pece, which is used
only for the final day of a trecena, is the Zapotec word for
"portent, omen". It is precisely at this point, at the end of
the or overturning of a passage of time, that omens are most
expected and feared. I should add that after the Conquest the
Devil became associated with the ominous god Bezelao "13 Xochitl",
which is the very last day of the divinatory cycle.

128
 The same pattern of prefixes turns up again in day-counts
recorded in the northern Zapotec district of Villa Alta in the 16th
and 17th centuries (Alcina Franch 1966), and on an unpublished
deerskin belonging to the Yale Map Library, which records a
genealogy of Zapotec lords up into the 16th century. Informed of
the latter's existence by George Kubler, I found it, erroneously
catalogued as a "Nauhtl. (sic) deerskin map c.1800?", to be a mine
of information on Zapotec calendrical names. Ironically, the
sequence of elements in a Zapotec lordly name, as reflected in this
document, is the exact opposite of that used by Nahuatl royalty:
(1) title, (2) personal name, (3) calendrical name, (4) numeral
coefficient.

Since none of the names is rendered in glyphs, the task of

analyzing their underlying meaning is all the more difficult. Some
calendrical names are nonetheless translatable owing to the
fortunate circumstance that neither the pref ix nor the day-name is
in the Spanish alphabet ambiguous: e.g. Huatela = 6 Itzcuintli,
Yaquela = 1 Calli, and Pezechila = 13 Cipactli. Others can be
reduced to two possibilities, e.g. Pilahuiya = 7 or 10 Malinalli,
while in extreme cases no means of selecting the correct
identification is likely ever to become available, e.g. Peloo = 2
or 5 (and perhaps 3 or 9) Ozomatli, Cozcacuauhtli, or Xochitl!

. Allowing for the inevitable ambiguities resulting from

imprecise phonetic renditions of such names, we are still left at
best with dual-referent prefixes, which are less likely to have
been differentiated by tone. This brings us again to the problem
of their function. Since numbers are dropped in names, the
prefixes can be assumed to have taken up the burden they carried,
that is, their influence on the fate of the day. A role as fate
indicators would explain the disproportionate frequencies of
particular prefixes in rulers' names, since we know from Aztec
so Lrces that decidedly negative birth dates could be overcome by
postponing the naming ceremony, and this would be all the more
desirable in the case of future rule rs. From a preliminary study
of these names I would suggest that 1, 7 and 10 carry a good fate,
3, 9 and 13 a bad one. The relationship between 9 and 13, the
numbers appearing with the lowest frequency on pre-Conquest
monuments, is evident also in the fact that 9 prefixes were applied
to the 13 coefficients at the time of the Conquest. My feeling is
that the prefixes are derived from associations with the Zapotec
Lords of the Night (see Whittaker 1983).

The symbolism of 9 and 13 crops up again in the present-day

calendar of the Southern Zapotecs, which was studied in recent
years by Carrasco (1951) and Weitlaner (1958, 1961). There the
divinatory calendar of 260 days is made up of 20 trecenas grouped
in five 52-day periods, a tradition which may descend from Monte
Alban IIIA, before the 65-day period was adopted.

In place of the expected 20 day-names this calendar runs a

parallel cycle of novenas, the nine day-names of which follow each
other in order except on the first day, on which the initial two
novena names double up. This adjustment at the beginning of the

129
divinatory cycle is somewhat reminiscent of the Preclassic Zapotec
novena cycle, which was adjusted to the solar year not by doubling
up, but by adding a day at the beginning of every second solar
year. In both cases it was the novena cycle which was adjusted in
length, not the divinatory or solar cycle.

Despite the many and varied developments in the history of

Zapotec culture, it is quite remarkable that the ancient
Mesoamerican institution of the calendar has endured so long and
remained so conservative. Not even the Southern Zapotec calendar
today, for all its losses, has altered its fundamental structure:
the divinatory cycle is still 260 days long, still divided into
trecenas and subperiods, and still preserves a novena cycle with a
doubling-up mechanism known elsewhere in Mesoamerica. Far from
being the obscure and seemingly aberrant system it has so long been
taken to be, the Zapotec calendar, in all its forms and all its
phases, can now be seen to have maintained its basic Mesoamerican
character at least 2500 years without major change.

130
References

Alcina Franch, Jose

"Calendarios zapotecos prehispanicos segun documentos de los

siglos XVI y XVII," in Estudios de cultura nahuatl, vol. 6.
( 1966)

Batres, Leopoldo

Exploraciones de Monte Alban. Mexico. ( 1902)

Carrasco, Pedro

"Una cuenta ritual entre los zapotecos del sur," in Homenaje

al doctor Alfonso Caso. Mexico. (1951)

Caso, Alfonso

Las estelas zapotecas. Mexico. ( 1928)

"Exploraciones en Oaxaca: quinta y sexta temporadas, 1936-

1937." Instituto Panamericano de Geografia e Historia,
Publicaci6n 34. (1938)

Calendario x_ escritura de las antiguas culturas de Monte Alban.

Mexico. ( 194 7)

"Zapotec Writing and Calendar," in Handbook of Middle American

Indians, vol. 3, part II. Uni ve rs i ty of Texas Press.
( 1965)

"Sculpture and Mural Painting of Oaxaca," in Hand book of

Middle American Indians, vol. 3, part II. University of
Texas Press. (1965b)

Cordova, Juan de

Arte del idioma zapoteco. Reedition of 1578 original.

Mexico. (1886)

Cruz, Wilf rido

El tonalamatl zapoteco. Mexico. (1935)

Grove, David C.

The Olmec Paintings of Oxtotitlan Cave, Guerrero, Mexico.

Dumbarton Oaks. (1970T

131
Marcus, Joyce

"The Origins of Mesoamerican Writing," in Annual Review of

Anthropology, S. ( 1976)

"Zapot ec Writing," in Sc ien t if ic American, 242, no. 2.

(1980)

Prem, Hanns J.

"Calendrics and Writing," in Observations on the Emergence ~

Civilization in Mesoamerica. Contributions of the University
~ Cal if ornia, Archaeological Research Facility, 11.
Berkeley. (1971)

Scott , John F.

The Danzantes of Monte Alban. 2 vols. Dumbarton Oaks.

(1978)

Seler, Eduard

The Tonalamatl of the Aubin Collection. Berlin and London.

(1901) - --

"The Mexican Chronology, with Special Reference to the Zapotec

Calendar." Smithsonian Institution, Bureau of American
Ethnology (Washington), Bulletin 28. (1904)

Gesammelte Abhandlungen zur amerikanischen Sprach- und

Altertumskunde, vol. S. Berlin. (1915)

Thompson, J. Eric S.

Maya Hieroglyphic Writing. 2nd edition. Norman. (1960)

Weitlaner, Robert J.

"Un calendario de los zapotecos del sur." Proceedings of the

32nd International Congress of Americanists. Copenhagen.
(19 61)

"La jerarquia de los dioses zapotecos del sur." Akten des 34.
- ------
Internationalen Amerikanistenkongress. Vienna. (1961)

Whittaker, Gordon

"On the Decipherment of Early Monte Alban Hieroglyphics."

Presented at Society for American Archaeology Annual Meeting,
St Louis. (1976)

"From Zapotec Hieroglyphics to the Mixtec Codices." Presented

at Society for American Archaeology Annual Meeting, New
Orleans. (1977)

132
The Hieroglyphics~ Monte Alban. Yale University Diss.
Ann Arbor: University Microfilms. (1980)

Los jeroglificos preclasicos de Monte Alban. Oaxaca:

I.N.A.H. (1981)

"The Tablets of Mound J at Monte Alban," in Coloquio

Internacional: Los indigenas de Mexico en ~ epoca
prehispanica y en la actualidad, ed. M. Jansen and Th.
Leyenaar. Leiden: Rijksmuseum voor Volkenkunde. (1982)

Calendar and Script in Protohistorical China and Mesoamerica.

Habilitationsschrift. University of Tubingen. (1983)

133
FRANZ TICHY

Observaciones del sol y calendario agrícola en Mesoamérica

En los altiplanos de México central y de Guatemala así como en

la costa del Pacífico reina un clima tropical que varía entre
precipitaciones estivales y una temporada seca de invierno. Se
cultiva el maíz como un cereal estival en combinación con frijoles,
calabazas y otras plantas bajo adecuadas condiciones. Estas se
alcanzan solamente en el supuesto de que las lluvias empiecen con
puntualidad y son esperadas a fines de abril y primeros días de
mayo, pero la incertidumbre amenaza siempre.

El año agrícola y con ello el calendario de trabajo de la

población est& adaptado en prbmedio a las condiciones atmosféricas.
Hay que insistir en "en promedio", primero a causa de la
incertidumbre del comienzo así como también del fin de la temporada
de las lluvias, y además porque los fases del tiempo no se pueden
esperar en todos los lugares de Mesoamérica en los mismos días o en
las mismas semanas. En regiones bajas la situación se presenta en
otra forma que en regiones altas, y en otra forma a sotavento que a
barlovento, en regiones del Pacífico en otra forma que en las del
Atlántico. Así se entiende, la razón por la cual la siembra del
maíz se hace en días diferentes adaptándose a las condiciones
locales • . A pesar de ello se halla una gran coincidencia en los
ciclos de trabajo agrícola. Esto es válido sobre todo para los
ritos agrícolas en combinación con las costumbres religiosas
típicas.

La descripción de la marcha del año agrícola en Tepotzlán/

Morelos, México, es un ejemplo de esta observación hecha por Osear
Lewis (1960). Esta es una región en donde se practica el desmonte
con quemadura. En sentido estricto no se trata aquí de un
calendario agrícola porque no se mencionan fechas, pero ellas
aparecen en las costumbres religiosas. Muy extendido están los
ritos agrícolas de la Fiesta de la Sagrada Cruz en los días 1 al 3
de Mayo. En Tepoztlán se celebra la bendición del maíz el día de
San Isidro el 15 de mayo. A mediados del agosto se hace una
procesión alrededor de los campos, y el 28 de septiembre se ponen
cruces en las milpas. En suma, el ciclo de labores puede
representarse en forma de calendario agrícola (véase Tabla 1 al
final del artículo).

Una estrecha conexión entre el viejo calendario, que aún

permanece latente, y el ciclo agrario lo encontró Wei tlaner entre
los chinantecos en Oaxaca (1969 S.538). El ciclo agrícola está
combinado con distintas fechas y además con distintos ritos
agrícolas, aún hoy en día en días de Santos o Santas del calendario
católico.

135
 Desde siempre se ha necesitado poner la siembra en el momento
oportuno para que las primeras lluvias puedan favorecer el
desarrollo del cultivo, es decir, no se espera a las lluvias por
sembrar después. Pero, ¿cómo fue posible en los tiempos
prehispánicos reconocer los distintos días en el curso del a~o,
p.ej. el 1 de mayo de hoy? ¿Cómo fue posible fijar el calendario
agrícola y el curso del ciclo de las labores? ¿Cuáles
observaciones del sol fueron en última instancia útiles para estos
asuntos.

Examinemos primero de todo un ciclo agrícola con fechas

conocidas. Se trata de un calendario de 260 días de duración,
levantado e interpretado por Rafael Girard en los pueblos chortí en
los Altos de Guatemala. El habla del "auténtico tzolkín de los
maya" o del "tzolkín original" .Y escribe en 1982:

"El tzolkín ele los chortí se computa desde el 8 de febrero al

24/25 de octubre. Es básicamente un almanaque agrícola ya que
su función principal consiste en regular las labores de cultivo, de
dos milpas consecutivas al año. Está dividido en dos secciones:
una de 80 días que corresponde al período estival y otra de 180 días
que es el ciclo de las lluvias... Los chortí cuentan la estación
estival del tzolkín por 'cuarentenas' o pares de uinales (véase
Tabla 2).

La primera actividad del campesino consiste en la tala de los

grandes árboles y debe finalizar en un plazo de veinte días.
"Durante el segundo sector de veinte días, el campesino se dedica a
la roza de 'montes bajos' o rastrojos." Después del desmonte
durante la primera cuarentena se realizan las quemas. Todas las
maderas deben estar en iguales condiciones de sequedad. "Las
quemas ••• principian a raíz de los ritos equinocciales en lugares
bajos y cálidos y se continúan hasta mediados de abril en tierras
altas... La ceremonia de mayor importancia se celebra cuando el
sol se realiza su primer paso por el cenit, que los chortí llaman
la 'Medianía del Mundo'; ceremonia que dura del 25 de abril al 3 de
mayo y determina las primeras lluvias."

El primer paso del sol por el cenit "divide el cuadrante

celeste en dos sectores de igual dimensión: el cielo claro y
luminoso de la estación estival y el cielo obscuro de 'invierno' •••
Las dos estaciones del trópico constituyen ciclos temporales
perfectamente simétricos de 180 días cada uno y suman el valor de un
tun maya de 360 días."

Durante el "reglamento de invierno", que dura 180 días, "el

cómputo del tiempo se realiza por novenas y trecenas... Parten dos
ciclos de 52 días divididos por el solsticio de verano. Es decir,
están enmarcados por los fenómenos más relevantes de la astronomía
maya, el primer y el segundo paso del sol por el cenit, divididos
exactamente en dos partes iguales por el solsticio. Entre el
primero y segundo paso del sol por el cenit se desarrolla todo el
ciclo de cultivo de la primera milpa. Entre el primero de estos
fenómenos y el solsticio de verano se cuentan cuatro trecenas.
Durante la primera se realiza la siembra, en la segunda trecena el

136
maíz está ya en 'alas de perico', es decir, con dos hojitas, al
comenzar la tercera trecena se procede a la primera limpia o
deshierbo y al final de la cuarta trecena, realizan los trabajos de
la segunda limpia. Durante el primer ciclo de 52 días, el maíz ha
llegado a su definitiva caracterización. Desde la quinta trecena,
el maíz está 'en elotes', acontecimiento que es celebrado con una
fiesta llamada aan kin."

"Los sacerdotes-astrónomos chortí observan el 25 de julio a la

Vía Láctea... Notan sus cambios de posici6n con relaci6n al sol.
Para el 25 de julio la trayectoria solar (la eclíptica F.T.) y la
Vía Láctea forman una gigantesca cruz en el cielo. Este fenómeno
astronómico señala el principio de la canícula, es decir, un
período de suspens1.on temporal de las lluvias y aumento de
temperatura, como resultado de las artes mágicas del sacerdote.
El Dios de la Vía Láctea se encarga de detener las lluvias durante
la canícula para permitir que los campesinos preparen sus tierras de
labranza para las siembras llamadas 'de segunda'."

"En la fecha del segundo paso del sol por el cenit del 12 al 13
de agosto, ocurre fuerte precipitación de lluvias de convecc1.on•••
el sacerdote ordena que se realicen las 'siembras de segunda'.
Durante el equinoccio del 22 al 23 de septiembre, los campesinos
hacen la única limpia de la segunda milpa. Desde el segundo paso
del sol por el cenit hasta la ceremonia de clausura del tzolkín,
que ocurre a la medianoche entre el 24 y 25 de octubre, median 76
días que corresponden a la segunda siembra y desarrollo del maíz.
Al finalizar los 260 días del tzolkín, comienza instantáneamente el
culto solar de 100 días que finaliza con los 5 días aciagos,
completándose así los 365 días del año maya."

De es.te modo se presenta la descripción del ciclo agrícola de

los chortí por Girard. Pero, ¿qué tan precisa es la relaci6n de
los hechos astronómicos con la realidad? La relación existe de
hecho para las regiones alrededor de 15 grados de latitud en donde
ocurre el primer paso del sol por el cenit al mediodía del 1 de
mayo, es decir, dos días después del período de la segunda
cuarentena, según Girard. El equinoccio de primavera sucede
también el 21 de marzo dos días después del período de la primera
cuarentena. El primer ciclo de 52 días finaliza el 19 de junio,
dos días antes del solsticio de verano. El segundo paso del sol
por el cenit ocurre el 12 de agosto cerca de los 15 grados de
latitud norte; el segundo ciclo de 52 días ya finaliza el 10 de
agosto. Con el ciclo agrícola según Girard llegamos al fin de los
ciclos siempre dos días antes de los fases astron6micos. En su
carta del 21 de agosto de este año Girard no puede por el momento
explicar bien en que consisten esos dos días de diferencia, sino
hasta cuando regresa a Guatemala, para consultar sus informantes,
los propios sacerdotes.

Ahora empieza la milpa de segunda con el primer ciclo de 40

días hasta el 19 de septiembre, una fecha elegida de una manera
interesante. Este es aquel día que se halla en el punto entre las
fechas del solsticio de verano y del solsticio de invierno, se
trata de un "día de la mitad del año". Claro que el equinoccio de
otoño sigue cuatro días después, el 23 de septiembre. Otros 36

137
días o 4 novenas siguen y el ciclo finaliza el 25 de octubre.

Siguiendo las fechas levantadas por Girard, pero mejorando su

cálculo por un traslado de dos días, el calendario se hubiera
organizado de tal manera que , pudiera ser controlado a base de
observaciones de los pasos del sol por el cenit. Girard no
menciona, si la observación es válida para el presente y no podemos
esperarla, porque las fechas no corresponden con los límites de los
ciclos descritos. La simetría y la coincidencia absolutas existen
solamente para un ciclo agrícola que empiece el 10 de febrero y
finalice el 27 de octubre. La situación en esta forma vale además
solamente para lugares entre la latitud geográfica de 15 grados y la
de 14014'.

En el calendario tzolkín de los chortí son de importancia los

equinoccios, pero no es posible verificarlos a través de
observa'c iones relativamente sencillas comparado con el método de los
solsticios como puntos extremos del sol en el horizonte. Con una
muy grande exactitud, sin embargo, es posible fijar la fecha de un
día a través de la observación de los pasos del sol por el cenit, es
decir, dos veces al año y solamente en los tr6picos. El control
de la primera observaci6n, p.ej. el 1 de mayo, es posible,
cuando hay sol, bajo 15 grados de latitud norte y el 12 de agosto,
104 días después. Por eso es posible hacer una correlaci6n del
ciclo agrícola, pero también de cualquier otro calendario que tenga
el curso del año solar como base. Como equipo o instalación basta
una pequeña apertura, p.ej. en el techo de una choza, de una casa
o de un templo. E's la misma si tuaci6n que nos hace saber Parsons
(1932) de un rito practicado por los sacerdotes de los Zuni (véase
Avení y Hartung, 1981, pág. 67). La puerta de la casa debe ser
cerrada para obscurecer el recinto. Antes se marca en el suelo el
punto originado por la vertical que empieza en la apertura, p.ej.
echando una plomada. En principio tenemos una cámara obscura y
podemos observar el curso de la luz del sol a través del punto
básico verticalmente señalado bajo la apertura en el techo. Una
construcción adecuada para la observación se halla a mi parecer en
el pozo de la cueva artificial de Xochicalco, Morelos (Tichy,
1980).

Hay grados de latitud señalados, en donde coinciden los pasos

del sol por el cenit con ciclos calendarios, como p.ej. en el caso
del tzolkín de los chortí. Bajo 15 grados de latitud la diferencia
entre los dos pasos del sol por el cenit asciende a dos veces 52
días. Bajo la latitud de 18 grados, p.ej. en La Venta, son dos
cuarentenas o cuatro veintenas, bajo la latitud de 19 grados, como
en Cholula o Xochicalco son 72 días u ocho novenas, bajo 20,5
grados, como en Uxmal, son tres veintenas.

Girard tiene razón en opinar que las mejores condiciones para

la invención del tzolkín agrícola están dadas en las regiones bajo
15 grados de latitud y además en la costa del Pacífico por causas
climatológicas. No obstante, no quiero dejar fuera la posibilidad
de que hayan existido ciclos agrícolas semejantes, bajo otros
grados de latitud ya señalados. Preguntemos ahora: ¿en qué forma
podría ser construido un calendario agrícola en Cholula o
Xochicalco?

138
 Podemos suponer los mismos ciclos parciales de 52, 40 y 36
días y una semejante simetría. Los pasos del sol por el cenit
ocurren respectivamente bajo 19 grados de latitud norte el 16 de
mayo y el 28 de julio. Entre el 17 de mayo y el 20 de junio hay 36
días y entre el 21 de julio, el solsti~io de verano, y el 27 de
julio hay otros tantos, es decir cuatro novenas. Los equinoccios
aquí no se pueden reconocer como límites entre ciclos. Un ciclo de
52 días empezaría el 26 de marzo y finalizaría el 16 de mayo. Más
agradable sería un ciclo de 40 días empezando el 7 de abril. Este
día es interesante porque el sol el 7 de abril desciende en el
horizonte con una desviaci6n de 7 grados de Oeste a Norte. La
desviaci6n es la que indica el eje de Tenochtitlan con su Templo
Mayor y de otros sitios más.

Por eso, dejemos empezar el calendario de 260 días con el

ciclo de 52 días el 14 de febrero y también dejemos finalizarlo con
52 días. Sin embargo, hacen falta 4 días para ser el 31 de
octubre y para cumplirse los 260 rlías. El primer día, el 14 de
febrero posee una buena relaci6n con respecto a una situación solar
y a los ejes piramidales que tienen la desviación de 13,5 grados de
Este a Sur. Direcciones de estas estructuras se conocen en
Teotenango, Kobah, Nohpat y Cozumel. En los días del paso del
sol por el cenit, el sol desciende en el horizonte con una
desviaci6n de 20 grados Oeste-Norte en el México central. Sin
entrar más en detalle, es digno de menci6n el hecho de que es
posible construir también para la latitud de 18 grados semejantes
ciclos astron6micamente relacionados. Se empieza de la misma
manera como en la latitud de 15 grados también el 10 de febrero con
un ciclo de 40 días.

Hay ba·stantes razones para hacer válida la opinión de que haya

existido una vez en Mesoamérica muy estrechas relaciones entre el
ordenamiento del espacio y el ordenamiento del tiempo en
coordinaci6n con la vida económica y el culto. A partir del curso
del sol se determinó la orientación del espacio y del tiempo en el
viejo mundo. Las variaciones del curso del sol determinan en suma
el curso de los fenómenos naturales, los de la atm6sfera y de este
modo se adapta el curso de las labores agrícolas de 260 días de
duraci6n de febrero a octubre. Probablemente se puedan hallar
otros también al Sur del Ecuador. Es posible controlar y
rectificar el calendario agrícola y sus ciclos con observaciones del
paso del sol por el cenit con una muy alta exactitud.

Referencias

Aveni, A.F. y H. Hartung, "The Observation of the Sun at the Time

of Passage through the Zenith in Mesoamerica", Archeostronomy,
3 (Journal ~ the History ~ Astronomy, 12), 1981, 51-70.

Beutler, G., "Algunas oraciones y ceremonias relacionadas con el

cultivo de maíz en México", Indiana (Berlín), 1, 1973, 93-
111.

139
Gi r a r d , R. , La e i v i 1 i z a c i ó n maya x._ sus e p i g o na l e s. Guatemala
C.A., s.~ (1982).

Larsen, H., "The 260-day Period as related to the Agricultural

Life of the Ancient Indian", Ethnos, 1, Estocolmo, 1936,
9-12.

Lewis, Osca r, Tepoztlán. Village Life in Mexico. New York,

1960.

Tichy, F., "Der ,Fes tkalende r Sahagún 's. Ein echter

Sonnenkalender? Lateinamerika Studien, 6, München, 1980,
115-37.

Weitlaner, R.J. y H.F. Cline, "The Chinantec", en Handbook of

Middle American Indians, 7 (1969), págs. 523-52.

140
 TABLA 1

El año agrícola en campos de quema en la montaña del centro y Sur de

M€xico. Ritos agrarios en TepoztlAn (L~wis, 1960).

Mes temporada labor ritos agrícolas

enero Última cosecha

febrero
marzo
abril
t. seca
1l
desmonte y roza
colocación de piedras
quemas
...
mayo siembra del ma1z San Isidro el 15 de mayo
bendición del maiz
junio lluvias siembra: maíz,
frijol, calabaza
julio deshierbo, limpia
pausa de
lluvias
agosto recalzar, acollar procesión en los campos
siembra de la a medidas de agosto
segunda milpa
septiembre lluvias primeras elotes cruces en las milpas
vigilar el 28 de septiembre
octubre (sachar 2. milpa)
noviembre cosecha: maiz,
frijol, calabaza
diciembre cosecha del maiz
transporte

141
 Tabla 2a El calendario agrícola de 260 días bajo 15º L.N.

A - meses cristianos
B los ciclos parciales en posiciones simétricas
C - tzolkin chorti (vease 2b)

Tabla 2b El tzolkin chorti desde el 8 de febrero al 25 de octubre

según Girard (1982) adaptado a los eventos solares
fechas días labor

10/2 21/3 equinoccio 40' desmonte

de primavera
22/3 - 30/4 primer paso 40 quemas· -verano-
del sol por
el cenit

:l
1/5 - 21/6 solsticio Milpa ciclo
de verano
de de
22/6 - 12/8 segundo paso
del sol por
primera las
el cenit
13/8 - 21/9 equinoccio
de estate
40} Milpa
de
lluvias

22/9 27/10 36 segunda -invierno-

260

11+2,
 Tabla 3a El calendario agrícola de 260 días bajo 19º L.N.
(Construcción a base de la distancia de los días
del paso del sol por el cenit)

.,., ,
._.,
Q. ,
I
I
I
I
I
I

'',~~r
✓✓/ •r, .• '
,,,,,,✓ -.

_.,,,..,..,
P DO\

IIX

Tabla 3b Los días del paso del sol por el cenit entre 15° y 23°N,
su distancia y las desviaciones de la puesta del sol

Latitud Pasos del sol Distancia Periodos Desviacibn de Lugar

Norte por el cenit dlas número la puesta dél
I II sol (acimut)

1sº mayo 1 - agosto 12 103 2 X 52 + 15°33' Copán/Honduras -

16° mayo 4 - agosto 9 97 + 16°40'
17° mayo 8 - agosto 5 89 + 17°48' Monte Albán
18° mayo 12 - agosto 1 81 2 X 40 + 10°58' La Venta
19° mayo 16 - julio 28 72 2 X 36 + 20º00· Cholula
20º mayo 20 - julio 23 64 + 21º21 • Tula
20°25 1 mayo 22 ·- julio 21 60 3 X 20 + 21°51 • Uxmal, c.Culiacán
21º mayo 25 - julio 18 54 + 22°34' Mérida
22º junio 1 - julio 11 40 2 X 20 + 23°50 1
AgUascalientes
23° junio 10- julio 2 22 + 2sºo1•
23°26,5' junio 21 o + 25°42' Alta Vista /JB/

143
JOHANNA BRODA

Ciclos agrícolas en el culto: un problema de la correlaci6n del

calendario mexica

Este breve estudio aporta da tos sobre la cor relaci 6n del

calendario mexica analizando la interrelaci6n que existía en este
calendario entre el clima y los ciclos agrícolas por una parte, y
los ciclos rituales y míticos por otra. Las fuentes son la
documentaci6n etnohist6rica del altiplano central en el último
momento antes de la conquista y la documentaci6n arqueol6gica de la
reciente excavaci6n del Templo Mayor de Tenochtitlan. A estas
fuentes se les aplica un ánalisis en términos de antropología social
vinculando la cuesti6n calendárica con las actividades econ6micas y
sociales. Nos hemos preguntado acerca de la funci6n social del
ritual calendárico. El culto mexica era, además, expresi6n de la
cosr.10visión; abarcaba un conocimiento práctico y una filosofía de
la naturaleza. En la última parte de este trabajo señalaré c6mo el
ciclo de Tlaloc en el Templo Mayor reflejaba a través del mito una
concepci6n filos6fica sobre los ciclos naturales de la lluvia y del
maíz.

Presento aquí las conclusiones de un estudio monográfico más

extenso sobre los ritos agrícolas (Ms.c), el cual, a su vez,
forma parte de una serie de investigaciones que he llevado a cabo
sobre la funci6n social del culto mexica.1 Este análisis ha
mostrado que existía, sin lugar a dudas, una correspondencia entre
los ritos, los fen6menos climatol6gicos y la agricultura. Las
cr6nicas del siglo XVI dan abundante testimonio al respecto. Hay
que concluir, por lo tan to, que si existía tal re laci6n (la
funci6n del culto consistía precisamente en reforzarla), entonces
debe haber existido algún método, aún desconocido, para mantener
el calendario en concordancia con el año solar.

La correlación del calendario mexica

Para el presente análisis, es fundamental contar con una

co rrelaci6n fija que nos permita relacionar los datos sobre
ceremonias y actividades econ6micas, con fechas exactas. Otro
elemento necesario para llevar a cabo el análisis, son ciertos
conocimientos básicos del clima, de la ecología y de los ciclos
agrícolas en el altiplano central. A este respecto, la
comparaci6n con da tos etnográficos modernos es de gran utilidad.2
Es de suponer que los factores climatológicos y ecol6gicos han
mantenido una constancia, a grandes rasgos~ desde el siglo XVI,
lo cual nos permite hacer tales comparaciones.
 CUADRO 1

Correlación del calendario mexica según Tichy y esquema

de orientaciones de los edificios pre-hispánicos

. ('
e, \\'o
(' ....,. Ss ":f
o co•·í•(S·
1/c.,~
{s.\\_
~ 1 //
·, ....
!:: • • 2' 15 ' \1/ .S\, l · •"'

~- ••
~- ·1'21'
1
,D
-~
.01,1 · •~

1 -SI ,l• •.;;

-OC ,C • •;;;
.e S_w
...........
~

146
 CLAVE DEL CUADRO 1

A- Calendario gregoriano para el año 1975

B- Los días con direcciones de la salida del sol correspondientes

al círculo C.

C- Desviaci6n axial de las pirámides contada a partir del E;

ángulos negativos de E a N indican la salida del sol en la
mitad estíal; ángulos positivos de E a S corresponden a la
parte invernal. Para la puesta del sol se aplican los ángulos
negativos en la mitad estíal y los ángulos positivos en la
mitad invernal. La orientaci6n hacia el O es determinante
para la mayoría de las pirámides.

D- "Tonalpohualli fijo", o calendario solar de 260 días + 100

días+ 5 días nemontemi, y sus subdivisiones.

E- Calendario mexica con 360 + 5 días. 18 períodos de 20 días+

5 nemontemi (19.-23.11.); más comúnmente, los nemontemi son
colocados antes de I Atlacahualo que comienza el 12 de febrero.
Este principio de- afio coincide con los datos de Sahagún,
cambiados a fechas gregorianas. Los 6 períodos de 60 días que
se registran en el círculo concéntrico exterior, indican los 6
rumbos del universo (E-N-0-S-Centro alto y Centro bajo).

F- Tonalpohualli de 260 días, dividido en 4 períodos de 65 días o

veinte semanas de 13 días.

Solsticios de verano y de invierno.

C19 Y C20,5 - Pasos del sol por el cenit en las latitudes de 19° y
20,5°N.

M- Días de la mitad del año, marzo 24 y septiembre 20.

CUADRO 1, según Tichy, diagrama 16.2.1977, basado en Tichy 1976,

fig.9; el texto con ligeras modificaciones de la autora.

147
 Esta tarea no se había emprendido anteriormente. Numerosos
investigadores han negado la correspondencia entre el culto mexica,
las estaciones y la agricultura. Un investigador tan destacado
como Karl Anton Nowotny que figura entre los mejores conocedores de
las fiestas del calendario mexica, negó, por ejemplo, la relación
entre los sacrificios de niños y los fen6menos climatológicos,
suponiendo que estos sacrificios, con el transcurso de los años,
se habían ido prolongando cada vez más desde el primer mes de I
Atlcahualo hasta IV Huey tozoztli. Si el año indígena perdía cada
cuatro años un d í a , ~ t a diferencia sumaba 25 días en 100 afios.
Es decir, en 100 años el calendario se hubiera desfasado por más
de un mes indígena. Nowotny rechazó rotundamente la posibilidad de
una correspondencia exacta entre las fiestas y el año solar debido a
la supuesta ausencia de métodos para corregir el calendario (1968a,
b).

Michel Graulich, partiendo de la misma suposición, propone

otra hipótesis difícil de comprobar: que los nombres de los meses
mexica tuvieran su origen en el año 682 d.C., durante la época
clásica, cuando habían correspondido con los fenómenos
climatol6gicos, mientras que en el momento de la conquista se
encontraban aproximadamente 5 meses europeos fuera de
correspondencia con el año solar. El ciclo calendárico mayor en el
que fiestas y fechas solares hubieran coincidido nuevamente, serían
1460 años. Graulich hace un análisis estructural del contenido
mítico de las fiestas, tomando como punto de referencia la supuesta
correspondencia para el año 682 d.C. (1976, 1980).

El presente trabajo se basa en la correlación fija entre meses

prehis%ánicos y fechas solares que ha sido propuesta por Franz
Tichy. No podemos explicar aquí en detalle la compleja
metodología aplicada por Tichy para elaborar esta correlación; sólo
sefialaré que en ella se combinan investigaciones arqueo-
astronómicas, arqueológicas, etnohistóricas, de etnografía y
geografía cultural moderna. El punto de partida es el principio de
orientación de pirámides y centros ceremoniales prehispánicos. En
la última década, Anthony F. Avení, Horst Hartung, Tichy y
algunos otros autores han emprendido un estudio sistématico de las
desviaciones de los puntos cardinales en los ejes principales de
edificios y sitios arqueológicos. 6 En base a estas mediciones de
campo, Tichy logró establecer un sistema de ángulos recurrentes en
las desviaciones, sistema que se vincula al mismo tiempo con
ciertas subdivisiones básicas del calendario y conduce así a una
correlación fija de éste con el año solar. (Véase Cuadro 1)

El particular interés de estas investigaciones para la

etnohistoria consiste en el hecho de que Tichy llegó, a partir de
datos obtenidos independientemente de la documentación
etnohistórica, a reconstruir la estructura interna del calendario
prehispánico, y a establecer su correlación fija. Es de notar que
esta correlación comprueba, además, la exactitud de las fechas que
dio fray Bernadino de Sahagún para los 18 meses aztecas en el Libro
II de su Historia General. Por lo tanto, la correlación de Tichy
que usamos en el siguiente análisis, coincide con las fechas de
Sahagún (Tichy 1980). (Véase Cuadro 2)

148
 CUADRO 2

Correlación fija del calendario mexica

(Según Sahagún, Libro II; Tichy 1976, véase Cuadro 1)

MESES PREHISP ANICOS CALENDARIO CRISTIANO (GREGORIANO)

I Atlcahualo 12.2. - 3.3.

II Tlacaxipehualiztli 4.3. - 23.3.

III Tozoztontli 24.3. - 12.4.

IV Huey tozoztli 13.4. - 2.5.

V Toxcatl 3.5. - 22.5.

VI Etzalcualiztli 23.5. - 11.6.

VII Tecuilhui tontli 12.6. - 1.7.

VIII Huey tecuilhui tl 2.7. - 21.7.

IX Tlaxochimaco-Miccailhuitontli 22.7. - 10.8.

X Xocotlhuetzi-Huey miccailhu.itl 1 l. 8. - 30.8.

XI Ochpaniztli 31.8. - 19.9.

XII Teotleco 20.9. - 9.10.

XIII Tepeilhuitl 10.10.- 29.10.

XIV Quecholli* 30. 10.- 18.11. *

XV Panquetzaliztli 19.11.- 8.12.

XVI Atemoztli 9.12.- 28.12.

XVII Tititl 29 • 12 • - 17 • 1 •

XVIII Izcalli 18.1. - 6.2.

Nemontemi 7.2. - 11.2.

* En la correlación de Tichy (Cuadro 1), se intercalan aquí los 5

nemontemi. El problema de los nemontemi está lejos de estar
resuelto. Sahagún los coloca entre XVIII Izcalli y I Atlcahualo,
sistema que seguimos aquí provisionalmente.
 Sabemos que los antiguos mexicanos conocían desde la época
clásica varios métodos que les permitían observar la verdadera
duraci 6n de 1 año tropical (365, 25 días), conoc imi en to que se
encuentra reflejado en inscripciones jeroglíficas mayas de Copán,
Palenque, Tikal y Quiriguá. 7 Entre los métodos de observaci6n del
año solar figura el observatorio de Xochicalco, orientado hacia el
cenit y el . solsticio de verano (Broda, Ms.b., fig. 18). La
verdadera duraci6n del año también se puede conocer en base a la
recurrencia de la posici6n de ciertas estrellas. Parece que los
mexica registraban con tal prop6sito la salida helíaca de las
Pléyades a fines de mayo que ocurría poco después del primer paso
del sol por el cenit. Tal coincidencia confería a estos fen6menos
una doble validez en la latitud geográfica de Tenochtitlan. 8

Estos métodos permitían mantener el calendario en

correspondencia con el año solar, las estaciones y los ciclos
agrícolas. Sin embargo, no se ha descubierto hasta hoy en día el
método exacto con el cual los antiguos mexicanos hacían coincidir
los fen6menos astron6micos con los ciclos calendáricos, los cuales,
como múltiples de 20 (260 días, 360 días; 52, 104 años, etc.) no
admitían las intercalaci6n de un día sin que se estropeara la
armonía interna del sistema. El presente estudio etnohist6rico no
contribuye a solucionar este problema desde el punto de vista
caltndárico-cronol6gico sino que más bien aporta datos sobre la
manera en que se fijaba, ritual y socialmente, en el calendario
mexica la recurrencia de los fen6menos climatol6gicos y agrícolas.
El culto era el puente entre astronomía, calendarios, ciclos
naturales, economía y cosmovisi6n, formando un todo que era
legitimado por medio de la religi6n.

Ciclos agrícolas y culto mexica

La característica geográfica fundamental de Meso américa ha

sido, desde siempre, su enorme variabilidad de alturas, climas y
suelos. Este hecho ya fue señalado muy pertinentemente a fines del
siglo XVI por el cronista fray Francisco de las Navas: "en estas
partes de las Indias hay diferentes climas que casi en todo el afio
se siembra y coje maíz en las tierras cálidas y templadas y en las
serranías" (CFN 150)

A pesar de estas condiciones ecol6gicas tan diversas que

perduran hasta hoy en día, se pueden distinguir, a grandes rasgos,
dos ciclos en el altiplano central: el de regadío y el de temporal.
Los sistemas agrícolas mesoamericanos han mantenido una continuidad
básica desde la época prehispánica hasta la actualidad. En el ciclo
de regadío se siembra en enero o febrero y se cosecha en junio y
julio; a los dos meses ya puede haber jilotes. En el ciclo de
temporal existe una gran variabilidad en la siembra según la altura
y la calidad de los terrenos, aún en los que pertenecen a un mismo
pueblo. Naturalmente, esta variaci6n es aún más grande entre
diferentes lugares. La siembra se hace hoy día entre marzo y
junio. Para la época prehispánica informa Durán que la siembra de
los montes se iniciaba ya en febrero, porque allí empezaban los
aguaceros antes que bajaran a los llanos: "y sembrábanlos tan
temprano a causa de la humedad de los montes, que, según ellos

150
dicen, siempre empi2zan por allí los ~uaceros muchos días primero
que bajen a los llanos" (1967 I: 292).

Lo más general era que se sembrara entre fines de abril y

principios de junio, y la cosecha tenía lugar entre fines de
octubre, noviembre y diciembre. Veamos la correspondencia entre
estos ciclos y los ritos de los meses mexica. Todos los datos que
aquí se presentan de manera resumida se basan en citas detalladas de
las cr6nicas del siglo XVI y su interpretaci6n se combina con
datos etnográficos actuales.lo

El ciclo de regadío

El ciclo de regadío tenía lugar en Tenochtitlan entre los meses

XVIII Izcalli/I Atlcahualo y VI Etzalcualiztli (enero/febrero hasta
junio). El primer mes del año, según Sahagún, I Atlcahualo,
tenía una serie de variantes de su nombre: se lla~aba también
Atlmotzacuaya-Quahuitlehua-Xilomanaliztli.

Atlcahualo "se detiene el agua"

Atlmotzacuaya "se ataja el agua"

Quahuitlehua "levantamiento de los postes"

Xilomanaliztli "ofrenda de jilotes" 11

Los tres primeros nombres con sus ritos correspondientes se

referían al hecho de la falta de agua en esta época del año, y a
ritos mágicos para atraer la lluvia. El último nombre,
Xilomanaliztli, contenía una referencia al comienzo del ciclo
agrícola de regadío, puesto que ya se hacían ofrendas de primicias
de jilotes o más bien, se pedía que éstos salieran pronto. Al
mismo tiempo empezaban los sacrificios de niños. Estos niños
representaban a los tlaloque, los pequeños servidores del dios
Tlaloc, y se identificaban con cerros específicos en los
alrededores de Tenochtitlan. Se trata de ritos mágicos para atraer
la lluvia en la época del máximo estío que comienza en febrero y se
prolongará hasta mayo. Al mismo tiempo, estos sacrificios se
relacionaban con el ciclo de regadío a través de la edad de los
niños que iba aumentando conforme crecía el maíz (Moto linía 19 6 7:
64, 65). Tenían lugar entre I Atlcahualo (a veces empezaban unos
meses indígenas antes) y IV Huey tozoztli; en este último mes,
correspondiente a abril, culminaban en la gran fiesta de Tlaloc en
el cerro del mismo nombre y en el remolino de la laguna, Pantitlan
(Broda 1971: 272-81).

El ciclo de regadío terminaba en Tenochtitlan con la fiesta de

VI Etzalcualiztli (23.5.-ll.6.). 12 A la mitad de este mes la gente
preparaba en sus casas la comida del etzalli que eran "unos boli tos
de masa de maíz guisados con frijoles". Durán informa que en esta
época del año los elotes eran ya bastante grandes y, debido a que
el año ofrecía buenas perspectivas, se daba permiso general para
comer el etzalli (1967 I: 259). Esta comida denotaba abundancia y
la recibían también unos limosneros que andaban de casa en casa,

151
bailando "el baile del etzalli" en representación del dios Tlaloc.
Estos limosneros traían prosperidad a las casas donde entraban.
Según el cronista Juan de Tovar:

En estos días, los labradores habían ya labrado la tierra y

salían en el hábito ••• (de Tlaloc)... diciendo al pueblo pués
por los labradores gozarían de tal semilla de pan como el maíz
cuya insignia traían, que era raz6n se lo gratificasen y así
todos les echaban en las ollas muchas cosas de comida,
especialmente de ésta de frijoles y maíz ••• y así todos se
holgaban estos días porque en ellos descansaban de haber
labrado y cultivado la tierra. (Kubler y Gibson 1951: 25)

En esta fiesta observamos otro fen6meno interesante: la

cosecha del ciclo de regadío se encuentra íntimamente mezclada con
ritos referentes al ciclo de temporal, puesto que en la misma
fiesta se celebraba la terminación de la siembra de temporal (o de
las primeras labores), y se hacían ritos para que descansaran los
instrumentos de trabajo. El ciclo de regadío hasta aparece
subordinado al simbolismo principal de Etzalcualiztli. Las
ofrendas de matas verdes de maíz, de jilotes, de elotes y la
comida de etzalli-- procedentes todos ellos del ciclo de regadío--
servían como analogía mágica mediante la cual se quería provocar un
igual desarrollo de la siembra de temporal. Así la comida del
etzalli denotaba abundancia y pronosticaba un feliz crecimiento del
maíz del temporal que apenas se había sembrado. Estas
circunstancias señalan a nivel simbólico que, de hecho, se
atribuía mayor importancia económica al ciclo de temporal, y que el
ciclo de regadío era considerado como secundario.

Otro hecho sumamente relevante tenía lugar durante el mes de

Etzalcualiztli. Era el comienzo pleno de la estación de lluvias.
Aunque solían ocurrir temporales esporádicos durante los meses
anteriores, éstos no se hacían verdaderamente abundantes hasta este
mes. Este acontecimiento era celebrado mediante el sacrificio del
dios Tlaloc y de su esposa Chalchiuhtlicue en el Templo Mayor, el
día último del mes; sacrificio que denotaba la culminación de los
ritos de la estación seca, principalmente de los sacrificios de
nifios por medio de los cuales se había pedido el agua. Los
ejercicios rituales que los sacerdotes de Tlaloc hacían a lo largo
de Etzalcualiztli, tenían asimismo la finalidad de provocar por
medios mágicos, el deseado comienzo de las lluvias (Broda 1971:
283-86, 290-91). Nos referiremos más adelante a la importante
cuestión de la división del año en estación seca y época de lluvias.

El ciclo de temporal

En cuanto al ciclo de temporal hemos visto que hasta la

actualidad, la fecha de la siembra depende de factores de altura y
del microclima en terrenos específicos; se 1n1c1a, por lo general,
cuando han caído las primeras lluvias. En el culto mexica se
pueden detectar a lo largo de los meses l Atlcahualo hasta VI
Etzalcualiztli (febrero-junio), ritos cuya finalidad era la de
fomentar el buen desenlace del ciclo agrícola de temporal. La
fiesta propriamente dicha que iniciaba ritualmente la siembra de

152
temporal, era IV Huey tozoztli (13.4.-2.5.); en ella se usaban las
matas verdes delmaíz de regadío, para atraer la fertilidad para el
ciclo de temporal. La fiesta giraba alrededor de la bendición del
maíz para la siembra (in cintli, xinachtli iez). Estas mazorcas
se habían guardado en manojos de a siete desde la cosecha del ario
pasado; tales manojos se llamaban ocholli, _ pero también llevaban
el nombre de "dios mazorca", cintli o cinteotl. Adornados con
papel rojo goteado de ulli líquido, estas mazorcas eran llevadas
por las doncellas (ichpopochtin) al templo de Chicomecoatl, el
Cinteopan. "Allí éstas... se convertían en corazones, en los
corazones de 1 granero... en maíz para la siembra" (CF II: 60-62).
Al volver a sus casas, la gente las ponía dentro de la troje,
convirtiendo así el demás maíz almacenado en maíz para la siembra la
cual iba a tener lugar pronto.

Con respecto a esta fiesta, la comparación con datos

etnográficos modernos es sumamente relevante: la fiesta de la Santa
Cruz que se celebra hoy en el día 2 de mayo en las regiones
indígenas de México y Guatemala, es junto con el día de los muertos
aquella fiesta católica que ha recogido más elementos de la
tradición prehispánica. Se trata precisamente de una fiesta de la
bendición del maíz para la siembra y del inicio del ciclo agrícola
de temporal, en la que se invoca, además de la deidad del maíz, a
la tierra y a las lluvias. La diosa prehispánica del maíz,
Chicomecoatl, merge con la deidad de la tierra y ambas llegan a
ser personificadas en el símbolo católico de la Santa Cruz (sin
embargo, generalmente se adoran cuatro cruces y su simbolismo
dista mucho del católico). Los lugares de culto son fuentes y
ríos, sobre todo, lo alto de ciertos cerros donde los ritos más
importantes tienen lugar durante la noche o al amanecer. 13

Es de notar que en el mes prehispánico de IV Huey totoztli

terminaba también el ciclo de los sacrificios de niños que se habían
hecho desde febrero hasta abril "para que los tlaloque diesen agua"
(CF II: 5). Aunque hemos visto que los sacrificios de niños
correspondían por una parte a las fases de crecimiento de las
sementeras de regadío, por otra producían la lluvia necesaria para
el inicio de la siembra de temporal. Esto es también la
explicación porque terminaban justo en el mes de la siembra de
temporal, aunque en esta época todavía no se había garantizado
plenamente la caída de las lluvias.

Curiosamente, según los datos etnográficos actuales de

Citlala, se hacen en la fiesta de la Santa Cruz ofrendas de ropa de
miniatura por los niños que han muerto en · el pueblo desde la fiesta
del año anterior. "No debe faltar ninguno, por que ellos son
'acompañantes' de la Santa Virgen de la Cruz', y así le ayudarán a
hacer que llueva pronto; los niños muertos 'ya han alimentado a la
tierra', están cerca de ella" (Olivera 1979: 154). De manera
sorprendente encontramos aquí la sobrevivencia de los dos elementos
básicos del culto de los niños muertos ("sacrificios de niños", en
la época prehispánica): su relación con la lluvia y con la
agricultura.

Durante el siguiente mes prehispánico, de V Toxcatl (3.-

22.5.), continuaban las súplicas y los ritos mágicos- por atraer la

153
lluvia que culminaban, como hemos visto, en la fiesta del dios
Tlaloc, en VI Etzalcual.iztli. Hasta entonces ya se había
sembrado, y la entrada de las lluvias estaba plenamente
garantizada.

Actualmente, los campesinos de Huexoyucan, en el altiplano de

Tlaxcala, conciben el año agrícola (de tem·poral) dividido en tres
ciclos de trabajo, de aproximadamente tres meses cada uno. El
primer ciclo es el de la siembra e incluye la siembra, la primera
labra, la segunda labra y el aterrado (abril-junio
aproximadamente). El segundo ciclo supone tres meses de descanso
del trabajo de la tierra (julio-septiembre). En esta época los
campesinos tienen tiempo para dedicarse a otras labores, hoy día es
el trabajo temporal y migratorio. La dobla que se ejecuta en este
tiempo, no exige una mano de obra fuerte. El tercer ciclo
significa tres meses de trabajo del segundo ciclo agrícola, o sea
todas las actividades relacionadas con la cosecha incluyendo el
amogotar, la pizca, la seca, la limpia, el almacenaje, el
barbecho y la vifl' que abarcan de fines de septiembre o octubre
hasta diciembre.

De acuerdo a la misma secuencia de ciclos que existe en la

actualidad, la siguiente fiesta prehispánica de un contenido
agrícola predominante era XI Ochpaniztli (31.8.-19.9.). En ella se
celebraba el casamiento ritual de la diosa madre de la tierra con el
dios solar, y como hijo nace Cinteotl, el dios mazorca. En
relación al abundante simbolismo agrícola y de la fertilidad
presente en Ochpaniztli (ofrendas de primicias de maíz y de otros
elementos), numerosos autores han sugerido que se trate de la
fiesta de la cosecha prehispáncica. Sin embargo, E. Seler quien
fue el primero en proponer esta interpretación, se basa en una
analogía equivocada con las condiciones climáticas europeas, donde
la cosecha efectivamente tiene lugar en esta época. En México sólo
se da la cosecha en tierras calientes (actual edo. de Morelos, por
ejemplo), pero faltan aproximadamente dos meses para que las
cosechas tengan lugar en las condiciones climáticas del altiplano
central.

Según lo que hemos visto, en relaci6n con las fiestas

anteriores, en el culto mexicano se celebraban tanto los fenómenos
acabados gracias a los dioses por haberlos concedido, sino que la
principal función de los ritos era la de provocar un buen desenla~e
de los fenómenos deseados. Así, los sacrificios de niños debían
atraer las lluvias y garantizar el crecimiento de las sementeras de
regadío. IV Huey tozoztli era la fiesta en preparaci6n de la
siembra donde las mazorcas secas eran transformadas mediante los
ritos en maíz para la siembra. En Etzalcualiztli hemos visto que
se celebraba el comienzo de la cosecha de regadío y al mismo tiempo
significaba el desenlace de la estación de lluvias. De la misma
manera, XI Ochpaniztli no era la fiesta de la cosecha, sino que su
propósito era e 1 de conjurar , mediante los ritos , e 1 bu en
cumplimiento de la maduración del maíz, . mientras que la cosecha
s6lo tenía lugar unos dos meses después, correspondientes a XIV
Quecholli (noviembre) y XV Panquetzaliztli (noviembre/diciembre)
respectivamente. Encontramos una comprobaci6n de esta
interpretación en el hecho de que en esta última época ya no se
hacía ninguna fiesta de la cosecha propiamente dicha, a diferencia
de la concepci6n cat6lica de expresar mediante el culto el
agradecimiento a dios por haber concedido los frutos agrícolas.

Con esto pasamos al tercero y último sub-ciclo agrícola, el

correspondiente a los meses de octubre y diciembre y la terminaci6n
del ciclo de temporal. Esta parte del año, a partir de octubre,
corresponde también climáticamente al final de la estaci6n de
lluvias. Es cuando se hacían en Tenochtitlan las ceremonias a los
cerros y a los dioses del pulque. Quizás, se puede explicar el
culto de los tlaloque en su aspecto de cerros en el siguiente
sentido. Cuando empezaba la estaci6n seca, el culto de los
cerros significaba que ahora éstos estaban reteniendo el agua para
luego soltarlo cuando empezaban de nuevo los sacrificios de niños.
Además de esto, los cerros tenían una relaci6n particular con
varias actividades que se iniciaban en es ta época de 1 año como era
la caza, el cortar madera para la construcci6n, el desmonte etc.
Todas estas actividades se derivaban de hecho de los cerros.

Los cerros estaban también específicamente relacionados con el

cultivo del maguey al cual los campesinos se dedican sobre todo en
la época seca. No s6lo se da el mejor pulque en la estaci6n seca,
sino que también es cuando se trasplantan los magueyes y después de
las cosechas se arreglan los bordes de los terrenos terraceados
donde están plantados los magueyes. El maguey tiene un sentido muy
unido a la agricultura de temporal, sobre todo al cultivo con
terracería donde se planta para evitar la erosi6n del suelo. 15 Al
mismo tiempo, el maguey sirve para la producci6n del pulque y son
bien conocidos los usos múltiples de esta planta mesoamericana que
la ha convertido en el símbolo absoluto de la fertilidad. Las
borracheras . rituales que tenían lugar en estos meses prehispánicos,
se vinculaban con el culto de los cerros, y en este aspecto eran
una expresi6n de la cosecha. Los dioses del pulque eran al mismo
tiempo dioses del desmonte (llevaban el hacha de carpintero en la
mano) y estaban íntimamente relacionados con los dioses de la lluvia
y de los cerros; ambos eran dioses patrones de lugares concretos
que compartían varios rasgos iconográficos. Estos dioses se
festejaban en la época del año apropiada para la serie de
actividades que representaban (XII Pachtontli y XIII Huey pachtli-
Tepeilhuitl) • 16

En el mes siguiente, XIV Quecholli (30.10.-18.ll.), empezaban

las cazas de venados, jabalíes, ansares, grullas y patos, y se
hacían ceremoniss al dios de la caza que culminaban en la gran
fiesta de Mixcoatl-Camaxtli (CFN: 167-69). Esta fiesta demuestra
la importancia que aún tenía la caza en la vida y mentalidad de los
mexica. Octubre-noviembre efectivamente son los meses cuando los
animales de caza ya no tienen cría, habiendo procreado en época
caliente; por lo tanto es cuando se les puede cazar. 17 Además,
llegan en esta época los pájaros migratorios del Norte (la Florida
etc.), entre los que se mencionan los ansares, las grullas y los
patos silvestres, y sobre todo el flamenco cuyo nombre en nahuatl,
quecholli o tlauhquechol, ha dado el nombre al decimocuarto mes
(Torquemada 1975-80, t.3, Libro X: 403, 426 y 427).

155
 Por otra parte, éste y el siguiente mes de XV Panquetzaliztli
(19.ll.-8.12.) eran la época apropiada para iniciarlas guerras que
seguían hasta XVII Tititl y XVIII Izcalli (Kubler y Gibson 1951:
32). El hecho de que se hicieran las campañas militares después de
haber terminado las cosechas y los trabajos agrícolas, demuestra
que la economía del México prehispánico e-staba todavía determinada
por la agricultura.

En XV Panquetzaliztli se celebraba la fiesta del patr6n del

grupo étnico mexica, Huitzilopochtli, con un destacado significado
político y guerrero, conectado con el culto solar. El solsticio
de invierno tenía lugar unos días después, durante el mes de XVI
Atemoztli (9.-28.12.), que en términos míticos significaba el culto
del sol en su paso por el río del inframundo (Cfr. Carrasco 1979:
57). En este filtimo mes se enlazaba el simbolismo solar con el de
la lluvia y de los cerros. Cuando caían los primeros aguaceros en
los montes, Tlaloc recordaba a los hombres que empezaran a pedir la
lluvia mediante los sacrificios de niños, iniciándose de esta
manera de nuevo el ciclo de ritos que se extendería a lo largo de
toda la estación seca (Broda 1971: 312 y 313).

En la siguiente fiesta, de XVII Tititl (29.12.-17.1.),

encontramos ciertos rasgos q1:1e parecen simbolizar al maíz "viejo" en
la época fría y seca del invierno. Se daba culto al dios del fuego
Xiuhtecutli-Huehueteotl, el más anciano de todos los dioses, así
como a Ilamatecuhtli, la "señora vieja' que, en cierto modo,
representaba al maíz viejo y al barbecho. Se hacían unas extrañas
ceremonias con su "troje de maíz' a la cual se prendía fuego (CF II:
144 y 145). Es de notar que en esta época del affo se limpian
actualmente los trojes, se saca el maíz viejo (del año anterior) y
se mete el maíz nuevo que acaba de secarse después de las
cosechas. 18

En XVIII Izcalli (18.1.-6.2.), mes cuyo nombre quiere decir

"crecimiento", se hacían ritos para el crecimiento de los niños,
estirándoles sus extremidades para que crecieran. Cabe pensar en
la posibilidad de que existía un cierto paralelo entre los niños y
las sementeras de regadío que se empezaban a cultivar en este
momento. El siguiente mes de I Atlcahualo, marcaba no solamente la
siembra de regadío y el inicio de los sacrificios de niños a gran
escala, sino que, además, para Tenochtitlan, era el principio
del año civil y cronol6gico, según informa Sahagún (CF II: 1).
Además, era un mes de borracheras generalizadas; hasta a los niños
pequeños se les hacía tomar pulque (CF II: 148-53).

Culto, cosmovisión y "filosofía de la naturaleza"

Finalmente, quisiera añadir otra reflexi6n sobre el contenido

mítico del ciclo anual de fiestas. De acuerdo con los fen6menos
climatol6gicos, la divisi6n básica del año se hacía en Xopan, lit.
"tiempo verde"-- la estación de lluvj_as-- y tonalco, "tiempo de
sol, y calor"-- la estaci6n seca. 1 ~ Tlaloc como el dios de las
lluvias, del agua y de la fertilidad en general, presidía sobre la
mitad del año que se iniciaba con la fiesta de VI Etzalcualiztli.
Era la estación de lluvias, la "época oscura" del año relacionada
con la noche, la luna, con Venus y las estrellas (incluyendo a las
Pléyades), así como con los muertos y el inframundo. Fue durante
esta época que la planta del maíz crecía y maduraba, y este proceso
en su complicada mitología fue representado en el culto durante los
meses de VII Tecuilhuitontli-XI Ochpaniztli (junio-septiembre). La
maduraci6n del maíz parece haber sido equiparada, a nivel mítico,
con el viaje del sol por el inframundo. Esta era la época del "Sol
Nocturno", símbolo de Tlaloc. 20

Por otra parte, la mitad del año llamada tonalco que se

iniciaba con XV Panquetzaliztli, era presidida por el sol como
deidad del cielo diurno. Esa deidad era Huitzilopochtli y su
nacimiento en el cerro Coatepec era celebrado en los ritos de
Panquetzaliztli. La terminaci6n de la cosecha de temporal
pertenecía a esta parte del año -tonalco, el calor del so1. 21

Esta divisi6n mítica del año tiene un particular interés ya que

se ve representada mediante numerosos símbolos en el dualismo del
Templo Mayor, fen6menos que pueden ser analizados en los hallazgos
de la reciente excavaci6n de este templo. En otro trabajo exploro
con mayor detalle este simbolismo (Ms.a), lo cual me ha permitido
concluir que esta di visi6n del año, en estaci6n. seca y de lluvias,
y su interrelaci6n con el ciclo agrícola del maíz del temporal,
imprimieron su huella profunda a la cosmovisi6n mexica,
reflejándose no s6lo en la estructuraci6n misma del ciclo anual de
fiestas, sino también en la arquitectura, las esculturas y las
ofrendas del Templo Mayor.

A la estaci6n seca pertenecía el culto de los tlaloque como

dio ses de los cerros. Los cerros que retenían el agua en su
interior, para soltarlo de nuevo en tiempo de lluvias. Los cerros
tenían también una relaci6n más profunda con el maíz, ya que éste
era guardado en su interior. Los tlaloque eran los dueños
originales del maíz y éste último tenía que ser adquirido por los
hombres mediante un contrato para con los dioses de la lluvia.
Este contrato por excelencia eran los sacrificios de nifios,
nextlahualli, "la deuda pagada", y ellos tenían lugar precisamente
en la época más seca del año (Cfr. Broda 1971: 272-76). Los
aspectos mencionados muestran una coherencia interna entre
observaci6n de la naturaleza, actividades econ6micas y sociales, y
la superestructura religiosa. En cierto modo, configuraban lo _que
podríamos llamar "la filosofía mexica de la naturaleza". En cuanto
a las actividades sociales y econ6micas, el calendario mexica
mostraba una planeaci6n tan detallada _que en él cada cosa y cada
actividad tenían su momento apropiado cuando debían llevarse a cabo
fielmente. S6lo era posible esta coordinaci6n de los ritos con la
vida social, si se usaba algún método para mantener el calendario
en correspondencia con el año solar.

La aportaci6n de este breve trabajo no ha sido la de derrocar

la modalidad de c6mo se hacían estas correciones peri6dicas del
calendario sino que más bien la de llamar la atenci6n sobre la
estructura interna del culto en su relaci6n con los ciclos
recurrentes del clima y de la agricultur~ El tema del contenido
de los ritos también debe ser tomado en cuenta al abordar la
cuesti6n cronol6gica del calendario mexica.

157
NOTAS

1 Broda 1971; 1978; 1979; 1982a, b; Ms.a, b.

2 Los datos etnográficos que usé para la comparaci6n provienen de

la regi6n chinampera del Valle de México (San Luis Tlaxialtemalco,
Xochimilco) (Peña 1978), con una agricultura de irrigaci6n, y del
altiplano de Tlaxcala (San Mateo Huexoyucan) (Bilbao Ms.), donde el
cultivo es de temporal. Los datos comparativos sobre la Fiesta de
Santa Cruz (véase adelante) se derivan de más lejos, de Citlala,
Guerrero, una regi6n nahua donde se ha conservado una cultura
indígena muy tradicional. (Cfr. Olivera 1979; Sepúlveda 1973;
Suárez 1978).

3 Con esto no quiero decir, de ninguna manera, que en otros

campos de la economía y estructura social y política los cambios que
ocurrieron a partir de la conquista, no hayan tenido una
trascendencia fundamental. Más aún que estamos hablando del Valle
de México, centro del estado mexicano moderno.

4 Los números romanos se refieren a la secuencia de los meses

según Sahagún (HG y CF), en la cual el año empieza con I Atlcahualo
(cfr. Cuadro 2)-.- --

5 Tichy 1976a, b· 1978; 1980; Ms.

'
6 Aveni ed. 1975, 1977; Avení 1980; Aveni, Hartung y
Buckingham 1978; Aveni, Hartung y Kelley 1982; Hartung 1975,
1977; Tichy cfr. nota 5, y la bibliografía citada en Avení 1980 y
Broda Ms. b.

7 Cfr. Aveni 1980: 170-73. Se han encontrado referencias que

reflejan tales conocimientos, sin embargo, no se conoce hasta el
momento ningún registro carente de ambiguedad que demuestre la
práctica peri6dica de correcciones del calendario solar.

8 Cfr. Broda 1982b. La observaci6n de la salida helíaca de las

Pléyades en combinaci6n con el primer paso del sol por el cenit,
parece haber sido importante desde la época teotihuacana.
Alrededor del año 150 d.C., ambos fen6menos coincidían en la misma
fecha en la latitud geográfica de Teotihuacán (19° 42': 18 de mayo)
(Aveni 1980: 225).

158
9 Hasta hoy en día, en la tan destruída ecología del Valle de
México, se conservan marcadas diferencias climatol6gicas. En las
lomas de la sierra del Ajusco, por ejemplo los pueblos de Ajusco,
Contreras etc., suelen caer aguaceros durante los meses de
diciembre y enero dada la diferencia de elevación de tmos 200-400m.
con las partes centrales del Valle.

10 Según he indicado arriba, esta ponencia presenta s6lo una

síntesis de un estudio más extenso que estoy preparando para su
publicación (Ms.c). La interpretaci6n general que se maneja de las
fiestas de los meses mexica se basa, naturalmente, en toda una
serie de estudios previos que me han permitido lograr esta
comprensi6n global. (Cfr. Broda 1971; 1978; 1979; 1982a, b;
Ms. a, b).

11 Las fuentes se citan en Broda 1969: 19, 20; y 1971: 269ss.,

nota 18.

12 Para una descripci6n detallada de esta fiesta basada en las

fuentes, véase Broda 1971: 282-98 y Ms.c.

13 Para esta interpretación del simbolismo de la fiesta actual de

Santa Cruz, me baso en datos de la regi6n nahua de Guerrero,
provenientes de Citlala, Ostotempa (Olivera 1979; también
Sepúlveda 1973 y Suárez 1978) y Ameyaltepec (observación propia, 2
de mayo 1980).

14 Los datos de San Mateo Huexoyucan, Tlaxcala, proceden de las

notas de campo de Jon Ander Bilbaoj 1978-79 (Ms.). Quisiera
expresar mi agradecimiento a Bilbao, profundo conocedor de la
ecología y la agricultura tradicional de esta región del altiplano,
por sus valiosas informaciones así como por sus comentarios a mi
trabajo.

15 Información proporcionada por J.A. Bilbao (Ms.).

16 Para una descripci6n detallada de -estos ritos, véanse Broda

1971: 301-312, y 1979. En aquellos trabajos analizo también la
participación de los grupos sociales en estas fiestas.

17 Información proporcionada por J.A. Bilbao.

18 San Mateo Huexoyucan, notas de campo de J.A. Bilbao (Ms.).

19 Malina 1970; Carrasco 1979: 53.

159
20 Cfr. Broda Ms.a. En cuanto a la iconografía de Tlaloc corno
Sol Nocturno, cfr. Pasztory 1974 y Klein 1980.

21 Mientras que el ciclo de cultivo de temporal empezaba a fines

de la estaci6n seca y se prolongaba a través de la estaci6n de
lluvias para terminar con las cosechas a principios de la época
llamada tonalco; el ciclo del cultivo de riego pertenecía
exclusivamente a esta última parte del año. Es de notar que los
cultivos de humedad se suelen llamar hasta hoy en día, tonarnil,
"milpas de estío" (Cfr. Carrasco 1979: 56). Además, el maíz de
regadío se llama tonalcentli, según indica Malina.

160
BIBL IOGRAFIA

Avení, Anthony F. (ed.)

Archaeoastronom y in Precolumbian America. Austin: University

of Texas Press. -(1975)

Native American Astronomy. Austin: University of Texas

Press. ( 1977)

Avení, Anthony F.

Skywatchers of Ancient Mexico. Austin: University of Texas

Press. (1980)

Avení, Anthony F., Horst Hartung y Beth Buckingham

"The Pecked Cross Symbol in Ancient Mesoamerica", Science,

202, 267-79. (1978)

Avení, Anthony F., Horst Hartung y J. Charles Kelley

"Alta Vista (Chalchihuites ): Astronomical Implications of a

Mesoamerican Ceremonial Outpost at the Tropic of Cancer".
American Antiquity, 47, 2, 316-35. (1982)

Bilbao, Jon Ander

"Notas de campo de S. Mateo Huexoyucan: tma economía campesina

tradicional en el altiplano de Tlaxcala". Investigaci6n
realizada para la Universidad Iberoamericana, Departamento de
Antropolgía Social, México, 1978-79. (Ms.)

Broda, Johanna

The Mexican Calendar. Acta Ethnologica et Linguistica.

Viena. (1969)

"Las fiestas aztecas de los dioses de la lluvia", Revista

Española de Antropología Americana (Madrid), 6, 245-74.
(19 71)

"Relaciones políticas ritualizadas: el ritual como expresión

de una ideología", en Carrasco, Pedro y J. Broda, 1978:
págs. 219-54. (1978)

"Estratif icaci6n social y ritual mexica", Indiana (Berlín),

5, 45-82. (1979)

"Astronomy, Cosmovision and Ideology in Pre-Hispanic

Mesoamerica", en Avení, Anthony F. y Gary Urton (ed.),

161
 Ethnoastronomy and Archaeoastronomy in the American Tropics.
Annals of the American Academy of Sciences-- (Nueva York: The
New York Academy of Sciences), 385, págs. 81-110. (1982a)

"La fiesta azteca del Fuego Nuevo y el culto de las Pléyades",

en Tichy, Franz (ed.), Space and Time in the Cosmovision of
Mesoamerica. Lateinamerika-Studie~lO(Munchen: Wilhelm
Fink Verlag), págs. 129-58. (1982b)

"The Templo Mayor as Ritual Space", en Matos, Eduardo,

Johanna Broda y David Carrasco, The Great Temple of Aztec
Mexico: Center and Periphery. University of California Press
(en prensa). (Ms.a)

"Arqueoastronomía y desarrollo de las ciencias en el México

prehispánico", en Moreno, Marco A., Historia de ~
Astronomía en México. México: UNAM: Insti'tuto de
Astronomía-Instituto de Investigaciones Históricas (en prensa).
(Ms.b)

"Ciclos agrícolas y ciclos rituales de la lluvia y del maíz"

(ms. inédito). (Ms.c)

CF, véase Sahagún.

CFN Calendario de Francisco de las Navas

Calendario de fray Francisco de las Navas, de D. Antonio de

Guevara y An6nimo Tlaxcal teca. --Colección Antigua del Museo
Nacional- de Antropología e Historia, 210. (Colección
Ramírez: Opúsculos Históricos, 21). México: Museo Nacional
de Antropología e Historia, págs. 93-203. (Ms.)

Carrasco, Pedro

"Las fiestas de los meses mexicanos", en Dahlgren, Barbro

(ed.), Mesoamerica. Homenaje al doctor Paul Kirchhof f.
México: SEP-INAH, pags. 51-60. (1979)

Carrasco, Pedro y Johanna Broda

Econ/mía política e ideología en el México prehispánico.

México: Nueva Imagen-CISINAH. (!978)

Durán, fray Diego

Historia de las Indias de Nueva España. Angel Ma. Garibay

(ed.). 2 tomos. México: Porrúa. (1967)

Graulich, Michel

"Les origines classiques du calendrier rituel mexicain".

Boletín de Estudios Latinoamericanos y_ del Caribe (Amsterdam),
20, 3-16. (1976)

162
 "La structure du calendrier agricole des anciens mexicains",
en Lateinamerika-Studien, 6 (München: Wilhelm Fink Ver lag),
págs. 99-114. (1980)

Hartung, Horst

"A Scheme of Probable Astronomical Projections in Mesoamerican

Architecture", en Avení, Anthony F. (ed.), 1975: págs. 191-
204. (1975)

"Ancient Maya Architecture and Planning: Possibilities and

Limitations for Astronomical Studies", en Avení, Anthony F.
( ed.), 19 77: págs. 111-29.

Klein, Cecilia

"Who Was Tlaloc?" Journal of Latín American Lore, 6, 2,

155-204. (1980)

Kubler, George y Charles Gibson

The Tovar Calendar. Memoirs of the Connecticut Academy of Arts

and Sciences, 11 (New Haven, Connecticut). (1951)

Malina, fray Alonso de

Vocabulario en lengua castellana y mexicana. México: Porrúa.

(1970)

Motolinía, Fray Toribio de Benavente

Memoriales. García Pimentel, Luis (ed.). México: Ed.

Avina Levy. (Edici6n facsimilar de la de 1903) (1967)

Nowotny, Karl Anton

"Die aztekischen Festkreise". Zeitsc,hrift für Ethnologie

(Braunschweig), 93, 1 y 2, 84-106. (1968a)--

"Zu den aztekischen Festkreisen". Tribus (Linden-Museum),

17, 31-39. (1968b)

Olivera, Mercedes

"Huémitl de Mayo en Citlala: ¿ofrenda para Chicomecoatl o

para la Santa Cruz?", en Dahlgren, Barbro (ed.),
Mesoamérica: Homenaje al doctor Paul Kirchhoff (México: SEP-
INAH, págs. 143-58. (1979)

Pasztory, Esther

"The Iconography of the Teotihuacan Tlaloc". Studies in Pre-

Columbian Art and Archaeology (Dumbarton Oaks: Washington,
o.e.). (1974)-

163
Peña Haaz, Elsa

El trabajo agrícola en un pueblo chinampero: San Luis

Tlaxialtemalco. Tésis profesional presentada en la Escuela
Nacional de Antropología, México. (1978)

Sahagún, fray Bernadino de

CF (Códice Florentino): Florentine Codex: General History of

the Things of New Spain. Anderson, J.O. y Charles E. Dibble
( trad.). Santa Fé, Nuevo México: Unive rs i ty of Utah and the
School of American Research. (1950-69)

HG: Historia General de las cosas de Nueva España. · Garibay,

Angel Ma. (ed.). 4 tomos-.- México: Porrúa. (1956)

Seler, Eduard

GA: Gesammelte Abhandlungen 4 tomos. Berlín. (1902-24)

Sepúlveda, Ma. Teresa

"Petici6n de lluvias en Ostotempa". Boletín INAH (México

Instituto Nacional de Antropología e Historia), Epoca II, 4,
9-20. (1973)

Suárez Jácome, Cruz

"Petici6n de lluvia en Zitlala, Guerrero". Antropología e

Historia: Boletín del INAH (México: Instituto Nacional de
Antropología e Historia) ~ o c a III, 22, 3-13. (1978)

Tichy, Franz

"Ordnun g un d Zuo rdnun g vo n Raum und Ze i t im We 1 t b i 1 d

Altamerikas, Mythos oder Wirklichkeit ?" Ibero-amerikanisches
Archiv (Berlín), N.F., 2, 2, 113-54. (1976a)

"Orientación de las pirámides e iglesias en el altiplano

mexicano". Suplemento Comunicaciones (Puebla: Fundaci6n
Alemana para la Investigación Científica, Proyecto Puebla-
Tlaxc~la), 4. (1976b)

"El calenda río so lar como principio de ordenación de 1 espacio

para poblaciones y lugares sagrados". Comunicaciones (Puebla:
Simposio de la Fundaci6n Alemana para la Investigación
Científica, Proyecto Puebla-Tlaxcala), 15, 153-64. (1978)

"Der Festkalender Sahaguns. Ein echter Sonnenkalender?", en

Wirtschaft und gesellschaftliches Bewusstsein in Mexico seit
der Kolonialzeit, Lateinamerika-Studien (München: Wilhelm
Fink Ver lag), 6, 115-37. (1980)

"Sonnenbeobachtungen und Agrarkalendar in Mesoamerik_a", en

Homenaje a Gerdt Kutscher (Berlín: Gebr. Mann Verlag, en
prensa). (Ms.)

164
Torquemada, fray Juan de

Monarquía Indiana. 6 tomos. México: UNAM: Instituto de

Investigaciones Hist6ricas. (1975-80)
GORDON BROTHERSTON

The Year 3113 BC and the Fifth Sun of Mesoamerica: an orthodox

reading of the Tepexic Annals (Codex Vindobonensis obverse).

The claim announced in this title is that the Goodman-Martinez-

Thompson base date of 3113 BC 1 (3114 BC for historians) extends
beyond the tun calendar and hieroglyphics of the Maya lowlands,
being likewise observed in texts written in the "Mixtec-Aztec" or
iconographic script of Mesoamerica; and that as the year "13 Reed"
engraved on the Aztec Sunstone it begins a 5200-year span equivalent
to the 5200-tun Era of the lowland Maya (Plate 1). Furthering the
argument set out in A Key to the Mesoamerican Reckoning of Time
(Brotherston 1982) and "Astronomical Norms in MesoamericanTime-=
Reckoning" (Brotherston 1982a), the paper concentrates on the
Vindobonensis obverse and related iconographic texts which Caso and
Nowotny have attributed as a group to northwest Oaxaca and the upper
Papaloapan drainage, the area named Anahuac in the Cuauhtitlan
Annals (cf. years 1064 and 1446 AD) and in the Florentine Codex Book
10. A means is found of correlating these texts with those of
Tenochtitlan and Tilantongo, and with certain hieroglyphic
narratives, which reveals their base-date to be 13 Reed 3113 BC.
This remarkable time span is spelt out by successive year dates in
the Vindobonensis while elsewhere it is covered by means of year
muttiples, chiefly the 52-year (Calendar) Round and the 400-year
tzontli or "Head". Our capacity to read year-narratives in general
is much improved with the proper recognition of how such multiples
function iconographically.

Throughout, with its characteristic solar design (originally

the Rain-god's headdress at Monte Alban) and with its Number-plus-
Series Sign (Fig. 2a), the year in question is taken to be the
seasonal period of agricultural tribute, i.e. to be of 18 weeks or
"Fasts" of 20 days plus epagomenal days, not the unvarying or
metric 365-day year of the hieroglyphic inscriptions. This last
point has been argued separately, on several counts, which tie in
with proofs set out elsewhere in this volume by Broda, Tichy and
others (Brotherston 1982: 9-16). 2

The Anahuac tradition

Some of the later studies completed by Alfonso Caso (1954,

1958, 1961, 1964) focused on the toponyms, events and rulers
common to the "lienzos" or Maps of Ihui tlan, Tlalpi tepee (Antonio
de Le6n), Coixtlahuaca and other towns in the Anahuac area.
Showing the parallels between them, he further related them to the
Selden Roll, the Gomez de Orozco fragment, the Baranda Codex and
above all the Vindobonensis screenfold (obverse). While indicating

167
 Fig. 1 Tepexic and surrounding towns

a Map
,\11,
~•',:_
,,,\.....
Gu.Lf of
; \'
~,,_ l tl. '" ,,...
,.,,,,
,,c:: •rec1.cc:L

O"1co \_
XlC oTehu.a..a:tl\

Hcv, \eo · 18oN

l~v.i.tlA,\
0
L"4'i te pe~

>
(oi.)cti.a..hu.~

I
98°w
1<to.11to1190
lllidt.tt1.to119:
0
Oct)(O.C<t.
ls¾ w
0

b Toponyms in native script

(centre) (east) (south) (north) (west)

I rl I
fJ}fI
V ~ I

Tepexic (Apoala) ( Mictlan tongo) Tepeaca Coatzinco Nacochtlan

£1 §; D ~
~ ~

Upper row: Tepexic Annals pp.32, 35, 39, 43, 48; lower row:
Mendoza p.42, Tilantongo Annals p.15, Selden p.6.

168
 Fig. 1 Tepexic and surrounding towns

c Diagram of Aubin ms 20

,..------------ 5 xvI...-------------~
12 R less
5 III 2 heptad 5 XVI II
805 AD _ __. leaves ( 611) 1416 AD

Tepexic Tehuacan

5 IV
5 xx
I
dots : 116 88(-1) Mictlantongo
Eight
Mercury: synodic sidereal Deer
21 21
sidereal 137 109 = 246 - - - - - 5 X I I - - - - - - - - - - - - - - - - '
moons: 5 4 = 9

d Tribute-town ciphers

Tepexic Tilantongo
head- all head-
towns towns towns

centre 6 centre 2

east 7 f
88 west 5
south 5 21 south 6
north 12 . 116 north 11
west 5· 21 east 7

29 246 29

Tenochtitlan
head-
gobernadores all towns
towns

Petlacalcatl 123 centre 9 13 10 26 16 26 7 10 7 9 = 124

Atotonilco 46 west 7 6 7 13 12 6 2 1 = 47
Tlachco 69 south 7 10 14 12 13 8 6 6 = 70
Chalco 82 east 7 6 22 11 11 3 22 8 = 83
Cuauhtochco 45 north 8 7 6 7 11 7 2 5 1 = 46

365 29 246

Tepexic A pp.32,35,39,43,48; Aubin ms 20. Zouche reverse pp.12-27.

Mendoza pp.22-55

169
a relative time-depth for such figures as Four Alligator, Four
Jaguar and Two Dog, 3 which appear in these sources, Caso
nonetheless mostly refrained from translating into the Christian
calendar the Series III year-dates attaching to them, intimating
rather that these texts might have to be dated en bloc, possibly
via a hitherto unnoticed correlation between Mesoamerican and
Christian years.

To this degree uncontentious and accepted, this work by Caso

on the Anahuac texts has completed studies by Nowotny, notably his
commentary on the Vindobonensis (1948) and his Tlacuilolli (1961).
Remarking on the toponym sequence common to the Vindobonensis, the
Aubin ms 20, and the Series II annals of nearby Cuicatlan (Codex
Porfirio Diaz), Nowotny interpreted it in terms of a historical
self-validation on the part of the ruling class and priesthood of
that area. Giving advice unfortunately ignored by such recent
commentators as Jill Furst in her basically synchronic reading of
the Vindobonensis (1978), Nowotny stressed that however ritualized
and to that extent "mythic" the events depicted in some of the
Anahuac texts might appear they should nonetheless be recognized to
have a firm basis in economic and material history, this history
being of considerable length and cosmic in its beginnings. The
historicity of the Vindobonensis has now been further vindicated by
Melgarejo Vivanco (1980), who reveals its connections not just with
the "Mixtec" capital Tilantongo but with the Gulf Coast, via the
Papaloapan. Also, in transcribing the Vindobonensis obverse dates
Melgarejo Vivanco uses the Tenochtitlan correlation rather than the
Tilantongo correlation known to be valid for the genealogy on the
reverse of the same screenfold.

Continuing this approach to the Anahuac group of texts we may

in the first instance further explore their geography. A main aid
to this end are the sets of toponyms which accompany five of the ten
fire-kindlings reported in the Vindobonensis, for they clearly
represent a tribute arrangement of head-towns (cabeceras) in the
centre and the four surrounding "fields" or quarters (pp. 32, 35,
39, 43 and 48; cf. Fig. 3a). The same standard arrangement is
found for Tenochtitlan in the Mendoza and for Tilantongo in the
Zouche (reverse pp. 54-68), where moreover the total of the
cabeceras is conventionally made to equal the lunar cipher 29,
exactly the Vindobonensis total: in "ritual" texts like the
Fejervary map (p. 1) and the Borgia tribute sequence (pp. 49..:.52),
this same total is esoterically achieved by means of four Quecholli
or feeding birds, one to each quarter with its characteristic
tree. 4 The consistency of the Vindobonensis with this convention
and its overlap in detail with the alphabetically-glossed toponyms
and tribute commodities in the Mendoza, and with other sources like
the Tilantongo Annals (Zouche obverse) and the Selden, enable us
surely to identify its central toponym as Tepexic and hence to name
the 52-page text the "Tepexic Annals" (Fig. 1 ). In this reading
not only the cabecera towns in the centre and the four tribute
fields fall into place but so does their respective tribute:
Tepexic with its cane--work (p. 32); Apoala in the east, the source
of cacao (p. 35); Mictlantongo to the south, the shrine of lady
Nine Tooth (p. 39); Tepeaca with its high mountains to the north
(p. 43); and Nacochtlan and Coatzinco to the west (p. 48).

170
 Such a geographical arrangement is further conf inned by, and
itself confirms, that found in other Anahuac texts. In Aubin ms
20 (Culte rendu au soleil), whose effaced central toponym may
fairly be reconstructed as Coixtlahuaca, Tepexic stands
appropriately to the northeast (top left) opposite what is possibly
Tehuacan (top right); below we again find Mictlantongo, to the
southeast (bottom right), matched with a toponym whose designation
"Eight Deer Defeated" must point to Tilantongo, in the southwest
(bot tom left; Fig. le). Succinct in conveying what appears to be
the literal dominance of Tepexic over Tilantongo, the Aubin ms 20
might in fact be more informatively referred to as the "Anahuac
Map". Finally, in the case of the Selden Roll there seems to be
no reason not to read its terminal toponym as "Ihuitlan", given its
resemblance to other versions of that town's glyph; standing on the
line between Tepexic and Coixtlahuaca, this place is the scene of a
fire-kindling orchestrated by the character Two Dog, also featured
at a near-contemporary kindling in the Tepexic Annals (p. 43; Fig.
3a, 6c).

Thus integrating the Tepexic Annals more fully into the Anahuac
group of texts on firmly geographical and economic terms opens the
way to a better understanding of just that material history insisted
upon by Nowotny and Caso, in contradistinction to idealist and
synchronic readings in the style of the structuralists. Indeed the
180 year dates recorded successively, in bous trophedon, over the
52 pages of these annals may for the first time be acknowledged for
what they are: moments in sequential time. They may be accepted as
no different from exactly analogo~dates found in texts of
identical format (cf. Plate 10) like the genealogy on the
Vindobonensis reverse, or the obverse and reverse of the Zouche
(here called the "Tilantongo Annals" and the "Bachelorhood of Eight
Deer") whose historicity has never been doubted.

Year correlations

Still in its infancy, the whole question of correlating

Mesoamerican annals or year-calendar texts has been hampered from
the start by two prejudices. Instead of allowing for even the
notion of self-awareness on the part of native scribes and
historians, within the iconographic script tradition which they
share, too often we have isolated this or that local convention of
style attempting to match it seriatim with the Christian calendar.
Then, instead of respecting the integrity of native sources,
especially those that have survived in iconographic script, we have
deferred to the often partial view of the conquering Spanish. The
case for the several Series III correlations proposed by Kirchhoff
and Davies (1977) would undoubtedly be stronger if it allowed more
for the idea of native calendrical competence.

So far, only three correlations between Mesoamerican and

Christian years can be accepted as firm, all three having been
decoded from native script. One, in Series II, is restricted to
Tlapa and the few texts extant from that area (cf. Toscano 1943).
The other two are both major Series III correlations: the one
involves large areas of the Aztec empire centred on Tenochtitlan;

171
 Fig. 2 Year Series and correlations

a The Twenty Si(JnS arranged in the five year Series I to V

I Alligator VI Death XI Monkey XVI Vulture

II Wind VII Deer XII Tooth XVII Quake
III House VIII Rabbit XIII Reed XVIII Flint
IV Lizard IX Water XIV Jaguar XIX Rain
V Snake X Dog xv Eagle xx Flower

b Progress through the year Series I to V (Borgia pp.39-40)

XI XVI
I VI XI XVI
I VI XII XVII
II VII XII XVII
II VII XIII XVIIJ
III VIII XIII XVIII
III VIII XIV XIX
IV IX XIV XIX
IV IX xv xx
V X

c Sample correlations of Reed years

C:C Sign Nos. AD Source

a 1:10 XIII 1:10 1035 Tlalpitepec Map
b 1:11 XIII 6: 3 1339 Tepexic Annals p.49
c 10: 1 XIII 9:13 1047 Zouche reverse p.35
d 11: 1 XIII 9:12 1423 Rios p.106
e 11: 10 XIII 7: 6 1499 Acuecuexatl Stone
f 11!12 XIII 11:12 1347 Cuauhtitlan Annals f.24
g 12:.11 XIII 3: 2 1455 Huichapan Annals p.29
a e directly equate the two year names, expected and alien;
d f g interject the alien name; b c invoke it (the Chichimec
New Fire year at Cholula, and the Era year, respectively).

d Two correlations involving Series III C1

00000
00000~
M g~g
0
g
0
-~

0
0 0 00
o 000

1 Reed twinned with 10 Reed 1035 AJJ, Tlalpitepec M;

9 Reed ousted by 12 Reed 1423 AJJ, Rios

172
the other Tilantongo, Nahua-speaking Tecamachalco, and "Colhua"
Texcoco. According to the former correlation, 1516 AD was a year
11 Flint, in the Round of 52 years; according to the latter 1516
AD was a year 10 Flint (cf. Caso 1967; Jimenez Moreno 1940). For
convenience they are referred to here as Series III "Cll" and "ClO"
respectively, after the example quoted. At the same time it is
clear that these were not the sole correlations to have existed, if
only because others can be seen to be alluded to if not decoded in
such texts as the Vindobonensis reverse (p. 6), the famous gold
pendant from Monte Alban analysed by Caso, and several stelae at
Xochicalco. In the first case an aberrant year Sign (13 Owl)
intrudes into an otherwise regular sequence of Series III dates,
whose equivalent value in Series III can by that token be
numerically ded u ced (13 House); this might be termed an "ousting"
device, with an unexpected year-date replacing the expected one.
In the second case, on the Pendant (cf. Caso 1960: 35), Series II
and Series III years are equated through the anatomical image of
lungs breathing in time with each other, 10 Wind with 11 House;
this might be termed a "twinning" device. In these matchings
between local varieties of the Mesoamerican year calendar, the
years substantially referred to matter less for the moment than the
f a ct that th ey exi st , as internal correlations, in native script,
and that th e y ex em p lify two distinct yet readily understandab le
principles of correlation, "ousting" and "twinning".

Taking into account the considerable range of year-calendar

texts in native script and their alphabetic transcriptions and
glosses in Nahua (occasionally Otomi), the following appears to be
the case. Several other aberrant year dates, dismissed hitherto
as consequences of the "scribal error" we so freely attribute to
native America, can more appropriately be read as examples of the
ousting and twinning noted above, and hence as further examples of
internal year correlations. These have been discussed at length
elsewhere ( Br otherston 1982: 17-20). Within Series III, they
either confirm the existing Cll and ClO, as in the twinning of 7
Re e d with 6 Re ed (1499 AD) on the Acuecuexatl Stone and the ousting
of 13 Rabbit by 12 Rabbit (1414 AD) in the Xolotl Maps (cf. Dibble
1951: 119); or they suggest two further correlations equivalent to
Cl and Cl2, i.e. 1516 AD as 1 Flint or 12 Flint (Fig. 2c). Most
important, in all cases these correlations reinforce each other
reciprocally, while being confined to just these four possibilities
in Series III: Cl, ClO, Cll and Cl2; and all four are
distinguished by the fact that they retain the same Sign for a given
year (e .g. Fli n t ) while changing the qualifying number (1, 10, 11
and 12), a "rule" independently established by Davies.

Four is t he limit of the great "usages" c, ~ power (ueue

tlamanitilztli, ueue tlatolloyan) acknowledged by Tenochtitlan:
Toltec, Chichimec, Colhua and Tepanec (Lehmann 1949: 105), an
arran gement echoed in Chimalpahin's Memorial breve. Two of these
readily correspond to two of the four calendrical correlations
identified above: Cll as Chichimec, and ClO as Colhua. Good
reasons for identifying the Cl2 convention with the Tepanec, who do
not immed i ately concern us here, can be found in the ousting of
Chichimec or Aztec Cll years by Cl2 years in texts from the Tepanec
are a, like the Huichapan Annals (1455 as 3 Reed rather than 2 Reed)

173
and the Techialoyan Books of Mimiahuapan and Tizayuca (1544 and
1596 as 1 Flint rather than 13 Flint), and in the Cuauhtitlan
Annals, with reference to the Tepanec hero Tezozomoc ( 134 7 as 12
Reed rather than 11 Reed). This leaves, then, only one of the
four correlations in question, Cl, and the possibility of its
being thought "Toltec" and hence older . and more august than the
other three.

Clear examples of internal correlations between Chichimec Cll

and "Toltec" Cl years occur in the Cuauhtinchan History where 3 Reed
ousts 13 Reed (1427 AD), in Rios where 12 Reed ousts 9 Reed (1423
AD; Fig. 2d), and on the Moctezuma Stone which twins 4· Rabbit with
1 Rabbit (1506 AD). In the Cuauhtinchan History the concern to
reconcile Toltec with Chichimec usage is announced in the very
subtitle of the work, "Historia tolteca-chichimeca"; and in Rios
concern with the ancient Toltecs and their tradition is quite
evident in the opening chapter, which is strikingly absent from the
closely-related Telleriano Remensis version of the same Chichimec
history, along with the 1423 AD 12 Reed equals 9 Reed correlation.
The same respect for a venerable "Toltec" antecedent to the
Chichimecs and Aztecs seems implicit in yet other Cll texts insofar
as they spell out the 16-year gap which formally separates their own
from Cl year names in the 52-year Round (e.g. in Cl 1 Flint equals
1516; in Cll it equals 1532). This gap is accounted for in the
pair of Nepopualco Maps, the one invoking the first Round year 1
Flint in Cl as 632 AD and the other invoking it in Cll as 648 AD
(see below and Plate 5); it appears also on the Mendoza title page,
between the 51 named years around the margin (1325-1375 AD) and the
67 year-leaves in the quarters (i.e. 1309-1375 AD) which themselves
incorporate the toponym Tula or Toltec (Plate 3). For their part
the Cuauhtitlan Annals offer to locate this Toltec tradition in
Anahuac since in exalting Coixtlahuaca as the capital of that region
and the key to the whole tribute wealth of the east they identify
its dynasty as both ancient and Toltec; even more straightforwardly
the Cuauhtinchan History announces the area as the Toltec enclave.

If the "Toltec" Cl correlation revealed in the Cuauhtinchan

History and Rios is in fact applied to the Anahuac group of texts
then it can immediately be confirmed as correct, both by reciprocal
correlation devices within the Mesoamerican sys 'tem and through
translation into dates AD. In describing the conquests of Four
Jaguar, the Tlalpitepec Map twins its own year name for 1035 AD, 1
Reed in Cl, with the name for the same year in Tilantongo's Cl0,
i.e. 10 Reed (Fig. 2d); the significance of this can be deduced
from the Tilantongo texts, the "Bachelorhood of Eight Deer" for
instance, which shows the intense military and political contact
between this hero and Anahuac's Four Jaguar at just this period
(Zouche pp. 52-79, 1045-47 AD), pace Smith (1973: 74, 124), who
fancies Four Jaguar never existed. ~ r its part the Baranda Codex,
associated by Caso with the Anahuac group, also corroborates the
validity of Cl. For if read in that correlation it ends devoutly
with the 16th Christian century, to the year (6 Flint or 1600);
and in the year 1 House or 1529 it shows a church and Cortes coming
to terms with a native leader whom an 18th-century native
interpreter firmly identified as Cosiobi of Tehuantepec, of the
Zapotec line driven from Zaachila by the Mixtecs (Caso 1958). It

174
was in that very year that the Dominicans began work in Oaxaca and
that Cortes became Marquis of the Valley, taking the royal line
from Cosiobi on the death of Cosiobi's father Cosihuesa.

Applying the "Toltec" Cl to the Anahuac texts makes possible an

extended chronological comparison between them, an essential
counterpart to the complex genealogical sequences that they contain;
and between them and the other two major Series III traditions, of
Tilantongo and Tenochtitlan. This work would continue that of Caso
and Nowotny, enhancing our overall view of the Mesoamerican year
calendar. Above all it enables us further to link this last with
the Maya tun calendar , with which it otherwise has so much in
common, in nomenclature, structure and sheer mechanics. For it
is generally agreed that in the tun calendar the Era lasts 5200 tuns
and has its start date in the year 3113 BC. Though little has been
made of the fact, in the year calendar the Era is clearly stated to
be 5200 years in length (Rios p. 8; cf. also the 2088 AD or 5200-
year date in the Tizimin, Edmonson 1982: 197), 5 while the Series
III year with which it begins is just as clearly stated to be 13
Reed (Cuauhtitlan Annals; Aztec Sunstone, Plate 1). In the
Toltec Cl, working back through the centuries, 3113 BC is in fact
a year 13 Reed (which of course it therefore could never be in other
Series III correlations like ClO or Cll); at the same time Tula,
from which the term Toltec derives, stands as the inaugural city of
Mesoamerica everywhere associated with the beginnings of the of the
5200-year Era span (cL Brotherston 1983). Indeed "Toltec" or Cl
texts like the Tepexic Annals can be shown to narrate an Era history
exactly commensurate with that found in the tun texts of the lowland
Maya.

Time depth

If the Tepexic Annals are read continuously from end to end,

strictly according to the norms that pertain to the boustrophedon
narratives of Mesoamerica, then there is no way that they can be
seen not to cover several thousand years (Fig. 3b). Opening with a
well-known formula which integrates the tonalpoualli with the year
(7 x 260 = 5 x 365 less 5 days) and which here issues into the first
year date, 5 Flint (p. 3), the first chapter alone covers 3451
years, culminating the fire-kindling in the year 6 Rabbit (p. 21).
One single irregularity in the boustrophedon reading occurs on p. 6
where the reading stream divides, the upper part covering two
Rounds less than the lower; these however are supplied by the upper
tributary stream from the previous page. Similarly, the narrative
continues quite regularly through the remaining nine chapters and
f ire-kindlings, over a further thousand years. Then, after the
tenth and final kindling, the reading stream is fully blocked or
staunched by a unique 2 House date inset into what will later be
shown to be a commemorative marker of year multiples (14 Rounds; p.
SO). This suggests that the remaining dates in the text project
into the future, in the fashion of certain tun narratives.
Arithmetically, this additional span, of 331 years, has the
effect of completing the calendrically significant overall total of
4800 years, from the first year date 5 Flint to the last, 8 Fli.nt.

175
 Fig. 3 Analysis of the Tepexic Annals

a Overview of the ten chapters (R = Round)

Page:

1-22 chapter 1
1 womb of heaven with four pairs each representative of a Sun
in descending order of time
2-3 ancestral and troglodyte pairs bearing day dates from 1 Deer
30 Mar~h 3113 BC to the first named year 5 Flint 3108 BC
4-6 'Beget am bear' birth sequence of Nine Wirrl, his descent
between sun and moon, arrl his raising of the sky in 10 HousP RS,
2843 BC
7-15 toponym history of the early period 5 House RS to 13 Rabbit R60
(2679 BC to 46 AD)
15-18 Tree-birth presided over by Nine Wirrl; appearance of bald- and
egg-head pair, Ten Death, Eight Wind etc.
19-22 1st or 'Classic' kiooling, by Nine Wirrl 6 Rabbit R66, 338 AD

23-32 chapter 2
23-26 preview of kioolers
27-29 luxuries of maize, pulque and mushrooms R68-70 (cf. topics
2-4 in Palenque Trilogy panel 1, 8.19.0 to 9.0.0.)
29-30 'hiatus' and restoration, R71-R73, 6th-7th centuries AD
31 start of 'dropped leap day' sequence R74 to R79
31-32 2nd kiooling, at Tepexic with its cane-work tribute: by Two
Dog Jaguar Claw, in 5 House R75, 805 AD

33-39 chapters 3-6

33-34 3rd kindling, 13 Rabbit R75, 826 AD
35-36 4th kindling, 7 Flint R76, 872 AD, by Xolotl. Eastern quarter,
seven toponyms inc. Apoala; cocoa tribute
36-38 5th kimling, 5 House R77, 909 AD
38-39 6th kindling, 1 Reed R77, 931 AD. Southern quarter, five
toponyms inc. Mi~tlantongo, with lady Nine Grass

40-52 chapters 7-10 (kindlings measured as Rounds)

40-41 7th klooling, 7 Reed R79, 1015 AD, with Four Alligator
41-42 8th kindling, 5 Flint R81, 1104 AD
43-47 9th kirrlling, 7 Reed R82, 1171 AD, with Two Dog. Northern
quarter, twelve toponyms inc. ?Tepeaca
48-52 10th kindling (drill unused), 5 House R84, 1273 AD. Western
quarter, five toponyms inc. Nacochtlan, Coatzinco, ?Tepexochocan
49 Eagle and Jaguar warriors at Venus temple 6 Reed-6 Reed,
1287-1339 AD
50 boustrophedon reading flow staunched at 2 House 1361 AD, the
likely date of composition, in R86, 14 Rounds (mouths) from
R72 arrl before R100
5? final projected date 8 Flint R92, 1692 AD, exactly 4800 years
or 12 Heads from 3108 ~ (p.3).

176
 Fig. 3 Analysis of the Tepexic Annals

b Transcription of the year-date sequence

(2) 1 VII 1 VII 1 II 7 II 12 II 13 II 4 V 7 V 4 I 11 II

(30 March ~113 BC)
(3) 7 XIX 7 XV 5 XVIII 0.5 0.42 1.39 (4) 1.40 2.18 3.10 4.10
BC 3108 2971 2922 2921 2891 2847 2895
(5) 6.19
{ 5.21 7.19 }
2832 2782 2730
(6) 4.19 5.10 5.49 6.5 6.20 7.5 7.44
2886 2843 2804 2796 2781 2744 2705
(7) 8.18 8.20 8.35 8.40 9.18 9.35 9.37 9.43 10.17 10.20
2679 2677 2662 2657 2627 2610 2608 2602 2576 2573
(8) 10.39 10.40 11.39 12.34 12.39 13.16 14.16 14.,40 15.35 16.27 16.42 17.20 18.19
2554 2503 2502 2455 2450 2421 2369 2345 2298 2254 2239 2209 2158
(9) 18.25 19.10 19.41 20 .. 16 20.32 20.44 21.36 22.20 22.25 23.20 23.40 24.40
2152 2115 2084 2057 2031 2029 1985 1949 1944 1897 1877 1825
(10) 25.40 25.42 26 .. 40 27.4 27.20 27.28 28.20 29.18 29.39 30.5 30.18 30.40 31.27
1773 1771 1721 1705 1689 1681 1637 1587 1566 1548 1535 1513 1-474
(11) 32.11 32.20 32.27 33.4 33.27 33.33 34.27 35.10 35.20 36.10 36.47 37.47
1438 1429 1422 1393 1370 1364 1318 1283 1273 1231 1194 1142
(12) 38.8 38.19 39.6 39.28 40.25 40.33 40.40 41.11 41.43 42.6 42.8 42.40
1129 1118 1079 1057 1008 1000 993 970 938 923 921 889
(13) 43.20 44.20 44.39 45.8 45.34 46.8 46.20 46.43 47.34 47.40 48.8 48.17 48.28
857 805 786 765 739 713 701 678 635 629 609 600 589
( 14) .49.20 49.45 49.46 50.21 51.8 51.20 51.40 52.40 53.20 54.5 55.5 55.40 56.20
545 520 519 492 453 441 421 369 337 300 248 213 181
( 15) 57.18 57.28 57.35 58.20 58.21 59.21 59.39 60.39 ( 18) 61.39 62.21 62.35
131 121 114 77 76 24 6 AD 46 98 132 146
(20) 63.10 63.39 64.20 64.39 ( 21) 65.39 66.19 (23) 66.35 66.39 66.44
173 202 235 254 306 338 354 358 363
(26) 67 .42 68. 35 68.39 68.44 (28) 69.35 ( 29) 70.8 71.,5 72.5 (30) 73.5 73.39
413 458 462 467 510 535 584 636 688 722
(31) 74.39 (32) 75.18 (33) 75.34 (34) 75.39 (35) 76.16 76.33 (36) 76.49
774 805 821 826 855 872 888
( 3 7) 77.18 (38) 77.35 (39) 77.40 (40) 78.20 (41) 79.20 80.5 (42) 81.5
909 926 931 963 1015 1052 1104
(43) 81.20 82.20 82.40 (48) 83.18 84.18 (49) 84.32 85.32 85.36 85.40
1119 1171 1191 1221 1273 1287 1339 1343 1347
(SO) 86.2 86.40 87.10 87.40 (51) 88.33 89.18 89.27 90.4
1361 1399 1421 1451 1496 1534 1543 1572
(52) 90.27 90.47 91.39 92.21
1595 1615 1659 1692

( 2) etc. = page number

1 VII etc. = opening day sequence (: 7 X 260 = 5 X 365 less 5 days)
0.5 etc. = Round in the Era and year in the Round with 1 Flint as the
first ye ar of the Round

177
 Fig. 3 Analysis of the Tepexic Annals

c Ritually significant data

Total of pages:
52, years of the Round

Total of year-dates used:

180, days between the equinoxes (9x20)

Total of toponyms in the quarters:

29, nights in the moon

Maxima of year-dates per page:

12 or 13, moons in the year

Overall year span:

4800, 12/13 of the Era

Maxima of year-dates used within the Round, by Series Sign:

7, orifices of the head, Signs House and Rabbit
9, orifices of the body, Signs Reed and Flint
32, teeth, all Signs, i.e. 2(7+9)
20, digits, unused, i.e. 52 - 2(7+9); cf. the Cospi formula
(pp.1-8): 5 X 52 = 2(72 + 92).

Most frequently used dates:

7 Reed (0.20), the 20- year Flag within the Round; designated
by mouth markers over 5-Round intervals at Rounds 17, 22, 27
and 32 (13x20=260=5x52).
1 Reed (0.40), the 40-year Cross Eagle or active life within
the Round, as with the heroes One Reed and Eight Deer;
designated by mouth markers over a 10-Round interval at
Rounds 14 and 24 (13x40=520=10x52)

Decimal features:
10 chapters
10 kindlings
10 characters named Two Dog (Sign X)

Dominant Round interval, of nine Rounds, marked by a) the year-

and-day date 2 Reed 2 Reed and b) Hero 9 (Quetzal-snake Nine Wind):
a) at Round 27 or 9x3 Rounds (27.28, the sidereal-year span)
at Rounds 39, 48 and 57
b) at Rounds 66, 75 and 84 (kindlings 1, 2 and 10).

178
 This considerable time depth has never been fully acknowledged
in the Tepexic Annals. For his part Nowotny contrived to deny it
by suggesting that though historical in origin the year dates given
had been absorbed into "timeless" ritual patterns; in addition, he
dispensed altogether with pp. 6-15 of chapter 1. These pages,
which alone cover over 52 Rounds (52 x 52 years), he demotes to
being a subsection (Unterabschnitt), a recapitulation of the
narrative before and afterwards ("Rekapitulation der ganzen
Bilderfolge ••• vor und nach"; 1961: 47, 256). On formal and all
other grounds there is absolutely no justification for treating the
narrative in this way: the pages in question are wholly continuous
with the general reading stream and not separated off at all, so
that the very use of the term (Unter)abschnitt or "cut" is
technically incorrect. 6 Formally speaking, there is no orthodox
alternative to the thousands of years sketched out above. It might
still be objected that Nowotny could be right in refusing to
recognize such a vast time-span in the Tepexic Annals, since it is
apparently anomalous among year narratives extant. As a group the
Cll texts never start before the mid 7th century AD while the ClO
texts, like Ixtlilxochitl's Segunda relaci6n or the Tilantongo
Annals, go back only a few centuries before that. Yet this
situation is more apparent than real. Far from being anomalous,
the Tepexic Annals upon examination provide the key to the Era
chronology of the whole year calendar, on the cosmo-historical scale
clearly detailed by early commentators like Garcia (1729). This
text proves to be the yardstick for the Anahuac Cl tradition
generally, other texts in it likewise taking the same early
starting date, though counting forward from it by means of year
multiples (see below and Figs. 5, 6) rather than by the continuous
year dates found in the Annals. And it is the chief guide to the
story examined elsewhere (Brotherston 1983) of the great Tula-
Nonohualco on the Gulf Coast with its cacao and cotton and rubber
and an economic history testified to in a wide variety of sources
from west and east Mesoamerica as being long prior to Teotihuacan
and to the Chichimec invasions.

To begin with and purely in terms of internal consistency, if

the Tepexic narrative is indeed read from the Era date 13 Reed 3113
BC then it yields an excellent self-dating through the ending year 2
House (p. 50). For counted out regularly from the first day-dates
in 3113 BC and from the first year date 5 Flint 3108 BC, 86 Rounds
are spelt out in the text before this 2 house date is reached,
which means it must equal 1361 AD: that is wholly consistent with
the dating attributed to the text on other grounds, including the
mid-fourteenth century ending of the ClO genealogy on its reverse,
already deciphered by Caso. Similarly, counted out from the 3113
BC base the narrative appropriately pinpoints two major figures in
Anahuac history, 4 Alligator and 2 Dog, presidents of the seventh
and ninth kindlings in 1015 and 1171 AD respectively (pp. 41, 43):
these dates, corroborated in other Cl texts discussed below,
concord well with the relative time-depth attributed to these
figures by Caso (cf. Fig. 3a). Beyond this, when read off from
the same Era base, whole sets of date-events in the text may be
shown to coincide chronologically with their counterparts in other
Cl texts, in other year-calendar texts, and even in tun or
hieroglyphic texts, which are known to be commensurate with the Era.

179
 Out of the vast and largely unedited corpus of hieroglyphic
inscriptions, texts from Palenque, the Classic Maya city closest
to Anahuac, have already suggested certain parallels with the year-
calendar tradition. In particular, attention has been drawn to
the god-like Nine Wind as a presiding patron in both the Tepexic
Annals and the Trilogy of panels at Palenque usually named "Cross",
"Sun" and "Foliated Cross" after their central designs. In fact
the so-called Cross of the first panel is really an anthropomorphic
tree; and beside it to the right a tiny baby-like creature is being
regally held (the idea of birth is further suggested by the
"Euripos" arrangement of the seven planetary symbols at the base of
the tree).7 This detail is doubly important. It helps to
integrate the first panel into the logic of the three and the story
of Classic Palenque's origins as it pertains to her ruling elite,
her warriors and her maize farmers. And it opens the way to the
first dated comparison with the Tepexic Annals version of the same
early history (Fig. 4). For on this first panel, standing between
the hieroglyphic columns to left and right, the tree scene as a
whole is placed by them at about the time of Christ. Counted out
year by year f ram 3113 BC, this is exactly the time of the famous
tree-birth in the Tepexic Annals (p. 16), Rounds 59 to 61 of the
Era, where supervised by Nine Wind the new-born ruling class
actually emerges from an anthropomorphic tree. Similarly, Nine
Wind's first kindling in the Tepexic Annals, the climax of the
first chapter in the early 4th century AD (6 Rabbit Round 66, 338
AD, p. 21), corresponds to the spread of the hieroglyphic stelae
cult in the lowland cities and the start of the Classic. Again,
after the much discussed ''hiatus" in the hieroglyphic inscriptions
in the late 6th century, which is matched chronologically in the
Annals by a revolutionary wheel (5 Flint Round 71, 584 AD, p. 29),
the Palenque panels celebrate the restoration of order in two
successive . "split sun" emblems (640 to 690 AD); 8 this same pair of
emblems recurs at similar dates in the Annals (636 to 688 AD),
being entirely absent from either narrative at any other point (Fig.
4b). Then just as the first kindling pointed to the rise of the
Classic and hieroglyphic stelae cult, so the second, at the end of
the second chapter (5 House Round 7 5, 805 AD, p. 32), points to
their decline and to the rise of towns like Tepexic, the scene of
this kindling, under the auspices of Two Dog Jaguar Claw.

Touching on major events verifiable in turn by archeology,

these parallels over almost a millennium strongly bear out the
orthodox reading of the Tepexic Annals proposed above, and the
notion that its span is indeed commensurate with that of the tun
calendar: they certainly could not be · explained statistically as
coincidence. And they could be further extended, to other
hieroglyphic texts, as well as within the Trilogy. For example,
the tilth island of Round 14 in the Tepexic Annals recalls similarly
dated agricultural glyphs in the Maize panel. Moreover, while once
thought "mythical", these remote events are now being shown
archeologically to be very · much part of the material history of
Mesoamerica (cf. Hammond 1982).

Within the year calendar a similar and yet more detailed

sequence of parallels may be drawn, in the first instance with the
42-page Tilantongo Annals (Zouche obverse), to which the Tepexic

180
Anna.LG are for rn
~l y close: ,c among native histor ies extant. So far
translations oi u1 . c: T•10 r k have mainly focused on the dense genealogy
beginning in the yea r 1 J House 789 AD, pe 23, a "new start" which
in date and format correspon d.s t o the end of the second chapter &..r; -'!
kindling in the Tepexic Annals (805 AD). Read once again strictly
according to bous t rophedon norms, the ea~lier pages likewise
closely echo those of Tepexic, at least from the time of the tre c·-
bi rth 15 Rounds previously (Round 60 of the Era). The very first
pages of the Tilantongo Annals are admittedly hard to date by
orthodox means since the boustrophedon is lacking on the first two
(where sequence depends rather on the logic of three successivly
more complex methods of raising tribute), while on the third
(which records the defeat of the "Stone men") some of the year dates
uncharacteristically are devoid of dependent days in the year.
After this, however, the reading proceeds normally and the very
chapter division reflects that of the Tepexic Annals and hence more
broadly the idea of the Classic: the closing date of the first
chapter which announces the Classic, 2 Flint 312 AD, in this sense
recalls its Tepexic equivalent, 6 Rabbit 338 AD.

Indeed, when they are laid out length for length, the
parallels between the Tepexic and Tilantongo histories do no less
than leap to · the eye (Fig. 4). The first parallel arises from the
tree-birth itself at around the time of Christ or Round 60, since
at the corresponding moment in the Tilantongo Annals (12 Flint - 6
Rabbit - 13 Rabbit, 16 BC - 30 AD - 50 AD, p. 4), which
constitutes the start of the narrative proper in that source, the
first born also include an unmistakable and distinctive bald- and
egg-head pair (Fig. 4c), along with companions whose names and
characteristics are in many cases identical with those given by
Tepexic: Ten Death, coloured grey and black; Seven Flower;
Eleven Alligator; the lady Eight Monkey, blue in the face; and
last in the list Eight Wind with his eagle headdress, to whom
Tilantongo devotes a brief biography (5 Flint to 8 Flint, 68 to 112
AD , pp. 5 -8). Sever a 1 of these names appear at the s tart of the
Bodley and other genealogies in the Tilantongo tradition, along
with the tree-birth, leaving little doubt that, as Nowotny noted
(1961: 257), they belong to the same event as that shown in the
Tepexic Annals and that this event indeed marked the beginnings of a
powerful and far-flung ruling class. What had been uncertain was
the date of the occasion: now, totally orthodox readings of the
year-date sequences in the Tepexic and Tilantongo Annals corroborate
each other in placing it at the time of Christ , Round 60 (to obtain
any other date would mean adopting some sort of unorthodox reading).

In particular, the Tilantongo Annals version of the event

helps us to appreciate its complex international nature, something
alluded to in such Spanish sorces as Burgoa, and hence the
difference in principle between the angles of Tilantongo and Tepexic
on what is a whole series of major happenings prior to and
throughout the Classic. For celebrated by Tepexic and Palenque
alike as the glorious triumph of Nine Wind's cause, the tree-birth
is exposed by Tilantongo as the climax of outright and bloody
oppression of the possibly Otomanguan Stone-men, whose "fall" is
only briefly alluded to in the Tepexic Annals (p. 15). This
different view is in keeping with Tilarrtongo's announcement of the

181
 Fig 4 Parallels in accounts of the Classic period
a Overview

&c. AD
..3000 2,f-00 () 50:J
1,. _ _ __,l____ _ _.L _ _ _J_ _ _ ___ _ . ..J. _ _ _ J _ _ J.__ _

Paler:qt.e * 0 t [ [ioo
Trila:gy 3,C9 I (~ntre) I . P1 1. , Q9,05

Te~c. 0 t [ [ loo )
Annals * 8 16 11 29 30 32
Ti~ntOf)Cp t [ [ ]oo ]
Annal.s 4 14 20 21 22

* smrt of tre Era , 311.3 BC

0 maize tooic
t t~ birth
[ start of the Classic
[] hiatus
o split-sun ' emblem
l Classic decline , rise cf western dynast-ies

b The split-sun emblem

Palenque Trilogy Borgia p.26 Tepexic Annals pp.29-30 Tilantongo

Annals p.21

c The bald- and egg-head pair

Tepexic Annals p.16 Tilantongo Annals p.4

182
Classic, three centuries later, not as Nine Wind's achievement
alone but the result of a triple alliance of which the same Stone-
men constitute one party (2 Reed - 2 Flint, 299 - 312 AD, pp. 11-
13); and it concords too with the attention paid prior to the
hiatus not to Nine Wind in his role as king-maker but to a line of
local matriarchs named Three Flint (7 Reed to 5 Reed, 343 - 575 AD,
pp. 14-20; cf. Plate 8). And when the hiatus arrives in Round 71
of the Era, late 6th century AD, Tilantongo reveals it plainly to
be the result of a massive uprising of the Stone-men, inspired by
one of the ladies Three Flint, against Nine Wind's egg-head elite
(3 Reed, 599 AD, p. 20): this much expands the mere hint of
revolution and unrest given in the Tepexic Annals at the same
chronological juncture (5 Flint, 584 AD, p. 29). Moreover with
the restoration in Round 72, marked for Tilantongo as for Tepexic
and for Palenque by the pair of distinctive "split-sun" emblems
which likewise appears only at this point in the entire narrative
(12 Flint - 1 Reed, 660 - 675 AD, p. 21; Fig. 4b), we see Nine
Wind strutting mercilessly among the corpses of the again defeated
Stone-men, who from this point on disappear as a military presence.
Overall, there can be little doubt about this continuity of echoes
not just between the Tepexic Annals and the Palenque Trilogy, but
between both and the Tilantongo Annals, a fact which further
justifies extending the tun base date of 3113 BC to the year
calendar, in particular the "Toltec" Cl convention of Anahuac.

Turning now to the Cll or Chichimec convention, we see that it

further corroborates the hypothesis. Here, the most important
single parallel stems from the fact that, just as the Round of the
tree-birth (60) operates as a kind of secondary base for Tilantongo,
so for "Chichimec" towns like Tenochtitlan the restoration of Round
72 (mid 7th century AD) assumes the same function. This is the
Round invoked as a base by the Nepopualco Maps, where it is
referred to as "yncalco chichimeca" ( the Chichimec base; Plate 5);
by Chimalpahin's Memorial breve with its start in the year 10 Rabbit
670 AD; by Ixtlilxochtli's Relaci6n sucinta; by the Zacatlan
History collected by Torquemada; and by the Cuauhtitlan Annals,
which make their start in the year 1 Reed 687 AD and count out every
single year between that date and Cortes's arrival in 1·Reed 1519
AD, exactly 16 Rounds (832 years) later. In his day Humboldt
(1810: 25) noted how consistently the mid-seventh century was
observed as a base date in native texts and referred specifically to
the year 648 AD, the first of Round 72 (1 Flint Cll). Elsewhere
this Round 7 2 base is invoked as such, in the Era, in counts
involving year multiples, discussed below (also cf. Fig. 7). The
most esoteric of these year-multiple _ counts is embedded in the
awesome figure who inaugurates the Cuauhitlan Annals and other
Chichimec texts: Obsidian Butterfly. For her name may be
numerically transcribed as Round 72, from 2 (iztli or obsidian,
second of the Nine Figures) plus 70 (papalotl or Butterfly, seventh
of the 13 Quecholli, who may carry their number value in lots of ten
Rounds). 9 Also we may note a particular parallel between the
Chichimec Cuauhtinchan History and the Tepexic Annals in the matter
of the ceremonial defeat of Quetzal snake at Cholula over the Ro\llld
1287 to 1339 AD (dated 3 Reed Cll in the former and, apparently, 6
Reed to 6 Reed Cl in the latter, p. 49). A quite special
confirmation, however, of the whole Era scheme is provided within

183
the Cll or Chichimec convention by the Mexicanus of Tenochtitlan,
page 9 of which is worth exacining in detail, since it openly
announces itself as a chronological key (Plate 4).

The dual wheel design on this page juxtaposes Christian with

Mesoamerican years. Of the former there are 28 or a solar cycle,
identified by the 4 x 7 Sunday or Dominical letters (a - g); of the
latter there are 52 or a Round, identified by the Thirteen Numbers
and Four Signs of a Series III Round. In the position shown, the
"b" year 1558 of the Christian wheel, seventeen years before the
"b" year 1575 (glossed as such in arabic numerals, Prem 1978: 275),
is just touching the first of the Rabbit years on the Mesoamerican
wheel, under the rabbit's nose: in the Tenochtitlan Cll convention
1558 was in fact a year 1 Rabbit, the last of the last Aztec fir -
kindling Rounds completed before the Mexicanus was written. 1O
Hence, the wheels not only are juxtaposed but mesh. In the centre
of the Christian wheel St Peter holds a triple-pronged key in his
right hand and in his left a book on whose pages five units (dots)
have been revealed; above his head a cross occupies an extra
division of the solar cycle wheel, raising from 28 to 29 the total
of the divisions that mesh with the 52 Mesoamerican years, should
the wheels be turned further in either direction. In the position
given, to alter or break the Christian cycle in this way seems to
point to the major calendrical event that occurred shortly after the
last letter recorded, "e" or 1578, i.e. the Gregorian Reform of
1582 otherwise alluded to in the Mexicanus, which did effectively
cause a break in the 28-year Sunday-letter sequence. At the same
time, within the Mesoamerican system this alteration ingeniously
points to the reasons behind the Reform in question, that is the
imperfect Julian measurement of the seasonal year and hence of the
need for leap-year days, insofar as this last was calculated in
Mesoamerica according to a formula involving the number 29 (rather
than 28). This formula, also found in the tun calendar,
specifies that over 29 Rounds or 1508 years the difference between
the metric year of 365 days (i.e. without leap-days) and the
slightly longer seasonal year itself amounts to a year (Brotherston
1982a: 122): and this is exactly the total of years produced by the
operation of the two wheels in question, since 1508 is the lowest
common multiple of 29 and 52. Complex enough in itself, this
still leaves unexplained the remaining numerical data in the design,
viz the three prongs of St Peter's key and the five dots on his
book.

In Christian calendrics the 28~year solar cycle has been

commonly used to produce larger spans, in conjunction with the
Metonic and Indiction periods for example, which gives the Julian
Cycle beginning in 4713 BC, or alone, to produce Usher's creation
date of 4004 BC (North 1977). Here in the Mexicanus, taking into
account St Peter's numerical clues, the emended 29-year cycle can
be seen to produce the distance between the start of the first Round
of the Mesoamerican Era in 1 Flint (Cl) 3112 BC and the end of the
last Aztec Round in 1 Rabbit (Cll) 1558 AD, thus linking and
confirming the two separate decodings, of date and cycle, made so
far. For fully turned thrice through all combinations, as the
lock imagery of the key suggests that they should be, the wheels

184
produce 4524 or 3 x 1508 years and the date 1412 AD, to which may
then be added five turns of the Christian wheel or 145 years, the
interval to the goal of 1558 AD. On the Aztec side this subsidiary
total is equivalent to the interval (exclusive) from 1 Flint Cl 1412
AD to 1 Flint Cl 1 1428 AD, plus two Rounds or 104 years to 1 Flint
1532 AD, plus the half-Round or 26 years to 1 Rab bit 1558 AD. If
novel, this reading of Mexicanus p. 9 violates none of the rules of
Mesoamerican calendrics and accounts satisfactorily for all the
numerical data supplied in the text.

Along with the evidence from Palenque and Tilantongo, this

apparent Era chronology in a text from Tenochtitlan strengthens the
argument that the 13 Reed on the Aztec Sunstone is indeed 3113 BC
and the start date of the Tepexic Annals. As a result, the Toltec
Cl convention observed in those Annals offers a unique means of
interlocking the firmest of the calendrical correlations so far
made, independently of each other, in Mesoamerican studies.
These are the 3113 BC base established for the tun calendar by
Thompson; the Tilantongo Series III ClO proposed by Caso; and the
Tenochtitlan Series III Cll known since Cortes. This evidence
becomes stronger again once the function of year multiples in the
Mesoamerican calendar is examined and taken into account.

Year multiples and time-structuring

Counting out time by means of year multiples may be deemed

fundamental to the very notion of the Mesoamerican year calendar,
such multiples being of three main kinds. The first derives from
tribute arithmetic and includes the 400-year tzontli (Head) or
ihuitl (Feather) along with such factors of the Head as the 20-year
pantli (Flag); in the tun calendar these have obvious analogies
with the 400-tun baktun and the 20-tun katun. The second derives
from the cycles formally produced by the shamanist and ritual
mechanisms used to name time, saliently the tonalamatl of 9
Figures, 13 Numbers and 20 Signs, the source of the 52-year Round
(13 Numbers x 4 Signs); again an analogy may be found in the tun
calendar of the Katun Count (13 Numbers with the katun-ending Sign
XX or Ahau). The third, less explored, derives from the sheer
measurement of time cycles perceived in earth and sky,
independently of tribute or ritual arithmetic, and it includes the
29-Round leap-day cycle noted above, the Metonic "Rain" (Sign XIX)
of 19 years, and the 11-year Monkey (Sign XI), also lunar-solar,
which equals in years what the epact is in days. Important in
principle (Brotherston 1982a), this third kind of multiple does not
materially impinge on the argument at this stage, which returns us
to the first two (Fig. 5).

While the existence and use of the two main kinds of year
multiple, typified by the Head and the Round, stand quite beyond
doubt, what is less clear is how many and various such multiples
are and, above all, how they are represented in iconographic script
(Fig. 5a, b). 11 This unsureness has much hampered the decoding of
year-calendar texts. It may be ascribed in part to the nature of
iconographic script, which unlike hieroglyphic makes no necessary
separation between arithmetic, script and picture. Yet more

185
 Fig. 5 Sources and markers for year multiples

a Tribute arithmeti c

~
4 20 80 400
Stone flag Jade Feather, Head

b The 52-year Round

------C)
Star-eye Lips, mouth Breath-scro ll, Tie
flame

c The Nine Figures or Ennead d The sevenfold pattern of the

Ennead as year guardians begin-
1 Fire or Year Lord ning with the 1st and 1 Flint;
2 Obsidian Borbonicu s pp.21-2
3 Embryo Sun
4 Maize God 1 7 5 2 9 6 3
5 Hell Lord 1 7 5 2 8 6 3
6 Jade Skirt 1 7 4 2 8 6 3
7 Cloth Goddess 9 7 4 2 8 5 3
8 Hill Heart 9 7 4 1 8 5 3
9 Rain 9 6 4 1 8 5 3
9 6 4 1 8 5 2
9 6 4

e The thirteen Quecholli or Fliers

1 Hummingbird 8 · Eagle ( sb:"i ped)

2 " (blue) 9 Turkey
3 Hawk 10 Eared Owl
4 Quail 11 Macaw
5 Eagle 12 Quetzal
6 Owl 13 Parrot
7 Butterfly

186
 Fig. 6 _E xamples of year multiples

a Jade, 3 Flint to 5 Flint, b Jade-Round (80 x 52 years),

1144-1224 AD; Boturini p.13 3113 BC to 1047 AD; Zouche
reverse p.35

n a.J,J, o C£.:rz.

c Ten Feathers less two stones

(10 x 400 less 2 x 4 years),
3113 BC to 1 Reed 879 AD;
Ihuitlan Map

d Cross Eagle, 13 Reed to 1 Reed

(to 2 Reed), 1219-1259 (-1299);
Cuauhtincha n History

e Three double Round f Lip Round; g Seven Caves;

scrolls, 1195-1507; Cuauhtincha n Laud p.21
Rios History

h Three star-eye Rounds, 1063- i Six star-eye Rounds equated with

ca. 1220; Vienna reverse p.9 66 in bars and dots; Selden p.1

45 ~ 21 ~

187
telling has been the lack of the Era framework, with its secondary
bases, within which the whole year calendar operates. Also there
has been a reluctance to admit much native sophistication outside
hieroglyphic in the handling of numbers, when in fact they are
repeatedly distinguished, in shamanist fashion, as odd or even,
prime or divisible, squared or to the power, as ratios or "sigma"
totals (e.g. sigma 5 = 1+2+3+4+5 or 15), and as ciphers by which to
shift between dimensions of time (days, years, Rour1s) or according
to number-base (decimal as well as vigesimal); and all this
within and beyond the Era which itself is the lowest common multiple
of the Head and the Round (52xl00 = 13x400 = 5200).

Often used after Cortes to register lots of 400 years in the

Christian calendar, in the native tradition the Head or Feather
appears preferentially in the Maps of the Chichimec or Cll
convention, like those of Quinatzin, Xolotl, Metlatoyuca and
Nepopualco (Plate 5), where two together (800 years) generally
point back to the seventh century and the Round 72 base; it is also
used in the Tlateloco Annals (Figs. Sa, 7). In the Toltec or Cl
convention it is implicit in the 4800-year or 12-Head span of the
Tepexic Annals, while in establishing the Toltec beginnings of the
Era, Rios uses the Head to give not just its full length, 13 Heads
or 5200 years, but that of previous ages as well. This last
example is further notable for bringing out both the importance of
ratio in these larger calculations and the flexibility of
iconographic script itself, since in Rios the first instalments of
these 400-year periods are shown as actual f ea the red water drops,
forming and falling from the "suckling tree" into the mouths of men
still innocent of agriculture - the theme of the chapter as a whole.
The overall total thus gained is notionally just under 26000 years,
in the following sequence:

400((20, 10 + 10: 12 + 13) ~ 2000(8:5)

In other words, if we recall the thirteen Numbers or Quecholli of

the tonalamatl (Fig. Se), we see that this and the previous era bear
to each other the lunar "Parrot:Quetzal" ratio 13: 12, while
together at 10000 years (25 Heads) they bear the Venusian-solar
"Eagle:Bald Eagle" ratio 5:8 to the preceding 16000 years (40
Heads). It will also be clear that as the fifth "Sun" of creation
this Era thus emerges as a fifth (13 Heads) of the full span or
cycle (65 Heads), whose astronomical meaning is touched on below.

Again in the Anahuac tradition and specifically as a Feather

(ihui tl), the 400-year period appears · to be incorporated into the
main toponym of the Ihuitlan Map (Fig. 6c), its ten Feathers giving
an appropriate date in the late tenth century AD, 4000 years from
3113 BC. In this same toponym the "source" of the flowing Feathers
is defined as a five-fold jade-like object, symbol of precious
water: that the jade could in fact be worth a fifth of 400, i.e.
80 years, is made abundantly clear from independent sources in the
Cll tradition. A "Jade" of 80 years is unmistakably marked out in
the Cuauhtinchan History between the years 1176, 1256 and 1336 AD,
9 Flint, 11 Flint and 13 Flint: these two 80-year periods are
designated by a Jade emblem and are glossed with the phrase "the
Jade (chalcatl) was completed". A good iconographic example occurs

188
in Boturini (p. 13) where doubling as the place-name Chalco "the
shrine of jade" it marks the period between the Aztec's arrival
in the first local town of their migration, the place of rushes
Tula-Xicocotitlan in 3 Flint 1144 AD, and their arrival in Chalco
with its maguey leaves in 5 Flint 1224 AD (cf. Sigllenza Map 116 7 to
1247 AD, and Fig. 6a, 7).

Once recognised for what it is, the Jade can be detected in a

variety of further sources. In the Zacatlan History collected by
Torquemada, the opening span of 720 years (which separates
Xihuitlapopoca of the late 14th century from the Cll Round 72 base)
consists of nine "model" reigns each one Jade or 80 years in length.
Then in the ClO tradition, in the Bachelorhood of Eight Deer, it
functions more complexly, in combination with the Round (equal to
the 52 divisions on the Jade's rim), at the highly resonant moment
in the Era chosen by that hero for his one fire-kindling, 1047 AD,
exactly 80 Rounds or 52 Jades (4160 years) from 3113 BC (3113+1047 =
4160; Fig. 6b). The probability of this reading is much
strengthened by the otherwise meaningless alphabetic gloss on the
"Jade-Round" emblem itself, which points to the Era base by citing
the Nahua name of the Era as such in the cosmogonical sequence of
five Suns, i.e. "nauhollin", "4 Movement" (Sign XVII: cf. Plates
1, · 2); the aptness of this gloss further increases when it is
considered that Eight Deer's kindling divides this 5200-year Era in
exactly the ratio, 4:1 (80+20=100 Rounds), which it itself is
defined by in the 26000-year cycle of five Suns (400+ 100=500
Rounds). The Jade appears again in the Borgia, on the
Tecuililhuitl page of the Fasts chapter (p. 30), which is dedicated
to Series I year bearers: there in its dominant central position
the Jade has its rim inset, analogously to the Eight Deer example,
with the tooth-total of 32 double bars. In sheer arithmetic this
again yields the five-Sun cycle of 26000 years, less one Head:
80x32x2x5 = 25600 years.

As the material jade, the 80-year period may readily be seen

as one set of a repertoire of tribute year-multiples mineral in
nature, extrapolated from the year itself in its homonym
"turquoise" (xiuitl). For in certain Era datings the Head seems to
occur as a large rock while the four-year Series Sign or leap-day
span may be identified with a stone or flint (Flint years appear to
be leap years in Series III). Reading the stone as 4 years makes
detailed sense of the difference between the two Nepopualco Maps, 4
stones or 16 years, which the second of them glosses in Nahua as
"iqhuac centlalicue", the "extra bit''; for this is exactly the
difference between 1 Flint of Round 72 in Cl (632 AD) and in Cll
(648 AD), Round 72 being the Chichimec base used by both Maps to
establish the date 1466 AD (written in arabic numbers in the text;
Plate 5). 13 Then in the Ihuitlan Map (Fig. 6c), the year given
above the toponym, 1 Reed, may be similarly pinponted if 8 years
or two stones are seen to be subtracted, literally, by the snake
that lies among the 400-year Feathers: for the overall distance
from 3113 BC to 1 Reed of Round 76, 879 AD, is 3992 years or ten
Feathers less two stones. It should however be emphasized that
these deductions are of a different order from those made concerning
the 80-year Jade, since this last is so unambiguously confirmed as
such in the Cuauhtinchan History.

189
 As for year multiples stemming from the ritual tonalamatl
rather than tribute arithmetic, the scholarly outlook is bleaker,
there being no generally agreed identification of markers for the
52-year Round or comparable periods. This must in part be
attributed to negligence. For example, little notice has been
taken of the 7-year sequence of year guardians, derivative from the
Nine Figures, which is so clearly spelt out and identified with ty~
Sign Deer (VII) in the Borbonicus screenfold (pp. 21-2; Fig. Sc).
Again, the same narrative that so cleanly defines the Jade, the
Cuauhtinchan History, likewise exposes a remarkable and no less
neglected 40-year period known as the Cuauh- or Eagle-Quecholli,
here rendered as "Cross Eagle" since this is how it is depicted
iconographically. Moreover, this period is just as cleanly
defined in other Chichimec sources. Attached always to Reed years
in Series III, the Cross Eagle is distinguished by the fact that
the Numbers qualifying this Reed Sign follow each other in ascending
order, to the limit of 13, which is 520 years or a decimal Round;
and that the Numbers-plus-Signs which spell out these lots of 40
years in the Round of years also spell out 40 days in the
tonalamatl, so that the distance from 1 Reed to 2 Reed is both 40
years and 40 days. As the two highest fliers among the 13
Quecholli (Fig. Se), the Eagle (5) and his bald companion (8) mark
out this 40-year period in the annals of the "eagle" towns
Cuauhtinchan and Cuauhti tlan, while the toponym Cuauhquechollan in
the Mendoza (p. 42) consists of the bald Eagle (8) together with
five dots (for Eagle). In the Cuauhtinchan History the two Eagles
cross their necks diagonally in a literal multiplication sign, L~
5 x 8 = 40 (Fig. 6d): placed with the years 1 Reed (1259) and 2
Reed (1299), this refers to the 40-year periods 1219 to 1259 and
1259 to 1299 which both in this text and in the Cuauhtitlan Annals
are glossed with the phrase "the Cuauhquechollan element was
completed". Beyond this, when qualified by each of the thirteen
Numbers in turn, these 40-year units build up into the full Cross-
Eagle Count of 520 years(= 13 x 40 = 52 x 10), the decimal Round
otherwise associated with the Quecholli (cf. note 9) and examplified
with the same 1 Reed date in the Tepexic Annals (Fig. 3c) and
possibly in the Borgia (p. 71, Plate 2).

Regarding markers for the 52-year Round (Fig. Sb), the

situation has been bedevilled mainly by the fact that there has been
little effort to concert such explanations and decodings as have
been offered, those of Caso being a notable example (1967: 16).
For his part Walter Lehmann drew attention long ago to the
iconography of the Round as a "year-binding", xiuhmolpilli in
Nahua; totals of five Rounds (260 years) are for example confirmed
by ties in narratives like the Mexican Annals (1247 to 1507 AD) and
the Tepechpan Annals (1299 to 1559 AD). And he went on to make the
quite reasonable suggestion that the knots and ties attached to
year names at one Round intervals in such cases could of themselves
denote a Round. He then applied this idea to such texts as the
Aztec Sunstone (Plate 1), where four such ties or molpilli are
bracketed on the outer band of fire snakes (Lehmann 1974: 322-23);
cf. Ibarra Grasso 1978). Permutated with the 18 dots which
accompany these four molpilli, Lehmann's reading can in fact be
satisfactorily related to the Chichimec Round 72 base, since 4 x 18

190
= 72 Rounds; and it can be applied also to the four lots of four
molpilli shown with the Aztec fire-kindling year 2 Reed in the
Borbonicus (p. 34; Plate 6), so as to identify it as the
Huizachetepec kindling of 1507 AD, 4 x 4 or 16 Rounds from the 72
Round base (the toponym Huizachetepec appears in the text).

Moreover, the all-important principle established by Lehmann

may be extended to other xiuhmolpilli iconography, notably the
fire-kindling scrolls of flame and breath (ayauitl) to which the
ties are related in these two sources and yet more systematically so
in Fejervary pp. 5-14. In both the Huitzilopochtli histories,
Rios and Telleriano Remensis, three double breath-scrolls of this
type confirm the six Rounds or 312 years that have elapsed between
1196 and 1508 AD, at the latter date (Fig. 6e). In the Boturini
screenfold nine such breath-scrolls or 468 years, uttered by
Huitzilopochtli from his cave in Colhuacan (p. 1), carry us tidily
from 1 Flint 648 AD (the Chichimec base) to 1116 AD, the start of
the continuous year count in that text (Plate 7). Then, from the
same base, the 14 flame-scrolls in the Azcatitlan Annals confirm
emperor Acamapichtli's inaugural year 1 Flint 1376, 14 Rounds or
728 years later (p. 22). Counting from the earlier base in Round
60, appropriate to its Cl0 convetion, the Acatlan Genealogy
(Egerton; Plate 9) shows 16 such scrolls as breath and 10 as flame
to place the 16th generation at 12 Flint 1388 AD (pp. 4, 20), the
date independently estimated by Burland (1965). In all these cases
the Round markers are so positioned in relation to the year dates
they locate as to leave little doubt about their relevance to them
or that they indeed have numerical value.

The consistency of these examples (which could be multiplied)

with the chronological framework established earlier encourages us
to turn t-0 what appears in context arid function to be an
iconography equivalent to that of the tie and the scroll, in other
Cll texts and in most major Cl and Cl0 texts (Fig. Sb).
Conceptually the key to this other iconography lies in the notion of
the body's nine orifices, "rounds" literally as that, presided
over by the Rain-god in his role as as the ninth of the the Nine
Figures and hence, with his full waterbag, of the nine moons of
pregnancy and gestation, during which these orifices form. It is
notable that the molpilli tie actually substitutes for the Rain-god
as the ninth of the Nine Figures in Cospi pp. 1-8, while in the
Azcactitlan Annals (pp. 28-42) the three fire-kindlings of
Tenochtitlan's hegemony in 1403, 1455 and 1507 AD, between
Huitzhuitl and Moctezuma II, are also marked by the Rain-god and
the cradle of his new-born. Similarly, the Rain-god faces the
tangle of 19-20 scrolls below the 8 Rabbit date in the Teozacoalco
Map, confirming it as 1046 AD, almost 20 Rounds from the Cl0 base;
and he appears to qualify as Rounds the 13 units on the opening
emblem of the Vindobonensis reverse, confirming the accompanying 7
Flint date as 720 AD, independently established by Caso, as 13
Rounds from the Cl0 base.

In addition to this, as properties of the Rain-god particular

orifices of the body can be seen to assert themselves as markers,
notably the lips or mouth, terrestrial as a cave, and the eye,
celestial as a star. For example, the toponym Chicomoztoc or

191
Seven Caves may be rendered specifically as the seven orifices of
the head when embodying the span of 7 Rounds or 364 years (Fig. 6g;
cf. Cuauhtitlan Annals f.l). As the commonest orifice markers,
the cave-mouth and star-eye are amply confirmed by their appearance
and use in the continuous year-narratives, where they ratify
significant totals of Rounds.

Sums of mouth-Rounds appear in the Tepexic and the Tilantongo

Annals alike, where on each occasion they pick out in the
continuous year narrative, a given period of calendrical or
historical significance. Hence the ten mouths which attach to the
second of two 1 Reed 1 Alligator dates at Rounds 14 and 24, 2345 to
1825 BC in the Tepexic Annals (pp. 8-9) ratify the distance between
them as a decimal Round or Cross-Eagles Count of 520 years; again
the ending date in the same narrative 2 House Round 86, 1361 AD (p.
50), is surrounded by 14 mouths which place it symmetrically both
14 Rounds after the restoration of Round 72 and 14 Rounds before the
Era's full complement of 100 Rounds (cf. Fig. 3a). At the start of
the Classic and of chapter 2 in the Tilantongo Annals (p. 14), the
year 7 Reed 343 AD is framed by 6 cave-mouths: these correspond to
the 6-Round distance from the tree-birth of Round 60, the ClO base
(Plate 8).

In these examples the mouth markers simply reaffirm dates

already established in a continuous year count: in the complex and
problematic Cuauhtinchan History, however, they make it possible
for the first time to calculate the proper time-depth of the
narrative, which begins in the great Tula on the Gulf Coast. For
prior to the continuous year count which begins only in 9 Flint 1176
AD (p. 51), time is measured by "years" and "days" which are
qualified by this same mouth marker, called "precious lips" or
quetzal-teueyac in the text (Fig. 6f). The 59 "years" and 13+10
"days" in question are said to "walk the road of Toltec Rounds" (p.
13: yn otli y xiuitl ynic ualneneque yn tolteca yn Icxicouatl yn
Quetzalteueyac), which suggests that they should be in fact counted
as Rounds. So that at the start, in the year-set, to move from 1
Flint to 2 House involves a span not of 1 but of 53 years, 3112 to
3059 BC, a 53-year kindling span having actually been observed
around Cholula (cf. Caso and Smith 1966: 30; also Molina on the
xiuhmolpilli of 53 years). Similarly, in the day-set 8 Eagle to 7
Deer, and 8 Rabbit to 4 Movement, the periods involved are not
days but Rounds-plus-days: even on historical grounds they cannot
be interpreted literally as days on account of all that happens in
them.

Read thus, the sequence of these mouth Rounds in the

Cuauhtinchan History makes perfect sense in the framework of the
Era. For the 59-year-Rounds and the 13 day-Rounds combine to
produce the Chichimec Round-72 base while the further ten-Round
distance to the start of tbe continuous year-count in 1176 AD is
provided by the second run of day-Rounds, ten in all. Once
accorded this order of time-depth the narrative, with its repeated
migrations and settlements, acquires a clarity it is otherwise
bound to lack; and as its subtitle suggests it should it can then
serve as an excellent guide to "Toltec-Chichimec history". For
after the initial diaspora it reports in detail such events as the

192
Toltec struggles against the Xicalanca Olmec at Cholula around Round
50 or 500 BC, a date corroborated by SahagGn, Ixtlilxochitl, and
the Tepexic Annals; the famous Chichimec emergence from Chicomoztoc
or Seven Caves in Round 65 (late 3rd century AD), exactly the date
given in the Tilantongo Annals (p. 10) and in the Cuauhtitlan
Annals, as 7 Rounds or 364 years before _Round 72 (f. 1); and then
the great circle of their clockwise progress down the Gulf Coast,
through Anahuac, Tepanec lands in the west, the highland valley and
finally the portals of Popocatepetl and Iztaccihuatl, eastwards
back to Cholula, at dates between Rounds 72 and 82 (mid 7th and
late 12th centuries AD) reciprocally confirmed at every stage by
documents from the towns involved. Along with the Cuauhtinchan Map
and the closely allied Borgia screenfold (cf. Nowotny 1961: 34 and
Brotherston 1982: 54) this text, with the aid of these Round
markers, becomes a major guide to the whole Chichimec historical
tradition.

More frequent and freer-standing than the mouth, the star-eye

readily identifies itself as a Round-marker in Tilantongo texts like
"Eight Deer in Tututepec" (Becker screenf old) and the Tilantongo
Annals. In both these cases the star-eyes reaffirm, from the
Round 60 base, dates that are already well established, like the
dyJ;).astic shift around 800 AD, Round 75 or (60+) 15 star-eyes
(Tilantongo Annals, pp. 4, 14 and 22; Plate 8) and Eight Deer's
kindling in 104 7 AD, Round 80 or ( 60+) 20 star-eyes (Becker). In
the Vindobonensis reverse (pp. 9-11; Fig. 6h) three Rounds of life
at Tilantongo after Eight Deer's death in 1063 and until the early
1220s are celestially noted as three star-eyes. A star-eye also
appears. _among the 14 flame- or Round-scrolls in the Azcatitlan
Annals and may be identified with the "ixtetlil comoloua" star-eye
marker inset into Itzpapalotl at Round 72 in the Cuauhtitlan Annals,
in the Cll tradition. Whole bands of star-eyes introduce texts
from the Anahuac and the Tilantongo domains (G6mez de Orozco,
Selden Roll; Selden p. 1, Bodley p. 2). And on each occasion the
dates they establish in Rounds fit perfectly into the Era scheme as
a whole. The 60 star-eyes in Gomez de Orozco lead straight from
the Era base to the time of the tree-birth; for their part the 65
star-eyes in the Selden Roll lead to the Chicomoztoc or Seven Caves
event which indeed appears in the text immediately afterwards,
Round 65 or the mid 3rd century AD being the date independently
recorded for this event in the Annals of Tilantongo and of
Cuauhtitlan and in the Cuauhtinchan History (Plate 11).

In the Selden Roll, the narrative continues after

Chicomoztoc/Seven Caves along a sky road to a barrier of rocks at
Ihuitlan, where a fire-kindling is orchestrated by Two Dog in the
year 10 House; and then it returns along the same road to
Chicomoztoc and the date 8 Flint. As before, the position of
these events in the Era is given in Rounds by star-eyes, 17 on the
way to Ihuitlan and 10 on the way hack. Added to the initial 65
star-eyes, the 17 place Two Dog's kindling year in Round 82, i.e.
1161 AD, which is just ten years prior to the kindling by Two Dog
recorded in the Tepexic Annals (7 Reed Round 82, 1171 AD, p. 43).
Moreover, Two Dog's kindling occurs three star-eye Rounds after the
appearance of Four Alligator which is exactly the interval given by
the Tepexic Annals, where Four Alligator presides over the seventh

193
kindling in 1015 AD (p. 41), and by other Anahuac texts (Maps of
Ihuitlan, Tlalpitepec and Coixtlahuaca ii). 15 Returning from
Ihuitlan the final ten star-eyes take us on to Round 92, so that
the year 8 Flint should be read as 1692 AD: this is identical with
the final projected date in the Tepexic Annal$ (8 Flint Round 92,
1692 AD, p. 52), the "12-Head" total at 12/13 of the Era which can
be shown to have marked a historical turning point for the
surviving priests and scribes of the Anahuac and Toltec traditions
to which all these texts belong (Brotherston 1982: 27). In any
case, the repeated coincidences between the Tepexic Annals and the
Selden Roll tend reciprocally to confirm the time-depth of the
former and the 52-year value of the star-eye in the latter.

In the Tilantongo texts, the opening star-eye totals of 6 and

15 mentioned above coincide with the start and end of the "Classic"
chapter common to the Tepexic and Tilantongo Annals, at 6 and 15
Rounds from the tree-birth of Round 60, Rounds 66 and 75 of the
Era. In this the Selden is especially gratifying since, like the
dual wheels in Mexicanus p. 9 and like the "nauhollin" gloss on
Eignt Deer's Jade-Round kindling, it provides its own small key to
Mesoamerican chronology (Fig. 6i). For the six star-eyes which
establish the Classic start date 4(?) Reed 327 AD, 6 Rounds from
the Round 60 Base, carry below them an unobtrusive but most
pertinent statement of their Round position in the Era. It
consists of the number 66 spelt out in bars and dots and subdivided
right to left as 21+45: the logic of this stems from the fact that
66 is one of the rare sigma numbers to have as possible components
two smaller sigma numbers (i.e. t6 + I:9 = tll, or 21+45 = 66).
Hence we may be confident in seeing here a reminder that Round 6 of
the Tilantongo convention equals Round 66 of the Era.

More and more examples of these orifice Rounds, mouth and eye,
could be winckled out of Mesoamerica 's iconographic texts and every
time could be shown to yield apt and significant datings. And the
discussion of Round markers as a whole could be extended to ·the more
esoteric varieties: the 20-Round or millennial tree five of which,
in the centre and the four tribute fields, complete the Era (cf. A.
Castellanos 's detection of the four "1040-year" markers at the start
of the Tlalpitepec Map; Corona Nunez 1964-67: iii: 114); the 13
Quecholli who according to their inherent values may individually
em-body decimal Round totals (cf. Itz-papalotl and Round 72 above,
and the Eagle 50 plus 5 star-eyes placed at 5 Flint Round 55, · 248
BC, Tepexic Annals p. 14; Plate 10); and the 13 Heroes who
according to their inherent values may individually embody ordinary
Round totals (cf. Sun 4 plus Quetzal-snake 9 as 13 Rounds or "676
years" in Historia de los mexicanos). Each carrying 100 flame
scrolls or snakes (mixcoa), two of these last on the Sunston€,
Hero Sun (4) and the Hero Fire-lord (1), supply in Rounds (i.e. 400
+ 100) the same 26000-year cycle noted above as 65 Heads: and as in
the Cuauhtitlan Annals, where "CCCC mixcoa" likewise establish the
same ratio and total of Rounds ( 400 + 100), the sheer arrangement
of the Hero Rounds, 400 pre and 100 post, pinpoints 13 Reed as
this Era's inaugural Series III year (Plate 1).

But such further exploration belongs elsewhere, along with the

fact that in the International Day Count of Mesoamerica this Era's

194
name as a Sun, 4 Movement, corresponds to the spring equinox of
3113 BC; and that the widely-reported 26000-year cycle, five times
this fifth Era's 5200 years, equals the full precession of the
equinoxes. The detail given already is more than enough to reveal
the use of year-multiples in the year calendar, within the
previously established scheme of the Era and its secondary bases.
As is shown in Fig. 7, year multiples persistently corroborate
dates independently established in continuous year counts; and
taken together they leave no doubt, in particular, about the true
formal length of the year span in the Tepexic Annals.

Conclusion

Over the four sections of this paper, the attempt has been
made to establish Anahuac as the home of the only remaining major
group of uncorrelated Series III texts; to ascribe to that group a
year correlation whereby 1516 AD = 1 Flint ("Cl"); to show that the
time depth in the chief text of the group, Vindobonensis obverse or
the Tepexic Annals, far from being illusory is commensurate with
the Era span of the tun calendar and extends to 13 Reed (Cl) 3113
BC; and finally to explore the use of year-multiples like the Head
and the Round, within the Era framework. In this last, certain
of the year multiples, like the 80-year Jade and the 40-year Cross
Eagle, though neglected are quite unambiguously defined in native
sources; others, however, like the "mouth" and "eye" Round have
had to be deduced by less direct means, which could of itself
prompt deb'a te. For that reason it should be emphasised again that
the main argument of the paper is not in any way obliged to invoke
these multiples: it is merely supported by them, including the
firmest of them, like the Jade. As for the correlating of 3113 BC
with the year 13 Reed, through Cl, this is achieved according to
norms generally respected in Mesoamerican calendrics, while the
time-depth from this inaugural year established for the Anahuac
texts results from a wholly orthodox and conservative reading of the
year dates in the Tepexic Annals, and from their incessant parallels
with those in tun and and other calendar year texts, even a few of
which would be enough to make the case in principle. And nowhere
is the notion of "scribal error" appealed to, to explain away
awkward discrepancies; rather anomalies previously deemed mistakes
an~ shown to be deliberate and meaningful.

Though in turn the source of further valuable confirmation,

certain aspects of the Anahuac tradition have been left unexplored,
for lack of space: their documentation of the lowland Tula on the
Gulf Coast, the first city named as such in east and west
Mesoamerica; their clues to the native distinction between the
International Day Count of Mesoamerica (tonalpoualli) and the day-
in-the-year count of the year calendar (cemilhuitlapoualli); their
astronomical data, like the sidereal year cycle of the Tepexic
Annals (Fig. 3c) which impinges on the whole concept of the 26000-
year precessional cycle; and the degree to which texts like the
Anahuac Map, with its hidden year-dates, render invalid the
customary division made between ritual (synchronic) and historical
(diachronic) texts, the former being more like condensed hence
poetic syntheses of both cosmic and political time. Other essential

195
Fig. 7 Compendium of Series III dates with year multiples
(A= Annals; M Map)

Correlation 1: year O Round O • 3113 BC 13 Reed

Er.:3 BC
2865 248 5 Flint, Eagle+S eyes+S years (5, 55x52); Tepexic A p. 14 (Pl.10 )
AD
3160 47 1 Reed, 60 eyes (40, 60x52); Gomez de Or0zco p.1
3420 307 1 Reed, 65 eyes (40, 65x52); Selden Roll (Pl.11)
3918 805 5 House, (60+) 8+7 eyes (18, 75x52); Aubin ms 20 (Fig.tc)
3992 879 1 Reed, 10 Heads less 2 Stones (4000-8); Ihuitlan M (Fig.6c )
4160 1047 (13 Reed), Jade-Round (80x52); Zouche reverse p.35 (Fig.6bl
4274 1161 10 House, 65+17 eyes (10, 82x52); Selden Roll
4474 1361 2 House, (72+) 14 mouths (2, 86x52); Tepexic A p.50
4525 1412 (1 Flint), (1+) 3x29x52 years; Mexicanus p.9 (Pl.4)
4529 1416 5 Flint, (805 AD+) 12 eyes less 2 heptad leaves (12x52 less
14 inclusive); Aubin ms 20 (Fig.le)
4805 1692 8 Flint, 82+10 eyes (21, 92x52); Selden Roll

Correlation 10: year O Round O • 11 AD 13 Reed

sum AD
228 239 7 Reed, 4 eyes (20,4x52); Yolotepec M
273 284 f 13 Flint, 5 scrolls (13, 5x52); Acatlan Genealogy p.1
316 327 4 Reed, 6 eyes (4, 6x52); Selden p.1 (Fig.6i)
332 343 7 Reed, 6 mouths (20, 6x52); Tilantongo A p.14 (Pl.8)
681 692 5 Flint, 13 eyes (5, 13x52); Bodley p.2
709 820 7 Flint, 13 units with Rain-god (33, 13x52); Vienna reverse p.1
768 779 1 Reed, 14 scrolls (40, 14x52); Nativitas M
816 827 10 Reed, 15 eyes (36, 15x52); Selden p.2
820 831 1 Reed, 15 eyes (40, 15x52); Tilantongo A pp.4-14-22 (Pl.8 )
1035 1046 8 Rabbit, 19 scrolls with Rain-god (47, 19x52); Teozacoalco M
1036 1047 9 Reed, 19 eyes (48, 19x52); Eight Deer in Tututepec
1214 1225 5 House, (20+) 3 eyes (18, 23x52); Vienna reverse pp.9-1 1 (fig.6r. l
1.377 1388 12 Flint, 26 units with Rain-god ( 25, 26x52) Acatlan Gene~logy
pp.4-20 (Pl.9)

Correlation 11: year O Round O • 647 AD 13 Reed

sum AD
469 1116 1 Flint, 9 scrolls (1, 9x52); Boturini p.1 (Pl.7)
496 1143 2 Reed, 9 scrolls (28, 9x52); Xochicalco New Fire Stone
529 1176 9 Flint, 10 mouths and days (9, 10x52); CUauhtinchan H p.51
(cf-Fig.6f)
548 1195 2 Reed, Hero 5 twice (28, 10x52); Rios p.87
577 1224 5 Flint, (1144 AD+) 1 jade (80); Boturini p.13 (Fig.Ga)
600 1247 (2 Reed), 1 jade and Hero 5 twice (80, 10x52); SigUenza M
'Chalco' (jade is chalchiuitl)
612 1259 1 Reed, hinge of Cross Eagle Count 1219-1259-1299-1339 etc:
CUauhtinchan H, CUauhtitlan A, Borgia p.71 (Fig.6d; Pl.2)
689 1336 13 Flint, (1176 AD+) 2 jades (80x2) CUauhtinchan H
708 1355 6 Reed, 13 eyes with 9 (32, 13x52); Santa Cecilia Tlaloc
728 1375 13 Reed, 7x2 eyes (14x52); Sunstone (Pl.1; cf. Mendoza Pl.3)
729 1376 1 Flint, 14 flames with eye (1, 14x52); Azcatitlan A pp.22-3
784 1431 4 Reed, 2 Heads (800) from 13 Reed C1 631 AD; Quinatzin M 2
811 1458 (5 R~bbit), 2 Heads (800) plus 10 years; Tlatelolco A*
816 1463 10 Reed, (1363 AD+) 5 Flags (100); Aubin
819 1466 (13 Rabbit), 2 Heads 3 Stones 6 years (818); Nepopualco Mii*(Pl.S l
825 1472 6 Flint, 15 ties (46, 15x52); Tula Caryatid
833 1480 (1 FlintJ, 16 eyes (1, 16x~2); Tizoc Stone, top (88 in the Era)
860 1507 2 Reed, (1247 AD+) 5 ties (Sx52); Mexican A
2 Reed, (1195 AD+) 6 scrolls (6x52); Rios (Fig.Ge)
2 Reed, 4x4 ties (28, 16x52); Borbonicus p.34 (Pl.6)
874 1521 (3 House), (1458 AD+) 3 Flags (60) 3 years; Tlatelolco A
912 1559 2 Reed, (1299 AD+) 5 ties (Sx52); Tepechpan A

*From 1 Flint 648 AD

196
factors that have had to be left on one side include the whole
question of the historical progress through the year Series (I, II,
III, IV; Fig. 2b) and the huge importance of the archaeological
Monte Alban sequence from at least 500 BC; the ritual patterning of
year multiples and their association with particular moments in the
Era (e.g. Seven Caves as the Rounds 65 to 72); and the principle of
shifting between number bases (e.g. the 5-year bar and the SO-year
stick in Laud, p. 44) and between dimensions in time, as via the
great year of 26000 years to the evolutionary scheme of five Suns,
for which 100s of millions of years are recorded hieroglyphically.

In sum, this is very much a first step, a new orientation

which if at all correct offers to reveal coherence in the year
calendar hitherto denied or not seen. Above all, read as they
demand to be, the Anahuac texts, like the tun texts of the Maya
lowlands, provide the "missing link" in Mesoamerican chronology, a
time-depth that conjoins the political history of such centres as
Tenochtitlan, Tilantongo and Tepexic with the awesome scheme of
Suns spelt out in the Cuauhtitlan Annals and the Popol vuh.

Notes

1 For a summary of the various correlations proposed for the tun

calendar, see Kelley 1976: 30-33. Added proof for the Goodman-
Martinez-Thompson correlation appears in Lounsbury's study of the
Dresden Venus tables, published in this volume, and in Edmonson's
translation of the Chilam Balam Book of Tizimin (1982; cf. note 5);
also the succession of katun dates in the first chapter of the
Chilam Balam Book of Chumayel (pp. 2-15) can be shown to support
Thompson's "11.16" katu,.-ri formula (cf. Brotherston 1982: 31).

2 The main reasons for believing the year of the year calendar to
be tropical-seasonal rather than metric are as follows:

a) as a set of four, the year-bearer days (e.g. 13 Reed) are

distinguished from other days by just the markers otherwise used to
identify the tropical year of agricultural tribute, while at Monte
Albin the set existed in its own right and was identified bf the
seasonal Rain-god (cocijo);
b) particular year Series Signs are identified with good or bad
rains and harvests, which of course depend on the tropical rather
than the metric year, so that in Series III for instance Rabbit
(Sign VIII) meant drought; also the 20-day divisions of the
agricultural (i.e. tropical) year are regularly intercalated with
the Series Signs (e.g. in the Series II annals of Tlapa or Azoyu,
and in the Rios from Quecholli 1 Reed to greater Tecuililhuitl 2
Flint, November 1519 to July 1520);
c) native chronographers like Chimalpahin correlate native with
Christian (i.e. tropical) years over spans in which more than a
year's difference accrues between the tropical and the metric year,
without regard for that difference;

197
 d) in native correlations, year names are made exactly
equivalent, for example as lungs breathing in time (see below),
which they could not be as tonalpoualli days of the metric year (on
the progression through the year-bearer Series cf. Fig. 2b and
Edmonson 1982: 195); also the tonalpoualli day given for Cortes's
arrival (8 Wind) is quite distinct from the cemilhuitlapoualli day
in the seasonal "month" Quecholli ( 1 Wind).

3 Over the vast span of the Tepexic Annals no less than ten
characters named Two Dog (Sign X) are featured in the narrative,
one of them twice, each being distinguished by a unique combination
of headdress, ear-plug and emblem-load. Five of them are
designated as principals by an emphatic eye-marking and by having
(with one exception) a contemporary and surrogate of the same name
(p. 23, 354-63 AD, start of chapter 2; p. 28, 510 AD, pulque
ritual; pp. 30-31, 722-24 AD; p. 32, 805 AD, Jaguar Claw
president of the second kindling, without partner; p. 43, 1171
AD, president of the ninth kindling); the first of them appears
alone near the end of chapter 1 (p. 20, 173 AD). When not
otherwise stated, the Two Dog referred to in this paper is the one
who lived in the latter half of the 12th century AD and who presided
over the ninth kindling in the Tepexic Annals, where he is shown
with a feather headdress, oblong ear-plug and a tobacco gourd
(yetecomatl).

4 The east-west-north-south sequence of the Quecholli (cf. Fig.

Se) is 12-1-3-13 in Fejervary (a tribute map consonant with the one
in Mendoza, Plate 3) and 12-5-8-4 in Borgia; in Dresden (pp. 25-
28) the four quarters give the bar-and-dot total of 59, twice the
lunar 29.5 days, and bear the moon glyph. For the Tepexic,
Tilantongo and Tenochtitlan totals see Fig. ld; in each case east
is the source of cacao. The elaborate tribute symbolism of four
fields, trees and feeding birds is explained to some extent in the
last chapter of the Museo de America Codex; on the lunar-solar
significance of the tribute ciphers 29, 246 and 365, see
Brotherston 1982a. The Tepexic toponym has been previously
identified by Jansen (1982), though with scant regard for the
actual positions of it and other Anahuac towns or their general
separateness from the southern Mixtec-Tilantongo tradition, and by
Chadwick (1971: 477) who, following Caso (1960: 58), points to its
dynastic importance for and rivalry with Tilantongo as part of an
Anahuac federation that included Xicotlan northeast of Coixtlahuaca
(Caso 1961: 242), Cholula, and the "Sun" town Tehuacan.

5 Edmonson discusses in detail the reasons behind the Valladolid

Ref or m of 1 7 5 2 which extended the tun Er a t o 2 0 8 8 (1 9 8 2 : 1 7 2 , 1 9 4 ,
197). However, while pointing out the lowland Maya's interest in
reconciling their 360-day tun with Christian and Mexican tropical
years, he does not specifically note that 2088, 4 Ahau in the tun
calendar, marks the completion of 100 Rounds or 5200 such years.

6 Besides Nowotny, both Caso and Melgarejo considered that to

read the Tepexic Annals normally would yield an "impossible" time-
depth (cf. Davies 1977: 68). It is noteworthy that precisely when
denying this time-depth Nowotny chooses to derogate the capacity of
Mesoamerican script, reducing it to the merely mnemonic: "Diese

198
Schrift ist, wie wahrscheinlich auch die der Maya, trotz ihrer
vielen phonetischen Elemente und trotz der exacten chronologischen
Angaben eine durchaus piktographische, d.h. sie erfordert einen
mllndlichen Kommentar" (1961: 48).

7 On the Euripos or Navel of the Sea mentioned in Plato's Phaedo,

cf.: "We meet the name again at a rather unexpected place, in the
roman circus or hippodrome, as we know from J. Laurentius Lydus (De
Mensibus, I, 12), who states that the centre of the circus was
called Euripos; that in the middle of the stadium was a pyramid,
belonging to the Sun; that by the Sun's pyramid were three altars,
of Saturn, Jupiter, Mars, and below the pyramid, altars of
Venus, Mercury and the Moon, and that there were not more than
seven circuits (kykloi) around the pyramid, because the planets were
only seven" (Santillana and Dechend 1970: 206-7). On the
apparently identical arrangement of sun plus planets, 3 superior
and 3 inferior, at Palenque, cf. Brotherston 1982a: 110.

8 At these dates in the Palenque panels the emblem used involves

either the sun glyph (kin) or Ahau (Sign XX) as in Fig. 4b; on this
cf. Kelley 1976: 278. On the Hiatus, see Willey 1974.

9 For the Quecholli, see Fig. 5e; and for other examples of
this convention, see Plate 2 (Quail, 4, as 40 Rounds), Plate 10
(Eagle, 5, as 50 Rounds), and Brotherston 1982: 56 (Owl, 6, as 60
Rounds). For more details of the Cll texts involved, see Nicholson
1978.

10 Widely commemorated in the Aztec tradition, this year appears

to be established on this same page of the Mexicanus in the
Christian ·Era, by data to the left of the two wheels, viz. 4
Feathers (1600 years) less 3 Flags (60 years, cancelled by shading)
plus 18 dots, i.e. 1600-1540-1558 AD; this is followed by year 2
Reed (1559) and a further 12 dots (Plate 3). The present reading of
Mexicanus p. 9 revises certain details of my previous attempt (1982:
87), of which Hanns Prem has kindly offered well-justified
criticism.

11 Important early notes on the use of the Head and the Round in
the year calendar include Gemelli 1700: 38 (re Hero Rounds in the
Sig\lenza Map), Borunda 1898 (1794-95), and Fabrega 1899 (1792-- 97,
drawn on by Humboldt 1810 and more recently by Broda 1969); see
also Chavero 1892, vol. 2, A. Castellanos 1912 (re Fabrega's 20-
Round or 1040-year period in the Anahuac texts, cf. Corona Nunez
1964-67, iii: 114), and the works by Lehmann and Caso quoted below.
A vocabulary of Mixtec terms for such periods has been generously
reported to me by M. Dllrr.

12 Good examples of these . techniques appear in the pages of bar-

and-dot numbers in the Fejervary (pp. 5-22) and the Cospi (reverse
pp. 1-11). Decimal value is well expressed in the angled-bars of
the former (e.g. p~ 14-15), in the 2460 sidereal moon total of the
latter (i.e. 10 x 246 or nine sidereal moons: cf. Fig. ld and note
4), in the Quecholli decimal Round convention and many other
aspects of the Mesoamerican ritual; sigma value is inherent in the
very position of Eagle in the Quecholli (5) and the Twenty Signs

199
(XV) respectively (cf. Brotherston 1982a). Modern survivals of
these techniques are reported though little commented upon by van
der Loo (1982), e.g. the squaring of 13 (p. 235) and the equation
of four moons with a Mercury year (4 x 29 = 116, p. 234).

13 A third copy of this Map, not mentioned in the Handbook of

Middle American Indians Census and equivalent to the 632 AD Map,
has been reported and described by Prem (1972), who, however, in
identifying all three Maps with Santa Maria Nepopualco, Puebla,
interprets the stones as boundary markers and the two 400-year
Feathers as ravines (he does not comment on the six xihuitl ends of
the boundary cord). In native terms both readings may of course be
possible, especially since the element -poal- in the toponym is
known to refer to counting in general and to counting by 400s in
particular (cf. the Christian Era count in Mexicanus, Pl. 4 and
note 10). This same multiple reading of native iconography is
after all clear enough in star-eye Round markers which actually
shine from the sky, and in cases like the Selden Roll where the
rocks as Ihuitlan (10 3 /4) and Chicomoztoc (12) prove to confirm in
Heads the date established for these places in Rounds (1171 and 1692
AD). On stone and Flint imagery attaching to "dropped leap-day"
formulae in Aubin ms 20, the Tepexic Annals (pp. 31-41) and the
Tilantongo Annals (p. 3), see Brotherston 1982a: 123.

14 Ritually a penance period (as for Tula's One Reed, Chadwick

1971: 477), seven years is a prime factor in such larger multiples
as Seven Caves (364 years) and the ideal lifetime of 70 years
recorded in the Mendoza and found throughout North American
chronology (Brotherston 1982: 39; 1982b; 1983a); in Boturini p. 1
(Plate 7) it appears with decimal angle-value to give three such
lifetimes or 210 years, the distance to the end of the text (p. 21).

15 The great calendrical importance of Four Alligator can be

judged by the fact that his kindling both completes the 5-Round
leap-day sequence (see note 13) and is the first to be measured as
an exact Round 963-1015 AD. He is exactly contemporaneous if not
identical with the founder of the second Tilantongo dynasty and
father of Eight Deer, Five Alligator, also a calendrical reformer.
(Shifts of one number in names are common in the Mixtec and Anahuac
traditions).

200
 Plate 1 Suns tone

3113 BC u Top centre: inaugural year of the Era 13 Reed (c1), 3113 BC;
possibly also 13 Reed (c11), 1375 AD (cf. Plate 3). Right
and left rim: 10 x 10 Round-flames in the Hero ratio 4:1,
i.e. 2o800 pre-Era years right, and 5200 Era years left; here
the Chichimec Round-72 base appears as 4 x 18 ties. Centre:
the tonalamatl name of the Era or Sun 4 Movement or Quake
XVII), incorporating the names of the four previous Suns.

oc 3113 13 Reed (C1)

.2.lM 4 X 18 X 52
AD 631 13 Reed

647 13 Reed (C11)

..11§. 2 X 7 X 52
1375 13 Reed

201
 Plate 2 Borgia p. 71

. . .. . •.· (' .
------.-- .- ',•-----· . ; .:;'

13 12 11 10 9
On his throne, Hero 4 (Sun) drinks the blood of a decapit-
1 Reed 1259 ated .Quail (flier 4) while the head of another lies below;
8 the Quecholli frame indicates a Cross Eagle Count to 1 Reed
1259 AD, in line with that in the Cuauhtinchan History and
Cuauhtitlan Annals: the position in the Era is given by
5x20R 4R 7 the two Quails (2 x 40 Rounds) plus the blood spurts from

Hero 4
•
40R
the neck of the upper Quai 1 ( 4). The blood flow then bapti-

•
40R 6
ses as Rounds the five Flags held by the Sun to give the
span of the Era ( 5 x 20 Rounds) , the name of which appears
4 Movement below his throne (4 Movement). As on the Sunstone (Plate 1)
this distance is then multiplied by his Hero number to give
the pre-Era span of 400 Rounds.
1 2 3 4 5
BC J11J 1J Reed (C1)
J2x40x52
4J68
t4x52
AD 1255 1J Reed ( C 1)
1259 1 Reed (C11)

202
 2 House 13 Reed
1. l° r, c:.,, ·- ~-
\
1325 AD 1375 AD

~ 19

south
west

House
leaves

Tenoch-
/
Tzom- north
titlan panco
19 leaves 16 leaves
.-I

.
0..
,6

eed

~ leaves
N
0 east
'Ea, CW'\
,:: 0
C\l
M 2 Reed
a,
.µ
1351 AD
~
.....
ll-t

Around the central toponym of Tenochtit lan, the relative pos-

itions of its four tribute fields are given by Tzompanco
(skull-rac k), due north and to the right. Around the rim run
the years from 2 House 1325 to 13 Reed 1375, 2 Reed .1351 being
picked out as a kindling year. Totalling 51, these years are
matched in the map area by the 51+16 years shown as leaves in
the four quarters~ the extra 16 reflecting the gap between
Cl and C11 year names; also the totals to west and south
amount to Metonic cycles. The 14-Roun4 distance from the Chi-
chimec base is recorded in the 7 cactus spikes pronounced to
be Rounds by Tenoch's scroll, these being multiplied by the
2 of the 4 years of the stone beneath that are actually named
on the map, House (west) and Reed (east).
 Plate 4 Mexicarrus p. 9

House
~ Peter\ ~
1575
b
key
l b (1558)Rab bit
I \ Flint

~ Reed
J
Of the two wheels, the Mesoamerican one (right) shows a Series III Round while the
Christian one (left) shows a solar cycle; they mesh at 1 Rabbit (C11) 1558 AD. In the
Era this date is reached through three full turns of the two wheels (St Peter's key,
Jx29x52 years) plus five turns of the Christian wheel (St Peter's book, 5x29 years);
the 400-year feathers, 20-year flags and dots to the left, at right angles, appear
to confirm the same date in the Christian Era, immediately prior to the 2 Reed kindling
year lower left (cf. note 10).

:OC 3112 1 Flint (c1)

.4lli 3 X 29 X 52
AD 1412 1 Flint
-14.§. 5 x 29 excl.
1558 1 Rabbit (C11)
 Plate 5 Nepopualco Map i i

1466

(~ X s)
4

AD 648 1 Flint (C11)

2 X 400
818 J X 4
{6
1466

cDrn:} X2
4 400
yncalco The Christian date 1 1466 1 is counted out by feathers,
ynchichimeca stones and grass cords from the Chichimec base (yncalco
yn chichimeca), Round 72. In the c0111plementary Nepo-
pualc.:> Map i, the count is from 632 AD, also Round 72
but 1 Flint in C1 rather than C11; the 16-year gap is
shown as four stones, here glossed as 'Yquhac oc centla-
licue', 'with the extra bit'.

205
 Plate 6 Borbonicus P• 34

4 2 Reed
Huizache
I 1507 AD -tepetl
4
I
4 4, 4
/
l
4 4/ '4 = 16R
I
4 - 4 - 4 = 28R (+72 = lOOR)

In the year 2 Reed 1507 AD, New Fire is brought from AD 648 1 Flint (C11)
Huizachetepetl to Tenochtitlan. The ties on the fire- 8J2 . 16 X 52
wood around the furnace give the distance in Rounds 1480 1 Flint
from the Round-72 base while the distance to the full 1507 2 Reed
100 Rounds of the Era appears in the supplementary
firewood.

206
 Plate 7 Boturini p. 1

9 scroll
Rounds
1 Flint
1116 AD

+ 7

:C 7 leaves

The distance to the year-date 1 Flint 1116 AD from AD 648 1 Flint (C11)
the Chichimec or C11 base (Round 72) amounts to nine 468 9 X 52
Rounds: these are shown as breath scrolls uttered by 1116 1 Flint
Huitzilopochtli from within the cave at Colhuacan. 209 '3x7x10 incl.
From there the frame of three heptad leaves, with 1325, 2 House
its 90° angles, projects to the end of the Aztec
migration in 1325 AD (cf Plate 3 and note 14).

207
 Plate 8 Tilantongo Annals p.14

cave-mouth Rounds 3+
star-eye Rounds 8+
7 3

lady
Three
Flint 7 Reed
343 AD

AD 12 1 Flint (C10)
The year-date before lady TI-lree Flint is 7 Reed (ClO)
J4J, which is measured from the Round 60 base by the )12 6 X 52
324 1 Fl int
surrounding six cave-mouths; this marks the start of
the Classic. The distance to the end of that period 34 3 7 Reed
is given in the band of 15 star-eyes (Round 75).

208
 Plate 9 Aca tlan Genealogy p. 20

~ r
•f
. ' . '
I

,.,, \

i(~ ~

.
.;-..w_~iJJ
,,~ n
/J ;- •.
)
I •
...
, ... -_,;4, ~, .
.
;'~ ....

... t .,,,,, , ..... "

5 12 Flint
4 7
1388 AD
3 2
?5
5 + 16 + 5 = 26 Rounds

The key afld fin~l date in the text, 12 Flint (CtO) AD 12 1 Flint (C10)
1388 AD. ritu~Uy attaches to the 16th generation~ 520 10 X 52
16 pa~s- and 16 Rounds from the introductory time- 532
marker which records 5 + 5 Rounds from the Round 60
base and anticipates the 16 Rounds to come (p.4). 832 16- X 52
Here, these Rounds appear as breath scrolls uttered 1364 1 Flint
by the two CQUples, while the flames issuing from 1388 t2 Flint
the temples behind them make up the overall 26-Rotmd
distance fr0111- ba,e.

209
 Plate 10 Tepexic Annals p. 14

5 XVIII 5 XVIII 8 XVIII 7 III

55.5 248 54.5 300 50.21 492 49.46 519

1 XIII 7 XIII 8 XIII 6 XVIII

55.40 213 53.20 337 51.8 453 49.45 520

7 XIII 1 XIII 7 XIII 7 XIII

56.20 181 52.40 369 51.20 441 49.20 545
(Round. year) (BC)
1 XIII
51.40 421
V
Following a boustrophedon line from 545 BC bottom right to 181 BC bottom' left (13 dates
and 7 Rounds in all), events in the Era are attached to years and to toponym symbols;
note the four snow mountains 520 BC which in fact are uniquely visible from Cholula
(cf. Fig.ta). Top left the date 5 Flint 5 Flint 248 ' BC stands exactly 55 Rounds from
the first year named in the text (5 Flint 5 Flint 3108 BC p.3); this distance is con-
firmed by the temples of the Eagle (i.e. 50) and the stars (5) to either side of the
t7 Flower' tribute house, which confirms· the year 5 Flint as the fifth of the Round.

210
 Plate 11 Selden Roll opening

One Deer One Deer

13 Rabbit 2974 BC
8
8
8
8
8
8
9
8
= 65 Rounds 1 Reed 307 AD
Between sun (left) and moon (right) the descent from
the sky-womb passes via 65 star-eye Rounds, from 13
Rabbit 2974 BC to 1 Reed 307 AD. To either side of BC 3113 13 Reed (C1)
the Wind-masked figure in the sky (named Nine Wind (2974 13 Rabbit)
in the Tepexic Annals) sits the ancestral pair of 3380 65 X 52
the Anahuac tradition, female and male One Deer, AD 267 13 Reed
the first pair named in the Tepexic Annals (cf. 307 l Reed
Fig. Ja).

211
References

a) Native sources

Numbers in parentheses refer to the Census of Native Middle American

Manuscripts (1975) by John B. Glass (native script, 1-699), Donald
Robertson (Techialoyan, 700-999) and Charles Gibson and John B.
Glass (alphabetic, 1000- ), which gives full bibliographical
details. In all cases the revised paginations follow Corona Nunez
(1964-67), Laud being the only exception. Bibliographical
references supplement those in the Census. MNA = Museo Nacional de
Antropologia, Mexico.

Acatlan Genealogy (279: Sanchez Solis, Egerton 2895); pp. 1-32

for 2-33. See Konig 1979

Acuecuexatl Stone; MNA

Aubin ms 20 or Anahuac Map (14)

Aubin ms of 1576 (13, 1014)

Azcatlitlan Annals (20)

Aztec Priests' Speech. See Lehmann 1949

Baranda (24)

Becker (27)

Bodley (31)

Borbonicus (32). See Nowotny 1974.

Borgia (33). See Nowotny 1976.

Boturini (34)

Castillo, Cristobal de: Fragmentos

Chilam Balam, Books of: see Chumayel, Tizimin.

Chimalpahin: Relaciones and Memorial breve.

Chumayel Book (60, 1146)

Coixtlahuaca Maps i and ii (70, 71)

Cospi 79

Cuauhtinchan History(359, 1129; Historia Tolteca-Chichimeca).

212
Cuauhtinchan Map (95)

Cuauhtitlan Annals (1033)

Cuicatlan History (255; Porfirio Diaz)

Dresden (113). See Thompson 1972.

Eight Deer Batchelor (240 verso; Zouche Nuttall); pp. 1-44 for 42-
84.

Eight Deer in Tututepec (27; Becker Ms. Also 72; Colombino)

Florentine Codex (274, 1104; Sahagun).

Fejervary (118); pp. 1-22, 23-44 for 44-23, 22-1. See Burland
1971.

Gomez de Orozco fragment (129)

Guevea Maps i and ii (130)

Historia de los mexicanos por sus pinturas (1060)

Huichapan Annals (142, 1042). See Al var ado 197 6.

Huitzilopochtli histories (270; Rios. Shorter version in 308;

Telleriano Remensis)

Ihuitlan Map (157)

Ixtlilxochitl: Relaciones (1043)

Laud (185; pp. 2-22, 23-46 for 46-25, 24-2)

Madrid (187)

Mani Book ( 1149)

Mendoza (196, 1053)

Metlatoyuca Map (199)

Mexican Annals (201)

Mexicanus (207). See Prem 1978.

Mimiahuapan Land-book (711)

Moctezuma Stone; MNA

Museo de America (229). See Wilkerson 1974.

Nativitas Map (232)

Nepopualco Maps (46). See Prem 1972.

213
Palenque Trilogy; MNA and in situ.

Paris (247)

Popul vuh (1179)

Quinatzin Map (263)

Rios (270). See Anders 1979.

Sahagun Tlatelolco Ms. (1099)

Santa Cecilia Tlaloc, Tenayuca, in situ

Selden (283)

Selden Roll (284)

Sigllenza Map (290)

Suns tone; MNA

Telleriano Remensis (308)

Teozacoalco Map. See Caso 1949.

Tepexic Annals (395; Vienna obverse); pp. 1-52 for 52-1.

Tepechpan Annals (317)

Tezozomoc: Cr6nica mexicayotl (1062)

Tilantongo Annals (240; Zouche-Nuttall)

Tilantongo Genealogy (395; Vienna reverse)

Tizayuca Land-book (728)

Tizimin Book (1157). See Edmondson 1982.

Tizoc Stone; MNA

Tlalpitepec Map (8; Antonio de Leon)

Tlapa Annals by sevens (21; Azoyu 1 obverse)

Tlapa Annals by eights (22; Azoyu 2 obverse)

Tlapa Genealogy (21; Azoyu 1 reverse)

Tlapa Tribute-book (22; Azoyu 2 reverse. Also Humboldt Fragment 1)

Tlateloco Annals (344)

Tula Annals (369, 1073). See Zantwijk 1978.

214
Tula caryatid; MNA. See Navarrete and Crespo 1971.

Vaticanus B (384). See Anders 1972.

Vindobonensis or Vienna (395). See Furst 1978; Melgarejo Vivanco

1980.

Xochicalco New Fire stone; Cuernavaca Museum.

Xolotl Maps (412)

Yanhuitlan Tribute-book (415)

Yolotepec Map (419)

Zacatepec Map (422)

Zacatlan History; in Torquemada, Monarchia indiana (1130), Book

III, ch. 18.

Zouche Nuttall (240; Nuttall)

b) Secondary works

Alvarado Guinchard, Manuel

El c6dice de Huichapan. Mexico (INAH). (1976)

Anders, Ferdinand

(ed.) Codex Vaticanus 3773. Graz. ( 19 72)

(ed.) Codex Vaticanus 3738 (or Rios). Graz. (1979)

Aveni, Anthony F.

(ed.) Archeoastronomy in the New World; Proceedings of the

Oxford Conference September ----r§'"81. London, New York: CUP.
(1982)

Borunda, Ignacio

Clave general de jeroglificos americanos. Rome. (1898)

Broda de Casas, J.

The Mexican Calendar as compared to other Mesoamerican Systems.

Vienna. (1969)

Brotherston, Gordon

Image of the New World: the American Continent portrayed in

Native Texts. London and New York. (1979)

215
 !_ Key~ the Mesoamerican Reckoning~ Time: the chronology
recorded in native texts. Occasional Paper of the British
Museum 38.- London. (1982)

"Astronomical Norms in Mesoamerican Time-Reckoning", in Aveni

(1982), pp. 109-42. (1982a)

"The Time-Record of America", in The Other America: native

artifacts from the New World, ed. V. Fraser and G.
Brotherston.~lchesterand London. (1982b)

"The Case of Tula: its script and history", in Voices of the

First Americans, ed G. Brotherston. Santa Barbara. (1983)

"The Time Remembered in the Winter Counts and the Walam Olum",
in Indians and Americans, ed. J. Fear. Washington. (1983a)

Burland, Cottie

(ed.) Codex Egerton 2895. Graz. (1965)

(ed.) Codex Fejervary-Mayer. Graz. (1971)

Caso, Alfonso

"El Mapa de Teozacoalco". Cuadernos Americanos, 47 (5),

145-81. (1949)

Interpretaci6n del C6dice Gomez de Orozco. Mexico. (1954)

"Comentario al C6 dice Baranda", in Miscelanea Paul Rivet

Octogenario dicata. Mexico, 1, pp. 373-89. (1958)

Interpretation of the Codex Bodley 2858. Mexico. (1960)

"Los lienzos mixtecos de Ihui tlan y Antonio de Leon", in

Homenaje a Pablo Martinez del Rio. Mexico, pp. 237-74.
(1961)

"El lienzo de Filadelfia", in Homenaje a Fernando Marquez

Miranda, pp. 138-44. Madrid. (1964)

Los calendarios prehispanicos. Mexico (UNAM). (1967)

Caso, Alfonso, and Mary Elizabeth Smith

Interpretaci6n del C6dice Colombino. Mexico. (1966)

Castellanos, Abraham

"La cronolog ia indiana". Anal es del Museo Nacion.al (Mexico),

ep. 3, 3 ( 8), 453-84.

Chavera, Alfredo

AntigUedades mexicanas publicadas por la Junta Colombina.

216
 2 vols. Mexico. (1892)

Corona Nuffez, Jose

Antiglledades de Mexico publicadas en la recopilacion de Lord

Kingsborough. 4 vols. Mexico. (1964-67)

Dahlgren de Jordan, Barbra

La Mixteca: su cultura e historia prehispanicas. Mexico.

(1954)

Davies, Nigel

The Toltecs until the Fall of Tula. Norman. (1977)

Dibble, Charles E.

Cadice Xolotl. Mexico. (1951)

Edmonson, Munro

The Book of Counsel: the Popol vuh of the Quiche Maya.

Tulan-;:-- ITT 71)

The Ancient Future of the Itza. The Book of Chilam Bamal of

Tizimin ( translatedand annotated~ Austin. ( 1982)

Fabrega, Jose Lino

"Interpretaci6n del C6dice Borgiano", Anal es del Museo

Nacional de Mexico (1st series), 5, 1-260.

Furst, Jill Leslie

Vindobonensis mexicanus I: A Commentary. University of New

York at Albany, Institute for Mesoamerican Studies. (1978)

Garcia, Gregorio

Origen de los indios de el Nuevo Mundo. Madrid. (1729)

Gemelli Careri, G.F.

Giro del Mondo. Vol. 6. Naples. (1700)

Glass, John B., Donald Robertson and Charles Gibson

"Census of Native American Manuscripts", in Handbook of Middle

American Indians. Austin, 14, pp. 81-252, 253-80;-15, pp.
322-99. (1975)

Hammond, Norman

Mesoamerican Archaeology: New Approaches. London and Austin.

(1974)

217
 Ancient Maya Civilization. Cambridge. (1982)

Hammond, Norman, and Gordon Willey

Maya Archaeology and Ethnohistory. Austin and London.

(1979)

Humboldt, Alexander von

Vues des cordill~res et monuments des peuples indigenes de

l'Anletique. Paris. (1810)

Ibarra Grasso, D.E.

La verdadera interpretaci6n del calendario azteca. Buenos

Aires. (1978)

Jansen, Maarten E.R.G.N.

"The Four Quarters of the Mixtec World". Lateinamerika

Studien, 10 (ed. F. Tichy), Munich, 85-96.

Jimenez Moreno, Wigberto, and Salvador Mateos Higuera

Cadice de Yanhuitlan. Mexico. (1940)

Kelley, David Humiston

Deciphering the Maya Script. Austin and London. (1976)

K8nig, Viola

lnhaltliche Analyse und Interpretation von Codex Egerton.

Hamburg. (1979)

Lehmann, Walter

Sterbende G8tter und christliche Heilsbotschaft. Stuttgart.

(1949)

Die Geschichte der ~onigreiche von Colhuacan und Mexico. 2nd

ed. Stuttgart. (1974)

Melgarejo Vivanco, Jose Luis

El c6dice vindobonensis. Xalapa. (1980)

Molina, Fray Alonso de

Vocabulario en lengua ~astellana y mexicana (1571). Madrid.

(1944)

Navarrete, Carlos, and Ana Maria Crespo

"Un atlante mexica", Estudios de cultura nahuatl, 9, 11-16.

(1971)

218
Nicholson, H.B.

"Western Mesoamerica: AD 900-1520", in Chronologies in New

World Archaeology. New York. pp. 285-325. (1978)

North, John D.

"Chronology and the Age of the World", in Yourgrau, pp. 307-

34. (1977)

Nowotny, Karl Anton

"Erlauterungen zum Codex Vindobonensis (vorderseite)". Archiv

fUr V8lkerkunde (Vienna), 3, 156-200. (1948)

Tlacuilolli. Die mexikanischen Bilderhandschriften. Stil und

Inhalt. Berlin:- (1961)

(ed.) Codex Borbonicus. Graz. (1974)

(ed.) Codex Borgia. Graz. (1976)

Prem, Hanns J.

"The 'Map of Chichimec History' Identified", in Acts of the

40th Congress of Americanists. Genoa, pp. 447-52.--(1972_)_

"Comentario a las partes calendricas del Codex Mexicanus 23-

24". Estudios de Cultura Nahuatl, 13, 267-88. (1978)

Sahagun, Fray Bernadino de

Florentine Codex: General History of the things of New Spain.

Ed& Charles Dibble and A.J.O. Anderson:- 11 vols. Santa Fe.
(1950-69)

Santillana, Giorgio de, and Hertha von Dechend

Hamlet's Mill: an essay on myth and the frame of time.

London. (1970)

Smith, Mary Elizabeth

Picture Writing from Ancient Oaxaca. Norman. (1973)

Thompson, J. Eric S.

Maya Hieroglyphic Writing. 3rd ed. Norman. (1971)

! Commentary on the Dresden Codex. Philadelphia. ( 1972)

Tichy, Franz

"Der Fes tkalend,er Sahagun' s: ein ech ter Sonnenkalende r?".

Lateinamerika Studien, 6, 115-37. (1980)

219
Toscano, Salvador

"Los codices tlapanecas de Azoyu". Cuadernos Americanos, ano

2, 10 (4), 127-36.

Van der Loo, Peter L.

"Rituales con manojos contados en el grupo Borgia y entre los

tlapanecos de hoy dia", in Coloquio Internacional: Los
indigenas de Mexico e n ~ epoca prehispanica z en la
actualidad, ed. M. Jansen and Th. Leyenaar. Leiden:
Rijksmuseum voor Volkenkunde, pp. 232-43. (1982)

Wilkerson, S. Jeffrey K.

"The Ethnographic Works of Andres de Olmos", in Edmonson, 27-

78. (1974)

Willey, Gordon

"The Classic Maya Hiatus: a 'rehearsal' for the collapse?",

in Hammond, 417-30. (1974)

Yourgrau, Wolfgang, and Allen D. Breck

Cosmology, History and Theology. London and New York.

(1977)

Zantwijk, Rudolf A.N.

(ed.) Anales de Tula. Graz. (1979)

220
GARY URTON and ANTHONY F. AVENI

Archaeoastronomical fieldwork on the coast of Peru

The archaeological reco~d of monumental architecture in

Pre-Columbian Peru testifies to a remarkably high level of activity
within the numerous river valleys which cut from the east to the
west across the dry strip of coastal desert. This activity dates
from several millenia B.C. until just before the time of the Spanish
conquest. The exposed ruins are now in varying states of decay, and
any information which can still be gleaned from the record is of
utmost importance in providing relatively accurate data for present
and future studies. In that light, the purpose of this report is
two-fold: first, to present an account of a series of precise
alignment readings, made with a surveyor's transit, of orientations
at 14 archaeological sites from the Lambayeque valley in the north
to the Rio Grande de Nazca in the south; secondly, to present these
data in the context of an argument, built on ethnoastronomical and
archaeoastronomical observations derived from the Peruvian coast and
highlands, for what appears to be a consistent interest in aligning
structures to the rise or set of a few specific celestial phenomena.

Table 1 gives a descriptive analysis of each alignment taken by

the authors. Descriptions of the specific features measured and any
additional explanatory comments are provided in the Appendix. Due
to the effects of precession of the equinoxes, the azimuths of
orientation to the rise or set of stars will change through time
(the effect is negligible for the sun and moon). Therefore, we have
provided approximate dates for the sites as determined from a review
of the archaeological literature. The table also includes the
azimuths of rise or set of the astronomical phenomena which, we will
suggest, may have been important in the orientation of these
structures.

At the beginning of our interpretation of the data collected 1n

the field, we were aware that it would be possible to attempt to
match site orientations with the rise or ~et of celestial bodies by
simply scanning computer charts containing the appropriate latitude,
time period and horizon elevation for each site. Given the number
of possibilities, however, one could be relatively certain of
finding an astronomical match for each orientation, given a standard
allowance of 1°-2°. Therefore, we were primarily interested in
the degree to which the site orientations would, or would not,
confirm a hypothesis for those alignments which should exist. After
a statement of our hypothesis, we will conclude with~ discussion of
how well it is supported by the evidence.

The Hypothesis

The study of archaeoastronomy and ethnoastronomy in Peru has

advanced most rapidly in the southern highlands, in the Dept. of
Cuzco. These studies, dealing with the astronomy of Inca, Colonial

221
and contemporary populations, have made it clear that a relatively
small number of celestial phenomena were central to the astronomical
and calendrical systems which were, and still are, adapted to this
region. As space does not allow a detailed account of how and why
each of the phenomena discussed below are -important, we will refer
the reader to a number of references (Aveni 1981; Urton 1981; and
Zuidema 1977 & 1981) and only mention the principal phenomena here.
They are:-

a) The rise and set of the Pleiades;

b) The sunrise and sunset on the day of passage of sun
through zenith;
c) The sunrise and suniet on the day of passage of sun
through anti-zenith.

These phenomena, important primarily because they relate to

critical periods in the agricultural and ritual cycles, were of
utmost importance in Incaic astronomy and calendrics, and they are
still employed usefully for precise astronomical observation among
contemporary Quechua-speaking populations in the Dept. of Cuzco.

The majority of alignment readings in Table 1 pertain to coastal

rather than highland sites; the one exception is the central
highland site of Chavin de Huantar. Therefore, it is significant to
find that, in addition to data supporting the central role of the
Pleiades and the zenith and anti-zenith sun from the highlands, the
ethnographic and ethnohistoric documents from the coast also attest
to the importance of these same phenomena. There are several
accounts referring to the role of the Pleiades in the astronomy and
calendars of the coast, beginning with the 17th century account of
Antonio de la Calancha:

They (the Chimu) do not count the year by Moons or by the

course of the Sun, but rather from the rise of the stars
which we call the Cabrillas (the Pleiades) and which they
call Fur. The reason for this is found in a long fable, which
is none of my concern. It was a law that they counted the year
thusly, because these stars gave them food and nurtured their
crops, for their livelihood, therefore, they had to begin the
year from the time they saw it appear and it gave them
sustenance. (Calancha, 1638:554).

Calancha's account relates to the Pacasmayo valley, on the

north coast, where the Indians were Yunga-speakers and descendants
of the Pre-Columbian Chimu culture. The term given by Calancha for
the Pleiades, Fur, is also recorded in the 17th century Yunga
dictionary of Carrera ((1644)1880). In a discussion of the
astronomical phenomena observed in the community of Huarochiri,
Avila (1966, Ch.29) also mentions the cabrillas (Pleiades) and
describes how they were observed in agricultural prognostications.
From the same period ( 1656), we have a document on "idolatries" from
San Pedro de Acas giving the testimony of one Hernando Hacas Poma in
which the Pleiades are referred to by the Quechua term oncoicoillor
("sickness star"). San Pedro de Acas is located near Cajatambo,
some 80 km inland on the central coast; the Pleiades were observed
here in the 17th century to time two annual festivals which were

222
attended by people from several coastal valleys; one festival in
November, Pocoimita, could have been timed by the heliacal set of
the Pleiades on November 18 while the other festival, Caruamita, was
held on Corpus and San Juan when, as the document says, the Pleiades
appeared and the frosts began (Huertas Vallejos, 1981: 50-53, 71,
105,106 and 115). In addition to these ethnohistoric references,
Gillin's ethnography of the communities of Moche and Huanchaco on
the north coast refers to the role of the Pleiades (las cabrillas)
in coastal navigation and time-keeping (Gillin, 1947:34). Zuidema
(nd.) gives additional early references to the Pleiades.

Documentation from the coast concerning the importance of the

zenith and anti-zenith passages of the sun is not as direct as that
for the Pleiades. It is possible that the zenith sun of Feb. 23 was
used to time the festival of Yue Yue in San Pedro de Acas; this
festival was related to the end of the rainy season and the plowing
(barbecho) of the potato fields. The other zenith passage on Oct.
23 could have been used to time one of four "solar" festivals held
in October in the nearby community of San Juan de Tulpay (Huertas
Vallejos, 1981: 52 and 71). October was also the month of
celebration of the one annual lunar festival in San Juan de Tulpay;
this month may therefore have been the time for correlating the
solar and lunar calendars. Similarly, the festival of oncoyllocsiti,
held in August in San Pedro de Acas, may have been timed by the
August 27 anti-zenith passage of the sun; the same anti-zenith
passage may have been used to time another of the four "solar"
festivals in San Juan de Tulpay, a solar festival said to have been
celebrated in August (Huertas Vallejos, 1981: 53 and 71).

It has also been argued, on the basis of the ethnographic

record, that the zenith and anti-zenith passages of the sun have
been important for some time in the agricultural and fishing
calendars of three north coastal communities, Santiago de Cao in the
Chicama valley and Moche and Huanchaco in the Moche valley (see
Sabogal Wiesse, 1975; Gillin, 1947 and the analysis of these data in
Urton, 1982).

On the basis of these data from the coast, as well as the

comparative material from the highlands, the hypothesis which we
wished to test with our field data was as follows: if astronomical
phenomena played any sort of role in the orientation of coastal
sites, then those events we should expect to find represented in the
alignments would be the rise or set points of the Pleiades and the
zenith and anti-zenith passages of the sun. It should be emphasised
that we do not necessarily assume that the sites were, in fact,
astronomically aligned; the supposition which qualifies our
hypotheses allows for the possibility that sites may have been
oriented in a particular direction for any number of reasons (e.g.
topography, the desire to take advantage of cool ocean breezes,
respect for sacred mountains, etc.). To re-emphasise a point made
earlier, the primary function of our report is to make the data in
Tabl~ 1 available, while a secondary purpose is to test one
particular hypothesis, out of several potential hypotheses, for
explaining why the particular orientations might have been chosen.

223
l'u
l'u
~

210- - - - - .... -- - - - - - -- - - - -- - - - - - - -- - - - - - -90

Fig. 1 Distribution in azimuth of alignments of structures

at sites surveyeQ along the coast of Peru
 Conclusions

That the documentation for the astronomical and calendrical

importance of the Pleiades is greater than that for either the
zenith or the anti-zenith solar passages is also reflected in the
alignment readings. This is especially true on the north coast, the
territory of the Chimu, who, according to Calancha, counted their
year by the Pleiades. The correlations between site orientations
and the azimuths of rise and/or set of the Pleiades, and the amount
of deviation between the two, can be summarised as follows: Batan
Grande (about 1° in 500 A.D.); Pacatnamu (about 2° in lOOOA.D.);
Huaca de la Luna (less than 1° in 200 A.D.); Ciudadela Tschudi
(less than 2° in 1400 A.D.); Ciudadela Bandelier (less than 1°
in 1350 A.D.); and possibly Chavin (here the errors are somewhat
larger owing to a high, variable horizon.). The correlations, given
the particular dates, are quite close.

There are fewer correlations between site orientations and

either the zenit h or anti -z enith passages of the sun, but the
existing correlations have a very small deviation. For either the
0
zenith or the anti-zenith, the correlations are: C. Sechin (½ ) ,
0 0
P a r am o n g a ( 1 e s s t h a n 3 ) , T am b o Co 1 o r a d o ( 1 e s s than ½ ), and
Cahuachi (about 1°). The orientation out the doorway of the Huaca
de Loro includes a range of azimuths from 58° - 79°03' which
would have allowed a view of the rising of both the Pleiades and the
anti-zenith sun.

Figure 1 displays the general distribution of alignments with

respect to the clockwise or counter-clockwise skewing from the
cardinal directions for all sites reported in Table 1. Though we
can provide no causal argument at this point in our studies, it
should be noted that in most cases the orientations are skewed
clockwise from the cardinal directions, the same pattern which has
been found in similar site orientation surveys in Mesoamerica (see
Aveni, 1980, Ch.5 for details). This result is in general agreement
with the results given in the only other coastal alignment survey,
that of Williams (1978) on U-shaped structures. It is perhaps of
some significance for the Peruvian case that many of the clockwise
skewed orientations in Mesoamerica can be accounted for by employing
astronomical hypotheses.

We conclude from previous observations that the data tend to

support the part of our hypothesis concerning alignments to the
Pleiades, especially for sites on the north coast. While the number
of sites having zenith/anti-zenith alignments is not large, it is
interesting to note that most sites which exhibit these orientations
are located either in the south or were influenced by the Incas who
had their capital in the southern highlands. These differences
between alignments in the north and south may be more apparent than
real (e.g. perhaps they are only a product of sampling), since we
have good evidence that both Pleiades and zenith/anti-zenith
alignments were incorporated in the Inca ceremonial architecture
(Aveni, 1981; and Zuidema, 1981 and 1982) and in the agricultural
and ritual calendars of the north coast (Urton, 1982).

We have suggested earlier that site orientations may have been

225
 sm31~F c!1 r□~
11

~~~~ ,v -- -
---= =,~
1 f§L
- - - ~ \ -\- - r l -
X

---- \
\
\
.-.. _, .
....__--
\ \ ,--
\
\
-----,
\
\,-" -- ----
\\
->---
_,, \ ----2"
\
tv \
tv L-____,

r
0\

•
. 1 lQ

\~ o 10 20 so M

Fig. 2 Tambo ·Colorad o (Plan after Harth-T erre (1938) reproduc ed

in Gaspar ini and Margol ies (1980) showing possibl e
geomet rical and astronom ical influenc e in the arrangem ent
of building s (dotted lines added by the authors) .
selected according to any number of astronomical or non-astronomical
considerations. In this regard, it is of interest to discuss
briefly the example provided by one site reported in Table 1: Tambo
Colorado. Tambo Colorado was an Inca coastal site which
incorporated at least one feature, an irregularly-shaped central
plaza, which was not uncommon in other coastal and highland Inca
sites (Fig.2). For instance, as Zuidema (p.c.) had suggested,the
central plaza in the Inca capital city of Cuzco was composed of two
adjacent plazas (divided by the Saphi River), which together
exhibited an irregular shape not unlike that of Tambo Colorado. It
has been shown that the arrangement of certain buildings within the
central plaza of Cuzco was at least partially determined by an
interest in orientation to the points on the horizon where the
zenith sun rose and the anti-zenith sun set (Aveni, 1981; and
Zuidema, 1981). At Tambo Colorado, the buildings along the northern
border of the large central plaza called the "esplanade"
0
(Harth-Terre, 1938) are aligned within ½ of the orientation to
the anti-zenith sunrise and the zenith sunset. While these
astronomical orientations may have been important in the alignment
of the buildings on the northern border of the plaza, it is possible
that certain geometrical principles that yet remain to be discovered
played a role in determining the overall deposition of buildings
forming the four borders of the esplanade.

One curiosity about Tambo Colorado is that the quadrangular

esplanade does not exhibit a single right angle. The absence of
right angles, however, also may be more apparent than real if we
consider the inter-relationship of other features among the borders
of the plaza such as the altar which is inset on the west (Fig.2,Z),
and the doorways of four buildings (Fig. 2B, C, G, H) which open
onto the esplanade. The orientation of the altar (Z), in Fig.2,
calculated on the basis of the two transit measurements taken on the
facades of the northern and southern buildings of the esplanade, is
0
54½. The eastward extension of this line of orientation passes
less than 1° (or a 2½ m. deviation over 155 m) to the north of
the opening on the eastern side of the explanade (Fig. 2,Y). The
north-eastern corner appears to have been the point where the old
Inca road entered the esplanade.

If a line is extended between the doorways of buildings C and

G, we find that the intersection of CG and ZY produces an angle of
0
88½ ; furthermore, the point of intersection occurs in the middle
of the esplanade. It is also of interest that the line CG is within
0
2½ of parallel with the facade of the building which contains
the altar (Z). Finally, a line extended between the doorways of
buildings Band H lies within 3/4° of the orientation of line CG
and also crosses the line ZY at riearly a right angle. However,
Hyslop (p.c.) suggests to us that structure B may date from an
earlier period; therefore this line must be regarded as a bit more
problematic than the others.

One can draw only the most tentative of conclusions when

referring to a site map such as that reproduced in Fig.2;
nonetheless, by combining such evidence with independent data, like
the alignment readings provided in Table 1, it is possible to arrive
at a number of creative and testable hypotheses. From the example

227
of Tambo Colorado, it would seem important to investigate more
thoroughly the relationship between astronomical and non-astro
-nomical (e.g. geometrical) principles of orientation and
architectural planning in Inca and pre-Inca sites along the coast
and in the sierra. The possibility of arranging structures in an
apparently irregular fashion while incorporating a set of rigorous,
underlying geometric principles, (e.g. describing right angles and
parallel lines in the open spaces between structures) may not have
been an uncommon architectural device in the Pre-Columbian world.
This is suggested by the fact that the organisation of buildings at
Tambo Colorado, and the "interstitial" geometry which results from
that organisation, is very similar to that found in the famous Maya
Nunnery at Uxmal (Aveni and Hartung, 1982).

We thank R. Feldman for providing us with more recent dates on

the structures.

228
 Table 1

!SE/SET AZIMUTH
RISE/SET AZIMU1ll. F SUN ON DAY OF:
LATITUDE HORIZON ALIGNM. ~ECIPROCAL(R) OR OF PLEIADES FOR ZENl1li NTI- ZENITH
SITE 1 (SOUTH) ATE ELEVATION AZIMU1li PERPENDICULAR ( P) ALIGNMENT GIVEN DATE 2 PASSAGE PASSAGE 2
--------
la Batan Grande 500A.D. 1° 11°26 1 107°26'-287°26'(P) 288°32'
lb 17°30' 107°30'-287°30'(P) 288°32 1
lOOOA.D. 11°26 1 107°26'-287°26'(P) 290°44'
17°30' 107°30'-287o30'(P) 290°44'
/

Pacatnamu 500A . D. 1° 202°40 1 112°40'-292o40'{P) 288°33'

lOOOA.D. II 290°45'
3a Huaca del Sol 400A.D. o0 296 °34 I 116°34'{hor.•11°){R)
3b 110 24°38 1 204°38'(hor,z0°)(R)
4a Huaca de la Luna ZOOA.D . o0 286°15' 106°15'(hor.•19°){R)
4b 10° 26°05' 206°05'(hor.•1°)(R)
Sa Huaca el Dragon 8°05' 1300A.D. 1° 289°12 1 109°12'(hor.•3°l{R)
0
Sb 16°43' 196°43'(hor.~o )(R)
Sc 287°57' 107°57'{hor.=6°)(R)
lOOOA . D. II
290°38'
0
---
6a 3
Huaca La Esmeral!a 8°05' 500A .D. o 284°50' 288°26'
6b 285°50 1 288°26'
lOOOA.D. II
290°38'
7a Ciu. Tschudi 8°06' 1400A.D. 0° 110°43' 290°43(R) 292 °20~
7b 201°12 1 21°12 1 (R)
7c 16°51 1 196°5t'(R)
1500A.D. II

8 C. Bandelier 8°06' 350 A.O. o0 22°10 1 112°10 1 -292°10 1 (P)

9 C.sechin ·sooa .r; . 2° 278°48' 98°48'(R)
10a Chavin 500B.C. 3° 13°16 1 193~16(R)
10b 30 13°29' 193°29' (R)
10c 23° 293°28 1 103°28' (R) ~9°08'
10d 23° 283°35' 103°35' (R) 289°08'
0
lla Paramonga 10°40' lOOOA.D. o 256°49' 76° 49' (R)
II
llb 1500A.D.
12a Tambo Colorado 14°26' lSOOA . D. 2° ~ 255°52' (R) ~
12b 60°39 1 240°39 1 (R)
13a Cahuachi 2°16 1 357°44' (R)
13b 103°35' 283°35' (R) 104°15
105 08'
13c 91° 271°(R)
13d 7°55 1 352°5' (R)
14 Huaca del Loro SOOA.D. 3° 79°03' 58°-79° 1 3(range)68°30'(ce nter} 70°15'

NOTES TO TABLE
1) Descriptions of site locations, the sped fie features measured and any additional co-nts are provided in the Appendix .
2) All astronomical matches were derived fro111 the tables of (Aveni, 1972). They are corrected for horizon elevation. General
agreement between azumth of alignment and astronomical event is indicated by underlining the azimuths.
3) Readings at Huaca La Esmeralda were made with a hand-held magnetic compass. The •agnetic readings were then correct e d
by a correction factor (+3°20') deten1ined by comparing magnetic and true azimuths detendned on the HIiie feature at
several different sites near Huaca la Es-ralda in the Hoche valley .
4) Unlike the case for stellar align11enta, the azimuth of sunrise and sunset remains relatively constant over
long periods of time.

229
 APPENDIX

The number entries in this appendix coincide with the numbers

to the far left in table 1.

la Wall 1, East Side, Las Ventanas Pyramid. We are

grateful to Izumi Shimada of Princeton University
for permitting us access to the site and for advising
us on where to make measurements.

lb Wall 1, East Side, Las Ventanas Pyramid.

2 West wall of unidentified building giving typical site

orientation; looking south.

3a Exposed row of adobe blocks along the southern edge of

the upper platform.

3b Exposed row of adobe blocks, roughly perpendicular to

3a, along the upper platform.

4a Adobe foundation blocks of a structure on the upper

platform of the large westernmost section of the Huaca • .

4b Adobe foundation blocks of a wall perpendicular to the

wall described in 4a, and part of the same structure.

Sa The north exterior compound wall.

Sb The west exterior compound wall.

Sc The north wall of the interior platform.

6a Edge of north wall of the platform.

6b The east-west "passageway" dividing the platform of

Huaca La Esmeralda into two sections.

7a The western section of the north exterior compound wall

(for maps of Cuidadela Tschudi and Ciudadela Bandelier,
see Moseley and Mackey, 1974).

7b The west exterior compound wall.

7c The west exterior wall of the "Gran Plaza

Ceremonial"; the first major plaza upon entering
the cuidadela.

8 The northwest exterior compound wall.

9 The elaborately carved, north exterior wall (see

Tello, 1956).

10a Northward along the eastern side of the Block and

230
 White Portal (see tht n..ip ()f r;havin de Huantar 1n
Rowe , 1962 ) •

10b The east wall of the large sunken square courtyard

to the east of the New Temple.

10c The west wall of the large sunken square courtyard

to the east of the New Temple.

10d Westward up the stairway leading from the sunken

circular courtyard to the Old -temple (see map in Llosa
1972).

11 South wall of structure, uppermost portion; looking

east.

12a The south wall of the central building bordering the

main plaza on the north; looking east (see map in
Harthe-Terre, 1973).

12b The north wall of the central building bordering the

main plaza on the southeast; looking east.

13a Exposed stucco-covered rubble wall along the eastern

edge of a large mound in the northwestern quadrant of
the site (see map by W. Strong, 1957).

13b The north face of the central platform of a platform

mound located north of the "Great Temple" (Strong map).

13c Foundation along the northern edge of a lower platform

of the "Great Temple".

13d Foundation along the western edge of a lower platform

of the "Great Temple".

14 From the centre of the Round Temple to the southern

edge of the doorway (see Paulsen, n.d.: Fig.3).

231
References

Aveni, Anthony F.

"Astronomical Tables Intended for Use in Astra-Archaeological

Studies". Am. Aq., 37(4), 531-40. (1972)

Skywatchers of Ancient Mexico. Austin: University of Texas

Press. ( 1980)

"Horizon Astronomy in Incaic Cuzco", in Archaeoastronomy in

the Americas, ed. Ray Williamson. Los Altos, CA: Ballena
Press, pp. 305-68. (1981)

Aveni, Anthony F., and Hartung, H.

"Precision in the Layout of Maya Architecture", in

Ethnoastronomy and Archaeoastronomy in the American Tropics,
ed. A. Aveni and G. Urton. Ann. New York Academy of Sciences,
385, 63-80. (1982)

Avila, Francisco de

Dioses y hombres de Huarochiri (1608). Tr. J.M. Arguedas.

Lima: Inst. de Estud. Peruanos. (1966)

Calancha, Antonio de la

Cor6nica moral izada del Ord en de San Agustin en el Peru.

Barcelona. (1638)

Carrera, Fernando de la

Arte de la lengua yunga de los valles Obispado de Trujillo

(1644Y:- Lima. (1880) - --

Gasparini, K. , and L. Margolies

Inca Architecture. Bloomington and London: Indiana Univ.

Press. (1980)

Gillin, John

"Meche: A Peruvi•an Coastal Community." Smithsonian

Institution, Institute of Social Anthropology, Publication
No. 3. (1947)

Harth-Terre, E.

"Tambo Colorado" (portion of an article by H.N. Urteaga).

Bol. de la Soc. Geog. de Lima. (1938)

232
Huertas Vallejo, Lorenzo

La re 1 i g i 6 n en u n a so c i e dad r u r a 1 and in a (Sig 1 o XV I I).

Ayacucho: Universidad Nacional de San Cristobal de Huamanga.
(1981)

Llosa F.

"Chavin Huaylas". Huella, 10. Lima. (19 72)

Moseley, Michael E., and Carol J. Mackey

Twenty-four Architectural Plans of Chan Chan, Peru.

Harvard University, Cambridge, Mass:-Peabody Museum Press.
(1974)

Paulsen, Allison C.

"Huaca del Loro: The Transition to the Middle Horizon on the

South Coast of Peru". (nd.)

Rowe, John H.

Chavin Art: An Inquiry into Its Form and Meaning. New York:
Museum ofPrimitive Art.--(1962) _ _ __

Sabogal Weisse, J.R.

"El rnaiz en Santiago de Cao", in Chimor - una antologia sobre

~ Valle de Chicama, ed. J.R. Sabogal Weisse et al.
Ediciones Especiales, 73. Mexico City: Instituto
Indigenista Interamericano, pp. 81-131. (1975)

Strong, William Duncan

"Paracas, Nazca, and Tihuanacoid Cultural Relationships in

South Coastal Peru". Memoirs, Society for American
Archaeology, 13. (1957)

Tello, Julio C.

Arqueologia del Valle de Casma. Lima. ( 1956)

Urton, Gary

At the Crossroads of the Earth and the Sky: An Andean

Cosmology. Inst i tuteof7::°atin Arner ican'studies, Monograph
No. 55. Austin: University of Texas Press. (1981)

"Astronomy and Calendrics on the Coast of Peru", in

Ethnoastronomy and Archaeoastronomy in the American Tropics,
ed. Anthony F. Aveni and Gary Urton. Annals of the New York
Academy of Sciences, 385, 231-47. (1982)

233
Williams Leon, c.
"Complejos de piramides con planta de U". Revista del Museo
Nacional (Lima), 44, 95-110. (1978)

Zuidema, R. Tom

"The Inca Calendar", in Nativa American Astronomy, ed.

Anthony F. Aveni. Austin: University of Texas Press, pp.
219-59. (1977)

"Inca Observations of the Solar and L\lllar Passages Through the

Zenith and Anti-Zenith at Cuzco", in Archaeoastronomy in the
Americas, ed. Ray Williamson. Los Altos, CA: Ballena
Press, pp. 319-42. (1981)

"Catachillay:;. The Pleiades and the Calendar of the Incas", in

Ethnoastronomy and Archaeoastronomy in the American Tropics,
ed. Anthony F. Aveni and Gary Urton. -Ann. New York Academy of
Sciences, 385, 203-30. ( 1982)

"Las Pleyades y la organizacion politica andina", in Actas de

la Asociaci6n Peruana de Etnohistoria (Lima). (nd.)

234
R.T. ZUIDEMA

Towards a General Andean Star Calendar in Ancient Peru

In two recent articles (Zuidema 1982a, 1982b), discussing the

calendar of the Incas in Cuzco at the time of the Spanish conquest
(~ 1530), I concentrated in my analysis on two groups of
constellations. The first group consisted of the Pleiades,
considered as the mother of all other stars and constellations.
The other group was formed of four constellations. The first of
these was a dark-cloud constellation-- interstellar dust within the
Milky Way-- representing a llama; also called in the province of
Huarochiri in Central Peru, Yacana. ,The eyes of the llama, known
today as llamac nawin (Cuzco) or llamapa fiawin (Ayacucho) "eyes of
the Llama," were formed by the stars Alpha and Beta Centauri. The
Llama stretches toward the east behind these two stars. The third
constellation, called Yutu-- yutu, a tinamou bird or, in
present-day Spanish, a "partridge"-- was represented, then and
today, by the constellation known in Western astronomy as the
"coal sack." Finally, the fourth cons tel lat ion-- but here our
knowledge is less definitive for the sixteenth century-- is the
Southern Cross, within which that of the Yutu as a dark-cloud
constellation is found. I will refer to the Southern Cross by this
Western name and accept that its astronomical calendrical function
coincides with that of the Yutu.

The Incas established a correlation between the period of

visibility of the Pleiades and that of a period of 328 nights,
beginning with the first heliacal rise in the morning of the
Pleiades (9 June) and ending on 3 May, some two weeks after their
last heliacal set in the evening. The period of 328 nights,
1
although based on a sidereal-lunar calculation (12 x 27 /3 = 328),
was used, first of all, for measuring periods of unequal length in
between succeeding astronomical observations in which the Incas were
interested. These observations were, for instance, the heliacal
rises and sets of stars, the sunrises and sets during the
solstices, the sunrises when the sun goes through the zenith
(zenith sunrises on 30 October and 13 February), sunsets when the
sun goes through the nadir (antizenith sunsets on 18 August and 25
April) and periods within the year equal to either a synodic month
(29 or 30 nights) or a sidereal month (rounded off to 28 nights).
Another use, combined with the former, was, however, defining
longer or shorter calendrical periods in terms of the socio-
political hierarchy of Cuzco and its valley. The period of 37 days
from May 3 to June 9 was not measured by way of this calendar. One
contribution of this article will be to clarify the Andean interest
in this period; a period to which I have already given some
attention in a third article (Zuidema Ms.).

235
 The Incas, and probably people of Southern Peru and Northern
Bolivia in general, observed a spatial opposit ion in the sky of the
Pleiades, on the one hand, and - the group of the four
constellation s, on the other. Thus, while the llama
constellation or that of its eyes was known under one name as
catachillay, the Pleiades were called by one of their names,
catachillay huarahuara, "the stars catachillay." The calendrical
positions of both groups of constellations were defined more
precisely withi n t he 328-night calendar. The Incas recognised the
following periods belon ging to this calendar in relation to their
observations of the Pleiades, the Yutu and Llamapa nawin (Note 1):

CALENDAR ASTRONOMICAL OBSERVATION

3 September - 30 September Period of lower culmination of Yutu

(3 September - 29 September)

1 October - 30 October Period of lower culmination of

Llamapa nawin (7 October-2 November)

1 November - 21 Novembe r Period of upper culmination of the

Pleiades (5 November - 18 November)

Chroniclers inform us in general terms that, except for the

Pleiades which were worshipped by all people, other stars and
constellations were worshipped by specific social groups during
their period of upper cul mination. These groups could be either
kin groups, groups of a special social rank, or professional
groups. The example of the Southern Cross and Alpha and Beta
Centauri shows that stars or constellations could be worshipped also
during their times of lower culmination. The practical use of
combining the observa tions of upper and lower culminations in our
example was that the spatial opposition of the two groups o f
constellations was defin ed more sharply by way of a calendrical
contiguity between the lower culmination of the Southern Cross and
Alpha and Beta Centauri and the upper culmination of the Pleiades.

One question to be answered in this paper is the following.

Given the data on the relationship mentioned, how was the 6ther
relationship expressed, the one between the lower culmination of
the Pleiades and the upper culminations of the Southern Cross (Yutu)
and Alpha and Beta Centauri (Llamapa nawin)? Our data from Cuzco
and Huarochiri clarified Andean concepts concerning the period of
lower culmination of the Pleiades. In Cuzco, the Incas
established a calendrical connection between the heliacal rise in
the morning of the Pleiades (9 June) and the June solstice (21 June)
and a similar one between the heliacal set in the evening(+ 15
April) and antizenith sunset (25 April). In Huarochiri I argued
that a feast, celebrated three times a year, also was timed to the
Pleiades. The first feast, correlating the disappearance of the
Pleiades with antizenith sunset (22 April, at the latitude of
Huarochiri, 12° south of the Equator), celebrated the death and
resurrection of the Thunder god, called Yaro or Pariacaca. He was
addressed as "our father" and was supposed to live in the snow-
capped mountains of Pariacaca. The second feast, at the time of
the reappearance of the Pleiades, was in honor of a female deity,

236
Chaupifiamca, addressed as "our mother." She was identified with
Pachamama, "mother Earth." The t hird feast, in colonial ti mes
celebrated close to that of San Andres (30 November), probably was
timed to the first heliacal set in the evening of the Pleiades (18
November; or 21 November in the Inca calendar).

The new data to be analyzed in this paper seem to tell us more

about the upper culmination of the Southern Cross in March; it is
found in documents of a certain type-- those on the extirpation of
idolatries-- that are particularly rich for various villages at the
latitude of 11 ° south of the· Equator. Historical legends support
their cultural unity, the region stretching over al 1 the mountain
chains at this latitude. Their mythology centers on the
rejuvenation within five days of the Thunder god; an event probably
timed to the antizenith sunset at this latitude on 19 April.

With the analysis of these new data, I hope to contribute to

the solution of a specific problem concerning the simultaneous use
in Cuzco of a solar year count, of a count using synodic luna r
months and of the 328-night coun t. The problem becomes acute if we
study calendrical oppositions like those of the solstices and those
of the zenith and antizenith sunrises and sunsets. Each opposition
divided the solar year into two half-years. Other oppositio ns ,
like those of the constellations discussed above, could not divide
the year exactly into two half-years. All this information had to
be recorded, however, by a calendrical system, that of the 328-
night count, that occupied an irregular place within the solar
year. I will suggest in the third section in schematic form how
the Incas managed to keep a certain correspondence between the three
counts , a correspondence to be argued in the fourth and fi fth
sections.

The solar year count

Before doing this, however, I shall have to analyze one other

corpus of data from Cuzco, one that refers exclusively to the
calendrical count of the solar year. The early chronicler Polo de
Ondegardo· (Polo 1916: 18-24), who recorded also the 328-night
count, tells us that the Incas divided the solar year into 12
parts, accomplishing this by adding one or two days to certain
synodic lunar months. The consequence of such a procedure is, of
course, a complete detachment of this type of calendar from the
count of synodic lunar months of 29 or 30 days each. Two other
chroniclers have similar data. Betanzos (1968: ch. 18), in 1551,
refers to a year of 360 (12 x 30) days, plus 5 extra days. The
late indigenous chronicler, Felipe Guaman Poma de Ayala (1980: 260/
62), in+ 1615 , probably copying Polo, mentions months of 30
days, with weeks of 10 days, months to which one or two days could
be added. If such a calendar existed, next to the 328-night
count, then it was probably calculated taking into account the
division of the year by way of the solstices and the equinoxes. We
shall recognize an effort of the 328-night count and of the synodic
lunar calendar to give these divisions of the solar year a
significant place in their own systems.

237
 The cultural and social expression of the solar year was
carried out by way of the following system of llama sacrifices.
Llamas were dedicated, either to the god Viracocha, identified by
the Spaniards as the Creator god, or to the Sun god, or to the
Thunder god. Guanacos, wild llamas, being described as brown,
were dedicated to Viracocha. White, woolly llamas belonged to the
Sun god and llamas with hair of different colors, that is probably
alpacas, belonged to the Thunder god. Polo de Ondega rdo and the
chroniclers who followed his description (Acosta, Cabello Valboa,
Murua, Guaman Poma and Cobo) mention that in each month one hundred
llamas of one class were sacrificed. For reasons to be explained
in a moment, I conclude that, when Polo talks about the month of
June, he means t~e period of 30 or 31 days ending at the June
solstice, according to the system set out in Table 1.

It is difficult to imagine that a system with diametric

oppositions, like that expressed here, could exist in any other
way than in a count of the solar year of 365 or 366 days. The
months around the solstices oppose to each other llamas of the same
category: guanacos or wild llamas. The months around the
equinoxes represent a more complicated system. The fourth and tenth
months involve a color opposition: sacrifices of white llamas in
the fourth and those of black llamas in the tenth. Additional data
on the other two equinoctial months support this conclusion and
allow us to lead this discussion into an astronomical one. In
October, the fifth month, a black llama was bound in the plaza to
a pole, depriving it of food and water, in order to make it "weep
for rain." From that time on, rains were expected to come for the
crops, especially the one of maize, that were planted from August
(early planting) a n d September (general planting) onwards. In
April, the eleventh month, a white llama was bound to a pole in
the plaza, a llama that was fed chicha (maize beer) and that was
expected to kick o ver a vessel containing beer with its feet. Here
the reference was to the repair of irrigation canals, needed after
the time of heavy rains. A ritual race, carried out from a
mountain called Anahuarque, enabled me to connect the ritual of the
black llama to calendrical considerations concerning the black llama
in the sky. As this constellation has its period of lower
culmination in October, that is its disappearance during par"t of
the night; we can suggest that the ritual of the black llama in
October referred to the period of lower culmination of the celestial
llama and the one of the white llama in April to that of its upper
culmination.

As the sacrifices of the guanacos and of the white llamas were

grouped into three seasons of three months each, we are presented
with the suggestion that the sacrifices of black llamas in March,
the tenth month, and those of different colours in the eleventh and
twelfth months also belonged to one season. The question therefore
becomes: is there a reason to oppose this whole group to that of
the white woolly llamas, sacrificed in the fourth, fifth and sixth
months? Although the rituals in these months include a strong
female connotation, it was the time when the sun grew stronger,
sending not only heat, but also rain to make the plants grow (see,
e.g., Guaman Poma 1980: 885/99). Priests called Tarpuntay-- - from
tarpuy, "to plant"; being priests, however, dedicated to the

238
 Table 1 Llamas and the solar year

The description of the My interpretation month names my interpretation

chroniclers. of their data. in chronicles. of these months.

Guanacos (wild June Months around

l
1) guanacos, brown, living

!
in the high mountains.
2) brown, color of vizcacha, . llamas) dedicated July June solstice
yahuarchumbi (reddish brown).
brown, without any spot, to Viracocha. August and mo , th after.
chestnut, color of vizcacha.
Septemher Months around

!
White woolly

!
4) white, woolly

5) nf the same llamas dedicated October Sept. equinox

6) of the same to Sun god. November and month after.

Guanacos December Months around

!
7) large llamas

8)

9)
brown, of a light color and
white
like
head and knees.
the other months,
chestnut color.
!dedicated to

Viracocha.

Llamas or alpacas
January

February

March
Dec. solstice

and month aft~r.

Months arourrl

! !
10) hlack.

11) moromoro, "pintado", dedicated to April March equinox

spotted, of different colors.
12) Thurrlergod. May arrl month after.

Table 2 Solar-lunar correlations in Cuzco

June snl stice full moon

(JS, 21 June)
sun in nadir; antizenith full moon
sunset (AZSS, 18 August)
September equinox full moon (17 September)
(21 September)
sun in zenith; zenith new moon
sunrise (ZSR, 30 October)
December solstice new moon (28 December)
(OS, 21 De,:emher)
sun in zenith; zenith full moon (11 February)
sunrise (ZSR, 13 February)
March equinox full moon (12 March)
(21 March)
sun in nadir; antizenith new moon
sunset (AZSS, 25 April)

239
Sun-- had to fast for two months until the maize plants were grown
two hands high. While these months expressed a connotation of
increase of the Sun, those around the solstices were more neutrally
defined. During the first three months (June to August) the earth
was dry, cold and without crops. From the seventh to the ninth
month (December to February) the rains were too strong for their
waters to be contained within irrigation canals. Therefore, maize
plants already had to be strong enough to withstand the destructive
forces of nature by themselves and without the help of man. In
March, the tenth month, people started their first actions to
regulate again the water received from heaven. In Huarochiri,
they would go up to the mountain lakes to see the level of water
there. They would then close the overflows to the irrigation
canals (Avila 1980: ch. 31). We can interpret the sacrifice in
Cuzco of the black llamas in the same month against the background
of this practice and against that of the general concern with water
from October to March. In October the black llama was starved but
not killed. In January, during the .heavy rains, people would go
out into the mountains around the town, weeping and doing penance
just as they forced the llamas to do. There would be less food.
Vegetables, eaten then, occasioned inflated stomachs and death.
But in March the first potatoes and maize arrived and vegetables
were good to eat. Guaman Poma (1980: 240/242; 241/243) mentions
specifically how it was that the high priests and the other priests
performed the sacrifices dedicating the hundred black llamas to all
the mountains that were recognized by the king to be worshipped.
The intention of killing the black llamas apparently was to induce
the end of the rainy season, a period that had been called up
initially by making black llamas weep.

In this context, Guaman Poma gives us a precious clue of how

to include the political function of the 328-night count within the
solar year. The high priests in Cuzco were descendants of the
legendary king Viracocha Inca, the king who derived his name from
the "creator" god Vi racocha. This group of descendants was cal led
Sucsu, a name ref erring to a sickly and consumptive condition, a
discolored and yellowish complexion and to a shrunken body or
shrunken arms and legs. I will analyze in another article the
myths related to this condition. We can, however, already
suggest the interpretation that it was the group Sucsu, with the
high priest belonging to it, who was in charge of the month of
March, ending the period of sickness and introducing the two months
of harvest. Calendrically, the group of Sucsu was in charge of
period III in the 328-night count (as I reconstructed its place
within the solar year), from 14 February to 17 March. This period
corresponds closely to that of the tenth month in the solar year
count, going from 19 February to 20 March. The data from the 328-
night count and the solar year confirm each other in this case.
(For the technical details of the 328-night count and the place of
Sucsu, see the third section.)

As was mentioned before, llamas with wool of different colors,

sacrificed in April and May, were dedicated to the Thunder god.
The data from the documents on the extirpation of idolatries will
tell us more about the calendrical association. However, the myth
from Huarochiri on the celestial black llama also has specific

240
details on llama wool of various colours. These details can
clarify not only the information on April and May, but also certain
other information of Polo de Ondegardo on the celestial llama and
her family.

In my first reading of this myth from Huarochiri, I paid

special attention to its first part (Zuidema and Urton 1976). The
llama drinks the waters of the earth, savl ng it from the Flood, in
the middle of the night , when nobody sees her. She suckles her
baby-- a dark-cloud constellation just below her-- when she, rut
not the baby, has already risen. By comparing this information
with that from Cuzco, I reached the interpretation that the myth
discusses the periods of lower culmination of the llama and that of
her baby. The myth also says, however, that sometimes, when she
drinks the water from all the springs, she falls on top of a man,
probably one sleeping in the fields next to a spring. Such a man
was considered lucky as he would be covered completely with wool.
Other men would shear the wool, this happening during the night.
In the morning they would discover that the wool was of various
colours: blue, wh ite, black and brown, all colours mixed. If
the lucky man had no llamas then he would immediately buy two, a
male and a female, and these would bear him a herd of two to three
thousand llamas. He would give cult to the wool where he had seen
the celestial llama (the Yacana) and where his friends had sheared
the wool.

We can search in two directions for an interpretation of this

part of the myth in terms of the Inca calendar. The first
interpretation relates it to the December solstice, and to the full
moon after. On the day of the December solstice, the noble boys
of Cuzco finished the rituals initiating them into manhood by going
into a cultivated field near a source of water and having their ears
pierced, an act f or boy s the timing of which was tied to the first
menstruation of girls as their initiation into womanhood. On the
next day, the pries ts of the "creator," the Sun, the Thunder and
the Moon, together with the herders of the king, began the feasts
of multiplication, counting the llamas belonging to these gods and
to the king. During the next full moon, men and women-- those
who were rich owners of llama herds and the parents of the rece·ntly
initiated boys-- took up a rope made of wool of four colors:
white, black, red and yellowish brown, a rope that represented a
llama. They danced around the king in the plaza, leaving the rope
at the erid in a spiral around him. The colours of the rope,
symbolizing a llama, suggest those of the wool given by the
celestial llama to future, prosperous owners of llamas. The
connection of the initiation rituals to the counting of llamas
suggests, moreover, an initiation of boys into economic prosperity
by obtaining llama herds.

But this second part of the myth from Huarochiri can be

understood also in terms of the llama sacrifices during the period
from the tenth to the twelfth month (March to May). The celestial
llama drinks the waters of the earth, preventing the Flood. The
act does not seem necessary in October, rut it is in March. Going
from sacrifices of black llamas in March to sacrifices of llamas of
different colors in April and May could reflect the bestowal of wool

241
of different colo r s by th e celestial black l l ama. The myth speaks
of a change from n i ght to day, but it symbolizes a change from wet
season-- when during six months, from October to March, black
llamas were important and when, moreover, black llamas, black
dogs and people wept-- into dry season around the March equinox.

This second inte r pretation is probably not in conflict with the

first. It leads, however, to further astronomical consequences,
if one compares the data from Huarochiri with those of Polo de
Ondegardo from Cuzco and those of Bertonio from Lake Titicaca.
Every author gives the celestial llama a kid that we can recognise
as a dark-cloud constellation near his mother. But while Bertonio
calls the young llama unuchillay (unu, "young"), Polo (1916: 3-4)
mentions also the star Vega in the constellation of Lyra as husband
of the black celestial llama. Moreover, he gives the father and
young llama the same name, urcochillay; urco meaning male,
specifically male llama. Vega was a male llama of different
colors, worshipped by owners of llamas and, probably, by their
sons aiming to o b ta i n a herd. If the information of Polo can be
trusted-- it is no t confirmed by any author independently of him--
it could express the astronomical relation between Alpha and Beta
Centauri and Veg a as t h e e yes of the female llama and the male
llama respectively. Vega has its heliacal rise in the morning
during the begin n i n g o f F eb r uary. That is, it occurs just after
the days when t h e ri t u al with the multicolored llama rope was
executed on a nig h t of fu ll moon following 28 December. (See
section III for t h e 328-night count; such a full moon could occur
between 28 December and 27 January). In February, when Vega rises
heliacally in the morning, Alpha and Beta Centauri have their
upper culmination a t midnight and Vega rises at the same hour. The
upper culmination of Vega at midnight is around the June solstice.
This fact may have been significant. While, on the one hand,
Polo de Ondegard o men t ions the sacrifices of the one hundred
guanacos in the months of the solar year count before and after the
June solstice, Molina, on the other, in his description of the
celebration of the June solstice itself during the synodic lunar
month of Intiraymi-- a month that could occur from 27 days before
to 27 days after t he day of the June solstice-- gives a list of
llamas of different colors which were then sacrificed, probably in
addition to the one hundred guanacos.

I wish to end the first part of this paper with three

observations:

1) Polo said that llamas of different colours were sacrificed to

the Thunder god in order "that there be no lack of rain." - We
discovered that their role in the calendar was of a far more complex
nature. They were sacrificed within the regular system of
sacrificing one hundred llamas a month, when the Sun was losing
strength, during the two months beginning the half year related to
the dry season. The dry season was needed for harvest, for the
drying and conservation of different foods, for travel and for
warfare. Nonetheless, there remained during this time a positive
interest in water. The irrigation canals were rebuilt, ensuring
that during next year's agricultural cycle water would not be
lacking. During these months, when the forces of the sun were

242
failing rapidly, the task of regenerating the year was given to the
Thunder god, at a time before the actual new agricultural cycle had
started. The myths to be dealt with next explain how this role of
the Thunder god was represented.

2) The diametrical oppositions as discussed here in the context of

the calendar-round of the solar year; the opposition of black to
multicolored llamas in terms of calendrical contiguity around the
March equinox; and the fact that llamas and men were made to weep
during six continuous months and not in the other months make it
imperative to think of these oppositions in terms of the solar year
cycle, divided by the solstices and equinoxes. But do we have
evidence, other than this interpretation of data from the Spanish
chroniclers, to support such a claim for a division of the solar
year into periods of 30 or 31 days? In 1977 I published a
photograph, with analysis, of a pre-Inca textile representing a
calendar of 12 months of 360 days with 5 extra days. Therefore,
not only did this type of calendrical structure exist in pre-Spanish
times, but it also carried out a regular repetitive count of the
days of a solar year as distinguished from the counting of irregular
calendrical distances between various astronomical observations, as
was done with the help of the 328-night count. The solar year
count apparently had a base also in astronomical observation.
Along with observing sunrise and sunset during the December solstice
and sunset during the June solstice, the Incas made important
astronomical observations of sunrise 30 days before and 30 days
after the June solstice (see Note 2). In addition to these
observations, the 328-night count recognized the date of 21
November, that is 30 nights before the December solstice, as the
beginning date of its second half-year period. These dates, then,
had a significant place in the count of the solar year in terms of
months with 30 or 31 days.

3) The celestial llama had its period of lower culmination between

about 7 October and 2 November and that of its upper culmination
around 20 April. The myth of the celestial black llama itself,
however, has no reference to the Thunder god, nor to his relation
with the wool of different colors that she gave. Moreover,
neither the mythology of Huarochiri, nor that of Cuzco, has given
us any clue to the constellation of the Yutu, the "partridge,"
although its period of culmination around 25 March covers part of
the time when llamas of different colors were sacrificed, that is
the time from 21 March to 21 May. The myths recorded in the
documents on idolatries assist us on this point: they discuss the
connection between a Yutu that is probably celestial and the Thunder
god.

The three calendrical counts

This paper is primarily concerned with testing the three

calendrical counts-- from Cuzco and elsewhere-- against each other
in their parts corresponding to the months from March to May. I
shall mention first the place of these counts in the year as
suggested in my earlier articles and indicate the type of
information that is used in support of each. This overview

243
enables me to define my terminology.

The solar year count

The year was divided into 12 months of 30 days, 36 weeks of 10

days, and 5 or 6 days extra, added either intermittently after
certain months or at the end of the year. The most important data
on these months are found in the system of llama sacrifices, one
hundred llamas in each month. These data on llamas are described
in the chronicles by way of the months of the western calendar. We
have to disentangle, therefore, data referring to the months of the
Inca solar year from those referring to the Western calendar and
those referring to the synodic lunar months. The count of the four
seasons of the solar year started 30 or 31 days before each solstice
and before two dates close to each equinox, respectively (see Table
1).

The synodic lunar month count

The correlation of this count to that of the solar year was

established by taking the lunar month including the June solstice as
the first of 12 lunar months. The average, and ideal,
correlation was having a full moon during the June solstice, 21
June. This average synodic lunar calendar started on the night of
the first visible moon, that is 9 June, some 17 days after the
beginning date of the solar year, 23 May. The latter date helped
to define the earliest possible date of a new moon in the movable
synodic lunar count, 25 May. The synodic lunar count is eleven
and a quarter days shorter than the solar year count. The
information on the synodic lunar months is given mostly in
descriptions of the agricultural system and of the feasts
celebrating events of the solar year on either a full or a new moon.
The associations whi c h the Incas observed in Cuzco between their
average synodic lunar month count and events of the solar year are
shown in Table 2. (Table 3 gives the names of the months). I
have added the equinoxes to this list, although I do not have
information of a correlation with the moon in their case. The data
on the feast of Coya raymi do, however, suggest such a
correlation.

The sidereal lunar calendar

Although closely related to the 328-night count, the sidereal

lunar calendar has to be distinguished separately, in a way
somewhat similar to the difference between the solar year count and
the synodic lunar month count. The sidereal lunar calendar can be
understood best as being based on the observation of 8 periods of 41
nights, each period being one and a half sidereal months of twenty-
seven and one third nights. As an inversion of this calendar, and
thus as another expression of the same, the Incas had a calendar of
41 weeks of 8 nights. As against the solar year calendar with male
associations, the sidereal lunar calendar had a female character.
The count of this calendar started with the heliacal rise in the

244
 Table 3 The pol itical
· organization of Cuzco

.r
85

37

-6.9
b
-
--
C
26 8

b 3

:rr
85
morning of the Pleiades, together with the first new moon of the
average synodic lunar month count, timed to 9 June. Apparently,
the 37 nights between 3 May, the end of the sidereal lunar
calendar, and 9 June were not counted in this system.

The ceque system as a system of astronomical observation

The ceque system consisted of 41 directions, radiating out

from the central Temple of the Sun in Cuzco. Comparison with other
similar systems in Andean towns suggests that the number 41 was a
constant, but that the directions were not exactly alike in each
local context. We can compare the ceque system best to a
concentric system of coordinates describing space around Cuzco.
Certain points on the horizon, indicated by way of ceques, were
also used for astronomical observation. But only two of the places
used to observe specific sunrises or sunsets coincided with the
Temple of the Sun in making their observation along a ceque.

The 328-night count

By way of recognizing certain locations along each ceque and

defining the number of locations admitted to each, a calendrical
sequence was built up, worshipping each location following a daily
order. A location, considered as huaca, "sacred place,"
because of its inclusion in the ceque system might have no
importance other than in the calendrical sequence it helped to
build. The role of a ceque in the total of 41 is indicated by the
combination of a Roman numeral (I, II, III or IV), an Arabic number
(1, 2 or 3), and a letter (a, b or c). Each combination serves
also two calendrical functions: one, for defining a date in the
328-night count and another for defining a period in relation to a
date. In II and IV the period is that of the number of nights
before the corresponding date. In I and III it is that after the
corresponding date. The context of the argument makes clear which
use is made of a certain · com bi nation (see Table 3). In the
political organisation of Cuzco, ritual functions in relation to
the huacas of a ceque were taken care of by a group of people living
in the direction of that ceque. I have described elsewhere this
system of 10 royal ayllus or panacas, of 10 non-royal ayllus, and
of certain non-Inca villages organized around the two other towns of
the valley of Cuzco: Safiu (today called San Sebastian) and Oma
(today called San Jer6nimo). In terms of these data on social
organization, IV is different from I, II and III. We have little
data on the panacas and ayllus associated with IV. Their places
are taken over by the Queen and women of the valley as one group,
and by the villages of Sanu and Oma. Sanu was divided into the
groups Safiu and Ayarmaca:=-- groups that still exist today as
moieties-rri-san Sebastian-- and the hamlets of Yacanora and Cari
(see Table 4).

246
Table 4 The association of a legendary king, panaca, and non-
royal ayllu to one group of 3 ceques (note exception in IV)

Group of non-royal ayllu Calendrical

3 cegues association

II 3 Lloque Yupanqui Haguayni CUycusa or

3rd king Aquiniaylla 9 June - 5 July

II 2 Capac Yupanqui Apu mayta Maras 6 July - 4 Aug.

5th king

II 1 Mayta Capac Uscamayta Sutic 5 Aug. - 2 Sert.

4th king

r
3 b, 2 Queen women 3 Sept. - 30 Sept.

IV 1, IVA 3 a,c village of Uma 1 Oct. - 30 Oct.

8

IVA 2. 1 (moiety) Ayarmaca, 1 Nov. - 21 Nov.

village of Sanu

I 1 Tupa Yupanqui Capac O,avin CUZCO 21 Nov. - 23 Dec.

10th king ayllu

I 2 Pachacuti Inca Inaca p., Arayraca ayllu, 24 Dec. - 21 Jan.

9th king Hatun ayllu CUzco callan.

I 3 Inca Roca Vicaquirao Huacaytaqui 22 Jan. - 11 Feb.

6th king

III 1 Viracocha Inca Sucsu Tarpuntay 14 Feb. - 16 March

8th king

III 2 Yahuar Huacac Aucaylli Saiiu (probably 17 March - 9 April

7th king moiety Saiiu of
village Sanu)

III 3 a Yacanora 10 April - 3 May

b Ayarmaca
C Cari

247
The myths from the province of Chinchaycocha (Lake Junin) and from
the village of Hacas, province of Cajatambo

The first myth, from the area of Chinchaycocha, today called

Lake Junin, deals with the adventures of a sacred mountain,
Tumayricapa-- also known from another source in the XVIth century
(Albornoz 1967: 30)-- who was widely worshipped as a god of
strength, industry, fire and luck. The character of the myth
demonstrates that we can consider him as the god of Thunder.
Duviols (1974-76: 289), who published and analyzed the small
chronicle in which this myth is found, registered from present-day
inhabitants of the area a modern version of the myth. I shall
refer to this version only in passing, interesting as it is, where
it helps to place the older myth calendrically. The modern version
mentions also a female deity known in the myth from Hacas,
establishing a connection between the older myths and the villages
where they are told. Duviols argues for this connection on other
grounds also.

The myth from Chinchaycocha (Duviols 1974-76: 275-78) tells

how Tumayricapa, together with his brother (who disappears
immediately), came down to earth near Chinchaycocha on a mountain
called Mamallqui jirca; a name that the document translates as
"plant, beginning or origin of the mountains." There he called
together all the Huacas, the sacred places. He went on to Bombon,
a plain 10 to 15 kms away, and changed himself into a richly
dressed child. A woman from the nearby village of Huaychau heard
the child crying and brought it to her village and gave it the
breast. Her name was Pullucchacua, interpreted by the document as
"skin of the partridge," that is "skin of the bird without its
feathers." Tumayricapa grew up within 5 days to be a strong man.
He called together two other snow-covered mountains from either side
of the valley. With them and with the other mountains and Huacas
he organized a communal hunt of vicuftas in the plain of Uirapampa,
close to Bombon. One mountain, Quirmuchan, wanted to grow higher
than Tumayricapa, but the latter defeated him bringing him down
with his rivi, a boleadora of which the three lead bullets had
been replaced by three heavy rocks. Quirmuchan is now a low ridge
of stones that divides the plain into two parts.

Concentrating my analysis of the myth only upon certain

elements, two female roles are identified as mother to Tumayricapa.
The first, who bore him, can probably be identified with Mamallqui
Jirca. The translation of "plant, beginning, or_i _gin" refers
correctly to the word Mallqui. As the text mentions mamallqui, we
might consider this word as a contraction of mama, "mother" and
mallqui "plant of origin," a plant in this case specified as
"motherplant," "origin of the mountains" like Tumayricapa and the
other mountains referred to. The second mother-role is occupied by
a woman who, because of her name, can be identified as a species
of Yutu, the tinamou. Various details of the myth, like that of
rebirth or rejuvenation, Tumayricapa as twin brother, the chaco or
communal hunt and that of the fight using the boleadoras, will
allow us to connect the myth, first, to the time of the passage of
the sun through the nadir, when it sets at its antizenith point of
19 April at the latitude at the latitude of Chinchaycocha, and,

248
second, to the time represented in the Inca calendar by t h e
eleventh and twelfth solar months, the months of April and May
around the date of 19 April. But ·1n order to come to an
identification of the woman Pullucchacua, I shall pass first t o
the myth from Hacas. The document where this myth is found belongs
to a group describing the idolatries of three villages close to each
other: 0cros, Chilca and Hacas (Duviols 1971: 367-86; Huerta s
1981: 50-59, 75-82 and 104-19).

The myth from Hacas, retold various times, explains a ritua l

carried out at a time specified calendrically in the following ways :

1) when people plow their fields; that is

2) when the rains stop (and when the water in the rive rs
diminishes);

3) at the time on or after Easter;

4) when people execute the dances of aylli aylli and ayrihua.

For the ritual, people catch the yucyuc, a bird with yellow be ak
and legs, dress him in a shirt and shouldercloth, carry him a r otmd
with girls following him, singing and playing small drums. Peopl e
offer sacrifices, thanking him for the cultivated plants that he
has bestowed on them. The myth tells how these plants had belonge d
to a woman called Mama Rayguana, "mother Rayguana." An eagle, a
fly catcher and the yucyuc stole, by a ruse, her baby, a bab y
which was · a conopa, a stone protecting cultivated plants. In
exchange for getting her child back, she gave up the plant s:
highland plants like potatoes etc. for the people living there and
lowland plants like maize etc. for the people on the coast.
Although the myth does not make it completely clear, we ma y
conclude that the highland people received their plants through the
yucyuc and the lowland people theirs through the flycatcher.

The description given of the yucyuc in the text does not allow
us to identify with confidence the species of the bird. However,
the Description of the West Indies of 1629 by Vazquez de Espinosa
(1942: para. 1738) mentions that on the northcoast of Peru there
existed three kinds of "partridges," tinamous: guayco, the big
ones; picasa (probably a mis-spelling of pisaca), the medium-sized
ones; and yuyo, the small ones (Rostworowski de Diez Canseco 198 1:
61). Vazquez de Espinoza mentions them for the coast; but in 1982
in 0cros I observed small tinamous kept in houses in a semi-
domesticated state, a fact that demonstrates their existence in the
highlands.

But who was Mama Rayguana? The myth says that she lived
elsewhere, in a village called Caina, probably further up into the
mountains. The modern version of the myth of Tumayricapa mentions
that the people north of Lake Junin (Chinchaycocha) worshipped
Rayguana as their goddess, at a time when there was still enough
water to cultivate lands. She lived in a place called Atojhuarco
("the hanged fox"). The myth explains how two eagles came down out
of the sky and how Rayguana advised Tumayricapa about the meaning of

249
the prognistication. As a result, the people of the Wankas,
living in the highlands, defeated those of the Waynas, obliging
them to retreat to warmer and lower valleys. We can possibly
identify Rayguana, as a huaca, with the place where Tumayricapa in
the older myth came down to earth: Mamallqui jirca, the '~lant or
beginning or origin of the mountains." The reference here to
Mallqui as "plant of origin" may have been similar to that of Mama
Rayguana in the myth from Hacas as the original female owner of
cultivated plants. But even if we discount a direct identification
of Mamallqui jirca with (Mama) Rayguana, the sequential place of
both is similar in the three myths. They precede either
Tumayricapa, the Thunder god, in the myth from Chinchaycocha; or
the eagle (one form of the Thunder god) in the myth from Hacas; or
Tumayricapa and the eagles together in the modern myth.

The two older myths seem to have a similar calendrical

reference and purpose. Before arguing this, I shall have to
mention some transformations between the two myths. Tumayricapa
acts in the myth from Chinchaycocha as a Thundergod, a god who in
other myths from Central Peru came down to earth as an eagle. If
we compare him to the eagle in-the myth from Hacas, then it becomes
an obstacle to comparison that in Chinchaycocha the yutu bird is a
foster mother of the eagle (i.e. Tumayricapa), but in Hacas his
male collaborator. However, while in Chinchaycocha the yutu is a
secondary mother to the eagle, in Hacas the yutu is the effective
secondary bestower of cultivated plants to the people there. Thus,
in Chinchaycocha Tumayricapa has an intermediary role between the
two mothers, on the one hand, and the people who worshipped them,
on the other; in Hacas, the eagle had an intermediary role between
the two bes towers of plants. In terms of the calendar, the role
of the Yutu (in Hacas called yucyuc) probably remains ' the same in
the two myths. The question, therefore, becomes: can the ritual
and mythic significance of the Yutu bird in these two myths refer to
the Yutu constellation; to the constellation which has its period
of upper culmination around 25 March?

With this question in mind, we are now in a position to

discuss the four calendrical references in the myth from Hacas as
they had been mentioned before:

1) The plowing

Guaman Poma, in his calendar, mentions, on the one hand,

that plowing of fallow lands was done in March; on the other hand,
the King in Cuzco would carry out a ritual plowing in May as well as
in August. At the moment our information on the agricultural
calendar of modern highland villages overlooking the coast, like
Ocros and Hacas, is still too scanty to draw specific conclusions.
There seem to exist, however, some remarkable differences between
this region and Ayacucho and Cuzco, from which latter places most
of our information derives. Therefore, I shall base my calendrical
argument for Hacas more on the following data.

2) The rains in Ocros and Hacas stop in about April. My

ethnographic information and that of Huertas indicate that then it
is time to repair the irrigation canals (Huertas 1981: 54).

250
3) Easter, as a movable feast, can fall on a day from 22 March
to 25 April.

4) The time of the feast of the ~ucyuc bird fell on a day with in
this period from 22 March to 25 April. Other data on the dan c es
aylli ay11i, ayrihua and that of the Pallas with their d ru ms
support the choice of this period. Data on the Inca calendar (e. g.
Cobo 1956: 214) indicate that the names of Ayrihua and aymoray
were applied to the eleventh and twelfth synodic lunar months
respectively. Their average place in the calendar is from 28 March
to 25 May, but Ayrihua could start as early as 13 March and as late
as 10 April. If we take into consideration the indications of Polo
de Ondegardo and Molina concerning the 12 months of the solar year
calendar, then we are dealing with the period from 21 March to 21
May. It is of further interest to recall that in the avera g e
synodic lunar calendar the new moon separating the months of Ayr i hua
and Aymoray coincided with the passage of the sun through the na di r,
when the sun is setting at the antizenith point, that is o n 25
April. In Ocros and Hacas, this passage occurs 6 days e a r lie r,
on 19 April.

In these villages, as in Cuzco, the name of ayrihua was

applied to the kinds of maize cobs that were considered as t wins ,
cobs having grains of two different colours. These were not e at en
or used for making beer, but burned on the fields as a sacrifice to
the new crops at the time of early harvest in April. Another name
of twin cobs was chuchu, a name given also to human twins and to
two small stars said to be close to the Pleiades. This informa tion
would mean · that the twin stars disappeared around the same time as
the Pleiades, these latter disappearing from+ 15 April to+ 8
June. We can conclude that the month of Ayrihua was tied to t he
time of the disappearance of the twin stars. Although in Cuzco the
name ayrihua is given normally to the eleventh month (eith er the
synodic lunar or the solar month) as a reference to the ritual o f
ayrihua, the ritual itself could be celebrated also in the twelfth
month (Guaman Poma 1980: 245/247; Molina 1943: 66). Probably we
have to interpret ayrihua in general as an agricultural ritual
celebrating the harvest during the last two months of the solar year
count, that is during the time when the twin stars disappear.

We arrive at similar calendrical results by analyzing the name

ay11i aylli. Aylli means "victory · song," also used in the
combination aucaylli (auca, "warrior, enemy"; ayl1i, "victo ry
song"). It was used ~war, as well as during planting and
harvest in agricultural rituals. The panaca associated in Cuzco
with III 2, representing the eleventh period in the 328-night count
(17 March to 9 April), was called Aucaylli panaca. This panaca
held land among the settlements of the non-royal, and probably pre-
Inca, ayllus of Saffu, related also to III 2, and of Yacanora,
Ayarmaca and Cari,~longing respectively to III 3 a, III 3 band
III 3 c. These last ceques together represented the twelfth period
in the 328-night count (10 April - 3 May).

Ceque III 3 a contained as its first huaca a plain, a

cultivated field, called Ayllipampa ("the plain aylli"), of which
"people said that it was the goddess earth called Pachamama ('mother

251
earth'), and they offered her small female dresses." The three
ceques were very close to the same direction, succeeding each other
out from Cuzco and pointing to the zenith sunrise as seen from the
Temple of the Sun. This direction had · been used in 1535 at some
date in the month of April (Julian calendar), that is between 11
April and 11 May (Gregorian calendar), for one of the last Inca
rituals carried out with full pre-Spanish pomp (Cristobal de Molina,
el Almagrista 1967: 81-83). The main event was a procession
celebrating the bringing in of the harvest. Cobo described such a
feast in more general terms occurring at the beginning of the month
of Aymoray; that is, in terms of the average synodic lunar
calendar, a date just after that of antizenith sunset, 25 April.
The use in this context of the names Aucaylli and Ayllipampa
confirms their calendrical importance, being related to the periods
of III 2 and III 3 (17 March - 3 May), overlapping with the months
of Ayrihua and Aymoray (Ayrihua having its earliest possible
beginning date on March 11), and overlapping also with the eleventh
and twelfth solar months dedicated to the Thunder god (21 March - 21
May).

The calendrical arguments given here, make it completely

plausible to arrive at the following conclusions:

1) The author of the document on Hacas was right in assigning a

time to the ritual of the yucyuc bird corresponding to the eleventh
solar ronth, the month after the March equinox.

2) The ritual reflected the upper culmination of the constellation

of the Yutu in the Southern Cross.

3) The male yucyuc bi rd of the ritual and the yutu-mot her of the
myth of Chinchaycocha probably had the same calendrical
significance.

4) The reference to the Ayrihua and the Aylli aylli dances

extended the correspondence with the Inca calendar to the whole
period when the Pleiades and the Twin stars disappeared and when the
eleventh and twelfth solar months were dedicated to the Thunder god.

5) The eleventh and twelfth periods of the 328-night count

corresponded to the beginning of the period mentioned in 4).

Mama Rayguana and the Thunder

Having made specific suggestions about the calendrical

questions, I can now come to a more general discussion of the
calendrical importance of the Thunder god in his relation to the
female goddess who was the origin of the cultivated plants. The
first question is: who was Mama Rayguana and can we give her a more
precise astronomical and calendrical identification.

Mama Rayguana and Mamallqui jirca apparently had a place in the

calendar related to an earlier date than either the celestial black
Llama or the celestial Yutu. While the Llama gave men wool of

252
different colors, but no plants, the Yutu had an intermediary
function between Mama Rayguana and men in terms of plants and also
between Mamallqui jirca as the "plant of origin" and the Thunder god
to whom the llamas with wool of different colors were sacrificed.
Although I do not have any direct evidence for the suggestion I want
to make, I suspect that Mama Rayguana and Mamallqui jirca both
might have a symbolic association with the toad, and that their
celestial correspondence was with the black-cloud constellation of
Hampatu, the "Toad." This constellation is mentioned in Cuzco
today, being found just before (to the west of) the Yutu in the
Milky Way (Urton 1981a, 1981b: 180-85). Its period of lower
culmination is just before the September equinox and that of its
upper culmination just before the March equinox; that is,
respectively, in the fourth and tenth solar months, when in the
first month 100 black llamas and in the second 100 white llamas were
sacrificed. The toad is identified in Mochica art as giver of
plants (Mariscotti 1978). Further iconographic research might
possibly suggest a relationship between its symbolic place there and
in Brazilian myths of the toad as mother of twins and as wife of the
jaguar. In Mochica art, one sees the animal either in a situation
of coitus involving a male jaguar and a female toad or one of two
toads (Mariscotti 1978). Today, in Cuzco, the toad is very much
identified with the rainy season, coming out of the cracks of the
earth in September and returning at the end of the rainy sea son
(Urton 1981b). In October the male toad is especially aggressive
sexually (Roca 1966), but otherwise the toad has an eminently
female role. Mariscotti has studied her identification with
Pachamama, "mother earth." In various places of Southern Peru the
toad is said to live in wet caves or under waterfalls, changing at
night into a beautiful woman, a siren, to whom men come "to have
their guitars tuned." In the last case, then, the physical
change of the toad is not seasonal, but daily. The myth of
Huatyacuri in Huarochiri (Avila: ch. 5)-- where the central deity
discussed is the Thunder god-- gives a crucial role to the toad.
Further research into Peruvian mythology may suggest, therefore,
the toad's importance in the calendar.

The possible significance of Mama Rayguana and of Mamallqui

jirca as the toad in its importance for agriculture may help to
understand further the rituals of the fourth synodic lunar month
(September) and the tenth (March) in their calendrical opposition.
The first, as the month of planting, was called Coyaraymi, the
feast of the queen, when women would invite men. Evil and illness
were expelled from the town, houses were cleaned and people would
invite each other~ Expelling evil was done by warriors,
brandishing torches and running out of town, following the 41
directions of the ceque system (Zuidema 1982c). The average
position in the calendar of Coya raymi means that a full moon fell
on 17 September, 4 days before the September equinox. The
earliest possible date of a new moon in this month was 19 August,
the time when the sun went through the central pillars, erected on
mountain Picchu for observing the antizenith sunset (that is, in
the 328-night count, after II b). The lunar month of Coyaraymi,
then, was correlated to the solar month before the September
equinox (the fourth), when white llamas were sacrificed to the sun,
and when priests of the sun, called Tarpuntay, started their

253
fasting, accompanying the germination and first growth of the maize
plants.

We observe a clear opposition between Coyaraymi and the time of

the year including the tenth solar month, as well as the tenth
synodic lunar month called Pachapucuy and the tenth period in the
328-night count (III 1, from 14 February to 17 March). The dates
of 14 February and 17 March correspond closely to the passage of the
sun through the zenith (13 February) and the equinox (21 March).
The name of the tenth month Pachapucuy (pacha, "earth," pucuy,
"rainy season, that of the ripening crops") probably refers to the
fact that at that time the king would address himself to the huacas,
the sacred places of the earth, in order to consult them about the
past, the future and distant places. The king, in his capacity
as consultant, was called Viracocha Inca. He identified himself
with the ancestor priest-king of this name. Guaman Poma (1980:
261-62/263-64), in a drawing of the event, clearly refers to the
ceque system, indicating his awareness of its importance. He does
not give the drawing in the context of his description of the
months, but in that of his description of the large temples and the
state ritu als of Cuzco. The description corresponds, however,
to the one of Pachapucuy, when the high priest, as descendant of
Viracocha Inca, carried out the consultation of the huacas. The
ceque system is referred to in two months, Coyaraymi an d
Pachapucuy, half a year apart from each other. Coyaraymi combines
the elements of female month and that of sacrifices of white llamas
to the male Stm god. We may ask, therefore, whether Pachapucuy
did not combine similar elements in a corresponding way; if this
month, when male rituals were carried out by high priests for the
huacas of the ceque system, was not dedicated also to female
deities like Mama Rayguana and Mamallqui jirca.

The suggestions made here may gain in strength if we see them

in the context, not only of the opposition between the fourth and
the tenth solar months, but also in that between the third and
fourth months together and the ninth and tenth months together.
The synodic months of Tarpuy quilla (3rd) and Goya rami (4th) formed
a group in the sense that they were both concerned with planting
(tarpuy). The dictionary of Gonzalez Holguin gives them similar
names, respectively: Kapak Sithua, "royal situa" and Antta
cittua. (He calls them August and July, respectively, but we
have to identify them as September and August). The name situa is
probably derived from an Aymara word (Bertonio): satav or sita,
being the root of verbs referring to planting, like Tarpu in
Quechua. The ayllu of the priests Tarpuntay belonged in the ceque
system, however, to III l; together with Sucsu panaca, the
panaca of Viracocha Inca and of the high priests. The tenth period
in the 328-night count (III 1, 14 February - 17 March) together
with the ninth period (I 1, 22 January - 13 February) corresponded
to the tenth and ninth synodic lunar months that had the name -pucuy
in common: respectively Pachapucuy and Haturtpucuy (hatun,
"large"). Thus, the two months of planting (tarpuy) were opposed
to the two months of ripening harvest (pucuy). While the priests
of Tarpuntay acted in the first two months, they belonged, in
terms of the ceq ue sys tern, to a period related to the other two.
Still today, the mother earth, Pachamama, is said to "open up" in

254
the months of August and February. It was probably in the time of
the year before the March equinox that Mama Rayguana was robbed of
her gift to humanity of cultivated plants, an act in which the
Thunder god played an active role and that was celebrated in the
month after the equinox.

With this calendrical understanding, we can venture now a

hypothesis of why, in the Inca calendar, one hundred llamas were
sacrificed to the Thunder god in each of the months Ayrihua and
Aymoray. Ayrihua was associated with twins; that is twins of
maize, animals and people, considered as being conceived by the
Thunder god. Twin children were dedicated to the Thunder god, one
was sacrificed, and seeds cal led ayri hua we re burned as an
offering to him. The months of Ayrihua and Aymoray were the months
of games, when the Inca king would enter into the binary relatio n
of a game of chance with people who otherwise were his inferi ors.
Finally, Ayrihua and Aymoray represented the time when the Ple iades
and the Twin stars were in the Underworld. The heroic time of t he
Thunder god, mentioned not only in the myth of Chinchaycocha bu t in
many other myths from Central Peru, dealt with the period when the
Twin stars represented him in the Underworld. A myth from Nor thern
Central Peru, Huarnachuco, about the Thunder god as a twin brother
(Agustinos 1952) is very similar to the myth of the twins in the
Underworld of the Popol Vuh in Guatemala. It lacks the ele ment of
the competitive games that the twins play in the Underworld there ;
but this element is not lost in the myth of Huatyacuri from
Huarochiri. This is another version of the theme of the her oic
Thunder god. The actions of the Thunder god accompany two
calendrical events: that of the passage of the Twin stars through
the Underworld and that of the moment when the Sun passes t hro ugh
the nadir. The Thunder god was reborn and grew up in five days
around the latter event, but his heroic actions in the myth of
Chinchaycocha extended over a longer period. It was at that ti me
that the Thunder god changed the crops of the year that was ending
into the seeds of the year to come.

The integration of the three calendrical counts

The mythological and astronomical information from

Chinchaycocha and Hacas helped to understand calendrical problems in
the period of the year from 14 February to 23 May and they concerned
not only this region of Pen1 but also that of the Incas in Cuzc o.
I have tried to understand in Cuzco data referring to the syn odic
lunar calendar and to the 328-night count in terms of the solar year
with months of 30 or 31 days and with seasons that each started with
the month before a solstice or an equinox. The comparison th at I
carried out of the three calendrical counts can lead us now to
suggest an integration closer than I hitherto suspected.

Aveni' s and my research into astronomy in Cuzco revealed that

the Incas had a specific interest in observing the sun in terms of
the solstices, of a date 30 days before the June solstice and of
the dates of the zenith sunrise and the antizenith sunset. The
integration of the average synodic lunar calendar and of the 328-

255
Table 5 The correspondence between the three calendrical counts

Months of solar Average synodic 328 night count (dates in

vear count lunar months (these II and IV indicate night
(estimated dates) months can occur 15 before; those in I aoo I I I
d~ys earlier or later) iooicate night after)

1) 22 Mav - 21 June 6 June - 4 July II 3 3rd king

( guanacos) Intiraymi 9 June - 5 July
taking maize ke~nels
from cob.
2) 22 June - 21 July Oiahuahuay guilla II 3 5th r:ing
(guanacos) 5 July - 3 August 6 July - 4 Aug.
( irrigation)
3) 22 July - 20 Aug. Tarpuy guil la II 1 4th king
(guanacos) 4 Aug. - 1 Sept. 5 Aug. - 2 Sept.
(early planting)
4) 21 Aug. - 20 Sept. Co ·ya raymi IV 3 b, 2 Queen, women
8
(white llamas) 2 Sept. - 1 Oct.
3 Sept. - 30 Sept.
(planting)
5) 21 Sept. - 20 Oct. Uma raymi IV 8 1, IVA 1 Uma
(white 11amas) 2 Oct. - 30 Oct.
1 Oct. - 30 Oct.
(expecting rains)
6) 21 Oct. - 19 Nov. Ayarmaca raymi IVA 2, 1 Ayarmaca
{white llamas) 31 Nov. - 29 Nov. 1 Nov. - 21 Nov.
(expecting rains)
7) 20 Nov. - 20 Dec. Caoac raymi I 1 10th king
(guanacos) 30 Nov. - 28 Dec. 21 Nov. - 23 Dec.
(rains, initiation)
8) 21 Dec. - 19 Jan. Capac raymi Camay guilla I 2 9th king
(guanacos) 29 Dec. - 27 Jan. 24 Dec. - 21 Jan.
(rains)
9) 20 Jan. - 18 Feb. Hatunpucuy guilla I 3 6th king
{guanacos) 28 Jan. - 25 Feb. 22 Jan. - 13 Feb.
(rains)
10) 19 Feb. - 20 March Pachapucuy guilla III 1 8th king
(black llamas) 26 Feb. - 27 March 14 Feb. - 16 March
(rebuilding lakes)
11) 21 .March - 19 April Inca raymi, Ayrihua g. III 2 7th king
(llamas different 28 March - 25 April 17 March - 9 April
colors) (rebuilding irr. canals)
12) 20 April - 21 May Ayrnoray guilla III 3
(llamas different 26 April - 25 May 10 April - 3 May
colors) (harvest from Ayrihua on)

256
night count-- the latter based on a sidereal lunar calendar-- into
the solar year add to those interests those of a date 30 days before
the December solstice (IV A la, 21 November), of a month-long
period around the antizenith sunset in August (II 1, 5 August - 2
September) and of the commencing date of a month-long period around
the antizenith sunset in April (III 3 a, 10 April). Furthermore,
the 328-night count seems to be more interested in the exact dates
of the zenith passages (IV A 3 a, c, 30 October; and III 1 a, 1
February) than in those of the antizenith passages. Although we do
not have any specific information about an interest in the
equinoxes, my discussion of the llama sacrifices seemed to reveal
such an interest. This interest might be supported by the
following dates in the 328-night count.

IV B 3 b , 1 7 S e p t ember and I I I 2 a , 1 7 Mar ch both o c cur f o ur

days before an equinox. The dates indicate that the Incas i n fact
might have been interested in observing equinoxes (see Note 3) ~
although in a way that led to a discrepancy of four days in rela t io n
to the exact observation of the equinoxes. If we take the two
dates of 17 September and 17 March into account calendrically , we
might arrive at a reformulation of the system of llama sacrifices
(one hundred llamas each month) in honor of the three gods,
Viracocha, the Sun and the Thunder.

The period from 23 May-- a date recognized by an observation

from the Temple of the Sun, although not in the 328-night count--
to 17 September (IV B 3 b) framed the time within which thefirs t
three months of the synodic lunar calendar could move from the ir
earliest to their latest occurrence in the solar year. The whol e
period covers 117 days. A similar argument can be made for the
third season of the synodic lunar calendar, accepting that the time
of I (85 nights) together with that of III 1 (31 nights) in the
328-night count framed those three synodic months. This who le
period was of a length similar to that of the first period
mentioned. This reasoning concerning the two periods of 117 and 116
nights would lead to the conclusion that the first and the third
seasons of the solar year were related more specifically to an
interest in observing synodic lunar months, induced by th e
obligation of celebrating the solstices in terms of full and new
moons.

In the articles previously mentioned (Zuidema 1982a, 1982b),

I arrived at the conclusion that the foremost interest during t he
second season of the solar year was in defining a sidereal lunar
count in terms of precise astronomical observation. The periods of
IV B 2 and IV B 1 together divide the period of one sidereal lunar
month of 28 nights in the regular way of 4, 4, 5, 5, 5, 5
nights. The constellations of Yutu and Llampa nawin (Southern
Cross and Alpha and Beta Centauri) have, as a group, their lower
culmination during this time (18 September - 15 October). We
observe on both sides of this period the half-month periods of IV B
3b (15 nights, 3 September - 17 September) and of IV A 3 a, c(l5
nights, 16 October - 30 October); periods that might have helped
to correlate the observation of the moon with that of the group of
the two constellations. During this period of the year (IV, 3
September - 21 November), the data support an argument for the

257
female associations in terms of stars (Southern Cross, Alpha and
Beta Centauri and the Pleiades), as argued previously.

The calendrical suggestions, presented here for the seasons

II, I and IV leave us with the final question concerning the
practical astronomical observations carried out in season III (14
February - 3 May) and the time not covered by the 328-night count (4
May - 8 June). I arrived at the conclusion that, in terms of its
constellations, this season was concerned especially with the upper
culmination of the Southern Cross (Yutu) first, followed by the
lower culminations of the Pleiades and the Twin stars. The central
concern during this season was, however, the Thunder god. We
are faced, therefore, with the question: can we assign to this
god a precise astronomical-calendrical association in a way similar
to that of the other seasons? One association that I determined
was with the antizenith sunset. But this date is not the same on
the various latitudes in the Andes (25 April for Cuzco; 19 April
for Chinchaycocha and Hacas). The Thunder god, the Twin stars and
the Pleiades were closely associated by Andean peoples with Venus as
the morning and evening star. In a future article, I shall
explore the significance of Venus in Andean astronomy and Andean
calend·a rs. I have to limit myself here to the suggestion that the
time from the March equinox and heliacal set in the evening of the
Pleiades and the Twin stars to the time of 23 May (being 30 days
before the June solstice) and heliacal rise in the morning of the
Pleiades was important especially for the inclusion of Venus in
calendrical considerations.

The calendrical dates presented in this final section reveal an

adaptation of the solar year count to those of the synodic lunar
count. The following periods, belonging to the solar year count,
were included in the calculations of the 328-night count: I 1 (33
nights) and I 2 (29 nights); II 1 (29 nights), II 2 (30 nights)
and II 3 (26 nights; that is a synodic month minus the nights of
the new moon that in this case were included in the period of the
year not covered by the 328-night count). While Polo de Ondegardo
referred in general terms to months of 30 and 31 days, the seasons
of I and II, each with 85 nights in the 328-night count, adapted
those concerns to other calendrical considerations, adaptat·ions
further accomplished with the help of III 1, 31 nights, and IV B 3
b, 15 nigh ts. It was the seasons of IV and III in the 328-nigh t
count that expressed the interest in observing the stars and,
possibly, Venus.

Notes

1 For an explanation of the periods in the 328-night calendar,

see pp. 245-46. The period of upper culmination indicates the
period when a star is seen during the whole night from the moment of
its heliacal rise to its moment of heliacal set at sunrise. The
period of lower culmination and of its invisibility extends from the
moment of heliacal set in the evening to that of its heliacal rise
in the morning. The constellations of the Southern Cross and of
Alpha and Beta Centauri, because of their relative proximity to the
south celestial pole, are defined as quasi-circumpolar. They
never disappear from the sky during the whole night. They can be
seen, at the latitude of Cuzco, and during their period of lower
culmination, to have their heliacal set at the beginning of the
night and their heliacal rise, during the same night, just before
sunrise. While the Pleiades are invisible during the whole night
for about 50 nights, the quasi-circumpolar stars are invisible for
only part of the night for about 25 nights in each case.

2 In 1980, A.F. Aveni and I made two types of measurement

concerning the direction of the Temple of the Sun in Cuzco facing
sunrise. The first consisted of the perpendicaular of the front
wall belonging to rooms A and B; the second of the center line in
between both rooms. The first gave a reading of sunrise 27 days
before and after the June solstice; the second of 30 days before
and after the June solstice. The first reading favors an attention
to the earliest possible date for a new moon of the month including
the June solstice; the second reading favors a date of interest for
the solar year count as discussed here. The discrepancy between
both readings consists of only some 15' of arc.

3 An Inca interest in observing the equinoxes is supported by the

measurements that Aveni and I carried out in Huanucopampa in 1980.

259
References

Agustinos

Religi6n en Huamachucho. Los pequenos grandes libros de

Historia Americana (Serie 1, vol. 17), Director: Francis.co
A. Loayza. Lima: Libreria e Imprenta D. Miranda. (1952)

Albornoz, see Duviols 1967

Avila, see Taylor 1980

Bertonio, Ludovico

Vocabulario de la lengua aymara (1612). La Paz: Don Bosco.

(1956)

Betanzos, Juan de

Suma y Narraci6n de los Incas ••• (1551), en Cr6nicas peruanas

de interes indigena, pp. 1-56. Biblioteca de Autores
EspaXoles. Madrid. (1968)

Cobo, Bernabe

Histo'ria del Nuevo Mundo (1653). Biblioteca de Auto res

Espanoles. Madrid. (1956)

Duviols, Pierre

"Un inedit de Cristobal de Albornoz: La instrucci6n para

descubrir todas las guacas del Piru y sus camayos y haziendas."
Journal de la Societe des Americanistes, 56, 7-39. (1967)

La Lutte contre les religions autochtones dans ~ Perou

co 1 on i a 1 , '1 ' ext i r pat ion de 1 ' id o 1 at r i e' en tr e 15 32 et 16 60.
Lima: Institut Fran~ais d'Etudes Andines. (1971)_ _ _ __

"Une petite chronique retrotivee: Errores, Ritos,

Supersticiones y Ceremonias de los Yndios de la Provincia de
Chinchaycocha y Otras del Piru." Journal de la Societe des
Americanistes, 63, 275-97. (1974-76)

Gonzalez Holguin, Diego

Vocabulario de la lengua general de todo el peru llamada lengua

Qquichua ~ del Inca ( 1608). Lim;;--Imprenta Santa Maria.
(1952)
Guaman Poma de Ayala, Felipe

El primer nueva coronica y buen gobierno (1615). Ed. J.V.

260
 Murra and R. Adorno. Mexico, D.F.: Siglo Veintiuno. (1980)

Huertas Vallejos, Lorenzo

La religi6n en una sociedad rural andina (siglo XVII)

Ayacucho: Universidad Nacional de San Cristobal de Huamanga.
(1981)

Mariscotti de Garlitz, Ana Maria

Pachamama Santa Tierra. Berlin: Gehr. Mann Verlag. (1978)

Molina, Crist6bal de, el "Almagrista"

Relaci6n de muchas cosas acontecidas en el Peru, en Cr6nicas

de interes indigena, pp. 56-95. Biblioteca de Autores
Espanoles. Madrid. (1967)

Molina, Crist6bal de, el "Cuzqueno"

Relaci6n de las fabulas y ritos de los Incas (1573). Los

pequenos grandes libros de-HistoriaAmericana (Serie 1, vol.
4). Lima: Libreria e Imprenta D. Miranda. (1943)

Polo de Ondegardo, Juan

Los errores x_ supersticiones de los indios sacados del tratado

x_ averiguaci6n que hizo el licenciado Polo. Ed. H. Urteaga and
c. Romero. Lima. (1916)
Roca, W. Demetrio

"El sapo, la culebra y la rana en el folklore actual de la

pampa de Anta." Folklore, Revista de Cultura Tradicional,
1, 41-66. (1966)

Rostworowski de Diez Canseco, Maria

Recursos . naturales renovables y pesca, siglos XVI x_ XVII.

Lima: Instituto de Estudios Peruanos. (.1981)

Taylor, Gerald

Rites et traditions de Huarochiri. Texte Quechua etabli et

--- - - - - - - - --
traduit par Gerald Taylor. Paris: L'Harmattan. (1980)

Urton, Gary

of the Earth and the Sky: an Andean

-At the Crossroads
- -
Cosmology. Austin: University of Texas Press. (1981a)

"Animals and Astronomy in the Quechua Uni verse." Proceedings

of the American Philosophical Society, 125, no. 2, 110-17.
(1981 b)

261
Vazquez de Espinoza, Antonio

Compendium and Description~ the West Indies (1629).

Washington, D.C.: Smithsonian Miscellaneous Collection.
(1942)

Zuidema, R. Tom

"Catachil lay: the role of the Ple iades and of the Southern
Cross and Alpha and Beta Centauri in the calendar of the
Incas," in Ethnoastronomy and Archaeoastronomy in the American
Tropics, ed. A.F. Aveni and G. Urton. Annals ofthe New York
Academy of Sciences, 385. New York, pp. 203-29. (1982a)

"The Sidereal Lunar Calendar of the Incas," in New World

Archaeoastronomy. Ed. A.F. Aveni. Cambridge: Cambridge
University Press, pp. 59-107. (1982b)

"The Inca Observation of the Solar and Lunar Passages through

Zenith and Antizenith,n in Archaeoastronomy in the Americas,
ed. R. A. Williamson. Los Altos: Ballena Press?Center for
Archaeoastronomy Cooperative Publication, pp. 319-42. ( 1982c)

"Las Pleyades y la organizaci6n politica andina." Paper

presented at the meetings of the Asociaci6n Peruana de
Etnohistoria, Lima, 1981. To be published in the Acts of
the Asociaci6n. (Ms.)

Zuidema, R. Tom and Gary Urton

"La constelaci6n de la Llama en los Andes Peruanos."

Allpanchis Phuturinqa, 9, 59-119. (1976)

262
 APPENDIX

Reconstruction and transcription of Madrid pp. 77-78

by M.Z. Alvarado.

p. 78 p, 77
Line 1

IMIX IX AKBAL KAN CHICCHAN CIMI MANIK LAMAT MULUC OC CHUEN EB BEN

~®@~~©~~~®~~GB
... ·······-· ......... ====
• •• •••

- ... -•• -•
Line 7

•••
Q
260
~
60
~
140
~
220
~
••
40
-
•
~
120
-- •--••
~
200
Q
260
s:._-..:w

~
100
- - - -
••••
~
180
--
••• •• •••
~
260
~
240
~
160

... --•• -• ••- -•

Line 9

•
:Ill(]
380
•• ..
~
80
--, ~-:::j
~
160
~
240
~
220
Q
140
~
220
••
~
40
~
•

120
I
.:--:..
,--7
-.:;
~

200
· -- -- -
•••• ••• ••••
~
280
~
360
~
180

A B C D E F G H J K L M

p. 78 p. 77

263