Marine Geology
Marine Geology
PROFESSOR OF GEOLOGY
Ph. H. KUENEN UNIVERSITY OF GRONINGEN
THE NETHERLANDS
Index 553
c H A p T E R o N E
is also slight (roughly YJo gram per liter). Yet the total amount
present in the oceans is 25 times the entire stock in the atmosphere
(60 X 10 12 tons as against 2.2 X 10 12 tons).' The chemistry of carbon
in its various compounds in sea water is highly complex and cannot be
treated here. The reader is referred to Sverdrup ('t al. (1942) .
No sooner, however, are surface waters carried away from direct
contact with the atmosphere by plunging currents than alterations in
the amount of the two gases, oxygen and carbon dioxide, begin to
develop. Photosynthetic activity of plants under the influence of
light causes the extraction of carbonic acid and the delivery of oxygen.
This process may ultimately lead to oversaturation with oxygen. As
the sunlight penetrates only to a few dozen meters' depth with suffi-
cient intensity to render plant life possible, the "euphotic" zone, in
which oxygen can increase, forms no more than a thin film covering
the ocean. At lower levels two processes tend to reduce tlle amount
of free oxygen. In the first place animals extract this clement from
the water by respiration while giving off carbon dioxide. In the
second place a similar exchange is produced oy the rotting of organic
matter.
Roughly speaking, the longer a body of \I'arer has been withdrawn
from contact with the surface the lower the oxygen content will be,
cete,.is pmibus. But as the original content itself is variable and as,
moreover, biological processes are nor of equal intensity everywhere
throughout the oceans, so that varying amounts of oxygen are con-
verted to carbonic acid in different places, no accurate value for the
"age" of the water can be deduced from the gas content.
The marginal layer between two bodies of water moving in opposite
directions is almost at rest, except where turbulence is strong. As the
consumption of oxygen appears to be independent of the oxygen
content, the more slowly a mass of water moves over a certain distance,
the more oxygen will be withdrawn from it, other things being equal.
Since only slight differences in gas content occur, and since large
bodies of water are involved, normal diffusion scarcely influences the
distribution of gases. Only diffusion by turbulent motion (eddies,
internal waves, and movements of swimming animals) brings in a sup-
ply of oxygen and precludes entire depletion. Consequently a low
percentage of oxygen should tend to mark such a stagnating boundary.
1 Compared with the amount of carbon incorporated 1n limestones of the earth's
cruSt, all the other stores of this element, even those contained in the depositS of
coal aDd petroleum, are negligible.
POORLY V E NTILATED BASINS 9
The poverty of oxygen is often sufficiently marked to aid oceanog-
raphers in tracing the limits of currents in the deep sea. Nevertheless,
inasmuch as low oxygen content may be due to other causes, such as
high consumption, great discretion must be exercised in deducing cur-
rent boundaries from gas content.
Conversely, a relatively high percentage of oxygen indicates the
center of a ventilating current. The deep basins in the Atlantic, for
instance, are replenished by a cold bottom current rich in oxygen
deriving from the surface in the neighborhood of the Antarctic conti-
nent.
in from outside and filled the entire depression of the basin. This
water remains stagnant, while the surface layer is generally diluted by
fresh water coming from the surrounding land. The brackish surface
Sediments
1. Fine grained 6. No bottom fauna
2. Highly reducing 7. Mixed (05sil assemblage
3. H,S present 8. OrganIC matter high
4. CoCO, high 9. Fossils only In certain layers
S. Laminated
Occasional rapid renewal causing catastrophic death
of all organisms
Sediments
1. Fine grained 3. Abnormaf bottom (auna 5. CeCO, hiih
2. Reducing 4. Organic matt.r relatively high
Periodic or continuous renewal
water remains at the top, owing to its lower specific gravity, even
during cold weather. It creeps Ollt slowly to sea, where it is dissemi-
nated and loses its identity. At depths the oxygen is gradually con-
sumed, and as 50011 as it is all used up hydrogen sulfide begins to form.
POORLY VENTILATED BASINS 11
From that time onwards organic life is excluded except for certain
sulfur bactcria. The dC\'elopment of hydrogen sulfide in the botrom
sediment begins at an evcn earlier stage (Fig. 4).
The conccntration of hydrogen sulfide in the water varies from one
place to an ther and depends on the degree of stagnation. In some
fiords, continuous though slow convection takes placc, so that a con-
stant condition is maintaincd with hydrogen sulfide content increasing
towards the bottom. ] n other places no stirring appears to occur
and the amount of hydrogcn sulfide gradually incrcascs from year to
year. Beyond thc threshold aperiodic variations in the properties of
FI<;. 3. A. Basin with local form at ion of basin water and outRow across the sill.
B. Basin with surfate outRow of 10\1" dcnsitv and occa sional ventilation b\' inflow
of dense watc r acro,s the sill. (After Sverdrup ct aI., 1942, T/Je Oceans: Fig. 37,
p. 148, Prentice-i-bll. )
\
1
750
.. Med llerranel'ln II 112 Tho,.
Lat, 36' 56' looS. 2' IS' E
\
1000 !
4. Vertical distribution of temperature, salinity, density, and dissolved oxy-
FIG .
gen in different types of basin. Baltic Sea ::: large basin in hi gh latitudes. Black
=
Sea = stagnant basin. Mediterranean large \'cntilared basin in area of excessive
evaporation. (After Fleming and Revclle, in Trask, 1939, Recent Marine Sedi-
ments, Fig, Q, p. 94, American Assoc iation of Petroleulll Geologists.)
tapped from the open sea. Eastward, the salinity gradually decreases,
reaching one-tenth of normal in the Finnish and Bothnian Gulfs.
The comparatively stable stratification of Baltic waters results in
poor ventilation of the bottom water. In the Bogskar Deep, hydrogen
sulfide is sometimes temporarily observed, and large areas are prac-
tically devoid of oxygen and have a high content of carbon dioxide.
o~----------~~~~~~~~~~--~~------~~
100
200
300
400
500
600
lOOg _lIiiiiiiiiiiiiiiiiiii;______iiiiiiiiiiil_ _ _ _ _ _ _ _ _liiiiiiiiiiiiiiiiiii_
fIG. 5. Section through Kaoe Bay, Halmahera, to Pacific. (True scale section
at bottom .) Showing oxygen in cci i and development of hydrogen sulfide in
depths. (Data gi\"en by van Riel, 1943.)
Hence the deposits are rich in organic matter (up to 670 ). In Chap-
ter 4 we shall return to the sediments formed in poorly ventilated
basins.
-20o__ _
0- -
_40
o
__ 10--
o
_- 20 -
0'
.50
1 ~
yL:l\
WS261
Cl~
p' V CI40 01210
750
VI Hiah latitude WS261 WilUGm
1000
(I , Cooml 01210 Dono
LaL 7' OrN Lona. 7S' 41 ' W
250
1~11 Cl40 r I.,
(' l~
C140
)
1\ WS261
01210
./
V
750
f"
1000 ~
FIG. 7. Vertical distribution of temperature, salinit}', density, and dIssolved oxy-
gen in the ocean. (Aiter Fleming and Revelle, in Trask, 1939, Recent Marine
Sedimems, Fig. 8, p. 93, American Association of Petroleum Geologists.)
cm/yr ~
150
Evaporation r- I
/ r---.. ~ ......
100 "-..,
-/ f- ..... ~
~< '-Rainfall -' '"
50
'"
'-
/ ' ---
-
-......... ./
-
'- ....
V--
i'-..
/' 96.
36
"-- v 35
~ "- ~
./ Salinity-
I 34
GO'S 50' 40' 30' 20' 10' O' 10' 20' 30' 40' 50' N
Latitude
FIG. 8. Distribution of evaporation and rainfall between 40' Nand 50' S over
oceans. High surface salinities are found in latitudes where c\'aporation exceeds
precipitation. Balance between rainfall and evaporation occurs at about 40 ' N
and S and close to equator. (After Fleming and Revelle, in Trask, 1939, Recem
Marine Sediments, Fig. 7, p. 88, American Association of Petroleum Geologists.)
18
PROPERTIES OF SEA WATER 19
blue, the tints change to intensive green; gray, brown, reddish, and
yellowish shades may also be observed (see Schott, 1942, Table XV,
for the coloring of the Atlantic Ocean) .
The color of the sea is also strongly influenced by cloudiness, waves,
and the position of the sun both as to height above the horizon and
direction with relation to the observer. The height of the sun is espe-
cially important, owing to the influence on reflection. When the angle
of incidence is 90 about 4-7% is reflected, but with an angle of 30
reflection amounts to some 2070 . In shallow waters the bottom also
shimmers through and influences the color that one observes.
In some regions the silt suspended in the waters has a marked influ-
ence on the color. Thus the Yellow Sea is named after the vivid hue
of the mud introduced by the Hwang Ho. The Amazon and Congo
PROPERTIES OF SEA WATER 23
cause a reddish color to distances of several miles from their mouths,
owing to the lateritic silt these huge tropical rivers contain. The Po
and Adige carry turbidity to the coastal waters, causing the Adriatic
Sea to lose its intensive blue color around their deltas.
Diatoms that thrive in untold numbers in the polar seas lend an
olive-green tinge to the water. Small jellyfish may lead to a brown
color, and Copepoda and algae to blood red. A milky white is also
observed from time to time.
Compressibility of Sea Water. Water is generally considered to be
incompressible. In comparison with gases, or when only small pres-
sures are considered, this is approximately true. In dynamic oceanog-
raphy, however, the effect of high pressures is found to be of great
importance as they cause compression. Although the ensuing altera-
tions in density are slight they playa decisive part owing to the huge
masses of water involved and to the often labile stratification in which
they occur.
A very important aspect is the rise in temperature due to compres-
sion, and vice versa. Where water is subject to vertical displacements
without the addition or abstraction of heat, the temperature is changed.
These changes of temperature are termed adiabatic. In a deep basin
that is filled from over the edge by inflow of water with a constant
temperature, a higher temperature will be found along the bottom
than at the intake. In spite of the higher temperature this bottom
water will show no tendency to rise, because every vertical movement
is compensated by an equivalent adiabatic change in temperature. By
expanding, a rising mass of water would automatically acquire the
same temperature as the surrounding masses in the new level and
finally revert to the original temperature of the inflow when it had
reached the level of the rim. Hence no energy would be set free to
overcome the friction due to the rising motion.
When oceanographical computations on the density of water masses
are carried out for ascertaining the degree of vertical stability, the
adiabatic influence must be eliminated. The concept of "potential
temperature" has therefore been introduced. By this term is meant
the temperature that would prevail if the pressure were released with-
out transfer of heat. In the example mentioned above, the water in
the basin would everywhere show the same potential temperature,
equal to that of the inflow. Hence the adiabatic rise in temperature
must be subtracted from the temperature in situ (read from the ther-
mometer) to obtain the potential temperature.
24 PHYSICAL OCEANOGRAPHY
FIG. 10. Diagram illustrating difference in shape of Antarctic and Arctic ice-
bergs. Moraine enclosed in latter.
CURRENTS 1
force exerted by the wind and the Coriolis force at right angles to the
movement are then equal and exactly opposed in direction. This state
is closely attained in nature. Yet there is some slight friction, and the
resulting force plus the Coriolis force is equal and opposite to the
force exerted by the wind. Figures 11 and 12 show these relations
and also the case in which a slope of the surface is the motive power.
It may be assumed that, after a certain time (a few days to many
weeks) after the wind has set in, a constant motion without further
-h2
V
-h3
-h
F C '
A B C D
FIG. I!. Balance of forces in ocean current due to slope of surface. V = veloc-
ity and direction of current, b 1 - b 4 = surface contours, Co :: Coriolis force,
F := frictional force. (A) in shallow water: the bottom friction is large and the
deflection by Co is relatively small ; (B) in deep water: the friction with deeper
layers is small and the deflection large ; (C) near the equator: Co is small; (D)
assuming no friction: the current flows at right angles to the surface slope .
acceleration has been reached. The primary and secondary forces
are then nicely balanced and a dynamical equilibrium under constant
conditions has been attained. Hence the larger systems of currents
that are driven by constant winds should show unaltering conditions.
Obviously this ideal is never attained, because all winds, even the
trades, are subject to variations in strength and direction.
When the wind just begins to blow, the first movement of the water
will be in the same direction as the wind, but as soon as motion of the
water sets in the rotation of the earth makes itself felt. In shallow
water, where the friction with the bottom is considerable, the current
will be less deflected by the Coriolis force than in the ideal case. In
regions like the North Sea and the Baltic, where the winds are variable,
where the water is shallow, and where the surrounding coasts are at
no great distance, wind-driven currents will generally follow the direc-
tion of the motive force fairly closely. In dealing with the tides of
CURRENTS 31
the former Zuider Zee in Holland the effects of wind will again be
referred to.
Along the Pacific coaSt of the United States the winds blow parallel
to the coast in a southeasterly direction. This causes a current with
a component directed away from the coast, because of the Corio lis
force directed to the right of the flow.
Even if the wind were directed obliquely
towards the coast the surface waters would Near equator At 45'
W W
still be gradually driven off shore and the
surface of the sea would be depressed along v
the coast. This suction causes deep water
to well up to the surface from depths of
100 or 200 m. Lower temperatures and a
supply of plant nutrients therefore char-
acterize the coastal waters off California.
For reasons to be discussed later there re-
mains a marked component in the direc-
tion of the wind.
Similar upwelling occurs along other
FIG. 12. Balance of forces at
coasts, as the west coast of South America surface in steady dtift cur-
and of Africa. The movements should rent set up by wind. W =
nOt be thought of as strongly developed stress exerted by wind, F =
currents, for there is merely an upward stress exerted on uppennost
component to much stronger horizontal water layer by that immedi-
ately heneath it owing to
flows. In the case discussed above, the difference in velocity and di-
vertical movement is of the order of 50 m rection, C == Coriolis force,
per month and, moreover, is intermittent. V = direction and velocity
On nearing the equator the Coriolis of current at surface, VI =
force decreases more and more until the direction and velocity of
layer JUSt below surface.
current approaches the direction of the
wind or flows at right angles to the surface
contours of the sea. At the equator the influence of the rotation of the
earth on the direction of the current is zero.
To give an idea of the strength of the currents developed it can be
stated that a wind of force 5 of the Beaufort scale at 45 latitude will
result in a current of about 15 em per second.
In spite of the great advances made in the last few decades with
respect to the calculation of currents, it has not yet proved possible
to explain completely the currents produced by constant winds such
as the trades and those of the "roaring forties" of the southern hemi-
sphere. Movements are found down to considerable depths having
32 PHYSICAL OCEANOGRAPHY
the same direction at all levels. This docs not agree with the theory
advanced above and is probably due to obstructing masses of water
and to macroturbulence generated by the wind.
In order to calculate currents due to unequal distribution of mass,
determinations of temperature and salinity at all depths are required,
preferahly at several stations not situated along a straight line. It is
then assumed that, below a certain depth, uniformity in pressure
prevails in any level plane and consequently a stare of no morion.
Further uniformity of pressure at the surface is postulated. The dif-
ference in weight of the columns between two levels must yield the
force to maintain the current.
If there were no friction and no accelerations the current would
flow parallel to the contour lines of the isobaric surfaces with the
Corio lis force at right angles and opposite in direction to the resolved
force of gravity. From this assumption it may be deduced that in
the northern hemisphere the lighter water lies on the right-hand side
of an observer looking in the direction of the surface current; in the
southern hemisphere the denser water lies to the right. This rule
explains why there must be a component of the current in the direc-
tion of the wind parallel to a coast, as discussed above.
Where the data on density are insufficient in amount or accuracy
to permit direct calculation of currents, a different procedure may be
followed. The distribution of temperature, salinity, or some other
variable is studied, and the investigator seeks to esrablish from a map
how certain masses move. Examples will be given later when treating
the circulation in basins.
The differences in density that cause currents should gradually dis-
appear in consequence of these very currents. Therefore in the case
of perm~nent flow, such as most density currents, external influences
must act continuously to maintain the differences. These influences
are the climatic factors already referred to. Many times these cause
vertical movements also, although the horizontal components generally
prevail. Lessening of the density at the surface has little direct influ-
ence on vertical movements, but an increase in specific gravity causes
the surface waters to sink. Thus dense surface water is developed by
strong evaporation at low to middle latitudes. However, as this water
is warm it cannot sink to great depths. The cold deep water beneath
will have an even greater density. Only in partly closed basins, that
will be dealt with later, can water of exceptional density be generated
by evaporation.
CURRENTS 33
The chief source of deep water filling the ocean basins must there-
fore be sought in polar regions. In the fall and winter intensive
cooling takes place and an increase of density occurs down to the
freezing point at - 2 0 C. The increase may continue even when the
formation of comparatively fresh sea ice adds to the salinity of the
remaining water. Under these circumstances sea water with the
greatest density is produced.
Circulation in the Deep Sea. A vertical section from north to south
through the Atlantic Ocean will be discussed in order to illustrate
some general features of deep-sea circulation. In the first place it
should be noted that the section (Fig. 13) is greatly exaggerated in
height, about 100 times. In a horizontal direction the properties of
the water vary extremely slowly, because each level is subject to
similar conditions over very wide areas. To reach a region with a
different climate distances of hundreds or even thousands of kilometers
must be covered. But in a vertical direction considerable variations are
encountered over short distances. Radiation penetrates only to shallow
depths. Gra\'ity tends to callse a stable stratification, with the heaviest
water at the bottom. For example, the vertical temperature gradient
in the Pacific is roughly 6000 times as great as the horizontal, and for
sa linity the same rule holds. In order to show both the vertical and
horizontal variations in the same figure, therefore, the vertical scale
must be greatly exaggerated. Moreover, the shape of oceanic basins,
relatively shallow in comparison with the enormous extensions, would
in itself force us to adopt the reduction of the horizontal scale.
In the second place our section clearly demonstrates that currents
may be deduced not only from calculation but also from the distribu-
tion in space of the properties of the water masses. The velocity can-
not be ascertained by this method, but the direction and extension of
the flow are frequently obvious. Consequently the major oceanic cir-
culations with vertical components are known qualitatively, although
quantitative calculations are not yet available.
In the third place the great predominance of the horizontal over the
vertical component of the major oceanic circulation currents follows
directly from our section, for even in this much-compressed represen-
tation the greater extent of the horizontal movements is still obvious.
Finally it becomes clear on inspection of our figure that the sinking
currents are in general much more clearly defined than the correspond-
ing returns to the surface. This is due to the fact that downward
flow takes place in more or less well-defined currents, whereas the
displacements back to the surface, that must, of course, be of exactly
34 PHYSICAL OCEANOGRAPHY
c
o
z .
o
'0
:;
u
....
'u
....
o
'"E
cu
~ '~ ~~~~~~---r7-~~
cutV
~al
~.fl" ,5 "r
VI ~,\--\---I~--+---=:--t--::~
.1l~~~L~=~~
CURRE NTS 35
equal volume, take place by mixing with overlying waters and are
therefore more diffuse.
Viewing the section in more detail, the most conspicuous featUre is
the great deep current that sinks in the north off Greenland and may
he followed southwards far across the equator. As explained above,
the polar surface water in the north sinks during the winter, so that
the current receives a pulsating impetus. But owing to the vast mass
1"<0. 14. The spreading of Amun:tic bOrtOill water into the deeps of all major
oceans. (Redrawn from Dcfum, 1940, Fig. 3, p. 6.)
FIG. 15. Schematic rcprescnration of the currents and water masses of the Anr-
arctic regions and of the distribution of temperature. (After Sverdrup et aI.,
1942, Tbe Occam, Fig. 164, p. 620, Prentice-Hall.)
FIG. 17. Left: Observed temperatures and salinities in the Straits of Florida.
Right: Velocities of the current through the Straits according to direct measure-
ments and according to computations based on the distributions of temperature
and salinity. (After Sverdrup et aI., 1942, The O ceans, Fig. 184, p . 674, Prentice-
Hall; according to Wiist.)
The South and North Equatorial Currents are the direct result of
the trade winds. The northern current carries the water along the
northern shores of the West Indies as the Antilles Current, but it also
forces its way into the Caribbean Sea and thence into the Gulf of
Mexico. The water is thus raised 19 cm in the Gulf. It flows out
through the passage between the peninsula of Florida and the Bahamas
40 PHYSICAL OCEANOGRAPHY
back into the Atlantic as "jet" current. This current in the Straits of
Florida shows an average surface velocity of no less than 6 km per
hour (2 m per second) and sometimes attains a maximum of 10 km
per hour (3 m per second). At a depth of 250 m a current of 4 km
per hour has been measured. The transport of water is about 26 mil-
lion cubic meters per second (Fig. 17).
Outside the Straits of Florida the current joins the Antilles Current
to form the Florida Current that flows along the east coast of North
America. The current reaches to
depths of no less than 600 to 800
01, whereas drift currentS arc only
100 to 200 m thick. This is prob-
ably due ro the fact that the Flor-
ida Current is not a pme drift
current iJut is also caused by dif-
ferences in level. Locally veloci-
ties of 160 ern per ~e~ond arc
reached at the surface, and the
total transport is estimated at 40
X JOG m a per second.
Off Cape Hatteras the current
diverges f10m the continental
Fl(;. ~ 18. S c he11l~tic representation of
the char:lcter of the Gulf Stream and slope. Here the name Gulf
accompanying eddies on slope. Inset: Stream is used. Complicated giant
Profile of the sea surface along the line gyrals appear to be developed es-
A -B. (After Sverdrup et aI., 1942. The pecially on the left, landward side
Oceans, Fig. 185, p. 680, Prentice-Hall.) of the Gulf Stream (Fig. 18). To
the sotltheast of Newfoundland
the Gulf Stream meelS the cold Labrador Current flowing south
through the Torres Strait (Fig. 19). From thence onw;lfds the Cllf-
rent is driven across the Atlantic as the North Atlantic Cnrrent bv the
prevailing southwesterly winds. The average velocity to Ireland'is 1h
to 1 km per hour. A part is here diverted to the south to pass along the
African coast as the Canary Current and thence to curve around the
Sargasso Sea and merge imperceptibly with the North Equatorial Cur-
rent. The other branch passes north, washes the coasts of Ireland,
Scotland, and Norway, and loses itself in the Norwegian and North
Polar Seas.
Although meteorological matters are not treated in this survey. we
must recall here that the warm water of the North Atlantic Current
influences the climate of northwestern Europe in a decided manner.
CURRENTS 41
FI~. 19. Schematic represe ntation of the mixi ng of the Labrador and Gulf Stream
currentS.
Flo. 20. Surface currents of the oceans in February-March. (After Sverdrup et a)., 1942, Chan Vll. Based on Goode Base Map No. 201HO. Copyright by the University of OUcago. Used by permission of the University of Chicago Press.)
CURRENTS 43
is of the same density. Here the Mediterranean water leaves the bot-
tom and spreads out north, west, and south, forming a layer that can
be recognized by the combination of high temperature and high
FIG. 21. Section over the sill in the Straits of Gibraltar. showing temperatures.
(Mainly according to Schott, 1942.)
sa linity. It has been traced beyond the equator in the south, the
Azores in the west, and Ireland in the north by oceanographical serial
measurements (Fig. 22).
N
tI 0Kilometers1000
FIG. 22. Chart showing the outward spread of Mediterranean waters at looo-m
depth in the Atlantic (temperatures and salinities) . (Mainly according to
Schott, 1942.)
1111111111111
430
PHYSICAL OCEANOGRAPHY
55,600 55 , 600
Station
48 39 23 3 14 35
O~~--~ ~~--~~---+----~------~--~--~--~--,
10
20 -~~~~~~_
30
~ 40
W
E_ 50
.<::
g. 60
o
70
80
90
100
F,c;. 24. Section along the Bosporus showing densities for September-October,
1917. (After Oefant, 1929, Fig. 16, p. n.)
FIG. 26. Chart of the Moluccan deep-sea basins showing directions of principal
ventilating bottom currents. (According to van Riel, 1934, Plate IV.)
current basin. This grade provides the force to drive the flow through
the narrow entrance. Thus a slightly higher temperature is encount-
ered in each successive basin even though the sill may lie deeper in
some cases than that of the preceding trough (Fig. 29).
CURRENTS 49
The Halmahera Basin has its own system of ventilation directly from
the Pacific.
The flow towards the central and southerly basins of the Molu ccan
region is found to be highly unexpected. Only the Timor and Aroe
F,G. 27. Section from Pacific CO Celebes Sea, showing temperature distribution
and ventilating currenrs. Relo'" true scale section . (Redrnwn from van Riel,
1934, Fig. 8, p. 24.)
Troughs receive their deep \\'atcr from the Indian Occan over a sill
of 1940-m depth to the south of Roti. But the Buru Basin, Ceram Sea,
and Banda Ba~ins all communicate with the Moluccan Passage over
a sill at 1880 m between Obi and the Sula Islands while the deeps to
the south of Buru and between this island and Ceram are fed from
the Banda Basin. The most surprising feature, however, is that not
only the Weber Deep between the Inner and Outer Banda Arcs
50 PHYSICAL OCEANOGRAPHY
SJ313W
o
'" M
tuJ
c:
.~
o
..c:
'"
I 1 1 I
1& bb
!.o
'"
b_
'"I ,, 'ry
, 1
1 1 1 1
CURRENTS 51
receives Pacific water bur that even the Savu Sea is supplied from
this source via the Weber Deep. The Flores and Salajar Troughs and
even the Bali Basin to the north of Bali are all fed from the Pacific.
The deeps of all these basins are cut off from the neighboring Indian
Ocean by shallow sills that shut out all deep water effectively, in spite
of the short distance across the submarine ridge connecting the Lesser
Sunda Islands. The 3000-km path from the Pacific is followed because
the greater depth allows colder and heavier water to pass this way.
This system of currents was deduced from the charts showing where
the deepest entrances exist and from the distribution of temperature
and salinity that demonstrates how water masses are connected. Excel-
lent confirmation is found in the oxygen content of the bottom water.
It was pointed out above that the deep-sea waters are replenished with
oxygen in the polar regions ("the lungs of the ocean"), where the
saturated surface waters sink and start their long wanderings through
the deeps of the oceans. If the picture obtained for the ventilation
of the basins in the Moluccas through the Moluccan Passage is correct,
the waters richest in oxygen should be encountered in the entering
oceanic currents. On the passage of the current through the Moluc-
can troughs and basins the oxygen would be gradually consumed and
the percentage should decrease along the lengthy route followed
through the successive deeps. This is actually found to be what
happens.
In the south a tongue of water rich in oxygen leaks into the Timor
Trough. The oxygen coment gradually diminishes from 3.4 cms per
liter to 2.2 cm s per liter in the Aru Basin. The flow southward
through the Moluccan Passage can be recognized by an oxygen con-
tent of 3.0 cm s per liter. A branch turns off to the west and finally
runs up north along the coast of Celebes. It is also clearly seen how
a gradually impoverished flow with a content of 2.5 cms per liter
passes into the northern Banda Basin. The offshoots in the Gulf of
Boni in southern Celebes and the Flores Sea contain only 2.3 cm s per
liter, and in the Bali Sea the oxygen has sunk to 2.0 ems per liter.
Finally in the Savu Basin a central mass is found where the oxygen
has been depleted to 1.7 cms per liter.
It is curious to note that the water flowing into the Celebes Sea
immediately attains the low oxygen content of only 2.3 ems per liter.
In a wide sweep it circles along the northern and western margins of
this large basin and passes southward into the Makassar Straits. It
finally abuts against the shallow banks in the south with a content of
52 PHYSICAL OCEANOGRAPHY
FIG. 30. Chart of the Moluccan deep-sea basins showing the oxygen content of
the bottom water in cell (according to van Riel). Note the influence of the
rotation of the earth combined with the centrifugal force as denoted by the
arrows giving the main direction of the ventilating currents: right-hand curve
north of equator, left-hand south of equator. (After Kuenen, 1948, Fig. 1.)
See also Fig. 31.
Finally the poor communication of the Sulu Basin with the South
China Sea is demonstrated by the extremely low oxygen content, which
in the southern deep is less than 1.5 cm8 per liter and locally sinks
even below 0.6 cm3 per liter. Here a stagnant condition is approached
in which the development of hydrogen sulfide begins.
CURRENTS S3
Kuenen has traced the apparent influence of the Coriolis force in
the distribution of oxygen (Figs. 30 and 31). The right-hand curve
in the flow through the Celebes Sea to the north of the equator is
striking, and the current passing from the Pacific into the Moluccan
Passage rises high up against the Snellius Ridge north of Halmahera
on its right-hand side. Thrusting towards the left in the southern
hemisphere is marked in the Gulf of Bani, Weber Deep, Sanana Strait
northwest of Buru, and especially in the Flores Trough where a loop
flc. 31. Sections through the troughs of Fig. 30, showing the position of maxi-
mum oxygen content deflected to the rig ht north of equator and to the left
south of equator by the force of Coriolis and the centrifugal force. Sections
seen in direction of the current (I, II) directed away from observer, (II ) addi-
tional centrifugal force, (III) towards observer, (A, B ) north of equator, (C-I)
south of equator. (After Kuenen, 1948, Fig. 2.)
passes southward. From the chart the cross section of this current
can be estimated at 6 km 2 '
Lek (1938), who studied the current measurements of the Snellius
in this strait, concluded that at 1500-m depth a current of 5 cm per
second is directed towards the Buru Basin (Fig. 32). More measure-
ments would be needed for a reliable value, however, because the
average velocity for the entire section is required. It has been pointed
Land
E3
0200
~
2001000
~
1000- 2000
III
20003000
_> 3000
0'1'020
Kilometel'$
FIG. 32. The sill of Lifomatola Strait, nonh of Buru, showing position and direc;-
tion of bottom current ventilating the Banda Basin. Depths in meters.
out above that current measurements made from an anchored ship have
inherent errors due to yawing. For a preliminary estimate of the
average velocity more than the measured value must be assumed, be-
cause the velocity was obtained near the upper limit of the current,
and the pressure gradient increases downward. Assuming double the
measured value, a velocity of 9 km per 24 hours is found, resulting in
a flow of 54 km 8 of water daily.
This flow must ventilate basins having a combined volume of 1,800,-
000 km 3, the estimated size below sill depths. A complete replenish-
ment would take roughly 100 years. For the wider portions of the
basins the velocity of the currents would be of the order of a few
dozen meters per day.
CURRENTS 55
It may prove possible to estimate the current velocities through the
other narrow entrances of the archipelago from the slope of the cur-
rent, which is indicated by the distribution of the densities of the water
at opposite sides of the sill. A considerable amount of data is available,
thanks to the soundings and oceanographical stations of the Snellius
expedition. A very rough estimate for the fl ow through Lifomatola
Strait just discussed is 30 cm per second sOuthward. Although this is
more than the 10 cm deduced from current measurements, the order
of magnitude appears to be correct.
Finally the question arises as to ,\'hat the cause of the currents may
be (Kuenen, 1948) . I t is ob\'ious that, if a sill were to be suddenly
lowered in consequence of diastrophic activity, an invasion of colder
water would enter the basin through the deepened inlet. But some
cause must operate to keep the current flowing after the entire deep
has been filled.
Since salinity cannot be altered in the deep, a gradual rise in tem-
perature must be assumed. This ,,'ould reduce the density and allow
new cold water to enter over the sill. Of course the process is con-
tinuous, and dynamical equilibrium prevails. It is not found possi ble
to trace the mechanism of this convection from the distribution of
temperature in the deep. In the oceans the sinking currents are far
more pronounced than the gradual return to the surface, and likewise
in the Indonesian basins there are no well-marked upward flows.
Evidently the movement is diffuse and takes place hy some ty pe of
mixing.
Several factors may be considered to take p:1rt in heating the deep
water: the flow of internal heat of the earth and turbulent mixing with
higher strata, frictional heat, and the consumption of oxygen. The
last two sources are probably considerably smaller than the others and
will be left our of account.
We will assume that the flow of heat through the ocean floor is
equal to that in continental areas, a postulate that still has to be proved.
On this assumption, however, 50 gram calories are conducted annually
upward through each square centimeter of the bottom. There are
reasons for assuming that about an equal amount is carried downwards
into the deeps by turbulent mixing. This turbulence is caused by the
tidal currents that affect all parts of the basins, also at greater depths.
The measurements of the Snellius show a slight rise in potential tem-
perature going upwards in the homothermal deeps. In the Celebes
Sea it amounts to 0.1 0 C. 1 it is assumed on this evidence that the
homothermal waters of about 2000-m thickness are heated during their
56 PHYSICAL OCEANOGRAPHY
TIDAL MOVEMENTS
Tidal Forces.It was shown by Newton that the tidal force is due
to the attraction of the moon and in lesser degree to that of the sun.
Since then many investigators have contributed to tht, development
of the tidal theory.l
If the earth did not rotate, it would still move around the mutual
center of gravity of earth and moon, a point situated within the earth.
The movement would be comparable to that given to a cup of coffee
to dissolve the sugar. Thumb and fingers describe similar circles situ-
ated beside each other. In the same way all points on and in the earth
describe identical circles in the course of one month. The mutual
center of gravity is displaced with respect to the earth and also
describes a circle in one month.
As all points of the earth move in equal circles at the same rate they
are subject to equal and parallel centrifugal forces. These forces are
also parallel to the line connecting the center of gravity of the earth
with that of the moon but are directed away from the moon (Fig. 33).
In addition each particle of the earth is attracted by the moon, bur
the strength of the attraction varies with the distance to the satellite.
In a section through both centers of gravity the attraction on the side
of the moon must for obvious reasons be greater than the average and
that on the opposite side smaller than the average. At the surface of
the earth the attraction of the moon amounts to only 1/ 300,000th of the
gravitational force of the earth. Further, the total attraction of the
moon must be equal and opposite to the centrifugal force, for if the at-
traction were greater the two bodies would approach, and if it were
smaller they would move apart.
On the front side of the earth the larger attraction of the moon
must predominate over the constant centrifugal force, and at the
1 See Thorade, 19i1; Defant, 1929; Schott, 1935, 1942; Sverdrup et al., 1942.
TIDAL MOVEMENTS 57
opposite side the reverse must hold. Hence on the front each particle
undergoes an accelerating force towards the moon, and on the back
one directed away. In both cases an observer would notice a force
directed upwards.
At points outside the line connecting the two centers of gravity the
attraction and centrifugal force are, moreover, not parallel. The paral-
lelogram of forces shows that the resultant is directed obliquely with
regard to the surface of the earth, except at the ends of the perpendic-
Cf. fro
C. gr.
Earth ~---;'; ;-; '-:-; j;:-:::=:::::::::::@
FIG. 33. The tidal force in a section through the line connecting centers of earch
and moon. E + M = center of gravity of earth plus moon. Atr. M. = attraction
of moon. Cf. fro = centrifugal force.
ular to the direction of the moon through the center of the earth.
As all sections through the connecting line are equal, a circle may be
drawn on the earth where the tidal force is directed vertically down-
wards. It is surrounded by a band with tidal forces pointing obliquely
downwards, while in front and behind are areas where the force is
directed obliquely upwards.
The tidal forces are extremely small. The centrifugal force is only
3.38 mg per kilogram; the attraction of the moon varies between 3.49
and 3.27 mg per kilogram. The tidal force is at the maximum in front
and behind, where it attains 0.11 mg per kilogram or 1/ 9,000,000th of
the force of gravity. For the sake of clarity the arrows in our figure
are drawn comparatively large. On the same scale the force of gravity
would be several dozens of kilometers long.
As the tides are related to the moon they run round the earth in a
month (27 .3 days). But the earth also rotates, and it revolves, as it
were, under the system of tidal forces. Twice daily each point on
58 PHYSICAL OCEANOGRAPHY
---B
Equator
P
Flc. J~. Diagram showing cause of daily inequality of tides. The outer circum-
ference denotes shape of ocean level. A and A', Band B', successive positions of
two points on the earth's surface at half daily intervals.
waves, a node is formed in the center off Port Sudan. There the tides
cause only horizontal movements while the height of sea level remains
constant (Fig. 36).
In shallow channels the friction with bottom and walls exerts a strong
influence and puts a brake on the tidal waves. A pure standing wave
then cannot develop, because the reflected wave is too weak to com-
pensate the primary wave. A transitional form between a standing
and a progressive wave is then developed.
In funnel-shaped estuaries the incoming tidal wave is forced into a
gradually. narro'wing channel. This results in increased height, and
40~~~h~a~dw-an------------------------~----~
E 20
u
+--_
301'_ ~Kase" m
4h 32 CT
]h2Smcr-- ...... , /Oschidda
oj 10 .........r Port Sudan
~O /'
C.lO
~ 20
30
40~~~~~~-L~-L~~~~L-~~~~~
o 1500 2000
Kilometers
FI G. 36. Diagram showing tidal range in Red Sea. Crosses show observed 3nlpli-
tudes, the full line the local tide, dotted line the wave entering from the lnc1ian
Ocean. (After Thoradc, 1941, Fig. 33, p. 55.)
the tidal range may ue still further enlarged when the period of free
oscillation in the estuary approaches the tidal period. The tidal range
in estuaries may exceed J 0 m (Bristol Channel in southern England,
and St. Malo, Normandy, ) 2.5 m; Bay of Fundy on the Canadian
Atlantic coast, 15 m). High tidal ranges tend to be accompanied by
powerful tidal currents.
In river mouths the tidal wave cannot be reflected because there is
no abrupt upstream ending to the channel. At the mouth an ingoing
and outgoing current will be observed in conjunction with the tides
at sea. The farther one proceeds up the river, the weaker the tidal
currents will be and the later they will set in.
Moreover, the farther upstream one goes, the greater is the influence
of the current of the river itself; the ebb current lasts longer and is
more powerful. Finally a point is reached where the current no
longer alternates in direction. The flow of the river there continues
day and night towards the sea. But at certain times the velocity will
decrease and at others increase in accordance with the tidal currents
62 PHYSICAL OCEANOGRAPHY
farther downstream. This rhythm in the flow also will abate upstream
until finally a region is attained where the waters continue their sea-
ward flow uninterrupted by any tidal influence.
In some fiver mouths the tidal wave is directly deformed on shallow
banks. The front is steepened until it resembles a breaker on the beach
and is followed by a few powerful oscillatory wa\'es. These may start
to curl over and break, and a foaming wall of water then proceeds up
current. This rush of water strongly resembles the swash on the
beach that will be discussed later.
This wall of water is followed by several normal waves. For rea-
sons to be dealt with later the steepness of the front is maintained, and
an observer, instead of witnessing a calm turning of the tides and
gradual rise of the surface, sees a wild, roaring breaker rushing up the
river at great speed. On a small scale this phenomenon IS developed
in the Severn and Trent in southern England and is there called the
"bore." In the Seine and Gironde the term "mascaret" was applied
to the bore, when it attained noteworthy dimensions. Since the shoals
in the entrances to these French rivers have been dredged away, the
mascaret has almost disappeared. The Amazon and the Tsientan in
southern China show the !Jore on an enormous scale. With indescrib-
able force a thundering wall of water several meters high dashes up
the riyer, attaining velocities of 6 to 7 In per second. The roar can
be heard at a distance of a few dozen kilometers, and this type of wave
can he very dangerous to shipping. The greatest height observed in
the Amazon is 5 m, and in the Tsientan even this figure is surpassed.
Bores on a miniature scale are sometimes developed by waves enter-
ing elongate pools on the beach from one end.
The direct tidal force develops only a very small range in the North
Sea (Fig. 37), but strong tidal waves enter from the Atlantic. In con-
sequence of the shallowness and complicated shape of the basin the
tidal waves are strongly deformed and forced up to great heights
locally. The wave entering through the Straits of Dover is of little
importance, but through the broad entrance in the north a powerful
wave passes into the North Sea. Along the Norwegian coast the range
in height is small, locally not more than 25 cm; the amplitude along
the British coast, however, is 4 m, and at certain points, as in the Wash,
it reaches 6 m. Along the Dutch coast the height is 2 m in the north,
1 m in the center, and greater again in the south, with a maximum of
4 m.
These great vertica l movements are coupled with strong horizontal
currents. Over extensive areas currents attain as much as 2 m per
TIDAL MOVEMENTS 63
second. In the Bristol Channel spring tides give rise to a current of
4% m per second. and between the Channel Islands velocities of 4 m
per second are reached with dangerous whirlpools. These are veloci-
ties hardly exceeded on navigable rivers. Although the velocity along
Tides
in th e
North Sea
- - - - Amplitude
- - Highwater In
\ h. after culm.
'. moon 31
"2m GreenWich
lOW. 5 lifO.
FIG. 37. Tides in the North Sea. (After Schott, 1942, Fig. 62, p. 218.)
Contours of
water suri3ce
- - - Hall meters
...... ... . Declmeter!t
FIG. 38. Chart of the Wadden Zee, north of the former Zuider Zee in Holland,
showing tidal currents and contoured sea level on May 31, 1919, 2 hours after
culmination of moon. (Redrawn after Lorentz and Thijsse, 1926.)
FIG. 39. Chart of same region as Fig. 38, showing calculated wind-driven and
tidal currents for the storm of December 22- 23, 1894, at 3 o'clock in the morning
if the dyke had then existed. Velocities in centimeters per second. Outgoing
currents dotted. (Redrawn after Lorentz and Thijsse, 1926.)
in some places exceeding 1 m per second (Fig. 39). The total flow
through the narrowest part forming the inlet into the basin of the
Zllider Zee attained 210,000 m3 per second, or double the maximum
flow of the Mississippi. Thus the wind-driven currents are found to
equal the tidal currents in velocity.
FIG. 40. Chart of the former Zuider Zee with contoured sea level during t he
stonn of November 6-7, 1921. Heights in centimeters above normal sea level.
(According to Lorentz and Thijssc, 1926.)
-15 - 10 -5 o 5 10 15
-7 --
A 5 4
5
=
r=7' r-- r-._
r- r- -.
~ :.._
~
--
i'-..
o
';-.. r-- / -- r- "K
-5
:-(- t-- r- r--- blo
IO II
Om
-5 o 5 -5 o 5 - 5 o 5
B C D
8
5
7. 8 6) 19 6~ i'..4
6'.\9 s..~10
~I
7~ I'.
o I~ ~IO 00 8~~
4'~1 ;t Lt-O
I ~~2
-5
3\ 1\0 '\
~I
2 I 400 m I'V ~O
100 m
"
-10 -5
+ em/sec
o
em/sec
5 10
3000 m
E
I 0 23
5 2
3( "'~..._ ~ 21
4~ ..._ ~
o ""- 19
_, 18
~~
-5
go; '"
'iO"- ~
II 12
. 17
~~
3
Om
FIG. 41. Current directions and velocities in centimeters per second for half-
daily and daily tides at various levels in meters, deduced from measurements by
the Snellius expedition, anchor station 364, Weber Trough, southeast of Ccram.
(Redrawn after Lek, 1938, Fig. 66, p. J 35.)
WAVES 69
coast is called the beach.l The beach is covered by normal high tides,
the dry beach only by storm waves. The beach can be quite narrow
and disappear completely on steep rocky cliffs. Sandy beaches are
generally not steep, slopes between 1 and 3 being normal. _Th_e
Q Q
WAVES 1
Oscillatory Waves. Waves form the third and last type of movement
to be treated.
A wave is described by its length, i.e., the horizontal distance from
crest to crest, and by its height, i.e., the vertical distance from trough
to crest. A wave is furthermore characterized by its period, i.e., the
time interval between the appearance or development of two consecu-
tive crests at a given position.
The simplest form of wave is found in deep water, after the wind
has died down. Then the waves are symmetrical with relation to a
vertical plane through the crest, and according to the theory of Gerst-
ner the particles revolve in circles (Fig. 42). The size of these circles
decreases with depth; at about one-ninth of the wavelength the diam-
eter has decreased to half, at two-ninths it is a quarter, at one-third
it is an eighth, and so on. Evidently a level is soon reached at which
. no perceptible motion occurs. It follows that a short wave dies out
downward sooner than a long one, even if the amplitude is greater.
1 Extensive bibliography of older observations in Johnson (1919) . See al.so
Thorade in: Dcfant (1940), Sverdrup et aJ. (1942), Hydrographic Office (1944) .
Bigelow and Edmondson (1947) .
70 PHYSICAL OCEANOGRAPHY
L
FIG. 42. Movement of wat~r particles in a deep-water \\aye of smu ll height. The
circles show the paths in which the particles 111()\e . The size of the orbits bel ow
the surface is sbown to scale, and tbe lower wavc lin e shows the amount of
vertical mO\ement at that level. The dashed WU\c profile and dashcd nearl y
\ertical lines show the position one-cighth of a period later than the elluivalent
full-drawn curves. L = wavelength.
narrow and steep; the trough is broad. Still water level is therefore
closer to the level of t he troughs than to that of the crests. In shallow
water, about three-fourths of the wave height, varying with the cir-
cumstances, is above still water level. Ac-
cording to theory the height may attain 6
one-seventh of the wave length, but in na-
ture the crest has often become unstable
and has started to topple over before this
degree of steepness is attained (for an ex-
planation see p. 82). Seldom is a wave
higher than one-twelfth of the length, and
the ratio of 1: 20 or 1: 30 is frequent. But
if two sets of waves bisect, pyramidal waves
develop that arc steeper than 1: 12.
The relation between wavelength L, ve-
locity V, and period T is V = LI T. With
V in meters per second, L in meters, and T
in seconds we have, for deep water (depth
> Y2L),
V = I.56T; L = 1.56P; V2 = I.56L
A characteristic of great importance to
the movement of suspended particles is that
a certain small mass transport of water
takes place in the direction of propagation.
A water particle moves in the direction
of progress when it is above its mean
,lepth, and in the opposite direction when
it is below its mean depth. Owing to the
decrease in velocity with depth, however, FIG. 43. Schematic represen-
it moves some'what faster in the direc- tation of successive shapes
tion of progress than in the opposite di- of a column of water in a
deep-water wave. (Mainly
rection. After having completed one rev- according to Thorade, in
olution in its orbit, the particle is advanced Defant et al., 1940, Fig. 51,
somewhat. This means that even in the p. 12),)
absence of wind an ::lcmal mass trans-
port of water takes place in the direction of propagation (Fig. 44).
The velocity of mass transport :s appreciable for high, steep waves
but is very small for low waves of long period.
72 PHYSICAL OCEANOGRAPHY
The surface layer of the water moves forward in the crest, and
energy is transferred from the wind to the wave. In the trough the
surface layer moves backward, against the wind, and energy is lost.
If these movements were equal the two effects would compensate
each other. But, as pointed our above, there is a slight advance of the
surface particles. Hence the transfer of energy from the wind to the
wave predominates. This is termed the pull or drag of the wind .
Even when the wave fonn moves faster than the wind and the push
has become negative, the drag can continue to transfer energy to the
waves because the water particles advance very much more slowly
than the wind. If this were not so no satisfactory explanation could
be given of the fact that waves sometimes have a higher velocity than
the wind producing them. According to computations by Jeffreys
this influence does not start until the wind has attained 4.8 m per
second.
Finally the wind acts on the water by suction and compression, as
was shown by Helmholtz and by Lord Kelvin. Over the crest the
wind is compressed and its velocity increased; hence the pressure must
be lowered. Conversely, an extra pressure must develop in the wave
74 PHYSICAL OCEANOG RA PHY
fIG . 47. Schematic representation of stream lines over a ",ave showing the suction
and pressure exerted on the water.
deflect the wind and thus increase the irregularities in the motion of
the air. Moreover, the wind almost always finds a surface on which
a system of waves is already present, but a system that does not con-
form either to the strength or to the direction of the wind. In this
manner two or more systems of waves are superimposed one on
another.
When the waves have grown to a larger size the wind continues to
generate smaller ones. But these continually pass into the shade of
the larger ones and are also flattened out by the spray driven off the
crests of the bigger waves. Hence, they receive insufficient energy
to grow, and they die away again. This is the reason why in a general
way the larger waves predominate, instead of all sizes down to the
smallest being represented.
The modern method of investigating wave shapes is by sLereoscopic
photography, using two cameras with a base several meters long.
The two photographs can be combined to give a chart with surface
contours of the area covered by both. Among others the Meteor
expedition brought home some exposures that have been worked out
with later additions in an atlas showing contoured charts (Fig. 50).
These maps show striking irregularity of the sea surface. In many
it is with great difficulty that any definite system of wave ridges can
be trattd. It is also apparent that a systematic distortion of the rear
part of the photographs is introduced by this method. Vertical stereo-
scopic exposures taken from a plane would therefore present great
advantages over these obliques.
The shorter the wavelength and the steeper the crests, the sooner a
set of waves must Jose its energy during propagation or be deflected
and suppressed by gusts and eddies of the wind. Moreover, the waves
always tend to lengthen during propagation. Hence the longest
waves developed in a stormy area must be propagated farthest beyond
the region of generation, while shorrer and possibly higher waves die
out much sooner or are changed to longer ones. Such couriers of far-
off storms are known to us in the shape of the ocean swell, already
referred to.
When deep water waves spread our into a region of calm, only half
the energy of the wave advances with wave velocity. As a result the
energy of a train of waves advances with a distinct "front" moving at
half the velocity of wave propagation. Wave~ are continuously run-
ning out in advance of the front but die out in a very short time.
Behind the front the waves retam the original amount of energy,
because loss by viscosity is small. However, the velocity (and there-
WAVES 79
fore also the period) gradually increases with advance while the height
decreases, because it has been found, as pointed out above, that this
alteration is a general characteristic of wave propagation, whether
wind waves or swell, and mainly as a result of the suppression of the
shorter waves by eddy viscosity.
A
B
- - Wave ridge
. . . Wave tro ugh
FI G. 50. Contoured chart and section of sea surface, showing two main trains of
wa'C5. The crest lines of one train arc almost parallcl to the wind . (According
to data given hy Schumacher, 1939.)
- - Wave crests
- - - - Depth contours In meters
F'G. 52. Refraction of waves around nurthern coast of San Clemente Island off
California as deduced from aerial photOgraphs taken on June 20, 1944. (Redrawn
from Scripps Institution Report, Hydrographic Office, U. S. Navy Dept., Pub. 234,
1944.)
orbital velocity of the water particles at the top of the crest exceeds
the velocity of the wave. The particles move forward f:lSter than
the wave and topple over onto the front slope. Theoretically this
must take place when the steepness increases to 71; the top angle is
WAVES 83
0
then 120 The wind dr:1g hastens the breaking of the wave. Hence
this phenomenon occurs before the theoretical maximum steepness has
been attained.
The development of surf along a coast is a second consequence of
a shelving sea bottom. It is most characteristically shown when a long
ocean swell advances over a gently sloping and smooth bottom (Fig.
53).
The lessening depth detracts from the velocity, causing the waves
to crowd together, and therefore to become steeper. In the crest of
the wave the velocity of the water particles in their orbital movement
approaches and finally exceeds the velocity of the wave. At this
moment the wave "curls" and the crest shoots forward in a circular
course, without there being sufficient material to fill the central con-
cavity. The crest is then no longer supported and crashes forward
in a foaming line of surf, sending a swirling mass of water, called the
"swash," up the beach. When the up'ward movement has ceased, the
water runs back again as the "backwash. " This is roughly the way
the surf i5 generated, hut a complete understanding of the phenomena
has not yet been reached (Fig. 54).
It has already been pointed out that the bore is of a similar nature,
bur the gradual narrowing of the estuary aids in increasing the height
of the wave.
The depth ae which a wave breaks varies according to its height
and steepness. Steep waves break like deep-water waves. White
water appears on the crest, and the wave top gradually rolls over.
The breaking continues for a considerable distance and does not take
place with a crash. The term "spilling breakers" is applied to this
type. Irregularity of the waves or the bottom, and gentle slope of
the bottom, tend to cause spilling. Not-so-steep waves in deep water
do not start to break when approaching the beach until the greater
part of the energy becomes concentrated in the crest. The descrip-
tion of surf given above applies to this type, and the term "plunging
breakers" is used. Steepness and regularity of the bottom, and the
84 PHYSICAL OCEANOGRAPHY
FIG. 54. Block diagram of a breaking wave. (After Davis, 1912, Fig. 185, p. 472.)
tom, steepness of the beach, and force of the wind. Off the Dutch
coast several rows of sandbanks are found, and each may cause a
separate line of surf. The heavier the waves, the farther out the
first row of whitecaps is observed. Over the following banks addi-
tional lines of surf are formed, so that three or four white strips may
be seen in front of the beach.
Heavy surf may be caused in two different ways, either by an
onshore storm or by ocean swell. In a storm the rhythm is compara-
tively swift, because the wavelength is not very great. Obviously the
size of the rollers increases with length of fetch and with degree of
exposure of the coast to waves from deep water. When the surf is
due to ocean swell the period is much longer.
Certain stretches of coast are ill-famed for their heavy surf; French
Morocco is among these. As the height increases with lessening of
depths a swell of a few meters' amplitude may result in breakers
9 to 10 m high at the moment they curl over. The length of the
crests is so great that sometimes a white line is formed almost simul-
taneously as far as the eye can reach. Communication by boat between
ships in the offing and the coast is excluded under these circumstances.
The treacherous nature of this phenomenon lies in the fact that it
WAVES 8S '
may set in suddenly without warning because the source is far out
in midocean.
Thanks to the results of wartime work, mainly by Scripps Insti-
tution, mentioned earlier, surf can now be predicted with consider-
able accuracy from weather charts.
Solitary Waves. Besides the oscillatory waves dealt with so far
another type may also occur in shallow water, namely, solitary or
translation waves. The particles of water in these waves do not revolve
in circular or elliptical orbits but undergo a single displacement in the
The volume per unit of breadth of the wave above still water must
equal the space between the positions of a set of particles in a vertical
section before and after the wave has passed.
At sea, solitary waves are developed gradually from oscillation waves
on nearing the coast, or suddenly after breaking of the surf. In the
first case the depth gradually decreases; in the second, the surf sud-
denly throws a large volume of water over the shallow strip fronting
the beach.
As noted above, the regular waves funning on a beach during calm
weather are formed from oceanic swell and have undergone important
modifications while approaching the coast. Instead of showing crests
and troughs, practically flat expanses arc seen between widely spaced,
regular welts on the surface of the sea. Even when coming in groups,
as is usual, the distance from crest to crest is much greater than corres-
ponds with the steep slopes of the waves, and it varies within wide
limits for the same height of crest. Such forms constitute almost pure
solitary waves.
The maximum height of a solitary wave above still water equals
1.3 times the depth; otherwise the 'wave starts to break. V cry few
observations have been made in nature of magnitudes of solitary waves
before breal<ing, because observers have generally not distinguished
between oscillatory and solitary waves. The irregularity of the shapes
in nature and the occurrence of transitional forms complicate matters
and render the making of quantitative observations exceedingly difficult.
Large volumes of water are piled up against the beach by the action
of breakers. The water returns to the open sea by so-called rip cur-
rents or by underto'w, dealt with in a later chapter. Longshore cur-
rents are set tip within the breaker 2K:me where the wave crests advance
obliquely towards the shore. From the formulas given in the Hydro-
graphic Office Report ( 1944) on this subject it follows that appre-
ciable velocities may be expected ( to 2 m per second under favor-
able conditions).
Bibliography
ALu:N,]. Scale Models in Hydraulic Engineering, 407 pp" Longmans, G(een,
London, 1947.
AMERICAN GEOPHYSICAL UNION, Symposium on the Geophysical Exploration of
the Ocean Bottom, Proc. Am. Philos. Soc., Vol. 79, pp. 1-44, 1938.
BIGELOW, H. B. Oceanography, 263 pp., Houghton Mifflin, Boston, 1931.
BIGELOW, H. B., and ,",'. T . EDMONDSON. Wind Waves at Sea, Breakers II7Id Surf,
177 pp~ Hydrographic Office, U . S. Navy, Pub!. 602, Washington, D. C., 1947.
B I BLIOGRAPH Y 87
BOURCART,]. Geographie du fond des mers, 1I:tude du relief des Oceans, 307 pp.,
Payot, Paris, 1949.
CHUN, C. Aus den Tiefen des T.v eltmeares, 592 pp., Fischer, l ena, 1903.
CLARKE, F. W. The Data of Geochemistry, U. S. Gaol. S1I1"',)., Bull. 770, 841 pp.,
1924.
CoKER, R. E. This Great and Wide Sea, 325 pp., Chapel Hill, N. c., 1947.
CORNISH, V. O cean Waves and Kindred Geophysical Phenomena, 164 pp., Cam-
bridge University Press, 1934.
CoRRENS, C. \;V., et al. Ticfseebuch, 144 pp., Mittlcr und Sohn, Bcrlin, 1934.
DAVIS, W. M. Die erkliirende Beschreilmng der Landfonl1en, 565 pp., Teubner,
Leipzig, 1912.
DI;;FANT, A. DynmllisciJe O~ea'l7ograpbie, 222 pp., Julius Springer, Berlin, 1929.
DEFANT, A., ct al. Wind, Welter und IVel/en aUf dem Weltmeere, 150 pp., Mitt-
ler und Sohn, Berlin, 1940.
EARTHQUAKE REsEARCH lNST. TOKYO IMP. UN IV., Bull. Supp., Vol. 1, Papers and
Reports on the Tsunami of 1933 on Sanriku Coast, Japan.
EMERY, K. 0., and A. R. CHAMPION. The Underway Bottom Sampler, Int. Hydr.
Rev., Vol. 25, pp. 61-62, 1948.
EWINC, W. M., A. C. VINE, and J. L. WORZEL Photography of the Ocean Bot-
tom, J. Optical Soc. Am., Vol. 36, pp. 307-321, 1946.
fLEMING, R. H ., and R. REVELLE. Physical Processes in the Ocean, in TRASK
et al., pp. 48-141, 1939.
HENTSCHEL, E. Das Leben des T.veltmeeres, 153 pp., Springer, Berlin, 1929.
HYnHOGRAPJ-IIC OrFlc. U. S. NAVY OEM'. Breakers and Surf, Principles in Fore-
casting, 52 pp., 1944; Supplement, 19 pp., 1945.
Wind Waves a'Tld Swell, Principles in Forecasting, 61 pp., no date.
Forecasting Longshore Currents, 6 pp., 1945.
Ice Atlas, 107 pr., Publ. 550, Washington, D. c., 1946.
JOHNSON, D. W. Shore Processes and Sborciinc Development, 584 pp., John Wiley
& Sons, New York, 1919.
KNUDSEN, M. Hydrographiscbc Tabelle7l, 63 pp., Gad, Copenhagen, 1901.
KON. NED. AARDR. GEN. De Zecen vaTi Ncder/andsch Oost Indie, 506 pp., Brill,
Leyden, 1922.
KRUMMEL, O. Handbucb der Ozeanograpbie, Bd. 1, Die riiumlichen, chemischen
und physikaljschcn Verhiiltnisse des Mceres, 526 pp., 1907; Bd. 2, Die Beweg-
ungsformen des Meeres, 728 pp., 1911.
KUENEN, PH. H. Influence of the Earth's Rotation on Ventilation Currents of the
Moluccan Deep-Sea Basins, Proc . Kon. Ned. Akad. Wet., Vol. 51, pp. 417--426,
1948.
De Kringloop van het Water, 408 pp., Leopold, The Hague, 1948.
KUENEN, PH. H., and G. A. NEB. Bottom Samples, The Snellius E:cpedition,
Vo]' 5, part 3, 268 pp., Brill, Leyden, 1943 .
KULLENBERC, B. The Piston Core Sampler, Svcnska HydrO-BioI. Komm. Skrifter,
S. 3, Bd. 1, Hf. 2, 46 pp., 1947.
LEK, L. Die Ergebnisse dec Strom- und Serienmessungen, The Snellius Expedition,
Vol. 2, part 3, 169 pp., Brill, Leyden, 1938.
LORENTZ, H . A., and ]. TH. THIJSSE, see Verslag Staatscommissie.
88 PHYSICAL OCEANOGRAPHY
MACDONALD, G. A., F . P. SHEPARD, and D . C. Cox. The Tsunami of April i, 1946,
in the Hawaiian Islands, Pacific Science, Vol. I , pp. 21 - 37, 1947. (SHEPARD,
F . P ., G. A . MAcDONALD, and D. C. Cox, The Tsunami of April I, 1946, in
preparation.)
MAcEwEN, G. F . Modern D y namical Oceanography, Proc. Am. Phil. Soc., Vol.
79, pp. 145-166, 1938.
MONK, W. H., and M. A. TRAYLOR. Refraction of Ocean Waves, a Process
Linking Underwater Topography to Beach Erosion, J. Geo/., Vol. 55, pp. 1-26,
1947.
MUlUlAY, J., and J. HJORT. The Depths of the Ocean, 821 pp., Macmillan, London,
19 12.
NATIONAL REsEARCH COUNCIL. Bulletin 85, Oceanography , Physics of the Earth,
Vol. 5, 581 pp., 1932.
RIEL, P. M. VAN. The Bottom Configuration in Relation to the Flow of the
Bottom Water, The SneiJius Expedition, Vol. 2, part 2, Chapter 2, 63 pp.,
Brill, Leyden, 1934.
The Bottom Water, Introductory Remarks and Oxygen Content, Tbe Snellius
Expedition, Vol. 2, part 5, Chapter I, 77 pp., Brill, Ley den, 1943 .
RIEL, P. M . VAN, et al . Voyage, Tb e Snellius Expedition, 177 pp., Brill, Leyden,
1937.
ROUCH, J. La mer, 252 pp., Flammarion, Paris, 1939.
M eteorologic et physique du globe, tome 2, Ph ysiq ue des mers, Soc. d':tditions,
Paris, 222 pp., 1941.
Traite ti'oceanographie pbysique, I Sondages, 256 pp., 1943, II L'eau de mer,
349 pp., 1946, TIl L es ?/I0llVC1I1e17ts de 10 m er , 41 3 pp., P ayot, P aris, 1948.
RUSSEl.L, F. S., and C. M . YO NCE. The Seas, 379 pp., F. W arne, London, 1944.
ScHorr, G. G eograpbie des Indischen und Sti/len Ozeans, 413 pp., Boysen, Ham-
burg, 1935.
G eographie des Atlantischen Ozeans, 438 pp ., Boysen, Hamburg, 1942.
SCHuMACHER, A. Stereophotogrammetrische W ellenaufnahmen, Deutsche Atlam.
Exp. Meteor, Wiss. Erg., Bd. 7, H. 2, L. I, 156 pp. and Atlas, 1939.
SElWELL, H . R. Results of Research on Surface Waves of the Western North
Atlantic, Pap. Phys. Ocean. Met., Mass. Imt, Tech., Woods Hole Ocean.
1mt., Vol. 10, 4, 56 pp., 1948.
- SHEPARD, F . P. Submarine Geology, 348 pp., Harper, New York, 1948.
SlETSON, H. C . Oceanography, Geo!. Soc. Am., 50th Anniv. VoL, pp. 45-69, 1941.
STR!iSM, K. M. Land-Locked W aters and the Deposition of Black Muds, in TRASK
et al., pp. 356-372, 1939.
SVERDRUP, H. U. Oceanography for Meteorologists, 246 pp., Prentice-Hall, New
York, 1942.
SVEIUlRUP, H . U., l\1. W . ]OIl NSON, and R. H . FLF;MI NG. The Oceans, Their
Physics, Chemistry, and General Biology, 1060 pp., Prentice-Hall, New York,
1942.
SVERDRUP, H . U., and W. H. MUNK. Theoretical and Eml.?itical Relations in Fore-
casting Breakers and Surf, Trans. Am. Geophys. Union, Vol. 27, pp. 828-836,
1946.
THORADE, H . Ebbe und Flue, 115 pp., Julius Springer, 1941.
TRASK, P. D., et aJ. Recent Marine Sediments, A Symposium, 736 pp., Am. Assoc,
Petro Geo!., 1939.
BIBLIOGRAPHY 89
VAUGHAN, T. W. International Aspects of Oceanography, 225 pp., Nat. Acad.
Sci., Washington, D. c., 1937.
VEEN, J. VAN. Onderzoekingen in de Hoofden, 252 pp., Landsdrukkerij, The
Hague, 1936.
Verslag Staatscommissie Zuiderzee 1918-15126, 345 pp., Landsdrukkerij, The
Hague, 1926.
WORZEL, J. L. Ocean Bottom Sampler for Ships under Way, Geophysics, Vol.
13, pp. 452-456, 1948.
c H A p T E R T w o
2 3 4 5 6
- - True shape
- - - Constructed
FIG. 58. Construction of bottom profile from successive echo soundings. 1-6
location of soundings.
B
F,G. 59. E xample of increase in knowledge concerning the topography of the
sea floor by echo sounding between Buru and Ceram. A, chart prior to Snelliu!
expedition, based on a few deep-sea soundings. B, chart based on several hun-
dred echo soundings of the Snelliu! expedition. (After Kuenen, 1941, Fig. 36,
p. 155.)
c~6000
FIG. 60. Bottom topography in the South Atlantic Ocean. A, Profile of the
bottom betwccn the South Shetland Islands and Bouvet Island based on 13 wire
soundings. B, Profile over the same course constructed from over 1300 sonic
soundings (Meteor) . C, Bottom configuration and the track of the Meteor.
Vertical exaggeration in A and B about 200: I. (After Sverdrup et a!.; 1942,
The Oeems, Fig. 1, p. 18, Prentice-Hail.)
heights and depressions into w hich this swell is divided along each of
the sections? In other words: is the Mid-Atlantic Rise diversified by
separate peaks or by lengthwise ridges? A few sections run close
together might help to solve this question, but without a complete
survey individual features will always remain open to more than one
interpretation. Kuenen (1935), by an example from the Moluccas,
has shown how two very different interpretations of the topography
can be based on the same set of comparatively closely spaced sounding
runs (Fig. 61). Veatch and Smith (1939) give similar examples from
the eastern continental slope of the United States. Later we will point
out how their interpretation of the bathymetrical character in that
area is apparently open to serious criticism because they overempha-
sized the valley pattern when drawing the'depth contours. van Riel
showed (1934) how in special cases the temperature recordings in
basins may be called upon to show the sill depths and tbus to aid in
drawing the contours at the entrances of deep-sea depressions.
SUBMARINE TOPOGRAPHY 97
SUBMARINE TOPOGRAPHY
Various systems for the classification of the oceans and seas have
been proposed. Some, as for instance those of Suess, Krummel, and
Andree, are founded partly on hypotheses concerning the genesis.
For a geologist a purely morphological arrangement would be of little
value. On the other hand the introduction of hypothetical elements
into a classification should be avoided as far as possible. In the fol-
lowing a compromise has been struck. It is unavoidable that all
manner of transitions will be found between the various elements of
our classification, for namre does not work according to a taxonomic
plan. (A discussion of the nomenclature of submarine forms may be
found in Pub!. Scient. No.8, Association d'Oceanographie Physique,
1940.)
First to be distinguished are the oceans-the Pacific, Atlantic, and
Indian oceans. The true oceanic basins (that is, in a geological sense)
are simated beyond the continental shelves, starting from the 100- or
200-m line. The continental slopes lead down to the deep-sea floor.
Certain parts of the major oceans are frequently considered separate
units, such as the North Polar Basin, the Antarctic Ocean, and the
Norwegian Sea.
Besides the oceans the adjacent seas can be distinguished. The water
circulating in these smaller units is less uniform in consequence of the
influence exerted by the surrounding land masses.
The seas can be divided into rl10rginal seas and inland seas. The
marginal seas are widely open to the oceans. The shallow marginal
seas or shelf seas are situated on the continental shelf and therefore
rarely exceed 200 m in depth, except off formerly glaciated coasts.
They form thin flanges to the oceanic basins. Examples are the North
Sea, Bering Strait, etc. The deep marginal seas are separated from the
true oceans by submarine ridges with or without islands. Examples
are the eastern Asiatic marginal seas.
The inland seas are largely surrounded by land or shallow waters,
so that the communication with the open waters is restricted t;o one
or a few straits. Inland seas also show great variations in depth.
Shallow inland seas or epeiric seas with less than 250-m depth as the
Baltic, Hudson Bay, Sunda and Sahul shelves (the latter two forming
transitions to shelf seas) may be looked upon as flooded continental
areas, although it is also possible that sedimentation has filled up
deeper depressions.
By far the greater number of fossil sedimentation basins belong to
98 THE SEA BAS INS
the groups of shallow adjacent and inland seas, for in few cases did
geosynclines and basins obtain greater depths and both were fonned
on continental areas. The temporary deepening of some geosynclines,
as for instance the Alpine Trough, clearly shows that this parallel and
the classification in general have only a relative value. We will return
to these matters in the next chapter.
The deep inland seas, mediterranean seas, or deep-sea basins show
depths from 200 m and more to depths exceeding those of the ocean
floor. Generally the floor is from 2000 to 5000 m below sea level.
As examples may be cited the Mediterranean, the East and West Indian
basins, the Red Sea, and the Black Sea.
All shallow seas, that is all seas covering part of the continental
blocks, may be classed together as epicontinental seas.
The deep-sea trougbs or deep-sea trencbes comprise all deep, oblong
depressions either of the ocean floor or between continental areas. A
not very sharp distinction can be made between deep~sea troughs or
deep-sea trenches with a V-shaped cross section (in the Moluccan
Passage, the East Asiatic deep-sea troughs) and the deep-sea troughs
or deep-sea tre11ches with a U-shaped cross section (Bartlett Trough
south of Cuba, Weber Trough in the Moluccas). Whether the Red
Sea shou ld also be classed in this group is doubtful. Some writers dis-
tinguish between troughs, showing gentle slopes, and trenches: show-
ing relatively steep ones. Basins are larger depressions with a more
or less circular or oval form. Small depressions may be termed pits,
as for instance in Strait Manipa between Buru and Ceram (Fig. 59),
and the Romanche Pit in the Mid-Atlantic Rise. The name trench
or trough is given to long and narrow, steep-walled depressions with-
out marked lengthwise slope, while submarine canyons are valleylike
trenches on continental or insular slopes.
The positive forms of the sea floor are classed as ridges that are
long and narrow, rises that are long and broad, and swells that are of
small relative height. Some are decorated by lesser ridges and hum-
mocks (Mid-Atlantic Rise, Whale Ridge, the ridge south of Java,
Carlsberg Ridge, etc.). There are also plateaus (Seychelles Plateau,
Azores Plateau, Albatros Plateau) and seamounts of lesser horizontal
extension, the latter generally submarine volcanoes or drowned atolls;
banks are elevations covered by less than 200 m of water; shoals
approach the surface and constitute dangers to shipping. Guyots are
flat-topped, deep-lying seamounts of more or less conical shape.
Mention should also be made of straits. The geographical tenn is
used for all narrow passages, but geologists would do better to restrict
SUBMARINE TOPOGRAPHY 99
the term to interruptions in isthmian links or chains of islands. In
this restricted sense the passages between the Lesser Sunda Islands or
the Japanese Islands, Strait Bab-el-Mandeb, the Straits of Gibraltar,
the Dardanelles, and the Bosporus fall under the term of strait. One
might also include the Straits of Dover and the openings between the
Danish Islands. On the other hand Makassar Strait between Borneo
and Celebes is a basin and not a strait in the geological sense.
Other terms frequently met with are fiord, estuary, bay, lagoon,
watt (tidal flat). They need not be discussed here.
The study and description of coasts and shorelines requires a system
of classification. Johnson (1919) formerly proposed a system based
on the emergence or submergence of the area in question, but Shepard
(1948) found that the recent eustatic movements of sea level during
the Pleistocene have obscured the features due to emergence or drown-
ing. He therefore proposed an entirely different classification of a
double nature. One system is meant to care for the smaller subdivi-
sions of the coasts and shorelines; the other pertains to the larger sub-
divisions of the coastal region as a whole. The main clements of the
first system are as follows:
I. Primary or youthful coasts and shorelines, configuration due primarily
to nonmarine agencies.
A. Shaped by terrestrial erosion and drowned by deglaciation or down-
warping. (Drowned river valley coasts; drowned glacial erosion
coasts.)
B. Shaped by terrestrial depositional agencies. (River deposition coasts;
glacial deposition coasts; wind deposition coasts; vegetation extend-
ing the coast.)
C. Shaped by volcanic activity. (Coasts with recent lava flows; shore-
lines due to volcanic collapse or explosions.)
D. Shaped by diastrophism. (Fault scarp .coasts; COasts due to folding.)
II. Secondary or mature coasts and shorelines, configuration primarily the
result of marine agencies.
A. Shorelines shaped by marine erosion. (Sea cliffs straightened by
wave erosion; sea cliffs made irregular by wave erosion.)
B. Coasts and shorelines shaped by marine deposition (straightened by
bars, prograded, longshore spits, coral reefs).
The classification of major subdivisions runs thus:
1. Coasts with young mountains.
2. Coasts with old mountains.
3. Coasts with broad coastal plains.
4. Glaciated coasts.
100 T H E S EA B AS INS
-tooo'
......... '.00'
Meters
.. .... 200
----- 2000
--4000
~6000
_ 9000
FIG. 63. Sketch map of the Solomon Basin with Bougainvillc-New Britain Trough .
2000 A
o 10
Kilometers
~ 1000 2'
.,
.!l
:IE
2000
FIG.64. Typical section of a guyot (S 51' N ; 163 10' E) from fathomerer record
(adjusted for slope). A , natural scale; D, vertical scale 5 X horizontal scale.
(Redrawn after Hess, 1946, Am. J. Sci., Vo)' 244, Fig. 28, p. 777.)
tively narrow, but the Mid-Atlantic Rise is more than 1000 km broad.
Besides ridges, more plateaulike elevations also occur. These are
nearly all situated as eminences on the crest of ridges, from which
they rise more or less abruptly. Some of them have complicated
shapes, such as the Azores Plateau; others have a rounded outline, as
for instance the Seychelles "Bank," Fiji Plateau, and the Toekang
Besi Plateau to the southeast of Celebes. A number of plateaus are
to be considered extensions of the continental masses, as the Birdwood
Bank off Cape Horn, the Bahama Plateau, and the Sokotra Bank. But
elevated areas, some of great extent, are also found independently of
ridges, as that of Wake Island or the Albatros Plateau to the west of
Central America.
Hess has drawn attention to the occurrence of many more or less
co ical seamounts in the northwestern parts of the Pacific that have
a flat horizontal top at depths of 1000 to 2000 m. He calls them
guyots (Figs. 64 and 65). They have since been encountered also
in the northeastern and southwestern Pacific. Perhaps some of the
104 THE SEA BASINS
I
",---_ ,
J'''
I I
" ....J;.
I
,~
.....r;/ ~
Eniwet .~.B:--_
atoll t ~'\,
\
B
\ I
\ I
'- -"
A
500
1000
~
1500,*
:;:
2000
Course
173' true
I t! I I , I
o 2 4 6
Kilometers
FIG. 65 . Situation of n;wctok atoll on two guYOts. (R edraw n after Hess, 1946,
Am. , . Scimce, Vol. 244, Figs. 4, 5, and 6, p. no.)
Continet)cal terrace
Total length of terrace = IOO,OOO km Average breadth of con-
Average breadth to break tinental slope = 75 km
in slope 70 km Average slope of continen-
Average depth at break in tal slope = J: 15
slope 130 km Total length of coastline = 400,OOO km
Average slope to break in
slope = 1: 540
From Tables 8-10 it is seen that the seas cover about three-fourths
of the surface of the earth and that the cubic content is equal to a
layer 2Y2 Ion thick covering the entire earth. Although the average
height of the land is % km above sea level, the frequency of each suc-
cessive zone decreases upwards. The lowlands below 200 m, for in-
stance, cover twice the area occupied by land lying between 1000 and
2000 m, although the latter zone is 5 times as high. The area of the
sea floor between the coastline and the 200-m depth contour is only
half that of the land less than 200 m high. Then follows the conti-
nental slope, that is in general much steeper, for the zone from 200 to
SUBMARINE TOPOGRAPHY 107
Land Sea
o 50 100
Nautical miles
FIG. 66. Chart of the eastern end of the Aleutian deep-sea trough. (After
Murray, 1945, Bull. Geol. Soc. Am., Vol. 56, Plate 3.)
111111111111111111111
o 5 10 15 20
Nautical miles
FIG. 67. Natural scale section of the Aleutian Trough, (After Murray, 1945,
Bull. GeoJ, Soc. Am" Vol. 56, Fig. 11, p. 777.)
difference between tbe two forms would then be due merely to ex-
ternal conditions. This view is substantiated by the relations of the
Aleutian Trough. The accompanying island arc continues uninter-
110 THE SEA BASINS
rupted in the North American Continent, and the trench itself passes
upwards at its eastern end onto the shelf towards the mouth of Yakutat
Bay (see Figs. 66, 67, and 68).
As far as present knowledge goes, therefore, the continental and
deep-sea surfaces show no structural differences of a fundamental
nature. All actual divergences may be attributed to differences In
FIG. 68. Generalized lengthwise profile of the Aleutian Trough and the adjacent
mountain arc. (After Murray, 1945, Bull. Geol. Soc. Am., Vol. 56, Fig. 12, p. 777.)
non-marine formations accumulated over a large area after the sea was
gone.
v'1t need hardly be emphasized, in conclusion, that sedimentation in-
fluences the ultimate size of a basin even if no dry land is formed at
the time, for, if shoaling by sedimentation has gone before, this must
lead to more extensive regressions when elevation or eustatic sinking
takes place later than if no deposition had preceded.
the surface of the earth is divided into low-lying ocean floors and
high-standing continental masses. Direct observation on the latter is
limited to small depths, and the substratum of the deep-sea floor below
the covering of recent sediment cannot be sampled at all, except on
bare seamounts, which may be quite foreign to their surroundings.
Hence inferences on the composition must be drawn from indirect
FIG. 69. Diagrammatic section illllsu ating inferred character of the earth's outer
shells according to Dal y, 19'12, Tbe Floor of tbe Ocean, University of North
Carolina Press, Fig. 30, p. 59.
I. Macquarie Island
2. Chatham Island
3. Kermadec Islands
4. Tonga Islands
5. Samoa
6: Fiji Islands
7. Santa Cruz
Islands
I
10. Palau Islands
11. Yap
., 12.
13.
Marianas Islands
Bonin Islands
'f
- 14.
15.
Kurile Islands
Aleutian Islands
16. Galapagos Islands
FIG. 70. The andesite line, mainly according to Hess (1948) and Smit Sibinga
(1943).
~ 3
.!l
Q)
E4
....
5
., 7
8
FIG. 71. Percentage of area of depth zones (in thousands of l1Ieters) in the three
oceans. excluding adjacent seas. Inset: average depth of the three oceans. Note
that the greater average depth of the Pacific is due mainly to low percentage of
small depths and high percentage of abnormal depths.
FIG. 72. Diagrammatic sections of oceans. A, Northern Pacific with great aver-
age depths; B, Atlantic with lesser average depth.
Flc. 76. The comparatively large distance between the African and South
American continental blocks prior to drifting according to Du Toit's recon-
struction (1937, Oliver & Boyd, Fig. 7, p. 64) .
in the orogenic rhythm of the earth's crust be traced between the ages
before the Mesozoic and since the beginning of the Alpine cycle.
For the special case of Indonesia the present writer showed how
the geological and bathymetrical data cannot be reconciled with
Wegener's synthesis (1935). The only possibility appears to be to
accept Du Toit's reconstruction and to suppose that the Moluccas
remained in contact with New Guinea from first to last, thus wheeling
round during the drift of Australia. Further details are mentioned in
the next chapter.
No doubt there are also arguments in favor of the theory of drift,
but they lie principally in the field of regional geology, biogeography,
and paleoclimatology, and therefore fall outside the scope of the
present volume. Many specialists on these subjects have also raised
grave objections.
If the conclusions of adherents to drift in some form or other,
such as Du Toit, Wegmann, Gutenberg, and Kirsch, are confronted
with the opinion of opponents, for instance Bucher, Umbgrove, Stille,
and Cloos, it becomes obvious that neither of the two camps can claim
a decisive victory. But the evidence favorabl e to drift often proves
illusive, or at least open to serious doubt, on closer inspection. For
the time being most geo logists appear to have lost faith in continental
drift as a sound working hypothesis.
On seismological grounds Rothe suggests that the Atlantic east of
the Mid-Atlantic Rise is of continental nature, while to the west it
is of the same ty pe as the Pacific. The latter part is supposed to have
been created by westward drift of the Americas in the sense of
Wegener. But the depths and the morphological structure of the
two halves of the Atlantic are practically identical and do not war-
rant postulating an entirely different origin.
Returning to the problem of permanency, an important argument
must be emphasized. It is concerned with the cubic content of the
oceanic basins. The present time is certainly not poor in volcanic
activity, yet the yearly production of extrusive materials is not more
than about 1 km s, according to Sapper. The intrusive magmas aver-
age roughly the same amount as far as can be ascertained. The highest
estimate that can be admitted for the loss of juvenile water by all
this material is 5ro. Hence, the yearly production of juvenile water
since the beginning of the Paleozoic can hardly have exceeded %0
km s; possibly it was considerably less. If this highest figure is assumed,
the amount of water in the oceans has increased but 50 X 106 km8
since the beginning of the Cambrian. As the present volume is
130 THE SEA BASINS
1370 X lOll km 8 ,
the amount of water at the close of the pre-Cambrian
must have been of the order of 1300 X 10 km 8 at the very least
(Fig. 77).
There is further no paleontological indication that marine waters
have become less saline through marked dilution in post-Algonkian
times. But this is what would have occurred if the salts had been
delivered gradually to the oceans through weathering, while the water
was produced principally during the last quarter of terrestrial history.
It is thus found to be a reasonable postulate that the total amount of
Twenhofel
sea water has not been augmented by more than a fraction since the
beginning of the Paleozoic (for further details see Chapter 5). This
water must have found room in the oceans, and the theory that the
deep'Jsea basins are of relatively recent formation is thus found to
encounter serious obstacles.
As the volume of the Pacific is but half that of all oceans put to-
gether, the other oceans must also have existed since early times;
otherwise there would have been no space to contain all the waters
at the surface of the earth. The alternative, that the Pacific showed
double the present depth and that all continents were connected by
dry land, where the Atlantic and Indian oceans now lie, is too unlikely
to be assumed.
It is of importance to note that, since as far back as the Cambrian,
at least, the upper surface of the continents has been close to sea level,
just as it is at present, being sometimes partly flooded, sometimes laid
almost completely dry. As the amount of water was almost equal to
that of the present seas, the cubic content of the ocean basins must
also have closely approached the conditions that now prevail.
The same reasoning can be applied to pre-Cambrian times, but the
farther back we go in geological history the less certain our conclu-
sions can be. The presence of conglomerates among the oldest rocks
THE PROBLEM OF PERMANENCY 131
known proves that land emerged above sea level, and that running
water, clouds, etc., were present from the earliest pre-Cambrian on-
wards. But concerning the salinity of the seas and the rate at which
juvenile water was given off by the crust of the young and active
earth only a guess can be offered. Still, it is justifiable to assume
that the volume of the oceans as well as the area and depth of the
oceanic basins were of the same order of magnitude as they are at
present.
Knothe (1933) also came to the conclusion that the volume of the
primitive oceans must have been considerable, even if a strong con-
traction of the earth has occurred since then. Schuchert likewise
expressed the opinion that the volume of the oceans has not increased
very drastically during geological history. He estimated the increase
at 25 0/0, without, however, presenting concrete arguments. But even
if we assume this comparatively high figure the enlargement since the
end of the Paleozoic would amount to only 2Y2ro . The opinion held
by the present author is expressed in one of the curves of Fig. 77.
In the writings of Plato a myth is recorded relating to Atlantis, a
large island, with a highly developed culture, that must have existed
outside the Straits of Gibraltar, and that was engulfed in the ocean
in prehistoric times. On the strength of this legend many writers
representing several branches of science have occupied themselves
with the problem of a foundered continent in the Atlantic Ocean.
A literaturt' of more than 2000 papers has appeared on this subject
(Hogbom, 1941). On geological grounds the possibility can hardly
be entertained, for, although some geologists who do not believe in
the permanency of the oceans have brought forward arguments in
favor of land bridges across the Atlantic from Europe to America,
these connections would be so ancient that there could be no question
of historic tradition. The mutual resemblance of the faunas and
floras on the Atlantic oceanic islands and the bordering continents is
so slight that no connection can have existed after about the middle
of the Tertiary. Many millions of years before the first appearance
of primitive man the Atlantic Ocean must already have been in full
existence. The occurrence of globigerina ooze dating from the last
interglacial period on the ocean bed where parts of Atlantis are recon-
structed is not only proof against a very recent subsidence but also
evidence of a great depth during at least the last 100,000 years.
Whatever point of view is taken by a geologist on these matters,
the most zealous opponent to the theory of permanency will readily
132 THE SEA BASINS
concede that there is no possibility of extensive areas having subsided
from above sea level to normal oceanic depths since late prehistoric
times.
Crystalline
crust
FIG. 78. Schematic representation of the origin of continents and oceanic basins
in the early pre-Cambrian by a process of buckling and drifting of an originally
siallic layer enveloping the whole e~rth. (After Umbgrove, 1947, Fig. 3, p. 175.)
heat continued, and a rigid crust gradually formed around the earth.
After that, great tangential displacements were excluded (Fig. 78).
It is obvious that all orogenic cycles determined by field obser-
vations-also the most ancient such as the Marealbian, Svecofennian,
etc.-are supposed by Umbgrove to have followed the consolidation of
the earth's rigid crust. No vestiges remain of the disturbances by
which the primeval sial layer was crumpled into a thicker stratum.
This evolution took place entirely in pregcologic times. This would
imply that the formation of the continents and oceanic basins must
136 THE S E A BA SINS
Meters
.. ........ .. .... 200
.................. 1000
,..,.....".",..---- 2000
3000
4000
5000
6000
FIG. 79. Bathymetric chart of the southern Atlantic Ocean. (Redrawn from
Schott, 1942, Plate V.)
138 THE SEA BASINS
Paleozoic, Vmbgrove crowds the entire process into an early stage of
terrestrial development, before the eanh had obtained a rigid crust.
Formulated thus the difference between the two conceptions may
appear only slight, but actually entirely opposed pictures of the geo-
logical development of our planet result from the difference in time
assigned to the process of drifting.
In this connection an interesting hypothesis evolved by Vening
Meinesz should be mentioned. On theoretical grounds this author
80,--,----.,----,.----,-----,,----,
Depths In meters
1
2000
8000-0nly ",.,I<td on Car/.~"1
~ggg &. Mid Indian Ridge
FIG. 81. Bathymetric chart of the Indian Ocean. (After Wiseman and Sewell,
1937, Fig. I, p. 221.)
144 THE SEA BASINS
Stearns (1945) summariz.ed the history of Pacific volcanic islands
and emphasized the widespread evidence of considerable subsidence.
In the discussion on atolls in Chapter 6 the importance of this sub-
sidence will be pointed out. Stearns and others have interpreted it as
proof of a general deepening of the Pacific Ocean floor. He attributes
it to the extrusion of the lava forming the islands. It is evident, how-
ever, that the volume of lava when spread out over the ocean floor
represents but a very thin stratum. Hess invoked the gradual relative
rise of sea level due to deep-sea sedimentation. This process, how-
ever, is very slow-probably of the order of 1 to 2 m in a million years.
We must assume a swifter subsidence that carries the volcanic cones
downwards in early stages in order to explain the observed drowning
of valleys. Vening Meinesz has shown that the gravity field denotes
regional compensation of oceanic islands. This means that a wide
area is slightly depressed by the weight imposed on the ..:rust by the
volcanic cones. Probably the depression keeps pace with the gradual
erection of the cone and cannot explain the physiographic drowning.
It is conceivable, however, that eventually the crust gradually gives
way and allows the cone to subside, while the surrounding depres-
sion returns to its original position. Compaction and squeezing aside
of the oceanic deposits on the ocean floor below the cone is another
possibility pointed out by Kuenen.
Ar~. the Continents Extending? A very important problem is whether
continents can be extended by the incorporation of new geosynclinal
belts along the margins. Much has been written on this subject, but
confusion as to the meaning of the terms continents, consolidated areas
of the crust, etc., has tended to obscure the points at issue. The word
continent will here be taken as meaning regions of the crust close to
or above sea level and so extensive that the weight of their mass pro-
truding above the level of the deep-sea floor cannot be borne by the
strength of the crust. A thick layer of light sial must therefore be
postulated to maintain isostatic equilibrium. There is a widely ac-
cepted view that the continents consist of old nuclei onto which
new orogenic z.ones are successively welded. By this process the
ancient shields are gradually extended to form the so-called kratogenic
areas. A comparatively recent example of this line of thought is to
be found in writings of Born (1933). This picture of continental
development, however, does not meet the facts. It is true that a geo-
syncline often migrates during the successive phases of an orogenic
cycle. But this movement frequently happens in a direction towards
the ancient nucleus. An example is the Molasse Trough along the
OR I G I N OF CONTI N ENTS AN D OCEANS 145
northern margin of the Alpine geosyncline on the side where this
chain borders on ancient Hercynian Europe. The simple picture of
a gradual spread of the nucleus from the Scandinavian shield to the
south is thus not confirmed. Admittedly it may hold a certain measure
of truth.
But when the spread of geosynclinal activity is interpreted as indi-
cating a gradual growth of the continental mass, more serious objec-
tions can be raised. The basement of all geosynclines is found to be
a continental area. The folding of a geosyncline therefore does not
imply an extension of the continent, but only an addition to the
ancient nucleus of intensely crumpled rocks. In fact, it means the
opposite to expansion, for an orogenic cycle results in thickening
and narrowing of a belt that already belonged to the continent.
This reasoning is borne out by the important fact that nowhere has
true oceanic sea floor been incorporated in the continents, because
no extensive pelagic deep-sea deposits are found in geosynclinal or
other sedimentary prisms. We do not know what agents cause the
labile conditions in a geosyncline, but in any event these troughs
develop from continental areas underlain by a thick sialic crust. The
writer has had occasion to emphasize this significant point (1935) and
to show that Schuchert's type of mesogeosyncline appears to have
been founded on a misinterpretation of European geology. His proto-
type was formed by the Alpine Geosyncline. According to Schuch-
ert it originated between two continents, the African and European.
But the word "between" should not be taken literally, for the Alpine
Trough was formed by the subsidence of a denuded land area. Every-
where Permian and Triassic are found to transgress over more ancient
denuded rock surfaces. Not until the Jurassic and Cretaceous were
greater depths developed through subsidence.
Neither is it true that a new geosyncline is always developed in a
region in which orogenic quiescence had reigned before. In the
center of the Alps ancient massifs have been laid bare that show
strong Hercynian orogenic structures. Wherever the pre-Cam-
brian is exposed, intensive tectonic disturbance prevails, yet these
formations appear to underlie all post-Cambrian mountain belts. Al-
though it must be admitted that generally a great length of time
elapses before an orogenic belt is again invaded by a geosynclinal
cycle, the repetition of compression in one and the same area thus
appears to be the rule. In spite of this the idea that a folded region
remains exempt from orogenic activity is prevalent.
146 THE SEA BASINS
The above considerations show that orogenesis does not lead to the
extension of continents but, conversely, to a certain measure of con-
traction and thickening. A development in structure has been mis-
taken for a growth in size by writers who advocate continental expan-
sion by orogenic cycles.
It now remains to be investigated whether growth is possible by
some other process. Born mentioned the extension of continents by
the dumping of sediment in geosynclines, but as we saw this means
only a local thickening of the sial. In the times following the pre-
Cambrian, growth of the continents has taken place only through
deposition on the continental slopes, and it is highly probable that the
same picture applies to earlier sections of terrestrial history. It is not
probable, however, that the prism of sediments building the conti-
nental terrace outwards has ever been folded and incorporated in the
kratogenic nuclei of the continents. Not only is it doubtful that
geosynclines ever originated right on the margin of continental blocks,
but also the true oceanic floor has never been carried upwards above
the level of the sea. Bucher, at any rate, denies that marginal geo-
synclines were developed at the very edge of a continent, for on the
side fronting the ocean there was always a denuding belt, now some-
times lost by subsidence (Appalachia, etc.). Umbgrove (1947) came
to the same conclusion, that the continental blocks do not expand by
incorporation of orogenic belts.
In his undation theory, van Bemmelen (1939) postulates a primary
salsima layer formerly enveloping the whole earth and now forming
the floor of the oceans. Differentiation has generated the sial of the
continents and a heavy sima layer underneath. This process is con-
tinuing in the salsima bordering on the continents. van Bemmelen
shows how this differentiation into a lighter upper part and a heavier
lower stratum must upset hydrostatic equilibrium. Consequently
sideways injection of salsima between the sial and sima of the conti-
nental sections must follow, resulting in elevation of the continents
and depression of the sea floor along the continental border. This
process causes a slope at the surface on which unconsolidated sediments
slide downwards, being thereby folded to tectonic structures.
van Bemmelen's line of speculation cannot be followed in detail here.
The reader is referred to the original publications. All that need
concern us is that it leads to the conclusion that the continents expand
by the incorporation of ocean floor into the margins of the sial blocks.
If van Bemmelen were right the lower strata of a geosynclinal prism
should be formed of pelagic oceanic sediments. Neither the growth
OR I GIN 0 F CON TIN E N T 5 AND 0 C E A N S 147
of the continents nor the incorporation of deep-sea floor into these
blocks appears to have taken place, as shown above. When dealing
with the continental terrace presently it will be shown that there is
strong evidence for the opposite to continental growth. It is difficult
to avoid the conclusion that marginal areas have subsided to form part
of the present deep-sea floor.
Patterns In the Earth's Crust. More than a century ago Dana and
Darwin remarked on the alignment of the Hawaii and Galapagos
Volcanoes in rows with constant interspaces. Since then many others
have traced certain regularities in the arrangement of volcanic cones
(Green; Friedlander, 1918; Kuenen, 1945). It is now obvious that
we are not dealing with a peculiarity of the sea floor but with a
general property of the crust. Evidently the suboceanic and conti-
nental sections of the earth's crust react in a similar manner to stresses
producing volcanic fissures.
Mecking (1940 ) draws attention to a number of similarities in the
surface relief of the earth. In the southern hemisphere a north-to-
south direction prevails not only in the forms of the continents but
also in the mountain systems and the great midoceanic rises. He
further concludes: "The less oceanic regions are hedged in, the more
extensive and Jess marked the subdivision in basins is found to be.
The narrower they are, the smaller and the more intensive and marked
the subdivision becomes." 1 But one has to admit that there are
many exceptions to these rules where intensive relief occurs in por-
tions of wide oceans or where simple, poorly marked forms are situ-
ated in narrow offshoots of the main oceans. The present writer
believes that many such homologies are to be expected from merely
chance distribution of forms and that no great value can be placed
on a repetition once or twice of any special feature when seeking for
a satisfactory geophysical picture of crustal development.
Some authors have gone much further and have attempted to trace
certain rules in the morphology of the crust: preferred directions,
constant intervals, etc. Thus Sonder (1939) advocated what he terms
Lineamenttektonik. He assumes that the ocean floor is rigid and
unable to react to tangential stress by normal orogenic processes.
Faulting is believed to form the main type of deformation, and vol-
canic eminences to playa major part in the surface relief. Swells
produced in this manner are said to intersect at fixed angles and at
1 Ie freier die ozeanischen Riume, desto groszziigiger und schwacher die Felder-
ung. je becmgter die Raume. desto Ideinziigiger. intensiver und lusgeprigter die
Felderung.
148 THE SEA BASINS
"Eo
z
...o
~
o
t.,
z
....ao
t:i
~--------------------------------------------~ ~
150 THE SEA BASINS
photographs, and the boundaries of continental blocks, fit his scheme
to a remarkable degree. However, there is always the danger of select-
ing the favorable evidence and overlooking misfits. An evaluation of
this hypothesis must await a detailed and statistical survey of the direc-
tion of the main morphological and structural features of the crust.
Whatever the result, a second system of preferred directions in
the earth's crust is of equal importance, a north-south east-west system.
According to Umbgrove (1947) this system requires a second cause,
and, being related only to the present and not to the supposed former
axis of rotation, it would necessarily be of later origin. Vening
Meinesz suggests that any stress tending to rupture the crust nearly
along the shear pattern will make use of these old lines of weakness,
but if diverging strongly new lines of disturbance will be formed.
This should automatically cause a second system of preferred direc-
tions bisecting the angles of the shear pattern and running roughly
north to south and east to west.
We must leave the problem of linear patterns in this early stage of
reconnaissance.
The Evidence from Earthquakes. Some authors have attempted to
fathom the structural nature of the oceanic basins by considering the
regional distribution of earthquakes. Clements and Emery ( 1947)
state that the seismic activity and the submarine topography in the
borderland oif southern California are related. They believe that the
major topographic features are of tectonic origin and that the forces
which brought them about are still active. Their map, however, indi-
cates that the great majority of shocks show no connection with sur-
face features. Probably they occur too deep in the crust to be accom-
panied by movements at the surface.
More far-reaching conclusions on oceanic structure were drawn
by Rehm (1936) on the basis of seismic data. Gutenberg and Richter
(1941, 1945) have since shown, however, that older maps giving the
location of epicenters contain a large number of errors, so that little
value can be attached to earlier reviews. Moreover, Rehm makes the
mistake of ascribing all shocks to fault tectonics. Even w hen only a
few epicenters are recorded in a region he excludes the possibility of
other tectonic happenings. The possibility that transverse horizontal
shifts, overthrusts, or faulting may cause earthquakes as minor accom-
paniments or aftermaths of folding is left out of consideration.
The fact that the great majority of earthquakes originate at several
dozens of kilometers below the surface (not to speak of deep-focus
ORIGIN OF CONTINENTS AND OCEANS 151
earthquakes) is sufficient reason for doubting that they can teach us
anything directly concerning the tectonic structure of the accessible
crust. For the time being the opposite procedure, as followed by
Gutenberg and Richter, is more promising, namely, to locate epi-
centers and observe whether a correlation can be established with
certain features of the crust. Not until such correlations have been
firmly established and a far greater volume of data has been collected
can a tentative determination of crustal structure be founded on
seismic data.
The careful sifting of data by the authors mentioned led to the
remarkable result that seismicity is concentrated in narrow belts to a
much greater degree than was hitherto suspected and that there are
large stable areas practically free of activity. This clarifies the picture
and renders the correlarions of certain types of seismicity with gravity,
volcanism, submarine rises, island arcs, and deep-sea troughs far more
pronounced than was formerly supposed. Thus the andesite line is
closely Jinked with a belt of excessive activity, and moderately seismic
belts follow the Mid-Atlantic and Easter Island rises and the rise in the
western Indian Ocean. But up to the present it is not possible to
deduce the structural nature of the oceanic basins or submarine ridges
and troughs from the distribution of epicente rs. Thus seismo logists
find rhe same absence of shocks in the Pacific as in the Canadian and
other continental shields, while the Tertiary orogenic belt of Central
Asia much resembles the Mid-Atlantic Rise and is not very different
from the East African Rift Valley.
It is not the writer's intention to detract from the value of regional
seismic studies, but again the conclusion must be that the time is not
yet ripe for far-reaching and well-substantiated conclusions.
Positive Forms of the Sea Floor. Thus far we have been mainly con-
cerned with the major depressions of the earth's crust, although
minor negative forms have also been mentioned. Turning now to
the positive forms of the sea floor we find that again little is known
with certainty and that each author makes his own guess as to the
nature of submarine ridges and plateaus. As, moreover, these prob-
lems are even more closely related to those of structural and regional
geology, fields of investigation we have avoided as far as possible,
the topic cannot be discussed in detail. A few aspects, however,
should be treated briefly.
Many positive forms are doubtless due partly or entirely to volcanic
activity. Countless oceanic isl:mds are known to be built up ex-
clusively of eruptive rocks, and a volcanic origin of many isolated
152 THE S EA B ASINS
F.(;. 84. A\'erage profile of the Hawaiian Swell. showing body of volcanics as
suggested by Betz and Hess ( 1942). Vertical scale about 40 X horizontal scale.
South of
Cape May Baltimore canyon
o
Coast line
Mississippi delta
I I I I I
o 10 20 30 40 50
Kilometers
FIG. 85. Two profiles of the continental terrace. Upper: On the east coast of
the United States from Cape May to a ridge south of Baltimore Canyon. Lower:
Off the mouth of the Mississippi. The approximate position is indicated of the
coastline from Barataria Bay to Chandeleur Sound at base of present delta .
.,
Umbgrove (1947) has reviewed the theoretical aspects and shown
how far we still are from a complete picture and general agreement.
Shepard (1~48) has summarized the data collected by sounding and
sampling.
Shepard (1948) arrived at the following general conclusions con-
cerning the morphology of the shelf. The average width is 70 km;
the depth at the edge averages 130 m but may attain 500 m. Hills of
20 m or more occur on 600/0 of the profiles, hollows of 20 m or more
on 35%. The average slope is somewhat steeper in its inner half than
in the outer half. In restricted bay entrances deeper channels are the
rule, and a shallow sill exists farther out, probably a sand bar. Off
formerly glaciated coasts the shelf is wide and deep with a very
irregular surface, due to glacial troughs and morainic deposits. Off
large rivers the shelves are shallow and broad, except where covered
by a protruding delta. In areas of active coral growth the shelves
are shallow and strewn with irregular shoals and banks. Off young
mountain ranges shelves tend to be narrow and deep (20 m deeper
THE CONTINENTAL TERRACE 155
than the average), or they are even entirely lacking. Where powerful
oceanic currents impinge along the coast the shelves are narrow or
missing. Thus Florida, a low fiat area, has a broad shelf on the west
but virtually no shelf on the south and southeast where the Gulf
Stream sweeps close along the shore, commonly attaining 1 to 2 m
per second. At greater depths of nearly 1000 m, however, the slope
is again gentle and terminates outward with an escarpment which is
in line with the edge of the continental shelf beyond the reach of the
current.
Shepard finds no simple relation between marginal depths and width,
or between depths and degree of exposure to storm waves.
The same writer further emphasizes that sand is the dominant sedi-
ment of the shelves. Mud is common off larger rivers and downcur-
rent from them. It is also found in embayments and depressions of
the shelf. Pebhles, cobbles, and rock bottom are common on most
shelves, especially towards the outer margins.
Shepard (1948) also gives a regional description of the continental
slopes and arrives at the following conclusions. The average slope
is 4 ~ 0 for the first 2000 m. Off large deltas the slopes are gentle and
smooth, averaging 1 Va to 2000 m, but with numerous irregularities
0
(due to slumping and pressing out of soft strata?). Off fault coasts
the average slope is 5%o . A long the Pacific coasts the slopes are
steeper than the average, but the presence of deep-sea troughs close
along the coast does not coincide with exceptionally steep continental
slopes. Mud covers no less than 60 70 of the slopes, 25 70 are covered
by sand, 570 by shells and ooze, and 10 70 are occupied by rock.
Veatch and Smith showed the existence of an ancient "Franklin
Shore" at 70 to 110 m depth on the Atlantic shelf of North America.
Bourcart found two or even three terraces below one another along
the eastern Atlantic border. He assumes that the deepest at 500 to
1000 m is of Mio-Pliocene age (Pontien), one at 200-500 m is sup-
posed to be Upper Pliocene (Villefranchienne), while the terrace at
0-200 m is of composite age from Mid-Paleolithic to Recent. On the
other hand elevated marine terraces occur along many coasts up to
heights of 100 m, perhaps even more. The lower of these may be
attributed to eustatic rises in consequence of the interglacial melting
of all ice on the continents. But this does not apply to the higher
ones, because a rise of only 40 to 50 m could be accounted for by
that mechanism. Further morphological evidence of recent updoming
of the continental border comes from many localities. Hence, one
cannot avoid the conclusion that the continental margin is unstable
156 THE SEA BASINS
"'- '"
lIoc:!" : 1-";
HENRY I
$ ['-....: CD
...
~ =! : ["'-...
~
SECTION
i
o
"rrUCTlte "OllUIO"'
CLCYA'TIONJ OtTl'_WlNCD IT
1I.t'"AtTtOH SElitIIOct""",
;j -
i --- ... - l:
80'Oa
:: ''b. :
STATUTE M ILEI EAST or PETERSeURG , ::c "'12000
0
20 40 eo eo 100 120 140 leo 180
FIG. 86. Two of Ewing's sections of the continental shelf, eastem United States.
(Note exaggeration of vertical scale.) (After Daly, 1942, The Floor of the Ocean,
University of North Carolina Press, Fig. 59, p. 109.)
o 50 100
Kilometers
FIG. 87. Sm.lcture of the continental terrace as revealed by deep wells, North
Carolina. Position arranged according to distance from coastline; drilling data
according to Swain, 1947.
FIG. 88. Diagrams illustrating various views on the constitution of the conti-
nental terrace. A. Due mainly to marine erosion of faulted structure with col-
lapse and some deposition. B. Due mainly to ourbuilding and moderate marine
erosion. C. Due to upbuilding on subsiding substratum. D. Due to combined
upbuilding and moderate outbuilding on subsiding substratum. E. Similar to D
but with greater outbuilding by a large river. T he vertical scale, and hence
also the amount of outbuilding, is exaggerated.
THE CONTINENT A L TERRACE 161
cult to explain the steeper front represented by the continental slope,
but otherwise this picture can account for the data so far obtained.
In the view represented by Fig. 88, D, the terrace is formed by up-
building on a substratum which subsides under the weight of the
accumulating deposits while outbuilding also takes place. This out-
building is due to sediment transported from the coast to the edge of
the shelf and deposited on the slope in the undisturbed water below
wave base. The isostatic depression it causes cannot be entirely local,
according to the theory of regional isostatic compensation. The
strength of the cruSt, although insufficient to carry large masses with-
out bending, is enough to cause depression over an area several dozen
kilometers (probably 50 to 150 km) broader than the burden imposed
on the crust. Hence the shelf surface is depressed together with the
continental slope. The space thus provided by deepening allows ac-
cumulation to proceed not only on the slope but also on the top. A
distinction can therefore be made between the material deposited on
the shelf and that laid down on the slope. These grade into each other
at the break in slope at the edge of the shelf. The point of transition
is gradually moved seaward as the terrace builds outwards. Owing
to the slight slopes and great breadth of the area involved, the result-
ing deposits approach horizontal strata and the relative amount of
outbuilding is much less than in the diagram of Fig. 88, D.
Supposing the outbuilding to be double the amount of upbuilding,
the total advance towards the ocean need only amount to some 10 km.
The whole structure from coast to deep-sea floor, however, is of the
order of 100 to 200 km. If the outbuilding is five times as much as
the upbuilding it would have totaled some 25 km. If the sinking is
due only to isostasy the present surface must lie on the average some-
what above the original surface, because the new deposits are less
dense than the substratum on which the crust is supposed to float.
It may be estimated that three-quarters of the volume of the terrace
sediments lies below the original slope surface.
Off the mouth of a large river the delta will advance further than
normal into deep water. The load of sediment on the crust is. then
greater, and an equivalent increase in the amount of depression must
result. The original surface can then be carried far below the level
of the deep-sea floor (Fig. 88, E).
The geology of the coastal land area can be explained by assuming
that the depression of the shelf formerly reached some distance inland
from the present coastline. Later the depressed area rose at the inland
side, causing emergence. This might be due in part to gradual shift-
162 THE SEA BASINS
ing of the main zone of accumulation towards the ocean as outbuilding
proceeded. The main cause, however, might be sought in a slow,
plastic bending of the crust, whereby the regionally depressed area
recoils to its original position, while the loaded area sinks farther.
The deep wells cannot show which of the n yo views on terrace
structure represented in Fig. 88, C and D, is nearer the truth. The
seismic results (Ewing et al., 193 7) show that more consolidated strata
approach the surface near the edge of the shelf, a finding that has been
interpreted as indicating basining of the terrace during Tertiary times
and rise of the Cretaceous deposits towards the edge. But this inter-
pretation is in need of confirmation. Sampling in the Georges Bank
submarine canyons demonstrated the outcropping of Upper Creta-
ceous rocks at depths of 550 m. But the position is 10 km in from
the 550-m depth contour on the slope beside the canyon. Hence.
the terrace may have been built out several kilometer!> during the
Tertiary by fore-set beds which lie in the blank area seaward of the
samples procured.
In the view put forward by Umbgrove (1947) the cause of subsid-
ence lies, not in external forces, but in reaction to internal processes.
As the observed subsidence can be accounted for by isostasy alone.
the present author prefers to assume purely local tectonic deformations.
Conclusive evidence as to the structure of the east-coast terrace is
thus found to be lacking, but the combination of upbuilding with
moderate advance towards the deep Atlantic is the most logical work-
ing hypothesis.
There is doubt what agent has caused the greater depth of the shelf
off glaciated regions. The trenches remaining in these terraces are
evidently due to glacial scour. But even if piedmont glaciers devel-
oped on the shelf it is not clear how a general abrasion could result
together with great deepening of the trenches. Another explanation
is that the depressed position is due to loading of the crust by the
Pleistocene land ice and that isostatic recoil will ultimately bring the
shelf back to normal depth.
Special mention should be made of the conclusion to which Emery
and Shepard came concerning the development of the continental bor-
derland off southern California, on the strength of sampling described
above: "The widespread distribution on the sea floor of rocks similar
to those found on land precludes the possibility that the continental
borderland off southern California is a wave-built terrace dislocated
by faults" (p. 448). They believe that post-Miocene pre-Pleistocene
diastrophism must have broken up the borderland into blocks and
THE CONTINENTAL TERRACE 163
basins. The submarine fault scarps are attributed to even later move-
ment in order to explain the absence of submarine canyons where the
scarps are found.
It would not be right, of course, to transfer these conclusions to
all shelves and continental margins, for the borderland off California
is evidently an exceptional feature.
149
8
FIG. 89. Frequency curves of terrestrial topography. Levels are shown in kilo-
meters. In the left part, a cumulative frequency curve; areas occupied by each
level are gi ven in millions of square kilometers. The same data are shown on the
right in a simple frequency curve.
'f
FIG. 90. Hypothetical models of the eartH's crust with deduced frequency curves
of elevation.
(Co .. ti.....d 0" facinll pat/I.)
THE CONTINENTAL TERRACE 165
in the isostatic equilibrium of the crust that calls for differences in
density.
If the crust consisted of a large number of units of differing densi-
ties and thicknesses one could reason that the frequency curve should
resemble a Gausz curve with one maximum at minus 2450 m, the
average level of the crust. For, according to the laws of probability,
the greatest deviations from the average should occur with least fre-
quency. Now that two maxima occur, is that not sufficient reason
for postulating uniform sial floating in uniform sima?
This, however, is only one out of two possibilities. The explana-
tion may be sought in the existence of two main types of crustal ma-
terial (Fig. 90, C). But the lighter type should then also show uni-
form thickness; otherwise there need still not be a second maximum
frequency (Fig. 90, B). There is the other possibility, pointed out by
Bucher, that denudation and deposition have caused the second maxi-
mum. If these factors had worked on uniform sial of variable thick-
ness or on sial of variable density and thickness, the result would in
both cases have been to bevel off the parts emerging above sea level.
Isostatic rise to counteract denudation would have continued until
each section of the continent was in isostatic equilibrium, with its
surface slightly above sea level. The denudation products should
form deltalike screes around the margins with a flat top at wave base
and a gradual slope towards the floor of the deep sea.
NaturalJy, with uniform sial, uniform thickness of the continents
would eventually prevail, while with sial of variable density the lower
surface should conform to the amount of hydrostatic support required
for maintaining the surface of the section in question at sea level (Fig.
90, D or E).
Which of the three suggestions (C, D, E, Fig. 90) resulting in a
double frequency maximum is the most probable? The view ex-
pressed in C cannot account either for the shape or for the position of
Bibliography
AsSOCIATION D'OciANOGRAPHlE PHYSIQUE. Report on the Criteria and N omencla-
ture of the Major Divisions of the Ocean Bottom, 124 pp., University of
Liverpool, 1940.
BAJtRELL,}. Rhythms and the Measurement of Geologic Time, Bull. Geol. Soc.
Am., Vol. 28, pp. 776-785, 1917.
BEMMELEN, R. W. VAN. Das Permanenzproblem nach der Undationstheorie, Geol.
Rundschau, Bd. 3D, pp. 10-20, 1939.
BETZ, F., and H. H . HESS. The Floor of the Nonh Pacific Ocean, Geographical
Review, pp. 99-116, 1942.
BORN, A. Die Entstehung der Ozeane, Meereskunde, Bd. 18, 3, Ht. 199, 31 pp.,
1932.
Obcr Werden und Zerfall von Kontinentalschollen, Fortschr. Geol. u. Pal., Bd.
10, Ht. 32, pp. 345-422, 1933.
BOUR~RT,]. Geographie du fond des mers, Etude du relief des oceans, 307 pp.,
Payot, Paris, 1949.
BoWIE, W. The Origin of Continents and Oceans, Sci. Monthly, Vol. 41, pp.
444-449, 1935.
BUCHER, W. H . The Defonnation of the Eanh's Crust, 518 pp., 1933.
Mechanics of Crustal Defonnation, Digest, Vol. 10, Tulsa Geol. Soc., pp. 50-61,
1941/42.
CLEMENTS, T., and K. O. EMERY. Seismic Activity and Topography of the Sea
Floor off Southern California, Bull. Scism. Soc. Am., Vol. 37, pp. 307-313, 1947.
CLoos, H. Zur Grosztektonik Hochafrikas und seiner Umgebung, Geol. Rund-
schau, Bd. 28, pp. 333-348, 1937
Hebung-Spaltung-Vulkanismus, Geol. Rundschau, Bd. 3D, pp. 403-527, 1939.
Tektonische Bemerkungen iiber den Boden des Golfes von Aden, Geol. Rund-
schau, Vol. 33, pp. 354-363, 1942.
DALY, R. A. ArchitectuTe of tbe Earth, 211 pp., Appleton-Century, New York,
1938.
For: Relevant facts and inferences from field geology, see Gutenberg, 1939.
The Floor of the Ocean, 177 pp., University of Nonh Carolina Press, 1942.
DANCEARD, L. Observations de geologie sous-marine et d'oceanograpbie relatives
a
fa Mancbe, 295 pp., These, Paris, 1929.
DIETZ, R. S. Some Oceanographic Observations on Operation Highjump, V. s.
Navy Electr. Lab., Rept. 55, 97 pp., 1948.
BIBLIOGRAPHY 171
DREHER, O . Besteht der Boden des pazifischen Oleans aus Sima oder aus Sial?
Geol. en Mijnb., pp. 39--48, 1942.
Du TOIT, A. L. Our WlI1Idering Continents, 366 pp., Oliver & Boyd, Edinburgh,
1937.
EMERY, K. 0 ., and F. P. SHEPARD. Lithology of the Sea Floor off Southern Cali-
fornia, Bull. Geol. Soc. Am., Vol. 56, pp. 431--478, 1945 .
ESCHER, B. G . Moon and Earth, Proc . Kon. Akad. 'II. Wetensch. Amsterda'm,
Vol. 42, pp. 127- 138, 1939.
EWING, M., A. P. CRARY, H. M. RUTHERFORD, and B. L. MILLER. Geophysical
Investigations in the Emerged and Submerged Atlantic Coastal Plain, Bull.
Geol. Soc. Am., Vol. 48, pp. 753-812, 1937.
EWING, M., G. P. WOOl.LARD, A. C. VINE, and J. L. WORlEL. Recent Results in
Submarine Geophysics, Bull. G eol. Soc. Am., Vol. 57, pp. 909-934, 1946.
FAIRBIlIDGE, R. W. The Juvenility of the Indian Ocean, Scope, ]. Sci. Un. Un. W .
AustT., Vol. I, 3, pp. 29-35, 1948.
FOURMARIER, P. Que faut-il penser de la notion de la permanence des oceans .
Bull. soc. beIge, d'Ctud. geogr., T. 10, pp. 26--48, 1940.
FRIEDt.ANDER, 1. Regelmassigkeit der Abstande vulkanischer Eruptionszentren,
Zeitschr. f. Vulkal1ologie. Bd. 4, pp. 15-32, 1918.
GEYER, R. A. Annotated Bibliography of Marine Geophysical and Geological
Surveys, Bull. Geol. Soc. Am., Vol. 59, pp. 671-696, 1948.
GLAESSNER, M. F., and C. TEICHERT. Geosynclines : A Fundamental Concept in
Geology, Am. J. Sci., Vol. 245, pp. 465--482. 571-591, 1947.
GUTENBERG, B. Geotektonische Hypothesen, Hal1db. d. Geophysik, Bd. 3, Lf. 1.
pp. 442-547, 1930.
GUTENBERG, B., et al. Internal Constitution of the Earth. Physics of the Earth,
Vol. 7, 413 pp., McGraw-Hili, New York, 1939.
GUTENllERG, B., and C. F. RICHTER. On Seismic Waves (second paper), Gerlands
Beitr. Geoph., Bd. 45, pp. 280-360, 1935.
GUTENBERG, B., and C. F. RICHTER. Seismicity of the Earth, Geol. Soc. Am., Spec.
Paper 34, 131 pp., 1941, and Bull., Vol. 56, pp. 603-668, 1945.
HAue,:t. Trai:e de geologie, T. 1,546 pp., Armand Colin, Paris, 1907.
HESS, H . H . Drowned Ancient Islands of the Pacific Basin, Am. J. Sci., Vol. 244,
pp. 772-791, 1946.
Major Structural Features of the Western North Pacific, Bull. Geol. Soc. Am.,
Vol. 59, pp. 417--446, 1948.
HILLS, G . F. S. The Formation of the Continents by Convection. 102 pp., Arnold,
London, 1947.
HOBBS, W . H. Mountain Growth, a Study of the Southwestern Pacific Region,
Proc. Am. Philos. Soc., Vol. 88, pp. 221- 268, 1944.
HOOBOM, A. G. Die Atlantislitteratur unserer Zeit, Bull. G eol. Inst. Upsala, Vol.
28, pp, 17-78, 1941.
HOLTEDAHL, O. The Submarine Relief off the Norwegill1l Coast, Norske Vidensk,
Akad. Oslo, 43 pp., 1940.
JOHNSON, D. W. Shore Processes and Shoreline Development, 584 pp., John
Wiley & Sons, New York, 1919.
KING, W. B. R. The Geology of the Eastern Part of the English Channel, Quart.
J. Geol. Soc. London, Vol. 104, pp. 327- 338, 1948.
172 THE SEA BASINS
KNOTHE, H. Die Entstehung der Ozeane und der Wasserhaushalt des Weltmeeres,
Peterm. Mitt., Bd. 79, pp. 126-129, 1933.
KOBER, L. Tektonische Geologie, 492 pp., Borntrager, Berlin, 1942.
KOSSINA, E. Die ErdoberfHiche, Handb. d. Geophysik, Bd. 2, pp. 869-954, 1933.
KUNKEL, E. Geologie Afrikas, I, 461 pp., Borntrager, Berlin, 1925.
KUENEN, PH. H. Geological Interpretation of the Bathymetrical Results, The
Snellius Expedition, Vol. 5, part 1, 124 pp., Brill, Leyden, 1935.
Quantitative Estimations Relating to Eustatic Movements, Geol. en Mijnb., pp.
194-201, 1939.
Kruistochten over de indiscbe Diepzeebekkens, 220 pp., Leopold, The Hague,
1941.
Volcanic Fissures with Examples from the East Indies, Geol. en Mijnb., 7, pp.
17-23, 1945.
The Formation of the Continental Terrace, British Assoc. Adv. Sci. (in press).
LAWSON, A. C. Insular Arcs, Foredeeps and Geosynclinal Seas of the Asiatic
Coast, Bull. Geol. Soc. Am., Vol. 43, pp. 353-381, 1932.
LEET, L. D. Practical Seismology and Seismic Prospecting, 430 pp., Appleton-
Cenrury, New York, 1938.
MECKING, L. Ozeanische Bodenformen und ihre Beziehungen zum Bau der Erde,
PetC1'1n. Mitt., Bd. 86, pp. 1- 10, 1940.
MOHOROVICIC, S. Das Erdinnere, Zeitschr. f. angew. Geophysik, Bd. 1, pp. 330-
383, 1925.
MURRAY, H. W . Profiles of the Aleutian Trench, Bull. Geol. Soc. Am., Vol. 56,
pp. 757- 782, 1945.
Topography of the Gulf of Maine, Bull. Geol. Soc. Am., Vol. 58, pp. 153- 196,
1947.
NATIONAL RESEARCH CoUNCIL. Bull. 85, Physics of the Earth,S, Oceanography,
5Bl pp., 1932.
NOLKE, F. Zur Tektonik des atlantischen Beckens, Geol. Rundschau, Bd. 30,
pp. 21-27, 1939.
NOVAK, V. J. On the Origin of the Continental Shelf, Mem. soc. roy. lettres et
sci. de Boheme, Cl. sci., Vol. 18, pp. 1-27, 1937.
PRATT, W. E. Petroleum on the Continental Shelves, Econ. Geol., Vol. 42, pp.
83-85, 1947.
REHM, H. Die Erdbebentatigkeit der Weltmeere sowie ihre Beziehungen zur
Tektonik, Veroff. Reicbsan. f. Erdbebenfor. lena, Ht. 27, 24 pp., 1936.
RIEL, P. M. VAN. The Bottom Configuration in Relation to the Flow of the Bot-
tom Water, The Snellius Expedition, Vol. 2, part 2, Chapter 2, 63 pp. +
chares, Brill, Leyden, 1934.
RITTMANN, A. Ober die Herkunft der vulkanischen Energie und die Entstehung
des Sials, Geol. Rundschau, Bd. 30, pp. 52-60, 1938.
ROM, J. P. Hypothese sur la formation de I'ocean atlantique, Compt. rend. acado
sci., T. 224, pp. 1295-1297, 1947.
ROUCH, J. Traite d'oceanographie physique, Sondages, 256 pp., Payot, Paris, 1943.
SAPPER, K. Vulkankunde, 424 pp., Engelhorn, Stuttgart, 1927.
SCHOTT, G. Geographie des indiscben und stillen Ouam, 413 pp., Boysen, Ham-
burg, 1935.
Geographie des atlantiscben Ouam, 438 pp., Boysen, Hamburg, 1942.
BIBLIOGRAPHY 173
ScHuCHERT, C. Gondwana Land Bridges, Bull. Geol. Soc. Am., Vol. 43, pp.
875- 916, 1932.
SENN, A. Die Geologie der losel Barbados B. W. I. (Kleine Antillen) und die
Morphogenese der umliegenden marinen Grossformen, Eel. Geol. H elv.,
Vol. 40, pp. 199-222, 1947.
SHEPARD, F. P. Submarine Geology, 348 pp., Harper, New York, 1948.
SHEPARD, F. P., and K. O. EMERY. Submarine Topography off the California
Coast, Geol. Soc. Am., Spec. Paper 31, 171 pp., 1941.
SMIT SIBINGA, G. L. On the Petrological and Structural Character of the Pacific,
Verb. Geol. Mijnb. Gen., Geol. Serie, Vol. 13, pp. 335-354, 1943.
SONDER, R. A. Zur Tektonik des atlantischen Ozeans, Geol. Rundschau, Bd. 30,
pp. 28-51, 1939.
STAUll, R. Der Bewegungmtecbanimtus der Erde, 270 pp., Borntrager, Berlin,
1928.
STEARNS, H. T. Late Geologic History of the Pacific Basin, Am. ]. Sci., Vol. 243,
pp. 614-626, 1945.
STETSON, H. C. Geology and Paleontology of the Georges Bank Canyons, Bull.
Geol. Soc. Am., Vol. 47, pp. 339-366, 1936.
STILLE, H. Geotektonische Probleme des pazifischen Erdraumes, Abh. Preusz.
Akad. Wiss., Math .-naturw. KI., Nr. 1I, 77 pp., 1944.
SVERDRUP, H. U ., M. W. JOHNSON, and R. H. FLEMING. Tbe Oceans, 1087 pp.,
Prentice-Hall, New York, 1942.
SWAIN, F. M. Two Recent Wells in Coastal Plain of North Carolina, Bull. Am.
Assoc. Petro Geol., Vol. 31, pp. 2054-2060, 1947.
TESCH, P., and TH. REINHOLD. De Bodem van het Zuidelijk Uiteinde der Noordzee,
Tijdscbr. ICon. Ned. Aard. G en., Vol. 63, 1, pp. 72-84, 1946.
UMBGROVE, ]. H. F. On the Origin of Continents and Ocean Floors, ] . Geol.,
Vol. 54, pp . 169-178, 1946.
The Pulse of the Earth, 358 pp., Nijhoff, The Hague, 1947.
VEATCH, A. C., and P. A. SMITH. Atlantic Submarine Valleys of the United States
and the Congo Submarine Valley, Geol. Soc. Am., Spec. Paper 7, 101 pp. +
12 chartS, 1939.
VENING MEINESZ, F. A. Gravity over the Hawaiian Archipelago and over the
Madeira Area. Conclusions about the Earth's Crust, Proc. Kon. Akad. V .
Wetenscb. Amsterdam, Vol. 44, pp. 2-12, 1941.
Topography and Gravity in the North Atlantic Ocean, Proc. Kon. Akad. V.
Wetensch. Amsterdam, Vol. 45, pp. 120-125, 1942.
De Verdeling van Continencen en Oceanen over het AardoppervIak, Versl. Kon.
Akad. V. Wetensch. Amsterdam, Afd. Nat., Vol. 53, pp. 151-159, 1944.
Shear Pattems of the Earth's Crust, Trans. Am. Geopbys. Union, Vol. 28, pp.
1-61, 1947.
VENING MEINESZ, F. A., ]. H. F. UMDCROVE, and PH. H. KUENEN. Gravity Expe-
ditions at Sea, 1923-1932, Vol. 2, 208 pp., Waltman, Delft, 1934.
WALTHER,]. Die Methoden der Geologie als historischer und biologischer Wis-
senschaft, Handb. Bioi. Arbeitmtetb., Abc. 10, 599 pp., Urban und Schwarz-
enberg, Berlin, 1926.
WAYLAND VAUGHAN, T. /mernlrtiollaJ Aspects of Oceanograpb y, 225 pp., N:lt.
Acad. Sci., Washington, 1937.
174 THE SEA BASINS
WEGENER, A. Die Entstehrmg deT Kontinente u1Jd Ozetmc, 242 pp., Vieweg,
Braunschweig, 1941.
WILLIS, B. Isthmian Links, Bull. Geol. Soc. Am., Vol. 43, pp. 917- 952, 1932.
WILSON, J. T. The Love Waves of the South Atlantic Earthquake of August 28,
1933, Bull. Seism. Soc. Am., Vol. 30, pp. 273-301, 1940.
WISEMAN, ]. D. H., and R. B. S. SEWELL. The Floor of the Arabian Sea, Geol.
Mag., Vol. 74, pp. 219-230. 1937.
WORZEL, J. L., and M. EWING. Explosion Sounds in Shallow Water, Mem. Geol.
Soc. Am., Vol. 27, pp. I-53, 1948.
c H A p T E R T H R E E
The Indonesian
Deep-Sea Depressions
the west, with connecting links in the southwest towards Java and
the Lesser Sunda Islands, a link eastwards to New Guinea, and north-
wards to Mindanao. Halmahera is joined to New Guinea and Min-
danao by island festoons, while Borneo and the Philippines are con-
nected by three rows of islands crowning submarine ridges. Vol-
canoes are met with from Celebes northwards and on Halmahera.
The shape of the depressions in between will be described separately
later.
The Moluccan islands are composed mainly of Paleozoic, Meso-
zoic, and Tertiary sediments and igneous rocks, showing moderate
to intensive orogenic disturbances. During Upper Tertiary times
overthrust sheets were developed on the Outer Banda Arc and in
eastern Celebes. A covering of young volcanic products characterizes
not only the volcanic areas but also large tracts of Celebe~. Recently
elevated reef limestones are encountered on most of the islands, locally
up to heights of 1000 m.
The gravity survey by Vening Meinesz revealed the occurrence
of large divergences from isostatic equilibrium. Along a narrow belt
there exists a great deficiency of gravity in which some of the most
extreme anomalies known on earth were measured. This belt of
negative anomalies follows all along the Outer Banda Arc and is con-
tinued northwards between Celebes and Halmahera in the direction
of the Philippines. An offshoot appears to follow the eastern arms
of Celebes. The positive anomalies are either arranged in belts fol-
lowing the negative strip on both sides, or they form separate broader
fields.
Indonesia forms a strongly seismic area. Although devastating
shocks are rare, the number of recorded epicenters is surprising. Shal-
low-focus earthquakes are clustered along the Outer Banda Arc; deep-
focus earthquakes are situated mainly on the Asiatic side of this strip.
An excellent summary of geological knowledge, especially from
the stratigraphical point of view, has been given by Umbgrove (1938).
Critical discussions of various theories on the structure and geophysics
may be found in Vening Meinesz, Umbgrove, Kuenen (1934), and
Kuenen (1935). A recent survey is provided by Umbgrove (1947).
A few dozen hypotheses concerning the geological origin of Indo-
nesia have been propounded during the last half century, but since
new data have come to hand most of them no longer hold and need
not be considered here. Thus many writers have postulated long
faultlines demarkating the island margins (Elbert, Voltz) (Fig. 92) .
The new bathymetr~cal chart, however, brings unequivocal proof that
ORI GI N OF DEPRESSIO N S 177
the morphology does not confirm the lines drawn by those authors.
Neither are indications revealed of any such faults continuing for
hundreds of kilometers.
Furthermore, the conception that all deep depressions are due to
vertical faults, by which continental areas were dropped to great
depths, certainly does not merit general application. Yet the sugges-
A B c
Flc. 92. G eotectonic map of the Moluccas according to Elbert. A . Mountain
arcs. B. Faults deduced from structure. C. Faults deduced from morphology.
(After Rutten, 1927, Fig. 155, p. 525.)
tion made by Verbeek (1908 ) , that Australia and Asia were formerly
connected via a continental Indonesian block finds support in several
observations. It will be sho'vn later that the Moluccan region was
formerly partly covered by epicontinental seas, in which depressions
were later developed by subsidence.
The picture first painted by Wegener (1941) and later developed
by Smit Sibinga of the evolution of the Moluccan Archipelago (192 7,
1933) is entirely different (Fig. 93). They suppose that insular arcs
were formerly suspended from the Asiatic continent, with uncovered
pools of sima in between fonning deep depressions. The mass of
Australia with New Guinea is thought to have drifted up against these
festoons, and to have warped nd distorted them into their present
shapes.
178 IN DON E 5 I AND E E P - 5 E A 0 E PRE 5 5 ION S
ORIGIN OF DEPRESSIONS 179
Serious objections can be raised to this conception (Kuenen, 1935),
the more important of which are as follows (Plate A). 1. Australia
and Asia are connected by shallow submarine ridges (New Guinea-
Obi-Sula islands; New Guinea-Halmahera-Snellius Ridge). This ab-
sence of a seam in the form of a continuous trough between the parts
welded together is not in accordance with the expected result of sial-
bergs drifting up against each other. If the latter were true the
morphological structure should be directed parallel, not at right angles,
to the joining line. Biogeography requires former dry-land connec-
tions between the Australian continent and the Moluccas. But there
is no apparent manner in which the line along which welding took
place could become bridged by land connections.
2. The trough from Ceram to Timor continues uninterrupted into
the Java Trough. Obviously they form one genetic unit, although
the depth varies considerably. From the first objection it followed
that no continuous seam marks the blocks supposed to have been
joined. The present argument clearly shows that the trough sup-
posed by Wegener to mark the line of junction can be traced far
beyond the region of contact, so that it cannot represent a junction
(Fig. 94).
3. Several ridges discovered since Smit Sibinga traced three dis-
rupted insular arcs cannot be fitted into the picture of former festoons.
Examples are the Snellius Ridge north of Morotai, the Luymes Ridges
south of Buru, the submarine ridge east of the Kei Islands, the ridge
upon which is situated Kisar, etc. But if not all submarine ridges are
parts of former arcs, the entire reconstruction loses its foundations.
For then the morphology must be younger than the time of junction,
in which event former festoons could not be deduced from the
present forms.
4. Some Mesozoic rock series are identical in fauna and facies on
New Guinea with Misool on the one hand (Australian block) and the
Moluccas (Asiatic) on the other, forming in fact connecting strips
between the two continents. This demonstrates that a close connec-
tion existed at a time when Wegener supposed several thousands of
kilometers of deep sea to have separated the two regions. The drifting
would then have brought these identical portions of the two blocks
directly opposite each other-an improbable coincidence.
5. The stratigraphy of the Moluccas proves the former existence
of denuding land areas adjoining the present islands. Moreover, the
present coastlines cut off the tectonic and stratigraphic structures
obliquely. Both facts testify to the subsidence of land directly along-
180 INDONESIAN DEEP-SEA DEPRESSIONS
side the present island arcs. According to the theory of drift, how-
ever, the islands have been bordered by sima pools since a compara-
tively remote past.
6. The last point shows that the deep-sea troughs did not develop
until intensive folding of the Outer Banda Arc had taken place in
Miocene times. Two explanations might be offered for this fact under
the theory of drift. (a) The meeting of Australia with Asia occurred
ORIGIN OF DEPRESSIONS 181
after the orogenic period mentioned. But this would mean that
the folding of the rocks on the Outer Banda Arc had taken place by
being pressed up against the surrounding sima. This is not in har-
mony with the postulated ductile nature of sima. Moreover, Wegener
explains the formation of the island arcs by the rupture of Asia drift-
ing away westwards from the ancient oceanic section of the Pacific,
while the island structures testify to compression. (b) The other
suggestion would be that the deep-sea relief of Indonesia dates from
after the joining of Australia to the Moluccas. In this event the
orogenic compression could be attributed to the pressure exerted by
the Australian block. But, as pointed out above, this would imply
that the present structure and morphology cannot serve as a key to
the geological history prior to the advent of Australia. The island
arcs would not represent Wegener's floating strips of sial, but secon-
dary structural elements. Obviously both suggestions lead to deduc-
tions that cannot be fitted into the theory of continental drift.
7. The supposed intensive buckling of island arcs, without their
having toppled over or even capsized, is hard to maintain and would
at any rate need a highly mobile and unresistant sima in between.
Under such conditions the pools of sima should show perfect hydro-
static equilibrium. Yet some, like the Weber Deep, are 2~ km lower
than the normal level of sirna.
8. If the island arcs are such narrow strips of sial, how could the
batholiths with sialic magma underlying the volcanic Inner Banda Arc
have been maintained? The volcanoes would almost resemble bottles
full of magma bobbing around in a sea of sima!
9. The gravity field is entirely discordant with the system of arcs
as traced by Smit Sibinga. Thus there is but one belt of negative
anomalies as against three postulated arcs. The belt disappears more
or less where Smit Sibinga connects Buru to the Tukangbesi Islands,
but is present, on the other hand, from the Banggai Archipelago
northward up the Moluccan Passage, where he infers no connections.
Halmahera is part of his arc but is not characterized by negative
anomalies. From these and other misfits it follows that the strip of
negative anomalies must have come into being after the buckling of
Smit Sibinga's island arcs. Enormous displacements of mass must have
occurred to bring about the negative anomalies. No less than 1000
to 1500 km s of sima must have been replaced by sial per kilometer
length of the belt, or several times the volume of the island arc above
the level of the deep sea. It is improbable in the extreme that such
drastic structural revolutions could have taken place partly in perfect
182 I N DON E S I AND EE P - SEA 0 E PRE S S ION S
coincidence with the present shape of the arcs (Buro to Timor),
partly entirely independent of them (north of Buru). And this must
be supposed to have occurred without influencing the surface struc-
tures, because they are taken as a trustworthy key to the buckling
and displacements of former simple arcs.
All these and several other arguments are opposed mainly to the
local details of Wegener's theory as applied to the Moluccan region.
They cannot carry much weight in our judgment of the theory as
a whole.
Although the application of continental drift to the Indonesian
region, as suggested by Wegener and others, appears to lead to
contradictions, an attempt can be made to apply the general principle
in a modified form. The present writer proposed (1935) a tentative
alternative to Wegener's reconstruction for southeastern Asia that now
appears to fit Du Toit's maps of the southern hemisphere fairly well
(Fig. 76). Australia is supposed to have remained in contact with
Asia via the Moluccas from the start, but to have swung around
when it came apart from Gondwana and Antarctica. During this drift
Indonesia must have been bent and crumpled. The deep-sea depres-
sions would have been formed by folding and subsidence in and
around the isthmus connecting Australia to Asia.
It was emphasized in the foregoing chapter that many geologists
and geophysicists are strongly opposed to the theory of drift (see
especially Umbgrove, 1947) and that the present writer is also skep-
tical of its merits. The survey just given shows that Indonesia, far
from giving eloquent testimony in favor of Wegener's views, as some
authors have held, presents many features strongly opposed to the
theory of drift. Only by assuming the admittedly farfetched recon-
struction outlined above can a flat contradiction be avoided.
A somewhat different conception of the structural development of
the Moluccan arcs was suggested by Brouwer. He believes that the
Banda Arcs were originally much farther away from the margin of
the Australian continent and simpler in outline. By strong hori-
zontal movements during the thrusting of the overthrust sheets the
arcs approached the border of the continent and adapted themselves
to the outline of the resisting block. The basins and troughs are thus
not considered to be new developments, but they have altered in
shape, breadth, and position (Fig. 95).
Some of Brouwer's arguments have been refuted by Stille, and
Kuenen attempted to show from the new charts of the Snellius expe-
dition that other ~jections can be raised. The distortion of the Outer
ORIGIN OF DEPRESSIONS 183
Banda Arc appeared to follow convincingly from the former chart~.
Two explanations could be offered: either the arc had gradually
adapted itself to the shape of the continental border, as supposed by
Brouwer-it was, so to speak, pressed into a mold; or the arc was
formed primarily parallel to the border, about in its present position.
The new chart confirms the second suggestion (Fig. 96). The
advanced position of the T animbar Islands was explained by Brouwer
from the indenture in the continent directly opposite. Now, how-
ever, this embayment has disappeared on the new chart, bringing the
advanced part of the arc closer to the opposite wall than elsewhere.
In Brouwer's conception this would imply that the arc had bent for-
ward in spite of a smooth wall opposite. The new chart shows,
moreover, that there is no question of an outward bend of this pan
of the arc, but that an en echelon arrangement of two interchanging
swells causes the sudden advance of the Tanimbar Group. This can-
not represent a gradual adaptation but must of necessity be a primary
structure. Thus tw o arguments arc found against the mechanism of
pressure molding.
Opposite the advanced portion of the Kei Islands there does actu-
ally exist an indenture in the continental border, the Aru Basin. But
instead of swinging outwards as assumed by Brouwer, the main mass
of the arc passes by in a smooth curve as may be seen by noting the
steep slope into the adjoining Weber Deep in the rear. Instead a
second (Groot Kei) and third (submarine) ridge have developed
along the outer side of the arc. Again all signs are missing that the
arc has advanced against the continent, and primary irregularities are
strongly suggested.
Kisar, according to Brouwer, formed a part of the Outer Banda
Are, but, in consequence of the sharp angle in the Australian block
just opposite, it was severed from the main mass and left behind.
Not only has the corner of the continent disappeared from the new
chart, but a separate ridge is found of which Kisar forms the culmi-
nation. It further becomes obvious that the depression in Manipa
Strait between Buru and Ceram (Fig. 59) can no longer be interpreted
as the surface expression of a transverse fault with horizontal dis-
placement (Brouwer's contention). A small broad basin, closed to
the south by a ridge, is now found on the chart.
The almost perfect symmetry of the Timor Trough does not accord
with the conception that one slope represents an ancient continental
border and the other the front of an advancing thrust (orogenic)
swell.
FIG. 95. The major tectonic axes of the Moluccas postulated by Brouwer.
Redrawn after Brouwer, 1925, The G eology of the Netherlands East Indies, Fig.
I, p. 58, University of Michigan Press.)
FIG. 96. Main elements of the morphological structure of the southern Moluccas
based on the cham of the Snellius expedition. (After Kuenen, 1935, Fig. 101,
p. 86.)
184
ORIGIN OF DEPRESSIONS 185
Finally it may be recalled that the Outer Banda Arc and the accom-
panying trough pass westwards beyond the Australian continent.
Here, if anywhere, Brouwer's moving geanticlines should have found
the minimum of resistance and could have swung southward unop-
posed. The chart shows, however, that there is no outward bulging
where the continental "resistance" is lacking.
The outcome of this review is thus found to be strongly opposed
to Brouwer's contentions.
A few words must be said concerning the comparison between the
Indonesian island arcs and waves entering a coastal embayment.
Argand was the first to draw this comparison, but merely in order
to describe the general shape. Staub, however, attempted to explain
the structure by this mechanism. To the writer this is inadmissible.
In the case of waves an undulation is propagated but the medium
remains in position. But, under Staub's contention that geanticlines
have adapted themselves to an indenture in the foreland, the rock
itself is pushed forwards, while the undulations remain fixed with
respect to this propagating medium. The physical laws governing
wave mechanics may on no account be applied to a moving geanti-
cline. The fan-shaped spread of waves in a bay is therefore no expla-
nation for the similar outward appearance of orogenic features of the
eanh's crust.
It was pointed out above that strong arguments are at hand to show
that the ad::tptation of the geanticlines to the continental border is a
primary feature. What is termed the border of the continent appears
to be merely the edge of the undisturbed pan of a formerly more
extended block. If that is granted, theorizing on the cause of present
parallelism of arcs and continental border is as futile as questioning
why a canal is parallel to its banks.
Molengraaff was the first to state clearly that the Moluccas may
be considered a folded mountain chain in statu nascendi. This investi-
gator called attention (1913, 1922) to the occurrence of overthrust
sheets of Miocene age and a block-faulted structure of Pliocene age
on the Outer Banda Arc in the Moluccas, especially on Timor and
surroundings. He suggested that during the Pliocene, and possibly
also recendy, the overthrusting may have continued at depth, while
the orogenic processes at the surface remained restricted to fault tec-
tonics. This younger thrusting is not exposed because erosion has
not yet cut deep enough into the crust. Wanner, Weber, and Argand
expressed similar ideas about exposed structures. Most later writers
followed Molengraaff, each postulating more or less original tectonic
186 IN DON E 5 I AND EE p. 5 E A 0 E PRE 5 5 ION 5
lines, guided by the morphology of the region as far as known through
the older soundings, but giving free rein to his imagination for the
areas where the deep-sea charts provided insufficient guidance. Since
publication of the Snellius charts, with their wealth of detail, these
tectonic diagrams may be looked upon as antiquated and their exact
forms will have to be revised. Hence, at present we only need to
inquire into the principles involved in the various structural hypothe-
ses, but not into their details.
Even the map (Fig. 97) given by Hobbs (1944), although based
on the new depth chart, shows a number of "anticlinal arcs" that run
obliquely across submarine ridges (connection between Borneo and
North Celebes, between the two southern arms of Celebes, and be-
tween Sumba and the ridge south of Java), while several major topo-
graphic elements are ignored (Halmahera with the Snellius Ridge to
the northwest, Buru with Sanana to the north, the ridges between
Halmahera and North Celebes), and the postulated ridge on the edge
of the Sahul Shelf does not occur on the deep-sea chart, although
there are a few reefs.
All new data indicate that a generalization for the whole region is
not warranted. From the bathymetrical charts, the distribution of
earthquakes, field work, and gravimetrical surveys it becomes ever
more: apparent that an active belt runs through the archipelago, along
Sumatra, Java, the Outer Banda Arc (possibly also the eastern am1S
of Celebes) , and the Moluccan Passage west of Halmahera. For this
strip the comparison may hold, drawn by Argand in his classical
treatments of the Alps in 1916, between the present conditions in the
Moluccan Arcs and the Mesozoic Alpine geosyncline. In both cases
a double island arc is encountered where the frontal margins of the
thrusting sheets come to the surface. To grasp the full measure of
similarity the map of the Alps must be held upside down (Figs. 98
and 99).
Later Kuenen (1935) viewed in more detail the points of agreement
(the similar arrangement of two major geanticlines with a smaller
element [Kisar-Monte Rosa] in between, crowned by denuding islands
and accompanied by deep troughs; the development of overthrust
sheets) and the points of difference (larger dimensions of the Indo-
nesian orogen, partly surrounded by oceanic basins; far greater in-
tensity of volcanic activity; absence of recent sediments comparable
to synorogenic Flysh deposits; longer duration and more leisurely
development in this region).
DEPRESSIONS 187
188 IN DON E 5 I AND E E P - 5 E A 0 E PRE 5 5 ION 5
The remaining parts of the Moluccas farther from rhe active belt
resemble the "Innensenken" or nuclear basins, or basins in the rear,
such as the Pannonian Basin of Hungary and the Po Basin of northern
Italy.
Umbgrove (1947) is also convinced that the Indonesian belt of
island arcs and troughs is of the same nature as the orogenic belts of
Kilometers
FIG. 98. The structural plan of the Alpine system of Europe (inverted ) accord-
ing to Staub. (Redrawn after Staub, 1928, Fig. 18, p. 33, and Fig. 24, p. 90.)
Kll o me ter ~
FIG. 99. Orogenic belt of Indonesia as indicated by gravity field, on the ame
scale as Fig. 98.
the continents. The great depth of the troughs is not due to abnormal
characteristics but merely to lack of sufficient detritus during their
subsidence.
Stille (1943) holds that the Moluccan orogenic area has passed
through the active period of compression and that the recent vertical
movements represent the final stages of development. Strong com-
pressive movements leading to the formation of Alpine structures are
not going on, nor will they take place in a later stage. Hence, this
region should not be compared to the Mesozoic stage of development
ORIGIN OF DEPRESSIONS 189
this crust found to have bulged upwards during the compression, for
erosion products do not become plentiful in the surroundings until
after the main compressive phase. Consequently the conclusion can-
not be avoided that the main part of the earth's sialic crust is forced
downward during orogenic compression.
On several gronnds it is generally inferred that in continental areas
the crust consists mainly of light rocks like granite, the sial, with
Geosy ncline
'r
1I0~ so
'.
.a
100--" .< 1/
3 0 0 / / .'iU~
700 "- 130
210 140
230
V
225",,~.:
'4- 40
-<7.
'" 70
100
.~150
ISO
40
140
220
~
\~ 0'\....,.1
50 25
76 300
l-'\214
440
much smaller and must have lain entirely within the present depres-
sions. But the uniformity of each of the present basins or troughs,
with their evenly modulated profiles, renders this supposition highly
improbable. Hence the conclusion is warranted that the deep-sea
morphology of the Moluccas is entirely of Upper and post-Tertiary
development.
The formation of an elongate trough with a synclinal cross section
can be readily explained by moderate lateral compression of a floating
crust. The development in close connection with the accompanying
belt of negative anomalies and the strong overthrust structures also
point to lateral compression. On the other hand it is probable that
these two phenomena were caused long before the troughs came into
being. A state of compression causing the older phenomena need no
longer prevail at the present time.
Another explanation is that before the elevation of the arc, when
the sial root had just been created, the adjacent crust was held by
its strength above its natural level on account of the negative anomaly
(Kuenen, 1936). This condition might be compared to a block of
ice locally buoying up a sheet of ice. During the gradual elevation
of the island arc in Pliocene-Pleistocene times in consequence of the
upward pressure of the root, the superelevation of the neighboring
portions of the crust should have been allowed to decrease, because
the buoyant root was gradually diminishing in depth. In other words
a subsidence would occur at both sides of the rising island arc.
This brings us to the moot question whether the present strong
negative anomalies are the gradually decreasing remains of an even
greater Upper Miocene disturbance of isostasy, or whether the com-
pensation occurred long ago but a renewed compression rejuvenated
the negative anomalies. However, to enter into this and similar mat-
ters would lead us into the maze of geotectonic problems and farther
and farther away from ascertained facts and direct deductions there-
from. The discussions berween Vening Meinesz, Bijlaard, Escher,
van Bemmelen, Kuenen, Hobbs, Umbgrove, and others will not be
gone into here.
It will only be pointed out in conclusion to this review that the
remaining deep-sea basins, beyond the region of negative anomalies,
present a U-shaped section and are probably due to purely vertical
forces. They are characterized by positive anomalies of gravity, espe-
cially the Celebes Sea and Banda Sea, less marked in Makassar Strait
and the Gulfs of Bani and Tomini around Celebes. This anomaly
should cause a further downward movement and is thus in accordance
I N DON E S I A N V S. F 0 S S I L BAS INS 195
with the view that these basins were recently developed by subsidence.
For the West Indies similar conclusions can be drawn as to recent
formation of the deep-sea depressions and a young orogenic ally active
area (Rutten, 1934; Hess, 1938).
~____~.""I
FIG. 103. Profiles of some Moluccan deep-sea depressions based on echo sound-
ings of the Snellius expedition. Vertical scale 10 X horizontal. 1,2,7, V-troughs j
3, 4, U-troughs; 6, shallower basin; 5, deeper basin.
have occurred can the correlation given above obtain a deeper signifi-
cance. On theoretical grounds there is no evident reason for denying
chat a geosyncline could ever have attained great depth. It has been
suggested by Hall, Russell, and others that the primary cause of
subsidence is the accumulation of a heavy column of sediments. The
following arguments have been brought forward against this view
(Grabau, Umbgrove, etc.). (1) Only part of the original area over
which sedimentation begins develops into the final geosyncline. (2)
The elongate, regular shape denotes an internal diastrophic cause and
not a chance exogenous cause. (3) The section is synclinal with the
largest depths and therefore the greatest subsidence at a distance
from the margin. As the detritus is normally brought in from one
side, weighting alone would result in a wedge-shaped cross section.
(4) The juxtaposition in the immediate neighborhood of a rising
geanticline indicates the operation of diastrophic forces in the area.
(5) The specific gravity of the sediments is much smaller than that
I N DON E 5 I A N V S. F 0 5 5 I L 8 A 5 INS 203
of the substratum that has to be displaced (even if the paramorphic
principle of Du Toit is given due consideration). Obviously the
weight of sediment deposited in the trough must encourage subsid-
ence, but an endogenous cause must operate as well to bring about
a wavelike deformation of the crust. According to the view of
Vening Meinesz, Bucher, and others this cause is compressional stress.
If this endogenous cause is admitted, it must also be conceded that
the primary subsidence must lead to an unfilled depression if no sedi-
ment happens to be supplied.
A second theoretical consideration is that the determination of
depths in which fossil sediments were deposited is a controversial
problem, especially in the case of bathyal formations (200-1000 m
deep). It is not improbable that some sediments, held to be bathyal,
have been laid down at greater, abyssal depth, provided that the dis-
tance to the coast was limited. The pelagic character of the oceanic
deep-sea deposits is taken far too lightly as indication of great depths,
and it is overlooked that actually the great distance from land has
caused the typical characteristics of the oozes. Thus radiolarian ooze,
pteropod ooze, and red clay can be formed only far from land, and
globigerina ooze is also very scarce nearer inshore. The recent oozes
of the Indonesian deep-sea depressions forming at depths of 5000 to
10,000 m are identical with some deposits from 1000 m or even less.
In fossilized state these would be readily classed as bathyal or even
neritic (0-200 m deep) formations that had been deposited some dis-
tance from the land. The investigation of the Snellius samples has
established that within the range of continental influence depth alone
plays no appreciable part in sedimentation beyond the 200-m line.
Ooze with a high percentage of lime was brought up from depths
of 5000 m in the Sulu Sea, but in the better-ventilated Celebes Sea
the deposits were found to contain but a trace of lime at similar
depths. The normal blue mud of the Weber Trough from 7400 m
was identical with that of the Timor Trough at 1000-m depth or less.
The present author knows of no property by which, in the absence
of typical Foraminifera, it can be proved with certainty that a fossil
hemipelagic sediment was formed in depths of 500 to 1000 m and
not in depths of 5000 to 6000 m (see also Chapter 5). When full
weight is given to this consideration it must be admitted that no
proof is forthcoming that some of the bathyal sediments of the Alpine
geosyncline were not laid down in depths exceeding 1000 m. There
are no grounds for denying the possibility that this geosyncline devel-
204 I N DON E S I AND E E P - SEA D E PRE 5 S ION S
oped depths several times as great as is generally assumed. The same
holds for a few other geosynclines.
The third argument to show that fossil sedimentation basins may
sometimes have grown to considerable depths is the occurrence of
equivalents to recent eupelagic deep-sea deposits among the Mesozoic
series of Timor and Roti in the Moluccas. This situation is the oppo-
site to that cited above, and we are confronted with arguments in
favor of actual deep-sea oozes formed at great distance from a coast
and at abyssal depths. The series contains typical red clay with man-
ganese nodules and Cretaceous sharks' teeth with the bony substance
dissolved so that only the enamel remains (Molengraaff, 1920). In
the recent seas this red clay is found only at abyssal depths, where
lime is dissolved. There are also silicified Monotis limestones, frag-
ments of which have been encrusted by iron-manganese oxides, and
heaped-up cephalopods with crusts of iron-manganese oxides. The
cephalopods represent a considerable portion of the Triassic although
the thickness of the limestones is but a few meters. Finally there
occur fine-grained pink limestones with manganese nodules. The
combination of all these phenomena leaves no doubt as to the pelagic
nature of the series. Considerable depth is also probable because
organisms thriving on the bottom of shallow waters are absent and
because recent equivalents are formed only in abyssal depths. Whethc:r
the alpine radiolarites also represent pelagic deep-sea deposits (Stein-
mann) has not yet been definitely established but appears to be rather
doubtful.
The various arguments presented appear to show that fossil geo-
synclines may have formed temporarily into deep depressions of the
sea floor under special conditions and by way of exception. There-
fore the great depth of the Indonesian deep-sea troughs and basins
cannot be urged as argument against the correlation with fossil sedi-
mentation depressions.
The problem of how far the similarity goes between deep-sea de-
pressions in general and fossil sedimentation depression is therefore
worth examining in more detail. First the deep-sea V -troughs will
be dealt with because they will be found to provide the clearest
evidence. Kuenen extended the arguments already presented by
Argand (1916) and Stille (1919) and showed how closely certain
troughs of the Alpine geosyncline resemble certain of the Moluccan
deep-sea troughs, and Umbgrove drew a parallel with his marginal
deeps.
I N DON E 5 I A N V S. F 0 5 5 I L 8 A 5 INS 205
As with primary and marginal deeps, a rising belt runs parallel to
these deep-sea troughs (Outer Banda Arc) and delivers detritus into
the depressions. At the other margin of the deep-sea V-troughs a
continental shelf or deep-sea floor is found. The ancient primary
and marginal deeps were likewise generally bounded on the outer side
by a shallow sea or an old continental shield. Both types show a more
or less symmetrical V-shaped cross section (synclinal), and both are
markedly elongate with varying depths along the axis. They similarly
form festoons several thousands of kilometers long. Length, breadth,
and depth are also of comparable magnitude (Fig. 104). The marginal
deeps and Indonesian deep-sea V -troughs are similarly situated along
the convex margin of a curved orogenic zone, and both were formed
after the orogenic compression and after this belt began to rise. As
far as known this also holds for deep-sea V -troughs outside the Indo-
nesian area. It appears highly probable that they may also be formed
in virgin areas, in which case they would then form a close parallel
with the simple eugeosynclines.
Formerly the present writer was of the opinion that the entirely
different character of the deposits in the Tertiary Molasse Trough
of the Alps did not allow of a correlation with the synclinoral type
of deep-sea depression. Umbgrove's exposition clearly demonstrates,
however, that this parallelization is warranted. The coarser nature of
the sediment in the Alpine Trough must be attributed to a higher
average level of the crust with larger areas of denudation. Hence it
is not of basic importance from our point of view. Possibly an even
closer similarity obtains with the most northerly of the Penninic
Deeps, the "geosynclinal dauphinois et valaison," or the Helvetian
Deep, as Argand has pointed out.
The deep-sea troughs with U-shaped section also strongly resemble
the corresponding type of sedimentation depressions, namely the in-
tramontane deeps. Both are decidedly elongate and are intimately
related to an orogenic belt. Thus the Indonesian deep-sea U-troughs
are situated between the Inner and Outer Banda arcs with their con-
tinuations to the west. Again in both types long festoons are devel-
oped. The varying breadth and depth are likewise typical of both.
It remains to be shown whether the U-shaped section of these recent
troughs is also a characteristic of the fossil deeps.
Following Staub, Kuenen correlated his basins of the Celebes Sea
type with "Innensenken," while Umbgrove drew a parallel with his
corresponding type of nuclear basins. The more or less equidimen-
sional shape and the framing by geanticlines, the steep walls, and the
206 INDONESIAN DEEPSEA DEPRESSIONS
flat floor are joint characteristics. We already noted that the great
depth of the Indonesian basins may form an objection to the correla-
tion, but that the absence of trustworthy indications concerning the
maximum depth of fossil sedimentation basins leaves us in some doubt.
The other basins (Gulf of Boni type) were correlated by Umbgrove
with his discordant basins. The less regular shape, the abrupt trun-
cation of older structures found on surrounding islands, and the ab-
sence of festoonal arrangement are points of resemblance.
Evidently there are fewer data on which to base a correlation of
the deep-sea basins with the fossil basins, than for the correlation of
deep-sea troughs with geosynclinal deeps.
The identity of the deep-sea troughs and geosynclinal deeps allows
an important conclusion to be drawn. It follows clearly from the
Snellius charts that the Timor deep-sea V -trough and the Java deep-
sea V-trough form one genetic feature. The original question whether
deep-sea troughs and geosynclines are identical should be answered in
the affirmative, if it is admitted that the Timor Trough is a certain
type of geosyncline. For if the shallower Timor Trough is a recent
geosyncline its deep continuation, the Java Trough, must be one also.
Two reservations have to be made, however. In the first place the
deep-sea basins are evidently ruled out from the correlation with
geosynclines, and the case for the deep-sea U-troughs is more open
to doubt than that for the deep-sea V-troughs. In the second place
the great majority of fossil sedimentni{'ln depressions has certainly
never exceeded neritic depths and has therefore never resembled deep-
sea depressions in a morphological sense.
Two interesting aspects may be emphasized in this connection. The
first is that judging by the results of gravity observations the ocean
floor offers the same resistance to horizontal compressive stress as the
continental blocks. The root developing in both cases is of the same
magnitude. It is true that the anomaly of oceanic roots is not so
strongly negative, for instance opposite Java and Sumatra, as in the
continental area of the Banda Arcs. But relative to the surrounding
area rhe deficiency of gravity is nearly as great because of the average
positive anomaly of the sea floor. There is a close parallel in the
morphology, for the accompanying V -troughs are as deeply depressed
with relation to the adjoining ocean floor as is the Timor-Ceram
Trough relative to the adjacent shelf. In this respect gravimetric
and bathymetric results again bear each other out satisfactorily. We
are led to confirm in a decided manner Bucher's opinion: "The origin
of 'welts' and 'furrows' is independent of the elevation of the earth's
I N DON E 5 I A N V s. F 0 5 5 I L BAS INS 207
Diameters in millimeters
Wentworth Atterber,
THE ATMOSPHERE
The particles dropping from the atmosphere are of three different
kinds. Meteoric dust of extraterrestrial origin forms an interesting
though small contribution to some sediments. The earth with its
atmosphere moves continuously through interstellar space and collects
a large number of dust particles, estimated at 10 to 20 million per day.
Three-quarters of them must drop directly into the sea. Some are
readily distinguished, namely minute spheres, that are formed from
"falling stars." When a meteoritr- shoots through the atmosphere a
212 SOURCES AND TRANSPORTATION
FIG. 105. Distribution of ash from the Oena Oena Volcano in bottom samples
from the neighborhood (shown by circles) according to data by N eeb, 1943 .
climates dense sand and dust storms are of frequent occurrence, and
it is certain that particles may be carried in large quantities hundreds
of kilometers from their source. The dropping of thick layers of dust
on board ships far out at sea during pitch-dark dust storms bears evi-
dence to repeated sedimentation of desert dust in the center of oceanic
basins (Fig. 106) .
In the Meteor R eports it was shown that fine Sahara sand, detect-
able by a film of iron oxide, occurs in noticeable amounts in the deep-
sea deposits of the Gulf of Guiana and off the Cape Verde Islands
(Barth, Correns, and Eskola, 1939). Close inshore it is strongly di-
luted by non-eolian material, but farther out it increases relatively.
Not until a few thousand kilometers offshore does the Sahara sand
diminish to imperceptible amounts, the grain size meanwhile gradually
214 SOURCES AND TRANSPORTATION
decreasing. In some fractions more than half the quartz grains were
unmistakably of desert origin. Thus at 300 km from the coast 6070
of the quartz was identified as of desert origin of which one-third was
from 0.1 to 0.5 mrn in diameter. Besides quartz, feldspar, and clay
minerals, a high percentage of calcite occurs in this Sahara sand. As
compared to the size of the equatorial Atlantic the area receiving larger
amounts of desert sand is small, however.
Lesser
Antilles
these, the number of individuals was always relatively small and the
span of life very great. Furthermore the materials building up the
skeletons is not very resistant, especially those of fishes. Consequently
vertebrate remains are seldom very much in evidence. But when in-
sufficient ventilation of basin waters causes anaerobic conditions, scav-
engers and normal bacteria are excluded and then the skeletons or
scales of fishes and other higher animals are excellently preserved.
Incidental or periodic mass mortality on t.he. shelf may have the same
result. The most resistant parts, the teeth (sharks, rays, etc.), ear-
bones of whales, finspines (sharks), or scales (ganoids), are sometimes
found concentrated in fossil deposits, although it is difficult to ascer-
tain the exact causes of this abundance. These most resistant parts
are relatively frequent where the rate of sedimentation is abnormally
low, as for instance in some Pacific areas covered by red clay.
In the deep sea, benthonic forms of life (bottom-living animals and
plants) do not playa very important part, although many species of
Foraminifera, deep-sea corals, holothurians, etc. flourish there. It is
the plankton, the driftmg population of surface waters, that contrib-
utes the bulk of material to sedimentation. Among the forms of life
with a siliceous (hydrated silica) frustule or test the diatoms play the
major role in high latitudes and the radiolarians in tropical waters.
Siliceous spiculas of sponges are of frequent occurrence, on an aver-
age forming 2 to 30/0 of deep-sea deposits. The pelagic Foraminifera,
especially the globigerinas, rank foremost among the suppliers of lime.
In about half of the recent deep-sea deposits the tests of these unicel-
lular animals predominate. The tests of pteropods and the loose plates
(coccoliths) of a certain type of planktonic algae, the Cocco litho-
phoridae, also have a wide distribution, although they hardly ever
attain to half the total bulk of a deposit.
In shallow waters planktonic forms are generally less in evidence.
Only under abnormal conditions, for instance in shelf regions where
upwelling water causes exceptional abundance of plankton, do they
become of greater importance. Examples from fossil deposits are
radiolarians in the formation of radiolarites in shallow water and plank-
tonic Foraminifera in the development of the chalk rocks along the
coasts of the English Channel. On the shelf and in beach deposits
benthonic forms of life predominate, such as gastropods, lamelli-
branchs, brachiopods, echinoderms, Foraminifera, sponges, and corals,
each of which may take the upper hand in turn. Less frequently
worms, crustaceans, cephalopods, Bryozoans, or calcareous algae occur
en masse.
216 SOURCES AND TRANSPORTATION
Not all the lime and siliceous tests delivered by dying organisms or
by shedding of tests is incorporated in the sediment, because part is
again dissolved. This topic will be treated in the next chapter (p. 277).
DECOMPOSABLE ORGANIC MAnER 1
Besides skeletons and tests, the decomposable organic matter of
plants and animals also plays an important part in sedimentary deposits
of the sea floor. Again it is the group of unicellular organisms that
contributes most, the enormous number of individuals compensating
for their small size.
Determination of the production of organic matter and of its disso-
lution is far from simple. Hence quantitative data can be given only
for certain restricted areas and even these should be looked upon as
rough estimates only (Sverdrup et aJ., 1942). For the present the
average production is estimated roughly at 1 kg per m2 per year in
coastal waters, but farther out it becomes less, and doubtless extreme
variations occur.
The composition of sea water, that is to say the proportions of the
principal dissolved elements, is constant except for some scarce nu-
trient salts used by plants in their metabolism. As nearly all elements
necessary for plant life are present in excess of requirement, the avail-
able amounts of a few essential constituents determine the intensity
of xegetable production. Generally phosphorus or nitrogen forms
the bottleneck, and in regions where radiolarians and especially di-
atoms occur in great abundance, available silica may perhaps limit the
production. In some regions, as the Antarctic Ocean, all nutrient salts
are provided in sufficient amounts, in which case other factors limit
plant growth, such as light or the consumption by herbivorous animals.
The distribution of plankton in the oceans clearly indicates that the
fertility of the water is of much greater importance than temperature.
Thus coastal waters that receive an ample supply of nutrient salts from
the adjacent land or from upwelling water, and colder environments
where a yearly overturn of the waters brings fertile water to the sur-
face, procluce greater quantities of plankton than regions far out in
the oceans and tropical waters with stable stratification the year round.
This relation is brought out strikingly by comparing maps of the
southern Atlantic showing the density of planktonic life and the per-
centage of phosphorus pentoxide (Fig. 6). Both indicate a rich zone
along the western coast of Africa with two tongues extending far
into the ocean at 10 north latitude and just south of the equator. On
1 See especially Trask. 1939.
PRECIPITATES 217
both maps a concentration is shown towards Antarctica and also an
isolated area in midocean off Trinidad, 20 south latitude (see Watten-
berg, in Correns, 1934).
Organic matter (possibly held by clay particles) is carried to the
sea by rivers. It has not yet been ascertained whether this organic
matter is incorporated in the sediments or how great a contribution it
makes to the organic content of various types of marine deposits, but
it appears probable that in certain localities this source is of consider-
able importance. In Arctic regions with extensive peat deposits, or
in the tropics where plant life on land is so intensive, this source of
organic matter in marine deposits should be reckoned with (Trask,
1939; Kuenen and Neeb, 1943).
But from initial production to final incorporation in bottom de-
posits there is a long way to go, and something like 98 to 99.98 CYo is
lost, according to Trask. Even after deposition a 10 to 20 % reduction
takes place. Some of the vegetable plankton is swallowed by animals,
and a rich fauna feeds on animal and vegetable plankton. The por-
tion that dies a natural death sinks to the bottom but is then devoured
by benthonic mud-feeders and is further attacked by bacterial decom-
position (rot). Only under anaerobic conditions, as described in
Chapter 1, is most of the organic matter on the sea floor preserved
(SrrpSm, in Trask, 1939; Brongersma-Sanders, 1948).
PRECIPITATES 1
Several substances can be precipitated from sea water, in many
cases by plants or animals, but concerning most of the chemical proc-
esses involved little has as yet been ascertained. Lime may be precipi-
tated when the assimilation of plants extracts too much carbon dioxide
from the water, by which process the solubility of lime is reduced.
Probably the precipitation of lime from sea water is generally a conse-
quence of bacteriological activity. It is supposed that these organisms
produce ammonia, with the result that the more alkaline water will
become oversaturated with lime. Apparently aragonite mud is pro-
duced in this manner in shallow waters of the Bahamas (Thorp, in
Trask, 1939). It has not yet been discovered to what extent bacteria
are also concerned in the formation of oolites. In a later paragraph
(p. 219) further details on the precipitation of lime are given.
Although precipitated iron has been observed in recent sediments,
at the present time no rich deposits of this element are being formed
anywhere on the sea floor. evertheless, concentration of iron in
1 See Correns, 1939.
218 SOURCES AND TRANSPORTATION
and to form a deep notch in suitable types of rock. When the wave-
cut notch becomes too deep the roof subsides, a slide often forming
a reentrant in the cliffs above. But usually the rocks forming the
coast are too incoherent or too strongly stratified or they weather
too quickly to produce a notch. Instead, a protecting heap of boul-
ders may be formed along the foot of the cliff. A balance is then
struck between the rate at which subaerial weathering attacks the
face of the cliff and that at which
the waves attack its base. Lessening
of wave attack will allow more time
for decomposition, solution, and
disintegration to weather back the
cliff to a gentler slope. This, in
turn, will allow more protective
waste to accumulate on the face of
the cliff and thus slow down the
rate of weathering. Conversely, a
slower rate of weathering will al-
low more time for the waves to
break up the blocks accumulating
F,G. 107. Diagram illustrating marine
at the base of the cliff and to carry erosion of various types of country
away the smaiJer fragments. This rock on a steep coast. A, hard, un-
will lead to steepening of the cliff stratified; E, the same when a broad
face (Fig. 107). marine erosion platform has devel-
The direction of dip of bedding oped; B, strata dipping away from
shore; C, strata dipping towards the
planes also influences the shape shore; D, more swiftly weathering
of the cliff. Where the dip is to- rock than A,
wards the land the steepness will be
found to increase. But where the dip is towards the sea portions will
tend to slide off the face of the cliff along the bedding planes, thereby
reducing the declivity. Furthermore, greater height of a cliff will
tend to lessen the slope, because more material has to be carried away
by the waves for each meter of recession. Obviously this must have
an adverse influence on the rate at which coastal erosion is a'ccom-
plished and leave more time for the action of weathering.
It goes without saying that the shape of the coastline is largely
determined by the height and durability of the rocks. The higher
the land and the more resistant the country rock, the more slowly
coastal erosion will advance. Capes will form where resistant rocks
crop out, while bays will correspond to the more easily attacked for-
mations. It is not merely the mass properties of the rock but to a
224 SOURCE S AND T RANSPORTATION
FIG. 109. Block diagram of arch. stack. and sea caves on a surface of marine
erosion.
But although coastal erosion works selectively and tends to cut out
an intricate pattern along the coast, other factors are also involved
that attempt to draw the coastline straight again. In Chapter 1 the
tendency was mentioned of waves to curve around and concentrate
their attack on protruding portions of the coastline, while on the
other hand they fan out in a bay and spend their energy on a larger
section of the shore. This must result in a stronger attack on head-
lands and a certain degree of protection in reentrants.
This explains why headlands often show steep cliffs or at any rate
a bare, rocky, wave-cut platform, whereas bays are almost always
characterized by a sand or shingle beach. As the jutting-out portions
of a coastline are closer to deep water than reentrants, the submarine
slope must be steeper. Naturally this greater steepness of the fore-
shore fronting headlands helps to increase the power and effect of
wave onslaught and the dispersion of the waste products.
226 SOU R C E SAN D T RAN S P 0 R TAT I O N
Where the irregularities of the coastline are not due to differences
in resistance, but to the drowning of a diversified land surface (rias
shoreline, fiord coast) the tendency to smoothing will be brought
out more strongly. In this case capes will not always be formed of
rocks with extra resistance to erosive action, and the bays need not
correspond with material of lesser resistance.
The platform developed in front of a marine cliff is termed marine
erosion platform or coastal terrace. The broader and shallower it is.
the less effective will be the activity of the waves beating on the
beach. But the coastal terrace is in itself also subject to wear and
tear, because the waves shove and roll the loosened blocks over its
surface. Boring organisms and solution, especially in limestone,
strongly support the mechanical abrasion.
As the incoming waves possess the greatest energy at the outer
edge of the platform and gradually dissipate their force while rolling
inwards, a perfectly horizontal surface would be attacked most
strongly at its seaward margin. Moreover, erosion has been able to
act here for the longest period, because it gradually extends land-
'"lards during the process of planation. From these two causes the
coastal terrace obtains a slight seaward slope, with the greatest depth
where the waves come from. As, however, the oscillatory move-
ment of the water quickly dies out below, as explained in Chapter 1,
the greater depth at the edge provides a certain degree of protection
from wave action. A balance is struck whereby at each distance
from the beach the wear on the bottom is just sufficient to produc
a certain slope. This slope is such that the amount of material
coming from the shallow side added to that produced locally equals
the amount delivered outwards to the neighboring area. Where the
depth for a time becomes too great, a certain amount of detritus is
deposited, while a too shallow portion is subjected to extra attack.
In the case of too great a slope more protecting material is brought
along from the coast, and, conversely, a flat area is open to more
effective attack because less sediment is delivered from the landward
side to protect it.
Hence, we find that, just as in a river, a profile of equilibrium tends
to develop. In neither of these two cases is erosion stopped by attain-
ment of this equilibrium, but from that time onward the shape of
the bottom curve alters but insignificantly. Whereas in a river the
position remains unaltered thereafter but the grade is gradually re-
duced, the profile of equilibrium in coastal erosion is slowly shifted
landwards while approximately the same shape and level are retained.
C 0 A 5 TAL AND B 0 TT 0 MER 0 5 ION 227
A number of factors influence the steepness and shape of the curve.
Local conditions as to climate, type and structure of the country
rocks, the degree of exposure of the coast, current action, and fauna
and flora all play a part in determining the shape of the profile of
equilibrium. In the following chapter the profile of equilibrium will
he dealt with in connection with sedimentation.
The attack on the rocks of the coast produces particles of all sizes
and shapes. The finest chips of the silt and clay fractions are washed
away, the coarsest fragments remaining where they were formed
until sufficiently reduced in size to become transported. The material
of middling sizes is gradually moved either sideways to build up
shingle and sand beaches in embayments or sea wards along the bottom.
The deeper the water to which it is carried, the slighter wave turbu-
lence at the bottom will be and hence the smaller the traction exerted.
This results in sorting, and one generally encounters sand or gravel
below low water line and sand or silt farther out from the coast. We
will return to the sorting action later. On a bold coast under wave
attack, for instance a volcano rising from the sea floor, the edge of
the erosion platform must be reached some distance from the shore,
where the original slope begins. The material transported across the
platform rolls over the edge and builds up a talus or suumarine scree.
By this process the platform is extended gradually seaward by a
wedge of dumped matter.
Normally the submarine slopes are slight and the area of erosion
gradually merges with the area of deposition. Nothing comparable
to a scree then develops, because only sand or silt is available. As a
rule even the areas undergoing marine erosion are largely cloaked
with sediment, at least during most of the year.
It may be asked to what depths mechanical erosion can reach, but
the question is found extremely hard to answer.
Many authors (Bourcart, Shepard) hold that erosive action reaches
only a few meters below sea level, because cobbles with delicate plant
and animal growth attached are sometimes dredged close to the shore.
According to Bagnold the depth to which shingle can be moved is
very smal1-of the same order as the wave height, whereas for fine
silt it is of the order of the wavelength, but this probably does not
apply to waves with a long period that cause only slow osci1latory
movements of the water. Otherwise ocean swell would disrurb silt
at depths of 500 or 600 m, which it almost certainly does not. From
the limitation of fine silt to depths exceeding 60 to 70 m off the New
228 SOU R C E 5 AND T RAN 5 P 0 R TAT ION
England coast, Stetson concluded that this is the limit to which waves
can stir the bottom sediments.
Others are of the opinion that erosion is possible to many dozens
of meters. In support of this contention it may be recalled that sev-
eral observations possibly indicate appreciable wave action down to
100 or 200 m below sea level. Discoloring of the sea water during
storms, the production of wave (?) ripple marks on the sea floor,
the hurling of sand on board ships during storm in comparatively
deep waters, and the absence of fine silt in central portions of the
North Sea all point to wave turbulence reaching depths of many
dozens of meters. One of the most striking instances is that off Land's
End in Cornwall stones up to half a kilogram in weight are sometimes
washed into lobster crawls at depths of over 60 m. In the English
Channel shells are occasionally injured by the movement of gravel at
depths of 70 m, and on the east coast, when ballast foreign to the
region is dumped in 20 to 35 m of water, the shore after storms is
strewn with these pebbles.
Even if waves are able to churn up unconsolidated debris at greater
depths, there can hardly be active abrasion of consolidated rocks
down to depths of 150 m. In any event, to carry erosion of firm
rocks to greater depths than 100 m would require such an extremely
long time that it is highly douutful that the relative level of land and
sea evl!r remained the same long enough to allow this depth to be
attained.
It should not be inferred from the above that erosion is every-
where active or at least that no deposition can take place at lesser
depths than 100 m. Over large areas of the shelf the bottom currents
are extremely weak and, at most, are able to retain clay in suspension.
Where wave turbulence is restricted by offshore winds, or in more
or less enclosed basins, the currents and turbulence are generally un-
able to prevent deposition. Depressions in the terrace, in spite of
being in themselves quite shallow, are generally sites of sedimentation,
even of clay particles. Where currents are absent or move in a closed
course even the most violent wave turbulence cannot stop sedimenta-
tion; it can only cause the deposition to be intermittent. In inland
seas this is the rule. Most fossil sediments, it should be recalled, were
laid down in shallow water. Here the combination of turbulence and
throughgoing currents must have been absent. These aspects will be
taken up again when dealing with sedimentation.
After what has been given above little need be said on the action
of waves on uncemc;_nted rocks as sand, marl, clay, etc., where these
COASTAL AND BOTTOM EROSION 229
between Java and Bali or the other Lesser Sunda Islands (Fig. 110),
between Taliaboe and Mangoli in the Moluccas, between the Orkney
and Shetland Islands, in the Seymour Narrows between Vancouver
Island and the mainland of British Columbia, such high velocities as
6 to 7 m are reported for spring tides. These currents are compar-
able in velocity with those of mountain streams. Although the meas-
25 50
Kilometers
FJC:. !l0. Chart of Strait Lombok and Strait Alas between Bali and Sumbawa,
Lesser Sunda Islands.
urements apply to the surface there can be no doubt that the velocities
along the bottom are also sufficient to cause erosion. Even in these
instances where deeper water is found at both ends of the passage
and no tools in the form of detritus are brought along by the cur-
rents themselves, the walls will be attacked and deliver fragments to
the bottom of the straits, while streams from the adjoining islands
will also contribute material. It may be safely assumed that powerful
erosion of the bottom and walls takes place. The wider the passage
becomes, the weaker the currents will grow. The formation by
erosion of very wide and deep straits between islands will therefore
not be possible uner these conditions.
C 0 A S TAL AND 80 TT 0 MER 0 S ION 231
In the Straits of Dover we have an example that has been studied
in great detail by van Veen (1936). At the surface, velocities of 2 m
are seldom if ever exceeded, while 15 cm above the bottom the maxi-
mum is 60 cm per second. Erosion of the bottom, that lies at about
40-m depth, does not take place, for not only are large areas covered
I I I I
1
FIG. 111. Depth chan of the Marsdiep, a tidal passage between the island of
Texcl and the Dutch mainland, showing deep scour and tidal delta.
south of Holland some tidal waters have scoured out gullies to depths
of 50 and 60 m, cutting through Tertiary deposits. Tertiary fossils
are regularly washed up on the shores of the Island of Walcheren.
When discussing submarine valleys on the shelf (Chapter 7) in-
stances will be given of channels scoured out between islands of the
shallow Sunda Shelf by tidal currents. These attain depths 30 m
below the surrounding sea floor.
On some coasts wind-driven currents of 1y:! to 2 m per second
are reported during storms. When they are combined with a tidal
component, surface velocities of well over 2 m may be expected.
That is equal to the flow of the Rhine in the middle reaches when in
spate. Similar velocities in consequence of wind action are observed
in Yucatan Strait between the Caribbean Sea and the Gulf of Mexico,
and in Florida Strait. Unhappily all these figures relate to surface
velocities. Obviously this calls for extensive observations on bottom
velocities, combined with measurements of sediment in the overlying
waters.
As a general result it is found that erosion of consolidated rock by
currents at sea is possible only in constricted passages of moderate
depth. The eroded material can therefore be of only local signifi-
cance. The erosion of unconsolidated deposits will be dealt with
later on, in connection with transport (p. 260).
Iii conclusion it should be pointed out that not all sedimentary
lime is directly delivered by organisms or precipitation. A certain
amount derives from bottom and coastal erosion and from river
transport. To this lime detritus should be credited only that portion
that had its source in limestone. Deposits of tests that have been
reworked and displaced should be counted as of biological origin.
RIVERS
Apart from marine erosion and the biosphere, rivers foml the chief
source of sedimentary material. Only very few mountain streams
debouch directly into the sea. The overwhelming majority of runoff
water first collects in rivers and reaches the ocean via sluggish lower
courses. Hence nearly all material carried into the sea is fine grained.
Even small pebbles are relatively unimportant in bulk; clay particles
greatly predominate. The Rhine, for example, carries no sand to the
coast, because only clay and fine silt are found in the last few dozen
kilometers of its course, except for sand washed in by tidal currents
from the sandy beach. At Lobith where the Rhine passes the Dutch
RIVERS 233
frontier only a million tons of coarse sand are carried along each year.
This entire amount is dredged out to ensure even flow and to avoid
silting up of the dyked-in river course.
The Mississippi dumps 1 km s of sand and silt into the Gulf of
Mexico each 3Y2 years, while the Nile carries one-tenth of that
amount of sediment to the Mediterranean. It is estimated that all
the rivers on the earth put together bring 12 km a of sediment into
the oceans per year. Spread out in an even layer over the entire
surface of the sea floor this would form a layer 3 cm thick in 1000
years.
Some rivers are exceptionally turbid. Thus the Rio Grande River
in the United States may carry 100/0 by volume of suspended matter.
Less abnormal is the Fraser River of British Columbia. At low water
1 part dry weight per 100,000 parts of water is carried, and during
floods 1 per 4000. The bottom load of the Rhine in it~ middle
course is several times the amount carried in suspension; at the mouth
of the Mississippi, on the other hand, it is but one-tenth. In this
respect also marked variations are thus encountered.
Another highly important function of rivers is the supplying to
the sea of dissolved matter that forms nutrient salts for plankton
(phosphates and nitrates especially) or is used as building materials
for tests and skeletons (lime and silica). The introduction of organic
matter into the sea water by rivers has already been mentioned, and
we must add floating vegetable remains sllch as pollen, seeds, and
tree trunks.
It is oftcn thought that the amount of silica carried to the sea in
solution by rivcrs is but a small fraction as compared to the contrib-
uted lime. This, however, is not the case, for the yearly figures are
respectively 320 X 10 tons of silica and 560 X 10 tons of lime.
But even less silica than lime is found dissolved in sea water (1 in
250,000). As a gradual increase does not take place an amount equal
to that brought in by rivers is evidently deposited on the floor. As
far as is known this deposition takes place in the form of plant .and
animal remains.
An interesting question is whether rivers or coastal erosion pre-
dominate in the delivery of sediment to the sea. In seeking for an
answer one should avoid confusion with local rates of operation.
On some coasts hanging valleys are met with in the cliffs, indicating
that the retreat of the shoreline is so swift that even with their low-
ered base level (or, rather, shortened course) the streams are unable
234 SOU R C E 5 AND T RAN 5 P 0 R TAT ION
WEATHERING
\cathering on the sea floor is much less active than on dry land.
The variations in temperature are much smaller and slower. Only in
the shallow parts of the offshore and on beaches of cold climates can
the important mechanical action of frost playa part. Burrowing,
boring, and mud-feeding activities are carried on by benthonic ani-
mals, but these factors are probably not more intensive than on land.
Chemical weathering is also deprived of some important tools that
are used on land. Most of the highly active acids, as humic acid, etc.,
are absent, and the sea water is slightly alkaline in reaction. The high
concentration of potassium, sodium, calcium, and magnesium ions in
sea water strongly counteractS solution of these elements from rock
particles.
Another circumstance tending to retard weathering is that currents
in submarine groundwater are almost entirely lacking. In finer sedi-
ment, groundwater is to all purposes stagnant and consequently satu-
rated with all soluble materials. In coarser-grained deposits and espe-
cially in shallow water a certain degree of ventilation of the ground-
water takes place.
WEATHERING 237
Another very important cause for the insignificance of submarine
weathering is found in the mineralogical nature of the deposits. In
general the particles have been formed or liberated during the disinte-
gration of rocks on land. Hence, they are either exceptionally resist-
ant, like quartz and muscovite, or they have been newly developed,
for instance clay minerals. Under external conditions they are there-
fore highly stable.
Taking the above into account it is not surprising to find so little
evidence of weathering on the sea floor. For example, Miss Neeb
(1943) showed that even the ash of the Indonesian volcano Tambora,
thrown out a century ago, showed no signs of weathering w hatever,
although similar material on Java is attacked on land in a few years'
time. Glacial pebbles are perfectly fresh when brought up from the
sea floor, though they may show a coating of manganese oxide on the
exposed upper surface. The many rock samples dredged off southern
California included only a few weathered rocks, nearly all of volcanic
material.
The possibility, however, that lava and ash erupted by submarine
volcanoes and reacting with sea water while still at a high temperature
may undergo intensive alteration has been suggested by several writers,
and some "weathered" basaltic and andesitic pebbles and fragments of
pumice have been reported from deep-sea samples. But the total
mass of material involved can hardly be very large.
Dietz (1942) has suggested that possib ly montmorillonite is slowly
converted to illite on the sea floor by the ahsorption of potassium from
sea water. But the alleged scarcity of montmorillonite in marine clays
is not in accord with the results of other workers (Correns, 1939) .
Future investigations will have to solve this important question.
On the beach, weathering takes entirely different forms than under
a permanent cover of water. The effects are genera lly masked, how-
ever, by the powerful abrasive action and swift reworking of the de-
posits by wave and current action. The result of solution may become
evident, especially on limestone rock; this tOpic will be dealt with, in
detail when treating coral reefs. A curious phenomenon occurring
locally is the development of rimmed solution pools in various kinds
of rock with a hardened edge, as the result of deposition of lime dur-
ing evaporation at low tide (Emery, 1946).
Recapitulating, we find that submarine weathering can hardly be
considered an important source of, or influence on, sedimentary
matter.
238 SOU R C E SAN D T RAN S P 0 R TAT ION
Inches
o 1
I I II I
o 234
Centimeters
.,
o
I III I I
o 234 5
Centimeters
Flc. 113. Two typical slump balls composed of shaly sandstone from a slump
sheet of shale. Carboniferous of southwest Wales. (After Kuenen. 1948, Fig. 5.
p. 370.)
1
Feet
FIG. 114. Slump sheet of shale, containing sandy slump balls. Carboniferous of
., southwest Wales. (After Kuenen, 1948, Fig. 7, p. 373.)
deep-sea deposits such as the recent red clay or radiolarian ooze (with
the possible exception of the Mesozoic geosyncline of Timor and sur-
roundings).
The fossil evidence of slumping having been reviewed, considera-
tion can now be given to the action of gravity on recent sediment
accumulating on the sea floor.
It is obvious that it must also be difficult to prove recent sliding by
bottom sampling. Fine lamination suitable for showing contortions in
a columnar sample is highly rare, and the samples are generally short.
Large-scale disturbances cannot show up in tube samples.
The most convincing evidence known to the writer has been offered
by Archanguelski. On the slope leading from the wave-cut shore
terrace to the deep basin of the Black Sea the recent sediments are
in many places absent on the upper part of the slope, leaving older
deposits exposed either at isolated spots or along long continuous
strips. On the lower slopes the slips were evidenced either by the
total pinching out of individual beds or by their intense deformation.
Fine illustrations of contorted lamination from core samples were
given. Slumping is mentioned as having occurred wherever the incli-
nation of the bottom attains 20 to 30 , occasionally even at 1 0.
Indirect evidence of slumping has been obtained from time to time.
Thus patches of coarse detrital material have been dredged by the
TTfwailleuT off the coast of Portugal from depths of 500 to 900 m
though the normal sediment there was fine mud. The littoral origin
was proved by the presence of a shallow-water, soft-shelled fauna .
Again, sliding is the most probable explanation for this anomalous
occurrence, although turbidity currents might also be invoked.
Indirect proof has been given by Bramlette and Bradley. In a series
of long samples taken by Piggot in the northern Atlantic some strata
were discovered which they attribute to slumping. The finer particles
are absent. The lower contact is sharp, and sometimes also the upper.
These strata contain Foraminifera indkating a warmer climate than
that prevailing at the time of deposition. These tests were probably
taken up by the slide from zones at some depth in the sediment higher
up the slope that had been formed during an earlier, warmer period.
The age determinations by Piggot and Urry later brought confirma-
tion, because these layers were deposited at an abnormal rate, pos-
sibly instantaneously. It is noteworthy, however, that some of the
beds are graded from coarsest at the base to finest at the top. This
is not indicative of a slump, and the writers mentioned invoke a
S L U M PIN G AND T U RBI D I T Y CUR R E N T S 247
SETTLING
The velocity with which a particle sinks to the sea floor can be
calculated theoretically or it can be measured experimentally. For
the calculation the law of Stokes is used, which says that for a sphere
the velocity of sinking depends on the specific gravity of the sphere
and its diameter, likewise on the density of the liquid and its viscosity.
Under given conditions v = C X r2, in which v is the velocity, C a
constant, and r ~e radius. As t he specific gravity of sea water is
SETTLING 249
fairly constant and the changes in viscosity with temperature are also
restricted (about 10910 per 5), the velocity for a given kind of
material can be taken as roughly proportional to the square radius of
the particle. Table 13 applies to quartz spheres in distilled water
of 20.
TABLE 13. V E LOCITY or SINKINO OF Q.UARTZ SPHERES IN WATER AT 20 C
Velocity of sinking,
Diameter, mm em / sec Time for sinking 10 em
1 6.6 1. 5 seconds
0 .5 4.4 2.3 seconds
0. 1 0.8 13 seconds
0.03 0 .09 2 minutes
0 .008 0 .005 72 hour
0 .001 0 .00009 1 day
0 .0001 0 .000001 3 months
larger the divergence from sphericity and the rougher the surface,
the greater will be the friction and consequently the slower the rate
of settling. Hence the formula will apply fairly well for quartz
grains, bur for flakes of mica or clay minerals the formula gives much
too large values for the settling velocity. Such disks tend to turn
their flat side at right angles to the direction of movement, that is to
say in a horizontal direction, resulting in a great resistance against
settling. In the same way hollow tests, or shells with marked decora-
tive irregularities and spines such as those of globigerinae or radio-
larians, will sink to the bottom very slowly.
By way of example the measured (Thoulet) and calculated settling
velocities of empty and broken globigerina tests are compared in
Table 14.
TABLE 14. MEA SU RED AND CALCULATED S~T"' LINO VELOCITIES OF OLOBIOERINA TEsn
greatly to influence the rate of settling. In any event the total time
of settling, even in deep water, would vary only slightly on account
of the high velocities.
The mechanical analyses having been made, the time required for
settling in still water of a given depth can be directly calculated fo r
each fraction. Combining this result with a given current velocity
the distance to which particles will be moved during settling over a
certain depth is readily computed.
However, this distance may be increased by the retarding influence
of Ulrbulent motion and by wave action, especially for small particles.
Hence the above computation will give only the minimum distance
of transport. In the following paragraphs these aspects of transport
are further elaborated.
HORIZONT AL CURRENTS
Currents may transport particles either completely in suspension,
or rolling along the bottom, or touching the bottom now and again
(saltation). In the first case the velocity is equal to the average cur-
rent velocity. In Chapter 1 a number of current velocities in nature
were given, and in discussing erosion on p. 229 several instances of
larger velocities are also mentioned. In open and deep water the
currents at the surface seldom exceed 1 m per sec. Exceptionally
swift currents such as the Gulf Stream may locally attain 3 m per sec.
In the deep sea, below 200 m, current velocities are generally not
accurately known. They appear to be of the order of 5 to 20 cm
per sec. The direction of flow in the various layers of the ocean is
not the same. Hence, the average velocity between the upper limits
of the deep sea and the bottom must fall far short of the velocity of
each separate stratum and cannot exceed a few centimeters per second.
Coarser grains can, therefore, not be moved more than 1 km and fine
sand not more than a dozen kilometers from the point where it left
the surface layers. For planktonic tests, these distances are greatly
increased by the buoyancy of the organic matter and the fine projec-
tions. These have generally disappeared by decomposition and abra-
sion before the mechanical analyses is made. The divergences from
the massive spherical shape are discounted by the method of analysis
(equivalent diameter).
In shallow water on the shelf the rate of flow is frequently greater
than in deep water. Thus the tidal currents in the North Sea may
exceed 1 m per sec, and in restricted channels values of half a dozen
meters per second have been observed. The currents sweeping over
252 SOU R C E 5 AND T RAN 5 P 0 R TAT ION
the Agulhas Bank of 100-m depth attain 2 m per sec. Under favor-
able circumstances wind-driven currents may flow at 1 m per sec,
as in the entrances to the Zuider Zee in Holland. All these figures
apply only to the surface velocities. Figure 117, p. 255, gives an
approximate idea of the decrease downwards where a current sweeps
over a shallow bottom. In the Golden Gate the velocity at 1 m
above the bottom was found to be half the surface velocity. Revelle,
Fleming, and Shepard carried out measurements with three super-
imposed current meters suspended from a tripod. In a tideway the
following data were obtained:
126 em above the bottom 26.3 em/ sec
51 em above the bottom 21.3 em/sec
21 em above the bottom 15.6 em/sec
Settling
Sediment Diameter velocity, Displacement
Time
in mrn em / sec
for
settling
I mISe< 10 em / sec
Sand { ~~:r8e 2
0. 1
25
0.8
7 minutes
J hours
0. 5 km
10
42 m
I km
Silt 0.06 0 . 35 8 hours 30 J
CIa. { coarse 0. 005 0. 002 2 months 5.000 500
) fine 0.0005 0.00002 14 years 500.000 50.000
"fern (pIan k tOnic)
f ora.mlOl . earger 0. 5 6 J.i hour 1J.i 170
II
sma er 0. 02 0.05 J day. 250 25
Radiolarians 0. 5 0.4 7 hour. 25 2.5
Diatoms 0.04 0.15 I day 80 8
But if the particles are heavier than the water they show a sinking
movement superimposed on the turbulence. This must cause a de-
crease in the density of the suspension towards the top. Hence
every turbulent mass of water moving upward will carry a greater
number of particles than the equivalent sinking mass of water. It
follows that no sooner does the concentration of the suspension
increase downward than a net upward transport of particles by the
turbulence 'will commence. The stronger the variation in concentra-
tion, the more effective will be this upward transportation. Finally
a stationary condition is reached in which the sinking in consequence
of gravity is exactly compensated by the upward transport by turbu-
lence. The upward velocity of the churning movement must be
greater, however, than the setding velocity of the particles.
This condition then remains stationary and can also be attained
when the sediment does not descend but is churned up from the bot-
tom. In the latter case the upward transport by turbulence is origi-
nally stronger than the influence of settling. In consequence of the
strong concentration at the bottom more and more sediment is caught
up and carried upward.
Because, in nature, the more frequent sedimentary particles are
roughly of the same density, only the (equivalent) size plays a part.
254 SOU R C E SAN 0 T RAN S P 0 R TAT ION
The minute particles of which clay (lutite) is composed show exceed-
ingly slow rates of settling. Hence the effect of slight turbulence is
sufficient to counteract the motion due to gravity. The sediment is
then distributed almost evenly right up to the surface. With sand
the settling velocity approaches
500 (if it does not exceed) that of the
vSil~__
ay vertical movements of the water.
Then a very strong upward de-
loam
crease in concentration is required
400 to attain a stationary condition.
Very fine
sand"-l
As the highest concentration is
limited, the layer of water in
E
<)
which sufficient decrease in con-
centration is possible must be very
trun. Even at a small distance
~
above the bottom a concentration
..
.8
:E2oo
of zero is reached and the higher
strata will receive no particles.
.,
.!!!I
If the concentration of the sedi-
:r
ment at 2 cm above the bottom
is considered 1, the concentration
100 at every level above the bottom
., _Coarse
~ sand \ for a given intensity of turbulence
and a given equivalent diameter
\
" - of the particles can be established
o '-.. theoretically. Our Fig. 116 thus
o 0.2 0.4 0.6 O.B 1.0 1.2 gives the curves for four grain
Silt contents
FIG. 116. Distribution of sediment (silt sizes up to 5 m from the bottom
content) in stream or current with re- with moderate turbulence. In a
spect to height above stream bed, as like manner the curves for vari-
computed from Schmidt's formula. ous states of turbulence can be
(After Hjulstriim, in Trask, 1939, Re-
drawn.
cent Marine Sediments, Fig. 2, p. 20,
American Association of Petroleum In the present treatment, how-
Geologists.) ever, one very important element
has so far been left out of account,
namely, the influence of the bottom on turbulence. It has been tacitly
assumed that the vertical component of turbulence remains equal
down to the very bottom. But this cannot be so, as the solid bottom
must hinder the upward and downward movements. Possibly this
will tend to increase the horizontal component, but, however this
may be, the vertical movements must decrease. The smoother the
HORIZONTAL CURRENTS 255
bottom, the more marked must be the decrease in vertical turbulence
along it. In a very thin layer directly in contact with the bottom
the flow must even approach a laminar movement (Fig. 117).
The scarce experimental data indicate that the bottom is in general
rough in a hydrodynamic sense, but with a smooth deposit of fine
Surface of water
Turbulent
1000
500
~I
300
200 ~ Erosion A ..
~~
till
100
50 =====
--
u
~ 30
1) 20
.~r7
~ 10
Qj
>
5
.,c'" 3
Tran sportation
:E
2 " I". -i'"
/
0.5
0,3 ~-
0.2
""r
Size of particles, mm
FIG. 118. Relationship between average current velocity in a river and sediments
of uniform texture showing velocities necessary for erosion, transportation, or
deposition. (After Hjulstrom, in Trask, 1939, Recent Marine Sediments, Fig. 1,
p. 10, American Association of Petroleum Geologists.)
~ 2 - .::
'E
I
'r V
.!l 1
E 0.8
~ 0.6
<:
'jij 04
l5 0:3 I
0.2
/" /
Y I
0.1
N ""..,. ID <08 ~ g Sj! fa ~8
Velocity, centimeters per second ....
FIG. 119.Suspension and traction curves, showing the relation of grain . size to
the velocities of currents that support debris in suspension (vertical currents)
and that carry debris along the bed by traction (horizontal currents). Curve A
is based on data from e1utriation tubes with turbulence; Curve B, on flume ex-
periments and traction-tube experiments. (Simplified after Nevin, 1946, Plate I,
p. 674.)
of the surface speed (Nevin gives 50ro). Where sand exists in quan-
tity all currents up to 75 cm per sec, according to the same investi-
gator, are ineffectual in moving shingle, whereas above that velocity
the current suddenly acquires the power of moving stones up to
nearly 7 ~ cm in diameter over a sandy bottom.
It is further believed that sand is set in motion at smaller velocities
in clear water than in muddy suspensions. This is explained by silt
and clay becoming wedged ben~een sand grains of the surface layer.
Thereby the surface becomes hydrodynamically more smooth and
approaches the state of a stratum of clay. On the other hand the
presence of pebbles and boulders facilitates the erosion of sand because
of the increased bottom turbulence.
The figures quoted show that there is still considerable divergence
between the results of different workers, and that data obtained under
natural conditions at sea are very scarce indeed.
Rocky
Sand
Clay and silt
~ Sand and clay
.:::::::: Sand and gravel
j! Shells and sand
Sand, shells, and clay
Sand, grovel, ond clay
:::::::; Gravel
PACIFIC COAST
OF THE
UNITED STATES
FIG. 120. Reconnaissance map of continental shelf sediments along Pacific Coast
of United States. (After F. P. Shepard, in Trask, 1939, Recent Marine Sediments,
Fig. 2, p. 224, American Association of Petroleum Geologists.)
than those formed during offshore winds. Not only is the wind
more powerful but the length of fetch is also greater. Obviously,
therefore, the transporting power of waves must be stronger towards
the coast than away from land when other factors are left out of
account.
Where a wave runs into shallower water it becomes asymmetrical.
The movement of the water along the bottom towards the coast is
then swifter than the returning flow and, although of shorter duration,
must exert a stronger action.
-----+---co.,,-{
i
eL.I"
FIG. pI. Tenninology applied to various parts of the beach profile. Benns are
small 1mpermanent terraces which are formed by deposition during calm weather
and by erosion during storms. (A fter Shepard, 1948, Submarine Geology, Fig.
H, p . 82, Harper and Brothers.)
But there are also transporting agents active in the opposite direc-
tion. Where the bottom slopes, gravity must exert a transporting
force downward, that is to say, in the majority of cases, away from
the beach. Sediment raised by turbulence also comes under the
influence of gravity, as explained above (p. 240). These gravitational
effects may compensate the forward impulse of the oscillatory waves
or even cause a net movement away from land. With an onshore
wind, light materials, especially floating ones, are then thrown upon
the beach and left high and dry when the tide falls (jetsam). With
an offshore wind, friction carries more water away from the coast
than is piled up against it by the weak oscillatory waves then running
inwards. A hydraulic undertow then sucks the water up along the
bottom back towards the beach. Hence during offshore winds more
and different species of shells are carried up the beach, and collectors
prefer these conditions for making their beach excursions.
TRANSPORTATION BY WAVES 269
It is generally held that the water carried landward by waves re-
turns along the bottom by what is termed the undertow. Although
the velocity cannot be great, all sediment held in suspension should
have a component away from the beach added to whatever other
forces arc working on it. Although the existence of an undertow is
assumed by most investigators, some doubt has recently arisen. A
different type of seaward transport of the excess water piled against
the shore by wind and waves has been detected.
FIG. 122. Diagrammatic sketch of a rip current showing its component parts
and direction of net water transport associated with the rip. Relati ve velocity
indicated by length of arrows. (After Shepard, 1948, Submarine Geology, Fig.
20, p. 45, Harper and Brothers.)
Flc. 124. Diagram showing the infl.uence of sandbanks on surf and resulting
currents.
The slope of the beach varies with the grain size of the material.
Shepard (1948) gives the figures shown in Table 18.
TABLE 18
Beach type Median diameter, mm Slope
Fine sand 0. 12 2
Medium sand 0.25 4
Coarse sand 0. 5 8
Gravei 2 12
Cobble 64 20
F,G. 126. Block diagram of beach drifting, showing path of large (dashed lines)
and small (full lines) particles.
RAFTING
The agents treated above may be looked upon as "normal" factors
governing the distribution of sedimentary material over the sea floor.
To these might be added swimming, crawling, floating, etc., of ani-
mals and plants whereby sedimentary source material is displaced be-
fore incorporation in the deposits. The problems involved in these
migrations are mainly of marine biological nature. To the marine
geologist they are of the greatest importance, because plant and ani-
mal detritus play so prominent a part in the formation of sediments.
However, in order to limit the extent of this book this vast field will
not be treated separately and the conception of "source" will be
ABRASION DURING TRANSPORTATION 277
limited to the moment when the animal or plant has died, leaving
treatment of earlier migrations to other writers. Displacements after
death are by the "normal" agents dealt with above.
There are other agents playing a part in the displacement of sedi-
mentary material that are of considerable interest, although seldom
becoming quantitatively important. They may be conveniently
classed under the heading of rafting.
Rafting can be brought about by floating ice (moraine in icebergs
or sediment carried by ice floes) or by trees with the materials lodged
among the roots. Particles from fine dust to large boulders may thus
be carried for long distances and dropped into sediment of entirely
different nature. Erratic boulders have frequently been dredged in
polar deep-sea deposits and even beyond the present limit of icebergs,
apparently having been carried there by bergs during the Ice Age
(see p. 235).
Living animals and plants can also act as "rafts" in this sense.
Emery, who has made a special study of this matter, has showed that
marine mammals frequently carry pebbles and cobbles in their stom-
achs, sometimes in considerable numbers. The action of birds and
other animals in collecting shells they had fed on, may be recalled
here, although these accumulations are normally formed on dry land
(Teichert and Serventy, 1947).
Kelp is a far more frequent agent in rafting stones and shells. Large
waves may wrench these plants from their hold on the bottom and
carry them to the beach with their "anchor" still attached. Large
kelps with floats may even carry stones across deep water. Smaller
algae and those without floats cannot lift stones, but they provide a
hold for waves and currents whereby coastal drifting is greatly fur-
thered, cases having been noted of displacements of 250 km from the
point of origin. A few examples are also given by Emery of pebbles
with kelp holdfasts still attached that were dredged from depths of
nearly 250 m.
From these studies it may be concluded that individual pieces of
stone found in fine-grained sediments, or among beach conglomerates
where they do not belong, may have been carried to this contrasted
environment by animals or plants.
axis. A flat boulder tends to remain lying on its larger side and to
shove about or rotate around its shorter axis. Upper and lower sur-
face are thereby worn, and a rotational ellipsoid around the shorter
axis presents the ideal shape. But when the size has been reduced to
a few centimeters the breakers can lift the cobble and roll it about on
the narrow edge. Then the flat shape is soon reduced and egg shapes
are evolved. These revolve around the long axis and do not tend to
revert to disks. On the other hand, when the original fragment is
oblong in shape and not too large either a triaxial ellipsoid or a rota-
tion ellipsoid around the longer axis tends to develop.
When an isolated boulder lies for a long time in a fixed position
on a sandy beach the sand blast of wave action may wear the side
facing the sea until a sharp ridge parallel to the shore is developed
along the exposed surface of the block. Kuenen (1947) has termed
such boulders aquafacts.
2 1 ~ ~
FIG. 128. Two methods of representing grain size: histograms and cumulative
curves. Note the more perfect sorting of the beach gravel from North Devon
as compared to the river gravel from Glencoe. (Redrawn from Hatch, Rastall,
and Black, 1938, Figs. 2, 3, 4, p. 44.)
RHYTHMIC ACCUMULATIONS
Several types of rhythmic accumulations of sediment are found on
the beach and the sea floor. They can be classified as ripple marks,
larger rhythmic accumulations, and beach cusps.
Many of the ripples exposed at low tides are due to the process de-
scribed, or to the longshore flow Out of beach troughs.
For lack of a better name they may be called surf ripples. Espe-
cially favorable localities for the development of surf ripples are the
shallow oblong depressions behind sand banks which are submerged
at high tide but fall dry at low tide.
A different type of ripple, 'which has been described by Timmer-
mans, is generated by backwash above the level of maximum wave
retreat. The thin film of the backwash sometimes stands up in ridges
of water a few centimeters high, parallel to the beach, and when the
water has retreated broad, flat ripples arc seen to have been formed
with narrow and shallow troughs in between, often some 30 em apart.
Shepard, who has also observed this phenomenon, suggests usc of the
term backwash ripples (personal communication).
A number of more intricate ripple patterns are also met with in
nature. They may be conveniently classed as "compound ripples."
In the first place an oscillation ripple mark may be altered by later
waves running in a different direction. Second, currents and waves
running in different directions may cause a combined pattern. Third,
current ripples may be superimposed on older current ripples formed
by a flow in a different direction. And fourth, both current and
oscillation ripples may be deformed by alteration in the strength of
the ge1\Crating agent. As the system of ripples is seldom preserved
when the force increases, the patterns we find are generally due to
decreasing force. One form is that of major oscillation ripples with
smaller crests in the troughs, formed by decreasing waves. Further
complications are introduced when ripples are laid dry during ebb
tide. Sharp crests may sag as they emerge above water or capillary
waves nibble at their flanks; water may trickle along the troughs and
produce linguoid ripples; blown sand may be checked on the wet
surface, forming a thin crust altering the profile and characteristic
sorting.
According to van Straaten (personal communication) longitudinal
ripples of great regularity are formed by currents in sand or mud under
special conditions, sometimes combined with transverse current ripples.
The following classification of water ripple marks may be offered:
I. Ripples formed by currents in one direction.
(a) Current ripples, formed between first and second critical veloci-
ties. Asymmetrical.
(1) Normal current ripples, tidal fiats, stream beds, marine cur-
rents, turbidity currents ( ?); (2) surf ripples; (3) linguoid
RHYTHMIC ACC U MU LA TIO N S 293
ripples; (4) rhomboid ripples; (5) longitudinal ripples; (6)
backwash ripples.
(b) Sand waves, formed above third critical velocity.
(1) Regressive; (2) progressive (much sediment in suspension).
(c) Meta and para ripples. Formed from sand waves by falling veloci-
ties.
II. Ripples formed by oscillation currents.
(a) Symmetrical oscillation ripples.
(1) Angular crests; (2 ) angular uoughs (rare).
(b) Asymmetrical oscillation ripples.
(I) On shelving bottom (probably due to one of the following
causes) .
(2 ) By asymmetrical oscillation currents.
(3 ) Combination with undertow or other currents.
III. Compound ripple marks.
(a) Reticular compound ripples.
(1) Two directions of oscillation form cd one after the other;
(2) two directions of current following each other; (3) oscil-
lation togcther with current; (4) oscillation and current fol-
lowing the one on the other; (5) longitudinal and uansverse
current ripples formcd simultaneously.
(b) Parallel compound ripples.
( I) By falling velocity of current; (2) by decrease in wave size.
and either merge gradually into the beach or are set off sharply
therefrom. The height may be very slight but may also attain more
than a meter, and the distance between is a few centimeters to sev-
eral dozens of meters. The apex may protrude several meters, and
the relative depth of the bays is also variable. In front of the bays
the foreshore is sometimes built out under water in a delta shape.
Although cusps may develop on many types of beach, they are
most frequent on slightly concave stretches of the coast. Observa-
FIG. 130. Block diagram of beach cusps, showing the horns (in this case con-
sisting of coarse material) a.nd the submarine deltas corresponding to the embay-
ments. (Mainly according to Timmerillans, after Kuencn, 1948 (a), J. Geo/., Vol.
56, Fig. I, p. 35 .)
't,tfl';,.
FIG. 131. Variations in the shapcs of beach cusps. (Aftcr Johnson, 1919, Shore
Processes and Sboreline Development, Fig. 141, p. 465, John Wiley & Sons.)
to carry the sediment along, the enlargement will continue both hori-
zontally and vertically. As the channel becomes deeper and its side-
ways slopes steeper the breadth also tends to increase. But when the
depth at the outer side approaches a certain limit, erosion gradually
slackens because the currents set up by the waves become too weak
to shift the sediment. Refraction of the swash as it fans out in the
embayments results in transport of material towards the sides and
prograding of the cusps (Fig. 132). The coarse material tends to
be pushed back up the beach of the embayment and Out along the
developing cusps. Growth of the bays and prograding of the cusps
must gradually decrease when the maximum depth in the central area
of the bay has been attained.
In the meantime adjacent channels have undergone the same process.
Where these are so close together that the two natural spheres of
growth overlap, a rivalry develops. This should tend to push the
296 SOURCES AND TRANSPORTATION
F'G. 132. Diagram showing the refraction of the swash. Pebbles arc rolled from
B to C. At A the powerful backwash of the foregoing wave impedes the swash
and increases the refraction. a-f, slope; /1' -f', direction of backwash due to the
piling up of water over the area DEF. (After Kncncn, 1948(a), 1. Geol., Vol.
56, Fig. 2, p. 37.) .
0,
Flc. 133. Successive stages in the development of cusps from irregular indentures
of the beach to an almost regular, rhythmic pattern. (After K uenen, 1948(/1),
J. Geol., Vol. 56, Fig. 3, p. 39.)
RHYTHMIC ACCUMULATIONS 297
bays farther apart, because the material eroded from the one and
dumped on the intervening cusp must tend to encumber the growth
of the other. The erusion in the latter will be shifted slightly to the
opposite side and thus cause the entire bay to move away (Fig. 133).
Where two neighboring bays are farther apart their adjacent cusps
will not coalesce and will leave a slight space in between. In this
space diffraction begins and a new bay will start to form.
As long as the maximum depth of "vater in a bay has not been
attained the tendency to enlarge will be greater than in an adjacent,
FIG.
-_-_- - _-Shoreline
-- -
- - -
134. Partially eroded older cusps and rcspaced larer series. (Afrcr Johnson,
1919, Sbore Processes and Shoreline Development, Fig. 142, p. 466, John Wiley
& Sons.)
shallower indenture. The larger one wjJJ grow at the cost of the
smaller one. This relation is reversed when the depth approaches its
maximum value. Then a smaller bay is more powerfully eroded and
tends to encroach on its overgrown neighbors. In this manner a
balance will be struck when all bays have attained the maximum depth
in relation to the size of waves playing on them. They will push
each other aside or obliterate each other until this stage is attained
and a regular rhythmic pattern has been evolved.
This working hypothesis appears to explain the main features of
cuspate beaches: the regularity of the pattern, the dependence of size
on the size of the waves, and the accumulation of coarse material in
the horns. But several problems remain to be solved. It is not clear
why cusps are a comparatively rare phenomenon. On the shore of
lakes their formation appears to depend on the breaching of a ridge
(Evans). The relations between slope of the beach, height of the
tide, size of waves, grain size of the beach material, and forms of the
cusps are unknown, and there is an almost complete lack of quantita-
tive data.
298 SOU RCE S A ND T RAN S PORTAT I ON
Bibliography
ANDREE, K. Geologie des Meeresbodens, Bd. 2, 689 pp., Borntrager, Leipzig, 1920.
Das Meer und seine geologische Tatigkeit, in W. SALOMON, Grundzuge der
Geologie, Bd. I, pp. 361-;33, Schwelzerbart. Stuttgart. 1925.
ARCHANGUELSKY, A. D. Slides of Sediments on the Black Sea Bottom and the
Importance of this Phenomenon for Geology, Bull. Soc. NatuT. Moscow,
Sect. Geol., T. 8, Nov. Ser. T. 38, pp. 32- 80, 1930.
AUSTIN. R. A. C. On the Valley of the English Channel, Quart. J. Geol. Soc.
London, Vol. 6, pp. 69- 97, 1850.
BAGNOLD, R. A. Lecture to the Geol. Soc. London, April 16, 1947.
BAILEY, E. B. New Light on Sedimentation and Tectonics, Geol. Mag., Vol. 67,
pp. 77- 92, 1930.
Sedimentation in Relation to Tectonics, Bull. Geol. Soc. Am., Vol. 47, pp.
1713-1726. 1936.
American Gleanings, 1936, Trans. Geol. Soc. Glasgow, Vol. 20, pp. 1- 16. 1938.
BAILEY, E . B., L. W. CoLLET, and R . M. FIELD. Paleozoic Submarine Landslips
ncar Quebec City, J. Geol., Vol. 36, pp. 577-614, 1928.
BALDREY, R. A. Slip-Planes and Breccia Zones in the Tertiary Rocks of Peru.
Quart. J. Geol. Soc. London, Vol. 94, pp. 347- 358, 1938.
BARTH, T. F. W., C. W . CORRENS, and P. ESKOLA. Die Emstelnmg der Gesteine,
422 pp., Springer, Berlin, 1939.
BEETS, C. Miocene Submarine Disturbances of Strata in Northern Italy, f. Geol.,
Vol. 54, pp. 229- 245, 1946.
BOURCART, J. Sediments quaternaires conserves sur la greve de la region de
Roscoif (Finistcre) , Compt. rend. acado sci., Vol. 221, pp. 357-359, 1945.
BOURCART, J., et }. JACQUET. Sur la repartition des sediments dans la baie du Mont
Saint-Michel, Compt. rend. acado sci. Paris, pp. 1507- 1508, 1946.
BRAMLETTE, M. N ., and W. H. BRADLEY. Lithology and Geological Interpreta-
tions, 34 pp., in Geology and Biology of North Atlantic Deep-Sea Cores.
U. S. Geol. SUTV. Prof. Paper 196, 1940- 1942.
BRONCERSMA-SANDERS, M. The Importance of Upwelling Water to Vertebrate
Paleontology and Oil Geology, Verh. Kon. Ned. Akad. van Wetensch .
Amsterdam, Afd. Nat., Sec. 2, 01. 45, 4, 112 pp., 1948.
BROWN, C. B. On a Theory of Gravitational Sliding Applied to the Tertiary of
Ancon. Ecuador, Quart. J. Geol. Soc . London, Vol. 94, pp. 359- 370, 1938.
CAIL.LEUX, A. Les actions eoJiennes periglaciaires en EUrope, 176 pp., Soc. Geol.
Paris, 1942.
CoRRENS. C. W. Die Sedimente des aquatorialen Atlantischen Ozeans, Meteor,
Bd. 3, T. 3, Lf. 2, 298 pp., 1937.
See BARTH et aI., 1939.
Leitminerale im Golf von Guinea, Nachr. Akad. Wiss. Giittingen, Math.-Phys.
Kl., pp. 1-3, 1947.
CoRRENS, C. W., et aI. Tiefseebuch, 144 pp., Mittler, Berlin, 1934.
DEMAREST, D. F . Rhomboid Ripple Marks and Their Relationship to Beach
Slope, f. Sed. PetT., Vol. 17, pp. 18-22, 1947.
DlETZ, R. S. Clay Minerals in Recent Marine Sediments, Am. Min., Vol. 27, pp.
219-220, 1942 .
BIBLIOGRAPHY 299
DIETZ, R. S., K. O. EMERY, and F. P. SHEP.~RD. Phosphorite Deposits on the Sea
Floor off Southern California, Bull. Geol. Soc. Am., Vol. 53, pp. 815 - 848, 1942.
DIXON, E. E. L. The Ouachita Basin of Oklahoma vis-a.-vis the Craven Lowlands
of Yorkshire, Geol. Mag ., Vol. 68, pp. 337-344, 1931.
EMERY, K. O. Transportation of Rock Particles by Sea Mammals, J. Sed. Petr.,
Vol. 11, pp. 92-93, 1941.
Transportation of Rock by Kelp, Bull. Geol. Soc. Am., Vol. 52, pp . 855- 862,
1941.
Transportation of Marine Beach Sand by Flotation, J. Sed. Petr., Vol. 15, pp.
84-87, 1945.
Marine Solution Basins, J. Geol., Vol. 54, pp. 209-228, 1946.
EMERY, K. 0., and F. P. SHEPARD. Lithology of the Sea Floor off Southern Cali-
fornia, Bull. Geol. Soc. Am., Vol. 56, pp. 431-478, 1945.
EVANS, O . F. Classification and Origin of Beach Cusps, J. Geol., Vol. 46, pp.
615-627, 1938.
The Classification of Wave-Formed Ripple Marks, ]. Sed. Petr., Vol. 11, pp.
37-41, 1941.
Effect of Change of Wave Size on the Size and Shape of Ripple Marks, J. Sed.
Petr., Vol. 13, pp. 35-39, 1943.
Further Observations on the Origin of Beach Cusps, J. Sed. Petr., Vol. 53,
pp. 403-404, 1945.
Ripple Marks as an Aid in Determining Depositional Environment and Rock
Sequence, ]. Sed. Petr., Vol. 19, pp. 82-86, 1949.
EVANS, O. F., and R. L. INCRAM. An Experimental Study of the Influence of
Grain Size on the Size of Oscillation Ripple Marks, J. Sed. Petr., Vol. 13,
pp. 117- 120, 1943 .
FAIRBRIDCE, R. 'V. Submarine Slumping and Location of Oil Bodies, Bull. Am.
Ass. Petro Geol., Vol. 3D, pro 84-92, 1946.
Coarse Sediments on the Edge of the Continental Shelf, Am. ]. Sci., Vol. 245,
pp. 146-153, 1947.
Possible Causes of Intraformational Disturbances in the Carboniferous Varve
Rocks of Australia, J. Proc. Royal Soc. New Soutb Wales, Vol. 81, pp. 99-121,
1947 .
GLANGEAUD, L. Etudes quantitatives et experimentales sur I'erosion et Ie transport
par les eaux courantes, Rev. Gliogr. phys. et geol. dyn., Vol. II, pp. 323- 370,
1938.
GRANT, U. S. Waves as a Sand-Transporting Agent, Am. J. Sci., Vol. 241, pp.
117-123, 1943.
GRUNAU, H. Geologie von Arosa mit besonderer Beruksichtigung des Radiolarit
Problems, Diss. Bern, 108 pp., 1947.
HANTZSCHEL, W. Bau und Bildung von Grosz-Rippeln im Wattenmeer, Seneken-
bergiana, Bd. 20, pp. 1-42, 1938.
HATCH, F. H., R . H. RASTALL, and M. BLACI{. Tbe Petrology of the Sedimentary
Rocks, 383 pp., Allen and Unwin, London, 1938.
HElM, ARN. Ueber submarine Denudation und chemische Sedimente, Geol. Rund-
schau, Bd. 15, pp. 1-47, 1924.
JONES, O. T . On the Sliding or Slumping of Submarine Sediments in Denbigh-
shire, North Wales, during the Ludlow Period, Quart. J. GeoJ. Soc. London,
Vol. 93, pp. 241-283, 1937.
300 SOU R C E 5 AND T RAN 5 P 0 R TAT I ON
The Geology of the Colwyn Bay District: A Study of Submarine Slumping
during the Salopian Period, Quart. f. Geol. Soc. London, Vol. 95, pp. 335-382,
1940.
The Geology of the Silurian Rocks West and South of the Carneddau Range,
Radnorshore, Quart. ,. Geol. Soc. London, Vol. 103, pp. 1-36, 1947.
JOHNSON, D. W. Shore Processes and Shoreline Development, 584 pp., Wiley,
New York, 1919.
The New England-Acadian Shoreline, 608 pp., Wiley, New York, 1925.
KINDLE, E. M. Notes on Shallow-Water Sand Structures, J. Geol., Vol. +4, pp.
861-869, 1936.
KRUMBEIN, W. C. Measurement and Geological Significance of Shape and
Roundness of Sedimentary Particles, ,. Sed. Petr., pp. 64-72, 1941.
KRUMBEIN, W. C., and F. J. PEmJoHN. Manual of Sedimentary Petrography,
549 pp., Appleton-Century, New York, 1938.
KUENEN, PH. H . Geological Interpretation of the Bathymetrical Results, The
Snellius Expedition, Vol. 5, part I, 123 pp., Brill, Leyden, 1935.
Water-Faceted Boulders, Am. f. Sci., Vol. 245, pp. 779-783, 1947.
(a) The Formation of Beach Cusps, , . Geol., Vol. 56, pp. 34-40, 1948.
(b) Slumping in the Carboniferous of Pembrokeshire, Quart. J. Geol. So c.
London, Vol. 104, pp. 365-385, 1948.
(c) Turbidity Currents of High Density, Eighteenth 5ess. Intern . Geol. Congr.,
1948.
(d) Troebelingsstromingen van haag soorcelijk Gewicht, Versl. K. Nederl.
Akad. Wet . Amsterdam.
(e) De Kringioop van het Water, 408 pp., Leopold, The Hague, 1948.
KUENEN, PH. H., and G. A. NEED. Bottom Samples, The Snelli71s Expedition,
Yo0I. 5, part 3, 265 pp., Brill, Leyden, 1943.
KUENEN, PH. H., and I. M. v. D. VLERK. Geheimschrift der Aardc, 370 pp., de
Haan, Utrecht, 1948.
MARR, }. E. Deposition of tbe Sedimentary Rocks, 245 pp., Cambridge University
Press, 1929.
MIGLIORINI, C. 'Sui modo di formazione dei complessi tipo Macigno, Boll. soc.
geol. italiano, Vol. 62, pp. 48-49, 19+4.
MILLER, W. J. Intraformational Corrugated Rocks, J. Geol., Vol. 30, pp. 587-{)10,
1922.
MILNER, H. B. Sedimentary Petrography, 3d ed., 666 pp., Murb)" London, 1940.
MooRE, R. C. The Origin and Age of the Boulder-Bearing Johns Valley Shale in
the Ouachita Mountains of Arkansas and Oklahoma, Am. J. Sci., Vol. 227,
pp. 432-453, 1934.
MUR.II.AY, J., and }. HJORT. The Deptbs of the Ocean, 821 pp., Macmillan, London,
1912.
N.EED, G. A. The Composition and Distribution of the Samples. See Kuenen and
Neeb, 1943.
NEVIN, C. Competency of Moving Water to Transport Debris, Bull. Geol. Soc.
Am., Vol. 57, pp. 651-{)74, 1946.
PEmJoHN, F. ]. Sedimentary Rocks, 526 pp., Harper, New York, 1949.
PEmJOHN, F. J., and J. G. LUNDAHL. Shape and Roundness of Lake Erie Beach
Sands, J. Sed. Petr., Vol. 13, pp. 69-81, 1943.
BIBLIOGRAPHY 301
PR.ATJE, O. Die Sedimente des siidatlantischen Ozeans, Meteor, Bd. 3, T. 2, Lf. I,
56 pp., 1935.
RETTGER, R. E. Experiments on Soft-Rock Deformation, Bull. Am. Assoc. Petro
Geol., Vol. 19, pp. 271-292, 1935.
RICH, J. L. Submarine Sedimentary Features on Bahama Banks and Their Bearing
on Distribution Patterns of Lenticular Oil Sands, Bull. Am. Assoc. Petro Geol.,
Vol. 32, pp. 767-779, 1948.
RUBEY, W. W. T he Force Required to Move Panicles on a Stream Bed, U. S.
Geol. Surv. Prof. Paper 189-E, pp. 120-141, 19311.
SCHOTT, G . Geograpbie des Atlantischen Ouans, 438 pp., C. Boysen, Hamburg,
1942.
ScHOTT, W. Die Foraminifercn in dem aequacorialen Teil des Atlantischen
Ozeans, Meteor, Bd. 3, T. 3, Lf. 13, pp. 43- 134, 1935.
SHEPARD, F. P. Non-Depositional Physiographic Environments off the California
Coast, Bull. Geol. Soc. Am., Vol. 52, pp. 1869- 1886, 1941.
Submarine Geology, 348 pp., Harper, New York, 1948.
SHEPARD, F. P., K. O. EMERY, and E . C. LAFOND. Rip Currents: A Process of
Geological Importance, J. Geol., Vol. 49, pp. 337- 369, 1941.
SHEPARD, F. P., and U . S. GRANT. Wave Erosion along the Southern California
Coast, Bull. Geol. Soc. Am., Vol. 58, pp. 919-926, 1947.
STEERS, J. A. Tbe Coastline of England and Wales, 644 pp., Cambridge University
Press, 1946.
STETSON, H. C. Geology and Paleontology of the Georges Bank Canyons, Bull.
Geol. Soc. Am., Vo)' 47, pp. 339-440, 1936.
STETSON, H. C., and J. F. SMITH. Behavior of Suspension Currents and Mud
Slides on the Continental Slope, Am. J. Sci., Vol. 35, pp. 1- 13, 1938.
SVERDRUP, H . U., M. W. JOHNSON, and R. H . FLEMING. The Oceans, Their
Physics, Chemistry, and General Biology, 1060 pp., Prentice-Hall, New York,
1942.
TElcHER'r, C., and D. L. SERVENTY. Deposits of Shells Transponed by Birds, Am.
J. Sci., Vol. 245, pp. 322-328, 1947.
TIMMERMANS, P. D. Proeven over de Invloed van Galven op een Strand (English
summary), Leidsche Geol. Med., dl. 6, pp. 231- 386, 1935.
TRASK, P . D. Sedimentation, chapter in Geology 1888-1938, 50th Anniv. Vol.,
Geol. Soc. Am., pp. 223-239, 1941.
TRASK, P. D., et al. Recent Marine Sediments, a Symposium, 736 pp., Am. Assoc.
Petro Geo!., Tulsa, 1939.
TWENHOFEL, W. H. Treatise on Sedimentation, 926 pp., Williams & Wilkins,
Baltimore, 1932.
Principles of Sedimentation, 610 pp., McGraw-Hill, New York, 1939.
The Rounding of Sand Grains, J. Sed. Petr., VA!. IS, pp. 59-71, 1945.
VEEN, J. VAN. Onderzoekingen in de Hoofden, 252 pp., Landsdrukkerij, The
Hague, 1936.
VEIllIE.EK, R. D. M. Krakatau, 546 pp., Landsdrukkerij, Batavia, 1885.
c H A p T E R F v E
The Formation
of Marine Sediments
, Whereas sea level provides a common base level for all rivers, the
horizontal part of the marine profile of equilibrium varies in depth
I according to local conditions. The size of incoming waves varies
with the climate (wind) and the length of fetch, and current action
also depends on local conditions. Hence, the depth to which erosion
is possible must also vary from one locality to another.
By the time equilibrium is attained, the profile is normally a very
flat one. The flatness of the curve is affected by many factors. In
this respect also the strength of waves and currents are important.
Obviously the nature of the country rock has great influence on
the shape of the profile. The degree of cohesion and solubility, the
size of the debris into which it disintegrates, and the hardness of its
constituents all count in the rate of supply, in the transport, and in
the abrasion of detritus.
The length of time since the profile was established is no less
important. When a river has reached its profile of equilibrium, ero-
sion does not come to a standstill. The bed continues to sink lower
and lower while a peneplain gradually develops. Neither does the
marine profile of equilibrium mark a final stage of development.
Erosion continues to carry the coastline land wards, and the sea floor
continues to be aggraded farther out. With rivers the profile ap-
proaches a horizontal line while the length remains constant. With
the marine profile the vertical range remains un'1tltered but the length
increases.
Up to this point our treatment of the marine profile of equilibrium
has been entirely theoretical. But when one seeks to apply it in
nature its value is found to be doubtful. Obviously, any relative
change in the position of sea level will entirely upset any approach
to equilibrium previously attained, Deposition or erosion can soon
adjust the profile of a river when diastrophic action has upset equi- t'
librium. But the vast area of the sea floor renders adjustment by I
sedimentation a lengthy process, and submarine erosion acts very
slowly. During the Ice Age sea level repeatedly swung up and down
many dozens of meters. Warping of the continental margins is con-
tinually changing the level of the land. Hence, conditions on all
coasts have been fundamentally altered recently, and we are left in
doubt whether the phenomena we now find represent an approach
to equilibrium or merely a stage in major readjustments.
The continental shelf and slope are generally looked upon as con-
forming closely to the marine profile of equilibrium for open oceanjc
coasts. However, as previously noted (p. 263), recent investigations
MAR I N E PRO F I LEO F E QUI LIB R I U M 305
have shown that "hard bottom" (rock and coarse detritlls) tends to
occur more especially along the outer edge of the shelf. This absence
of recent sediment on an area situated between environments closer
to the shore and farther out which show deposition can hardly repre-
sent an advanced stage of adjustment under the conditions now pre-
vailing. Neither can the continental terrace have been formed under
similar conditions, because oil wells, dredging, and seismic reconnais-
sance on the shelf along the east coast of the United States have
revealed a thick pile of sediment underlying the non-depositional edge
'.' , . ' ; ',::'.':' .: .... :. '::: : ',,'.: .:. '.:.' , ', ... :.: ",::.' ...,: ::':. ', . ', :" .. . .
::'.: :.': :.:.: ',.
_-_-__-_-_-_-_-_-_-.:--_-_-_-_-_-_-_-_-_---_-_-_-_-_-_-_-:_-==:.:.:_-.:_-_-_-_-_-_-_-_---_
A - - __-
Preserved
Deposited
Time_
Flc. 137. Schematic representation of sedimentary record. A, general rise of
top level of sedimentation, determining thickness of sedimentary series, shown
on the right; B, oscillations due mainly to diastrophism, giving disconformities
I-IV; C, minor oscillations due mainly to climatic and physiographic changes,
giving major diastems 1-6. Minor diastems due to incidental causes. At the
top of the diagram the time occupied by deposition and the time ultimately re-
corded are shown in black.
pelagic nature; that is to say, they show little direct influence from
the neighboring land. This means that coarser and frequently also
relatively fine segimentary detritus from the continents are reduced
ENVIRONMENTS OF SEDIMENTATION 315
FIG. 139. Schematic block diagram of the growth stages of a small delta in a
lake Of inland sea. TBA, deposited beyond shoreline; TDF, top-set beds; FDE,
fore-set beds; S, bottom-set beds.
The fore-set beds are deposited on the steep subaqueous slope. The
bottom-set beds are part of the normal deposits forming on the floor
of the lake or inlet. The fore-set beds gradually extend the delta
front over the bottom-set beds. The fore-set beds, in turn, are cov-
DELTA ENVIRONMENT 325
ered by the top-set beds as the river extends its course to the edge
of the delta front.
Large, lowland rivers carry only fine sediment to the coast. Either
no delta develops, or the sediment is dropped over wide areas. Hence
the fore-set beds slope much less steeply, while the top-set beds are
much attacked by wave action. The differences between the various
types of bed are then much less distinct. In general the rule holds
; , \
~..:. \ \
t: )
I', , , ,: (
/
\
,, \
I
/
20 40 /
Kilometers
I
FJG. 140. Map of the Mahakan Delta on the cast coast of Borneo with broad
estuarine distributaries. Depths in meters.
that, the larger a delta is, the sma ller are the surface slopes and the
finer the sedimentary particles.
The marine delta environment is characterized by the interplay of
marine and terrestrial influences. The inorganic material is almost
entirely supplied from land by the river that forms the delta. But it
may be submitted to wear, sorting, drifting, etc., by 'waves and cur-
rents before it is finally deposited. The organic sedimentary ma-
terials, both vegetable and animal, as also tests and skeletons, are partly
supplied by the river, e.g., mammalian bodies. trees, humic colloids,
etc. The remainder is derived from plants and animals that lived in
the lagoons and open waters of the delta, e.g., molluscans, forami-
niferans, echinoderms, seaweeds, etc.
With deltas forming in the sea the deposition is in part purely
marine, such ::IS that of the fore-set beds, in part purely terrestrial
such as that of the inner portions of the top-set beds. Bars of sandy
326 FOR MAT ION 0 F MAR I N ESE 0 I MEN T S
material may begin to develop beyond the coastline, but they fre-
quently change to islands with dunes and are ultimately incorporated
in the land. The lagoons enclosed between such offshore bars and
levees are originally of normal salinity. Later, however, the water
becomes brackish and finally entirely fresh. In this manner, with the
growth of the delta, the irregular boundary between terrestrial and
marine sedimentation is gradually pushed out seawards. Matters may
be complicated by subsidence that goes on permanently over the entire
area of a large delta. During outbuilding the river tends to alter its
,. 1838
- 1859-1870
and is gradually planed off. Only the bottom-sees and the deeper
lying fore-sets will ultimately remain.
The modern investigations of the Mississippi Delta (Russell et aL,
1936; Russell, 1948), however, have shown that subsidence may inter-
vene and result in the safe conservation of delta deposits in the earth's
crust. Probably the increasing weight of the delta causes subsidence
of the entire crust, while compaction of the deeply buried strata adds
to the sinking movement of the surface. Locally the rate of subsi-
dence of the Mississippi Delta attains no less than 2 to 3 m per century.
The amount of sediment added to this delta is estimated at 2 X 106 tons
per day. At the mouth of important distributaries the delta front is
328 FOR MAT ION 0 F MAR I N ESE 0 I ME N T S
built up and the coastline is carried gradually seawards to the edge of
the shelf because accumulation outweighs subsidence and marine ero-
sion. Owing to increasing length the grade of the river becomes
gradually smaller and the bed is aggraded by the deposition of sedi-
ment. Finally the river breaks through its own levees and forms a
new bed beside the abandoned course or it seeks an entirely new
short cut to the coast over the delta . Around the old mouth the
supply of sediment is suddenly cut off. Waves and currents build
an offshore bar. In the meantime the secular subsidence continues,
and the bar is converted into a row of islands. Ultimately these arc
submerged together with the lagoons and levees behind, and the
new coastline is established a great distance back of its original position.
This interplay of growth and loss shifts from one point to another
along the periphery of the delta. The ultimate result is an almost
stationary condition as to horizontal extension of the ddta. Hut the
entire volume of the structure is steadily enlarged, and the strnta are
becoming gradually more compact. Thus the thickness of the Ter-
tiary column of the Mississippi Delta is estimated at 10,000 III and the
Quaternary at 1000 m. This vast accumulation lies in n geosynclinal
(?) depression in an east-to-west direction along the Gulf coast of the
southern states.
Evidently in a large delta the great weight causes isostatic and
compactional subsidence, to which geosynclinal subsidence may he
added. The normal sequence of the physiographic delta cycle, con-
sisting of growth followed by erosion, will not be developed. The
occurrence of widespread and thick series of delta deposits in the
stratigraphical column in which top-set beds are also incorporated
thus finds a natural explanation.
The foregoing discussion explains why the deposits of a delta, espe-
cially the top-set beds of a large one, are the most variable formations
known to geologists. There is an intimate intermingling of marine,
fluviatile, lagunal, terrestrial, and limnic environments. The distribu-
taries are continually changing their comses, cutting off new lagoons,
scouring out fresh channels now here, now there, filling up old
lagoons, and pushing the beach outwards. False bedding is frequent.
Birds drop marine shells in terrestrial and lagoonal environments. The
river washes terrestrial material out into the marine environment.
During spate the river causes freshening of the waters in brackish
lagoons and along the coast, and the result is mass destruction of life.
Marine erosion is nibbling at the top-set beds and pushing back the
coastline. Bars may be thrown up temporarily. only to he destroyed
DELTA ENVIRONMENT 329
later. Invasions by sea water into brackish and fresh lagoons and
across dry land are the result of subsidence or tsunamic waves. Slump-
ing frequently displaces and distorts the strata because many deposits
are laid down at steep angles.
A separate type of delta is represented by tidal deltas (Fig. 143) .
The river is replaced by an opening in an offshore bar or similar nar-
2'
I. '0
..
10
2 4 6
KIIQ",ote"
7.
FIG. 143. Double tidal delta of Ocracoke Inlet, on the east coast of the United
States. Depths in fathoms.
rows in which strong tidal currents sweep in and out. Some of the
sediment is eroded from the walls and floor of the tidal channel, but
the greater part is provided by coastal drifting that carries material
from the adjoining beach and foreshore into the tidal channel. The
lagoon will also supply a certain amount of fine terrestrial and plank-
tonic sediment.
In two important aspects the tidal deltas differ morphologically from
the normal type of river delta. The surface cannot emerge above sea
level because the tidal currents are the active elements. The entire
330 FOR MAT ION 0 F MAR I N ESE DIM E N T S
mud cracks, etc. Only where seasonal variations in sea level occur in I
combination with a very low coastal plain may flooding occur at long
intervals, leading to thorough drying in between.
It is also ObVlOUS that an advancmg sea will generally rework its
own littoral deposits before final deposition can occur. The littoral
deposits left by a retreating sea are likely to be destroyed before
becoming covered by later sediment. The chance of shore formations
being preserved is therefore only slight.
Bourcart showed that some fossil beach conglomerates and also
boulder tracts dredged in deeper water off the Atlantic COast of
Europe are reworked river deposits. This follows from the varied
composition. Instead of consisting of reworked fragments of the
underlying rock formations, they contain a great variety of rock
types, among which vein quartz may predominate. Bourcarr believes
that the boulders were carried to the coast and scattered along the
beach by coastal drifting. The same explanation might be offered
for the non-local composition of many recent cobble beaches. In
glaciated regions the reworking of a cover of boulder clay to form
a cobble beach is a common feature. Bourcart also pointed out that
many basal conglomerates resting on a surface of marine abrasion
show the same polymict composition derived from the waste products
of extensive inland areas.
Oii the other hand many pebble and cobble beaches are composed
entirely of local material, identical with the rock exposed in the
adjoining cliff sections. It would evidently be wrong to suppose
that beach material is always dominantly composed of reworked de-
nudation products from the land.
The littoral environment presents special characteristics where man-
grove vegetation occurs. The force of the waves is entirely broken,
and all properties due to their activity are lacking. The closely spaced
roots tend to entrap lutite, and in this manner fine deposits are formed
on open coasts. In a fossil state, when the roots have rotted away,
such a deposit would hardly suggest a littoral facies.
Flat, low-lying areas of great extent may be occasionally or repeat-
edly invaded by the sea. It is merely a matter of definition whether
they are to be incorporated in the beach or not, but this possibility
should at any rate be borne in mind when interpreting fossil deposits
showing evidence of temporary marine invasions.
An area illustrating this point is the Rann of Cutch to the south-
east of the Indus Delta. According to Lyell (1868), a former land
area of 5000 km 2_ was inundated by the sea and converted into a
EUXINIC ENVIRONMENT 335
lagoon after the earthquake of 1819, while a low ridge was formed
in the neighborhood. Btlt the remainder of the Rann, 17,000 km2, is
frequently invaded by the sea when the monsoon winds raise the
water in the Gulf of Ciltch and the creeks at Luckput. On evapora-
tion of the sea water a crust of salt is left. The Rann is also likely
to be flooded occasionally in some parts by river water of the Indus.
Another cause of temporary flooding by the sea is found in tsu-
namis, which may spread over vast areas, carrying marine shells far
inland. Large deltas are especially liable to this kind of marine deluge,
as Grabau (1940) has pointed out. He coined the term "marining"
for very short-lived spreading of marine waters over a surface of pri-
marily floodplain or delta deposits, and he called the resulting sedi-
ments "huangho" deposits after the Huang Ho of China, the best-
known modern example.
MARINE SEDIMENTS
Shelf Sediments. For these sediments also the reader is referred to
the above treatment of the neritic environment.
Large tracts of the shelf are covered by sediments identical or
strongly related to the terrigenous sediments of the bathyal and abyssJ-l
zone described below (e.g., Bourcart, 1947). These deposits will not
be dealt with separately here.
However, one feature of recent marine sediments from the neritic
zone needs special attention. Although deposits rich in lime are com-
mon among the fossil representatives, they are much less frequent on
the present sea floor, especially in the temperate and cold regions. But
even in tropical waters lime is rare, except in the neighborhood of
coral reefs where calc~reous sand and mud are generally forming on
the sea floor.
One of the few recent calcareous deposits forming in higher lati-
tudes is the so-called tongue of the bay of Mont Saint Michel on the
western coast of France. Bourcart has shown that the lime may
reach 7510 and is derived from a great variety of organisms that
belong in depths of 10 to 30 m. A sub recent deposit of these cal-
careous elements on the foreshore is undergoing erosion, and the
resulting fine products are carried to the shallow regions of the bay
and deposited there in the form of a fine gray mud.
Another, di fferent type of recent calcareous deposit of limited
extent is found in the Irish Sea. It consists of calcareous algae,
echinoderms, and molluscs (Herdman, 1895).
Two explanations for this contrast between recent and fossil neritic
sediments may be offered. It is possible that the spread of warm
climates to high latitudes in the geological past favored the formation
of calcareous deposits. But even then the scarcity of equivalents from
the Indonesian seas remains unexplained.
A different cause may be found in the loss of lime by deposition
in the deep sea since the Cretaceous, resulting in a lower concentration
in the surface waters of the oceans. This hypothetical view will be
treated later (p. 392).
Hemipelagic or Terrigenous Sediments. For many types of sediment
belonging in this class the term hemipelagic is most appropriate, be-
338 FOR MAT f 0 N 0 F MAR f N ESE 0 f MEN T 5
cause they contain mainly pelagic and but little terrigenous matter.
Some, however, are almost free of planktonic remains, and for these
the term terrigenous is more suitable. Submicroscopic particles, al-
though deriving largely from the land, are here classed as pelagic
matter because they are carried to midocean.
It was pointed out above that the hemipelagic deposits are roughly
equivalent to bathyal sediment. But, whereas the sediments accumu-
lating at depths of several thousands of meters in the Indonesian and
West Indian deep-sea depressions, even the typical blue muds and
volcanic muds, should actually be incorporated in the abyssal deposits,
there is no objection against classifying them as hemipelagic sediments,
because the adjacent islands play a predominant part in their forma-
tion. For the depth at which accumulation of these sediments is
taking place proves to be of minor importance and doe~ not appear
to influence the nature of the deposits fundamentally. It is only
bottom-living microorganisms that may differ according to the depth.
Blue mud is the most common representative of the terrigenous
deposits. It consists of Rne terrigenous material with an admixture of
pelagic products, such as planktonic Foraminifera. In general it
accumulates at moderate distances from land, where coarse fractions
are lacking owing to the absence of strong currents, and where the
production of lime is not dominant. The highest percentage of lime
allowed is 35 parts of calcium carbonate in 100; otherwise the term
calcareous mud is used. The color is lead-gray to dark blue-gray in
moist condition, changing to light yellowish gray or bluish gray on
drying. This'tint is due to the low degree of oxidation of the con-
tained iron compounds (hydrotroilite, ferrous sulfide) and finely di-
vided organic matter. The upper layer of 0.5 to a few centimeters
thick is dark reddish brown, owing to the oxidized state. In the
interior of the deposits reduction takes place to ferrous sulfide, which
ultimately forms marcasite or pyrite. It is found that in general the
color becomes darker with increasing content of organic matter,
while lime tends to cause lighter shades.
The predominance of terrigenous matter is not due to lack of
planktonic tests, for these are plentiful at moderate distances from
the coast. The abundant supply of material from the land is evi-
dently responsible for a dilution of planktonic products. This con-
tention is borne out by the observed high rate of sedimentation as
compared to that of the organic oozes forming at greater distance
from the coast. Illustrations of the point in question are to be found
HEMIPELAGIC SEDIMENTS 339
340 FORMATION OF MARINE SEDIMENTS
., 119 120
Station No .
121 122 123 124
% minerals > 20 /.I 9.8 5.1 3.7 1.0 0.5 0.4
% lutite < 2/.1 30.0 33.9 37.6 39.2 27.1 21. 5
% CaCCa 17.0 27 . 1 27.5 46 . 1 54.0 76.6
Sahul shelf
Timor 119
120 121 122 123 124
1000 ~
2000~
3000
150
Kilometers
FIG. 145. Stations of the SneJ/ius expedition south of Timor to Sahul Shelf, show-
ing gradual decrease of mineral grains larger than 0.02 mm away from the coast,
independent of the slope of the bottom. (Data from Neeb, 1943.)
FIG. 146. Deep-sea deposits of the Moluccas. V, volcanic mud; volcanic-terrigenous mud; T, terrigenous mud; RC, red clay;
G/, calcanous mud, mostly globigerina ooze. Full-drawn line =' of 3% volcanic ash in globigerina ooze. Dashed line = limit
of volcanic ash. Shelf sediments white. Soundings by dot. (Simplified from Neeb, 1943, Plate l)
HEMIPELAGIC SEDIMENTS 343
The ash of several volcanoes in this area could be recognized by
some distinctive features. The Batu Tara ash northeast of Flores
shows clear crystals of leucite. The Una Una ash (Gulf of Tomini,
Celebes) from the 1898 eruption is characterized by small biotite
crystals and a low glass content. It has been recognized in the muds
of Makassar Strait at the other side of the northern arm of Celebes
--A
" .. .. B
ec
OD
@E
FIG. 147. Distribution of Tambora ash of 1815 in bottom samples. A, limit of
ash fall; B, limit in samples; C, 60-80% ash; D, 3-30% ash; E, less than 3% ash.
(Redrawn from Neeb. 1943. Plate II.)
fine terrigenous fractions. Close to the shore of Groot Kei, 600 m. (Cf. the
well-sorted St. 102 and 363.)
St. 102. Terrigenous mud, very coarse, little lutite. Close inshore to Groot
Kei, 350 m. (Cf. st. 103.)
St. 363. Terrigenous mud, very fine. East slope Weber Trough, 950 m. Far
from the coast but not deep.
St. 197. Terrigenous-volcanic mud. A, fine Tambora ash; B, lutite. North
of Flores, 5100 m. (Cf. the almost identical diagrams, st. 82 and 107.)
St. 82. Hemipelagic globigerina ooze (36% CaCOs); globigerina ooze accord-
ing to Neeb. A, pelagic Foraminifera; B, terrigenous lutite. Northeast of Obi,
1000 m. (Cf. st. 197.)
St. 107. Globigerina ooze. A, broken shells; B, terrigenous Iutite. East of
Tenimber Islands, 400 m.
St. 260. Terrigenous mud; rather fine grain. Mindanao Trough, 7950 m.
346 FORMATION OF MARINE SEDIMENTS
found in the north polar arctic seas. During the Pleistocene this type
of deposit covered considerable areas that are now receiving no glacial
matter or only inconsiderable amounts.
Glacial marine sediments deposited close to the termination of a
glacier tend to consist almost entirely of glacial materials. An ex-
ample of recently elevated glacial marine sediments has been described
from Alaska by Russell (1893). They consist of stratified morainic
matter, with sandy clay containing large ' quantities of glaciated boul-
ders distributed irregularly throughout the 1500 m of sediment. In
finer portions seashells are numerous and boulders with attached cases
of annelids also occur. These stones must have remained exposed on
the bottom of the sea for some time before being wholly buried.
The typical muds forming in poorly ventilated basins, such as the
Black Sea, were mentioned earlier. The main characteristics are the
high organic content, the presence of hydrogen sulfide, and the rich-
ness in ferrous sulfide. The color is gray, owing to a considerable
admixture of calcium carbonate.
Yellow mud and red mud are merely special types of blue mud that
owe their exceptional color to the climatic conditions of the adjoining
land. For example, the Yellow River carries enormous quantities of
fine yellow sediment from the deserts and regions of loess in China
out into the China Sea. The resulting yellow mud on the sea floor
hardly differs from normal blue mud except in its brilliant color and
the nature of the iron compounds. In environments receiving sedi-
ment from lands covered by lateritic soil the deposits also tend to
show a reddish brown hue or even bright red colors. Neither of the
types of terrigenous muds mentioned has a wide distribution on the
sea floor.
TABLE 20. AREAS COVERED BY HEMIPELAGIC SEDIMENTS
Total 73 20
Green mud is similar to blue mud except that it is green. The color
is occasionally due to chlorophyl1 or green terrigenous minerals but
usually to the pre~ence of glauconite, in which event the term glauco-
PELAGIC SEDIMENTS 347
nite mud is more appropriate. According to Revelle, glauconite muds
are generally found off coasts lacking important rivers. Glauconite
is often associated with phosphorite nodules in areas of slow deposition
or none.
Pelagic Sediments. The pelagic (or eupelagic) sediments are char-
acterized by the absence of terrestrial mineral grains larger than the
colloidal fraction. The most common constituents are clay minerals
and remains of planktonic unicellular organisms.
The term "pelagic sediments" corresponds very nearly to "abyssal
sediments," because on the floor of the deep sea beyond the bathyal
zone the direct influence of the continents in the form of microscopic
mineral grains is seldom evident. But w here strong currents or winds
carry continental waste far out to sea, or wherever large rivers spread
their turbid waters to dozens of kilometers from the coast, the terrig-
enous admixture is found far beyond the 1000-m-depth curve. The
same holds for inland seas or along steep continental slopes. It has
already been pointed out that, in the abyssal depths of the Moluccan
deep-sea depressions, no true pelagic deposits are accumulating, not
even below the 5000-m line. Conversely, pure pelagic sediments are
found on banks of less than 1000-m depth in midocean. These excep-
tions clearly demonstrate the necessity of distinguishing sharply be-
tween pelagic and abyssal deposits.
The finest terrigenous matter, the colloidal lutite particles, do not
flocculate when diluted sufficiently. This explains why a minute
quantity is found even in the purest waters of midocean. Sea water
in midocean of average depth contains 75 mg of clay in each column
of I cm 2 cross section. On deposition this would form a layer 0.3 mm
thick. Owing to their relatively large surface the flakes are carried
in suspension almost indefinitely by ocean currents and the circula-
tion of the deep sea. Not until a particle happens to be carried by a
slow bottom current along the deep-sea floor does it find opportunity
of settling somewhere. Meanwhile in the course of hundreds of years
it may have been carried through all the oceans before it dropped to
the bottom. Hence, the finest fractions of pelagic deposits bear a
universal character and should be practically identical everywhere,
provided that conditions of deposition are otherwise equal.
Naturally this does not exclude a certain amount of sorting accord-
ing to different environments. Thus Revelle found an appreciable
difference between the lutite fraction of red clays on the one hand
and globigerina ooze on the other. But the evidence he obtained
favored the view that the particles are supplied by the continents and
348 FORMATION OF MARINE SEDIMENTS
not by weathering of volcanic matter on the sea floor. Correns was
even able to show that in the equatorial Atlantic the mineralogical
nature of the lutite shows regional variations which do not conform
to the distribution of the red clays, globigerina oozes, and the blue
muds of that region (Fig. 149). He attempted to explain this by
differences in neighboring sources. This would imply, however, that
in many cases the_ clay particles were not carried more than a few
PELAG I C SED I MENTS 349
hundred kilometers by oceanic currents, and that the rate of accumu-
lation should vary with the abundance of local sources. These deduc-
tions cannot be granted, and the present author is inclined to assume
tentatively a selective deposition from a more or less uniform supply
in the bottom waters, or local diagenetic alterations of clay minerals
on the sea floor (see p. 237). Moreover, Correns himself showed that
grains of O.l-mm. diameter are carried by surface currents at least
2000 km away from the southern coast of the Gulf of Guinea (Fig.
150). The lutite, then, should be spread out far and wide and mixed
with the clay fraction of other sources long before reaching the
ucean bed.
On the sea floor this fine material is diluted by truly pelagic matter
of organic origin and further by volcanic ash, terrestrial and cosmic
dust, etc. Depending on which of these materials predominates, dif-
ferent types of pelagic deposit are formed. Like the hemipelagic
muds, the pelagic oozes are again classified according to the coarser
fractions, because only these can be conveniently studied under the
microscope.
It is not generally realized that the often-reproduced illustrations
of deep-sea deposits given by Murray and Philippi are of washed
samples from which the colloidal fractions have been removed. Less
clear, but more truly representative, are the photomicrographs of the
actual samples given by Correns (1939). (Compare with the photo-
micrographs of hemipelagic samples given by Neeb, 1943.)
Calcareous Oozes. The pelagic sedimentary particles consist for the
greater part of tests of plankton organisms. Among these the princi-
pal representatives in warmer waters are Foraminifera, coccoliths, and
rhabdoliths (minute calcareous disks forming the protective structure
of Coccolithophoridae, a kind of calcareous algae) and siliceous tests
of radiolarians. In colder climates the siliceous frustules of diatoms
are paramount. In those deep-sea regions where the tests of globi-
gerinas, the most important planktonic Foraminifera, do not dissolve
either during settling or on the bottom, a calcareous ooze is formed,
called globigerina ooze. Obviously globigerina ooze may grade 1ater-
ally into blue mud or other types of mud near the coast. Generally
3070 of foraminiferal tests is taken as the lower limit for globigerina
ooze. Schott proposed counting the number of individuals larger than
0.1 mm and taking 6000 per cubic centimeter as the lower limit justi-
fying the term globigerina ooze. Although there is something to be
said for this method, it is tiJ:Tle consuming and requires careful treat-
ment so as not to crush the tests. It would also be inconvenient to
350 FOR MAT ION 0 F MAR I N ESE 0 I MEN T S
.
II)
PELAGIC SEDIMENTS 351
change the original definition after half a century of use. A minor
alteration was made by Miss Neeb in describing the Snellius samples,
namely of taking 30% lime instead of Foraminiferll as a limit, but with
the condition that more than half the calcareous matter is composed
of planktonic Foraminifera (see p. 342). This is far simpler than
counting the tests and more accurate than estimating the percentage
Diameter, microns
Sand Silt I Lutlte
-E...
>
Cf.)
e0
......
"S
~
"8~
E~
>, ::I:
"P 0
..I: U
bIlp
:.:: C
~e
~
ci ....
c: 8
...
~ci.
-
~aO
... ;::;
2co."t>O
E'"
.s
......
2
i
0
"B
.
'0
bIl
.
8
c
~
.
~
N
~
.;
fZ
354 FOR MAT ION 0 F MAR I N ESE 0 I MEN T 5
pelagic sediments in tropical regions than in polar areas. Production
is also at a peak where upwelling water brings a rich supply of nutri-
ents into the photosynthetic zone. Although production is mainly
due to planktonic organisms. there might theoretically be additional
supply by precipitation from oversaturated surface water. Another
factor in lime production is the mass destruction of life where warm
and cold surface currents meet and the tests produced by growth of
organisms over a wide area are concentrated after their death on a
relatively small patch of the sea floor.
In the second place dilution by other sedimentary matter helps in
determining the lime content. Broadly speaking this admixture in-
creases towards the coast, especially where large rivers debouch or
deserts deliver dust. Volcanoes, whether subaerial or submarine, may
also strongly dilute calcareous matter. Oceanic currents are seldom
sufficiently turbulent to hold globigerina tests in suspension, but they
may carry away clay particles where they sweep along the floor. In
such positions a pure globigerina ooze accumulates owing to the lack
of pollution by fine terrigenous matter. This frequently occurs on
submarine ridges, seamounts, and other topographic highs.
In the third place solution influences the lime content. This factor
comes into play during settling of calcareous tests and on the sea floor.
Roughly half the number of tests produced by planktonic Forami-
nifera are abandoned by the animal while reproducing. The other
half either passes through the digestive organs of plankton-feeding ani-
mals (where it is dissolved?) or dies and sinks to the sea floor. In the
latter case solution may be retarded during the settling, owing to pro-
tective covering by organic matter. But solution is also active on
the sea floor, inasmuch as the bottom waters are hardly ever perfectly
stagnant or saturated. Decomposition of organic matter contained in
the sediment produces carbon dioxide, which, in turn, results in
activating the solution of calcium carbonate.
The author is inclined to assume that by far the greater amount of
solution takes place on the sea floor and not during the settling because
the time of settling is so much shorter than the time left before a test
becomes covered and protected from further attack on the sea floor.
The theory of lime solution is not yet satisfactorily established.
It was pointed out in an earlier chapter that the distribution of car-
bonic acid in sea water is yet insufficiently known, and this is the
main factor determining the solution of calcium carbonate. A detailed
discussion of these _matters is given by Sverdrup et a!. (1942).
PELAGIC SEDIMENTS 3SS
Solution is found to be more active with increasing depth, partly
because the distance of settling is longer, partly because the solvent
action becomes more powerful owing to the lower temperarure and
salinity at great depths, and probably also to higher pressure.
Theoretically solution of lime should also increase with the velocitv
of the bottom current. It is observed, however, that in general the
Per cent
00 ill W ~ ~ ~ ~ ro w ~
3000
Near -shore ,S] \
World _...:., "-./
J "
- --
,9 .~
<II
~ 4000
jL. East Indies
// ~ Challenger
:::e
5000
~/ __ 0 ~ ..... , /
~
y; _::..... 0 ~lantic Ocean
Kl"-
7000 I
8000
FIG. 153. Relation between depth and lime content of deep-sea deposits. The
line for Indonesia was constructed from all available data, the Atlantic from data
given by Pia (1933), the world average from data given by Trask (1937). This
author gave separate data for pelagIc (more than 800 km) and near-shore de-
posits (less than 800 km from the coast), making allowance in each group for
unequal distribution of samples by taking an equal number from each repre-
sentative area. In combining these twO groups for obtaining a world average
the present writer gave different weights for eaeh depth to ncar-shore and pelagic
deposits respectively as follows: 90 m 1:0; 315 m 20:1; 450 m 10:1; IHO m 2:1;
2250 m 1:1; 3150 m 1:5; 4050 m 1: 10; 4950 m 1:25; 5850 m 0:1; 6750 m 0:1.
OOr---~--~-r~--~~--r---~~
50
_ 40
.,c:
:: 30
~
20
10
Diameter, microns
FIG. 154.Equal area distribution curves of grain size in red clay and blue mud
of four Meteor stations. (According to Correns, in Barth, Correns, and Eskola,
.. 1939, Fig. 33, p. 172.)
10
O~~~~T-~~~~ ~~~~~~~~~~--~
0.02 0,2 2 20 200 0,02 0.2 2 20 200 2000
Diameter. microns Diameter. microns
FIG. 155. Distribution curves showing composition and grain size of red clay,
Meteor station 305, and blue mud, Meteor station 222. A, montmorillonite; B,
mica; C, feldsp ar; D, quartz; E, rest; F, halloysitc; G, siliceous tests. (According
to Correns, in Barth, Correns, and Eskola, 1939, Figs. 38 and 39, p. 178.)
c..
~>
(J)
u
.
'6'.0
"&.
362 FORMATION OF MARINE SEDIMENTS
TABLE 21. AREAS COVERED BY MARINE SEDIMENTS
.,
and therefore gives the direction in which the beds are younging.
Care should be taken, however, to use only such parts of current-
bedded series as show evidence of concavity together with a clear-cut
break in slope at the other end of the laminae. Otherwise the upper
convex part may be mistaken for concavity the other way up (Fig.
157, C). Luckily the convex laminae hardly ever show a break in
slope at the lower end but do show a more or less pronounced con-
cave slope back to horizontal. On account of the possible misinter-
pretation of upper convexity for a lower concavity it is safer not to
rely on a single observation but to search for several examples in a
single exposure. _
STRATIFICATI O N 365
Other phenomena that may indicate the directions of younging are
(1) graded bedding (p. 366) and (2) washouts (Fig. 158). Washouts
are channels eroded in deposits during the accumulation. They may
be characterized by unconformable contacts at their edges, a coarser
fill than the surrounding deposit, or the inward slope of their sides.
See also Shrock (1948) for many other guides to interpreting the
sequence in sedimentary deposits.
The ultimate cause of stratification is to be sought either in inter-
ruptions of deposition or in variations in the quantity or nature of
rather surprising in view of the rich bottom life of the seas, and the
cause is not properly understood.
Now and again stratification may be rhythmic in nature and display
cycles. Instances are known of the regular alternation of coarse and
fine beds, of lime and clay, of sand and marl, and many other repeti-
tions. Threefold repetitions of various rot:ks are likewise found in
some rock series. In several cases the cause is to be sought in excep-
tional storms, that result in the churning up and the resettling first of
sand and later of silt and clay; in others, in the cycle of seasons, the
rhythm of tides, etc. Other possible causes of rhythmic stratification
are: (1) longer climatic periods, such as "BrUckner's period" (the
existence of which is not admitted by all investigators); (2) the pre-
cession of the equinoxes (21,000 years); and (3) the variation in the
eccentricity of the earth's path around the sun (91,000 y ears).
Highly complex cycles in the Pennsylvanian system III the central
and eastern states of the United States have been described by Wanless
and Weller as cyclothems. A complete cyclothem is composed of
the following members in descending order:
Marine sediments:
8 Shale with "ironstone" nodules and bands.
7 Limestone with marine fossils .
.r
Black shale with large concretions.
Continental sediments:
5 Coal.
4 Underclay.
3 Limestone without marine fossils.
2 Sandy shale.
Sandstone unconformable on lower beds.
These cyclothems are traceable over wide areas and are found to
be of remarkable constancy. The authors mentioned attribute them
to diastrophic movements w hich must have affected the entire eastern
half of North America. Later, Wanless and Shepard (1936) sug-
gested eustatic movements of sea level in connection with glacial stages
of the Upper Paleozoic as the major cause.
Stratification in Recent Sediments. Present knowledge concerning
stratification of recent marine deposits is limited by the technique of
sampling. Earlier instruments either disturbed the strata or took
samples only a few inches long. Hence, the impression was gained
that uniformity is the rule. When longer and longer samples were
obtained variations in color and composition were repeatedly en-
S T RATIF I CATION 369
countered. In the Atlantic deep sea, for instance, Philippi (1910)
discovered widespread stratification that was later shown to obtain
over t he entire Atlantic (Schott, 1939). In modern samples of a few
meters' length stratification is even found to be a rule to which there
are few exceptions. One of these, a core of red clay 246 em long,
was taken a few hundred kilometers off the coast of California.
The most typical arrangement for pelagic deposits is a stratum of
globigerina ooze 20 to 40 cm thick resting on red clay of similar
thickness, and this alternation is probably repeated several times at
greater depths (Fig. 161, p. 377). In the red clay certain Foraminifera
indicative of warm surface waters are absent though occurring in the
underlying and covering stratum of globigerina ooze. They are re-
placed by forms belonging to higher latitudes. This indicates lower
surface temperatures during the deposition of the red clay. The
lower lime content of this stratum is apparently due both to smaller
supply of lime and to a more pronounced bottom current with strong
solvent action. These relations point to a colder climate during the
accumulation of the red clay, and the conclusion is warranted that
this stratum represents the last Ice Age. The lower globigerina ooze
can be attributed to the last interglacial or interstadial period.
In the long samples obtained by Piggot from the northern Atlantic
as many as four cold periods are represented (Figs. 159 and 160).
Age determination by radioactive methods indicates that these arc all
stages of the last Ice Age (\ Visconsin = Wiirm), because the oldest
stratum ,vas deposited not more than 70,000 years ago. In the Carib-
bean the same investigator took a long sample of globigerina ooze that
proved to contain several strata with Foraminifera indicating about
five warmer and colder periods. The age determinations described
on a later page show that the second (Mindel-Risz) Interglacial period
was reached. Further investigations are needed, especially of the very
long samples Pettersson has s:;ollected, before a correlation between
glacial variations of climate and strata on the ocean floor can be estab-
lished with any certainty.
Closer to the continents the buried stratum of red clay is replaced
by blue mud and the thickness increases (Fig. 161, p. 377). The direct
influence of the land becomes gradually more noticeable. In cores
taken in the regions where blue mud and red clay are accumulating
at the present time no stratification is visible, but examination of the
contained Foraminifera generally shows the presence of the stratum
with cold-water forms. Close to the foot of the continental slope
the postglacial stratum becomes so thick that samples of less than 1 to
370 FOR MAT ION 0 F MAR I N ESE 0 I MEN T 5
1Y2 m no longer reach the deposits of the Ice Age. The same strati-
fication has been observed in the western Indian Ocean, and the
phenomenon may be universal.
It is more difficult to find an explanation for the opposite sequence,
in which red clay is found resting on a layer of globigerina ooze
(Philippi, 1910; Murray and Hjort, 1912). The suggestion has been
Hallf..
--.~.
-0- ____ - _---.-0--0-
4 5
____--~--------------___ tJi,;
6 7 8 9 10 11
-o-o-.--Fllmouth
12 13
3 t~
10680"'" . 10,620 6,420
II,~" 11,6~~11 IJ,~ "3 ...II~,JOO
\.!. J~..;_70_ _ _!.::I
_ _I... I,~~
"
F,,;, 159, North Atlanric cores, showing depth and location and the relation of
the glacial and volcanic strata, (After Piggot, 1938, Scientific Monthly, Vol. 46,
Fig. 6, p. 214.)
offered that swift subsidence of the sea floor took place from the
level in which globigerina ooze may accumulate to depths in which
solution of lime causes the formation of red clay. But this hypothesis
requires movements of questionable magnimde and speed. It seems
more probable that the globigerina ooze accumulated at lesser depths
somewhere in the neighborhood and slumped down into depths where
only red clay could form in situ. After the slump, normal accumula-
tion of red clay continued on top and formed the upper layer of the
samples. Until the topographic forms around the sites of such ab-
normal stratification have been accurately ascertained no definite
answer can be given. The explanation offered here is even more
probable for the cases in which three or more strata with different
STRATIFICATION 371
o. '00 Tt.u
'1,.500
",400
f.Ofll. '0
DlPTH 1100 UtDO
..
elll 10 ISOO
uoo
' 4100
liDO
...,,00 11,700
".:SOO
HS,'OO
1000 ''',400
",100
to.7OO
..5.
"DO
z
.
z"
Zo
::~
1100
U"IOO
F ".
14,s00 A
11,'00 n,IOO z
METE'"
~~oo !OO
....
I
800
I
.b.
1000
I
'000
I
I
1100
I ,
obo 81)0
NAUTICAL 11'1..(1
000 1000
ITB ZONE 0' ALIt.U,IC VOI.CAHIC OLASS SHoUOI zo.n: 0' "'IiIO".~OU' 0['011''1
FIG. 160. Chronology of three North Atlantic cores. The right-hand profiles
show the fluctuations in the temperature of the surface waters at the site of
each core as deduced from planktonic tests. (After Piggot and Urry, 1942, Bull.
Geol. Soc. Am., Vol. 53, Fig. 2, p. 1195.)
one case, and this lends support to the explanation of these sands by
turbidity currents given on an earlier page (240), because both phe-
nomena may be attributed to the presence of topographic highs at
not too great a distance.
No glacial stratification was found in the samples 1% to 2 m long
that the Snellius expedition took in the Moluccan area. This might
be due to lack of change in the tropical climate of this region during
the Pleistocene. However this may be, another factor is that sedi-
372 FOR MAT ION 0 F MAR I N ESE 0 I MEN T 5
mentation goes on at so high a speed that much longer samples would
be needed to pierce through the postglacial layer in most if not all
localities. For instance, Kaoe Bay between the northern arms of
Halmahera is 500 m deep, but the entrance is only 50 m deep. During
glacial low sea level this basin must have changed to a lake in which
deposits quite different from present ones accumulated. The samples
of J48- and J 68-cm length failed to reach these fresh-water deposits.
In a later paragraph evidence will be presented for assuming an accu-
mulation of many meters since the close of the Ice Age even in the
wide Moluccan basins. In Kaoe Bay a few dozen meters may have
been deposited.
Apart from the Quaternary stratification just discussed little is
known concerning the formation of strata in hemipelagic and pelagic
sediments. Alterations in color and composition are met 'with here
and there, but nothing can be said as to which of all the possible
causes is responsible in any particular deposit. In the Moluccan deep-
sea basins alterations in color have been observed that were some-
times repeated many times. But these cases were always found to
be of local occurrence and no cause was evident. Only when coarse
layers of volcanic ash and pumice were encountered in the neighbor-
hood of active volcanoes was the explanation obvious. Similar occur-
rences in the northern Atlantic must be attributed to violent pre-
histdric eruptions probably on Iceland.
In connection with the latter volcanic stratum an interesting feature
of deep-sea sedimentation has been treated by Bramlette and Bradley
(1940). These authors pointed out that the volcanic ash is not con-
centrated in thin layers, but part occurs in one or several ill-defined
strata, while the remainder is scattered at random through a consider-
able thickness of the sediment above. This cannot be attributed to
differential settling of the shards, because they are mixed in with a
layer representing nearly all of postglacial time, although the size of
the ash particles is such that they should have reached the bottom in
five years.
A similar case has been noted by Miss Neeb in the Moluccas. The
ash deriving from the Tambora eruption (see p. 343), although thrown
out 115 years prior to the collecting of the bottom samples, is found
right up to the surface of the deposit. According to their size the
grains should settle to the bottom in a few years' time at the most.
As there is an increasing admixture of clay from the bottom of the
volcanic stratum to the surface (some 20 to 30 cm thick in a few
samples) it cannot be doubted that accumulation of non-volcanic
STRATIFICATION 373
matter has gradually proceeded since the eruption to an amount of
at least a dozen centimeters. Hence, part of the ash is mixed with
material that has settled to the bottom recently. Neither can the ash
in the upper part of the samples be referred to material still washing
off the neighboring islands, because it is fresh, while similar ash on
Java is known to weather in a few dozen years. The suggestion
offered was that turbulent currents have retarded the settling. Not
only, however, is turbulence of sufficient strength highly improbable,
but the currents causing this turbulence would have swept all sus-
pended ash away and scattered it far and wide.
A more plausible explanation is offered by Bramlette and Bradley,
who attribute this phenomenon to mud-feeding animals. The newly
deposited ash is supposed to have passed through the intestines, and
succeeding generations of the animals must have picked up progres-
sively fewer shards as the sediment became more and more diluted
by the continual influx of the constituents belonging to the normal
sedimentation. Each time an animal scooped up mud and later ex-
creted it on the sea floor, most of the particles were deposited at a
level a little above the original position. The same should apply to
coarse grains deriving from glacial layers and scattered through the
foraminiferal marl above. The amhors mentioned attribute this ac-
tivity especially to holothurians and further to echinoids, annelids,
and ophiurids.
This explanation is tht: most logical one that has yet been suggested.
But, as pointed out above (p. 369), the general- not to say universal-
occurrence of clear-cut stratification in deep-sea deposits could not
be accounted for if the action of burrowing organisms were of gen-
eral importance. Neither is there any apparent reason why they
should be more abundant where ash deposits occur. Hence confirma-
tion of the explanation offered is needed before it can be accepted
without reserve.
Search in fossil and recent marine sediments for a yearly rhythm,
equivalent to the varves in glacial fresh-warer deposits, has nearly
always been made in vain. The only occurrence of marine varves
known to the author is described by Moore from the Clyde Sea
(1931). But they are caused by biological factors. The scarcity of
marine varves due to inorganic precipitation is attributed by Fraser
(1929) to the flocculation of clay in sea water, causing it to sink as
rapidly as fine sand. Moreover, in shallow marine environments dis-
turbance by waves and burrowing benthonic life may have destroyed
varves. Kuenen (1943) attributed the absence of a yearly rhythm in
374 FOR MAT JON 0 F MAR J N ESE D JMEN T S
RADIOACTIVITY 1
The average radium content of acid igneous rocks (granites) is of
the order of 2.5 X 10-12 gram per gram of rock, and of basalts 1.0 X
10 - 12 gram per gram, while sedimentary rocks contain less, from 0.2
to 0.8 X 10- 12 gram per gram. But the radioactivity of pelagic de-
posits is considerably greater. Although different investigators have
not arrived at the same values, they all agree that the ocean bottom is
strongly radioactive. Pig got found in 1933 an average of 6.5 X 10 - 12
gram Ra per gram for 28 samples. Red clay is richest, containing
9.5 X 10- 12 gram Ra per gram as the average of 13 samples. Pettersson
(1943) found l3 .5 X 10- 12 in 19 samples of red clay, 13.1 X 10- 12
gram Ra per gram in two radiolarian oozes and 3.3 X 10- 12 gram Ra
1 See especially_ Evans; Piggot and Urry.
RADIOACTIVITY 375
per gram in 7 samples of globigerina ooze. But he points out that
there is a wide range and that some samples of red clay fall well within
the range of acid igneous rocks. The outer layer of manganese
nodules is even more radioactive and attains values around 50 X 10- 12
with a maximum of 135 X 10- 12 gram Ra per gram. Evans and Kip
reported 2.5 X 10 - 12 gram Ra per gram in terrigenous muds. Five
terrigenous muds with and without volcanic ash from the Moluccan
area (analyzed by the latter workers) contained an average of 2.2 X
10- 12 gram Ra per gram, of which the sample from 10,050 m in the
Mindanao Trough showed the lowest value, 1.38 X 10- 12 gram Ra per
gram, although the depth of water was twice that of the other samples.
Pettersson also arrived at the conclusion that activity is independent
of depth.
We cannot enter here deeply into the problem of what causes the
strong activity of the sea floor, but a few general aspects may be
mentioned. In the first place, it is conceivable that the radioactive
elements are precipitated independently on the ocean floor directly
from the sea water, the sedimentary particles acting merely as dilutent.
Evidence against this conception is to be found in the activity of the
Moluccan muds, which are accumulating some 50 times as quickly as
oceanic pelagic deposits, but in which no dilution of the radioactive
matter appears to occur. Russell arrived at the same conclusion from
a study of fossil sediments, but he pointed out that with very swift
accumulation insufficient radioactive elements may be present to cause
high activity.
Studies by the same author and by Beers have shown that the bulk
of the radioactive elements are deposited from solution together with
organic compounds. Hence marine oil shales are strongly radioactive.
Recent work by Piggot and Urry has shown that in long samples
the activity decreases with depth in the deposit and soon attains nor-
mal low values for fossil sediments. Pettersson showed that the same
generalization applied to manganese nodules, the strong activity of
the surface quickly diminishing inwards. From this it follows t,hat
the abnormal activity of the superficial layer must be due to the
short-lived elements ionium and/ or radium. The final low values
are caused by a normal percentage of uranium, When the sediment
is freshly deposited it contains more of the strongly active but short-
lived daughter elements than are generated spontaneously by the par-
ent element uranium. Very little is known concerning the distribution
of other radioactive elements such as thorium and potassium, and they
will be left out of account in the following.
376 FOR MAT ION 0 F MAR I N ESE DIM E NT S
In the course of time the short-lived elements in excess of radioac-
tive equilibrium are lost by spontaneous disintegration until no more is
left than is in radioactive equilibrium with the uranium. Uranium
disintegrates so slowly that no appreciable reduction in activity takes
place in the few hundred thousands of years represented in the longest
samples of red clay yet examined. In some samples it proved to be
ionium; in others, radium that had been deposited in excess of radio-
active equilibrium.
Pettersson (1943) showed that the radium and ionium content of
sea water is far below radioactive equilibrium. The conclusion is war-
ranted that the excess of these elements in deep-sea sediment is due to
the sedimentary particles picking up certain amounts of radium and
ionium and carrying them to the sea floor. The ionium and radium
generated by the uranium dissolved in sea water is continuously being
adsorbed to these particles and carried to the sea floor, there to dis-
integrate relatively quickly.
In the next paragraph an account will be given of the manner in
which ages can be calculated from these data.
In complete accordance with the results jl1st mentioned, E vans
found that a Cretaceous deposit from Timor, that is generally held
to be a fossil red clay with manganese nodules, shows normal low
activity. If, as seems probable, this red clay was also strongly radio-
active at the time of deposition, the normal amount of uranium is all
that remained soon after.
RATE OF SEDIMENTATION
Several methods have been proposed for ascertammg the rate of
accumulation of recent marine sediments. Twenhofcl used the amount
of matter suspended in sea water, but as the rate of settling and the
degree of turbulence are unknown the results are more in the nature
of a shrewd guess than an actual calculation. Nevertheless he arrived
at figures comparable to those derived from other methods.
Theoretically sediment traps could be used to demonstrate the rate
of accumulation. This method has been followed with success in
Alpine lakes, but at sea the technical difficulties are formidable. More-
over, the conditions in a trap differ from those on the adjoining sea
floor, because benthonic life is hampered and because a trap will tend
to catch more of the suspended matter than a flat expanse without a
rim, while on the other hand particles rolling along the bottom are
excluded.
RATE OF SEDIMENTATION 377
An estimate has been based on the degree to which transoceanic
telegraph cables become covered by sediment in the course of years.
But a heavy cable sinks deeply into the soft deposit and gives no in-
dication of the amount of subsequent covering. No wonder the re-
sult arrived at of 250 cm per century of globigerina ooze is prepos-
terously large.
Schwinner (1936) estimated the yearly fall of meteoric dust at
500 tons. Using the number of cosmic nickel-iron and silicate spher-
Africa
FIG. 161. The post-glacial, glacial. and interglacial stratum of deep-sea deposits
cncounrercd in one of the Meteor traverses of the tropical Atlantic. (Data
from Schott, 1935.)
Globigcrina Diatom
Blue mud ooze ooze Red clay
{ Atlantic 36 24 17
Average thickness, Ocean
postglacial IDdiaD 12 11
Ocean
FIG. 162. Diagram illustrating sources of Post Algonkian sediments. 1-6, land
surface at beginning of Cambrian; 2, Tertiary sediments; 3, Mesozoic sediments;
4, Paleozoic sediments; 5, intrusive rocks; 7, Recent volcano; 8, .continental slope;
9, de'ep-sea sediments; 10, Recent volcano sunk by regional isostatic compensation
and pressure on deep-sea deposits.
Differences between columns VII and VIrr in percen t age of each constituent of the
source ma teri al.
t L arge di screpancy, explained in text.
0.980/0 Ti0 2 , while the continental rocks show a much lower per-
centage. For shale, sandstone, and limestone the respective amounts
are 0.65, 0.25, and 0.06 0/0 . The same applies to globigcrina ooze, with
only 0.290/0. On the other hand rcd clay and blue mud are relatively
rich in Ti0 2 , showing 0.98 and 1.27 0/0 respectively. A large amount
of blue mud must be assumed to compensate the continental sediments
and the globigerina ooze, while the amount of red clay is irrelevant.
The percentages of P20 5 in continental deposits are 0.17, 0.08, and
0.04, the source rocks containing 0.28 0/0. Blue mud and globigerina
ooze are also below the last figure with, respectively, 0.21 and 0.180/0.
From these figures a very large amount of red clay, that contains
0.300/0 of P 2 0 5 can be deduced as necessary to compensate the low
percentage in all other sediments.
The valucs for titanium and phosphorus are hardly sufficiently trust-
worthy to permit an exact calculation of the total amounts of sedi-
ment. But two facts emerge from the above analysis. First, the total
mass of deep-sea deposits must be several times as large as that of con-
tinental deposits, for all types of the latter are deficient in Ti0 2 and
P 2 0 5 and no comhination of these deposits call explain the geochemis-
try of titanium and phosphorus. Second, a welcome confirmation is
found of the rel iability of the geochemical computations. The source
and sediments of our table show a fair measure of agreement for the
major constituents for which the calculations were made. The two
remaining elements are likewise found to be in good agreement when
the various ty pes of sediment are combined in the same ratio.
Summarizing the outcome of the different and in part independent
mcthods for estimating the total mass of deep-sea deposits contained
in the earth's crust, the following result is obtained:
From the rate of recent accumulation 9 X loS km 8
From the ratio shale: sandstone: limes tone 8 ~ X loS km B
From denuda ti on and volcanism 12~ X loS km 3
From the sodi um content of the oceans 5 X loS km 3
From the geochemistry of Ti~ and P 206 Several times 2 X lOS km 3
z
~
'" z
0
"
a:
!:::
a:
...
0
I/)
Z
...
I-
z ~
u
0 '"::E z
a:
z is
~
.... '" S'"
I/) :I:
:::> ~
ID ~
ii
....
I/)
0
CLIMATE AND SEDIMENTATION 403
An example occurs in Whale Bay on the west coast of Africa. Twice
yearly an immensely rich plankton crop of dinoflagellates grows there
(Brongersma-Sanders, 1948). The e organisms secrete a poisonous
substance that kills off the fish fauna. Dead fish and plankton settle
on the bottom together and form a sapropele-like deposit rich in
hydrogen sulfide. Benthonic life is not able to colonize this area, and
only anaerobic bacteria are encountered. In consequence of this
peculiar cycle extensive deposits are formed on the sea floor with an
140
130
Caco, 0 r\ Clay a Depth II
120 1,\
110 ~
100
90
~
['..0
o I"
"
1\
~
"
r\.11
II\..
~
80
"1l ~
70
60
50 '-
\
\.
"
40
30 A B '\ C
o~~~~~~~~~~~~~~~~o
Per cent Per cent
Meters
FIG. 164. Relation between organic matter (expressed in 1/1000% nitrogen) and
various factors in the Moluccas. (After Kuencn, 1943, Fig. 3, p. 18.)
Bibliography
ANDaU, K. Sedimentbildung am Meeresboden, Geol. Rundschau, Vol. 3, pp.
324-338, 1912; Vol. 7, pp. 123- 170, 249-301, 329-337, 1916/17; Vol. 8, pp. 36-79,
1917.
Geologie des Meeresbodens, Bd. 2, 689 pp., Borntrager, Leipzig, 1920.
Das Meer und seine geologische Tatigkcit, in W. SALOMON, Grundziige der
Geologie, Bd . 1, p. 361-533, 1925.
ARCHANGUELSKY, A. D. On the Black Sea Sediments and Their Importance for
the Study of Sedimentary Rocks, Bull. Soc. Nat. Moscow, Sec. Geol., T . V.,
Nouv. Ser., T. 35, pp. 199-289, 1927.
BAILEY, E. B. Sedimentation in Relation to Tectonics, Bull. Geol. Soc. Am., Vol.
47, pp. 1713-1726, 1936.
BARRELL,]. Rhythms and the Measurement of Geological Time, Bull. Geol. Soc.
Am., Vol. 28, pp. 745-904, 1917.
Marine and Terrestrial Conglomerates, Bull. Geol. Soc. Am., Vol. 36, pp.
279-341, 1925.
BEERS, R . F. Radioactivity and Organic Content of some Paleozoic Shales, Bull.
Am. Assoc. Petro Geol., Vol. 29, pp. 1-22, 1945.
BoURCART,}. La sedimentation dans la Manche, in La geologie des terrains recems
dans l'ouest de l'Europe, Scss. Extr. Soc. Belges de Geol., September 1946,
pp. 14-43, 1947.
Sur les vases du plateau continental fran~ais, Compt. rend. acado sci., Vol. 225,
pp. 137- 139, 1947.
BouRCART, ]., and C. FRANClS-BoEUF. La vase, 67 pp., Hermann, Paris, 1942.
BRAJNIKOV, B., C . FRANCIS-BoEUF, and V. ROMANOVSKY. Techniques d'etude des
sediments, 110 pp., Hermann, Paris, 1943.
BRAMLETTE, M. N., and W. H. BRADLEY. Lithology and Geological Interpreta-
tions, in Geology and Biology of North Atlantic Deep-Sea Cores, U. S. Geol.
Surv. Prof. Pape, 196, 1940-1942.
410 FORMATION OF MARINE SEDIMENTS
BRONGERSMA-SANDERS. M. The Importance of Upwelling Water to Vertebrate
Paleontology and Oil Geology. Verh. Kon. Ned. Akad. Wet. Amsterdam,
AId. Nat . Sec. 2. Dl. 45. 4. 112 pp . 1948.
CAILLEUlC, A. La disposition individuelle des galets dans les formations detritiques.
Rev. geogr. phys. et geol. dyn., Vol. 11. pp. 171-196. 1938.
CLARKE. F. W. The Data of Geochemistry. Bull. U. S. Oeol. Surv. 770. 841 pp .
1924.
CoRRENS. C. W. Leitminerale im Golf von Guinea. Nachr. Akad. Wiss. Oottingen,
Math.-Phys. Kl . pp. 1-3. 1947.
CoRRENS. C. W. In T. F. W. B... RTH. C. W . CORRENS. and P. ESKOL .... Die Entsteh-
ung der Gesteine, 422 pp . Springer. Berlin. 1939.
CoRRENS. C. W . and \V. SCIIOTT. Die Sedimente des aquatorialen Atlantischen
Ozeans. Meteor, Bd. 3. T. 3. Lf. 1. 2. 298 pp . de Gruyter. Berlin-Leipzig.
1935 and 1937.
DIETZ. R . S. Clay Minerals in Recent Marine Sediments. Am. Min., Vol. 27.
pp. 219-220. 1942.
DIETZ. R. S. K. O . EMERY. and F. P. SHEP ... RD. Phosphorite Deposits on the Sea
Floor off Southern California, Bull. Geol. Soc . Am., Vol. 53. pp. 815- 848. 1942.
DIRCKSEN. R. Dar Wattemneer, 220 pp . Bruckmann. Munich. 1942.
DOEGLAS. D. J. Interpretation of the Results of Mechanical Analyses. ,. Sed.
Petr., Vol. 16. pp. 19-40. 1946.
EHRHARDT. A. Dar Watt, 96 plates. Ellerman. Hamburg. 1937.
EMERY. K. 0 . and R. S. DIETZ. Gravity Coring Instrument and Mechanics of
Sediment Coring. Bull. Geol. Soc. Am., Vol. 52. pp. 1685- 1714. 1941.
EMERY. K. 0 . and F. P. SHEP"'RD. Lithology of the Sea Floor off Southern Cali-
fornia. Bull. Geol. Soc. Am., Vol. 36. pp. 431-478. 1945.
ESCfIR. B. G. Beschouwingcn ovcr het Opvullings-Mechanisme van Diepzee-
Slenken. Verh. Geol. Mijnb . Gen. Nederl., Vol. 3. pp. 79- 88. 1916.
EVANS. R. D., and A. F. Kn'. Radium Content of Marine Sediments from the East
Indies, etc . Am. ,. Sci., Vol. 36. pp. 321-336. 1938.
FRANCIS-BoEUF. C. Recherches sur Ie milieu Ruvio-marin et les dep~ts destuaire.
Ann. Inst. Ocean., Vol. 23. pp. 149-344. 1947.
FRASER, H. J. An Experimental Study of Varve Deposition. Proc. Trans. Royal
Soc. Canada, Vol. 23. Section IV. pp. 49-60, 1929.
a
GLANGEAUD, L. Transport et sedimentation dans l'estuaire et I'embouchure de la
Gironde, Bull. soc. geol. France, 5e Sec., T. 8. pp . 599-631, 1938.
Sur la formation et la repartition des facies vaseux dans les estuaires, Compt.
rend. acado sci., T. 213. pp. 1022-1024. 1941.
GRABAU. A. W. The Rhythm of the Ages, 561 pp., Vetch. Peiping. 1940.
RANTZSCHEL, W. Die Schichtungsformen recenter Flachmeer-Ablagerungen im
Jade Gebiet. Se1lckenbergiana, Bd. 18, pp. 316-356. 1936.
HERDMAN. W. A . et al. The Marine Zoology, Botany. and Geology of the Irish
Sea. Report British Assoc. Adv. Sci., pp. 455-467. 1895.
HEss, H. H. Drowned Ancient Islands of the Pacific Basin, Am. ,. Sci., Vol. 244,
pp. 772-791, 1946.
Hn.Ls, G . F. S. The Formation of the Continents by Convection, 102 pp., Arnold,
London, 1947.
BIBLIOGRAPHY 411
JOHNSTON, J., and P. D. Wn.L1AMSON. The Role of Inorganic Agencies in the
Deposition of Calcium Carbonate, ]. Geol., Vol. 24, pp. 729-750, 1916.
KVENEN, PH. H. On the Total Amount of Sedimentation in the Deep Sea, Am. }.
Sci., Vol. 34, pp. 457--468, 1937.
The Cause of Coarse Deposits at the Outer Edge of the Shelf, Geol. en Mifnb.,
pp. 36-39, 1939.
Geochemical Calculations concerning the Total Mass of Sediments in the
Earth, Am. ]. Sci., pp. 161-190, 1941.
Rate and Mass of Deep-Sea Sedimentation, Am. ]. Sci., Vol. 244, pp. 563-572,
1946.
De Kringloop van het Water, 408 pp., Leopold, The Hague, 1948.
KVENEN, PH. H., and G. A. NEEB. Bottom Samples, The Snellius Expedition,
Vol. 5, part 3, 268 pp., Brill, Leyden, 1943.
KVENE.N, PH. H., and I. M. v. D. VLERK. Geheimschrift der Aarde, 370 pp., de
Haan, Utrecht, 1948.
KVENEN, PH. H., and C. I. MIGLIORINI. Turbidity Currents as a Cause of Graded
Bedding (in press).
LYELL, C. Principles of Geology, Vol. II, 649 pp., 10th ed., 1868.
MARR, J. E. Deposition of the Sedimentary Rocks, 245 pp., Cambri4ge University
Press, 1947.
MARSHALL, P. The Wearing of Beach Gravel, Trans. Proc. New Zealand Inst .,
Vol. 58, pp. 507-532, 1928.
MOORE, H. B. The Muds of the Clyde Sea Area. III Chemical and Physical Con~
ditions; Rate and Nature of Sedimentation, and Fauna, }. Marine Bioi. Assoc.
V. K., Vol. 17, pp. 325-358, 1931.
MURRAY, J., and J. HJORT. The Depths of the Ocea'll, 821 pp., Macmillan, London,
1912.
MURRAY, J., and E. PHILIPPI. Die Grundprobcn der "Deutschen Tiefsee-Expedi-
tion," Wiss. Ergebn. Valdivia, Bd. 10, pp. 77-206, 1908.
MURRAY, J., and A. F. RENARD. Deep-Sea Deposits, Challenger Reports, 525 pp.,
Longmans, London, 1898.
MYERS, E. H. Rate at Which Foraminifera are Contributed to Marine Sediments,
]. Sed. Petr., Vol. 12, pp. 92-95, 1942.
NATLANll, M. The Temperature and Depth Distribution of Some Recent and
Fossil Foraminifera in the Southern California Region, Scripps Inst., Tech.
Ser. Bull., Vol. 3, pp. 225-230, 1933.
NEEB, G. A. The Composition and Distribution of the Samples. See Kuenen and
Neeb, 1943.
NEVIN, C. Competency of Moving Water to Transport Debris, 8uJl. Geol. Soc.
Am., Vol. 57, pp. 651-674, 1946.
PETTERSSON, H . Manganese Nodules and the Chronology of the Ocean Floor,
Goteborgs K . Vet. Vito Samh. Hand. Sjiitte Foljden, Ser. B, Bd. 2, No.8,
pp. 1--43, 1943.
Iron and Manganese on the Ocean Floor, ibid., Bd. 3, No.8, pp. 1-37, 1945.
Three Sediment Cores from the Tyrrhenian Sea, ibid., Bd. 5, No. 13, pp. 1-94,
1948.
PET'l1JOHN, F. J . Sedimentary Rocks, 526 pp., Harpers, New York, 1949.
412 F O RMAT I ON OF MARINE SEDIMEN T S
PHILIPPI, E. Die Grundproben der deutschen Stidpolar-Expedition, 1901..{)3, II,
Geographie und Geologie, Bd. 6, pp. 411-616, 1910.
PHLEGER, F. B. Foraminifera of Three Submarine Cores from the Tyrrhenian Sea,
Goteborgs K. Vet. Vito Samh. Hand. Sjatte Foljden, Ser. B, Bd. 5, No.5,
pp. 1-19, 1947.
Foraminifera of a Submarine Core from the Caribbean Sea, ibid., Bd. 5, No. 14,
pp. 1- 9, 1948.
PHLEGER, F . B., and H . PETl'ERlSSON. Foraminifera of Three Submarine Cores from
the Tyrrhenian Sea, ibid., Bd. 5, No.5, pp. 1-19, 1947.
PIA, J. Die rezenten Kalksteine, Zeitschr. f. Krist., Min., Petrogr., Abt. B, Erg.
Band, 420 pp., 1933.
PICGOT, C. S. Core Samples of the Ocean Bottom and Their Signilicance, Sci.
Monthly, Vol. 46, pp. 201-217, 1938.
Factors Involved in Submarine Core Sampling, Bull. Geol. Soc. Am., Vol. 52,
pp. 1513- 1524, 1941.
PlGGOT, C. S., and W. M. D. VRRY. Time Relations in Ocean Sediments, Bull.
Geol. Soc. Am., Vol. 53, pp. 1187-1210, 1942.
PRATJ, O. Die Sedimente des siidatlantischen Ozeans, Meteor, Bd. 3, T . 2, Lf. I,
56 pp., 1935.
Die Ausdeutbarkeit der Sedimente, Geol. Rundschau, Bd. 29, pp. 168-174, 1938.
REVELLE, R . R. Marine Bottom Samples Collected in the Pacific Ocean by the
Carnegie on its Seventh Cruise, Carnegie Inst. Wash ., Pub. 556, part I, 180 pp.,
1944.
RUSSELL, I. C. Second Expedition to Mount Saint Elias, in 1891 , 13th Ann. R ept.
U. S. Geol. SUT'll., pp. 7-91 (see pp. 24-26), 1893.
RUSSELL, R. J. Coast of Louisiana, Bull. Soc. Beige Geor., Vol. 57, pp. 380-394,
,1948.
RUSSELL, R. J., et al. Lower Mississippi River Delta, Louis. Geol. Surv., Geol.
Bull., Vol. 8, 454 pp., 1936.
RUSSELL, W. L. Relation of Radioactivity, Organic Content, and Sedimentation,
Bull. Am. Assoc. Petro Geol., Vol. 29, pp. 1470- 1494, 1945.
RUTTEN, M. G. Actualism in Epirogenetic Oceans, Geol. en Mijnb., Vol. 11, pp.
222- 228, 1949.
ScHOTT, G. Geographie des Indischen und Stillen Ouam, 413 pp., Boysen, Ham-
burg, 1935.
Geographie des Atlantischen Ozeans, 438 pp., Boysen, Hamburg, 1942.
ScHOTT, W. Die Foraminiferen in dem aequatorialen Teil des Atlantischen Ozeans,
Meteor, Bd. III, T. 3, Lf. I, B, pp. 43-134, 1935.
Rate of Sedimentation of Recent Deep-Sea Sediments, in Trask et aI., pp. 409-
415, 1939.
ScHUCHERT, C. Geochronology, or the Age of the Earth on the Basis of Sediments
and Life. Bull. Nat. Research Council, No. 80, pp. 10-64, 1931.
ScHW1NNER, R. Lehrbuch der physikaliscben Geologie, Bd. 1, 356 pp., 1936.
Sediment-Heft, Geol. Rundscbau, Bd. 29, pp. 145-461, 1938.
SHEPARD, F. P., and K. O. EMERY. Submarine Topography off the California
Coast, Geol. Soc. Am., Spec. Paper 3 I, 171 pp., 1941.
SHROCK, R. R. Sequence in Layered Rocks, 507 pp., McGraw-Hill, New York,
1948.
SITTER, L. U. DE. F~cies analyse, Geol. en. Mijnb., Vol. 3, pp. 225- 237, 1941.
BI BLIO G RAPHY 413
SVERDRUP, H . U., M . W. JOHNSON, and R. H. FLEMING. The Oceans, 1087 pp.,
Prentice-Hall, New York, 1942.
TERCIER, J. Depots marins actuels et series geologiques, EcJog. Geol. Helv., Vol.
32, pp. 47-100, 1939.
TRASK, P. D. Origin and Environment of Source Sediments of Petroleum, 323 pp.,
Gulf, Houston, Texas, 19n.
Relation of Salinity to the Calcium Carbonate Content of Marine Sediment,
U. S. Geol. Surv. Prof. Paper 186-N, pp. 273-299, 1937.
TRASK, P . D ., and H. W. PATNODE. SouTce Beds of Petroleum, 566 pp., Am. Assoc.
Petr. Geo!., Tulsa, 1942.
TRASK, P. D ., et al. Recent Marine Sediments, A Symposium, 736 pp., Am. Assoc.
Perr. Geol., Tulsa, 1939.
TWENHOFEL, W. H . Treatise on Sedimentation, 926 pp., Williams & Wilkins.
Baltimore, 1932.
Principles of Sedimentation, <510 pp., McGraw-Hili, 1932.
UMIlGRO\,E. ]. H. F. Periodal Events in the North Sea Basin, Geol. Mag., Vol. 82,
pp. 237-244. 1945.
WANLESS, H. R., and F. P. SHEt>ARD. Sea Level and Climatic Changes Related to
Late Paleozoic Cycles, Bull. Geol. Soc. Am., Vol. 47. PI'. 11 77-1206, 2008-2014,
1936.
WANLESS, H. R., and ] . M. WELLER. Correlation and Extent of Pennsylvanian
Cyclothems, Bull. Geol. Soc. Am., Vol. 43, pp. 1003- 1016, 1932.
WEIBULL, W. The Thickness of O cean Sediments Measured by a Reflection
Method, Medd. Ocean. Inst. G oteborg, Vol. 12, 1'1" 2- 17, 1947.
WISEMAN, J. D . H., and H. BENNETT. The Distribution of Organic Carbon and
Nitrogen in Sediments from the Arabian Sea. l ohn Murray Expedition,
Vol. 3, 4, pp. 193- 221, 1940.
WOOLNOUGH, "If. G. Geological Extrapolation and Pseud-Abyssal Sediments,
Bull. Am. Assoc. Petro Geol., Vol. 26, pp. 765-792, 1942.
Zo BELL. C. B. Marine Microbiology, 240 pp . Waltham. Mass., 1946.
c H A p T E R 5 x
Coral Reefs
I '
!l 10
FIG. 16S. Semi-diagrammatic view of half a simple coral polyp in extended con-
dition. The calcareous skeleton is shown with a dark line. (After Woods, 1947,
Fig. 26, p. 80, Cambridge Universiry Press) .
Depth, m
29-36 37-45 46-54 55-63 64-72 73- 90
Number of dredgi ng, 29 65 55 33 31 23
Number of species 36 61 22 15 15 1
Nu mber of genera 16 19 13 8 8
r:;:::::;::] Coral fragments, branching and massive colonies ~ Mud with andesite detritus
o Corallragments, branching colonies 0 Terrestrial deposits
~ Mud containing coral debris [::::::::::1 Andes:~ e
FIG. 166. Section through the fringing reef of Emmahaven, Sumatra, based on
borings. Illustrating growth of a reef on a muddy substratum, probably by out-
growth on scree. Vertical rwice horizontal scale. (Modified after Umbgrovc,
1931, Plate 4.)
'f
Blow hole
Closed
opening
c:
0
A B 'c
~
..
E
(5
-5
:.:J
~
:.:J
FIG. 167. Lithothamnion edge of Funafuti Atoll, showing buttresses and trenches.
(Redrawn after Funafuti Report, 1904, Plates 17 and 19.)
isms and help to fill in the crevasses between fixed animals and plants.
On the other hand fishes and boring organisms cause damage to the
reef structure and living corals, while holothurians and other mud-
feeders assist the mechanical wear of detritus, especially in the lagoons.
A B
'----'
5km
3 '----' 5 km 4 '----' 5 km 5 6
i1
FIG. 168. Seveh barrier reefs (depths in meters). A, Tahaa, Society Islands,
typical barrier reef with strongly em bayed island and fringing reef. B, Fiji
Islands: (1) Wakaya, strongly asymmetrical barrier reef with wide lagoon, pass-
ing into fringing reef; (2) Naitomba; (3 ) Budd Reef, an almost-atoll with in-
complete barrier, deep lagoon, and half a dozen lagoon reefs; (4) Mango, narrow
lagoon, partly fringing reef; (5) Kanathea, asymmetrical barrier, partly fringing
reef; (6) Vanua Vatu, barrier joined to island by several patches of reef.
many cases the reefs curve inwards and jut out into the lagoon for a
few hundred meters at both sides of a pass. In the lagoon may be
found smaller or larger numbers of reef patches, generally rising to
the surface, while the coast is commonly bordered by a fringing reef.
Atolls are reef rings around a lagoon in which there are no islands
of non-coral origin (Fig. 169). In other respects they are similar to
barrier reefs. The average depth of the lagoons is 45 m, according to
Shepard.
FIG. 169. Three atolls (depths in meters). A, Nilandu Atoll, Maldive Archi-
pelago. Rim consists of faros (depth 20-30 m) . Many lagoon reefs, correlated
with large number of passes in the rim. B, Etal Atoll, Caroline Islands. Rim
closed, no lagoon reefs shown on charco C, Ngatik Atoll, Caroline Islands. Rim
with onl y one passage showing horns, few lagoon reefs.
FIG.170. Plan and profile of reef between Eniwetok and Porky islands, Marshall
Group. (After Stearns, 1945, Bull. Geol. Soc. Am., Vol. 56, Fig. 3, p. 787.)
!
iI / 1
,
\
,
\
,I
I
I
!
\
i
\
right angles to the reef edge on some flats. More than one rampart
may be found running parallel (Fig. 172).
An interesting record of changes which shingle ramparts have un-
dergone in 17 years on Low Isles, Great Barrier Reefs, is given by
Fairbridge and Teichert (1947).
Elevated Islands. Elevated islands are found on most reef flats
throughout the coral seas. They rise to a few meters above high-
tide level and consist either of sand and debris or of solid reef rock.
430 CORAL REEFS
They have evidently been formed during a recent relatively higher
stand of sea level. Erosion is nearly always occurring around parr or
the whole circumference, thus testifying to the ephemeral nature of
these elevated islands.
------------
N
o Sand
~ Fine shingle
~ Coral blocks
-~
'f I.o..ll 1Mangrove ~-
I'
,.
I'
Pelo
I
mJ
tD
Co.". shlngl.
Cemented sand and shinjle
1"" "" B,
FIG. 175. Sketch map of two coral islands on reef fiat west of Morotai, Moluccas,
showing elevated reef rock, cay sandstone, and dipping beach sandstone where
island has been eroded. (After Kuenen, 1933, Fig. 49, p. 36.)
Cay
sandstone
o 100 200
Meters
FIC. 177. Sketch map of Jarngoer Roeal, a sand cay near Jamdena, Moluccas,
showing beach sandstone and cay sandstone. (After Kucnen, 1933, Fig. 65, p. 50.)
West East
FIG. 178. Section through Jarngoer Roeal, showing dipping beach sandstone
and horizontal cay sandstone. (Redrawn after Kuenen, 1933, Fig. 66, p. 50.)
just mentioned, lie about 5 m above sea level and were probably de-
veloped by a world-wide sinking of ocean level some 3000 to 4000
years ago. In part the terraces had already been removed by solution
and to a much smaller extent the surrounding reefs had grown out
to seaward. The resultant reef flat is 200 m broad on the former and
75 on the latter island. From this it follows that about 0.2 mS of rock
per meter length of coast has been removed per annum. Further ob-
r"" I
........... ll
""'III
FIG. 179. Sections through the elevated rim of Maratua Atoll. Vertical 5 times
horizontal scale, g' = natural scale. I, living reef; II, sand; III, reef rock. D =
site of borings. (After Kuenen, 1947, Fig. 2, p. 9.)
servations are needed before this figure can be relied upon and the
variations in magnitude under different conditions evaluated.
I ! I ! I ! I
o 2 4 6
Kilometers
FIG. 180. Cocos Keeling, an atoll showing the effect of the prevalent winds.
(Modified after Wood Jones, 1910, Fig. 62, p. 260.)
Sand cays may be small and barren, and then they tend to accom-
modate their shape to the varying directions of the wind. A crescent-
shaped form resembling that of barchans may be assumed under the
regime of strongly prevailing winds. Alternating monsoon winds. are
able to force the crescent into the opposite orientation twice yearly.
On the reefs of oceanic atolls the spits joining to cays are generally
bent backwards toward the lagoon and may even curve round to meet,
thus enclosing a shallow pond. Islands covered by vegetation are,
of course, less mobile, but barren spits are often developed that alter
in position and shape with the time of year. It takes some time for
the waves to shift the sand, and a change of direction for a few days
will not influence a cay to any appreciable extent.
442 CORAL REEFS
Sand cays are strongly developed in regions that are somewhat pro-
tected from the full force of oceanic swell and violent storms. Indo-
nesia and the West Indies and the lagoon reefs of the Great Barrier
Reefs are the classical regions of cays. On more exposed reefs most
islands owe their existence to a relative fall of sea level. They consist
of elevated reef rock with added accumulations of coarse detritus and
some sandy material.
The action of wind and waves tends to heap up sand on the lagoon
side of exposed atoll reefs in a high ridge on which sand dunes may
develop. A remarkable feature of many islands in the Indian Ocean
is the occurrence of vast accumulations of pumice, which may be
attributed to the Krakatau eruption of 1883. Using this datum, aver-
age subsequent growth to an amount of several decimeters per year
could be ascertained in some cays.
Cays are sometimes found to migrate across the reef flat in the
direction of the atoll lagoon. The fact that this process is at present
in full swing strongly indicates that the islands owe their formation
to a comparatively recent alteration of conditions. It will be shown
later that a recent world-wide sinking of ocean level, which probably
occurred three or four thousand years ago, is responsible for the emer-
gence of the reef flats and initiated the development of the cays. The
co,nsequences have not yet developed to a stable condition.
On Coorser Detritus. The influence of the wind on the accumulation
of shingle and boulders is no less evident than in the case of sand but
different laws can be established. The larger weight and rough inter-
locking shape render the shingle less mobile than the sand. Hence, the
coarser material tends to accumulate closer to the edge of the reef
flat, where it is directly supplied by the living reef. Once it has been
thrown up in ridges to the height of a few decimeters to 1 or 2 m
these accumulations are relatively immobile.
Nevertheless these shingle ramparts are not permanent structures.
Several cases have been described in which the ramparts migrate
slowly away from the reef edge and are forced across the fiat. Partly
buried mangrove trees may testify to this movement. In other cases
a new rampart is in course of formation closer to the edge, and half-
buried remnants of the original rampart, consisting of blackened
shingle, emerge below its inner slope.
The conclusion is warranted that the ramparts are destroyed in the
course of time. Scattering of the shingle, solution and disintegration
of the coral fOlgments, will gradually result in the disappearance of
WIND AND CORAL ISLANDS 443
an old rampart and the formation of sand and silt that is swept away.
Under favorable conditions a new rampart is then built up at the sea-
ward edge of the reef flat.
Umbgrove showed that the position of the shingle ramparts is
closely related to the direction of the prevailing winds. In the Bay
of Batavia (Fig. 181) the ramparts are restricted to the side of the
reefs, whence the prevailing and strongest winds also blow. On the
Duizend Eilanden in the Java Sea the influence of both monsoons is
NW
N
- ~
~~D
PURMERENO
fi)
W E tIfJ
SW . tP
S
~ Coral sand IIIID Coral shingle
FIG. 181. The reef island of Purmerend in the Bay of Batavia, showing the dis-
tribution of sand and shingle under influence of prevailing winds. Scale 1: 12,000.
(After Umbgrove and Verwey, 1929, Fig. 1, p. 9.)
clearly visible, for here this investigator met with ramparts at opposite
sides of the reef flats. Finally, in the Gulf of Tomini, Celebes, he
found ramparts to be entirely absent in the remarkably calm waters
of this bay.
Similar observations have been made elsewhere in Indonesia, the
Great Barrier Reefs, and the West Indies (Steers and others). On
oceanic reefs the ramparts are generally found to consist of large
boulders or to be entirely absent, owing to excessive force of the
waves.
Here and there the shingle accumulates in tongues at right angles
to the edge of the reef flat. These ridges are low and flat but may
attain considerable length. Probably they are formed more especially
on reef flats that lie relatively low, allowing somewhat more scope to
waves running across the surface.
On most reef flats, where one of the two types of debris (sand or
shingle) accumulations is found the other is also present. The dis-
tance between the shingle and sand may be large, and they may lie at
444 CO RA L REE FS
opposite sides of the flat. Elsewhere the two types are grown to-
gether, but generally a shallow lagoon with tidal waters lies between,
either entirely enclosed or communicating with open waters.
Mangroves on coral reefs are generally restricted to the reef flats
in little-exposed positions. They grow either in a narrow belt along
the coast of the cay or in dense forests behind the rampart. In the
Boo Islands east of Halmahera a sandy ridge close to the edge of the
reef encircles an extensive mangrove swainp on several of the reefs.
It would lead us too far to describe the various shapes of coral
islands. The reader is referred to the reports by the authors men-
tioned and the account and beautiful maps given by Spender of a
few Australian islands.
FIG. 182. Chart of the Sibutu Archipelago. north of Borneo. a group of oblong
atolls grown on a submarine ridge under the influence of currents between the
Sulu and Celebes seas. Slight tilting of the foundation has elevated Sibutu
Island and partly drowned the large atoll to the west of Tumindao Atoll. (After
Kuenen, 1933, Fig. 38, p. 26.)
FIG. 183 . Chart of the central pan of the Thousand Islands in the Java Sea,
showing deep swales between the reefs caused by currents. (Redrawn after
Umbgrove, 1929, Plate I.)
Flc. 184. Boulder ramparts on Pclokan, Sapuka Atoll, south of Celebes. The
section is viewed from the north. (After Kuenen, 1941, Fig. 24, p. 118.)
the strand line. The sand cays and shingle ramparts lying on the
surface of the flat must also have developed after the relative emer-
gence. It would even appear that practically no cays and ramparts
could have been formed hut for the preceding emergence. This is
beyond doubt true of all islands formed of reef rock that abound on
the atolls and barriers of the coral seas. It is almost universally ob-
served that all islands on a reef are being attacked by erosion and
doomed to be devoured by the waves. It has already been men-
tioned that the islands on the windward side of the Maldive and Lacca-
dive archipelagoes have disappeared. Although some reef islands are
being added to on the lagoon side, the coasts are rapidly receding on
the outer side. The resultant migration will probably end with their
being pushed back across the reef into the depths of the lagoon. Even
CORAL REEFS
in the calm waters of the Moluccan seas coral islands are almost
always found to be on the wane.
It was Gardiner who especially emphasized the importance to the
development of coral islands of a relative rise of the reefs throughout
the coral seas. The fact that they are often larger and closer together
on the leeward side of atolls is a strong argument against their being
due only to the action of waves and in favor of the opposite conten-
tion that waves may be temporarily forced to build islands on an
emerged flat but will ultimately succeed, together with solution, in
destroying all or the vast majority of resultant islands.
It is of interest to note that Darwin observed the destruction going
on among the coral islands and attributed it to subsidence. This
appeared to argue in favor of his theory of atoll and barrier formation.
If he had pondered the question how these islands could first have
been formed, he would soon have come to the conclusion that only
a relative emergence could have caused the former period of con-
struction. The exact opposite to what Darwin attempted to prove
has taken place.
Daly later drew attention to the large number of coasts on which a
subrecent emergence of nearly uniform amounts has left its mark in
the shape of elevated notches and erosion terraces. He attributed them
all to a recent world-wide sinking of ocean level by 5 or 6 m. Since
then evidence has been added from many more localities, especially
from practically all the reef groups investigated. Daly 's postulate has
thus been confirmed and may be accepted as practically established,
especially when the less obvious but no less pertinent testimony of all
reef flats that'lie above the growth limit of reefs is added. It should
be admitted, however, that subsidence of the land leaves much less
evidence, and it might be claimed that an equal length of coast may
have been drowned. In that event sea level should have remained
stable and deformation of the crust would have caused the deleveling
noted. This contention is unreasonable, however, in view of the fact
that all the emerged beaches are found at roughly the same altitude
above present sea level.
As soon as an attempt is made, however, to ascertain the exact
amount of the depression of sea level, difficulties are encountered in
spite of the fact that coral reefs are the best geological tide gauges
available. The advantages offered by reefs for studying former stands
of sea level are due to the following properties: (I) the solid reef
cannot be built up above average sea level; (2) the reef can develop
into a hard, coh~rent mass; (3) sharply chiseled notches are cut in
SEA LEVEL AND CORAL REEFS 449
elevated reef rock even in protected positions; (4) the reefs tend to
grow up to low tide level; (5) erosion of emerged reefs generally ends
when low water level is reached ; (6) sand cays are built up to a
small height above high tide and many of them possess a flat surface.
Uncertainties are introduced when the range of the tides is larg ,
or where strong onshore winds pile up the water against the coast.
_--- ~
""
.. .
"..,.,., "", """,
FIG. 185. De\'elopmcnr of the isla nds in the Bay of Batav ia by lowering of sea
level in three successive stages, according to Kuenen. (After Kuenen, 1933,
Fig. 85, p. 72.)
I !
o 10 20
Kilometers
Flc. 186. The dro wned barrier of Waigeo in the Moluccas. The lagoon floor
is 170 m below the crest of the barrier. (After Kuenen, 1933, Fig. 100, p. 107.)
reduce an entire atoll to the lowered sea level in the time available
during an ice age. But without further confirmation this conclusion
is still highly speculative.
Examples of elevated atolls have been described among others from
the Lau Archipelago. But Ladd and Hoffmeister showed that the
limestone is generally of Tertiary age and consists of coral rock for
a minor portion only. The basin-shaped surface appears to be due
mainly to denudation, whereby the central hollow is formed by solu-
tion. Denudation processes are also responsible for the barrier-shaped
rampart of coral limestone in Kisar northeast of Timor and the so-
called Makatea of several Pacific islands (Kuenen, 1933).
rhe sea is only a few dozen meters, local eminences as flat as a billiard
rabie, and several dozen kilometers across, must be postulated on the
anticJines-a fact neglected by Stearns' text. His diagrams indicate
outward growth on screes as one of the processes involved. One can
hardly believe that there was sufficient time for so slow a proces~ .
Besides, the greatly exaggerated vertical scale of these diagrams tends
to obscure the abnormal shape of the anticlines postulated by Stearns,
especially for atolls first formed on somewhat deeper anticJines during
low levels of the Pleistocene.
The above discussion shows that atolls and barriers cannot be ex-
plained in the manner proposed by Ladd and Hoffmeister. It is
admitted, however, that, if the postulated stratum of fine sediment is
omitted, there may have been a few exceptional cases in which bar-
riers or atOlls with shallow lagoon, slight outer slopes, and cliffed
islands formed on pre-existing platforms without intervention of sub-
sidence after colonization started. But the overwhelming majority
show features that cannot be thus explained and call for subsidence as
the dominant factor in development. There may be antecedent plat-
forms at great depth below the coral formations, and there are very
probably platforms at glacial sea level cutting across most coral for-
mations. To this extent the importance of antecedent platforms may
be readily admitted. But this should not be allowed to obscure the
evidence for subsidence that is so clear and so widespread.
Foremost among the coral-reef theories postulating a relative rise
of sea level is the subsidence theory of Darwin. Darwin started from
a fringing reef that grows upward during slo'" subsidence of the
foundation (Fig. 18 7) . As the corals tend to grow vertically upwards,
and owing to the more favorable conditions on the outward side of
the reefs, the periphery is gradually raised while a lagoon forms
between it and the central island. Thus a barrier is developed w hich
mounts until the island disappears, leaving an atoll. In this manner
Darwin was able to arrange the major reef forms in a closely related
series. The logical simplicity of this theory forms a strong attrac-
tion, and soon after it was propou'nded Dana added an important
observation in confirmation. Nearly all the islands surrounded by
barriers show deeply embayed shorelines, testifying to a submergence
of a few hundred meters at least. Borings on Oahu, Hawaii, for
instance, have since indicated a subsidence of at least 300 m.
Further arguments in favor of the subsidence theory can be enu-
merated. In the Moluccas a number of lagoons of abnormal depth
amounting to 150 or even 200 m have been found. These call for
456 CORAL REEFS
subsidence to an amount of 150 m. The case of Waigeo may here
be recalled (see p. 451).
Many atolls and barriers show abnormally steep submarine slopes.
Down to depths of 200 m they sometimes approach the vertical, and
declivities of 50 reaching to a depth of 600 m have been recorded.
As the slopes around volcanic islands seldom exceed 35 the only
logical explanation of the steep foundations is by vertical upgrowth
Flc. 187. Block diagram showing the development of barrier and atoll reefs from
a fringing reef on a subsiding island according to Darwin's theory.
..0
N
ci..
~
.!?P
t...
~
~
0-
.,cC
.,
~ '"
.,...
~
. .;
U
<I>
"Ec
0
N
c
0
..c
.,
<I>
E
'0
.'e.
N
:>
ci..
...0'"
C-'
]
..
-a
b.O
C
f-;
.
-
.....0
~b.O
:e
..
..:.:u
0
iii
00
~
..;
ri:
458 CORAL REEFS
and fine calcareous lagoon deposits. The last contrast strongly with
the coarse terrigenous sand found on the surrounding insular terrace
of Borneo and consist mainly of fine, chemically precipitated calcite
needles identical with some recent lagoon sediments. Evidently this
lime mud was deposited in the protected, badly ventilated waters of
a deep lagoon. The bore hole reached to 550 m below the crest of
the atoll rim and brings unequivocal evidence of subsidence to an
FIG. 189. Natural slope of elevated islands with fringing reefs and of atolls and
reefs in the Tukangbesi Group southeast of Celebes. (After Kuenen, 1933,
Fig. 88, p. 94.)
existing reefs and thus deprived islands and continental coasts of the
equatorial belt of their protecting fringe of reefs. This result was
strengthened by the churning up of vast quantities of mud that
formerl y lay beyond the reach of waves but could now be activated
by the lowering of sea level. The turbidity caused further havoc
among coral life. Erosion set in, and pre-existing banks were beveled
off at glacial low level. Thus a vast number of foundations were
smoothed and cleansed and prepared for colonization.
In later writings Daly laid more stress on the "sandpapering" of
submerged banks and island shelfs formed by marine action on ancient
volcanic islands. The adverse conditions of the ice age should then
have prevented the establishment of corals during the cold periods,
thus allowing the waves to smooth the platforms.
As soon as a milder climate returned coral larvae began to colonize
these platforms. They came from the more favorable localities, the
Moluccas for instance, where corals had hibernated during the cold
spells. The conditions most favorable to coral life were offered by
the outer edges of the platforms, where food and oxygen were most
abundant and the water of special cleanness. Hence, corals here soon
gained the upper hand of those established farther inwards. They
460 CORAL REEFS
were able to grow upward after the gradual return of sea level to its
normal station. In this manner atolls and barriers became rooted at
the periphery of the platforms that were gradually submerged to a
uniform depth of about 80 m, much deeper than the limits of coral
growth. Aggradation of the lagoon floors took place. Owing to the
smaller area within atolls of lesser extent as compared to the circum-
ference, the accumulation was greater. This explains the observed
fact that, the smaller the lagoon, the less the depth at which the floor
tends to lie.
The theory of glacial control evidently gives a satisfactory expla-
nation of the extreme scarcity of atoll and barrier lagoons exceeding
80 to 90 m in depth and the relatively small and uniform volume of
the encircling reefs, while the lesser depths of small lagoons is also
accounted for .
Davis was much impressed by the morphological evidence of
drowning encountered in barrier-reef islands. A wide experience
in exploration and an exhaustive study of charts led 'him to accept
the subsidence theory for explaining atolls and barriers, hut to intro-
duce an important modification. He showed that the straight shore-
line and plunging cliffs, absent on barrier islands in the central coral
seas, prevail on islands in the marginal belts towards the northern and
soutl~ern limits of coral growth. He concluded that the zone of
flourishing reefs was narrowed down by the cooling of surface waters
during the glacial low levels. In the marginal belts thus abandoned
low-level erosion succeeded in planing off preglacial reefs and in
cliffing central islands. In shorr he concedes the principles of glacial
control for these marginal belts, but not for the torrid central coral
seas. His explanation of the uniformity of lagoon depths is that they
represent base level for the waves. For w ider lagoons this level
should lie deeper on account of the larger waves raised. Any sedi-
ment introduced into a lagoon in excess of the ordained depth is
churned up and carried to deeper spots or swept out through the
leeward passes and scattered far and wide over the deep-sea floor.
Kuenen went yet a step further in combining the main elements of
the two theories (1933), later proposing the "glacially controlled sub-
sidence theory" (1947). His contention is that, although the argu-
ments in favor of subsidence are almost irrefutable and go far toward
proving the formation of the foundation of atolls and barriers by slow
sinking of the substratum, no one can deny that all Tertiary reef
structures must have suffered attack during glacial low levels (Fig.
191). Existing reefs therefore owe their present shape largely to
THEORIES OF ATOLL FORMATION 461
One of the main points in which Kuenen differs from Daly, how-
ever, is that he assumes as many coral structures before the ice age
as at present and that he ascribes the destruction of the limestone ex-
posed during low sea levels to chemical marine erosion and denudation.
All emerged reef rock is supposed to have suffered, while the corals
462 C ORA L REEFS
too deep for colonization, or at any rate they would not have offered
particularly favorable sites for reef establishment. Extensive terraces
somewhat deeper than glacial sea level would jut out here and there
beyond the periphery of the atoll and barrier reefs. However, in
the central coral seas and wherever a distinct barrier or atoll reef is
developed such platforms are conspicuous by their absence.
With the postulates of glacially controlled subsidence the continued
protection of the platform below sea level by thriving coral growth
not only accounts for the absence of greater marginal than central
CORAL REEFS
depths in wide lagoons but it also offers a logical explanation for the
consistent external position of present reefs. The corals were already
L.nd
o Reel
N
Meters
63
-----7 2
r --83
0'4S' N
\
U J
~\ ,..;,r;:.
.' , ,'
13 til
"
;- -. ..0...... 0)
- , .
I --
"I I
73 Kilometers
FIG. 193. Chan of Suva diva Atoll. Note slight basin shape of lagoon floor.
Depths in meters. (After Kucnen, 1947, Fig. 10, p. 22.)
established along the margins of the platforms, and the firm sub-
stratum of abraded reef rock offered an ideal foundation for coloni-
zation, far better than the abraded or aggraded lagoon floor farther
inwards.
THEORIES OF ATOLL FORMATION 465
ISO
. :~\(~) Bell reef
Klmbobo .'....... ,~.I.~ .... / W
(G;J:' Mallma
Malevuvu reef
<;~.
fs})
o 10 20
Kilometers
FIG. 194. Parr of Lau Archipelago, Fiji, showing the deep American Passage.
Depths in meters. (After Kuenen, 1947, Fig. 9, p. 21.)
PlslrTenph
Atoll
'r
o 1
Kilometers
FIG. 195. An atoll of the Togian Group, Celebes, showing a rim without pas-
sages and basin-shaped lagoon floor. Depths in meters. (Modified from Umb-
grove, 1939, Fig. 10, p. 151.)
these are applied separately. Yet one should keep in mind the want
of quantitative data on chemical marine erosion, reef growth, etc.
The weakest point is the assumption that exposed reefs were largely
destroyed although the processes invoked may prove to act very
slowly. It may also be that in some reef groups other theories offer
a better explanation.
EVIDENCE FROM CORAL REEFS 469
Maratua, Muaras, and Kakaban. These three raised atolls have been
repeatedly mentioned on foregoing pages, but no account has been
given of the geological history of the group. It is situated in the
Celebes Sea to the northeast of Borneo (Fig. 196). Maratua is raised
to 110 m above sea level and is partly abraded, presenting a crescentic
rim, while Kakaban attains about 45 m in a complete ring. Muaras
lies at sea level and appears to have been entirely eroded, except for
one small island at the northwestern end. These reefs rise abruptly
from an almost flat shelf that gradually dips outward from the main-
land. An intermittent barrier reef is found along the 100-m line, but
the atolls are situated farther out in water that is about 250 m deep.
470 CORAL REEFS
Vanua Levu
Kilometers
.,
FIG. 197. Parr of the barrier of Vanua Levu, Fiji, wirh large faros. Depths in
meters.
into atolls as the great distance to the coast tended to equalize condi-
tions all around the circumference of each reef. The curious barrier
of Vanua Levu (Fig. 197), consisting of large atoll-like faros, may
illustrate an intermediate stage of development. The reefs and atolls
on the northwestern edge of the Sahul Shelf, rising from a platform
at 540-m depth, also represent this stage of development (Teichert
and Fairbridge, 1948) . In the meantime accumulation of terrigenous
deposits aggraded the shelf by an amount of 200 to 300 m. During
the ice age an independent barrier took root farther in in shallower
water. The final act was the upwarping of the edge of the sub-
merged Shelf whereby the atolls were raised to their present high
station.
The Tijger AtCill1. This atoll rises steeply from a deeply submerged
ridge connecting Celebes with a group of islands to the south (Figs.
198 and 199). The atoll tops this geanticJine like a chimneypot stand-
ing on a roof, bllt it is much broader than the crown of the ridge and
EVIDENCE FROM CORAL REEFS 471
has evidently spread outward on its own screes. Its small neighbor
Taka Garlarang sprouts up abruptly from the side of the ridge, from
a depth of more than 2000 m. It is hard to imagine tectonically
_ Land
Meters
Meters - _ 2000 o Important soundlni
" ..........- 200 .... -- ..... - 2500
... ... 500 - - - 3000 - Echo - sounding hne
- - --1000 ---4000
1500 - 5000
F1G. 198. Chart of Tijger Atoll and Taka Garlarang Atoll south of Celebes.
(After Kuenen, 1933, Fig. 93, p. 102.)
Jo""
Iti/o~:,O 0 10 20 30
!'~ Kilometers
FIG. 199. Block diagram of Tijger Atoll and Taka Garlarang Atoll south of
., Celebes. (After Kuenen, 1938, Fig. 4, p. 96.)
for the regular shapes and fairly constant sizes, and especially the close
juxtaposition with narrow deep passages in between. Neither tec-
tonic nor volcanic activity tends to produce such forms. But growth
on a gradually subsiding ridge again offers an adequate explanation of
the configuration (Fig. 200).
The Fiji Islands. The Fiji Islands form the most complicated and
richest group of reef islands of the world. Barrier reefs and atolls
fringing reefs and elevated coral structures abound. Ladd and Hoff-
meister described the elevated limestones at length and offered the
suggestion that after marine erosion of pyroclastic cones the platforms
were blanketed with fine-grained limestone, consisting of algae, Fora-
minifera, etc. On Numuka these rocks are 70 m thick and covered by
at least 25 m of reef limestone. As the lower strata were deposited
in shallow water a subsidence of about 100 m is proved.
It is difficult to imagine conditions under which these fine-grained
deposits would cling to a submarine platform in shallow water. The
small depth is proved by some isolated reef corals and reefs and by
EVIDENCE FROM CORAL REEFS 473
cross-bedding. But if we assume a protecting barrier and slow subsi-
dence, as in Darwin's theory, the limestones could be interpreted as
FIG. 200. Part of the group of Maldive atolls in the Indian Ocean. Note the
deep channels dividing Malosmadulu Atoll into separate partS, and the faros.
Depths in meters.
---... - - - --
Principal wind "" ""
"
,""
"
... '--='''''''-I~:-'''''''------I
FIG. 201. Diagram showing development of barrier islands of the Lau Archi-
pelago, Fiji, acco rding to glacially controlled subsidence. 1. Marine erosion of
volcanic islnnd with reefs on the platform edges. 2. Slow subsidence, blanket-
ing during Tertiary times of island by fine lagoon deposits, with some reefs
and scree material. 3. Stepped elevation accompanied by erosion and final
er sion at glacial sea level. 4. Postglacial growth of barrier reef and sedimen-
tation on the platform. (After Kuenen, 1947, Fig. 4, p. 14.)
tween. In this picture the long period between the deposition of the
limestones and the present is occupied by outgrowth of the atolls
probably before they were elevated and by the elevation itself.
Davis attempted to bring all these forms into a logical and compara-
tively simple system by postulating a north-to-south geanticline wan-
dering westward through the group. But field examinations, espe-
5 TAB I LI T Y 0 F R E E F F 0 U N D A T ION 5 475
cially by Ladd and Hoffmeister, have shown that most of the so-called
raised atolls and barriers consist of Tertiary limestones and that sub-
aerial solution is responsible for the basin shape. Renewed testing of
Davis's views is required.
Bibliography
ARMSTRONG, H. E., et al. Tbe Atoll of Funafuti, Royal Soc. London, 428 pp., 1904.
DALY, R. A. A Recent Worldwide Sinking of Ocean Level, Geol. Mag., Vol. 57,
pp. 246-261, 1920.
The Cbanging World of tbe Ice Age, 271 pp., Yale University Press, 1934.
Coral Reefs-A Review, Am. J. Sci., Vol. 246, pp. 193-207, 1948.
DARWIN, C. Tbe Structure and Distribution of Coral Reef! , 2nd ed., 278 pp.,
Smith, Elder, London, 1874.
DAVIS, W. M. The Corlll Reef Problem, Am. Geogr. Soc., 596 pp., 1928.
EMERY, K. O . Submarine Geology of Bikini Atoll, Bull. Geol. Soc. Am., Vol. 59,
pp. 855- 860, 1948.
EMERY, K. 0., J. 1. TRACEY, and H. S. LADD. Submarine Geology and Topography
in the Northern Marshalls, Trans. Am. Geopb. Union, Vol. 30, pp. 55-58,1949.
FAIRBRIDGE, R. W. Notes on the Geomorphology of the Pelsart Group of the
Houtman's Abrolhos Islands, J. Royal Soc. Western Australia, Vol. 33, pp.
1-43, 1948.
F~IRBRIDGE, R. W., and E. D. GILL. The Study of Eustatic Changes of Sea Level,
Austr. J. Sci., Vol. 10, pp. 63-67, 1947.
FAIRBRIDGE, R. 'V., and C. TEICHERT. The Rampart System at Low Isles, 1925-45,
Rept. Great Barrier Reef Comm., Vo)' 6, I, pp. I-J6, 1947.
FLINT, R. F. Glacial Geology and the Pleistocene Epoch, 589 pp., John Wiley &
Sons, New,York, 1947.
GARDINER, J. ST. Coral Reefs and Atolls, 181 pp., Macmillan, London, 193 J.
HESS, H. H . Drowned Ancient Islands of d1e Pacific Basin, Am. J. Sci., Vol. 244,
pp. 772- 791, 1946.
HOFFMEISTER, J. E., and H . S. LADD. The Foundations of Atolls, ] . Geol., Vol. 43,
pp. 653-665, 1935.
A Criticism of the Glacial-Control Theory, ] . Geol., Vol. 44, pp. 74-92, 1936.
The Antecedent-Platform Theory, f. Geol., Vol. 52, pp. 388-402, 1944.
Geology of Lau, Fiji, B. P. Bisbop Museum Bull., Vol. 181, 399 pp., 1945.
KOENEN, PH. H. Geology of Coral Reefs, Tb e Snellius Expedition, Vol. 5, part 2,
126 pp., Brill, Lcyden, 1933.
Submarine Slopes of Volcanoes and Coral Reefs in the East Indian Archipelago,
C. R. Congr. Intern. Geogr. Amsterdam 1938, T. II, Scc. U", pp. 93- 98.
Kruistochten over de indische Diepzeebekkens, 220 pp., Leopold, The Hague,
1941.
Two Problems of Marine Geology, Atolls and Canyons, Verh. Kon. Ned.
Akad. v. Wet. Amsterdam, afd. Nat., 01.43, 3, 69 pp., 1947.
Reports of the Gl-eat Barrier Reef C07mnittee, Vols. I- V, 1925-1942.
BIBLIOGRAPHY 479
SAVILLE KENT, W. The Great Barrier Reef of Australia, 387 pp., Allen, London,
1893.
SEWELL, R . B. SEYMOUR. Studies on Coral and Coral Formations in Indian
Waters, Mem. Asiatic Soc. Bengal, Vol. 9, pp. 461 - 540, 1935.
An Account of Addu Atoll, An Account of Hornsburgh Atoll, Jolm Mltrray
Exp., Vol. I, 3 + 5, pp. 63-93, 109-125, 1936.
SPENDER, M. A. Islands Reefs of the Queensland Coast, Geogr. J., Vol. 76, pp.
193- 214,273-297, 1930.
STEARNS, H. T. Decadent Coral Reef on Eniwetok, Marshall Group, Bull. Geol.
Soc. Am., Vol. 56, pp. 783-788, 1945.
An Irttegration of Coral-Reef Hypotheses, Am. J. Sci., pp. 245- 262, 1946.
STEERS, J. A. The Queensland Coast and the Great Barrier Reefs, Geogr. I.,
Vol. 74, pp. 232-257, 341-370, 1929.
The Coral Islands and Associated Features of the Great Barrier Reefs, Geogr. J.,
Vol. 89, pp. 1-28, 119-146, 1937.
The Coral Cays of Jamaica, Geogr. ]., Vol. 95, pp. 30-42, 1940.
STEERS, J. A., et al. Sand Cays and Mangroves in Jamaica, Geogr. J., Vol. 96,
pp. 305-328, 1940.
TEICHERT, C. Contributions to the Geology of Houtman's Abrolhos, Westem
Australia, Proc. Linn. Soc. N. S. Wales, Vol. 71, pp. 145- 196, 1947.
TEICHERT, C., and R. W . FAIRBRIDGE. Some Coral Reefs of the Sahul Shelf, Geogr.
Rev., Vol. 38, pp. 222- 249, 1948.
TRACEY, J. I., H. S. LADD, and J. E. HOFFMEISTER. Reefs of Bikini, Marshall Island,
Bull. Geol . Soc. Am., Vol. 59, pp. 861- 878, 1948.
UMBGROVE, J. H . F . (a) The Influence of the Monsoons on the Geomorphology
of Coral Islands, Proc. Fourth Pac. Sci. Congr. Java, Vol. 2, pp. 49-54, 1929.
(b) De koraalriffen der Duizend Eilanden (Java-zee) , Wet. Med. Dienst Mijll-
bouw Ned.-Indie, No. 12, 47 pr., 1929.
De koraalriffen van Emmahaven ("V. Sumatra), Leid. Geol. Med., Vol. 4, pp.
9-24, 1931.
De atoHen en barriere riffen der Togian Eilanden (summary in English), Leid.
Geol. Med., Dl. II, pp. 132-187, 1939.
Coral Reefs of the East Indies, Bull. Geol. Soc. Am., Vol. 58, pp. 729-778, 1947.
UMBGROVE, J. H. F., and J. VERWEY. The Coral Reefs in the Bay of Batavia,
Fourth Pac. Sci. Congr., Exc. A2, 30 pp., 1929.
VER\.\'EY, J. Coral Reef Studies, Trmbia, Vol. 13, pp. 169-198, 199- 215, 1931.
WOOD JONES, F. Coral and Atolls, 392 pp., Lovell Reeve, London, 1910.
WOODS, H. Palaeontology, Invertebrate, 477 pp., Cambridge University Press,
1946.
c H A p T E R s E v E N
Geomorphology
of the Sea Floor
,(80
SHELF CHANNELS .481
SHELF CHANNELS
The shelf channels are of three different kinds. One class resembles
river valleys; another, glacial troughs; the third is due to tidal scour.
Drowned River Valleys. Various shelves are crossed by winding chan-
nels of moderate depths, which are evidently the seaward continua-
tions of land river valleys. The best-known example of a simple val-
ley is the Hudson channel that runs in a slightly winding course from
"J/)
o 15 30
Kilometers
FIG. 202. Chart of the Hudson channel and Hudson Canyon. (Redrawn from
Veatch and Smith, 1939; see also Fig. 217, p. 494, and Fig. 224, p. 506.)
50\ 70\
'Borneo "
~ . -. '
.> . ~
FIG. 205. Chart of a branching glacial trough off the Norwegian coaSt. (After
Shepard, ] . Geoi., 1931, Fig. 6, p. 349.)
Borneo show few species and genera in common with the rivers drain-
ing to the east, their fauna is intimately related to that of the Sumatra
rivers. At present 500 km of sea water separate these two islands, but
they should have joined during the Pleistocene stages of low level,
484 G E 0 M 0 R P H 0 lOG Y 0 F THE S EA Fl 0 0 R
their streams then forming the branches of one master drainage system.
It is of interest to note that the shelf channels are a few dozen
kilometers wide, evidently representing shallow valleys, ~ot actual
streambeds. The streambeds may have been filled in by sedimenta-
tion during the rise of sea level and later, or it may be that they will
be detected later by more accurate surveys.
Channels Due to Tidal Scour. Many deep swales in shallow regions
are evidently due to tidal scour (see, for instance, Fig. 111, p. 231).
o 10
200 400
Kilometers
Depths In ,.thoms
ContO l.;r Interval
100 tathoms
so
FIG. 206. Chart of the South Norwegian Trough, showing a large depression
close to its origin south of Oslo. (After Shepard, 1931, J. Geol., Fig. 9, p. 352.)
o 5
Kilometers
"
Fl(l. 207. The three canyons of Georges Bank, drawn by Shepard. Contour 25 fathoms. (After Stetson, 1936. Bull. Geol. Soc. Am., Vol. 47. Plate 3.>
SUBMARINE CANYONS 485
the middle. As some of these tidal channels link up with the den-
dritic system on beyond, however, it appears likely that they origi-
nated by river erosion and have merely been deepened by the action
of the tides since.
Drowned Glacial Troughs. Shepard ( 1931) and Holtedahl (1929)
pointed out that off coasts that were glaciated during the Pleistocene
a different type of shelf channel is met with. The walls are relatively
steep and straight, and the base is broad (Figs. 204, 205, and 206). The
transverse section is similar to that of fiords. The longitudinal profile
is undulating, with the deepest parts frequently in the inner portions.
Most of these troughs are seaward continuations of estuaries, espe-
cially of the nord type. Coalescing and branching troughs, deep-
rimmed depressions, and hanging valleys on the side are all met with.
Evidence of the presence of moraines within the troughs and along
the border has been discovered. Holtedahl is of the opinion that
faults have played an important part in locating both fiords and
drowned glacial troughs.
The evidence in favor of glacial erosion is overwhelming, and one
can hardly doubt that Shepard's interpretation is correct.
There remains one aspect that is in need of further elucidation.
Though it is easily understood why glaciers should cut troughlike de-
pressions out of pre-existing river valleys, the cutting of troughs on
the continental shelf needs some explanation. Why did the ice com-
ing out of the fiord onto the flat shelf not spread out to form a pied-
mont glacier? Nansen assumed that valleys had been cut across the
shelf prior to glaciation. As sea level was lowered before the glacier
reached the coast, there is much to be said for thjs suggestion. Most
troughs are wider than the fiords with which they connect. Hence,
there evidently was a certain tendency to expand beyond the narrow
valley.
SUBMARINE CANYONS
Data on Submarine Canyons. Submarine valleys of great depth cut-
ting the continental slope became known towards the end of the nine-
teenth century, and an explanation by the drowning of rivers was
offered. More general interest, however, was not excited until
Shepard took up the subject in the early thirties. This investigator
has also been most active in the collecting of data and the testing of
theories to explain these intriguing features of the sea floor. Although
the hydrographic surveys of many countries have published charts
of the continental slopes exhibiting these deep gashes that are now
486 G E 0 M 0 R P H 0 LOG Y 0 F THE SEA FLOOR
generally called submarine canyons, the U. S. Coast and Geodetic Sur-
vey has contributed by far the greatest number of soundings on which
accurate charts have been constructed (Fig. 207). Other oceanog-
raphers and geologists who should be mentioned are Stetson, who
investigated canyons off the east coast of the United States, and
.~ .
.,
. u!!--
_ _- - -- ----
,ttt&l io U
SC:AL(: N ST,IITu"( ~L[5
[UVATlON! IN r(T
VERTICAL x ~
Veatch and Smith, who compiled a set of contoured charts for the
whole eastern coast. Johnson reviewed all available data up to 1938.
He and several others mentioned below offered hypotheses concern-
ing the origin.
The wealth of information now available renders a detailed regional
description out of the question. But thanks to the publications cited
anyone interested can easily gain access to all the facts that have been
established so far.
SUBMARINE CANYONS 487
A typical canyon starts as a steep, narrow gorge cutting across the
continental shelf for a few dozen kilometers and running straight
down the continental slope to great depths (Figs. 208- 210) . At the
edge of the shelf the bottom may lie many hundreds to a thousand
meters below the adjoining sea floor. The walls are steep, in some
cases locally exceeding 1000/0 slope, and may rise more than 1000 m
on both sides. The transverse section is V -shaped, and in ground
plan a moderately sinuous course is followed. Longitudinal profiles
--Coast
o 10 20
Kilometers
FIG. 209. Longitudinal sections of rwo west coast and one east coast canyons,
with approximate shape of adjoining slope. Vertical 10 times horizontal scale,
depths in meters. (After Kuenen, 1947, Fig. 1, p. 38.)
United States they can still be traced to well over 2000 m, as far out
as the surveys have been carried, but the number of sounding lines is
insufficient to establish the true pattern.
Veatch and Smith show on their charts of the continental slope off
the east coast a valley system that is similar in pattern to that of youth-
fully dissected continental areas (Fig. 211). Shepard has criticized
.,
FIG. 210. Transverse sections of submarine canyons measured at the edge of the
continemal terrace (100 fathoms ). Redondo and Monterey on the west coast.
For comparison, three land "nl le),s are added. Vertical scale is 2~~ times hori-
zontal scale. Note asymmetrical profile of Monterey Canyon in bend, steep on
outward side; the other section is across a straight portion upstream from
point A. (Sec Fig. 232.) (After Kuenen, 1947, Fig. 14, p. 67.)
F,G. 211. Dendritic parrern of canyons off east coast of the United States ac-
cording to chart of Veatch and Smith. (After Kuenen, 1947, Fig. 13, p. 66.)
ing lines parallel to the valleys were available, that is, at right angles to
the section. The entirely unwarranted exaggeration of the relief by
the contours as shown by the dotted line is brought out in a striking
manner.
Until more detailed surveys are made, the true nature of the 'relief
must remain unknown, and in discussing submarine canyons only the
major, well-established features can be taken jnto account. Jt is to
be deplored that the authors, according to their own admission, set
out to draw the contours on the postulate that a valley pattern pre-
vails. The result is consequently biased, and their charts should not
be used without full allowanc for this fact.
Along the west coast of the Uruted States several canyons fade out
toward the Roor of the basins off southern California at about lOOO-m
490 G E 0 M 0 R P H 0 LOG Y 0 F THE SEA F LOOR
FIG. 212. Topography shown on part of chart by Veatch and Smith of the
co~tincnral slope off the east coast of the United States. Note that the valleys
are ' constructed parallel to sounding lines (soundings shown by dots) and
oblique to the general trend of the contours (shown, in fathoms, by thick,
dashed lines) . Saddles where sounding lines cross the ridges are also shown.
700
CIt
E
0 800
:5
~ 900
~ 1000
I , , , , I
o Kilometers 5
FIG. 213.Sections across the continental slope roughly parallel to the east coast
of the United States. I, according to Veatch and Smith. II, same locality
according to subsequent sounding line. III, elsewhere according to Veatch
and Smith. IV, along sounding line shown on their chart 6 km down the
slope. Note the exaggeration of the relief by the dotted sections based on
contours. (After Kuenen, 1947, Fig. 2, p. 40.)
SUBMARINE CANYONS 491
depth. More to the north they can be traced to at least 2500 m, but
farther out they lose their original character and are continued in
shallow, troughlike depressions of the sea floor. These troughs
strongly resemble the basins of the neighboring ocean floor. Shepard
points out that the direction also accords well with the supposed
diastrophic nature of these outer troughs. The manner in which the
narrow canyon enters obliquely, often at the side and not at the head
of the trough, is further evidence that the outer portions are not the
products of erosion but merely tectonic depressions into which the
canyons debouch.
Some of the canyons off California, where the shelf is generally
narrow, head in close to the beach (Fig. 215). Others start on sub-
marine banks with or without an island, but significantly no examples
of typical canyons are known on banks that remain below depths of
100 m.
Most of the eastern continental slope of North America is serrated
and, according to Veatch and Smith, shows practically no flat-topped
ridges between the furrows (Fig. 216). But it has just been shown
that their interpretation is doubtful. On the other hand it appears
492 G E 0 M 0 R P H 0 LOG Y 0 F THE 5 E A FLOOR
FIG. 215. Chart of dendritic canyons off the California coast. Inferred original
shape of contours before canyons were cut is also shown. Data from chart by
Shepard. Depths in feet. I, Davidson Seamount; 2, Lucia Canyon; 3, Sur
Canyon; 4, Partington Canyon; 5, Monterey Trough; 6, Monterey Canyon;
7, Cannel Canyon; 8, Soquel Canyon; 9, Ascension Canyon.
SUBMARINE CANYONS 493
NORFOLK
Flc. 216. Chart of the Norfolk Canyon and vicinity according to Veatch and
Smith. Outside the loo-fathom isobath the contour interval is 25 fathoms.
(After Daly, 1942, The Floor of the Ocean, Fig. 66, p. 120, University of North
Carolina Press.)
pattern than that off the east coast. This may be attributed to the
greater irregularity of the continental slope on the western side of
North America, but it appears doubtful whether there is any funda-
mental contrast in the nature of the valleys. Figure 210 brings out
the similarity in transverse section of the canyons on both slopes, and
Fig. 209 in longitudinal profile.
Outside the United States there are no areas explored with sufficient
accuracy to show the nature and number of submarine canyons.
Nevertheless more than a hundred examples have become known, and
apparently the occurrence of canyons may be considered the rule
494 G E 0 M 0 R P H 0 LOG y 0 F THE SEA FLOOR
rather than the exception. One of the best-known examples is that
of the Congo, which is exceptional in that it runs directly into the
estuary of that river.
Dredging and coring operations by Stetson in the eastern canyons
and by Shepard in the Californian gorges have brought to light many
,.
,('
1
r
Kilometers
I I I I I I I
o 4 8 12
Stlltute miles
FIG. 217. Chart of Hudson Canyon, based on chart by Veatch and Smith. (Sec
Fig. 202, p. 481, and Fig. 224, p. 506.)
-....
Flc. 219. Section ac ross Monterey Canyon compared to the Grand Canyon.
Same scale and spacing of observation points. Vertical 5 times horizontal scale.
(After Shepard. 1948. Submarine Geology, Fig. 72. p. i16. Harper and Brothers.)
topographic forms, and that it was partly filled by mud from the Mis-
sissippi after having been cut as a normal canyon. Several authors
have attributed the formation to a great submarine slide (Fig. 220).
Hypotheses on the Formation of Submarine Canyons. Several attempts
have been made to explain the formation of submarine canyons.
Shepard has carefully weighed with commendable impartiality the
evidence pro and con, coming to the conclusion that several agencies
have probably played a part in originating, altering, and maintaining
the canyons, but that as yet no definite decision can be made as to
what factor was chiefly responsible. On the whole he favors stream
erosion during a vast lowering of sea level in early Pleistocene times.
The present author, on the other hand, is strongly inclined to reject
SUBMARINE CANYONS 497
FIG. 220. Chart of the Mississippi Trough. Contour interval 150 feet. In sec-
tion: vertical 5 times horizontal scale. (Simplified after Shepard, 1937, Bull.
Geol. Soc. Am., Vol. 48, Fig. 2, p. 1351.)
Only the main arguments can be given, because Shepard lists about
one hundred for and against the various theories and even then many
more could be introduced. First a number of suggested causes will
be dealt with, that appear to the present author to be ruled out as the
main factors responsible. For the latter reason it is deemed superflu-
ous to sum up the points in favor of these views.
498 G E 0 M 0 R P H 0 LOG Y 0 F THE 5 E A fL 0 0 R
Diastrophic Origin. Several authors havc suggested that submarine
canyons have been developed mainly by faulting and local subsidence
by which narrow troughs were produced. The chief opposing con-
siderations are: the dendritic and sinuous pattern, the graded longi-
tudinal profiles, the direction transvcrse to the tectonic trends of the
adjoining COast (California), the evident linking \,-ith major aggrading
rivers on the adjoining land, the restricted timc of formation (late
Plioccne to late Glacial), the great number ofT the cast coast of the
United States that is tectonically inactive.
Although diastrophism is hereby ruled out as the main cause, it may
have helped to locate canyons, to altcr their course, and to form the
broad troughs out towards the ocean floor. Thus Shepard points to
the evident structural control of Monterey Canyon. The present
writer is strongly inclined with Lawson and others tlJ attribute the
cutting of Carmel Canyon to a pre-existing fault zone across the
granite mass into w hich this gorge is incised. The direction is in ac-
cord with the strike of the many faults in the neighboring Santa Lucia
Range, and on the opposite wall of Monterey Canyon complicated
features indicate some abnormal influence in action. Besides, the time
available for the cutting is extremely short even if subaerial erosion
by an extension of Carmel River is assumed.
Shepard places the deep valiey off San Clemente Island in a separate
class, calling it the San Clemente "Rift" Valley. It lies in the direct
continuation of the submarine escarpment that cuts off the bank
around this island on the northeastern side. Pointing to the abnor-
mally straight course and the position, he argues that a tectonic origin
is highly probable, although the nature of the faulting cannot be
ascertained.
The singular, broad depressions in the Bahamas have many points
in common with submarine canyons, but the major depressions can
be more readily explained by diastrophic processes. On the other
hand the regular outward grading and sharp V-shaped trench along
the floor of the "Tongue of the Ocean" might be due to some type
of erosive current. (See Hess, 1933.)
Warping of the Continental Borderland. Various authors have
assumed great changes of level in the areas now forming the conti-
nental slopes, so as to dro~ n canyons that had previously been exca-
vated by subaerial rivers. Daly and Umbgrove have raised objections
to these views. As long as canyons were thought to be exceptional
features and to be limited to a few incidental gorges off the mouths
of some major rivers, this explanation held the attraction of simplicity.
SUBMARINE CANYONS 499
Now that canyons have been found by the dozen, are known from all
over the world, and will evidently be found to be more the rule than
the exception, the conception of valleys drowned by warping has lost
most of its appeal. Not only must one assume that most continental
slopes were drowned to an extent of a few thousand meters in very
recent times, but before the period of high stand, sedimentation was
proceeding. From this it follows that up to the end of the Tertiary
the continental slopes were roughly in the same position as at present
and a temporary emergence during the cutting of the canyons has to
FIG. 221. Diagram showing the peripheral welt (1) which must have existed if
submarine canyons (2) had been fomled by warping of the continental border.
FIG. 222. Diagram illuStrating the artesian spring sapping hypothesis. I-IV,
successive positions of spring heads if the postulates of that hypothesis are
granted. A, aquifer. (After Kuenen, 1947, Fig. 3, p. 44.)
forming initial depressions that other factors have altered to the pres-
ent canyons, and in keeping the canyon walls and head free of recent
sediment.
To assess quantitatively what the action of gravity sliding has been
is as yet beyond us. It may even prove to have been a major factor
in setting up turbidity flows able to erode the bottom of canyons. It
also appears likely that the minor tributaries starting some distance
down the slope and the furrows in the walls of major canyons owe
their shape to density flows set up by slumping or even to slumping
alone. We shall return to this mechanism when discussing the action
of turbidity currents.
Tsunamis. Bucher invoked the action of tidal waves formed, by
earthquakes, so-called tsunamis, as the principal agents in canyon
formation. There is much to be said for and against this opinion. The
main objections are that these great waves are oscillatory and would
carry nearly as much sediment up as down the slope, even if we are
prepared to admit that the engendered currents are sufficiently power-
ful to erode the continental slopes to any marked degree. Once
gullies were formed, one would expect the intervening, shallower
portions of the slopes to be more severely attacked than the floors of
502 G E 0 M 0 R P H 0 LOG Y 0 F THE 5 E A Fl 0 0 R
the canyons. The dendritic pattern is a further obstacle, and the evi-
dent link between canyons and land rivers is also hard to explain.
Two other important facts are difficult to account for: (1) that
the canyons were cut since the late Pliocene, and (2) the develop-
ment in regions where tsunamis are at present unfrequent (southern
Atlantic, California coast, Shepard and Emery, 1941).
Bucher suggested that the Pleistocene was perhaps a particularly
seismic period. Little can be brought forward in support of this
ad hoc postulate. Even then the available time is so short that an
almost uninterrupted string of seismic waves would have been re-
quired to effect such gigantic work of erosion.
FIG. 223. Diagram illustrating the hydraulic currents caused by onshore winds
that cannot sweep down the slope in consequence of density stratification of
the ocean, (After Kuenen, 1947, Fig. 4, p. 46.)
" The present author is inclined to assume that tsunamis may play an
important part in the prOduction of slumping (see p. 247), but only
when a canyon has once been formed by other agents and steep walls
are already present.
Hydraulic and Tidal Currents. Davis attempted to explain the
mock valleys, as he called them, by hydraulic currents. Onshore
winds pile the sea water up ag'a inst the coast, and an outflow along the
bottom must take place. These or other throughgoing currents, how-
ever, cannot be held responsible on account of the density stratifi-
cation in the oceans. The warm surface waters are unable to force
their way down beyond the edge of the continental shelf, because they
would have to replace heavier cold water (Fig. 223).
Nevertheless, it is probable that an important part is played by the
hydraulic effect of onshore winds in starting suspension currents, as
will be explained below.
Tidal currents and eddies have been measured close to the bottom
both on the east and west coasts of the United States in and around
the canyons. The maximum velocities found were only a few dozen
centimeters per second, and water samples have entirely confirmed
the inability of such currents to churn up even fine sand or mud.
SUBMARINE CANYONS 503
Neither was the direction closely correlated to the shape of the can-
yons. Again, the limited geological time in which cutting was effec-
tive rules out normal currents as agents in the development of the
canyons. Several objections raised against the hypotheses already dis-
cussed apply to this view also.
Effects Due to the Ice Ages. One of the most significant facts dis-
covered about submarine canyons is the short period during which
they must have been formed. Many instances of Miocene strata crop-
ping out along the walls have been encountered during dredging oper-
ations, and even Pliocene deposits have come to light. Evidently the
formation of the canyons may have begun earlier and furrows may
have developed by sliding, diastrophism, etc., in earlier periods. But
there can be no doubt that, during the Pleistocene, excavation sud-
denly became universal and proceeded at great speed. Whereas in
earlier periods deposition was the rule on the continental slopes, the
ice ages saw a reversal of conditions and erosion in gullies became
widespread off the Californian and the Atlantic coasts of North
America.
No less significant is the return to deposition since the last cold
stages of the Pleistocene. Not only does this prove -a time limit to
the period of active erosion, but it also is conclusive evidence that
deposition is the normal phenomenon beyond the edge of the shelf
and in depressions.
The question thus arises of the cause of abnormal conditions on the
sea floor during the Pleistocene. This period does not appear to have
been exceptional in diastrophic or any other internal respect. Neither
can the general cooling of the oceans have resulted in erosional proc-
esses on the sea floor. There was probably increased erosion on the
continents, but it is hard to see how this could have called forth di-
rectly an excavation of furrows on the continental slopes. All other
causes being thus eliminated, we are forced to attribute the canyons
to the lowering of sea level during the ice ages. This having been
ascertained with great probability one is left with tWO possible ex-
planations. Either the sea level was lowered suffi~iently to leave the
sea floor dry, thus allowing normal subaerial erosion to cut the can-
yons, or a submarine type of current was engendered in consequence
of a moderate fall of ocean level.
Excessive Lowering of Sea Level. Soon after beginning his prolific
studies on the problem of submarine canyons; Shepard carrie to the
conclusion that only subaerial excavation by rivers could' acc6unt for
their formation. Although he later admitted that slumping, dias-
504 GEOMORPHOL O GY OF THE S EA flOOR
'r
,
FIG. 224. Chart of the lower end of Hudson Channel ending on a submarine
delta, based on chart by Veatch and Smith. Depths in fathoms. Note that
there is no direct connection with the head of the submarine canyon. (See
also Fig. 202, p. 481, and Fig. 217, p. 494.)
But the relation is not a very close one. There are many large
rivers without a canyon. Many canyons occur where only small
streams reach the coast, or where only doubtful glacial melt water
streams can be postulated (Georges Bank). Thus the largest Portu-
guese canyon, the Nazare, is located off a small river, and there is
little to be said in support of Shepard's suggestion that the Tagus may
have reached the coast at this point during the Pleistocene. The cor-
relation between the size of the river and the size of the adjoining
canyon is poor. Canyons also tend to come in groups, often quite
close together, where there is only one river on the land. In other
SUBMARINE CANYONS 507
cases the canyon has a forked head, only one of the two gorges
showing a relation to a land river.
Moreover, with only few exceptions the canyon head is situated a
few or even many kilometers out from the coast. The topography
of many canyon heads is definitely against the supposition that they
I
! ! I I I I
2 4 6
Kllometor1
Hercules Bay
Depths In fathoms
Flc. 225. Chart of three canyons off the northeastern coast of New Guinea.
(Redrawn from Sprigg, 1947, Fig. 2, p. 300.)
formerly extended to within the present coastline but have since been
infilled by river deposits. The canyon usually narrows and shoals
gradually towards its head and sometimes ends in a sharply cut small
furrow. Had a much deeper and broader gorge been silted up, one
would expect a rounded head and a flat bottom (for instance, like the
Mississippi Canyon). The ending is evidently a primary unaltered
erosion form. Where the wall rock is consolidated the same conclu-
sion can be drawn. This point is strong evidence against the hypothe-
508 G E 0 M 0 R P H 0 LOG Y 0 F THE SEA FLOOR
sis of river cutting, because in that event the river could not have
failed to sink a deep valley across the shelf.
Shepard has also sought for other sources of evidence showing that
sea level stood much lower in comparatively recent times. He men-
tions the widespread occurrence of rounded gravel and shallow-water
shells in canyons and elsewhere. The present writer believes that
turbidity currents of high density may also account for these phe-
nomena, while rafting, winnowing by currents on seamounts, and sub-
sidence may be responsible in other
localities.
Guyots are also held by Shepard
as evidence of a low sea level, be-
cause they are evidently beheaded
volcanoes, and many have flat tops
FIG. 226. Diagram of the formation at about 1300-m depth. However,
of guyOtS by low-lcI'cl marine plana- the depth varies between 1000 and
tion, as postulated by Shepard. A
should have been eroded while B
1700 m, and probably the range
was being beheaded. will expand with increasing knowl-
edge of the sea floor. While the
deeper-lying guyots were being beheaded, the more lofty ones evidently
escaped erosion, for there are no terraces on the slopes corresponding
t'6 the wave action that found time to smooth off their neighbors to a
perfectly level bank 10 to 20 km across (Fig. 226).
If, on the other hand, we assume deleveling since the truncation,
which must have occurred less than half a million years ago in Shep-
ard's scheme, the deduced movements are exceedingly swift (at least
1 mm per year) and surprisingly large. Moreover, sea level must have
stood for a long time at its extreme level to truncate also lava flows,
plugs, etc. Finally, if such swift deleveling is common, guyots or at
least the sinking ones must represent young volcanic cones, for all
pre-Pliocene ones would have disappeared (see also p. 477).
Then, in the case of Bikini the drilling has demonstrated that the
atoll structure commenced to grow at least as early as the Miocene.
Hence the guyot below was beheaded long before the Pleistocene.
The alternative view put forward by Hess is that guyots are of
all ages and tend to subside very slowly, the deeper ones having
been submerged farther than the shallower ones (see p. 397). This
picture accounts more satisfactorily for the observed facts.
Two possibilities can be suggested for the conditions in the Medi-
terranean during excessive lowering of sea level. If precipitation and
runoff were in excess of evaporation, the sea should have been ponded
SUBMARINE CANYONS 509
at sill depths in the Straits of Gibraltar at 400 m. For about two-
thirds of the time of lowered sea level Mediterranean waters thus
should have stood at 400 m below the present position. During this
long time considerable deltas should have been built by the major
rejuvenated rivers. But nowhere does the 500-m line project appre-
ciably beyond the 200-m line, not even on the Nile delta. Moreover,
the submarine canyons of this inland sea descend at least 500 m below
still depths.
The other possibility is that evaporation remained in excess of pre-
cipitation. Then the surface of the Mediterranean would have sunk
below sill depths. Even if one is prepared to grant this possibility,
still more seriolls objections are presented by the Japan Sea. Here,
in this colder climate, precipitation must have remained in excess of
evaporation, and ponding at less than 200 m would result. There are
canyons, however, in this basin also, passing more than 1000 m below
sill depths.
The outcome of this review is that even if the highest possible
values are assumed for glacial lowering of sea level only half the mini-
mum depth to which canyons are cut can be attributed to this cause.
The remainder must be explained by assuming a highly speculative
upbulging of the continental slopes. Even then many features of the
canyons themselves can hardly be reconciled with logical deductions.
Some further points will be mentioned in the next paragraph. To all
these objections must be added the evidence from geology and biology
against any major reduction of sea level.
Turbidity Currents. Soon after the problem of submarine canyons
had again excited the interest of geologists Daly proposed an ingenious
mechanism by which submarine cutting might be explained. During
the moderate lowering of glacial sea level great quantities of mud on
the continental shelves are supposed to have been churned up by storm
waves.
When sediment is taken up in suspension the specific gravity of
the water is increased. Thus the density of water to which 1% of
silt is added becomes roughly 1.017, while for sea water the density
is raised from 1.026 to 1.040. Hence, the turbid water of the shelf
during storms would have formed a heavy suspension. This weighted
water would act just as though the density had been increased by
greater salinity, and it wou ld tend to flow out from the shelf and to
run down the continental slope to the floor of the ocean. In this
manner rivers of muddy water were set up that could have eroded
the bottom and cut the submarine canyons.
510 GEOMORPHOLOGY OF THE SEA FLOOR
1 2
Kilometers
FIG. 227 . Chart of channel on Rhone Delta in Lake Geneva. (Redrawn from
chart by Dclcbecque, in Collet, 1925, Fig. 42, p. 189.)
the velocity of flow in rivers also holds for turbidity currents, but
with increasing density allowance must be made for higher viscosity.
The flow collects in slight indentations of the slope and draws in turbid
waters flowing down the slope along the sides of the trench. The
flow, even when showing a very slight effective density, runs along
the bottom, and, although turbulent mixing with surrounding clear
water is active, the currents continue for large distances as separate
bodies. Owing to their momentum they do not come to a standstill
when they reach horizontal areas but continue for a long way and
can even ascend slopes if they happen to encounter them.
Turbidity currents of low velocities are known to occur in reser-
voirs and lakes. In lakes they form trenches on the deltas, but pos-
sibly more by lack of deposition than by erosion (Fig. 227). In
reservoirs no trenches appear to develop. It is frequently argued
SUBMARINE CANYONS 511
can sometimes have exceeded that of sea water and that turbidity
currents were thus generated beyond the mouth. The present writer
is therefore inclined to assume that the tendency of canyons to occur
off river outlets is partly due to this phenomenon. On the other hand
the link between rivers and canyons is not so close, as one would
expect if Sprigg's suggestion were the principal process involved.
Some of the criticism of Shepard's hypothesis applies also to this
explanation. Hence the extra sediment is held to have been supplied
indirectly by rivers during low sea levels and to have been first
deposited before being churned up by waves or set in motion by
slumping.
One of Shepard's many important discoveries is that slumping takes
place at the head of some Californian canyons. By careful repetition
of lines of soundings he found that sudden deepening up to amounts
of nearly 20 m may occur. In one place a hump that was found to
have settled on the Roor had later disappeared, evidently having slid
out to sea.
At Redondo a slide was actually witnessed from the pier that had
recently been extended to the head of Redondo Canyon. Numerous
people fishing along the pier with a calm sea and no wind felt their
leads being pulled out to sea and had to put out more line to get
bottom. After a short interval boiling masses of mud appeared in
the water. In about an hour the deepening had ceased but not until
depths had increased from 3 to 12 m. Shepard also points out that,
where great masses of sand and silt are yearly introduced into the
heads of canyons close inshore, slumping of some sort must be con-
tinually keeping these depressions from being filled in.
There can be small doubt that the movement in these cases was of
the nature of slumping. But some features indicate that the slumps
tend to change to turbidity currents. More than once, where sudden
deepening was actually witnessed, currents have been seen to flow
"into the head of the canyon. This proves that a large volume of
water was passing down the canyon away from the coast. If only
sediment had taken part in the movement, no extra Water would have
been sucked into the canyon. Thus evidence is found that the slump-
ing results in a Row of watery mud or muddy water.
Then, during a local northwest blow rapid deepening to the amount
of 8 m has been observed in Redondo Canyon. Here the churning-up
of sediment and a turbidity flow are indicated as the causes of the
erosion.
514 G E 0 M 0 R P H 0 LOG Y 0 F THE S E A F LOO R
Shepard's finding may here be recalled that, while canyons heading
well out to sea have received a fill of recent fine sediment apparently
ever since the last glacial stage, the gorges that run in close to the
shore have hard rocky bottoms or coarse deposits along the floor,
with only a thin film of mud on top . He rightly draws the conclu-
sion that some type of sliding intermittently sweeps out the latter
group of canyons.
A strong case can be made for the contention that turbidity cur-
rents are the chief agent in this process. Neither the depth, slope,
nor shape of the canyons shows systematic differences in these two
groups. The only difference that can logically be claimed as the
cause of the inshore type being swept out is the smaller depth of the
surrounding sea floor and the shorter distance to the zone of breakers.
In other words, the fact that mud and sand are occasionally churned
up around the head of a canyon might well be the cause of non-depo-
sition farther out. A turbidity current is then set up into the canyon
and sweeps it out. This conclusion is supported by the observation
that the sweeping is not restricted to the inshore parts but continues
far out to sea beyond the line along the steep heads of offshore can-
yons with unswept floors.
It is true that sliding could also account for the phenomena, but
tho slope is rather small. Although it is known that a body of mud
can slide on a slope no larger than that of a canyon bed, it may be
doubted that it can remove all mud on so slight a declivity without
having changed to a turbulent current.
Whether the sweeping out is attributed to a pure slide or to a
turbulent current, the difference between the two groups of canyons
at any rate shows that the lowering of sea level during the ice age
must have brought all types under more favorable conditions for
erosion.
Further Aspects of Turbidity Currents. It was shown above that the
explanation of submarine canyons is to be sought in the effects of the
ice age and that either a great lowering of sea level or turbidity cur-
rents may be the main cause. A number of objections against a great
depression of sea level that must have amounted to more than 1000 m
have already been enumerated. In the following an attempt will be
made to show that the hypothesis of turbidity currents can account
for some remarkable features of the canyons that are inconsistent with
the alternative explanation by subaerial currents.
In the first place, there is the formation of some canyons in positions
where no hinterland exists that could have provided large volumes of
SUBMARINE CANYONS 515
water for river erosion (Georges Bank off the east coast of the
United States, Spanish and Delgada canyons off California, Tanner
Canyon off Tanner Bank to the southwest of San Clemente Island,
Nazare Canyon off the Portuguese coast).
FIG. 228. Cha.rt showing position of La Jolla and Scripps canyons to the north
of a cape on the Califo.rnia coaSt. Based on chart by Shepard and Emery, 1941,
Plate 9, p. 64. T-T, apprOXimate original edge of shelf. Depths in meters.
to the north of a line at right angles to the general trend of the coast-
line as it is at the present day.
In the third place, the forked head of many canyons is easily ac-
counted for by considering how the sllspension is drawn in from the
sides off the shelf. Had the erosion been due to a river crossing the
shelf from the hinterland only a single trench would be expected.
According to Shepard ( 1941, p. 86), "The extraordinary thing
about this pairing is that in all cases except r.a Jolla Can:von the head
of the southern member of the pair
approaches closer to the coast than
docs the northern mcmber. Also in
all cases the southern member is the
larger. Furthermorc in most of
these cases the southern canyon is
found directly north of either a
point or of the termination of one
of the short coast ranges. Finally, in
the majority of cases both canyons
FIG. 229. Diagram illustrating the head into a broad indentation of
twn types of offshore and inshore the coast." These characteristics
canyons on the \fest coast of the are readily explained by the sup-
U n\~ed States. The inshore type is
cunrinually swept out by slumping posed development of tllrhidit.I' cur-
and turbidity currents. Develop- rents where the waters were piled
mellt norrh of headland is due to up by oblique storms. Once the
ponding of muddy waters against canyon had deeply indented the
obstruction to longshore currents,
shelf a tributary was formed to the
due to oblique winds. (After
Kuenen, 1947, Fig. 7, p. 56.) north by inflow off the shelf, but it
failed to develop as far as its older
partner through lack of time and sediment. It is most unlikely that
any other theory could explain these rules more satisfactorily.
To the mind of the present writer one of the most significant points
in choosing between submarine and subaerial erosion is concerned
with the disposition of materials excavated from the canyons. Shepard
mentions three canyons (Coronado, Redondo, and Dume) in front
of which a delta like bulge of the continental slope is revealed by the
soundings. Either theory could explain the occasional occurrence of
deltas where conditions \vere favorable to deposition. But, as already
pointed out above, a more general development would be expected
if subaerial erosion had cut out the canyons. Moreover, the canyon
usually continues down along the side ' of the "delta"-a feature that
SUBMARINE CANYONS 517
does not accord with the view of delta building during lowered sea
level (Fig. 230).
With turbidity currents, however, the exceptional deltas might be
due to preglacial slumping or local circumstances in connection with
slope, relatively coarse materials, etc. But there is no need to expect
FIG. 230. Chart showing delta-shaped bulge at lower end of Coronado Canyon
with the channel passing down the side of the "delta." The delta may be due
to a major slu mp, later furrowed by turbidity currents. If it were a delta
formed during excessive lowering of sea level, it should lie deeper, as the San
Diego Trough, into which the canyon flows, would not have been ponded.
(Based on ('hart by Shepard and Emery, 1941, Plate 8, p. 62.)
FIG. 231. Chart of the Santa Cruz Canyon ending rather abruptly on the slope
at "ponding level" of Santa Cruz Basin. Note absence of delta and slight indi-
cadon of slump smothering lower end of canyon. (Based on ehart by Shepard
and Emery, 1941, Plate 17, p. 88.)
Shepard maintains that some canyons end above the floor of basins
into which they debouch. They die out at the level at which lowering
of sea level would have caused ponding of the waters. The Santa
Cruz Canyon is an example. On the one hand there is no indication
of the delta, at this level, that should have resulted from ponding.
On the other hand there are features on the adjoining slope that might
indicate a slump that smothered the former, lower end of the canyon
where it passed below sill depths (Fig. 231).
Another argument of considerable weight in choosing between sub-
aerial and submarine erosion is the degree of angularity of the canyons.
SUBMARINE CANYONS 519
Shepard has repeatedly emphasized the complete similarity between
submarine gorges and the canyons formed by running water on land.
There is, however, one respect in which the two forms show a sig-
nificant difference. Water is able to flow on land around relatively
sharp bends on account of gravity that holds it down even where
a strong centrifugal force is developed in a sharp turning. A turbidity
~F
~
I I I II I I I I
o 20 40
Kilometers
flow, on the other hand, has a very small effective density; conse-
quently in a sharp bend it would be forced high up against the outer
wall by its inertia and jump right our of a shallow canyon. There-
fore only slight curves can be followed by a turbidity flow, and where
faults would tend to induce a sharp turning the current wi)) round
off the bend and erode a wide curve.
When the submarine canyons are compared to subaerial canyons a
remarkable difference immedia ely catches the eye. In Fig. 232 the
two most sinuous canyons off California and the east coast are drawn
to the same scale as twO land canyons. The sounding lines are indi-
520 G E 0 M 0 R P H 0 L OG Y 0 F THE 5 E A F LO O R
cated to show the large number of data on which the shapes of the
marine canyons are based. When the great depth of the submarine
canyons is taken into account it is obvious that, although minor irregu-
larities need not come out in a chart, the major curves are accurately
known. If forms comparable to those of the land canyons existed,
these should cause abrupt turns in the depth curves. It appears safe
to conclude that the submarine canyons nowhere present the sharp
turnings of the land canyons. Even a much larger river than the sup-
posed stream in Monterey Canyon can twist and turn in quite sharp
bends, as exemplified by Colorado Canyon. The structural control
of Eel and Monterey canyons is expressed only in wide bends. Knee-
bends like the one in Niagara gorge have not been developed. The
greater slope of the continental terrace may be held responsible for
the generally straight course of the canyons. But the e'(amples shown
in our figure clearly demonstrate that this is not a rule without excep-
tions. Had the agents cutting the submarine canyons been able to
form harp bends there would doubtless have been found examples
by now.
Daly suggested that quite possibly a few of the California trenches
were actually river-cut far back in geological time, then drowned,
afterwards more or less completely filled with sediment, and, finally,
recently re-excavated by turbidity currents. Although this is a rea-
sOl1able suggestion the difficulty is to account for the outcropping
of Tertiary rock high up on the walls. The correlation with the
shapes of the coastline would also remain obscure. Is it likely that
some canyons are of a different origin from others off the same stretch
of coast? It is not an impossible but hardly a satisfactory explanation.
The most weighty argument that has been brought forward against
the hypothesis of turbidity currents is that they cannot have devel-
oped sufficient erosive power to cut into the continental slope. This
point must therefore be treated in some detail. Dredging in the east-
coast and west-coast canyons has shown that the majority of rock
encountered is less consolidated than rocks, in general, encountered
on land. The absence of percolating groundwater, the principal agent
in the induration of sediments below the land surface, readily explains
this difference.
Two sets of observations bear out this conclusion. In the first
place, geophysical investigations of the Atlantic continental shelf by
seismic methods have demonstrated that the materials building up the
continental terrace are poorly or moderately consolidated. Below a
few hundred meters of unconsolidated sand, mud, and gravel, a thick
SUBMARINE CANYONS 52 1
.6
....
c:
r
o
c:
'u
.g
.
c
~
522 G E 0 M 0 R P H 0 LOG Y 0 F THE 5 E A FLOOR
wedge of semi-consolidated sediments, possibly interbedded with thin
layers or lenses of hard rock, was deduced.
In the second place, the absence of "waterfalls" along the course
of the canyons and of steep cliffs of great height, either on the canyon
walls or where faults are inferred on the continental slope, forms
strong evidence for the comparative weakness of the rocks.
The scarps of what are probably submarine faults are all of mod-
erate steepness, although the fault planes must in most cases have been
mUch nearer vertical. Thus the steepest parts of San Clemente, Cata-
lina, and Coronado escarpments show slopes of 30-40% (1 :3.3 to 2.5).
The steeper sections of Gorda Escarpment are 26 % (1: 3.8) j the aver-
age is only 4.370 (I :23) (Figs. 233 and 234).
0F=~~~~~======~~==~~~~~~==~
--.... Slurt)""P:d
ilOOO
2: :::-.;.
I I I I
3000
Turbidity curren~---'="-- Meters
2000
or river 500 meters
FIG. 235. Transverse section (namral slope) of the Congo submarine canyon
at edge of shelf, showing small proportion of eroded matter as against slumped
matter.
1730
012345
Kilometers
FIG. 236. Block diagram of the extinct volcanic islands of Tidore and Maitara
in the Moluccas.
Kilometers
5 10
0,
FIC. 237. Sketch, true-scale section, and schematic block diagram of the active
volcanic island of Gunung Api, north of Wetar in the Banda Sea. (Kuenen,
1941, Fig. 32, p. 146; Fig. 33, p. 150.)
are reasons for supposing (Lawson) that a fault zone had crushed
and weakened the rock beforehand, thus greatly facilitating erosion.
We need but look at the abnormal direction of this gorge, closely
according with the direction of fau lts in the adjoining Santa Lucia
Range, the complicated physiography of the wall of Monterey Can-
yon, opposite the lower end of the tributary Carmel Canyon, and the
structural control of the whole system of furro'ws as emphasized by
Shepard to be convinced of the high probability of fa ulting having
aided in locating Carmel Canyon and preparing the rock for facile
SUBMARINE CANYONS 525
erosion. La Jolla Canyon also appears to coincide with a fault on the
coast. In other cases the existence of a fault is demonstrated by the
contrasted nature of the opposite walls proved by dredging. The
slope is then found to be much steeper on the hard-rock side.
When seeking for evidence on the origin of the canyons it should
be kept in mind that evidence from depths less than 100 m is question-
able, because of the emergence during the glacial lowering of sea level.
Concluding Remarks on the Origin of Submarine Canyons. On the
foregoing pages the problem of submarine canyons was discussed at
some length. The following main conclusions were reached.
Flc. 238. Sketch of the active volcanic island of Sangean north of Sumbnwa,
seen from the south southwest.
That the formation was almost entirely brought about during the
ice ages appears to be firmly established. This rules out most pro-
posed agents as constituting the main cause and can be explained only
hy attributing the excavation to the influence of lowered sea level.
Diastrophic action, especially faulting and the development of troughs,
has played a part, but only a minor one. Slumping and sliding are of
importance, both in forming initial furrows and in widening the can-
yons from vertical trenches to open valleys. Probably turbidity flows
were engendered by this mechanism. Tsunamis may have caused
some of these slides. Spring sapping and excessive lowering of sea
level are not believed to have played a part.
The main cause is sought in turbidity flows, weighted by mud and
sand. During the ice ages this action was greatly strengthened owing
to the sinking of ocean level by 80 to J 00 m below present level, so
that virtually the canyons were developed and modeled during this
period. It brought all canyons into even more favorable positions for
the action of turbidity flows than the present inshore canyons of Cali-
fornia. These canyons are being continually swept out by this mecha-
nism at the present day. The chief points in favor of turbidity flows
are: absence of sharp turnings and waterfalls, the almost universal
absence of deltas at the lower end, the forked heads and preferred
position close to the north of promontories (California), the only
526 G E 0 M 0 R P H 0 LOG Y 0 F THE 5 E A FLOOR
partial correlation with rivers, the passage far below sill depths and
absence of deltas in the Mediterranean and Japan Seas, the close
adherence of the heads to the shelf and shallow banks, and the for-
mation during the Pleistocene.
The main point of doubt is whether the rocks forming the conti-
nental terrace are so poorly consolidated that they can have been
swiftly eroded by currents of a few meters per second. This doubt
is so serious, in view of the many places in which hard rock has been
encountered on the canyon walls, that the hypothesis of turbidity cur-
rents cannot be definitely accepted, in spite of its many outstanding
merits. But the swiftly accumulating mass of data may bring con-
viction one way or another in the near future.
the study of submarine slopes reveals that the concave profiles of larger
subaerial volcanoes are due to erosion and deposition, and not, as has
been suggested, to collapse.
A test of the above deductions is found in the shape of the sub-
marine profiles. Here the fact that the slopes are almost straight
down to near [he sea floor is in accordance with the view that con-
cavity is due to subaerial influences.
Ternate
FIG. 240. Sketch of Ternate (1715 m) and Maitara (357 m), two volcanic islands
west of Halmahera, Moluccas, seen from the south.
I ~~
FIG. 241.
~ I
Sketch of Paluweh (875 m), a deeply eroded, active volcanic island.
north of Flores, Moluccas, seen from the northeast.
steepness, height, and volume. Soon the upper portion assumes the
natural slope. The growth will continue upward and outward, until
over the whole pile the natural slope predominates. With the varying
force of eruptions and the range in size of ejectamenta a slight con-
cavity may result. When the cone emerges above sea level the dry
part will be slightly steeper on account of the higher angle of repose
for dry material. With increasing height of the subaerial part, erosion
gains in relative importance. On an average (in Indonesia) it
SUBMARINE VOLCANIC SLOPES 529
will compensate for the difference between the wet and dry material
when the cone rises to 600 m. The upper part loses more and more
of the fine volcanic materials through the washing out by rain. The
lower parts are built up in ever-increasing degree by water-borne
3~~ ________________________________________ ~
F1G. 242 . Two sections through the volcanic island of Paluweh with submarine
slope based on four soundings (see Fig. 241).
FIG. 243. Diagram showing the growth of a composite volcanic cone from the
sea floor [0 above sea level.
Bibliography
BAILEY, E. B. American Gleanings, 1936, Trans. Geol. Soc. Glasgow, Vol. 20,
pp. 1-16, 1938.
BELL, H. S. Density Currents as Agents for Transponing Sediments, ,. Geol.,
Vol. 50, pp. 512-547, 1942.
BoURCART, J. Sur les rechs, sillons sous-ma.rins du plateau continental des Albcres
(Pyn:necs-Orientales), Compo rend. acado sci., Vol. 224, pp. 1175-1177, 1947.
BUCHER, W. H. Submarine Valleys and Related Geologic Problems of the North
Atlantic, Bull. Geol. Soc . Am., Vol. 51, pp. 489-512, 1940.
CoLLET, L. W. Les lacs, 320 pp., Gaston Doin, Paris, 1925.
DALY, R. A. Origin of Submarine Canyons, Am. ]. Sci., Vol. 31, pp. 401-420,1936.
The Floor of the Ocean, 177 pp., University of North Carolina Press, 1942.
DAVIS, W. M. Submarine Mock Valleys, Geogr. Rev., Vol. 24, pp. 297-308, 1934.
DUTOIT, A. L. An Hypothesis of Submarine Canyons, Geol. Mag., Vol. 77, pp.
395-404, 1940.
HESS, H. H . Interpretation of Geological and Geophysical Observations, The
Navy-Princeton Gravity Expedition to the West Indies, 54 pp., 1933.
HOLTEDAHL, O. Some Remarkable Features of the Submarine Relief on the North
Coast of the Varanger Peninsula, Northern Norway, Avb. Norske Vid. Akad.
0510, I Mat.-Nat. K1., Vol. 12, 14 pp., 1929.
The Submarine Relief off tbe Norwegian Coast, Norske Vid. Akad. Oslo, 43
pp., 1940.
JOHNSON, D. W. Origin of Submarine Canyons, ]. Geol., Vol. I, 1938, Vol. 2,
1939. (Also: Columbia University Press, 126 pp., 1939.)
KU&NEN, PH. H. Geological Interpretation of the Bathymetrical Results, The
Snellius Expedition, Vol. 5, Geological Results, part I, 123 pp., Brill, Leyden,
1935.
Experiments in Connection with Daly's Hypothesis on the Fonnation of Sub-
marine Canyons, Leidsche Geol. Med., Vol. 8, pp. 327-351, 1937.
Density Currents in Connection with the Problem of Submarine Canyons,
Geol. Mag., Vol. 75, pp. 241-249, 1938.
Two Problems of Marine Geology, Atolls and Canyons, Verh. Kon. Ned. Akad.
Wet. Afd. Nat. 2e Sect., Vol. 43, 3, 69 pp., Amsterdam, 1947.
Turbidity Currents of High Density, Int. Geol. Congr., 18th Session, 1948.
LEWIS, R. G. The Orography of the North Sea Bed, Geogr. ,., Vol. 86, pp. 334-
342, 1935.
MOLENCRAAFF, G. A. F. Geologie, in De Zeeen van Ned. Oost Indie, pp. 272-357,
Brill, Leyden, 1922.
RICH, J. L. Review of Johnson'S book, J. Ge,..ol., Vol. 49, pp. 107-109, 1941.
SHEPARD, F. P. Glacial TroughS of the Continental Shelves, ,. Geol., Vol. 39,
pp. 345-360, 1931.
"Salt" Domes Related to Mississippi Submarine Trough, Bull. Geol. Soc. Am.,
Vol. 48, pp. 1349-1362, 1937.
Non-Depositional Physiographic Enviroruncnts off the California Coast, BUll.
Geol. Soc. Am., Vol. 52, pp. 1869-1886, 1941.
Imaginary Submarine Canyons, Science, Vol. 98, pp. 208-209, 1943.
Sulmutrine Geology, 348 pp., Harper, New York, 1948.
BIBLIOGRAPHY 531
SHEPAlID, F. P., and C. N. BI::"RD. Submarine Canyons; Disrribution and Longi-
tudinal Proliles, Geogr. Rev., Vol. 28, pp. 439-451, 1938.
SHEPARD, F. P., and K. O. EMERY. Submarine Topography off the California
Coast, Geol. Soc. Am. Spec. Paper, No. 31, 171 pp., 1941.
SMITH, P. A. Submarine Valleys, U. S. C. G. S. Field Eng. Bull., No. 10, pp. 15()'-
158, 1936.
The Submarine Topography of Bogoslof, Geogr. Rev., Vol. 27, pp. 625-636,
1937.
Lands beneath the Sea, Sci. Monthly, Vol. 53, pp. 393-409, 1941-
SPRIGG, R. C. Submarine Canyons of the New Guinea and South Ausrralian
Coasts, Trans. Roy. Soc. S. Austr., Vol. 71, pp. 296-310, 1947.
STETSON, H. C. Geology and Paleontology of the Georges Bank Canyons, Pt. 1,
Geology, Bull. Geol. Soc. Am., Vol. 47, pp. 339- 366, 1936.
STETSON, H . C., and J. F. SMITH. Behavior of Suspension Currents and Mud
Slides on the Continental Slope, Am. J. Sci., Vol. 35, pp. 1-13, 1938.
TERRY, R. A. Notes on Submarine Valleys off the Panamanian Coast, Geogr. Rev.,
Vol. 31, pp. 377- 384, 1941.
UMBGROVE, J. H. F. De Koraalriffen der Duizend Eilanden (Java-Zee), Wet. Med.
Dienst Mijnb . Ned. Indiif, No. 12, 47 pp., 1929.
The Pulse of the Eart}" 358 pp., Nijhoff, The Hague, 1947.
VEATCH, A. C., and P . A. SMITH. Atlantic Submarine Valleys of the United States
and the Congo Submarine Valley, Geol. Soc. Am. Spec. Paper, No.7, 101 pp.,
1939.
c H A p T E R E G H T
Eustatic Changes
of Sea Level
532
RECENT EUSTATIC MOVEMENTS 533
but one side testifies to their not being shifted under present condi-
tions. They appear to occur in narrow strips parallel to the coast.
Bourcart interprets them as ancient beach conglomerates. In many
cases they are associated with a break in slope. While most occur-
rences belong to the edge of the present continental terrace and can
be attributed to Pleistocene changes of sea level, there are also occur-
rences in depths of 200 to 500 m and even more. Bourcart believes
that they may be the consequence of a Late Pliocene regression but
prefers to assume warping of the continental border. Although there
is much to be said for this opinion, the fact that the boulders are nOt
covered by recent sediment needs explanation. Until this question
has been solved we must reserve a definite conclusion on the meaning
of these boulder tracts. According to Stearns (1945) an intermediate
level at 20-m depth is also indicated on Pacific coasts. This might
represent a stage in the retreat of the land ice, when sea level must
have remained approximately stable for some length of time.
In conclusion attention may be drawn to a paper by Hoffmeister
and Wentworth (1939), who point our the complicated processes at
work in the production of benches on rocky coasts and the difficulties
encountered in ascertaining the height of sea level at the time such
features were developed. In some benches wave attack dominates;
in others, solution or organic activity plays a major part. The nature
'of the rock undergoing erosion is highly variable, and in consequence
the level at which the bench is cut or the beach is thrown up varies
within wide bounds. The authors cited have found, from personal
experience, that prolonged application to the subject in the field is
essential to the collecting of trustworthy data.
One serious omission in this otherwise excellent discussion is that
of benches formed at low tide level in limestone, especially cora lifer-
ous limestone, by means of solution in the upper stratum of the sea
(chemical marine erosion). The great importance of this activity,
which gradually curs off the emerged part of elevated atolls and other
limestones, was discussed in Chapter 6.
L..-_ _ _-...J ~
AMOUNT OF WATER ON EARTH 543
leveling, whether isostasy is maintained or not, would cause a eustatic
rise of only 250 m. Neither the distribution of depths nor the posi-
tion on the sea floor of the denuded material would influence this
figure.
The smallness of the figure is surprising. Sea level might, of course,
rise more than this amount with respect to the level of the continents,
if the sea floor were to be raised and the land to be depressed with
respect to each other by some internal cause. But external agents,
that is base leveling of the land and sedimentation in the sea, could
not result in a larger shift of sea level than the 250 m mentioned.
Umbgrove points out that the eustatic movements of the Pleistocene
amounted to three times the value assumed by the present writer for
earlier eustatic movements. He concludes that a larger amount should
be assumed for the earlier swings of sea level. But Umbgrove, like
loly, does not bear in mind that for most of the geological past the
continents were much flatter than at present and at a slightly lower
level in consequence of there being no land ice. During those times
smaller swings of sea level would result in far greater transgressions
and regressions than those of the Pleistocene. In the Pleistocene sea
level started to sink from an already regressive stage when shallow
seas were at the minimum. On the comparatively steep slopes of the
emerged continents a further reduction of sea level could bring about
but minor changes in coastal configuration.
The next step in our inquiry must be to calculate the amount of
water that would have to be added to the sea to cause a transgression,
or subtracted for a regression. For a rise of 50 m, roughly 20 X 10 6
km3 is required. This amount must be added to or abstracted from
the volume of water, or the cubic content of the ocean basins, in about
20 to 30 million years to explain the pulsations deduced by stratigraph-
ers for the Paleozoic Era.
Having ascertained roughly the magnitude involved in the average
eustatic variations of the past, we are now in a position to consider
how this change in volume may have been brought about. Several
explanations can be suggested.
SEDIMENTATION
Suess attributed transgressions to the influence of sedimentation, his
contention being that the dumping of sediments into the sea must
cause a rise of level. Grabau () 936) is less explicit. When considered
more in detail the process is found to be complicated. To start with,
a distinction must be made between sedimentation in the deep sea and
that in geosynclinal troughs, as the effects are opposite in the two
locations.
Sedimenta tio n in Geosynclines. The normal type of geosyncline is
a large trough, the bottom of which sinks while sedimentation con-
tinually keeps the hollow brim full. Whether we ascribe the sinking
to horizontal or vertical forces, or to the combined action of isostasy
coupled with alterations in the paramorphic zone, and whether the
water depths show marked variations or not, the outcome must always
be loss of water from free circulation. This loss is in the pore space
of the detrital rocks. All that sedimentation does from our point of
view is to reduce the inflow of water into the newly forming depres-
sion.
To prove this contention it should be borne in mind that alterations
in the shape either of the emerged part of the continents or of their
parts submerged in the sima have no influence on the level of the con-
tinent. They merely form a shifting of the load. But when a defor-
mation takes place such that the sea can flood the new depression, the
continent will sink deeper into the sima in the neighborhood. T he
consequent slight rise of the sima elsewhere merely floats up the crust
with respect to the center of the earth but not with respect to the
surface of the sima. The resu lt is that the area covered by water is
enlarged, and, owing to the constant total amount, the depth of the
oceans is decreased. This fina lly results in a decreased sinking of the
remainder of the continents into t he sima. As net outcome of the
whole process there results a slight increase of freeboard of the con-
tinents outside the geosynclinal area (Fig. 246).
To cause a loss of 1,000,000 km s of water, a volume of 5,000,000
km s of sediments is required (assuming an average pore space of
20%). A geosyncline w ith an average depth of 5000 m and an area
of (2500 X 400 =) 1,000,000 km 2 is needed for this amount. In other
words, we require 20 such geosynclines forming in 20 million years
546 EUSTATIC CHANGES OF SEA LEVel
Geosynclinal sedimentation
FIG. 246. Diagram illustrating the opposed effects on sea level of geosynclinal
development and deep-sea sedimentation.
BASIN FORMATION
Umbgrove pointed out that the deep basins of Indonesia, the West
Indies, and the Mediterranean have probably sunk away in "recent"
times and that the formation of these new receptacles might constitute
a cause of regression.
A rough estimate of the volume of water contained in these basins
and troughs within the Indonesian archipelago is 6,000,000 km 8 As
there were basins there before the recent subsidence, and as elevation .
has also taken place, the whole of that amount is not available for
causing a regression. The maximum estimate would be 5,000,000 km 8
This would cause a eustatic shift of only 13 m.
The total volume of all the Indonesian and West Indian basins, the
Mediterranean, and the East Asiatic basins is roughly 27,000,000 km 8
It does not appear improbable that some are to sink still further, as
many are characterized by strong positive anomalies of gravity. On
the other hand they are not pure gain. If we assume, for the sake
of argument, that they are all of recent origin, then their combined
influence would cause a eustatic depression of sea level of 60 to 80 m.
We have at last found an influence that is at least of the correct order
of magnitude.
It is a different question, however, whether this mechanism can be
relied upon to explain repeated eustatic movements. In the first place
it must be assumed that there was not a corresponding rise of the
ocean floor elsewhere. In the case of oceanic deep-sea troughs one
would expect such a rise of neighboring parts of the sea floor, and for
this reason these troughs have not been added to the estimated volume
of the possible recent sinkings. In other words, most of the mass
abstracted from below the floor of the basins must have flowed in
below the continents, causing them to rise if a fall of sea level was to
result. In the second place the youth of all or most of the basins
548 EUSTATIC CH AN GES OF SE A lEVEL
this book to enter into the merits of this hypothesis. For those who
are inclined, like the present author, to reject joly's hypothesis, yet
another cause must be sought for. Possibly the conception of con-
vection currents in the substratum affords an explanation.
The investigations of gravity at sea by Vening Meinesz have shown
that there exists a general positive anomaly over the oceanic basins
with a deficiency over the continents. The difference is of the order
of 30 milligals, corresponding to the attraction of a layer of rock
about 300. m thick. The strength of the crust is insufficient to bear
this weight, and Vening Meinesz finds that the existence of convec-
tion currents in the substratum is the only hypothesis that can account
for the phenomenon.
It is obvious that, if this mechanism of convection currents includes
variations in magnitude, these currents may be sufficient to cause major
eustatic movements. To work out this line of thought in detail is a
separate problem that will not be attempted by us.
Our review of possible explanations of the eustatic swings of the
past has led us to the unexpected conclusion that even the lesser shifts
cannot be accounted for by processes operating either in or at the
surface of the crust. As external processes and diastrophism are in-
adequate, we are forced to assume that only some pulsating, sub-
crustal influence, such as alternating convection currents, can be of
sufficient magnitude to explain the rhythmic rise and fall of sea level
during the geological past.
A few words may be added on the position of sea level through
longer periods. When a folded geosyncline is raised and gradually
eroded, the contained water is brought back into circulation. Part of
the water in the migmatized zones is given off as "juvenile" magnetic
exhalations.
In the long run the sedimentation in the deep sea and the addition of
truly juvenile water might cause a small rise of sea level. Our esti-
mates would lead to an addition of about Y7 km s per year, or 175 m
since the beginning of the Cambrian. This increase is shown on the
diagram, Fig. 77, p. 130.
On the other hand orogenesis is generally supposed to thicken the
continents, thus giving them more freeboard. Intrusion of differen-
tiated acid magma would have the same influence. The fact that since
at least the beginning of the Paleozoic the continents have remained
on the average at the same level as compared to sea level proves that
the net outcome of all these processes must be to compensate each
other. To correspond with a deepening of the oceans by 175 m, the
550 EUSTATIC CHANGES OF SEA LEVel
.LAND
.-.
200m.
hF)',~ <H 2002000""
1000""
2000-<4000",-
3000""
04000 - 6000 m,
15000 M\,
_ :>6000 tn,
~III
TeRTIARY BASINS OF SUBSIDENCE AND SEDIMENTATION
(IDIOGEOSYNCUNES); MOOERATE fOLDING TOWARDS THE END Of
THE PLIOCENE
r:"777:IlIIa SUBMARINE RIDGES IN III
!L.L.:J
~ III b SUBMARINE CONTINUATION OF III
l"XX'l,.)( IV REGIONS THA"L WERE ABOVE SEA-LEVEL DURING THE. GREATEST
L..!......J PART Of THE TERTIARY r fARlY-CIMMERIAN fOlDING)
V ~f~h~~SA~~E:O~ri~N~TERTIARY STRATA SUffERED NO OR
VOLCANOES
.....
'.' .
. .
... :_.:..... ... .. .
REGIONAL ISOSTATIC . . ' ., ' .. ','
_
ANOMALIES
-250 to-200mllfisal
.... .
_ -200. -150 "
.... .. "
_ -ISO. -100
m. -100. - 50
IIIIIIlIIIJ - 50. 0
CJ 0 5 0 .
t:."':'J SO " + 100
" ...':',.
f:1.:~.:tl -1-100 " + 150 . . . :...
"
......
.. .
"
: ', " ' . , 0"', ~ . . '..
PLAu B. Geology and gravity field of Indonesia. (Alter Umbpove. 1947, Plate 8.)
Index
Pages in italic type indiclrtt illustrations.
Abrasion, see Marine erosion Aquafacts, 282
of particles, 277-282 Aragonite mud, 217, 341
Abraumsalze, 221 Archanguelsky, 246, 298, 335, 409
Abyssal-benthic zone, 315 Arctic Basin, 27, 97
Abyssal enviromnents, 313 crust below, 121, 122
VI. pelagic environments, 313 topographic forms of, 119
Abyssal sediments, 347 Arcric icebergs, 26
Adiabatic changes of temperature, 23 Argand, 185, 186, 200, 204, 205, 208
Adjacent seas, 97 Armstrong, 422, 478
Adriatic, color of water in, 23 Ash, see Volcanoes
Agulhas Bank, 219, 252 Atlantic, 104, 378, 381, 382
Ahlmann, 534 age of, 125
Airy, 84 color of water in, 22
Aleutian Trench, 100, 109, 110 crust below, 122, J29
Allen, 86 currents in, 33, 34, 38, 43
Amazon River, 41 depth of, 107, 119, 121
bore in, 62 icebergs in, 25, 27
Anaerobic conJitions, see Euxinic en- lime in sediments of, 356
vironmems linear patterns of, 148
Anchibenthic zone, 315 ocean swell in, 77
Anchor stations, 5 origin of, 135
Andesite line, 11 7, IJ 8, 151 plankton of, 16, 216
Andree, 97, 239, 245, 298, 409 salinity of, 15, 20
Angular unconfoTlnity, 308 soundings in, 96
Animals, action in depositing mud, 318 temperatures in, 24
burrowing and mud-feeding, 217, 367, topographic foons of, 9), 119, J2I
373, 406 Atlantis, 131
rafting by, 277 Atmosphere 3S source of sediment, 211 -
Antarctic circumpolar current, 41 214
Antarctic convergence, 36, 37 Atolls, 398, 42S, 427, #1, 44f, 4)7, 4)8,
Ancarctic icebergs, 26, 27 459, 464, 468, 469, 471, 472, 473
Antarctic Ocean, 97 antecedent-platform theory of, 454-
currents in, 37 455
icebergs in, 26, 27 elevated, 438, #9
plankton of, 216 glacial-control theory of, 459, '163
Antarctic water, 35 glacially controlled subsidence theory
Antecedent platform theory of atoUs, of, 460, 461, 474
454-455 marginal-belts theory of, 460
Appalachia, 126, 146 principal theories of, 453-469
554 IN DE X
Atolls, subsidence theory of, 455, 456 Beach sandstone, 4:H, 434, 435
Atterberg, 211 Beaches, 69, 83, 84, 267, 268, 332
Austausch coefficient, 29 abrasion on, 277
Austin, 266, 298 concentrations on, 186
slope of, 173
Backvvash, 83, 270, 291 sotting on, 285-287
Backwash ripples, 292 Beard, 531
Bacteria, 218, 220, 407 Beers, 375, 409
anaerobic, II, IJ Bects, 241, 298
Bagnold, 227, 270, 286, 298 Bell, 510, 530
Baile)" 239, 298, 366, 409, 530 Bemmelen, van, 146, 170, 194, 248
Baldrey, 242, 298 Bennett, 400, 413
Balls, see Sand banks Bering Sttait, 97
Baltic, 97 shelf of, 105
currents in, 30 Betz, 152, 170
salinity of, 12, 19, 20 Bigelow, I, 28, 69, 86
tides of, 60 Bijlaard, 194
ventilation of, 13 Bikini, 421, 440, 456, 458, 467, 477, 508
Banda Arcs, see Moluccas, deep-sea Birdfoot delta, 326
troughs and ridges of Bishof, 220
Banks, 98, 306 Bjerknes, 28
Barbados, 126 Black. 299
Barrell, 114, 170, 234, 309, 409 Black Sea, 12, 44, 98, 335, 374, 401
Barrier reefs, 42'1, 4H, 455, 4S6, 466, 474 rate of 'Sedimentation in, 380
Bars, see Sand banks slumping in, 246
Barth, 298, 409, 410 tides of, 60
Bardett Trough, 98, 201 ventilation of, 9, 12, 13
Bartrum, 538 Blue mud. 218, 338, 3SJ, 358, 359, 369;
Basins, S, 10, II , 12, 14, 47, 98 see also Terrigenous sediments
alterations in shade of, Ill- lIS Boat channels, 426
filled with sediment, 197 Boggild, 342
Borderlands, lOS, 126, 146, 169; see also
in.Buence of formation on sea level,
Continental borderlands
547-548
Bore, 62, 83
of the oceans, 119
Born, 144, 146, 170
poorly ventilated, 9, 10, 11, 12, 13, 14, Boschma, 420
311; see also Euxinic environments Bosporus, circulation in, 44, 'If
temperatures in, 23; see also Moluccas, Bottom contours, 95
circulation in basins of Bottom erosion, see Erosion
with deep entrance, 46-56; lee a/so Bottom friction, 29
Moluccas, circulation in basins of Bottom load, 258
with shallow entrance, 42-46 Bottom samplers, 6-7
Bathyal environments, 313, 316, 321- reduction of core length in, 383
322 Bottom-set beds, 324
Bathyal sediments, 338 Bottom transport by waves, 267- 276
Bay of Biscay, 235 Boulders, 281, 428, 4;0, 443, 447
Bay of Fundy, 61 abrasion of, 281
Beach cusps, 293, 294, 295, 296, 297 moved by storm waves, 222
Beach drifting, 274, 27f, 276 shape of, 281
INDEX 555
Bourcare, 87, 90, 106, 155, 158, 170, 227, Centrifugal force, 28
234, 298, 318, 331, 334, 337, 409, Challenger, 3, 356
495, 499, 530, 539, 550 Champion, 6, 87
Bowie, 132, 170 Chlorinity, 15
Bradley, 246, 263, 265, 298, 372, 373, Chun, 87
409 Clarke, 87, 387, 389, 390, 410
Brajnikov, 409 Clay, submarine conversion of, 237
Bramlette, 246, 263, 265, 298, 372, 373, VI. pelice or lutice 211
409 Clements, 150, 170
Breakers, plunge point of, 269; see also Cliffs, 223
Wa,' es on headlands, 225
Bristol Channel, 61 Climate, influence on sedimentation,
tidal currents of, 63 403-405
British Channel, floor of, 110 Climatic facmcs causing currents, 28, 32
Brongersma-Sanders, 217, 298,403,410 Cloos, 108, 129, 140-141, 153, 157, 170
Brouwer, 182- 185, 208 Coarse sediment, on beaches, 284
Bro,vn, 242, 298 on seamounts and shelf edges, 263
Bruckner, 368 rounding of, 280
Bubnoff, 199 Coastal drifting, 273, 274-
Bucher, liS, 127, 129, 140, 142, 146, 165, Coastal erosion as source of sediment,
170, 197,201,203,206,208, 501,502, 221 - 232, 22J, 224, zzr
530, 544, 550 Coastal erosion VI. rivers as source of
Buckling of crust, 189 sediment, 233, 234
Bullard, 156 Coastal terrace, 226
Burrowing and mud-feeding animals, Coastline, shape of, 223
217, 367, 373, 406 Coasts, classification of, 99
Byerly, 122 Coccolith oozes, 352
Cocos Keeling, 440, 441
Cailleux, 280, 298, 333, 410 Coker, 87
Calcareous mud, 21 7, 341 Collet, 298, 530
Calcareous ooze, 349 Compound ripples, 292
Calcium, tOtal amount in sediments, 392 Concentration on beaches, 286
Calcium carbonate, Iee Lime Conductivity, 21
Caldera, submarine, 152 Constructive waves, 270
Canary Current, 40 Continental borderlands, 105. 158, 162,
Oapacit)' of currents, 158 339
Cape Lopatka, tsunamis of, 80 Boor of, III
Carbon, 8 organic matter in deposition of, 401,
Carbon dioxide, 7, 8, 56 402
Carbonic acid, 8 Continental drift, 125, 126-129, 128, 13r,
Caribbean Sell, 381 178, 399
Car ion eaters, 13 in Indonesia, 177-182
Carnegie, 384 Continental shelf, 104, 304
Cascadia, 126 coarse sediment on, 26J
Caspian, 46 data on, 154
Catastrophic waves, see Tsunamis See also Continental terrace
Cay sandstone, 433, 434, 4Jr Continental slope, lOS, 154; see IIlso
Cayeux, 316 Continental terrace
Celebes Sea, 47, 51, 356 data on, 155
556 IN 0 EX
Continental terrace, 1M, 153-170, 30S Corals, rypes of, 416
data on, 153,15'1, lS5, If6, If7 Corers, 6, 7, 383
disappearance of, 168 Coriolis force, 28, 30 31, 32, 53, 65
oil in, 168 Cornish, 87
origin of, 157-159, 160, 161-162, 163, Correns, I, 28, 87, 213, 217, 237, 257,
164, 167 263, 298, 348, 349, 410
See alro Continental Shelf Cox, 88
Continents, extension of, 144--147 Crary, J7J
origin of, 115-153 Cross bedding, 308, 363, 364
Coprolites, 218 Crust of the earth, 116, 121, 122, 135,
Coral mud, 341 138, 149
Coral reefs, 423, 427, 428 Cumulative curves, 283
almost-atolls, 426 Current, Labrador, 40
as geological tide gauges, 448, 540 Current bedding, 308, 363, J64
bank-inset reefs, 426 Current erosion, 229-232
barriers, 424, 'I)], 455, 456, 466, 474 Current ripples, 290
beach and cay sandstone, 433, 434, 435 Currents, 28- 56
boar channels, 426 as acti~e in submarine canyons, 502
cays, 428, 429, 430, 434, 435, 440--442, as transporting agents, 251 - 267, 350
443, 449 capacity of, 258
classification of, 423-430 causes of, 28
drowned reefs, 426, 451 combined with waves in transport,
faros, 42), 426, 461, 467, 470 262
foundations of, 420, 475-478 influence on coral reefs, 444, 445, 446
fringing, 423, 424 jet, 40
influence of currents on, 445, 446 longshore, 86
intluence of sediment on, 438-440 primary forces causing, 28, 30
influence of wind on, 440, 'I'll, 442, secondary forces causing, 28, 30
443, 444, 447 surface, 37-42, 42/43
passages through, 425, 461, 466, 468, velocity of, 251-252
473
ventilating, 9, 48, 49, 50, 52, 53, 54,
pinnacles and lagoon reefs, 42 f
215; see also Basins, poorly venti-
ramparts 428, 430, 443, 447
lated
reef flats, 426
wind -driven, 31, 38, 6J
reef patches, 426
seaward slopes of, 427, 458 Cusps, 293, 294, 295, 296, 297
shingle on, 420, 428, 4JO, 442, 443, #7 Cyclothems, 368, 373
uplifted reefs, 193, 426, 433, 438, 457,
459 Daly, 115, 116, 125, 134, 139, 158, 170,
Su also Atolls; Lithothamnion 197, 208, 220, 266, 394, 448, 459, 461,
Coral sand, 341 462, 463, 465, 475, 478, 498, 509,
Corals, biological data on, 414-423, 415 511, 520, 530, 537, 538, 539, 550
depth of growth of, 417 Dana, 147, 195, 455
foundations of, 419 Dangeard, 110, 170
influence of silt on, 418 Dardanelles, circulation in, 44
oxygen requirements of, 418 Darwin, 132, 147, 397, 448, 453, 455, 461,
rate of growth of, 420-421 473, 475, 478
saliniry requirements of, 4 J8 Da,is, 87, 439, 453, 460, 465, 473, 47f
temperature requirements of, 416 477, 478, 502, 530
INDEX 557
Decomposable organic matter, see Or- Discordant basins, 197, 198
ganic matter Dixon, 241, 299
Deep sea, circulation in, 33-37 Doeglas, 408, 410
influence on sea level of sedimenta- Dolomite, 221
tion in, 546-547 Doodson, 59
topography of, 108 Dredging, 7
Deep-sea basins, 98, 200 Dreher, IB, 171
Deep-sea deposits, 145, 315 Drew, 220
clay minerals in, 348, 351 Drift currents, 29
diastems in, 309 Drowned glacial troughs, 483, 484, 485
fossil, 126, 192, 204, 376, 394 Drowned river valleys, 481,482/483,506
meteoric dust in, 212 Dunbar, 245, 266
pelagic character of, 203 Du Toit, 125, 128, 171, 182, 203, 208,
radioactivity of, 374 499, 530
rate of accumulation of, 378
See also Abyssal; Pelagic Earthquakes, 120, 121, 122, 150
Deep-sea depressions, 102, 201 as cause of submarine slides, 247
classification of, 199, 200-201 Echographs, 4
Deep-sea expeditions, 3 Echo soundings, 91, 92
Deep-sea sands, 360 Eddies, see Gyrals; Turbulence
deposition of, 240 Eddy vi cosity, 21, 29
Deep-sea sediments, thickness of, 396 Edmondson, 69, 86
Deep-sea troughs, 98, 100, 101, 102, 109, Ehrhardt, 410
110, 199, 201, 202 Ekman, 28, 29
in Indonesia, 175, 199, 201, 202 Elbert, 176
compared to geosynclines, 195- 208, Emery, 6, 7, 87, 108, 111, 150, 152, 162,
202 170, 171, 173,237,247,248,277,287,
Deep-sea U-troughs, 200, 205, 206 299, 301, 383,410, 412, 425, 477, 478,
Deep-sea V-troughs, 199, 200, 201, 202, 502, 531
204, 206 Eniwetok, 104, 427, 477
Defant, I, 28, 29, 56, 69, 87 Environments, classification of, 313, 314
Delta environments, 324-BO
definition of, 312
Deltas, 161, 324, 325, 326, 327
Eolian sedimentary matter, 213
Demarest, 298
Density currents, 32, 238; see also Epeiric seas, 97
Turbidity currents Epicontinental seas, 98
Deposition, minimum current velocities Equilibrium, profile of, 226, 302-310,
for, 260 305, 406
of sand and mud, 317-319 Equivalent diameter, 250, 251
Deposits, organic content of, see Or- Erosion, as source of sediment, 1.21-232
ganic matter by currents, 229
Depth curves, 94 by waves, 221, 227
Destructive waves, 270 minimum current velocities for, 260
Diastems, 308, 309, 310 of coral limestone, 435-438, 452, 461
Diatom ooze, description of, 357 Erosion platfonns, 226
Dietz, 108, 170, 219, 237, 298, 299, 383, Escher, 132, 171, 194, 245, 410
410 Eskola, 298, 410
Dircksen, 410 Estuarine environments, 330-331
Disconfonnity, 308, 310 Eugeosynclines, 197, 205
558 INDEX
Eulittoral zone, 315 Frequency curve of crustal elevation,
Eupclagic, see Pelagic see Hypsographic curve
Euphotic zone, 8, 15 Friction depth, 29
Eustatic movements, see Sea level Friedlander, 147, 171
Euxinic environments, 10, 11, 215, 217, Frosted sand, 214
218, 317, 335- 336, 401; see also Funafuti, 422, 423, 431
Basins, poor! y ventilated
E"ans, 288, 290, 291, 294, 297, 299, 374, Galle, 29
375, 376, 410 Gangcbin, 248
Evaporation as cause of precipitation, Gardiner, 417, 420, 448--450, 478
220 Gases in sea water, 7- 9
Ewing, 7, 87, 156, 162, 171, 174 Gaskell, 156
Geoeconomy, annual amount of detrital
Facies, definition of, 312 sediment, 233
Fairbridge, 125, 171, 241, 265, 299, 429, cubic content of oceans, 124, 129, 130
436, 450, 470, 478, 479 drowning of guYOtS, 397-399
False bedding, 308, 363, 364 estimate of marine erosion, 113
Faros, 425, 426, 461, 467, 470 eustatic shifts, 540-550
Faults, submarine, 108, 521, 522 of calcium, 392
Ferrous sulfide, 338 of carbon, 7
Fetch, 75 of manganese, 390, 395
Field, 298 of phosphorus, 390
Fiji Islands, 424, 425, 466, 470, 472, 474 of silica, 233, 395
Fiords, ventilation of, 12 of sodium, 388
Fischer, 132, 163 of titanium, 390
Fleming, 2, 28, 87, 88, 173, 229, 252, of volcanoes, 388
Jill, 478, 535, 539, 550 ratio of sediment types, 387
Florida Current, 39, 41, 306 size of continental terrace, 166
Flotation, 287 total volume of sediment, 386, 388
Foraminifera, 319, 381, 535 See also Sea level
as indicating depth, 316 Geosynclines, 98, 109, 111, 112, 114, 124,
145, 190, 191, 307, 546
as indicating temperatures, 369
compared to deep-sea troughs, 195-
as lime producers, 394
208, 202
as source of sediment, 210, 215
definition of, 195- 197
rate of production of, 380 fail ure to rise above sea level, 207
See also GJobigerin3; Lime rate of sedimentation in, 385
Foredeeps, 114, 198 recent, 202
F orel, 25, 59 subdivision of, 197
Fore-s t beds, 324 volume of sedimentation in, 545, 546
Fossil deposits, interpretation of, 315, Gerstner, 69
323, 333, 408-409 Geszti, 134
Fossilization, 13 Geyer, 171
Fourmarier, 125, 171 Gilbert, 163
Francis-Boeuf, 318, 331, 409, 410 Gill, 450, 478
Franklin Shore, 1S5, 539 Gilluly, 544, 550
Fraser, 410 Glacial control theory of atolls, 459,
Fraser River, 233 463
Freezing point, 19, 20 Glacial marine sediments, 235, 345
I NDEX 559
Glacial troughs, drowned, 4t15 H awaiian Islands, 147, 152, 153
Glacially controlled subsidence theory tsunamis of, 80
of atolls, 460, 461, 474 Heim, 220, 245, 299, 351
G laciers as sources of sediment, 235- 236 Helland-Hansen, 29
G laessne r, 171, 196, 208 Helmholtz, 73
Glangcaud, 25 9, 299, 331, 410 Hemipelagic-abyssal environments, 316,
G lauconite, 346 320--321
formation of, 218 Hemipelagic environments, 313
Glauconite mud, 346 Hemipelagic sediment, 337-347
Globigerina mud, 342 Hentschel, I , 87
Globigeri na 007.e, 319, 349-357 Herdman, 410
anrage composition of, 389 Hess, 100, 103, 117, 144, 152, 170, 171,
radioactivity of, 375 195, 207, 208, 397,410, 477, 478,498,
rare of accumulation of, 377 508, 530
Globigeri nas, settling velocities of, 250 Hiller, 122
Golden Gate, 252 Hills, 139, 171, 393, 41 0
Gondwana, 125 Hinterdecps, 198
Grabau, 80, 197, 202, 208, 335, 4 10, 540, Histograms, 283
545, 550 Hjnrt, 1,88,212,300,370,411
Grade scales, 21 I Hjulstr~jm , 229, 259
Graded bedding, 240, 242, 366 Hobbs, 11 7, 171, 186, 189, 194,208
Grain size, sorting according to, 282- Hoff, Van't, 220
285 Hoffmeister. 422, 438, 453, 455, 472 , 475,
Grain sizes, classification of, 21 I 478, 479, 540, 550
Grant, 234, 299, 301 Hiigbom, 131, 171
Gra\'i ty, 189 Holredahl, lOS, 158, 171, 485, 530
action in transport, 238-248, 268 Horn, 245
in tecton ics, 248 Huangho deposits, 335
Sec also Isostasy Hudson Bay, 97
Green, 147 Hydrogen sulfide, 10- 14, 335
Green mud, 346 Hydrotroilite, 338
Greenla nd, icebergs, 25 H ypsographic curve, 134, 163-170, 16J,
Grunau, 245 , 299 16<#, 541
Gulf of California, 374, 380
Gulf of Guiana, 213, 350 Ice, 20
Gu lf of Mexico, circulation of, 39, 44 as source of sediment, 235-236
G ulf Stream, 26, 39,40, 41, 155, 159, 251 Ice Ages, 381
G utenberg, 115, 120, 123, 125, 129, 134, as cause of submarine canyons, 503
150,151, 171,533,534,550 causing stratification, 369 .
Guyots, 98, 10J, 104, 306, 397-399, 477, influence on shelf sediments, 264, 266
J08 See also Sea level, glacial eustatic
Gyuls, 40, 45 movements of
Ice Barrier, 26
Haarmann, 248 Icebergs, 25, 26, 27
Halimeda, 422 Idiogeosynclines, 198
Hall, 195, 202 Indian Ocean, 378
Hiintzschel, 299, 407, 410 age of, 125
Hatch, 299 crust below, 122, 142
Haug, 112, 124, 133, 171, 195, 197, 208 currents in, 36
560 INDEX
Indian Ocean, depth of, 107, 120 Kermadec-Tnnga Troughs, 100
icebergs in, 27 Kindle. 272. 300
origin of, 135 King. 110. 171
salinity of, 15 Kip, 375, 410
temperatures in, 24 Kirsch, 129
topographic forms of, 119, 143 Knothe. 131, 172
Indonesia, 414-479, 547 Knudscn, 87
age of deep-sea depressions of, 192 Kober, 133. 172
continental drift of, 129, 178 Kossina. 113, 172
salinity of sea water, 19 Koszmat, 548, 550
shelf of, 105, 482/483 Krakatoa. 212
tides of, 67 . Kratogenic areas, 144
tsunamis of, 80 Krempf,44O
\'olcanic slopes of, 523, 524, 525, 527, Krenkel, 136, 140. 172
528, 529 Krumbein, 211, 249, 250, 300
See a/so Moluccas Krummel, 2, 87, 97
Ingram, 299 Kuenen. 53, 55, 87, 96, 102. 104, 108. 124.
Inland seas, 97 129. 145. 147, 153, 157, 172, 173 , 176.
Innensenken, 198, 205 179, 182, 186, 189, 191 , 194, 197-208,
Internal heat, SS 209,217,239. 241, 244, 247, 265, 282.
Internal waves, 8 291,294,300,366, 373, 379, 389,401,
Intraformational distortions, see Slump- 411. 434, 436, 437, 439, 44<5, 449, 452.
ing 453, 460, 461, 462, 466, 478, 510, 511,
Intramontane troughs, 197, 198 526, 530, 533, 534, 540, SSO
Iron, 14 Kullcnberg, 6. 87, 383
Iron-sulfur compounds. 13 Kuroshio, 41
Isostl\.6y, 118, 176, 188, 189, 206, 396,476,
537, 538 Labile shelfs, 199
in continental terrace, 161 Labrador Current, 26, 41, 381
subsidenQe of deltas, 327 Ladd, 422, 438, 453, 455, 472, 475, 478,
Isthmian links, 125
479
LaFond, 301
Jacquet, 298
Lagoon environments, 331-332
Japan. tsunamis of, 80
Lake Geneva, seiches in. 59
Jeffreys, 72
Johnson, 2, 69, 87, 88, 99, 171, 173, 221, temperatures in, 24
294, lOO. 301, J 11 , 486, 500, HO, H8, Laminae, 363
sso Laminar flow, 21
Johnston, 220, 411 Landslides, submarine, 241
Joly, 541, 543. 548, 550 as origin of submarine canyons, 500
Jones, 24 1, 244, 299 See also Slumping
Lawson, 141, 172, 498. 524
Kaisin, 220 Leet, 172
Kaoe Bay, 14, 372, 401 Lek, 46, 54, 87
ventilation of, 9, 12, 13 Length of fetch, 75
Kara Bougas, evaporation in, 45 Leuchs, J97, 208
Kay, 196, 197, 208 Lewis, 270, 482, 530
Kelp, rafting by, 277 Lime, 14, 233
Kelvin, 73, 90 average in sediments, JB, 35J
INDEX 561
Lime, in oozes, 349-357 Marine erosion platform, 223, 224, 22f,
in shelf sediments, 33 7 226,234
precipitation of, 217, 219 Marining, 335
solution of, 320, 321, 354 Marmer, 551
See also Geoeconomy of calcium Marr, 300, 411
Limestones, 8 Marshall, 278, 411
groundmass of, 220, 394 Mascaret, 62
muine solution of, 435, 452, 461 Mecking, 106, 147, 172
Linear patterns, 147- 150 Mediterranean, 98
Linguoid ripples, 290, 292 circulation in, 42, 43
Lithofacies, 312 salinity of, 12, 19
Litbotblmmion, 423, 430-433 submarine canyons of, 495, 508
biological data on, 422 thickness of sediments in, 392
Litbotbll'11mion ridge, 422, 427, 431, 432 tides of, 60
Littoral environments, 313, 317,332-335 waves in, 76
Lohmann, 377 Mediterranean seas, 98
Longitudinal ripples, 292 Meta ripples, 291, 293
Longshore drifting, 274 MeteoT, 36, 67, 78, 96, 211, 213, 264, 340,
Longshore transportation, sorting dur- 358, 359, 371, 384
ing, 285 Meteoric dust, 377
Lorentz, 64, 87 Migliorini, 239, 300, 366, 411
Lows, see Deep-sea troughs Miller, 171, 243, 300
Lugeon,248 Milner, 300
Lundahl, 300 Minette orcs, 218
Lutitc, 211, 317 Mississippi, 233
in pelagic sediments, 347, 348, 351 Mississippi delta, 326, 327, 385
X-ray analysis of, 343 Mohn,28
See also Mud Mohorovicic, 132, 172
Lyell, 334, 411 Molengraaff, 185, 192, 193, 197, 204, 208,
476, 482, 483, 530, 536
MacDonald, 80, 88 Moluccas, 175-209, 414-479
MacEwen, 88 circulation in basins of, f, 46-56, 48,
Maldives and Laccadives, 417, 420, 42f, 49, fa, 52, H, 54
464, 472, 473 currents in, 68, 229
Manganese, 218 deep-sea troughs and ridges of, 93, 96,
geoeconomy of, 395 98, 102, 175, 177, 178, 180, 182, 184,
total amount in sediments, 390 187, 188, In, 199, 201, 202
Manganese nodules, 360 oxygen in deep water of, 51, 213,
fossil, 376 341, 342/343, 343, 344, 3JJ
radioactivity of, 375 sediments of, 340, 342, 356, 371-373,
rate of growth of, 381 375, 378, 401, 403
Mangroves on coral reefs, 430, 444 submarine volcanic slopes of, 526-529
Mareographs, 533 Monogeosynclines, 198
Marginal-belts theory of atolls, 460 Moore, H. B., 218, 373, 380, 411
Marginal deeps, 197, 198, 200, 205 Moore, R. C., 241, 300, 312
Marginal seas, 97 Mountain roots, 190
Marine erosion, Ill, 234 Mud, definition of, 211
estimate of amount of, 113 deposition of, 317-319
forming of benches, 538, 540 off luger rivers, 155
562 I N 0 EX
Mud-feeding animals, 217, 367, 373, 406 Oceans, area and depth of, 106-108, 120,
Mud flats, see Tidal flats 121
Mud flows, 239 Ochsenius, 46
as origin of submarine canyons, 100 Oena Oena, 212, 213
Mud grains, 318 Oolites, 217, 218, 220
Munk, 76, 81, 88 Ooze, definition of, 211
Murray, I, 88, 100, 172, 209, 212, 300, Organic matter, 8, 13, 335, 399-401, 402,
315, 349, 370, 377, 411 403, 407
Myers, 379, 380, 411 derived from land, 217
determining of, 400
~an en, 29, 485 source of, 216-217
~ational Research Council, Organisms, influence on sedimentation,
atland, 316, 411 406-408
~atural bridges or arches, 225 Oscillation ripples, 288
~eeb, 208, 212, 217, 235, 237, 300, 321, Oscillatory waves, 267
340, 342-345, 349, 351, 372, 378, 379, Oseen, 249
411 Outlier, 225, 225
~egative forms of sea fioor, definitions Owens, 261
of, 97 Oxygen, 7, 12, 14, 18, 418
description of, 100
~egro heads, 428 Pacific, 104, 381
~eritic environments, 313, 316, 322-324 age of, 125
~evin, 249, 260, 262, 300, 411 andesite line in, 117, 118
~ewton, 56 crust below, 120, 121, 122, 129, 133,
~ile, 233 144
~ile delta, 326 curren ts in, H, 36
~itragen, 7 depth of, 107, 119, 120, 121
~itrogen compounds, 14 lime in sediments of, 356
~olke, 133, 172
linear patterns of, 148
~orth Equatorial Current, 38, 40
origin of, 134
~orth Pacific Current, 41
salinity of, 15, 20
temperatures in, 24
~orth Sea, 62, 97
topographic forms of, 101, 102, 103,
currents in, 30, 251
104, 109, llO, 119, 121
sandy bottom of, 228, 229 volume of, 130
shelf of, 105 Pacific coast of United States, 31
submerged river valleys of, 482 Pack ice, 27
tides of, 62, 63 Pangea, 127, 134
waves in, 222 Para ripples, 291, 293
Notches, 223 Parageosynclines, 198, 199
in coral limestone, 435 Patnode, 399, 413
~ovak, 157, 172 Pelagic-abyssal environments, 316, 319-
~uclear basins, 197, 198, 205 320
Nullipores, see Lithotham1Zion Pelagic deposits poor in lime, 357
Nutrients, 14, 17, 21, 36, 216, 233, 354, Pelagic environments, 313
357 'UI. aby al environments, 313
Pelagic sediments, 347-360
Ocean floors, constitution of, 115-123 distribution of, 362
Oceanic basins, cubic content of, 129 See also Deep-sea deposits
INDEX 563
Pelite, see Lutite Profiles of equilibrium, complications
Pcnck, 541, 548, SSI of, 303
Permanency of continents and oceans, Proudman, 533, 534, 551
123- 132 Prouty, 159
Persian Gulf, 15 Psammites, 211
salinity of, 19 Psephires, 211; see also Coarse sedi-
Pettersson, 374, 375, 376, 378, 381, 395, ments; Boulders; Coral reefs, ram-
411, 412 parts; Coral ree.fs, shingle on
Pettijohn, 211, 249, 250, 277, 300, 409, Pteropod ooze, 352
411
Philippi, 349, 369, 370, 411, 412 Radioactivity, 374- 376
Philipsite, 219 showing rate of sedimentation, 380
Phleger, 412 Radiolarian ooze, description of, 358
Phosphate nodules, 219, 264 radioactivity of, 374
Phosphates, 14 Radiolarians as source of sediment, 215
Phosphorus, total amount in sediments, Radiolarite, 204, 215, 245, 359
390 Rafting, 276-2 77
Photosynthesis, 8, 14 Rann of Cutch, 334
Phy toplankton, 14 Rastall, 299
Pia, 412 Rate of sedimentation, 376-385
Pickering, 132 for calculating total amount, 386
Piggot, 6, 246, 369, 374, 375, 380, 382, Ratio of sediment types, 387
383 , 412 Rech, set Submarine canyons
Pits, 98, 201 Red clay, 192, 204, ;51, ;$8, ;59, 404
Plankton, as source of organic matter, average composition of, 389
216 description of, 359
3S source of sediment, 215 fossil, 376, 394
distribution of, 16, 216 radioacth'ity of, 374
ill pelagic sediments, 319 rate of accumulation of, 377, 382
Plants, action in depositing mud, 318 stratification in, 369
rafting by, 277 Red mud, 346
Plateaus, 98 Red Sea, IS , 98
Platform of marine erosion, 226 salinity of, 19
Plato, 131 tides of, 60
Plunge point, 269 ventilation of, 13
Plunging waves, 83, 270 Reefs, see Coral reefs
Polar seas, color of water in, 23 Refraction of waves, 81, 82
Positive forms of sea fioor, definitions in cusp building, 295
of,98 Rehm, 150, 172
description of, 103 Reinhold, 110, 173
origin of, 151-153, IH Renard, 315, 377, 411
Potential temperature, 23 Rettger, 301
Pratje, 301, 409, 412 Revelle, 2, 28, 87, 252, 336, 342, 347, 351,
Prlltt, 168, 172 357, 380, 412
Precipitates as source of sediment, 217- Rhine, 232
221 Rhomboid ripples, 291
Prielen, 330 Rhythmic accumulations, 288-297
Profiles of equilibrium, 226, 302-HO. Rhythmic stratification, 368, 373
;05, 406 FUch, 293 , 301, 500, 530
564 INDEX
Richter, 120, 125, 150, 151, 171 Sargasso Sea, 21
Ridges, 98, 103, 306 Saumtiefen, 48
Riel, van, 47, 88, 96, 172, 209 Saville Kent, 479
Rift valleys, 498 Scandia, 126
Rimmed solution pools, 237 Schaffemicht, 59
Rio Grande, 233 Schott, G., 2, 20, 22, 24, 25, 28, 56, 88,
Rip currents, 86, 269 106, 172, 301, 360, 412
Ripple marks, 288, 289-293 Schott, W., 264, 301, 349, 369, 378, 410,
classification of, 292 412
Rittmann, 141, 172 Schuchert, 125, 131, 145, 173, 198, 199,
River valleys, drowned, 481-484 209, 412, 540, 551
Rivers, as sources of sediment, 232-235 Schumacher, 88
sediment transport in, 21-1, 2H, 2S9 Schwinner, 377, 412
sorting by, 283 Scott-Russell, 85
vs. coastal erosion as sources of sedi- Screes, 227
ment, 233, 234 Sea ice, 27
Rocky sea floor, charting of, 110-11 I, Sea level, amount of glacial lowering of,
339 535-537
Rolling of sediment particles, 258 correlation with glacier volume, 533
Romanovsky, 409 estimate of instability of, 541- 543
Roots of mountains, 206, 207 excessive lowering of, 477, 503- 509
Rothe, 129, 172 glacial eustatic movements of, 170,
Rotting of organic matter, 8 264, 266, 535- 540
Rouch, 2, 3, 28, 88, 90, 106, 172 influenced by amount of water on
Roundness of sediment particles, 278 earth, 543- 544
Rubey, 261, 301 influenced by basin formation, 547
Rus~il, F. S., I, 88 influenced by internal causes, 54S-550
Russell, I. C., 346, 412 influenced by sedimentation, 545, 546
Russell, R. D., 277, 282 influence on coral reefs of changes of,
Russell, R. J., 327, 412 445-453, 449
Russell, W. L., 375, 412 in former periods, 1 II, 11 3, 124, 130,
Rutherford, 171 155, 398, 540-550, 542
Rutten, 195, 209, 316, 412 interglacial, 539
profile of equilibrium influenced by
Sahara dust, 213 movements of, 405
Salinity, 12, 15, 18, 19, 42-44, 45, 418 progressivc rise of, 549
Salt water vs. fresh water, 19 recent changes of, 533, 534
Saltation, 257, 286 subrcccnt sinking of, 448, 449, 537
Salts, concentration of, 9 Sea water, color of, 21 -23
formed by evaporation, 220 composition of, 15
in sea water, 14-19 compressibility of, 23
See also Salinity density of, 19
Samplers, 6, 7, 383 properties of, 19- 23
Sand, deposition of, 317-319 temperatures of, 24-25
Sand banks, 272, 273, 293 transparency of, 21 -23
development by waves of, 270, 272 Seamounts, 98, 103, 152
Sand bars, see Sand banks Seas, adjacent, 97
Sand waves, 290 epciric, 97
Sapper, 129, 172, 388 epicontinental, 98
IN D EX 565
Seas, inland, V7 292, 299, 301, 316, 368, 380, 410, 412,
marginal, 97 413, 425, 454, 465, 477, 485 , 488, 491,
mediterranean, 98 494, 496, 497, 498, 502, 503-509, 513,
shelf, 97 514,516,518,519,523,524,530, 531,
Seaweed, 11 535
Sediment traps, 376 Shoaling by sedimentation, II4
Sedimentation, causing shoaling, 114 Shoals, 98
influence of coastal topography on, Shorelines, classification of, 99
310-311 Shrock, 365, 412
in partly closed basins, 307 Sial, 116, 118
influence on sea level, 545, 546, 547 Sial line, 117, 118, lSI
of mud, 317 Silica, 14, 233
rate of, 376-385 geoeconomy of, 395
Sedimentation basins, 197, 200 solution of, 320
Sedimentation troughs, 197 Silt, see Lutite
depths of water in, 202-204 Sima, 116, 118
Sedinlcnts, amount provided by rivers Sinking, 248, 249
and coastal erosion, 233 Sitter, de, 408, 4 I2
classification of, 336 Sliding of sedimentary particles, 258;
distribution of, 360 see also Slumping
in suspension, 2)4 Sluiter, 421
011 steep slopes, 247, 248 Slumping, 240-248, 242, 2#, 370
organic matter of, 399 in submarine canyons, 500, 512, 5U,
volume of, 386-399 522
Seiches, 59 Smith, 96, ISS, 173, 239, 301, 486, 488,
Seismological evidence on nature of 491, 531, 539, SSI
ocean floor, 120, 121, 122, 123, 150 Smit Sibinga, 119, 173, 177-182, 209
SeiweU, 88 Snappers, 6, 7, 383
Senn, 126, 173 Snellius, 4, S, 13, 46, 54, SS, 67, n, 94,
Serial observations, 5 96, 104, 175, 182, 18!J, 186, 201, 203,
Serventy, 277, 301 206, 212, 229, 247, 340, :141, 342, 343,
Settling, 248-251 351, 367, 371, 381, 401
velocity of, 248, 2'19 Sodium, 388
Sewell, 142, 174, 434, 437, 479 calculation total volume of sediment,
Seymour Narrows, 230 388
Shape of grains, influence of abrasion See also Salinity; Salts
on, 277 Solution pools, 237
influence on settling, 249 Sonder, 147, 153, 173
influence on sorting, 285- 286 Sorby, 220
Shelf, Continental, 104, 304; tee also Sorting, 227, 282-288
Continental terrace absence of, 317
Shelf channels, 481--485, '181, 482/48!J, according to mineralogy, 286
S06 according to shape, 285-286
Shelf deposits, 337, 374 according to size, 282-285
Shelf seas, 97 Soundings, accuracy of, 92-94
Shepard, 2, 7, 88, 90, 99, lOS, 106, 108, by thennometers, 92
111, 152, 154, 15S, 157, 158, 159, If)], by wire, 90-92
168, 171, 173,227,229,234,240,247, position of, 94
248, 252, 256, 263, 264, 269, 272, 273, sonic, 91, 92
566 IN 0 EX
Specific gravity, 19, 118 Submarine canyons, hypotheses on,
Spender, 444, 479 496-499
Sphericity, 277 in Mediterranean, 495, 508
Spilling waves, 83, 270 link with rivers, 481, 491, 505, 506, 507
Sprigg, 239, 506, 512, 531 nature of wall rock, 494, 520-525,521,
Stacks, 225, 22) 522
Stagnant water, 9, 10, 11; see also sampling in, 157, 162
EUlonic environments slumping in, 500, 512, 513, 522
Stassfurt, 220 Submarine landslides, 241; see also
Staub, 173, 185, 205, 209 Slumping
Stearns, 144, 173, 450, 454, 459, 461, 476, Submarine photography, 7
479, 540, 551 Submarine topography, 97- 110
Steers, 261, 270, 275, 301, 441, 443, 479 Submarine weathering, 236-238
Steinmann, 112, 204 Subsidence theor), of atolls, 455, 456
Stereoscopic photography, 75, 78 Suess, 97, 124, 133, 540, 545, 548
Stetson, 88, 157, 173, 228, 229, 239, 248, Suffolk scarp, 539
263, 301, 486, 494, 531 Sulu Sea, 46, 52, 356
Stille, 125, 129, 173, 182, 188, 195, 196, Sanda Ri"er, 482, 492/'183, 536
197, 198, 204, 209, 540 Surf, see \Vaves
Stokes, 248 Surf ripples, 29J - 292
Straaten, van, 288, 292 Surface tension, 72
Strait of Gibraltar, 42, 43 Surface water, temperature of, 20
Straits, 98, 230 Surry scarp, 539
Straits of [Jover, 231 Sverdrup et aI., 2, 8, 19, 25, 28, 29, 36,
tidal currents of, 63 69, 88, 173, 216, 301, 354, 356, 360,
Stratification, 362-374 362, 392, 413
abn9rmal, 247 Swain, 157, 173
general considerations on, 362-368 Swash, 83 , 270, 291
in pelagic deposits, 320, 369, 370, 371, in cusp building, 295
377 See a/so Waves
in recent sediments, 368--374 Swell, 77, 78
rhythmic, 368, 173 modified to solitary waves, 86
Str~m, 9, 88, 217, 380 See also Waves
Sublittoral zone, 315 Swells, 98
Submarine canyons, 98, 105,248,485-526
absence of waterfalls in, 522 Talus, 227
age of, 494, 503 Tambora, 212, 237, 343
bends in, 518, 519 Tangue, 337
currents in, 252 Taylor, 125
data on, 481, 484/485, 485-496, 486, Teichert, 171, 196, 208, 277, 301, 429,
487, 489, 499, 492, 493, 494, 495, 497, 470, 478, 479
506, 507, 515,517, 518, 519, 521, 522 Tercier, 413
deltas off, 510, 516, 517 T ereaces, 104
diastrophic origin of, 498 coastal, 193, 226
downwarped river valleys, 498, 499 See also Continental terrace
formed by mud flows and landslides, Terrigenons sediments, 337-347, 339,
500 341, 342/343, 344
formed by spring sapping, 500, rOl ayerage composition of, 389
formed by tsunamis, 501 Terry, 494, 531
INDEX 567
Tesch, 110, 173 Turbidity currents, 238-240, 265
Tests as source of sediment, 214-216 as cause of graded bedding, 366-367
Thijsse, 64, 87 as cause of submarine canyons, 509-
Thomson, 3, 90 526
Thorade, 2, 56, 69, 77, 88 Turbulence, 8, 21, 22, 29, 2H
Thorarinsson, 533 influence on sediment transport, 252-
Thorp, 217 256, 2H
ThouJet, 250 Twenhofel, 210, 211, 261, 279, 301, 376,
Tidal deltas, 231, 329 392, 403, 409, 413
Tidal flats, 318, HO-331
ti des of, 64-69 Umbgrove, 102, 115, 119, 125, 129, 132,
Tidal waves, lee Tsunamis 133, 134-139, 141, 146, 150, 154, 156,
Tides, 5tHl9 158, 162, 173, 176, 182, 188, 189, 191,
daily inequality of, 58, 58 194, 197- 208, 209, 385, 4J3, 419, 437,
geological significance of, 59 439, 443, 479, 482, 498, 531, 536, 540,
in inland seas, 60, 61, 62, 63, 64, 68 543, .117, 551
in lakes, S9--{)0 Unconformity, 308
tidal currentS, 62--{)7, 68, 502 Undertow, 86, 268, 272
tidal forces, 56, 57, 58, 59 Underway bottom samplers, 6, 7, 383
tidal period, 58 Upwelling, 31
tidal range, 58, 61, 62, 63, 64 Urry, 246, 374, 375, 380, 412
tidal scour, 231, 482/483, 484-485 Ushant, 229
tide gauges, 533
Timmermans, 292, 301 Varves, 373
Titanium, total amount in sediments, Vaughan, 2, 4, 89, 94, 106, 173
390 Veatch,96, 155, 173,486,488, 491,531,
Tizard, 84 539, 551
Topography, continental VI. oceanic, Veen, van, 63, 89, 231, 256, 301
J08 Vening Meinesz, 70, 120, 127, 138, 144,
of the sea floor, 97 148, 153, 173, 176, 189, 190, 191 , 194,
Top-set beds, 325 200, 203, 209, 549
Trabert, 134 Ventilation, see CurrentS, ventilating
Tracey, 422, 478, 479 Verbeek, 80, 177, 209, 212, 301
Traction tube, 260 Verwey, 418, 479
Tractional force, 258, 261 Vine, 87, 171
Translation waves, 83, 85, 270 Vlerk, van der, 300, 411
Trask, 9, 28, 88, 210, 216, 217, 238, 263, Voelzkow, 220
301, 352, 360, 379, 380, 399, 400, Volcanic mud, 342
401, 413 Volcanoes, as cause of stratincation, 372
Travtrilltur, 246 as source of sediments, 212, 2B,
Traylor, 76, 81, 88 342/343, 343, 388
Trenches, see Deep-sea troughs on deep-sea floor, 388, 397, 529
Troughs, set Deep-sea troughs submarine slopes of, 524, 526, 527, 529
filled with sediment, 197 Voltz, 176
on beaches, 272 Volume of sediments, 386-399
Tsientan, bore in, 62
Tsunamis, 79, 265, 276, 335 Walther, 124, 173, 220
as cause of slumping, 247 Wanless, 368, 413
QS origin of submarine canyons, 501 Wanner, 18S
568 INDEX
Wattenberg, 217 Wegener, 125, 126--129, 134, 136, 140,
Wave erosion, 221-229 163, 174, 177-182, 209
depth of, 227 Wegmann, 129
Wave ripples, 289 Weibull, 392, 413
Wavecut cliff, 223 Weller, 368, 413
Wavecut notches, 223; see also Erosion Wenrworth, 21J, 540, 550
of coral limestone Whale Bay, organic matter in sedi-
Waves, 69-86, 79, 83, 84, 8S ments of, 403
abrasion by, 279 Whitecaps, 74, 270
action in transPOrt, 240, 262-276 Williamson, 220, 41 J
combined with currents in transport, Willis, 125, 174, 209
262 Wilson, 122, 174
conStrUctive, 270, 271 Wind, abrasion by, 279
destructive, 270 as source of sediment. 213. 214
effect of wind on, 72, 73, 74 effect on waves, 72- 77
energy of, 77 inRuence on coral reefs, 440--444
in deep water, 69-86, 70, 71, 72j see sorting by, 282, 286
also Swell Wind waves, 77; see also Waves
in shallow water and surf, 71, 80-85, Wiseman, 142, 174, 400, 413
72, 83, 84 Wood Jones, 439, 440, 479
internal, 5, 8 Woods, 479
mass transport by, 71 Woollard, 171
maximum size of, 74-77 Woolnough, 316, 413
oscillatory, 69 Worzel, 6, 87, 89, 156, 171, 174
Wiist, 29. 36, 104
plunging, 270
refraction of, 81, 295, 82
Yellow mud, 346
sideways transport by, 273-275 Yellow Sea, color of water in, 22
solitary, 85-86, 270, 83, 8S Yonge, I, 88
spilling, 270
tsunamic, see Tsunamis ZoBell, 413
velocity of progagation of, 73 Zoppritz. 28
wind, 77 Zuider Zee, 66, 252, 331
Weathering, 236--238 currents in, 31, 64, 65
Weber, 185 tides of, 64-67
1\1\11\1\111
430
\..\'30
CENtRAL INStitUTE OF flSHERlfS EDUCATlON
LIBRARY.
"uthor~Den Ph H
...