Ss f -

lH

»

M,

> t?^-"

*£*£*

w

>w*K

Tef*

irlJiW

r

iW '.- ~ r"

pg«

•,f.

§ I

SJ

1 =

5:

*• —

B-

_D

; o
1 m
' i-n
• a
I o

i □

: m

I D

DISCOVERY REPORTS

VOLUME IX

Cambridge University Press
Fetter Lane, London

Neiv York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo
Maruzen Company, Ltd

All rights reserved

DISCOVERY REPORTS

Issued by the Discovery Committee

Colonial Office, London

on behalf of the Government of the Dependencies

of the Falkland Islands

//£T

VOLUME IX

CAMBRIDGE

AT THE UNIVERSITY PRESS

'934

PRINTED IN GREAT BRITAIN BY WALTER LEWIS, M.A., AT THE UNIVERSITY PRESS, CAMBRIDGE

CONTENTS

HYDROLOGY OF THE BRANSFIELD STRAIT (published 22nd February, 1934)
By A. J. Clowes, M.Sc, A.R.C.S.

Material page 3

Introduction and Bathymetric Features 3

Horizontal and Vertical Sections of Temperature, Salinity and Density . . . 10

Water Movements 38

The Differences in Temperature and Salinity in February and November 1929 along the

North-Eastern Line of Stations 55

Note on the Presence of Ice in the Bransfield Strait and on the Currents through the

Straits separating the South Shetland Islands 59

Summary 62

List of Literature 64

DISTRIBUTION OF THE MACROPLANKTON IN THE ATLANTIC SECTOR
OF THE ANTARCTIC (published 10th April, 1934)
By N. A. Mackintosh, D.Sc.

Introduction page 67

The Physical Environment of the Antarctic Macroplankton 68

Methods 70

Organisms Identified 72

Cruises in the Period 1927-31 77

The Antarctic Convergence 83

Diurnal Variations 87

Relative Abundance and Distribution of Individual Species 97

Distribution of Rich and Poor Plankton 109

Distribution of Warm- and Cold- Water Species 121

Plankton Communities 137

The Macroplankton and the Distribution of Whales 153

Summary 157

List of Literature 158

THE SUB-ANTARCTIC FORMS OF THE GREAT SKUA (CATHARACTA SKUA
SKUA) (published 14th June, 1934)
By J. E. Hamilton, M.Sc.

Introduction page 163

Plumage 164

Measurements 165

Geographical Groups 171

Conclusions 173

Note on the Distribution of McCormick's Skua 174

Literature 174

Plate I following page 174

THE MARINE DEPOSITS OF THE PATAGONIAN CONTINENTAL SHELF
(published 16th August, 1934)
By L. Harrison Matthews, M.A.

Introduction ' page 177

Topography of the Sea-Bottom of the Region 177

Distribution of the Sampling 179

43988

CONTENTS
THE MARINE DEPOSITS OE THE PATAGONIAN CONTINENTAL SHELE

(continued)

Collection of Samples .

Analysis of the Samples

179
180
184
189
190

i93
196

200

201

202

203

following page 206

Distribution of the Grades .
Types of Deposit .
Description of the Types

DzrTz : z ss™« ™ ™ op ™ *». « * wko^

Discussion

Summary

References

Tabulated Data of the Samples Examined

Plates II to XIV

THE DEVELOPMENT OF RHINCALANUS (published 6th September, i934) ^ ^

By Robert Gurney

NEMERTEANS EROM THE SOUTH ATLANTIC AND SOUTHERN OCEANS

(published 21st November, 1934)
By J. F. G. Wheeler, D.Sc. page 2I?

Introduction . 217

219

225
. 225

"STL™;* *T WH.CH »«n» ™ CO^CTHO, WT„ TK, M ObT^EO

Systematic Account

Part I. Nemerteans from Saldanha Bay, South Africa . . •

Part II. Nemerteans from the Falkland Islands, South Georoia and the Islands and

Banks of the Western South Atlantic Ocean
Part III The Pelagic Nemerteans .

Notes on the Distribution of the Southern Nemerteans

List of Literature

Index

Plates XV and XVI

247
280
288
290

293
following page 294

: SEA-FLOOR DEPOSITS. I. GENERAL CHARACTERS AND DISTRIBU
TION (published 20th December, 1934)
By E. Neaverson, D.Sc., F.G.S.

Introduction

Classification of the Deposits
General Distribution of the Deposits
Description of the Samples .
Plates XVII to XXI I

. page 297

• 298

• 300

• 312
following page 349

ON THE STOCK OF WHALES AT SOUTH GEORGIA (published aoth December, „34^ ^
By J. F. G. Wheeler, D.Sc

ERRATUM
P. ,56, in Table XVIII, a, bottom of left-hand column:/*, "Early January" ,«« "Early February .

[Discovery Reports. Vol. IX, pp. 1-64, February, 1934]

HYDROLOGY OF THE BRANSFIELD STRAIT

By
A. J. CLOWES, M.Sc, A.R.C.S.

CONTENTS

Material P°ie 3

Introduction and bathymetric features 3

Horizontal and vertical sections of temperature, salinity and density

April 1927 I0

February 1929 :7

November 1929 25

December 1930 32

Water movements 3°

November 1929 and December 1930 41

February 1929 5°

April 1927 . • • • 55

The differences in temperature and salinity in February and November 1929

along the north-eastern line of stations .... 55

Note on the presence of ice in the Bransfield Strait and on the currents through

the straits separating the South Shetland Islands .... . 59

Summary °2

List of Literature *4

HYDROLOGY OF THE BRANSFIELD STRAIT

By A. J. Clowes, M.Sc, a.r.c.s.

(Text-figs. 1-68)

MATERIAL

THEhydrographic material on which this report is based was obtained from a series
of stations made by the ships of the Discovery Committee in the Bransfield Strait
and in the adjacent sea. The observations used were taken in April 1927, February 1929,
November 1929 and December 1930, and are summarized as follows:

Table I

Stations

Date

196-200
202-206

iv. 1927
iv. 1927

King George Island to Trinity Peninsula
Livingston Island to Trinity Peninsula

WS 382-WS 388
WS 389-WS 393
WS 395-WS 398

ii. 1929
ii. 1929
ii. 1929

King George Island to Trinity Peninsula
Livingston Island to Trinity Peninsula
Brabant Island to Low Island to Smith Island

WS 476-WS 482
WS 483-WS 487
WS 488-WS 491

xi. 1929
xi. 1929
xi. 1929

King George Island to Trinity Peninsula
Livingston Island to Trinity Peninsula
Brabant Island to Low Island to Smith Island

537-541
542-547
549-553

xii. 1930
xii. 1930
xii. 1930

Elephant Island to Joinville Island

Cape Melville, King George Island, to Trinity Peninsula

Snow Island to Trinity Island

In addition to the above lines of stations, odd stations, taken at varying times in the
Bransfield Strait and de Gerlache Strait, have been used in the construction of the hori-
zontal sections. The positions of all stations used are plotted in Fig. 1 . Ice data have been
obtained from our own observations, from a collection of ice reports made by Mr Risting
of the Norwegian Whaling Association, and very largely by the kind help of Mr Nielsen
of the Hector Whaling Company, whose personal experience of the Bransfield Strait has
been of great value to us.

INTRODUCTION AND BATHYMETRIC FEATURES

The Bransfield Strait is a long, narrow strip of water approximately 70 miles wide,
whose general direction lies north-east and south-west. It is bounded on the north-west
side by the various islands of the South Shetland group, and on the south-east side by
Graham Land and Trinity Peninsula, which together form the most northerly part of
the Antarctic Continent. Fig. 1 shows the position and extent of the Bransfield Strait
with its boundaries.

4 DISCOVERY REPORTS

Many expeditions have visited and made hydrographic investigations in this area.
The Expedition Antarctique Beige in the S.Y. ' Belgica ' crossed the southern end of the
Bransfield Strait, passing between Snow Island and Smith Island to de Gerlache Strait
on the way to the Bellingshausen Sea in January 1898. The first expedition to do any
considerable hydrographical work in the Bransfield Strait itself was the Swedish South
Polar Expedition in 190 1-3. The Deuxieme Expedition Antarctique Francaise in
1908-10 also worked in and south of the Bransfield Strait. Later the German Atlantic
Expedition, 1925-7, visited the area in January 1926 and was the first expedition to use
echo-sounding. Since March 1927 very considerable work has been done in the

60°

58°

56°

537J

ElepTicmtlkJ

Clareruce

:Ktnq George I.

•542 -171
• 543

pBridcpruxn I.

•WS475

538

539

540

•WS384

Snow I-Cy%? ^?
Smith I\.

W5492

-Sft

< / WS493 202

•548

WS393
•WS4&3
..WS392
203- WS484 joi
94 H

197" "w00^ *544
.WS385
I98'|wS479 -545
•WS386

•546
. .WS387

WS^° -WS388

-Saiws4a_J54z_

193

541

tvl

Fig. 1 . Chart of the Bransfield Strait showing positions of stations.

Bransfield Strait and the neighbouring seas by the various ships of the Discovery
Committee, i.e. the R.R.S. 'Discovery', the R.R.S. 'William Scoresby' and the
R.R.S. 'Discovery II'. Altogether over a hundred stations have been taken by these
ships in the Bransfield Strait area. The soundings at these stations, mostly obtained
by use of the Lucas sounding machine, have been considerably augmented by over
1400 echo-soundings taken by the R.R.S. ' Discovery II '. Thus by combining all the
information at hand a fairly accurate bathymetrical chart can be drawn of this area ; it

HYDROLOGY OF THE BRANSFIELD STRAIT 5

is shown in Fig. 2.1 It will be seen that the contours of the sea-bottom in the Bransfield
Strait are very irregular, but the soundings show that the strait itself is practically cut
off on all sides by land masses or by submarine ridges which will be described later.
Since soundings of over 2000 m. are recorded and confining ridges, varying between
600 and 250 m. in depth from the surface, have been discovered, it follows that the
Bransfield Strait consists of a deep basin shut off on all sides. These ridges restrict
the deep water in the Bransfield Strait from contact and free circulation with that
of the seas outside. This has a marked effect on the temperature of the water in
the Strait below the level of these submarine ridges. The existence of very low

Fig. 2. Bathymetric chart of the Bransfield Strait. Soundings in metres.

temperatures in the water below 300 m. has been noted by many expeditions, and
the probable existence of submarine ridges has been postulated by reason of the pre-
sence of these temperatures, taken in conjunction with the few soundings available before
echo-sounding was used. Thus O. Nordenskjold (1917, p. 9): "Deuten also schon
die Lotungen an, dass hier ein allseitig abgesperrtes Tiefenbecken vorliegt, so wird
dies in eklatanter Weise durch die Messungen der Wassertemperaturen bestatigt, und

1 This chart is based on that published by Herdman, 1932, pi. xlvii.

6 DISCOVERY REPORTS

schon J. G. Andersson zog aus der gefundenen Bodentemperatur von — i-6° bis
— i-8° diese Schlussfolgerung ". Similarly J. Rouch (1913, p. 28): "...a partir d'une
profondeur voisine de 500 metres, la temperature de l'eau de mer est constante et
assez basse, ce qui laisse supposer que le detroit de Bransfield forme un bassin special
ferme par un seuil dont la profondeur serait voisine de 500 a 600 metres". Also
G. Wiist (1926, p. 243): " ...das Bransfield-Meer. Dieses abgeschlossene Meeresbecken,
dessen grosste Tiefe wir auf unserem Kurse mit etwa 2000 m. feststellten, ist unterhalb
300 m. mit einem sehr kalten und fast homohalinen Wasser erfullt, das offenbar an den
Boschungen des Graham-Landes abgesunken ist. Der Temperaturunterschied
korrespondierender Tiefen innerhalb und ausserhalb des Beckens erreicht in 1000 m.
den hohen Betrag von 3*05° ".

In February 1929 the R.R.S. 'William Scoresby' observed a temperature of i-ii° C.
at 1500 m. at St. WS 400, in 620 07' S, 620 33' W, i.e. just outside the strait, whilst
5 days previously at St. WS 385, inside the strait at 620 32' S, 570 55' W, a temperature
of — 1 -63° C. was recorded at the same depth.

The above comments or results from four different sources show the very great con-
nection between the contours of the sea-bottom of this area and the peculiar hydro-
graphical conditions to be found there.

Previous to the echo-sounding work of the German Atlantic Expedition in 1926 and
the soundings taken during the Discovery investigations in 1927, 1929, 1930 and 1931,
the bathymetric chart published in 1917 by O. Nordenskjold was probably the most
accurate. The Swedish South Polar Expedition in 190 1-2 had taken a line of soundings
between McFarlane Strait and Astrolabe Island, some soundings on the continental
shelf of Graham Land and some soundings to the north-east of the Bransfield Strait.
These soundings showed the presence of a longitudinal basin which was closed at the
north-east end by a rise of the sea-bottom to the common shelf depth of 400-500 m., to
fall again rapidly on the eastern side to the great ocean depths. The Deuxieme Expedi-
tion Antarctique Francaise, 1908-10, confirmed the existence of the basin in the
Bransfield Strait. The echo-soundings of the German Atlantic Expedition, 1925-7,
showed the rise of the continental shelf of the South Shetland Islands from the great
ocean depths of the Drake Strait. According to O. Holtedahl (1929, p. 95) these
soundings of the German Atlantic Expedition showed the existence of very marked
planes of abrasion, thus proving a former higher level of the land and consequently an
increased width of the South Shetland Islands in the past. The echo-soundings of the
' Meteor ' again showed the presence of a deep basin in the Bransfield Strait, with an
indication of a rise in the sea-bottom to the north, some 13 miles west-south-west of
Elephant Island and approximately 12 miles north of Aspland Island.

Such was the state of knowledge when our own observations began. Lines of
stations from King George Island and Livingston Island across the strait to Trinity
Peninsula, and between Smith Island and Brabant Island were worked in 1927, 1929
and 1930, and the sounding data from these stations were augmented by over 1400 echo-
soundings in this area by R.R.S. 'Discovery II'. In the bathymetric chart, Fig. 2, all

HYDROLOGY OF THE BRANSFIELD STRAIT 7

the above information and as much of the older information as possible has been
used.1

At the south-west end of the Bransfield Strait lie the islands of Snow, Smith, Low,
Hoseason, Intercurrence, Brabant, Trinity and Deception. The soundings show that
these islands are connected by submarine ridges with different saddle-depths. To the
east of Smith Island a channel of greater depth than 600 m. occurs, which connects with
the south-west basin of the Bransfield Strait. A channel, possibly deeper than 500 m.,
may occur south of Low Island, also giving access to the south-west basin which is
described later. Between Snow Island and Deception Island the depth is nowhere
greater than 250 m., although between the latter island and Livingston Island a depth
of over 500 m. exists. The submarine ridge between Deception and Trinity Islands has
a saddle-depth of probably between 700 and 800 m. There is also a possibility of a
slightly increased depth near Trinity Island. However, in the area contained by the
triangle formed by the islands of Deception, Low and Hoseason, depths between 1300
and 1400 m. are recorded. Thus at the south-west end of the Bransfield Strait a basin is
present whose boundaries are the islands and ridges connecting the islands of Snow,
Smith, Low, Hoseason, Trinity and Deception, and whose greatest depth is probably
not more than 1400 m.

The second and larger basin in the Bransfield Strait runs south-west and north-east.
It begins at the shoal on which Deception Island stands, and continues until that of
Clarence Island is reached. A slight constriction of its width occurs south-south-east
of Bridgeman Island where the sea-bottom rises to a depth of 1200 m. from the
surface ; on both north-east and south-west sides depths greater than 2000 m. are
reached. This basin lies closer to the South Shetland Islands than to Trinity Peninsula
and Graham Land owing to the steepness of the very narrow continental slope on the
south-east side of the South Shetlands, and the more gentle sloping of the extensive
continental shelf on the coast of Graham Land and Trinity Peninsula. At the north-east
end of this basin there is evidence of two confining submarine ridges. To the north the
'Meteor' obtained an echo-sounding of 199 m. some 13 miles approximately west-
south-west of Elephant Island, whilst the R.R.S. ' Discovery II ' on a course just west
of north in the gap between Elephant Island and King George Island obtained
gradually diminishing echo-soundings with a minimum depth of 223 m. at 6i° 14' S,
560 39' W. Thus there is some evidence for assuming the existence of a continuous
continental shelf between Elephant and King George Islands. To the east the wide
continental shelf north of Joinville Island in the direction of Clarence Island has been
shown to be considerably extended and a large area of just over 300 m. was found by
echo-soundings. Thus to the east of the Bransfield Strait there is also some evidence of

1 It is to be noted that a number of the earlier soundings cannot be plotted on recent charts. Surveys
undertaken during the course of our work have shown that the existing charts were inaccurate, and though
much still remains to be done, especially off Trinity Peninsula, the coast-line of some part of the South
Shetlands has been corrected and the positions of a number of the islands have been revised. The corrections
have been incorporated in the latest Admiralty charts. A number of the early soundings were fixed by land
bearings, and they cannot be transferred to the new charts unless the original data are available.

8 DISCOVERY REPORTS

a submarine ridge between Clarence and Joinville Islands, with the saddle-depth a short
distance south of the former island. Many more soundings in this area are necessary
before the bottom contours at the northern end of the strait can be drawn accurately.
For a more detailed description of the soundings in the Bransfield Strait reference
should be made to the paper on soundings by H. F. P. Herdman (1932).

Thus the Bransfield Strait consists of two basins enclosed by shelves of the South
Shetland Islands, Elephant and Clarence Islands, Graham Land and Trinity Peninsula
and by submarine ridges at either end of the strait. The effect of these ridges is to
restrict the horizontal circulation of the water masses which lie below the depth of the
confining ridges.

Fig. 3. Temperature-salinity diagrams for Sts. WS 400, WS 384 and WS 385.

Before discussing the hydrology of the Bransfield Strait and the manner in which
it is influenced by the topography of the sea-bottom, the following comparison of con-
ditions inside and outside the strait will be of service. For this purpose three stations
from February 1929 have been selected: St. WS 400 situated in 620 07' S, 620 33' W,
i.e. in the Drake Passage just north of the Bransfield Strait, and Sts. WS 384 and
WS 385, situated both at the north-east end of the strait in 620 25' 40" S, 580 06' 10" W
and 620 32' S, 570 55' W respectively. The temperature and salinity data from these
stations have been plotted in Fig. 3 as temperature-salinity diagrams. The figures on
the curves represent the depths in metres of the observations.

In the Antarctic Zone of the South Atlantic Ocean the water masses may be divided
into three layers. The uppermost consists of Antarctic surface water, the middle layer

HYDROLOGY OF THE BRANSFIELD STRAIT 9

of warm deep water and the lowest layer of Antarctic bottom water. The effect of the
relatively warmer middle layer is to produce a temperature inversion below the cold
nucleus of the surface layer.

If we consider firstly the temperature-salinity curve for St. WS 400 it can be seen
that it consists of three distinct portions AB, BC and CD. The line AB represents the
effect of summer conditions on Antarctic surface water, which has come from the
Bellingshausen Sea and has been warmed by the sun and diluted by melting ice in its
upper portion. The point B represents the relation of temperature and salinity at the
cold nucleus of this layer. Between B and C both the temperature and salinity increase
rapidly with depth and this portion of the curve represents the mixtures between these
depths of Antarctic surface water and warm deep water, the latter water being of Pacific
origin. At B only Antarctic surface water and at C only warm deep water is present and
the proportion of each type of water in the mixture represented by the line BC may easily
be calculated. From the point C the temperature commences to fall rapidly to the lowest
observation and the salinity increases very slowly as far as 1000 m., where it remains
constant until 1500 m., below which it decreases very slightly to the two lowest
observations at 2000 and 3000 m. If the warm deep water were a homogeneous layer
the portion CD would accurately represent the mixtures of this layer with Antarctic
bottom water. The latter, however, does not commence to influence the salinity until a
depth between 1500 and 2000 m. is reached. Thus only below 1500-2000 m. does the
amount of Antarctic bottom water in the mixture commence to be appreciable. At this
station there is a very smooth transition from the warm deep water to the bottom water
of the southern part of the Drake Passage.

The temperature-salinity curve for St. WS 384 shows essentially the same features
as that for St. WS 400, with, however, some modifications, due to the fact that St. WS
400 is situated outside and WS 384 inside the strait. The summer warming and dilution
of the surface layer is seen between o and 200 m. The cold nucleus of the Antarctic
surface water is, however, deeper than usual and the salinity at this point is high, which
indicates the effect of vertical mixing in winter. Between 200 and 300 m. both salinity
and temperature increase, but the temperature does not rise beyond — 0-31° C, which
is found at 300 m. Compared with the corresponding temperature at St. WS 400 this
temperature is 2-03° lower and is situated 300 m. nearer the surface. Thus a large
difference is found in the temperature and depth of the maximum temperature of the
warm deep water at stations inside and outside the Bransfield Strait. This is perhaps the
chief effect of the topography of the sea-bottom of the strait. The flow of warm deep
water into the strait is restricted by submarine ridges, and as a consequence the circula-
tion with the warm deep water of the outside sea is greatly modified. At the majority of
the stations inside the strait the temperature inversion due to the presence of the warm
deep water is either absent or very weak. Between 300 m. and the lowest observation
the warm deep water at St. WS 384 is mixed with increasing amounts of Antarctic
bottom water until at 1500 m. the temperature reaches the low value of — 1-56° C.

St. WS 385 was selected as one of the many stations inside the Bransfield Strait at

io DISCOVERY REPORTS

which no temperature inversion is present owing to the poor development of the warm
deep water inside the strait. In the temperature-salinity curve for this station it is seen
that the surface water has been warmed by the sun and considerably diluted. Between
the surface and a depth of 40 m. the salinity increases rapidly, and from 40 to 80 m. the
salinity increase is less rapid but the temperature decreases rapidly. Between 80 and
300 m. the salinity increases rapidly, but the temperature does not fall very much. This
layer between 80 and 300 m. represents the effect of vertical mixing in winter, with the
consequence that the salinity in this layer is less than at the corresponding depths in the
two other stations, and all evidence of a temperature inversion has been eliminated.

The temperature-salinity curves for these three stations show a gradation from con-
ditions in the outside sea to conditions inside the Bransfield Strait: in the latter the
amount of warm deep water is greatly restricted, as is seen either by the small maximum
temperature of this layer at some stations, e.g. St. WS 384, or by the complete absence
of a temperature inversion at the majority of stations, e.g. St. WS 385. As a consequence
of the decreased amount of warm deep water inside the strait the level of the Antarctic
bottom water is much higher inside the strait than outside.

The restricting influence of the submarine ridges is especially pronounced in the low
temperatures inside the strait below 300 m. In the Bransfield Strait only traces of the
warm deep water from outside are present, and this water in very reduced amount is
found mainly at the south-west end of the strait and at stations close to the south-east
coast of the South Shetland Islands (except in November 1929), where the extent of its
presence is shown by weak intermediate maxima in temperature. Farther towards the
eastern part of the strait nearly all thermal trace of the warm deep water is missing.
This is due to mixing with the highly saline and cold Antarctic water whose origin is the
Weddell Sea ; part of this water is carried round Joinville Island to flow across the strait
at its northern end, and some down it in a south-westerly direction on the Trinity
Peninsula side of the strait. The maximum temperature of the warm deep water, which
in the adjacent sea is found near the upper surface of this layer at approximately 500 m.,
is situated inside the strait at least 100-200 m. nearer the surface.

The bottom water from the deep ocean outside the strait cannot enter the Bransfield
Strait because of the submarine ridges. Even in the great depths of the Drake Passage
and the Weddell Sea the temperature never approaches the low value of — 1-72° C,
which has been found inside the Bransfield Strait at a depth of 1500 m. The bottom
water inside the strait is essentially formed within the strait.

HORIZONTAL AND VERTICAL SECTIONS OF TEMPERATURE,

SALINITY AND DENSITY

APRIL 1927

From the results obtained in April 1927 a more complicated system of water
layering was sometimes found than had hitherto been observed by us in the south-
west Atlantic south of the Antarctic convergence. Three layers of water are usually

HYDROLOGY OF THE BRANSFIELD STRAIT

found in these latitudes, Antarctic surface water, warm deep water and Antarctic bottom
water. In the Bransfield Strait the layering is often more complicated, although the
warm deep layer is frequently absent. Factors introduced by the presence of ice and its
melting, and by the cooling and intense vertical mixing in winter, give rise to patches
of old water which appear out of place in the vertical column as reflected particularly
by the vertical temperature series. Unusual inversions of temperature and sometimes
of salinity are seen at some stations. Close to Graham Land and Trinity Peninsula
practically all trace of the warm deep water, as usually shown by an intermediate
maximum in temperature at about 500 m. depth in the neighbouring sea, is absent.
Figs. 4-6 give the surface salinity, temperature and density (at). Light water can be

Fig. 4. Surface salinity: April 1927.

seen in the surface close to the South Shetland Islands, and heavier colder water
spreading out from the north-west coast of Trinity Peninsula. The light water, which
flows as a north-east going set, consists of Antarctic surface water from the Bellings-
hausen Sea, and as later surveys have shown, enters the Bransfield Strait between
Smith Island and Snow Island and between Low Island and Smith Island. A branch
of this stream is seen as an eastward set reaching approximately 35 miles south-east of
Livingston Island, corresponding to the easterly set noted slightly more to the east of
this position on the Admiralty chart of the area. The heavier colder water close to the
Trinity Peninsula coast has its origin in the Weddell Sea; it has spread out round

12

DISCOVERY REPORTS

60'

Fig. 5. Surface temperature: April 1927.

Fig. 6. Surface density (o-,): April 1927.

HYDROLOGY OF THE BRANSFIELD STRAIT

13

Joinville Island and reaches about as far as Trinity Island, where it meets the slight
north-east set from de Gerlache Strait in Orleans Channel between Trinity Island and
Graham Land. The effect of the cold dense Weddell Sea water towards the north-east
end of the Bransfield Strait and on the Trinity Peninsula side is such that all or
practically all traces of the temperature inversion due to the warm deep water, which
usually lies immediately below the colder Antarctic surface water, have disappeared
from Sts. 197, 198, 199, 200, 201, 204, 205 and 206, which are enclosed in the shaded
area in Fig. 7.

The difference between the water on either side of the strait is emphasized in this
series of observations because the water on the north-west side of the strait has been

Fig. 7. The stippled patch shows the area in April 1927 in which no thermal evidence of the

warm deep water was present.

warmed more by the sun than that on the south-eastern side, which is much colder
and has its origin in the Weddell Sea. In winter this difference of temperature of
the surface waters in these currents would not be shown to such an extent owing
to surface cooling and vertical mixing. In midsummer in calm weather a discontinuity
surface is sometimes found with a temperature difference of over 20 C. between the
surface and a depth of 10 m.

Considerably less saline surface water is to be seen in and south-west of the de Gerlache
Strait, and this water shows a very striking contrast with the heavy colder surface water
farther to the north-east along the coast of Graham Land and Trinity Peninsula.

14

DISCOVERY REPORTS

The vertical sections of the salinity, temperature and density (at) of the line from
King George Island to Trinity Peninsula are given in Figs. 8-10. 1 Near Trinity
Peninsula the surface water consists of cold dense water forming part of the south-west
set on this side of the strait. As King George Island is approached the surface water
becomes warmer and less dense and forms part of the north-east going current which has
entered the Bransfield Strait between Snow Island and Smith Island and between Low
Island and Smith Island.

At St. 198, approximately 25 miles south-south-east of King George Island, the
temperature and salinity of the water below 200 m. are lower than at corresponding
levels at Sts. 197 and 199 on either side. From the shape of the lines of equal density
and from dynamic analysis it appears that St. 198 is the centre of a circular motion, the
water moving in opposite directions on either side of this station. The charts of dynamic

STATION 196

K.GEORGE

33-90&

TRINITY PENIN*

50CK

I000u

I500n

200Ck

Fig. 8. Vertical section of salinity: King George Island to Trinity Peninsula, April 1927.

topography show that the level of the surface of the sea at St. 198 is higher than at
Sts. 197 and 199 on either side. The movement of water has a more or less constant
velocity down to 200 m. at St. 198 and then decreases with depth. In a later part of this
report considerable doubt is thrown on many of the temperature and salinity results of
the line of stations from King George Island. In such an enclosed basin as the Brans-
field Strait it is surprising to find so much horizontal circulation below 200 m. as

1 In considering these sections, some observations of salinity have been neglected, and in particular
St. 197, 800 m., 35-06 °/00, St. 198, 1500 m., 34-87 °/00, and St. 199, 700 m., 35'0I700- ^ is considered
that these salinities, abnormally high for the latitude, are due to the partial freezing of the sea water in
the reversing water bottles before a sample could be taken, with the consequence that the salinity of the
remaining water became considerably increased. During these particular stations air temperatures of - 7°
to — io° C. were recorded. All three salinities were the result of duplicate titrations.

HYDROLOGY OF THE BRANSFIELD STRAIT

15

is needed to account for the shape the lines of equal density assumed in April
1927.

STATION 196

K.GEORGE I.

TRINITY PENIN*

500

1000m-

1500

2000

Fig. 9. Vertical section of temperature: King George Island to Trinity Peninsula, April 1927.

STATION 196

K.GEORGE I. 27 20

197

198 199

27302240

200

2750

TRINITY PENIN6

500m-

1000

1500m

EDOOi

Fig. 10. Vertical section of density (o-4): King George Island to Trinity Peninsula, April 1927.

The vertical temperature section shows the cold water at all depths at the north-east
end of the strait and the slight trace of the warm deep water only at St. 196 closest to
King George Island. The vertical salinity section shows the lower saline water below

i6

DISCOVERY REPORTS

200 m. at St. 198 which has already been noticed. In general the salinity of the deeper
water in the Bransfield Strait is lower than at corresponding depths in the seas outside
the strait. This is due to the effect of the bounding ridges surrounding the Bransfield
Strait in restricting the flow of warm deep water into the strait and to the very consider-
able vertical mixing which occurs throughout the strait in winter. At depths below the

STATION 202 203

LIVINGSTON I. 33-90&

500n

lODOrv

TRINITY PENIN^

Fig. 11. Vertical section of salinity: Livingston Island to Trinity Peninsula, April 1927.

STATION 202
LIVINGSTON 1.

0-0°

203

204

205

206
-0-5°

TRINITY PENIN*

\ >oo°:

0 0'

1

— -n-s° '

i

~~ 1 ,

_^y

— h0°. L_ /

500m-

<

-1-0° .

nnn..^

Fig. 12. Vertical section of temperature : Livingston Island to Trinity Peninsula, April 1927.

STATION 202 203 204 205

LIVINGSTON I. 2720 2730

206

TRINITY PENIN*

500s

1 000s

\t

2760^.

J- 2740 -^
-r27-50^>

^___y

-"27-70

V

2780 —

Fig. 13. Vertical section of density (<r,): Livingston Island to Trinity Peninsula, April 1927.

traces of warm deep water inside the strait the water is formed from the dense water
of Weddell Sea origin which mixes with the surface water in the remainder of the
strait and in winter is rendered more saline by the removal of water in the form of ice.
The more saline water then sinks to form the bottom water.

HYDROLOGY OF THE BRANSFIELD STRAIT

i7

The vertical sections of salinity, temperature and density (c7() for the line between
Livingston Island and Trinity Peninsula are shown in Figs. n-13. The surface
salinity is least and the temperature highest at the station nearest Livingston Island.
This station, 202, is in the north-east set and is the only station to show a positive
temperature range below the surface water ; at 300 m. it has an intermediate temperature
maximum of 0-36° C. At the other stations barely a trace of a temperature inversion
occurs below the surface and the influence of the warm deep water may almost be said
to stop at St. 202. The water below 300 m. at all stations consists of a mixture of
Antarctic surface water or warm deep water, as at St. 202, with increasing amounts
with depth of Antarctic bottom water.

Fig. 14. Surface salinity: February 1929.
FEBRUARY 1929

The hydro logical work in the Bransfield Strait which had been done in April 1927
by the R.R.S. 'Discovery' was repeated and augmented by the R.R.S. 'William
Scoresby' in February 1929, when, in addition to repeating the lines of stations from
King George Island and Livingston Island to Trinity Peninsula, a new line of four
stations was made at the southern end of the strait from Brabant Island to Smith Island,
with one station between the islands of Smith and Snow, and another between Living-
ston and Deception Islands. Figs. 14-16 show the distribution of the surface salinity,
temperature and density (<r<).

i8

DISCOVERY REPORTS

Fig. 15. Surface temperature: February 1929.

kLAF

Fig. 16. Surface density (a,): February 1929.

HYDROLOGY OF THE BRANSFIELD STRAIT 19

The temperatures and salinities of the surface water at the stations in this survey show
the effect of summer conditions in the Bransfield Strait. The surface water is probably
at its maximum temperature in February and the approach of autumn is not reflected in
the surface temperatures until a month later in the year.

The surface charts show a common agreement: warm light water is seen near the
coast of the South Shetland Islands and at the south-west end of the strait, with dense
colder water close to Trinity Peninsula. Poorly saline water of comparatively high
temperature, whose origin is the Bellingshausen Sea, is seen entering the strait between
Low Island and Smith Island and continuing into the strait, causing warmer and less
dense water at St. WS 390 on the line between Livingston Island and Trinity Peninsula.
Another inflow of poorly saline warm water occurs between Smith and Snow Islands. On
the line from King George Island to Trinity Peninsula the salinity of the surface water
increases normally towards Trinity Peninsula as far as St. WS 384. At Sts. WS 385 and
WS 386 an accumulation of water of considerably lowered salinity occurs in the surface
layer. Thus a difference of 0-43 °/00 is seen in the salinity observations between the
surface and 30 m. at St. WS 386. From St. WS 386 to the coast of Trinity Peninsula
the surface salinity then increases as usual. The accumulation of low salinity water in
the upper layers at St. WS 385, and in particular at St. WS 386, is also reflected in the
lower phosphate contents of the surface waters at these stations. It is remarkable
because normally the surface salinity at these two stations would reflect the influence
of the cold and more saline Weddell Sea water. No trace of the low salinity is found in
the section from Livingston Island. The water in the surface layer at Sts. WS 385 and
WS 386 is therefore surrounded by more saline water, and we think its occurrence can
only be explained by the assumption that considerable ice melting must have recently
occurred in this vicinity and that the diluted water had not yet been transported away.
Consequently slack water conditions prevailed and layering occurred, leaving, in the
absence of eddy motion and vertical mixing, a shallow discontinuity layer. A very
strong north-east setting current was observed at St. WS 383 but was no longer
present at St. WS 384. The topographical charts show that heavy water is present at
St. WS 384 and that this station is on the right-hand side of the north-east flowing
current at St. WS 383. An anti-clockwise eddy at St. WS 386 is shown in the topo-
graphic chart of the surface, and in these circumstances light water would be accumu-
lated at this station. The whole temperature series at St. WS 386 is also remarkable, the
many changes between relatively warm and cold water exhibited by the vertical
observations are confusing. A similar occurrence, however, has been noticed by us
before in the Bransfield Strait and is probably a direct consequence of a succession of
ice freezings and meltings at this particular station whereby an abnormal temperature
series was established.

The vertical sections of salinity, temperature and density (o-() of the section from
King George Island to Trinity Peninsula are shown in Figs. 17-19. St. WS 385 was
taken in a very strong north-east setting current which caused large stray on the water-
bottle wires. Unprotected thermometers which give an indication of the true depth were

3-2

DISCOVERY REPORTS

unfortunately not fitted to the reversing water bottles used below 400 m. Consequently
the isolines below this depth in the vertical sections at this station are situated too low.

STATION WS382 WS383 W53B4 W5385
KING GEORGE I. 3400& 33;30&3370&

WS386
3360S&

WS387 WS388
34-1 0%o_

:3420°4~
"3430&,

TRINITY PENIN6

500

1000

1500

2000m

Fig. 17. Vertical section of salinity: King George Island to Trinity Peninsula, February 1929.

TRINITY PENIN*

STATION WS382 W5383 WS384 WS385

KING GEORGE I. _ _2Q° |5°_

ii

WS38G WS387 WS388
10°

500

1000

1500

2000M

Fig. 18. Vertical section of temperature : King George Island to Trinity Peninsula, February 1929.

Nevertheless the slope of the lines of equal density between Sts. WS 382 and WS 383
shows the presence of the north-east going current which sets up the strait on this side.

HYDROLOGY OF THE BRANSFIELD STRAIT

In the topographical chart (Fig. 59) St. WS 384 is shown as occurring on the right-hand
side of a strong current close to St. WS 383. In the southern hemisphere the lines of
equal density in a vertical section will slope downwards to the left-hand side of the
current direction provided the current decreases in velocity downwards from the surface.
Consequently the lighter surface water will form a deeper layer on the left-hand of the
current, whilst to the right, denser water will appear close to the surface. The vertical
section of density distribution along this line of stations shows lighter water on either
side of St. WS 384.

Accumulations of light water in the surface layers at Sts. WS 385 and WS 386 are
seen in the vertical sections. Towards the coast of Trinity Peninsula the mean density
for all depths increases. The influence of the warm deep water is very small and is
restricted to a poor development at St. WS 382, except for rudimentary traces of

WS386 W5387 WS388

STATION WS3B2 WS383 WS384 W5385
KING GEORGE I. 2720 27201710 2700

W5387
2730

= 2740-

TRINITY PENIN^

500m

1000

1500

EOOOm

Fig. 19. Vertical section of density (o>): King George Island to Trinity Peninsula, February 1929.

intermediate maxima in temperature at some of the other stations in the section.
St. WS 382 is the only station on this line at which Antarctic bottom water is absent.
At all the other stations Antarctic bottom water is present at depths below 300-400 m.
Owing to the very considerable vertical mixing which takes place in winter in the
Bransfield Strait, the salinity of the Antarctic bottom water is at all times less inside the
strait than in the surrounding seas outside.

In the section from Livingston Island to Trinity Peninsula, Figs. 20-22, similar
conditions to those of the more north-easterly section appear, with the exception that
the Weddell Sea influence is lessened. Also the warm deep water is much more
apparent at the station nearest to Livingston Island, St. WS 393, than at the corre-
sponding station, St. WS 382, in the north-east section, despite the fact that St. WS 393

DISCOVERY REPORTS

is farther from the South Shetland Islands side of the strait than is St. WS 382. The
influence of the colder water on the Trinity Peninsula side of the strait is seen as far as

STATIC

)N WS393
JVINGSTON 1. 3390&

WS332
34-00%,

WS39I W5390
3400%.

WS383
34I0Z

TRINITY PENINA

\34Z0Z :=^====-

] |

— S

-3430^; ^.an*..

500m-

3450?; /

1000m-

Fig. 20. Vertical section of salinity: Livingston Island to Trinity Peninsula, February 1929.

WS39I W5390 WS2

10

STATION WS393

LIVINGSTON I.

WS392
1-5°

TRINITY PENlN*

500

1000,

Fig. 21. Vertical section of temperature : Livingston Island to Trinity Peninsula, February 1929.

STATION WS393 WS392

LIVINGSTON I. 2710 2720

WS39I WS390 W53B9

272Q__ 2730

41

TRINITY PENIN*

500

I 000m

Fig. 22. Vertical section of density (at): Livingston Island to Trinity Peninsula, February 1929.

St. WS 392, and here no trace at all of the warm deep water is observable, the tempera-
ture falling continuously from surface to bottom. The difference in composition of the

HYDROLOGY OF THE BRANSFIELD STRAIT

23

water in the north-easterly current near the South Shetland Islands, and that in the
remainder of the Bransfield Strait, influenced by the Weddell Sea and by the production
in winter of Antarctic bottom water, is nowhere more clearly demonstrated than by a
comparison of the temperature at 300 m. at Sts. WS 393 and WS 392. St. WS 393,
situated in the north-easterly flowing current, has a temperature of 0-70° C. at 300 m.,
whereas St. WS 392 has a corresponding temperature of— 1-05° C. Thus a temperature
difference of 1750 C. is shown at 300 m. between two stations both inside the strait and
approximately 12 miles apart. The temperature of 0700 C. at 300 m. at St. WS 393 is
due to the presence of warm deep water, and that of - 1-05° C. at 300 m. at St. WS 392
is due to mixed water containing a high proportion of Antarctic bottom water. The

-20

338 3 34-0

■2 -3 4 5 G

2 3 -4 5 -6 7

336 9 340 I

CD /oo

Fig. 23. Temperature-salinity diagram for Sts. WS 392 and WS 393.

diagrams showing the relation between temperature and salinity for these two stations
are shown in Fig. 23. The figures alongside the curves represent the depths in metres
of the observations. The portion BC of the diagram for St. WS 393 represents the mass
of mixed water occurring between the cold nucleus of the Antarctic surface water at
40 m. and the intermediate maximum temperature of the warm deep water at 300 m.
The diagram for St. WS 392 shows that the mass of mixed Antarctic surface water and
warm deep water is missing at this station.

At St. WS 390 an accumulation of slightly less saline but relatively warmer water
in the 0-100 m. layer corresponds to an inflow of surface water from the south-west.

24

DISCOVERY REPORTS

Figs. 24-26 show the vertical sections of salinity, temperature and density (at) for
the section between Brabant Island and Smith Island. The surface temperature and
salinity of the stations in this section at the south-west end of the Bransfield Strait show
that the surface layer is composed of warmer and considerably less saline water than in

STATION

SMITH

W5398

33-80 &

33 70Xo

W5397 W539G W5395

33'B0lo 33-50&

BRABANT I.

250m-

500»

STATION
SMITH

Fig. 24. Vertical section of salinity: Brabant Island to Smith Island, February 1929
WS398 W5397 WS39G WS335

1-5'

BRABANT

500

STATION
SMITH

Fig. 25. Vertical section of temperature: Brabant Island to Smith Island, February 1929

W539B W5397 W5396 WS395

2690 2680

2700

BRABANT I.

500m

Fig. 26. Vertical section of density (ct(): Brabant Island to Smith Island, February 1929.
other parts of the strait. It has originated in the west and consists of part of the
Antarctic surface water which flows out of the Bellingshausen Sea. The greater part of
this stream passes north of the South Shetland Islands and flows north-east through
the Drake Passage. Some, however, enters the Bransfield Strait between Low and

HYDROLOGY OF THE BRANSFIELD STRAIT 25

Smith Islands and between Snow and Smith Islands. All trace of Weddell Sea influence
is absent at this end of the strait. At all stations in this section the layer below the
surface water consists of a mixture of Antarctic surface water with increasing amounts
of warm deep water with depth. The shallow depth at these stations precludes the
presence of Antarctic bottom water. Table II gives a comparison of the salinities and
temperatures for the midway station of the three lines of stations taken in February 1929.

Table II

Depth

North-eastern section

Middle section

South-western section

in m.

St.WS38S

St.WS39i

St.WS396

S%o

(°C.

S°/oo

fC.

S%o

fC.

0

33-69

1-30

34-02

1-85

33'46

2-12

10

33-69

1-30

34-oi

1-89

33-47

2-05

20

33-86

!-i5

34-02

1 -87

33-49

I-9I

30

34-08

0-93

34-n

1-09

33-69

I-I3

40

34-12

0-85

34-15

°-59

3377

o-77

50

34-19

-0-05

34-i8

0-17

33-85

0-56

60

34-25

-o-37

34-21

0-04

33-95

0-29

80

34-3°

— 0-64

34-25

-0-31

34-05

-0-83

100

34-43

— 0-71

34-30

-0-85

34-!3

-0-63

15°

34H7

-o-8i

34-42

-o-86

34'32

— o-oi

200

34-52

-o-86

34-45

-0-98

34-43

0-32

300

34-59

-o-88

34-52

— 0-96

34-66

i-i8

400

34-59

— 1 -oo

34-54

-0-99

—

—

600

34-61

— I-I2

34-56

— 1-07

—

—

800

34-62

- i-33

34-57

- 1-13

—

—

1000

34-63

-1-52

—

—

—

—

1500

34-67

- 1-63

—

—

—

—

The temperatures and salinities below 100 m. on the north-eastern and middle
sections may be contrasted with those of the south-western section: at the midway
station in the two former there is a complete absence of warm deep water, whereas in
the latter, at St. WS 396, a temperature of 1-18° C. is recorded at 300 m.

An extra set of observations was made at St. WS 399, between Snow Island and
Smith Island. This station is situated in the relatively deep channel between these
islands and has a depth of 738 m. From the surface to a depth of 100 m. Antarctic
surface water is present ; below this depth it gradually mixes with the warm deep water,
which occurs in considerable volume and has an intermediate thermal maximum at
400 m.

NOVEMBER 1929

In November 1929 the hydrological survey made in February of the same year by
the R.R.S. 'William Scoresby' was repeated. As previously, horizontal sections of
salinity, temperature and density (at) have been drawn and are given in Figs. 27-9.
Just as the surface temperatures and salinities in February 1929 reflected summer
conditions, so the effect of late winter or early spring is seen in the surface values in
November. The low salinity and relatively high temperature of the surface water in

26

DISCOVERY REPORTS

Fig. 28. Surface temperature: November 1929.

HYDROLOGY OF THE BRANSFIELD STRAIT

27

February have been replaced by a high salinity and low temperature in November.
Thus on the King George Island to Trinity Peninsula line a difference of 0-29 °/00 and
2-20° C. exists between February and November in the surface water at corresponding
stations, WS 385 and WS 479, approximately 24 miles from Trinity Peninsula. In the
horizontal sections the conditions at the surface show the same characteristics that
prevailed in February when allowance is made for the increased salinity and decreased
temperature due to the difference in season. Lighter water is found at the surface at the
south-west end of the strait and at stations close to the South Shetland Islands, whereas
towards Trinity Peninsula the surface salinity increases and reaches the high value of
34-51 °/00 at the station nearest the peninsula. The surface temperature shows very

Fig. 29. Surface density (a(): November 1929.

little variation throughout the strait, the difference between the highest and lowest
temperatures recorded amounting to only 0-53° C.

The vertical sections from King George Island to Trinity Peninsula, Figs. 30-32,
show a much greater development of the warm deep water at this end of the Bransfield
Strait than in February. In February 1929 the only station on this line with a positive
intermediate maximum temperature in this layer was St. WS 382, some 4 miles from
King George Island, where a temperature of 0-53° C. was recorded at 400 m. The other
stations on the same line in February recorded only traces of intermediate thermal
maxima and all of negative temperature. In November, however, on this line, positive

4-2

28

DISCOVERY REPORTS

intermediate maxima are present at the following stations, Sts. WS 476, WS 477,

WS 478, WS 479 and WS 480, covering a distance of approximately 42 miles from

King George Island. At the next station, St. WS 481, the whole water column has been

STATION WS476 WS477 WS4-7B WS473 WS4Q0 WS4BI WS4B2

KING GEORGE I. 3400& 3400& 34I0Z 3420& 3430% 3440%, 34 47& TRINITY PENIN*

500m-

I000r<

I500v

2000n

Fig. 30. Vertical section of salinity: King George Island to Trinity Peninsula, November 1929.

STATION WS476 W5477 WS478 W5479 WS480 W548I WS482

KING GEORGE

-10° -1 16° TRINITY PENIN4

500

1000,

1500m

2000m

Fig. 31. Vertical section of temperature: King George Island to Trinity Peninsula, November 1929.

subjected to vertical mixing and from the surface to a depth of 400 m. the total variation
in temperature and salinity amounts to only o-io° C. and 0-06 0/oo respectively. Thus
in November the warm deep water was present almost as far as St. WS 481, where

HYDROLOGY OF THE BRANSFIELD STRAIT

29

completely mixed and homogeneous water of Weddell Sea origin was present. In the
vertical temperature section the warm deep water is seen to extend across the strait as
an attenuated wedge-shaped layer, the apex of the wedge sloping up towards Trinity
Peninsula.

STATION WS476 WS477 WS478
KING GEORGE I.

WS479

2740

WS4B0 WS4BI WS482

2750 2760 3770 2775 TRINITY PENIN*

'27 78

27791

500i

ID0O

1500

2000

Fig. 32. Vertical section of density (<r(): King George Island to Trinity Peninsula, November 1929.

The vertical sections of the line between Livingston Island and Trinity Peninsula are
shown in Figs. 33-5. As with the section farther to the north-east the densest surface

STATION

LIVINGSTON I.

W5483 WS484

WS485

5C»

ID00M

1500m

WS486 WS487
3410%. 34-20%

TRINITY PENIN*

Fig- 33- Vertical section of salinity: Livingston Island to Trinity Peninsula, November 1929.

water is again found at the station closest to Trinity Peninsula, St. WS 487. The com-
plete vertical mixing which was found at the corresponding station, St. WS 480, on the
King George Island line is not seen at St. WS 487. The warm deep water has its

30

DISCOVERY REPORTS

maximum effect at St. WS 484, where an intermediate maximum temperature of
0-25° C. was found at 150 m. At the other stations in this section only rudimentary
traces of intermediate maxima are found. The proportion of Antarctic bottom water in
the depths below 150 m., besides increasing normally with depth, also increases towards
the coast of Trinity Peninsula.

The vertical sections of the stations at the south-west end of the strait are shown in
Figs. 36-8. The surface temperature at St. WS 488, close to Brabant Island, is but

STATION

LIVINGSTON I.

WS483

WS4B4

WS485

W5486 WS467

TRINITY PENIN*

500m-

1000m

1500m

Fig. 34. Vertical section of temperature: Livingston Island to Trinity Peninsula, November 1929.

STATION

LIVINGSTON 1

WS4B3

WS484

WS4B5

2740

WS4BG WS487
2750

TRINITY PENIN*

1—

_===J3==

2760 -i

. /

500m-

~2780 ^^

■/

~2785.__

y ; _____

1000m-

icnn. ._

Fig. 35. Vertical section of density (<j(): Livingston Island to Trinity Peninsula, November 1929.

slightly higher than that at the east end of the strait close to Trinity Peninsula. The
surface salinity at the other stations of the section from Smith Island to Brabant Island
is very similar to that of the rest of the strait, except at the stations close to Trinity
Peninsula which show the influence of dense Weddell Sea water. Dilution by melting
ice in the spring has not yet taken place farther to the south-west in sufficient amount to
cause surface salinities of less than 33-80 °/00 to appear at this end of the strait, and more-
over this minimum value is only found at St. WS 488 close to Brabant Island. At all

HYDROLOGY OF THE BRANSFIELD STRAIT 31

stations in this section the layer between the Antarctic surface water and the sea-
bottom consists of warm deep water, there being no Antarctic bottom water present.

As in February a station, St. WS 492, was taken between Snow Island and Smith
Island. At this station the water below 200 m. consisted of warm deep water. It is

STATION

SMITH I

WS49I

W5490

W5489

WS488

250m-

500^

3400%,

33-90*.

BRABANT

STATION

SMITH

Fig. 36. Vertical section of salinity: Brabant Island to Smith Island, November 1929.

WS49I W5490 W5489 WS488

BRABANT I.

250,

500,

Fig- 37- Vertical section of temperature: Brabant Island to Smith Island, November 1929.

STATION

SMITH I.

WS49I

WS490

W5489 WS489

2730 BRABANT I.

250m

500

Fig- 38- Vertical section of density (at): Brabant Island to Trinity Island, November 1929.

probable that most of the warm deep water which exists in the Bransfield Strait enters
the strait in the channel 600-700 m. deep between Snow and Smith Islands and flows
south of Deception Island, possibly through a gap in the supposed ridge between

32

DISCOVERY REPORTS

Deception and Trinity Islands. Thus the two deep basins of the Bransfield Strait may
be connected by a flow of warm deep water. Further soundings on the line between
Deception Island and Trinity Island are urgently needed.

DECEMBER 1930

In late December 1930 a line of stations was attempted by the R.R.S. ' Discovery II '
from midway between Elephant and Clarence Islands to Joinville Island. Unfortunately
after St. 541 heavy pack-ice, which stretched as far as could be seen, prevented any
further stations being made in this direction. A new line of stations was then made
between Cape Melville and Trinity Peninsula and another between Snow Island and

Fig. 39. Surface salinity: December 1930.

Trinity Island. Temperature observations and samples of salinity and phosphate
content were taken at all stations, with samples for oxygen content determination at
alternate stations. Horizontal sections are given for temperature, salinity and density in
Figs. 39-41. The three lines of stations are rather too far apart to show the surface
movements in any detail. In general the surface salinity and density increase in all
three sections towards Trinity Peninsula, and as usual the surface water at the north-
east end of the strait is colder than that at the south-west. Compared with the observa-
tions made by the R.R.S. 'William Scoresby' in November of the previous year, the
surface in December is composed of warmer but more dense water. This is explained

HYDROLOGY OF THE BRANSFIELD STRAIT

33

Fig. 40. Surface temperature: December 1930.

Fig. 41. Surface density (<r(): December 1930.

34

DISCOVERY REPORTS

by the fact that the 1930 observations were closer to summer conditions by over a
month. The increased surface salinity is probably due to the presence of much more

STATION 537

539

540 541

3440& 3450%, 3453%.

500m

IDOOi

1500

Fig. 42. Vertical section of salinity: Elephant Island towards Joinville Island, December 1930.

STATION 537

500

1000

1500m

Fig. 43. Vertical section of temperature: Elephant Island towards Joinville Island, December 1930

STATION 537 538 539 540

"I

500

541

2779

1000m

1500

Fig 44. Vertical section of density (cr<): Elephant Island towards Joinville Island, December 1930.

HYDROLOGY OF THE BRANSFIELD STRAIT

35

pack-ice in this area in the spring of 1930 than in November 1929. Thus in December
1930 the open water was left more saline before subsequent dilution by ice melting
occurred. The vertical sections, Figs. 42-4, on the line from between Elephant and
Clarence Islands towards Joinville Island, show the isolines sloping up towards St. 541,
the nearest station to Joinville Island that could be worked. St. 541 is situated on the
broad continental shelf north-east of Joinville Island, and the water here, throughout
the whole vertical column, has been completely mixed and consists of Antarctic surface
water of Weddell Sea origin. In this section the presence of warm deep water in traces
is indicated by the slight inversions of temperature which occur at all stations on this
line except in the completely mixed water at St. 541. Sts. 538 and 540 show two
temperature inversions at different depths ; the intermediate maximum temperature at
the greater depth of St. 540 occurs at a depth of 500 m., which is within 10 m. of the

544 545

STATION 542 543

C.MELVILLE king george

500m

1000.,

1500.,

34£0£_!

547

34-53%,TRINITY PENIN*

2000n
Fig. 45. Vertical section of salinity : C. Melville, King George Island, to Trinity Peninsula, December 1930.

sea-bottom. Thus at this station the bottom water itself contains an admixture of warm
deep water. The increase in salinity with depth and the presence of temperature
inversions show a stable layering at all stations in this line, except at St. 541 where the
whole column is practically homohaline.

The vertical sections of the line from Cape Melville to Trinity Peninsula, Figs. 45-7,
show as usual the increase of salinity and decrease of temperature in the surface water
as the line progresses towards the coast of Trinity Peninsula, except at St. 543, some
15 miles from Cape Melville. At this station the current is flowing more to the north,
and positive temperatures are found in the first 30 m. The greatest increase in surface
salinity occurs between Sts. 545 and 546. At the latter station the 0-400 m. layer is
practically homohaline, whilst the uppermost portion of the layer has been warmed by
insolation. At St. 547 the whole vertical column, consisting of Antarctic surface water,

5-2

36

DISCOVERY REPORTS

is nearly homothermal and homohaline. At the other stations on the line more stable
conditions are observed and both warm deep water and Antarctic bottom water are

STATION 542 543 544

C. MELVILLE. KINGGEORGE I 00°

545 54G

-0-5°

547

-10° TRINITY PENIN*
0E°-

500m-

lOOQ

I50O

2000m-

Fig. 46. Vertical section of temperature: C. Melville, King George Island, to Trinity Peninsula,

December 1930.

STATION 542. 543 544

C.MELVILLE, king george i. 2750

545 546

2760 2770 2776

547

2779 TRINITY PENIN*

500i

1000m-

1500

2000

Fig. 47. Vertical section of density (ct(): C. Melville, King George Island, to Trinity Peninsula,

December 1930.

present. The warm deep water has its greatest influence at St. 542, where an inter-
mediate maximum temperature of o-oo° C. is found at 300 m. No thermal evidence of

HYDROLOGY OF THE BRANSFIELD STRAIT

37

this layer was observed at the next station, but at Sts. 544 and 545 small temperature
inversions below the surface layer indicate the influence of the warmer water. At St. 545
a second though small temperature inversion is seen at 500 m., or some 100 m. above the
sea-bottom. This corresponds to a temperature inversion, also at 500 m., which occurs
below the vertically mixed surface layer at St. 546. The presence of such an irregular
temperature series at Sts. 538, 540, 545 and 546 and the proximity of these stations to
the flow of pack-ice from the Weddell Sea, would seem to indicate that the level of the
upper boundary of the warm deep water may be depressed during a season when
vertical mixing is particularly easy, i.e. when the surface is freezing or being cooled by
pack-ice. This would account for the depth of the lower temperature inversion. Sub-
sequently new relatively warm water may appear at a higher level and form the upper
inversion. An alternative explanation may be that warm deep water from the Weddell
Sea flows over the ridge between Elephant and Clarence Islands to Joinville Island, and
thus at these stations warm deep water from both the west and the east is present.

STATION

SNOW I.

549

550

551

552

553

SOOn

IOOOn

1500s

34 I0%„

TRINITY I

Fig. 48. Vertical section of salinity: Snow Island to Trinity Island, December 1930.

Antarctic bottom water is found in considerable quantity at depths below the
intermediate temperature maxima.

The vertical sections of the line between Snow Island and Trinity Island are given in
Figs. 48-50. The line cuts across the smaller basin at the south-west end of the
Bransfield Strait. The surface water at all the stations has been warmed by the sun.
A cold nucleus of the Antarctic surface water, with temperature below — 0-50° C, is
seen in the two most northern stations and at the station nearest Trinity Island. The
depth of the minimum temperature increases southwards from Snow Island. In general
the surface salinity increases towards Trinity Island. Near Snow Island, below the
depth of the cold nucleus of the surface layer, the temperature increases and reaches the
relatively high value of 0-50° C. in the warm deep water at 400 m. at St. 550. The rise
in the sea-bottom at Sts. 551 and 552 apparently prevents any appreciable amount of
warm deep water from appearing at further stations in the section, as the intermediate

38

DISCOVERY REPORTS

temperature maxima at stations other than St. 550 are very faint. Probably the greatest
part of the warm deep water inside the strait at the south-west end flows eastwards past
St. 550, south of Deception Island, through a gap in the supposed ridge between
Deception and Trinity Islands. Below the level of the intermediate temperature
maxima at Sts. 550 and 553 the water contains increasing proportions of Antarctic
bottom water with depth.

STATION

SNOW I

552

553

500:

1000m-

500

STATION

SNOW

Fig. 49. Vertical section of temperature: Snow Island to Trinity Island, December 1930.

549 550 551 552

500

IOOQi

1500-

Fig. 50. Vertical section of density (ct(): Snow Island to Trinity Island, December 1930.

WATER MOVEMENTS
The treatment by dynamical methods of the data obtained from the various surveys
of the Bransfield area has been a difficult task. As may be seen from the bathymetric
chart, very great differences of depth occur at adjacent stations and it is impossible from
the present data to obtain sufficient figures from which satisfactory topographical charts
may be drawn of the sea surface in relation (say) to the 2000 or 1500 decibar surfaces.
Consequently some arbitrary treatment is necessary on account of the great differences

HYDROLOGY OF THE BRANSFIELD STRAIT 39

in depth at the stations. In future work in the Bransfield Strait and similar areas the
number of stations must be greatly increased, particularly on the steeply shelving
continental slopes leading to the main channel through the strait.

If two stations are considered whose depths differ greatly, it cannot be assumed that
along the potential surface cutting the bottom at the station of lesser depth the difference
of pressure is nil. That is to say the physical characteristics of the water between the
stations at depths below that of the shallower station affect the height or topography of
the various isobaric surfaces at all points between the stations. Consequently, when
determining the difference of height in the sea surface or any other isobaric surface, a
method must be employed in which the specific volume of the sea water below the
potential surface referred to above is taken into account. Such a method was worked
out by Jacobsen and Jensen (1926). These workers, however, assumed a linear variation
in the specific volume in both a horizontal and a vertical direction. A glance at the
vertical section of density in the Bransfield Strait shows that this is not always correct.
When the difference of depth between the stations is great the correction factor C in the
Jacobsen and Jensen formula generally becomes great and may be dubious. However,
by using this method the main currents in the Bransfield Strait were determined as
follows. Water from the Bellingshausen Sea flows in a north-easterly direction at the
south-western end of the strait, is diverted by the shelves of the islands of Brabant, Low
and Smith, and enters the Bransfield Strait between Smith and Low Islands. Another
inflow is in the relatively deep channel between Smith and Snow Islands. This water
continues up the strait in a north-easterly direction, after having passed north or south
of Deception Island. The north-easterly current along the southern coast of the South
Shetland Islands is relatively strong. Part of the water which has passed Deception
Island makes a characteristic bend to the south-east and sometimes reaches more than
half-way across the strait before it turns again to the north-west to join the current along
the South Shetland Islands. At the eastern end of the strait the influence of a current
from the Weddell Sea is apparent. Having thus obtained a general idea of the water
movements, it was decided to try and find a more exact method for determining the
topography of the various surfaces required.

Professor B. Helland-Hansen, Director of the Geophysical Institute at Bergen,
kindly gave me particulars of a method which he and Nansen had previously worked
out. A vertical section of the anomaly of specific volume is constructed. At depths
greater than 1500 or 2000 m. in an area such as the Bransfield Strait it is seen that the
isosteres are horizontal. In this case it may be assumed that the isobaric and isosteric
surfaces lie parallel to the level surfaces, and hence there are no acting forces and no
convection currents. The depth at which the isosteres become horizontal is then taken
as the datum level of zero current for all stations in the section. The isostere at this
datum level is then extended horizontally on either side to cut the bottom contour and
continue along the line of stations. Above the datum level the isosteres are usually
inclined ; they are continued until they cut the bottom contour, and from this point the
lines are drawn horizontal. In this way a series of levels and values is obtained which

40

DISCOVERY REPORTS

allows the computation of the dynamic heights of the sea surface and other isobaric
surfaces at shallow stations relatively to those at deep stations. Prior to the construction
of the vertical section of specific volume anomaly, the curves of the distribution of
temperature, salinity and density with depth were drawn. In a few cases the values of
temperature and salinity were smoothed when this was necessary, and this was done in
such a manner that the interpolated values of temperature and salinity were in good
agreement with the density and specific volume anomaly curves. The dynamic depths
of the various isobaric surfaces are calculated in the usual manner, adding to the known
values of the specific volume those which have been artificially obtained in the manner
described above. Fig. 51 illustrates the method for obtaining the extra values of
specific volume anomaly.

From the dynamic calculations one station could be selected at which the level of the
surface of the sea was lowest, i.e. the most dense water occurred at this station. By
taking the dynamic depths above the datum level it was then possible to obtain relative

7

Fig. 5 1 . Sketch showing method of extrapolation used in
obtaining specific volume anomalies.

heights of the sea surface at the various stations and find the difference between the
other stations and the station selected. Occasionally a datum station was chosen which
was not that of the most dense water. When the height of the sea surface above the
datum isostere at a station was lower than that of the datum station, this was expressed
as a minus difference of level in dynamic centimetres.

The two principal lines of stations in the Bransfield Strait were from King George
Island and from Livingston Island to Trinity Peninsula. By use of the Helland-Hansen
and Nansen method the relative height of the sea surface at stations along each line was
found. It was then necessary to find a connection between the two lines in order to be
able to construct dynamic charts. This has been done by combining a station from one
line with a station from the other, either by the Jacobsen and Jensen method, if the two
stations in the link were selected so that the difference in depth of the stations was small,
or by a direct comparison in the case of the deepest stations where the values of the

HYDROLOGY OF THE BRANSFIELD STRAIT

4i

specific volume anomalies were the same at the lowest depths. Table III gives the
stations in the two principal lines which were linked by means of the Jacobsen and
Jensen method.

Table III

Date

Stations

Difference in

depth of stations

in m.

April 1927
February 1929
November 1929

205-200
WS 387-WS 390
WS 480-WS 486

no
109

47

In the observations made in December 1930, two of the lines of stations were from
Cape Melville, King George Island, to Trinity Peninsula and from midway between
Elephant and Clarence Islands in the direction of Joinville Island.

NOVEMBER 1929 AND DECEMBER 1930

After a preliminary survey of the data it was decided to commence with the November
1929 observations and try to build up a composite picture of the lines of flow in the
strait by adding to the November charts the values in December 1930. This representa-
tion is open to criticism because of the fact that observations from two different years
are brought together. It is, however, impossible to give a picture of the flow, either into
or out of the strait, of Bellingshausen Sea water and of the influence of the water from the
Weddell Sea, unless stations on either side of the two principal lines of November 1929
are considered. The presence of Weddell Sea water at the north-east end of the strait
and its influence across the strait had been previously suspected, but it was not until
the values of December 1930 had been added to those of the November 1929 survey
that the turn of this water to the north across the strait was demonstrated. Table IV and
the explanation below show how the relative heights of the sea surface at stations in
November 1929 were calculated.

By using the Jacobsen and Jensen formula

C=-(Vb2- V<h),

where C = correction to be added to the difference of the depths of the isobaric surface
at the depth of the shallower station ; z = the difference in depth in metres of the two
stations ; Vb2 and Va2 = specific volumes in situ at the two stations at a depth equal to
that of the shallower station ; we obtain

WS 480-WS 486 = — 2-03 dynamic cm. (1).

By a direct comparison of the dynamic depth of the 1800 decibar surface we have

WS 480-WS 482 = 0-05 dynamic cm. (2).

Subtracting (1) and (2) we obtain WS 482-WS 486 = - 2-08.
Assuming WS 482 to be the basic station we obtain WS 482 = o-oo and WS 486 =
2-08.

42

DISCOVERY REPORTS

These values have been used to obtain the relative heights of the sea surface at the
stations of the November survey, and have been entered in column 4 of Table IV.

Table IV

(2)

(3)

(5)

(1)

Dynamic depth to

Height in

(4)

Relative height

Station

1800 decibar surface

dynamic cm.

in dynamic cm.

in dynamic m.

above St. WS 482

above St. WS 480

WS476

1744-0561

1 1- 1 1

11*11

11-06

WS477

1744-0263

8-13

8-13

8-o8

WS478

1744-0192

7-42

7-42

7-37

WS479

i743-9923

4-73

4-73

4-68

WS480

I743-94S5

0-05

0-05

o-oo

WS481

I743-9534

0-84

0-84

0-79

WS482

I743-9450

o-oo

o-oo

-0-05

Dynamic depth to

Height in

1300 decibar surface

dynamic cm.

WS483

in dynamic m.

above St. WS 487

S-oo

7-95

1261-0522

8-07

WS484

1261-0129

4-14

4-07

4-02

WS485

1261-0230

5-!5

5-08

5-°3

WS486

1260-9930

2-15

2-08

2-03

WS487

1260-9715

o-oo

— 0-07

— 0-12

Now St. WS 482 is a very shallow station, the bottom depth being 152 m., and in the
preparation of a series of topographic charts it is better to choose as a basic station one
whose depth is at least equal to the greatest depth represented in a chart of dynamic
topography. Since all the values in column 4 are purely relative, any station may be
arbitrarily chosen as the basic station, and in column 5 of the above table St. WS 480
has been thus selected and the values at the other stations correspondingly altered.

The lines of stations in December 1930 from Cape Melville to Trinity Peninsula and
from between Elephant and Clarence Islands in the direction of Joinville Island were
linked by direct comparison of the dynamic depths to the 2000 decibar surface. In this
case St. 543 was chosen as the basic station, and Table V contains the heights of the sea
surface of the other stations relative to this station.

It has been previously mentioned that the results from November 1929 would if
possible be linked with those of December 1930, so as to obtain a more complete
picture of the water movements. By inspection of the vertical sections of the anomaly
of specific volume for the two lines from King George Island in these two years, it was
assumed that at 2000 m. the specific volume anomaly in each case was 10 x io~5 and
that no current existed at this depth in either year. Then, by a direct comparison of the
dynamic depth to the 2000 decibar surface for St. WS 478 in November 1929 and
St. 543 in December 1930, we have the following relation for the difference of level in
dynamic cm. of the sea surface at these two stations :

WS 478-St. 543 = 7-25 dynamic cm.

HYDROLOGY OF THE BRANSFIELD STRAIT

43

Then, if we allow the relative heights for November 1929 to remain, we can use the
value WS 478 = 7-37, which was obtained in Table IV, column 5, in the above relation
between St. WS 478 and St. 543. Thus we obtain St. 543 = 0-12 relative to the
November 1929 values. As we have already obtained relative values for December 1930
(Table V) this new value (St. 543 = 0-12) can be substituted in these values, so that in
Table VI we have a direct comparison between November 1929 and December 1930.

Table V

Height in dynamic

Station

cm. above surface

of sea at St. 543

538

7-17

539

3-60

540

0-14

S4i

-o-57

542

5-59

543

o-oo

544

— o-io

545

- 1-27

546

-2-57

547

-4-64

It only remains to link the line of stations from Snow Island to Trinity Island taken
in December 1930 with the line of November 1929 from Livingston Island to Trinity
Peninsula. This was done by a direct comparison of the dynamic depths to the 1300
decibar surface of Sts. 550 and WS 483 as follows:

St. 550-WS 483 = 3-53 dynamic cm.;
but in Table IV, column 5, we have WS 483-WS 480 = 7-95 dynamic cm. By addition,
therefore, St. 550-WS 480 = 1 1-48 dynamic cm. Thus we have now a complete relative
relationship for the level of the sea surface for the majority of the stations in November
1929 and December 1930 above that at St. WS 480. These figures are given in Table VI
below.

The values in Table VI were entered on a chart of the area and dynamic isobaths were
drawn at intervals of every dynamic centimetre. In the construction of the isobaths
attention was paid to the law of Ekman which states the change of direction of flow
of a convection or gradient current when approaching shallowing or deepening water.
In the southern hemisphere a gradient current will turn to the left on approaching a
shallower part of the ocean and to the right when approaching deeper water. Topo-
graphical charts for surfaces other than that of the sea surface were made on the basis
of the above table by using the differences of dynamic depth to these surfaces of the
various stations and the basic station WS 480. In this way topographic charts for the
levels of o, 50, 100, 200, 300, 400 and 600 m. were constructed and are given in
Figs. 52-8. On account of the fact that observations from two different years have been
used in these charts, it is not strictly accurate to represent the dynamic isobaths as
continuous lines. Consequently the junction of the November and the December

6-2

44

DISCOVERY REPORTS

results is shown by a definite break in the lines on these charts, indicating the uncertainty
in the combination of the results from two years. Arrows have been added to the
dynamic isobaths to indicate the direction of the currents.

In the topographic chart of the sea surface the same current features are shown as
were previously mentioned, but the great influence of water from the Weddell Sea is
especially emphasized. When the line of stations from Elephant and Clarence Islands
was being taken, it was impossible to make observations as far as Joinville Island because
very heavy pack-ice was met just beyond St. 541. This pack-ice must have come from
the Weddell Sea, and its present position was the resultant of wind and current direction.
The combined observations from November 1929 and December 1930 show a current
of Weddell Sea water setting to the west between Sts. WS 481 and WS 480, which
turned back close to the latter station and flowed to the north, joining the current along

Table VI. Relative heights in dynamic centimetres of the level of the sea surface

above that at St. WS 480

Novembei

1929

December 1930

St. WS 476

11 -06

St. 538

7-29

WS477

8-o8

■539

372

WS478

7-37

54°

0-26

WS479

4-68

S4i

-o-45

WS480

o-oo

542

57i

WS481

079

543

0-12

WS482

-0-05

544

0-02

WS483

7-95

545

- I-I5

WS484

4-02

546

-2-45

WS485

5-°3

547

-4-52

WS486

2-03

549

I0-O9

WS487

— 0-12

550

11-48

55i

5-91

552

4-42

553

5-46

King George Island from the south-west. On the present evidence it is difficult to
estimate how much of the current close to Cape Melville flows counter-clockwise round
the South Shetland Islands, and how much flows north of Bridgeman Island and then
to the east. Between Sts. 538 and 541 a current towards the east is shown, but the
dynamic isobaths to the north of St. 539 have not been joined up with the corresponding
lines close to King George Island because it is impossible at present to estimate the
direction of flow of water across the probable ridge between Elephant and King George
Islands. There is no doubt of the anti-clockwise circulation around the islands of the
South Shetland group. The observations at St. 537 between Elephant and Clarence
Islands have not been included in these charts because this station is situated in a
relatively deep basin between these islands and is shut off to the south by a ridge which
prevents free circulation with the water to the north-east of the Bransfield Strait.
The northerly set of the current between Joinville and King George Islands explains

HYDROLOGY OF THE BRANSFIELD STRAIT

45

Fig. 52. Relative topography of the sea surface: November 1929 and December 1930.

Fig. 53. Relative topography: 50 m., November 1929 and December 1930.

46

DISCOVERY REPORTS

Fig. 54. Relative topography: 100 m., November 1929 and December 1930.

Fig. 55. Relative topography: 200 m., November 1929 and December 1930.

HYDROLOGY OF THE BRANSFIELD STRAIT

47

Fig. 56. Relative topography: 300 m., November 1929 and December 1930.

Fig. 57. Relative topography: 400 m., November 1929 and December 1930.

48

DISCOVERY REPORTS

the existence of the belt of pack-ice which the whaling fleet finds in this position at the
commencement of the season.

The representation of the surface currents of the Bransfield Strait from the present
data is not complete, as the gaps between the lines of stations are too great. It must be
regarded as being purely tentative and will no doubt be modified in the light of further

data.

Between the cold surface layer and the bottom layer the relatively warmer and more
saline layer exists in the Bransfield Strait at a level closer to the surface than in the sea
outside. In November 1929 this warm intermediate layer was found in greatest amount

Fig. 58. Relative topography: 600 m., November 1929 and December 1930.

at the west end of the strait between Snow and Smith Islands (maximum temperature,
1-07° C. at 400 m.) and between Low and Brabant Islands (0-55, o-88 and 0-97° C. were
recorded at 200-300 m.). In the same month, on the line between King George Island
and Trinity Peninsula, this water extended from the station nearest the former island,
St. WS 476, where a temperature of 0-75° C. was found at 300 m., to St. WS 480
nearly two-thirds of the distance across the strait, where a maximum temperature of
o-io° C. was found at a depth of 100 m. In December 1930 this warm intermediate
layer was again in greatest amount at the south-west end of the strait, a temperature of
0-50° C. being recorded at 400 m. at St. 550 between Snow and Trinity Islands. On the
other hand, all the other stations in the December 1930 survey show poorly developed

HYDROLOGY OF THE BRANSFIELD STRAIT 49

and minus temperature maxima, or none at all. Thus both the November 1929 and
December 1930 observations show that the intermediate layer of warm deep water
occurs at the south-west end of the strait in greatest amount ; in the former year in
November there is evidence for its presence over a large area between King George
Island and Trinity Peninsula, whereas in December 1930 its existence over this area
is not shown by well-marked temperature maxima. This is one of the difficulties
encountered in combining material from two different years from an area where there are
large variations from year to year in the physical characteristics of the water masses. It is
possible that the warm intermediate layer present in the Bransfield Strait comes from both
the west and the east, but the flow from the western end is undoubtedly much the greater.
Owing to the irregularity and steepness of the bottom contour there are not many
observations below 600 m. to show the movement of the bottom water in the strait. As
previously mentioned the movement of this water must be restricted by the submarine
ridges. It might be supposed that the bottom water in the deepest parts of the deep
basins would be stagnant. This, however, is far from the actual case, as the oxygen
content at great depths shows. Thus at St. 543, approximately 15 miles from Cape
Melville, the oxygen content increases from a depth of 600 m. to the lowest observation
at 1500 m., where it is 6-43 c.c. /litre, equal to a saturation of 76-3 per cent. This
indicates that the bottom water must have been lately renewed, and movement of the
bottom water must take place whenever the older water is replaced by a fresh supply of
highly aerated water. In considering the movement of the bottom water within the
strait, its mode of formation must be sought. H. U. Sverdrup, in his report on the waters
of the North Siberian Shelf (1929), noted that instabilities were found in the surface
layer and that the accuracy of the salinity and density determinations was such that
these instabilities could not be otherwise than real. Moreover, the instabilities were
only observed in the cold season when freezing takes place, and never in summer. A
series of tables and graphs was given by the same author which show that the formation
of these instabilities is clearly connected with the freezing processes. He also states
(1929, p. 67) "The salinity of the water increases when ice forms, but the heavy brine
does not sink immediately to the level of equilibrium, the sinking may be very slow and
irregular and instabilities are so frequent that they are found in more than one-third of
all cases. It is characteristic that the greatest numbers of instabilities are found in the
earliest part of the winter when ice is freezing over many open lanes, while the greatest
numerical values of the instabilities are found in the middle of winter, when the
number of openings is smaller but freezing is more rapid ". It is probable that a simdar
state of affairs occurs over the continental shelf of western Graham Land and Trinity
Peninsula, and indeed over the shallower portions of the strait, in winter. In 1929, even
in November, the surface values of temperature, salinity and density (a,) at St. WS 482
were - 1-18° C., 34-51 °/00 and 2778, and in December 1930 at St. 547 the corre-
sponding values were - 1-02° C, 34-53 7oo and 2779- These two stations were
situated on the continental shelf of Trinity Peninsula. The freezing-point of sea water
of salinity 34-50 °/00 is approximately - i-88° C, and thus the surface water at these

So DISCOVERY REPORTS

two stations has already been considerably warmed; it has also without doubt been
diluted. Typical temperature, salinity and density values of the bottom water at 1500 m.
in the deepest part of the strait were found at St. 543 to be - 1-72° C, 34-65 °/00 and
27-91. The temperature of this water is higher than the freezing-point of water of
salinity 34-50 °/00. It is not difficult to imagine that in winter the surface salinity
increases to a value greater than that of the bottom water and the temperature decreases
to the freezing-point of such water, with the consequence that in areas where lanes of
open water are found, sudden and intense freezing takes place with the formation of
instabilities. In November and December we have noted that the water on the conti-
nental shelf is still very saline and cold from surface to bottom. In the winter months,
water still heavier and of uniform temperature and salinity will be formed on the shelf.
Within this layer intermittent sinking of water of higher salinity and lower temperature,
formed by freezing processes, will take place and this heavier water will follow the slope
of the bottom contour. In this way the bottom water in the strait is renewed and
maintains its high oxygen content. This renewal of the bottom water will cause the
water above to flow out of the strait, and it is significant that the oxygen content is at a
minimum at a depth of 400-600 m. at St. 543 in the deepest part of the strait. This
depth almost corresponds with the depth of the gap in the ridge south of Clarence
Island through which the displaced water must flow.

Thus in the Bransfield Strait the formation of the bottom water is considered to take
place in winter as a result of freezing processes mainly on the extensive continental shelf
of west Graham Land and Trinity Peninsula. This mode of formation probably takes
place also in other parts of Antarctica, and in particular some of the Weddell Sea
Antarctic bottom water is undoubtedly formed on the continental shelf in the southern
part of this sea as suggested by Brennecke (1921, p. 38). From the observations hitherto
made it cannot be determined whether a renewal of the bottom water in the deep basins
of the Bransfield Strait may also be affected by horizontal flow across the ridges,
particularly across that to the east of the strait.

FEBRUARY 1929
The movement of the surface water in February 1929, as shown by the topography of
the surface of the sea for the two principal lines of stations (Fig. 59), shows in general the
two main features noticed for November 1929 and December 1930; there are, however,
some modifications. The characteristic bend of the dynamic isobaths on the line from
Livingston Island occurs almost as far across the strait as Astrolabe Island, so that the
north-westerly flow of the surface water back to the north-easterly flowing current
along the South Shetland Islands is found over a greater distance than in November
1929. The north-easterly flow on the South Shetland side of the strait is concentrated
chiefly between the second and third stations from King George Island. A difference
of height of 9-6 dynamic cm. occurs between these two stations, which are only 5-8 miles
apart, and this corresponds to a surface current of 69 cm./sec. The gradient may be too
great owing to considerable stray on the water-bottle wire at St. WS 383 causing the

HYDROLOGY OF THE BRANSFIELD STRAIT

Si

Fig. 59. Relative topography of the sea surface, February 1929.

Fig. 60. Relative topography: 50 m., February 1929.

7-2

52

DISCOVERY REPORTS

deeper samples to be obtained from lesser and incorrect depths. However, when the
vertical section of the anomaly of specific volume was constructed, an attempt was made
to correct for this by drawing the isosteres at this station closer to the surface, and using
the interpolated values so obtained in the calculation of the dynamic depths to various
surfaces at this station. There is no doubt that in February 1929 the current to the
north-east in the strait was concentrated close to St. WS 383 in contrast to the wide-
spread nature of this current in November 1929.

The lack of data between the lines from King George Island and Livingston Island
gives an impression of the absence of movement in this area. In all probability whirls of
movement actually do exist here, although it is impossible to furnish any proof. The

Fig. 61. Relative topography: 100 m., February 1929.

influence of the Weddell Sea water on the current in the surface is shown by the
northerly component of the dynamic isobaths towards King George Island. This
influence is shown to occur somewhat farther into the strait and to stretch over a larger
area across the strait than in November of the same year. A circular movement in an
anticyclonic direction has been drawn round St. WS 386, which is in agreement with the
very light water found in the upper layers at this station. Figs. 60 and 61 show the
relative topography at 50 and 100 m.

The movement of water in February 1929 (Figs. 62-5) differs considerably from that
in November of the same year at depths below 200 m. In February at these depths very

HYDROLOGY OF THE BRANSFIELD STRAIT

53

Fig. 62. Relative topography: 200 m., February 1929.

Fig. 63. Relative topography: 300 m., February 1929.

54

DISCOVERY REPORTS

54°

Fig. 64. Relative topography: 400 m., February 1929.

Fig. 65. Relative topography: 600 m., February 1929.

HYDROLOGY OF THE BRANSFIELD STRAIT 55

little flow is shown coming from the south-west except close to the South Shetland
Islands, and the movement from the east end of the strait appears to predominate. This
would account for the entire absence of any intermediate temperature maximum at all
stations along the line from Livingston Island in February, except at St. WS 393, which
is nearest to the South Shetland group and owes its positive temperatures between
100 and 300 m. to water from the south-west. In the discussion of the vertical sections
of temperature and salinity attention was drawn to the great difference in the tempera-
ture-salinity diagrams of Sts. WS 393 and WS 392.

APRIL 1927

A discussion of the water movements in April 1927 is rendered extremely difficult by
the conflicting observations on the line from King George Island to Trinity Peninsula.
Attention has already been drawn to the fact that several salinities are abnormally high,
probably due to the partial freezing of the sea water in the water bottles before a sample
was withdrawn. The observations were made in very unfavourable conditions and it
is probable that some of them are unreliable. When the vertical section of the anomaly
of specific volume was constructed the salinity and temperature curves were smoothed,
but the result as shown in the topographic charts is not happy. The levels of the
various surfaces at St. 197 as shown in the charts are too low and those at St. 198
too high, with the consequence that the dynamic isobaths show a large whirl at all
depths at St. 198. This is not at all likely in view of the results from other years.
The characteristic bend of the dynamic isobaths along the line from Livingston Island
which was noted in both of the surveys in 1929 is once more shown by the observations
in 1927. For this season one chart only, that of the surface movements, is reproduced
(Fig. 66).

It must again be stressed that in the absence of more complete data the movements of
water as outlined above are to be regarded as purely tentative. It is recognized that the
dynamic isobaths could have been drawn in other ways, and it is possible that future
work will considerably modify the directions of the flow which have been deduced.

THE DIFFERENCES IN TEMPERATURE AND SALINITY
IN FEBRUARY AND NOVEMBER 1929 ALONG THE NORTH-
EASTERN LINE OF STATIONS

Two hydrological surveys were made in 1929 : one in February when typical summer
conditions were noted, and the other in November which reflected late winter or early
spring conditions. It has already been noted in the discussion on water movement that
the chief differences in these two surveys occurred at the north-eastern end of the strait.
For this reason it was thought that a comparison of the observations in February and
November at stations which were as nearly as possible in the same position on the line
from King George Island would be of interest. The stations which are available are the

56

DISCOVERY REPORTS

following pairs: WS 382 and WS 476, WS 383 and WS 477, WS 384 and WS 478,
WS 385 and WS 479, WS 387 and WS 480. Attention has previously been drawn to
the fact that the observations at St. WS 383 are inaccurate owing to the great stray on
the water-bottle wires. Consequently this station and its corresponding station in
November, WS 477, have been omitted from this discussion. Table VII gives the
difference, multiplied by 100, of the temperatures and salinities of the February stations
from those of November. The figures in round brackets indicate that one or both of the
observations in question have been interpolated.

From this table it will be noticed that at all stations the surface layer is warmer in
February than in November, which is mainly a seasonal difference. The thickness of the

Fig. 66. Relative topography of the sea surface, April 1927.

layer of higher temperature in February is greater to the north-west (Sts. WS 382 and
WS 476) and becomes thinner across the strait as far as Sts. WS 387 and WS 480 ; this
is due to the greater development of the relatively warmer north-east current along
the north-western side of the strait in February.

At all stations the layer immediately below the surface layer is warmer in November, be-
cause of the increased thickness and spread across the strait of the warm deep water in that
month. The excess of temperature in November over that of February is least marked
at the station closest to King George Island, where a slight negative difference (February-
November) is found only at 300 m. (At = — o-oi° C). The thickness of the layer in

HYDROLOGY OF THE BRANSFIELD STRAIT

57

which the November temperatures are higher increases at first and then gets less towards
Trinity Peninsula and the layer is also closer to the surface in the same direction.

At all stations the layer below the warm deep water consists of warmer water in
February than in November. This is probably due to the effect of the formation of
Antarctic bottom water in winter and consequently the temperature of this layer is less
in November than in February.

Close to the land at Sts. WS 382 and WS 476 the salinity in the 0-28 m. layer is
higher in November than in February. This may be (i) because in November the
surface water is not so much diluted by ice melting as in February; (ii) because in
February the surface water may be showing the effect of dilution from the coast; or
(iii) because the surface water just south of this position is moving in February at a
higher rate than in November and consequently the surface water to the left of this
current forms both a lighter and a deeper layer in February than in November.

Table VII

WS382

-WS 476

WS 384-

WS478

WS 385

-WS 479

WS 387-

WS480

A

A

A

A

A

A

A

A

100 f

i°°S°/00

100 f

iooS°/00

100 f

100 S%0

100 f

iooS°/00

0

227

- n

299

3

220

-29

207

— 12

10

226

- 10

302

3

220

-28

192

- 13

20

213

1 1

291

3

205

- 11

129

(-8)

3°

174

2

202

10

184

11

38

-8

40

153

5

149

18

i77

14

2

-9

50

152

8

i34

18

87

4

— 20

— 20

60

142

9

106

23

55

3

-53

(-29)

80

139

8

57

13

25

0

-96

(-30)

100

133

1 1

3

17

1 1

9

- i37

-26

15°

82

-4

-83

5

— 21

5

-87

(-15)

200

(33)

(-0

-96

-3?

-49

2

(-61)

(-9)

300

— 1

— 2

-42

(5)

-87

(3)

14

(-8)

400

33

5?

(-40

4

- 15

1

— 1

(-6)

600

—

—

19

(3)

23

0

12

(-5)

800

—

—

32

1

24

1

—

—

1000

—

—

9

2

15

(2)

—

—

1500

—

1

10

6

At the next pair of stations to the south-east (Sts. WS 384 and WS 478) the water
throughout, with the possible exception at a depth of 200 m., is more saline in February
than in November. As far as the surface layer is concerned this cannot be due to ice
melting in November because the greatest difference occurs at a depth of 60 m.,
whereas if it were due to ice melting in November the greatest difference would be at
the surface itself. The dynamic charts for February, besides showing the Weddell Sea
influence occurring farther across the strait towards King George Island than in
November, also show an extremely steep gradient close to St. WS 383. In this case the
velocity of the north-east current at St. WS 383 is very great in February and conse-
quently the layers of equal density will slope down towards the north-west and up

S8 DISCOVERY REPORTS

towards the south-east. Consequently in February denser water throughout the column
will appear closer to the surface on the right-hand side of this current, i.e. at St. WS 384.
At the next pair of stations (Sts. WS 385 and WS 479) the 0-25 m. layer is much more
saline in November than in February owing to a local accumulation of low salinity
water in the upper layers in February. The salinity throughout is greater in November
than in February at the pair of stations Sts. WS 387 and WS 480. As far as the surface
layer is concerned this is due to the greater salinity of Weddell Sea water in November
than in February.

Below the first 20 or 30 m. in the surface layer, the salinity is greater in February than
in November at all stations except at the pair Sts. WS 387 and WS 480. With the
exception of the pair Sts. WS 382 and WS 476 this is explained by the greater penetra-
tion of Weddell Sea water in February across the strait towards King George Island,
which is shown by the dynamic charts. These charts, however, do not show the effect
of mixing and sinking of water masses and hence the higher salinity of February over
November in the position of St. WS 382 on the left-hand side of the strong north-east
current may be due to mixture of the Weddell Sea water in February with the Bellings-
hausen Sea water on the western side of the strait.

The vertical sections show that in November the warm deep water, which lies
immediately below the surface layer, was present over a greater area of the strait than in
February; it has its maximum thickness at Sts. WS 476 and WS 477 on the northern
side of the strait. If we compare the salinities of this layer for February and November
we find that slightly more saline water is present at St. WS 476 in November than at
St. WS 382 in February, which is due to the greater development of this water in
November. At the next pair of stations (Sts. WS 384 and WS 478) the greater flow of
warm deep water below the surface layer in November is shown chiefly by the higher
temperatures of November over February in the layer approximately between 100 and
560 m., whereas only at 200 m. is the salinity higher in November. The dynamic charts
show that the stations in the middle of the north-eastern line are affected in February
by Weddell Sea water to a much greater extent than in November, when Sts. WS 478
and WS 479 are influenced by water from the south-west. The pair of stations closest to
Trinity Peninsula (Sts.WS 387 and WS 480) show that throughout the whole column
the salinity of November is higher than in February, no doubt due to the higher
salinity of Weddell Sea water in November than in February.

At the stations in the north-western half of this line the bottom water is more saline
in February than in November, whereas in Sts. WS 387 and WS 480, which are on the
continental slope of Trinity Peninsula, the reverse is the case, i.e. November is higher
than February. The life history of this water mass is not sufficiently known to enable a
discussion of the variations in salinity of the bottom water to be made.

In the absence of data from the same months in other years the differences in
temperature and salinity which have been noted above cannot be assigned solely to
seasonal differences, as the influence of the great change in current movements in the
strait from February to November may be due to variations other than seasonal. It is

HYDROLOGY OF THE BRANSFIELD STRAIT 59

by no means certain that the same differences in temperature and salinity noted would
be repeated every November and February, although it may seem probable that the
current from the Weddell Sea is stronger and reaches farther westward in February
than in November. Our material does not allow of a more detailed discussion of seasonal
variations.

NOTE ON THE PRESENCE OF ICE IN THE BRANSFIELD

STRAIT AND ON THE CURRENTS THROUGH THE STRAITS

SEPARATING THE SOUTH SHETLAND ISLANDS

The operation of scientific vessels and the whaling fleet inside the Bransfield Strait
has been limited to the short season November to April, and as far as our knowledge
goes no ships have been present in this area in the southern winter. Much ice prevails
in the strait in early spring and late summer and it is not known whether the strait is
entirely closed to navigation in winter. According to a private communication from
Mr Nielsen of the Hector Whaling Company, who has spent many whaling seasons at
Deception Island, it is his opinion that the strait is never completely closed to navigation.

The approach of winter conditions in the Bransfield Strait is marked by the appearance
of pack-ice at the north-eastern end of the strait ; the ice then spreads across the strait at
this end and down the Trinity Peninsula coast. When this accumulation of pack-ice has
rendered whaling too difficult the whaling fleet has usually ceased operations for the
season and has left the area. On their return in spring the whalers find, in years when
not much ice is present, that the Bransfield Strait is either free or easily navigable at the
end of October or the commencement of November. When, however, much ice is
present the whaling factories have been considerably delayed in their endeavours to
reach their anchorages in Deception Island or Admiralty Bay in King George Island,
and may only be able to force their way into the strait in late November.

The position of the pack-ice in the south-western part of the strait in autumn and
early spring appears to be regulated chiefly by the direction of the wind. Thus in the
whaling season 1920-1 northerly and north-easterly winds prevailed from the end of
November 1920 to the beginning of February 1921 and the ice which earlier had been
carried north-east in calm and fine weather by the current close to the southern side of
the South Shetland Islands was driven back in a south-westerly direction. The ice was
still present in the eastern part of the Bransfield Strait in late February. On the other
hand in the season 192 1-2 the south-east side of the strait was free on November 25
1921 while the whole of the northern side was blocked, these conditions prevailing until
January 1 1922, when the ice finally moved out in a south-westerly direction. The ice in
the Bransfield Strait normally moves out of the strait to the north-east. In December
1923, contrary to usual, ice began to form in the sea east of Trinity Peninsula and
Joinville Island towards Cape Melville. In March 1924 ice was moving to the south-
west right into the middle of de Gerlache Strait. Had the north-east current not been
flowing in the Bransfield Strait the whaling ships would have had to leave Deception
Island as at certain times the ice pressed right up against the island. The easterly parts

8-2

60 DISCOVERY REPORTS

of the strait and the Trinity Peninsula coast were blocked throughout nearly the whole
season. Thus the ice conditions inside the Bransfield Strait vary very considerably from
year to year.

The pack-ice outside the strait on the northern side of the South Shetland Islands
usually exists in a semicircular shape running north-east and south-west of these islands.
This ice is usually still present until after the Bransfield Strait is either navigable or free
from much hindrance.

When a succession of north-easterly winds is absent, or when no fresh accumulation
of pack-ice is formed at the north-east end of the Bransfield Strait, the speed with which
the ice moves out of the strait is remarkable. Thus on November 7 1923 the Bransfield
Strait was reported full of thick pack-ice and large icebergs, whilst on November 16
only scattered ice remained which was no hindrance to navigation.

The whaling fleet has usually been able to work in and south of the de Gerlache Strait
in January.

The Antarctic Pilot (1930), in giving the opinion of the whaling captains, states
" . . .during November and the first half of December only a narrow strip of ice-free
water exists from Martins Head [King George Island] to Deception Island, enabling
vessels to pass through Nelson Strait and make Admiralty Bay ; but in some seasons the
bay itself is shut off by ice extending from Telefon Rocks [King George Island], in which
case Port Foster, Deception Island, may be reached instead. This ice is sea ice and bergs
drifting east-north-east close in to the south side of King George Island and thence to
the south side of Elephant Island. ... All this ice appears to come from outside Palmer
Archipelago. The main pack comes eastward, past the south side of Deception Island,
but a little may pass northward of the island" (p. 71). "The general current in de
Gerlache Strait is to the north-east. Ice from de Gerlache Strait goes on north-eastwards
through Orleans Channel, where many bergs are stranded. . . " (p. 84).

Two main current features are to be seen in the Bransfield Strait ; part of the water
from the Bellingshausen Sea enters the strait between the islands of Snow, Smith and
Low and flows both north and south of Deception Island, where it makes a characteristic
bend to the south-east before returning to the north-west, and then flows as a north-east
set along the southern shores of the South Shetland Islands. The second great current
movement comes from the Weddell Sea, bringing a large amount of cold dense water
round Joinville Island which then spreads across the north-eastern end of the strait and
down the coast of Trinity Peninsula.

The positions of some of the stations worked by the research ships have obviously
been affected by these currents, which, however, are variable in strength. Thus during
the time the R.R.S. 'Discovery II' was in the Bransfield Strait in April 1930 an
allowance of 1-2 knots was made for the speed of the north-east flowing current close to
the South Shetland Islands, during a passage across the strait from south-east to north-
west. The result was that the final position of the ship was too much to the south-west
and no allowance need have been made at all. A very considerable belt of pack-ice had
been observed to the east and this may have had the effect of slowing down very greatly

HYDROLOGY OF THE BRANSFIELD STRAIT 61

the north-east setting current. Similarly an accumulation of pack-ice to the south-west
might also have had the same effect.

Normally, however, the north-easterly current piles up water against the South
Shetland Islands and through the various narrow straits separating these islands. Thus
the main part of the current along the South Shetland Islands sets to the north-east and

Fig. 67. Diagrammatic sketch of circulation round the South Shetland Islands

and position of overfalls.

some passes through the individual straits of the islands curving to the south-west on
the northern shores of the islands, causing an anticlockwise circulation. This is shown
diagrammatically in Fig. 67.

Fig. 68. Effect of north-easterly wind on currents in Nelson Strait.

The south-west going stream on the northern side of the islands has a tendency to

flow southwards down the various straits, just as the north-east setting current in the

main Bransfield Strait has also a tendency to pass through the individual small straits

but in a contrary direction. Both currents then meet outside the narrow straits on the

62 DISCOVERY REPORTS

northern side of the South Shetland Islands, with the result that at the junction heavy
overfalls are seen. When, however, north-easterly winds predominate, the south-west-
going stream on the northern side of the island is considerably strengthened, with the
result that this current overcomes the northerly set passing up the narrow straits, and
passes into the individual straits. When this happens the two contrary directed currents
meet inside the narrow straits and it is here that the heavy overfalls are now found.

In this latter condition a prominent line of demarcation is to be seen between the two
currents, showing that on the western side of the narrow straits there is a north-going
set and on the eastern side there is a south-going set. During the time the R.R.S.
' Discovery II ' was lying at anchor in Nelson Strait we observed this last condition very
well. A north-east wind commenced to blow and later the northerly set could be seen
to be gradually overcome by the now strengthened southerly current.

SUMMARY

i. Previous workers have suggested the probability that the Bransfield Strait consists
of an enclosed basin. This is confirmed and it is shown that two such basins exist,
bounded by islands and submarine ridges. The effect of the bathymetric features is such
that the strait is cut off from free circulation with the outside seas. As a result the
temperature below the surface layer is abnormally low, the horizontal circulation is
restricted very considerably at depths below those of the confining ridges and the
amount of warm deep water within the strait is thus considerably less than exists outside,
the lowest layer, consisting of Antarctic bottom water, being correspondingly increased.
Temperature-salinity diagrams for stations both inside and outside the strait are given
in illustration of the differences in composition of the water masses in this area.

2. Vertical sections of temperature, salinity and density show in general an increase
of salinity and density and a decrease of temperature in the surface layer from the South
Shetland Islands towards Trinity Peninsula. The amount of warm deep water is greatest
at the south-west end of the strait and close to the South Shetland Islands, but decreases
towards Trinity Peninsula. In February 1929, on the line from King George Island,
thermal evidence of the presence of warm deep water was restricted to the station nearest
the island, whereas in November of the same year the vertical section of temperature
showed a wedge of this water reaching to at least within 24 miles of Trinity Peninsula.

3. Apart from the effects of the bottom topography the chief features of the circula-
tion in the strait appear to be due to two main currents. At the south-western end water
from the Bellingshausen Sea enters the strait between Low, Smith and Snow Islands
and flows both north and south of Deception Island, finally appearing as a north-east
going set close to the southern shores of the South Shetland Islands. At the eastern end
of the strait a current of cold, more dense water, whose origin is the Weddell Sea, flows
round Joinville Island and across the strait, some of it spreading down the Trinity
Peninsula coast. In the available data there is one feature of the water which later forms
the north-east set close to the South Shetland Islands, which appears to be permanent.
The surface water after passing Deception Island makes a characteristic bend to the

HYDROLOGY OF THE BRANSFIELD STRAIT 63

south-east before returning to the north-west and then flows north-east past the South
Shetland Islands.

4. Charts of dynamic topography were constructed according to a new method of
Professor Helland-Hansen. By combining the hydrographic results of November 1929
and December 1930 in these charts the two chief currents and in particular the Weddell
Sea influence in the strait were demonstrated. Charts showing the movement of water
at the levels of o, 50, 100, 200, 300, 400 and 600 m. are given for February 1929,
November 1929 and December 1930, the two latter being combined. In February 1929
the characteristic bend of the lines of flow of the surface water on the line from Living-
ston Island extend almost as far across the strait as Astrolabe Island, whereas in
November of the same year the south-easterly movement did not reach so far. In
February 1929 the north-easterly current was concentrated about 10 miles from King
George Island over a very narrow path, in contrast with the widespread nature of this
current in November 1929. At depths below 200 m. the water movements in February
1929 differed from those in November of the same year in that the predominant move-
ment appeared from the east in the former month. The results for December 1930
showed in particular the great influence of the Weddell Sea at the north-eastern end of
the strait.

5. The temperature of the Antarctic bottom water inside the strait is much lower
and the salinity slightly less than outside the strait. The Bransfield Strait Antarctic
bottom water is formed in situ as a result of freezing processes in winter on the wide
continental shelf of western Graham Land and Trinity Peninsula. This bottom water
has a very high oxygen content which indicates its recent renewal.

6. No attempt at discussion of seasonal variation has been made owing to lack of
suitable data, but the differences in temperature and salinity along a line of stations at
the north-eastern end of the strait in February and November 1929 are discussed.

7. A note is added on variable movements of pack-ice in the strait and on currents
between the South Shetland Islands.

In conclusion the author would like to express to Professor B. Helland-Hansen,
Director of the Geophysical Institute, Bergen, Norway, his very great appreciation of
the kindly help and advice given him throughout the preparation of this report and for
valuable criticism of the manuscript.

64 DISCOVERY REPORTS

LIST OF LITERATURE

Antarctic Pilot, 1930. London.

Brennecke, W., 1921. Die ozeanographischen Arbeiten der deutschen Antarktischen Expedition 1911-1912.
Archiv der deutschen Seewarte, xxxix, Nr. 1, p. 38, Hamburg.

Discovery Reports, 1929. Station List, I, pp. 1-140, pis. i-vi, Cambridge.

Discovery Reports, 1930. Station List, III, pp. 1-132, pis. i-x, Cambridge.

Discovery Reports, 1932. Station List, iv, pp. 1-232, pis. i-v, Cambridge.

Herdman, H. F. P., 1932. Report on Soundings taken during the Discovery Investigations 1926-1932. Dis-
covery Reports, vi, pp. 205-36, pis. xlv-xlvii, charts 1-7 (separately).

Holtedahl, Olaf, 1929. On the Geology and Physiography of some Antarctic and sub-Antarctic Islands.
Scientific Results of the Norwegian Antarctic Expeditions 1927-8 and 1928-9, No. 3, p. 95, Oslo.

Jacobsen, J. P., and Jensen, A. J. C, 1926. Examinations of Hydrographical Measurements from the Research
Vessels 'Explorer' and 'Dana' during the summer of 1924. Rapp. Cons. Explor. Mer, xxxix, p. 57.

Lecointe, G., 1905. Travaux hydrographiques et Instructions nautiques. Expedition Antarctique Beige.
Resultats du Voyage du S.Y. 'Belgica' en 1897-9. Anvers.

Nordenskjold, O., 1917. Die oseanographischen Ergebnisse der schzvedischen Siidpolarexpedition 1901-3.
Wiss. Ergebn. schwed. Siidpolarexp., Bd. 1, Lief. 2, p. 9.

Rouch, J., 1913. Oceanographie Physique. Deuxieme Expedition Antarctique Francaise 1908-10, p. 28.

Spiess, F., 1928. Die ' Meteor '-Fahrt. Forschungen und Ergebnisse der deutschen Atlantischen Expedi-
tion 1925-7, p. 190, Berlin.

Sverdrup, H. U., 1929. The Waters on the North-Siberian Shelf. The Norwegian North Polar Expedition
with the 'Maud' 1918-25. Scientific Results, IV, No. 2, pp. 65 et seq.

Wiist, G., 1926. Zweiter Bericht uber die oseanographischen Untersuchungen der deutschen Atlantischen
Expedition. Zeit. Ges. Erdkunde Berlin, Nr. 5/6, p. 243.

[Discovery Reports. Vol. IX, pp. 65-160, April, 1934]

DISTRIBUTION OF THE MACROPLANKTON
IN THE ATLANTIC SECTOR OF THE

ANTARCTIC

By
N. A. MACKINTOSH, D.Sc.

CONTENTS

Introduction page 67

The physical environment of the Antarctic plankton 68

Methods 70

Organisms identified 72

Cruises in the period 1927-31 77

The Antarctic convergence 83

Diurnal variations 87

Relative abundance and distribution of individual species 97

Distribution of rich and poor plankton 109

General distribution 109

Variations in abundance 1 1 1

South Georgia 1 1 1

Drake Passage and Scotia Sea 116

Distribution of warm- and cold-water species 121

South Georgia 121

The South Sandwich and Weddell Sea region 124

The Orkney-Shetland region 132

The Bellingshausen Sea 133

The Northern Antarctic region 134

Plankton communities 137

The Weddell Sea 138

The Bellingshausen Sea 141

The Orkney-Shetland region 144

The northern zone 146

Faunistic divisions T50

The macroplankton and the distribution of whales 153

Summary 157

List of literature 158

DISTRIBUTION OF THE MACROPLANKTON
IN THE ATLANTIC SECTOR OF THE

ANTARCTIC

By N. A. Mackintosh, d.Sc
(Text-figs. 1-48)

INTRODUCTION

IN the course of the oceanographical work which the Discovery Committee is con-
ducting in the south, tow-nets of four different kinds have been used systematically
for the study of the plankton. These are the 50 cm., 70 cm. and 1 m. nets, and the
young-fish trawl. Of the four the 1 m. net, towed obliquely (N 100 B), is the most
suitable for investigating the horizontal distribution of the larger plankton organisms.
The following report is concerned with the samples taken by this net in the Antarctic
surface waters, and its purpose is to find how the Antarctic environment of the whales
is reflected in the distribution and variations of the macroplankton.

During the first commission of the R.R.S. ' Discovery ' in 1925-7 the 1 m. nets were
towed horizontally at routine stations, but it was found that a net towed obliquely from
about 100 m. to the surface gives a better representation of the plankton, and the latter
method was therefore adopted during the work of the 'William Scoresby' from 1927
to 193 1 and of the 'Discovery I Tin 1930 and 1931. The catches of the early horizontal
nets, many of which have already been examined by Hardy and Gunther (1934), are
not strictly comparable with those of the oblique nets ; we are thus concerned here
with the period 1927-31, and the stations taken into consideration are those in the true
Antarctic water, together with a few on the north side of the Antarctic convergence.
Of the species taken at these stations only one or two are confined to coastal regions,
and we are in fact dealing with the plankton population of the open ocean.

During the period 1927-31 the majority of samples were collected in the neighbour-
hood of the Falkland Islands Dependencies, but lines of stations also extended from
Bouvet Island in the east to a point west of Peter 1st Island in the Bellingshausen Sea.
The positions of most of these stations are shown in Fig. 1 .

The present report is based on the examination of about 600 samples from the 1 m.
nets. From such a large number it is possible to obtain a great deal of information about
the plankton, and the material is by no means exhausted by the results put forward in
the following pages. It is the purpose of this report, however, not so much to elucidate
the individual distribution of the various species as to examine the distribution of the
macroplankton as a whole, and to distinguish individual communities whose constitution
is dependent on the hydrological and geographical features of their environment.

68

DISCOVERY REPORTS

THE PHYSICAL ENVIRONMENT OF THE
ANTARCTIC PLANKTON
A summary account of the hydrology of the Atlantic sector of the Antarctic and the
Bellingshausen Sea has been published by Mr G. E. R. Deacon (1933), and I am

Fig. 1 . The Atlantic sector of the Antarctic, showing stations at which N 100 B samples were taken. Stations
included in intensive work around South Georgia and in the Bransfield Strait, and a line of close stations off
Adelaide Island, are omitted. Surface currents are indicated by arrows.

indebted to him and to Mr A. J. Clowes for much valuable information. There are
four primary divisions of the oceans of the southern hemisphere : the Antarctic, sub-
Antarctic, sub-Tropical and Tropical Zones. The Antarctic Zone is characterized by
a cold, poorly saline surface layer of an average depth of about 200 m. This surface
water flows away from the melting pack-ice to which its low temperature and low salinity

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 69

are due, and gradually works northwards (with a strong easterly component) until it
meets the more saline, but warmer and lighter water of the sub-Antarctic Zone, beneath
which it sinks. The line along which this sinking takes place is the northern boundary
of the Antarctic Zone and is known as the "Antarctic convergence".

Beneath the cold surface water is a much thicker layer of warmer water whose origin
is in the North Atlantic or Pacific. There is a southerly component in the direction in
which this water drifts, and when it reaches high latitudes it becomes cooled. Part of it
rises to the surface to replenish the cold superficial layer, and the rest, which has been
cooled but not diluted, sinks to form the Antarctic bottom water. This layer flows in a
northerly direction. There is in fact a movement of cold water away from the pole at the
surface and bottom, and towards the pole at intermediate depths. The warm inter-
mediate layer is very rich in nutritive salts, and it is the upwelling of these salts into the
cold surface layer which enables the richest plankton in the world to flourish in the
Antarctic surface water.

In the lower latitudes of the Antarctic Zone the surface waters are drifting in an
easterly direction, and Drake Passage may be described as a constriction in this drift.
Graham Land and the associated islands and archipelagos, which form the southern
promontory of this constriction, separate the Weddell Sea to the east from the Bellings-
hausen Sea to the west. North of the Weddell Sea lies the Scotia Sea, bounded to the
west by Drake Passage and on all other sides by the Scotia Arc. The Scotia Arc is a
loop of elevated land projecting eastwards into the Atlantic. It comprises — as existing
land masses — Tierra del Fuego, Staten Island, the Shag Rocks, South Georgia, and the
South Sandwich, South Orkney and South Shetland Islands. These, as Herdman
(1932) has conclusively demonstrated, are connected by well-defined submarine ridges.
The Burdwood Bank, to the east of Staten Island, forms part of the arc. Water passing
eastwards through Drake Passage is deflected a little to the north-east and flows
mainly through a gap in the Scotia Arc between the Burdwood Bank and the Shag Rocks.
The Antarctic convergence passes through the middle of Drake Passage, and the water
to the south of it is derived mainly from the Bellingshausen Sea.

In the Weddell Sea there is a clockwise circulation. This great bay in the Antarctic
Continent receives water drifting westwards in the higher latitudes, and the flow is
directed northwards and north-eastwards by the Graham Land Archipelago, and flows
out as a very cold current towards the South Sandwich Islands and South Georgia.
It joins the drift from the Bellingshausen Sea along a line running from the South
Shetlands to South Georgia. The drift from the Weddell Sea carries with it much pack-
ice and innumerable icebergs, so that in longitude 20 or 300 W the pack frequently
extends far to the north of the South Sandwich Islands, while around 60 or 700 W it is
never found very much north of the South Shetland Islands.

We are here dealing only with the larger plankton organisms and their horizontal
distribution in the cold surface layer. These are the true Antarctic species. At greater
depths we find species which are common to the deep waters of the Antarctic and the
tropics, but with these we are not concerned.

7o DISCOVERY REPORTS

METHODS

The construction and method of working the i m. oblique net have been fully de-
scribed by Kemp and Hardy (1929). The essential points are as follows. It is a conical
net with an opening 1 m. in diameter and the principal fishing part of the net is of
stramin. The filtration of the stramin is roughly equivalent to that of silk bolting cloth
having 15 meshes to the inch. " For oblique hauls open N 100 and N 70 were attached
to the warp close together and 3 or 4 m. above the lead. With the ship steaming at
2 knots 200 m. was paid away, and as soon as this had been done hauling commenced.
The rate of hauling was 10 m. per minute; each haul thus took 20 min. and the
distance covered was two-thirds of a mile" (Kemp and Hardy, he. cit.). By this method
the net is fished obliquely from a depth of about 100 m. to the surface. The samples
are preserved in formalin and stored in screw-top bottles.

The method of analysis of a plankton sample must be adapted to the type of net used.
The vertical N 70, for example, is fished at a very uniform speed through an accurately
measured column of water, the organisms caught are small and a comparatively refined
technique can be used for quantitative estimations. The depth and speed at which the
oblique N 100 is fished, however, are greatly affected by the difficulty of adjusting the
ship's speed to varying weather conditions, so that precise quantitative comparisons
between different samples are of questionable value, and a sufficiently accurate estimate
may be obtained by less refined methods.

There is much variation in the richness of the Antarctic plankton, and a single haul
may yield anything from a dozen to 200,000 organisms. Different methods must there-
fore be used for different catches. In small samples, containing up to 200 or 300
organisms, the total number of each species is counted without difficulty. Samples of
average size, however, are much bigger than this, and a typical one might contain about
5000 organisms of which perhaps 4000 would be copepods. These have generally been
treated as follows. The bulk of the formalin is strained off and the sample is washed
into a glass dish with plenty of water. The large animals such as Amphipoda, large
Siphonophora, Polychaeta and Salps can generally be quickly picked out. It is then
necessary to go very carefully through the whole sample to find any of the species
which are both small and present only in small numbers. It is very easy to miss such
organisms as Limacina or the small siphonophore, Dimophyes arctica, whose presence
or absence may be of some importance, and it is often advisable to turn over the whole
sample several times before dealing with the more numerous organisms. The medium-
and large-sized organisms such as Euphausians can all be picked out and counted if
there are not more than (say) 60 or 70 of them. For the more numerous organisms sub-
samples are taken, and it may be necessary to take first a quarter or an eighth and count
perhaps the Euphausians and Chaetognatha in that, and then to take anything from
1/16 to 1/256 to count the various species of Copepoda. For the Copepoda it has
generally been customary to take a fraction which will contain 100 to 200 specimens.
Even then there is a possibility of missing one or two of the rarer species. For taking

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 71

sub-samples a rough and ready method has been found to be best. The catch is put in
a large Petri dish and the water strained off until the sample has the consistency of a
loose paste. It is well mixed up and the dish is shaken until the sample lies in a flat,
even layer. It is then divided into quarters (or some other fraction), and a quarter is if
necessary removed to another dish and further divided. The method is crude, but if
carefully done can give surprisingly accurate results. An Einar Lea apparatus can be
used, but with large organisms it is not more accurate than the method described
above.

In the analysis of these samples no hard and fast line of procedure can be followed.
The method must be suited to each sample, and the worker must endeavour to satisfy
himself that he has not missed any of the rarer forms, and that he has estimated the
approximate numbers of each species present. The analysis of a large sample of perhaps
20,000-30,000 organisms may occupy several hours, especially if there is a great variety
of species. The smallest samples can be analysed in about 20 minutes.

The object of plankton work is to determine the nature of the plankton in a given
area from the analysis of samples taken at selected points in that area. The inferences
thus drawn are liable to certain errors which may arise from the distribution of the
plankton itself in the water, from the method of collecting the samples, or from the
method of analysing the samples. The subject has been dealt with by Hardy and
Gunther (1934), whose remarks apply in a large degree to plankton collected by the
N 100 B, and need be discussed only briefly here. The following are some points to be
borne in mind: (i) Heterogeneous distribution of the plankton organisms may give a
false impression of the fauna of an area. Thus allowances must be made for species
which tend to a specially patchy distribution, (ii) Active avoidance of the net must
sometimes take place, especially by such species as Euphausia superba. I have been able
from the ship's side to watch a net being towed a few feet below the surface and passing
through a shoal of this species. The Euphausians could clearly be seen to leap backwards
out of the way of the approaching net. (iii) Variations in the speed of towing and depth
of the net have already been mentioned, (iv) Where a number of species of many
groups have to be quickly identified, complete accuracy is not always to be depended
on. (v) Big catches of shoaling species such as E. superba and Salpa may swamp the
sample and make an estimation of the other species almost impossible. Such catches
must sometimes be disregarded except for the Euphausians or Salps themselves.

It may seem that some of these factors must lead to serious errors in the analyses,
but Hardy and Gunther (1934) have shown that fluctuations in the numbers of the
smaller plankton organisms are so great that an error of 50 per cent has little sig-
nificance, and Hart (1934) points out, in connection with the analysis of samples of
phytoplankton, that "in practice differences of 100 per cent and over are the smallest
that can be regarded as of much significance". The same applies, possibly with even
greater force, to the catches of the N 100 B, for although it may be too much to say that
the important features of a sample are those which may be recognized at a glance, it is
at least probable that the most obvious features will be the important ones, and that a

72

DISCOVERY REPORTS

significant difference in the numbers of a species in two samples will be great enough
to swamp any minor inaccuracies which might arise from the methods of working the
net and analysing the catch. At the same time conclusions must be drawn with caution,
and the character of the plankton in a locality must not be judged from a single haul.

ORGANISMS IDENTIFIED

Samples taken by the N ioo B in the Antarctic contain a relatively small number of
species, the majority of which are easily recognized. It is therefore possible for a single
worker to familiarize himself with the common species and work straight through the
samples, avoiding the delay which would be involved if all the catches were sorted and
the groups distributed to a number of experts for separate identification. I have, how-
ever, received much assistance from Mr F. C. Fraser, formerly of the Discovery staff
and now on the staff of the British Museum, and from Dr H. E. Bargmann, Mrs
M, E. White and Mr A. H. Laurie, of the Discovery staff, each of whom has undertaken
the analysis of a number of the samples. The Copepoda from some of the samples were
identified by the late Dr Andrew Scott, who has also furnished a valuable reference set
of named specimens. Other named specimens have been provided, of Amphipoda by
Dr Barnard of the South African Museum, of Siphonophora by Capt. Totton of the
British Museum, of Polychaeta by Mr Monro of the British Museum, and of Pteropoda
by the late Miss Massy. Capt. Totton and Mr Monro have also kindly given me personal
assistance from time to time. Under the guidance of Dr Kemp I have been able to
identify the various species of Enphausia.

Even with such assistance it has not been possible throughout to identify by any
means all the species which occur, but I have attempted to estimate the number of the
following in every sample.

Siphonophora.
Diphyes antarctica, Moser.
Dimophyes arctica, Chun.
Pyrostephos vanhoffeni, Moser.

Medusae.

Sibogita borchgrevinki, E. T. Browne.
Solmnndella sp.

Polychaeta.

Tomopteris sp. (large).
Tomopteris sp. (small).
Vanadis antarctica (Mcintosh).

ECHINODERMATA.

Auricularia antarctica, MacBride.

Copepoda.

Calanus acutus, Giesbrecht.
C. propinquus, Brady.

Copepoda (cont.).

C. simillimus, Giesbrecht.
Rhincalanus gigas, Brady.
Pleuromamma robusta (F. Dahl).
Metridia gerlachei, Giesbrecht.
Haloptihts occllatus, Wolfenden.
Haloptihts sp.
Parenchaeta sp.
Heterorhabdus sp.
Encalanus sp.
Euchirella sp.
Candacia sp.

Amphipoda.

Parathemisto gaudichaudi (Guerin).
Primno macropa, Guerin.
Vibilia antarctica, Stebbing.
Eusirus antarcticus (Thomson).
Cyllopus spp.

Fig. 2. Species of the Antarctic macroplankton.

a, Diphyes antarctica, anterior nectophore, x 2.

b, Dhnophyes arctica, anterior nectophore (after

Moser), x 7.

c, Sibogita borchgrevinki (after E. T. Browne), x ij.

d, Pyrostephos vanhoffeni , nectophore, x 5.

e, Solmandella sp., x 2.

/, Tomopteris carpenteri, x 1 .
g, Auricularia antarctica, x 8.

//, Vanadis antarctica, anterior portion, x 1.
i, Cleodora sulcata, x 3.
j, Limacina helicina, x 5.
k, Limacina balea, x 15.
/, Spongiobranchaea australis, x 4.
m, Clione antarctica, x 4.
n, Salpa fusiformis f. aspera (after Ihle),
x ii.

\ bi

Fig. 3. Species of the Antarctic macroplankton.

a, Calanus acutus, x 8.

b, Calanus propinquus, x 8.

bi, C. propinquus, basal joint of th. 5.

c, Calanus simillimus $, x 9.

ci, C. simillimus, basal joint of th. 5.

d, Rhincalanus gigas, x 6.

e, Pareuchaeta antarctica, x 6.

/, Pleuromamma robusta (after Sars), x 10.

g, Metridia gerlachei, x 18.

h, Haloptilus ocellatus (after Wolfenden), x 6.

i, Heterorhabdus sp., x 8.

j, Eucalanus sp., x 8.

k, Euchirella sp., x 8.

/, Candacia sp., x 8.

Fig. 4. Species of the Antarctic macroplankton.

a, Parathemisto gaudichaudi, x 4.

b, Cyllopus sp., x 3.

c, Eusirus antarcticus, x 4.

d, Vibilia antarctica, x 4.

e, Primno macropa (after Stebbing), x 4.
/, Antarctomysis maxima, x 2.

g, Euphausia superba, x ij.
"i, Rostrum of E. frigida.

gz, Rostrum of E. vallentini.

£3, Rostrum of E. triacantha and E. crystallorophias.

£4, Abdominal segments 3-5 of E. frigida and E.

crystallorophias.
£5, Abdominal segments 3-5 of E. vallentini.
g6, Abdominal segments 3-5 of E. triacantha.
h, Thysanoessa macrura, x 3.

76 DISCOVERY REPORTS

Mysidacea. Mollusca.

Antarctomysis sp. Cleodora sulcata (Pfeffer).

Limacina helicina (Phipps).

Euphausiacea. L. balca, Moller.

Euphausia superba, Dana. Spongiobranchaea australis, d'Orbigny.

E.frigida, Hansen. Clione antarctica, E. A. Smith.

E. crystallorohhias, Holt and Tattersall. T

„ . , „ . , m ,, LUNICATA.

E. triacantha, Holt and Tattersall

E. vallentini, Stebbing

Thysanoessa spp. Chaetognatha

Salpa fusiformis f. aspera (Chamisco).

Each of the above species and genera, and the group Chaetognatha will be treated in
this paper as a separate unit. The specimens of Solmiindella are likely all to be S.
mediterranea. Specimens of Haloptiliis, apart from H. ocellatus, may include H. oxy-
cephalic and possibly others. Pareuchaeta sp. is generally P. antarctica, but may include
P. biloba. Heterorhabdus sp. probably includes only H. austrinus. Some specimens of
Eucalanus have been identified as E. acus, and some of Euchirella as E. rostromagna.
Candacia sp. may include more than one species. The identity of Eusirus antarcticus
has not been determined with absolute certainty. Cyllopus spp. includes C. lucasi and
C. magellanicus, and Thysanoessa spp. includes T. macrura ; Antarctomysis is probably
A. maxima. There are two common Antarctic species of Tomopteris, T. carpenteri and
T. septentrionalis. When adult the former can always be distinguished from the latter
by its greater size, but the identification is otherwise difficult. Those distinguished in
this paper as "large Tomopteris" are nearly always T. carpenteri, and those included
under "small Tomopteris" are generally T. septentrionalis, but may include an unknown
proportion of immature T. carpenteri. The vast majority of the Chaetognatha are
Eukrohnia hamata, but Sagitta gazellae, S. maxima, and S. planctonis occur in small
numbers. Naturally those units identified only as genera or as a group have little value
compared with those identified as species, but nearly all of them occur only in small
numbers. Only Thysanoessa and the Chaetognatha appear in large numbers, but im-
mature specimens, especially of the former, make up so high a percentage of these two
units that complete specific differentiation would have been very difficult, and would
have greatly prolonged the period occupied in the analysis of the samples.

For description of the species shown in the above list the following authorities may
be consulted.

Moser (1925): Diphyes antarctica, Pyrostephos vanhoffeni.

Chun (1897): Dimophyes arc tic a.

Browne (1910): Sibogita borchgrevinki.

Benham (1921): Vanadis antarctica.

MacBride (1912, 1920): Auricularia antarctica.

Wolfenden (1908): Calanus acutus, C. propinquus, C. simillimus, Metridia gerlachei.

Schmaus and Lehnhofer (1927): Rhincalanus gigas.

Sars (1903): Pleuromamma robust a.

Wolfenden (1911): Haloptiliis ocellatus.

Barnard (1932): Parathemisto gaudichaudi.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 77

Stebbing (1888): Primno macropa, Vibilia antarctica.

Stebbing (1906): Eusirus antarcticus.

Tattersall (1908): Euphausia superba, E. crystallorophias, E. triacantha, E. vallentini.

Hansen (1913): Euphausia frigida.

Massy (1920): Cleodora sulcata.

Massy (1932): Limacina helicina, Spongiobranchaea australis.

Bonnevie (191 3): Limacina balea.

Eliot (1907): Clione antarctica.

Ihle (1912): Salpa fusiform is f. aspera.

Other species which occur from time to time, but which are mostly uncommon and
are disregarded here, are certain Medusae, Ctenophora, Polychaeta, Ostracoda and one
or two small Amphipoda.

This method of taking only certain units into consideration might be criticized as
arbitrary ; but the use of any net with a particular aperture and mesh is also arbitrary,
and the best we can do is to study certain organisms, and find from them what we can
of the general behaviour of the macroplankton.

CRUISES IN THE PERIOD 1927-31

The following notes are not perhaps essential to the purposes of this paper, but since
the catches taken in the various cruises will not be dealt with in strict chronological
order, I have felt that a brief statement of the order in which the stations were taken
might be useful for occasional reference.

During the first commission of the 'Discovery' and the associated work of the
' William Scoresby ', the 1 m. nets were towed in horizontal flights of three. On the
return of the 'William Scoresby ' to South Georgia in February 1928, and in all subse-
quent work, the oblique 1 m. net (N 100 B) was used at all routine stations. The greater
part of the plankton work in the Antarctic has been done during the summer months
from October to April, and it will therefore avoid confusion if we speak of summer
seasons (which are equivalent to the whaling seasons) rather than of years, and say that
the regular work with the N 100 B began in February 1927-8. Full details of all the
stations carried out, together with charts, are published in the Station Lists.1

Season 1927-8 (Figs. 5 and 10). The work of the ' William Scoresby' began with a line of stations
from the Falkland Islands to South Georgia (WS 1 39-43*) and was followed by a survey of the South
Georgia whaling grounds (WS 144-93). A line of stations was then worked from South Georgia to
the vicinity of the South Shetlands and another line northwards to the Burdwood Bank and the
Falkland Islands (WS 196-205).

Season 1928-9 (Figs. 6, 7, 1 1 and 15). The ship was next engaged in a trawling programme, but
plankton work was resumed in August 1928. After a line of stations from the Falkland Islands to
South Georgia (WS 253-6), the South Georgia survey was repeated, with the exception of the two
southern lines, in incessantly adverse weather conditions, during the end of August and the greater

1 Discovery Reports, HI, pp. 1-132, and IV, pp. 1-230.

2 Where a line of stations crosses the Antarctic convergence, the station numbers given here in brackets
include only those taken in Antarctic water together with the first station on the north side of the convergence
if it is still quite close to the latter. Stations at which the N 100 B was not used are omitted in the
charts shown in Figs. 5-17.

•WSI9I
•V«5I90
■WSI67

•WS'66 .WB3

•WSI65 I

•W5I64
1&S63

•WSI6I

•W5I60

•W5I59

•WSIS8

•V^L57 .^^g

•WSI56 .wsi6Z -W5I48

■W5IS5 I .WSI47

•WSI54 — ;W5H6.

■WSI52
•WSI5I
WSI50

■WSI44

•WS273
•W5274

•WS275
•W5276
■WS277

■WS265

-WS266
I
--WS267-

•WS266
•W5263
■W5270
•W5Z7I
•W5272

■W5264
•W5263
W5262
■WS26I
-WS260 j
WS259
W52S8 |

\hT///t///,

Fig. 5. South Georgia survey,
1927-8.

February-March

Fig. 6. South Georgia survey, August-October
1928-9.

■ ., „ . „■ . 31

.

-

■WS347
•W534G
■WS345
•WS344

1

•WS3.

17 L

-

■w

S326
W53Z5
•W5324
•W5322
•W52

•WS34:i
1

•WS342
•WS341

•WS327-WS340

5321 'WS338
L~W-- I WS37'"

■v\
•W5

•ws?.:-

•W533I

•WS330
/VS3Z9

■WS336
•WS335

5334
»3

-

-

s

*" -WS367 L '.
■WS368 . ^

28

^•-wsm

h

•W535

iff*

J73

7

j * * I

•

•WS370

WS4IB
WS419

•WS

•

358-
WS359

•WS360
•WS36I
•WS3t
•W

2 .

J363 I
W5364

:

51- ".- W

r

1

•300

1

•301

•302

1

.

•320
• 32J

•304

•305

057
.35a

1

.■32;

'323
3I9-' -316

■306
•307

•313
■312

-

.317 -315 .311
• 316 .3|; '310

3

4 -325

■3
■330
£331

■-'■-_

•323 ~

29 *

'y-„ >M

MO ' .
■341 L

-339 ^

. . *-?.-,„ -'355 '"356

>■ "048

■337
:336

•34

7

346
•34S

.335 -'344
■334 -343

Fig. 7. South Georgia survey, December-January
1928-9.

Fig. 8. South Georgia survey, January-February
1929-30.

Fig. 9. South Georgia survey, November 1930-31.

50"

WSI39

WSI40

WSI4

WS200

50°

4.0

Fig. 10. Stations in the Scotia Sea, 1927-8.

Fig. 11. Stations in the Scotia Sea, etc., 1928-9.

50

40

WS530

WS468 WS529

WS520

WS466
WS52I ' ' WS522

WS473

~WS523 WS465

WS525
• I WS464

WS526

WS528

50°

Fig. 12. Stations in Scotia Sea, etc., 1929-30.

Fig. 13. Stations in Bellingshausen Sea, 1929-30.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 81

part of September (WS 257-96). A series of stations was then worked along the ice-edge to the
south-east of the island (WS 287 and 288, and 297-310), and a second line was laid from South
Georgia to the Falkland Islands (WS 314-17). After the return of the ship to South Georgia a com-
plete survey was again carried out (WS 321-72 in December-January), and she proceeded to the
South Shetland Islands (WS 374-81), where intensive work was done in the Bransfield Strait in
February (WS 382-99). A line of stations was then taken across the Drake Passage to Cape Horn
(WS 400-5), and then to South Georgia via the Falkland Islands (WS 413-16). Here two more
lines of stations were taken in April on the south side of the island (WS 417-26) and a fourth line
between South Georgia and the Falkland Islands (WS 427-31). The season closed with the ship's
departure from South Georgia to Capetown, when two more stations were taken in Antarctic water
(WS 434 and 435).

Season 1929-30 (Figs. 8, 12, 13 and 16). The 'William Scoresby' returned to South Georgia in
October and proceeded thence to the Falkland Islands (WS 464-6). From November to February
she visited the Bellingshausen Sea, and, though primarily engaged in other work, was able to con-
tinue some plankton investigations. These included a line of stations from the Falkland Islands to the
South Shetlands (WS 468-74), intensive work in the Bransfield Strait (WS 476-93), various stations
off the coast of Adelaide Island and the Biscoe Islands (WS 496-501), five stations along the ice-
edge in the Bellingshausen Sea as far as 1000 W (WS 502-8), and a line of stations at short intervals
running north-westward from Adelaide Island (WS 509-17). The ship then returned in February
to the Falkland Islands. In the meantime the 'Discovery II' reached South Georgia in January,
and proceeded with the usual survey of the whaling grounds (Sts. 300-59). This was followed by a
visit to the South Sandwich Islands where six plankton stations were taken (Sts. 360-9), and some
time was spent in topographical surveying. In March a line was begun from the Sandwich group
to the Burdwood Bank (Sts. 372-5), which was later completed by the 'William Scoresby' (WS 527-
30) after the latter ship had worked more stations between the Falkland Islands and South Georgia
(WS 520-6). In April the ' Discovery II ' visited the South Shetlands, took a station in the Bransfield
Strait (St. 376), a line across the Drake Passage (Sts. 378-85), and two more stations on the way back
to South Georgia (Sts. 390 and 392). The 'William Scoresby' now left for Europe and the 'Dis-
covery II ' for Capetown. At the beginning of the voyage an attempt was made with the latter ship
to carry out continuous observations throughout a 24-hour period near South Georgia. Flights of
six closing N 100 B were to be taken every four hours (St. 393), but the series was interrupted by
bad weather before it could be completed. One more station (St. 394) was taken in Antarctic water
on the way to the Cape.

Season 1930-1 (Figs. 9, 14 and 17). The ' Discovery II ' sailed from Capetown in October, reached
the ice-edge south of Bouvet Island (Sts. 452-60) and followed it westwards to South Georgia
(Sts. 462-72). A 24-hour station with closing N 100 B (St. 461) was worked off the ice. The South
Georgia survey was repeated in November (Sts. 475-525), and a cruise to higher latitudes begun.
The ship followed the ice-edge from a point near the Sandwich group to the Bransfield Strait
(Sts. 528-41), taking fourteen stations in the Strait (Sts. 542-55), and continuing the voyage from
the South Shetlands to Adelaide Island (Sts. 556-60), and westwards along the ice-edge of the
Bellingshausen Sea to about ioo° W, 700 S, and back to Adelaide Island (Sts. 561-82). A line of
stations was worked north-westwards from Adelaide Island (Sts. 583-92) and a number of other
stations off Adelaide Island, the Biscoe Islands, the South Shetlands, and the South Orkneys
(Sts. 593-617, January-February)— a period largely occupied also with survey work. The voyage
was continued to the South Sandwich Islands (Sts. 618-29), westward towards the Falkland Islands
(St. 631), south again to the South Orkneys (Sts. 633-6) and Bransfield Strait (Sts. 637-44), north
towards Staten Island (Sts. 646-9), and back to South Georgia at the end of March (Sts. 655-9).
In the meantime the 'William Scoresby' returned to South Georgia in January and made a cruise
south-eastwards past the South Sandwich Islands, and southwards to the ice-edge in the eastern part
of the Weddell Sea, making the return journey on much the same route (WS 534-65, January-

C/3

P

J3

CO

BO
C

c

u

-d

o
o

en

BO

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

83

February). In March she repeated a line of the November survey off South Georgia (WS 567 75)
and the season's work was over.

By the time the present paper was ready for publication the ' Discovery II ' had re-
turned from a second commission, and a large collection of new N 100 B samples was

Fig. 15. Stations in Bransfield Strait, 1928-9.

Fig. 16. Stations in Bransfield Strait, 1929-30.

Fig. 17. Stations in Bransfield Strait, 1930-1.

available, but it was thought desirable to publish some results of the four previous
seasons' work to avoid the delay which would be entailed if the new material was
incorporated with it.

THE ANTARCTIC CONVERGENCE
The Antarctic convergence, which divides the Antarctic from the sub-Antarctic
surface waters, is an important faunistic boundary. In the period 1927-3 1 a number of
lines of stations crossed this convergence, and it is necessary first to determine which
of these stations lie in Antarctic and which in sub-Antarctic water. The Antarctic
convergence can be located by a sudden change in surface temperature and is usually

3-2

§4

DISCOVERY REPORTS

found at the point at which the coldest part of the Antarctic water sinks below the level
of 200 m. (see Deacon, 1933, p. 192). Applying these criteria to the data given in the
Station Lists (Discovery Reports, vols. 111 and iv) the following can be quoted as the
stations nearest the convergence in each line of observations made across it. The
position of the convergence is shown by an arrow: sub-Antarctic stations are above
the arrows and Antarctic stations below them.

WS 139 Cold layer at 750 m.
WS 140 Cold layer at 200 m.

1927-28

WS 205 Cold layer indistinct.
WS 204 Cold layer at 200 m.

1928-29

WS 253 Cold layer at 750 m.

WS 254 Cold layer at 100 m.

WS317 Cold layer indistinct.

WS 316 Cold layer at 80 m.

WS 405 Cold layer indistinct.

WS 404 Cold layer at 150 m.

WS 412 No temperature records.

>WS 413 No temperature records.

WS 414 No temperature records.

WS 431 Faint minimum at 600 m.

>WS 430 Faint minimum at 400 m.

WS 429 Cold layer at 200 m.

WS 437 Cold layer indistinct.

►WS 436 No temperature records.

WS 435 Cold layer at 150 m.

WS 467 No temperature records.

■»WS 466 No temperature records.

WS 465 No temperature records.

WS 468 Cold layer indistinct.

->WS 469 Cold layer at 200-300 m.

WS 470 Cold layer at 150 m.

St. 385 Cold layer indistinct.

St. 384 Cold layer at 200 m.

St. 389 Surface temperature 4-35° C.

■»St. 390 Surface temperature 4-85° C.

St. 392 Surface temperature 3-90° C.

1929-30

WS ^20 Faint minimum at 600 m.

WS 521 Cold layer at 200 m.

WS 530 Cold layer indistinct.

WS 529 Cold layer at 150 m.

St. 396 Surface temperature 7-45° C.
St. 394 Surface temperature 4-15° C.

1930-

St.

451

St.

452

•St.

633

St.

632

-St.

631

St.

630

Cold layer indistinct.
Cold layer at 100-150 m.

Surface temperature 5-14° C.
Surface temperature 4*85° C.
Surface temperature 5-80° C.
Surface temperature 3-40° C.

St. 650 Cold layer indistinct.

St. 649 Faint minimum at 400 m.

St. 648 Cold layer at 80 m.

St. 655 Faint minimum at 800 m.

St. 656 Cold layer at 300 m.

St. 657 Cold layer at 150 m.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 85

Where a pair of stations is shown in the above list the convergence lies between them.
Thus at WS 254 the cold layer lies at 100 m., while at WS 253 it lies at 750 m. : WS 253
is therefore in sub-Antarctic and WS 254 in Antarctic water. At WS 404 the cold layer
is at 150 m., while at WS 405 it has become obscured through sinking and mixing with
deeper water. This indicates that WS 405 is well on the north side of the convergence.
At such stations the cold layer can be detected only as an irregularity in a curve showing
the rate of decrease of temperature as the depth increases. WS 469 lies just about on the
convergence itself. At some stations no temperatures, or only surface temperatures, are
given. Thus WS 413, 466 and St. 390 probably lie very near the convergence. Sts. 630-3
lie near an eddy, which is roughly indicated by the bend of the convergence in Fig. 1.
From its surface temperature it is evident that St. 630 is in Antarctic water. Sts. 631
and 633 are probably very close to the convergence, and St. 632 in Antarctic water.
Sts. 391 and 395,* at which the N 100 B was not fished, are omitted from the list.

A full account of the differences and resemblances which exist between the plankton
of the surface waters of the Antarctic and sub-Antarctic Zones would be a large subject
and will probably be dealt with in subsequent publications. I will give here only a brief
indication of the effect of the convergence as a faunistic barrier, and for this purpose
have examined the N 100 B analyses for twenty stations lying between 100 and 200
miles north of the convergence. The Copepoda have been identified in only eight of
these analyses, but they will serve for a purely qualitative comparison.

Antarctic species occurring north of the convergence can be divided into the following :

Normal inhabitants of sub-Antarctic water

Calanus simillimus. Occurs at five out of eight stations, sometimes in moderate
numbers.

Rhincalanus gigas. Occurs at seven out of eight stations, and is generally the most
numerous copepod.

Pleuromamma robusta. Appears to occur at four out of eight stations, but its specific
identity at these sub-Antarctic stations has not been checked with absolute certainty.

Eticalanus sp. Occurs at four out of eight stations, usually in small numbers.

Parathemisto gaiidichandi. Occurs at seventeen out of twenty stations. Barnard (1932,
pp. 6-19) records this species in surface waters at various stations in comparatively low
latitudes.

Primno macropa. Occurs at nine out of twenty stations, a high proportion for this
species. Occurrence in sub-Antarctic waters confirmed by Barnard.

Vibilia antarctica. Occurs at four out of twenty stations — a sufficiently high pro-
portion. Recorded by Barnard in two sub-Antarctic surface hauls. It is evidently less
common than Primno in these latitudes.

Etiphausia vallentini. Recorded in only eleven out of the twenty stations, but occurs
sometimes in large numbers and is actually a typical sub-Antarctic species which only
occasionally strays into the Antarctic.

86 DISCOVERY REPORTS

Cleodora sulcata. Recorded at four of the twenty stations. No doubt less common in
sub-Antarctic than in Antarctic water.

Lhnacina balea. Occurs at eleven out of the twenty stations. Massy (1932) gives its
distribution as the "temperate zones between Arctic and Antarctic and circumtropical
zone".

Spongiobranchaea australis. Occurs at five out of the twenty stations. This is a fairly
high proportion.

Dimophyes arctica. This species occurs in large numbers only in the coldest Antarctic
water. However, it is recorded at one of our twenty stations, and I am informed by
Capt. Totton that it occurs quite commonly at various depths throughout the Atlantic,
and there are instances of its occurrence in tropical and sub-tropical surface waters.
The latter record can hardly be attributed to accidental straying into warmer waters.

O illy juvenile stages apparently common in sub- Antarctic water
Calanus propinquus. At least one of the eight stations has an example of this species
and at six of the eight there are varying numbers of young copepods which appear to
be C. propinquus.

Species probably belonging only to the Antarctic water, but which occasionally

stray into sub-Antarctic water

Calanus acutus. Several occurred at one of the eight stations, but it is probably rare
everywhere north of the convergence.

Metridia gerlachei. One specimen was recorded at two of the eight stations, but it is
really typical of the colder Antarctic water.

Euphausia frigida. Two doubtful records in the twenty stations. This species is not
usually taken anywhere to the north of the convergence.

Euphausia triacantha. One example recorded in the twenty stations. Not uncommon
in stations only a short distance north of the convergence.

Limacina helicina. Recorded at two of the twenty stations, but like Metridia is found
mostly in the colder Antarctic water.

Salpa fusiformis f. aspera. Occurs at one of the twenty stations. It is said to have a
very wide distribution (Ihle, 19 12), but it seems commonest in Antarctic water.

Among the organisms of which only the genus is identified, Tomopteris, Pareuchaeta,
Candacia and Thysanoessa commonly occur in sub-Antarctic water, and Cyllopus and
Euchirella occur once each. The Antarctic species which do not occur at these stations
are Diphyes antarctica, Pyrostephos vanhoffeni, Vanadis antarctica, Auricularia ant-
arctica, Haloptilus ocellatus, Eusirus antarcticus, Euphausia superba, E. crystallorophias,
Clione antarctica, Solmundella sp., Antarctomysis sp., and Haloptilus sp.

These results may need some revision if a larger body of material is taken into con-
sideration, but they are enough to show that some of the common Antarctic species are

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 87

also normal inhabitants of sub-Antarctic surface waters, while others are sufficiently
rare in the latter zone to be regarded as intruders if they are found there.

DIURNAL VARIATIONS

During the daytime some species sink to a depth which is beyond the reach of the
net, while others do not. Therefore, in order to trace the distribution of the macro-
plankton we must have some idea of the hours between which a haul will be indicative
of the presence or absence of each species. The following section of this paper is, how-
ever, confined to the study of diurnal variations as they affect the catches in the N 100 B,
and should not be taken as an attempt to investigate the vertical migrations of the
different species. Diurnal variation is, so to speak, an adventitious phenomenon which
is caused by vertical migrations, and the proper study of the latter should depend mainly
upon hauls taken at different depths with closing nets. Hardy and Gunther (1934) have
studied in this way the vertical migrations of certain Antarctic species, and reference to
their results is made on p. 96.

The majority of the N 100B samples from Antarctic water have been collected in-
discriminately at all hours of the day and night, so that an estimation of the diurnal
variations could be worked out for each species if a comparison were made of the average
number per haul for various times of day. The accuracy of the results of this calculation
might be disturbed by three factors: (i) the irregularity of distribution of the plankton,
(ii) the possible effect of the difference in the period of darkness in different latitudes,
and (iii) the varying depths from which the net is fished. These difficulties cannot
altogether be disposed of, but the quantity of data is sufficient to swamp any serious
error arising from distribution, and we can afford to restrict the estimation almost
entirely to samples taken between 52 and 6o° S. This will include the great majority of
stations without too great a range of latitude. Errors arising from the different depths
at which the net begins its oblique passage towards the surface will also be largely dis-
counted through the abundant data. The calculation will of course be rough, but
sufficient for our immediate purpose.

It must be remembered that the diurnal variations revealed by this method are those
which result only from the more extensive vertical migrations. There may, for instance,
be certain species having a well-defined vertical migration within the limits of the
Antarctic surface layer, and these might seem here to show little or no diurnal varia-
tion.

Of previous work on vertical migrations that of Russell (1925-31) is the most im-
portant, but this was done on a much finer scale in the shallow water of the English
Channel. It revealed movements of a kind which could not be detected in the N 100 B
and took into account various subsidiary factors which must be ignored here.

In working out the variations for each species I have omitted the following stations:
(i) All those south of latitude 6o° S. (ii) Those in which it was not possible to make a
reliable estimate of the numbers of the species in question (such as sample? which were

PLEUROMAMMA
ROBUSTA

00 04 ok 12 16 20 DO 04 OS 00 04 08 12 16 20 00 04 08 00 04 08 12 IS 20 00 04 08

200

100

EUPHAU5IA

VALLENTINI/
v T

00 04 08 12 16 20 00 D4 08 00 04 06 12 16 20 00 04 OS 00 04 08 12 16 20 00 04 08

do 04 o'a li is 20 do 04 o'a oo 04 os a is 20 00 04 os 00 04 os 12 is 20 00 04 08

200

00 04 08 12 16 20 00 04 08 00 04 OB 12 16 20 00 04 08 00 04 08 12 16 20 00 04 08

400

200-

do 04 da i2 ife 2b do 04 oe do 04- de <z ife 20 do 04 os 00 04 oe 12 ie 20 00 04 oa

Fig. 18. Diurnal variations of macroplankton species. The curves show the average number per haul at
four-hour intervals. Midnight and midday are accentuated by black and white triangles respectively. The
species are arranged roughly in order of the magnitude of their diurnal variation (see Table I, p. 95).

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

89

swamped with a shoal of Euphausia superba or Salpa). (iii) Those in an area which was
evidently altogether devoid of the species in question, (iv) Those in an area in which
the only samples were from a series of daily stations all taken at the same time of day.

1000-

500

DO 04 08 12 16 20 00 04 08

/

00 04 08 12 IS 20 00 04 08

00 04 08 IE 16 20 00 04 08

800

400

00 04 08 12 16 20 00 04 08 00 04 08 12 IG 20 00 04 08 00 04 08 12 16 20 00 04 08

400

200

00 04 08 12 16 20 00 04 0B

00 04 08 12 16 20 00 04 08

00 04 08 12 16 20 00 04 08

2000-

1000

2000-

00 04 08 12 il 20 00 04 08 00 04 08 12 16 20 00 04 08 00 04 08 12 16 20 00 04 08

PARATHEMISTO GAUDICHAUDI

▼ V T

00-

^\ r—^\

50-

00 04 08 12 16 20 00 04 08

00 04 08 12 16 20 00 04 08

00 04 08 12 16 20 00 04 08

Fig. 19. Diurnal variations of macroplankton species (see legend to Fig. 18).

Figs. 18, 19 and 20 show the variations for each species. With the available data the
average per haul for periods of one hour, or even two hours, gives figures which are a
little too erratic. Four-hour periods, however, give more regular results, and have

90

DISCOVERY REPORTS

therefore been adopted in the construction of the curves. These figures are given in
Table I, p. 95.

The following pages contain notes on the variation of individual species. It has
already been mentioned that during a cruise of the 'Discovery II' a 24-hour station,
461, was worked south of Bouvet Island. At this station seven flights, each of six oblique
closing 1 m. nets, were towed every four hours at a series of depths down to about
700 m., and I have made a preliminary analysis of the samples. Although the present
paper is really concerned with the routine N100B samples, it will perhaps be per-

00 04 08 12 16 20 00 04 08

20-

00 04 08 12 IB 20 00 04 08

00 04 03 12 16 20 00 04 08

00 04 08 12 16 20 00 04 OS

00 04 08 12 16 20 00 04 08

00 04 08 12 16 20 00 04 08

00 04 08 12 16 20 00 04 08

00 04 08 12 16 20 00 04 08

PRIMN0 MACR0PA

T V

▼

2-0

10

00 04 08 12 16 20 00 04 06

Fig. 20. Diurnal variations of macroplankton species (see legend to Fig. 18).

missible to refer here and there to the results of St. 461 in so far as they apply to the
results shown in the table on p. 95. There is also a line of stations (618-25) at which the
net was hauled twice a day at about 10 a.m. and 10 p.m. through a region of uniform
plankton, and these constitute a useful piece of independent evidence.

Diphyes antarctica (Fig. 20). The material is limited, but it is enough to show that there
is no significant variation. Specimens taken at St. 46 1 show no apparent vertical migration.

Dimophyes arctica (Fig. 20). The curve for this species is unreliable, not so much be-
cause the species is uncommon north of 6o° S as because it occurs in sharply fluctuating
numbers. Thus the peak at 0400 is due to only two stations at which the species appeared
in larger numbers than usual. The instances of its occurrence are quite evenly distributed

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 91

throughout the day, and it is very improbable that there is any real variation. The dis-
tribution of this species at St. 461 strongly suggests that there is no variation.

Pyrostephos vanhqffeni (Fig. 20). The curve is actually made out for the numbers of
nectophores, and not the individual colonies. It is a patchy species and the peak at
midday is due to only one station, at which an abnormal number occurred. The actual
occurrences are equally distributed through the 24 hours, and the species can be regarded
as having no significant diurnal variation.

Sibogita borchgrevinki (Fig. 20). Only fifteen specimens were taken north of 6o° S, so
that a reliable curve cannot be drawn. The fifteen were spread equally through the day
and night, but the stations at which it was counted as absent were mostly day stations.
Several specimens occur at St. 461, but there is no sign of any vertical migration.

Solmundella sp. (Fig. 18). This is not a very common form, but in contrast to the four
preceding species, it has a very clear diurnal variation, no specimens having been taken
in the daytime between 0500 and 1859, except in high latitudes. It is not represented
at St. 461.

Tomopteris sp. (large) (Fig. 19). There is a slight suggestion of variation, about
equivalent to that of Calanus aattus, but it is not of much significance. No conclusive
results are to be found in St. 461 or in Sts. 618-25.

Tomopteris sp. (small) (Fig. 19). Similar to the large Tomopteris. All these curves
which show a slight variation have peaks which might be regarded as accidental, were
they not all at or near midnight.

Vanadis antarctica (Fig. 20). As in the case of Sibogita we have insufficient material
for a trustworthy curve. The genus is recorded at only thirteen stations north of 6o° S.
They are spread throughout the day, so that there is not likely to be any very significant
variation. On the whole slightly more specimens were taken at night.

Auricidaria antarctica (Fig. 20). Again there are very few specimens north of 6o° S, but
these show no sign of any variation ; nor do those from farther south. It must perhaps still
be regarded as a doubtful species, but any important diurnal variation is very unlikely.

Calanus acutus (Fig. 19). The averages of a large number of hauls and vast numbers of
specimens show a slight diurnal variation. The species, however, is taken at all hours, and
some of the biggest catches have been in the daytime. There is perhaps a faint suggestion
of a tendency towards vertical migration at St. 461, but there is no sign of any variation
at Sts. 168-25, and it: maY for the Present be regarded as a species without significant
diurnal variation.

Calanus propinquus (Fig. 19). The exact amount of diurnal variation cannot properly
be estimated without a very large body of data. There is evidently rather more than in
C. acutus, but the largest hauls of this species also can be taken in daytime as well as at
night. St. 461 gives no definite indication, but Sts. 618-25 confirm the existence of

some variation.

Calanus simillimus (Fig. 18). The curve shows a clear variation. The species has been
caught at nearly as many day stations as night stations, but the night catches are larger.

It is not represented at St. 461.

4-2

92 DISCOVERY REPORTS

Rhincalanus gigas (Fig. 19). This species again can be taken in the greatest abundance
during the day, but on balance there seems to be some sort of variation. St. 461 and
Sts. 618-25, however, do not suggest any definite migration or variation.

Pleuromamma robusta (Fig. 18). Of all species found in the Antarctic surface water,
this has on the whole the most pronounced diurnal variation. Examples occurred in only
three hauls out of 196 between 0600 and 1959. The vertical migration is shown very
clearly at St. 461. It appears to sink below 600 m. during the day, and rises to within
reach of the N 100 B only within a few hours of midnight.

Metridia gerlachei (Fig. 18). Here again there is a strongly marked migration, but
specimens have occasionally been taken in daytime. At St. 461 there was a wide range
of vertical distribution, but the vertical migration was not well defined.

Haloptilns ocellatus. This species occurred only once north of 6o° S. Specimens from
farther south show no sign of diurnal variation, and St. 461 suggests that there is no
vertical migration.

Haloptilns sp. (Fig. 20). Occurrences are spread equally through the 24 hours, and
the figures from St. 461 suggest no vertical migration.

Pareuchaeta sp. (Fig. 18). The diurnal variations are very clearly defined, and examples
are rarely taken during the hours of daylight. It is evident that this genus lives at a
considerable depth and comes within reach of the net only at night. This supposition
is confirmed by the catches at St. 461.

Heterorhabdus sp. (Fig. 18). The genus is uncommon, but great variation obviously
takes place. St. 461 shows that it lives at a deep level but the series of hauls does not
bring out the vertical migration very clearly.

Eucalanus sp. (Fig. 19). The material again is limited and the numbers in which the
species occurs are variable. There is evidently a certain amount of variation, but there have
been nearly as many occurrences in daytime as at night, and it is difficult to know what
would result from a larger body of data. A vigorous vertical migration is improbable.

Enchirella sp. (Fig. 18). There is little material, but there is evidently a marked
variation. St. 461 shows a considerable amount of vertical migration.

Candacia sp. (Fig. 18). This is another uncommon organism, but there is no question
that it is normally beyond the reach of the net during daytime. The few specimens
occurring at St. 461 were taken in the deepest nets.

Parathemisto gaudichaudi (Fig. 19). There is only a slight variation, roughly equivalent
to that of Calanus acutus. Curiously enough Sts. 618-25 suggest quite a marked diurnal
variation. At successive stations the catches were as follows: 21, 4, 260, 7, 19, 8, 68, 1.
The alternate large catches were those from the night stations. It was absent from
St. 461 except for some doubtful juvenile stages. At all events this species has actually
been caught in both large and small numbers almost as often during the day as during
the night.

Primno macropa (Fig. 20). The behaviour of this species is most remarkable. From
the figures for diurnal variation it seems impossible to avoid the conclusion that it rises
to the surface in the daytime and sinks at night, for although the species is of common

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 93

occurrence in small numbers, none is recorded between 2200 and 0159 except for a
single batch of twenty. On the other hand there is a clear maximum about midday.
Few were caught at St. 461, but there is no sign of such a reversed migration. Diurnal
variation is not of course necessarily proof of a corresponding vertical migration. It is
conceivable, for instance, that a species like this might become localized at night into
small shoals which are missed by the net.

Vibilia antarctica (Fig. 18). The material is not really adequate and the numbers are
patchy. However, there is a most distinct variation. The fact that the peak comes rather
before midnight is due to only two or three samples which were abnormally large.

Ensirus antarcticus (Fig. 19). A very doubtful species. Among those occurring south
of 6o° S there is no evidence of any diurnal variation. Those occurring north of 6o° S
suggest a sharp variation, but there are so few of them that the apparent maximum
around midnight may be quite accidental.

Cyllopus sp. (Fig. 18). There is not very much to go on, but there seems to be a marked
variation. As in Vibilia the peak comes before midnight, but with the available material
we cannot be sure whether this is accidental or whether it represents the actual state

of affairs.

Antarctomysis sp. (Fig. 18). There is a pronounced diurnal variation, though the genus
has been taken practically as often during the day as at night. This is perhaps connected
in some way with the neritic distribution of this genus.

Enphausia superba (Fig. 19). This is a difficult species to deal with owing to its extreme
patchiness and tendency to form shoals. Many of the big shoals have been seen at the
surface during both the night and the day, but the deeper shoals and the more scattered
individuals might undertake vertical migrations. For the estimation of the average per
haul at different hours, samples containing over 1000 E. superba have been disregarded.
This should eliminate the disturbing influence of heavy shoal catches and give some idea
of the general behaviour of the species. The resulting curve suggests only a minor degree
of diurnal variation. At St. 461 the majority seemed to remain near the surface, while
those living at greater depths gave some signs of moving up and down. At Sts. 618-25
there was quite a marked diurnal variation. The explanation would seem to be that while
a section of the population of this species undergoes some vertical migration, the greater
part remains at the same level, especially perhaps when forming shoals.

Euphausia frigida (Fig. 18). The variation is strongly defined, but fair-sized catches
appear now and then during the day and the period of abundance at night is a little
more extended than in such organisms as Pareuchaeta and Metridia. St. 461 suggests
a vigorous migration, and the diurnal variations are clearly shown at Sts. 618-25.

Euphausia crystallorophias. This is a neritic species of which few examples have been
taken, and none north of 6o° S. The few that occur south of 6o° S give some suggestion
of diurnal variation, and since most of the other species of Euphausia undertake vertical
migrations, it is probable that E. crystallorophias does so too.

Euphausia triacantha (Fig. 18). Here the variation is almost as sharply defined as in
Pleuromamma. Examples occurred in only three out of 127 hauls between 0600 and 1759.

94 DISCOVERY REPORTS

At St. 461 very few specimens were taken and the vertical migration is not well denned,
but it is clear that this species lives at a considerable depth.

Eiiphansia vallentini (Fig. 18). Since this really belongs to the sub-Antarctic water
there is not much material to go on. All that can be said is that there is a very definite
variation which seems slightly more marked than that of E. frigida.

Thysanoessa spp. (Fig. 19). There is a clear diurnal variation, but it is not nearly so
marked as in Euphausia triacantha and E. frigida. Examples are taken just as often in the
middle of the day as at night, but the night hauls are on the average larger. At St. 461
Thysanoessa seems much scarcer in daytime, at all depths down to 600 m., than at night.
The variations at Sts. 618-25 are °f tne same order as those shown in the curve.

Cleodora sulcata (Fig. 19). There is evidently a distinct variation. However, the species
occurs quite commonly in small or moderate numbers during the day. Three large catches ,
which were evidently abnormal, have been disregarded in the construction of the curve.
There is evidence of vertical migration at St. 461, but no data from Sts. 618-25.

Limacina helicina (Fig. 20). This species seems to be commoner at midday than at
midnight, but there is not a great amount of material and it is possible that the peak at
midday is accidental. The actual occurrences are distributed through the 24 hours as
evenly as could be expected. There is no definite evidence from St. 461 or Sts. 618-25.

Limacina balea (Fig. 1 9). Like Euphausia superba, this species is difficult to place on
account of its patchy distribution and tendency to form shoals. The curve is constructed
from samples containing less than 1000. This represents the great majority of hauls and
shows quite a modest variation, but it is noteworthy that the catches of 1000 to 6000 all took
place at night, while two or three catches of 20,000 to 1 60,000 seemed to bear no particular
relation to the time of day. There is no definite evidence from St. 461 or Sts. 618-25.

Spongiobranchaea australis (Fig. 19). Here the rather limited material suggests that
there is a slight variation, but the numbers of catches in which the species occurs are fairly
equal through the day. There are no examples at Sts. 618-25, and too few at St. 461 to
justify any conclusions.

Clione antarctica (Fig. 19). The figures suggest a slight variation of the same order
as that of Calanus propinquus. There are few specimens but no indication of vertical
migration at St. 461.

Salpa fusiformis f. aspera (Fig. 18). This is another of those species of which the vast
majority of specimens actually caught have been from a limited number of swarms. If
shoal catches are included in the calculation of the mean values for the different times
of day they produce an unfair distortion of the curve of variation. All samples of 100 or
more Salps have therefore been disregarded here, and the resulting curve shows a regular
and well-marked variation. If all samples of twenty or more Salps are also disregarded,
the shape of the curve is not much affected. There were not enough specimens at
St. 461 to provide additional evidence, and none occurred at Sts. 618-25.

Chaetognatha (Fig. 19). The variations of the group as a whole are equivalent to
those of Calanus propinquus. Large catches are liable to be taken at all times of the day.
St. 461 gives inconclusive results, but there is a distinct variation at Sts. 618-25.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

95

Table I . Mean diurnal variations

Average nurr

bers per haul for 4-hour periods

Hours be-

tween which
hauls are

0000

0400

0800

1200

l600

2000

(2200-

(0200-

(0600-

(1000-

(14OO-

(1800-

regarded as

°J59)

0559)

0959)

!359)

J759)

2159)

valid

Pleuromamma robusta

J45

130

6

0

0

23

2100-0259

Euphausia triacantha

17-9

^■5

0-3

0

0

H7

2000-0359

Solmundella sp.

3-8

2-0

0

0

0

4-6

2000-0359

Euphausia vallentini

no

29O

7"3

0

0-2

I I-I

2000-0359

Euckirella sp.

46-5

5-2

o-i

1-2

6-o

447

2100-0259

Pareuchaeta sp.

183

75

2

4

5

97

2000-0359

Candacia sp.

24-3

0-4

i-8

o-i

0

8-4

2100-0259

Heterorhabdus sp.

25-9

9-2

1-2

0

0-2

13-3

2000-0359

Euphausia frigida

366

458

33

8

39

302

1800-0559

Metridia gerlachei

258

39

!5

n

6

74

2000-0359

Salpa fusiformis f . aspera

8-4

4-0

0-9

o-6

i-o

4-9

2000-0359

Antarctomysis sp.

196

6-2

2-4

2-1

19-6

5°

2000-0359

Calanus simillimus

446

161

82

53

58

166

1800-0559

Vibilia antarctica

2-1

0-7

I-O

0-3

°"5

4'5

1900-0459

Cyllopus spp.

I-I

0-3

o-o

o-6

i-i

26

1900-0459

Cleodora sulcata

87

i-o

i-4

1*4

2-4

5-8

1900-0459

Thysanoessa spp.

1318

779

221

435

385

729

1800-0559

Eusirus antarcticus

2-5

o-o

0-3

07

0-7

2-0

1800-0559

Euphausia crystallorophias

37

1-2

1-2

i-8

4-1

O

All hours

Limacina balea

70

57

38

51

11

3°

1700-0659

Eucalanus sp.

73-2

*57

58-0

10-2

16-4

18-9

1700-0659

Calanus propinquus

779

7*3

261

331

57i

559

1700-0659

Clione antarctica

2-4

2-1

1-2

i-6

i-5

2-3

1700-0659

Chaetognatha

45 r

408

37°

301

220

333

1700-0659

Spongiobranchaea australis

i-o

o-8

o-8

0-4

o-8

o-8

1600-0759

Rhincalanus gigas

1733

2898

2086

1208

2030

2006

1600-0759

Euphausia superba

88

46

46

61

53

96

All hours

Calanus acutus

325°

3969

3453

1849

2890

3021

Tomopteris sp. (large)

0-9

o-8

07

0-7

0-4

0-9

Tomopteris sp. (small)

6-3

4-0

37

4-2

2-9

6-4

Parathemisto gaudichaudi

103

92

73

61

90

92

Sibogita borchgrevinki

0-4

0-2

o-i

0-2

0-2

0-3

Vanadis antarctica

i-o

!"5

—

0-3

0-2

2-0

Dimophyes arctica

2-6

17-7

2-2

7-5

4H

2-5

Diphyes antarctica

I-I

2-2

o-6

2-5

0-4

3-o

Auricularia antarctica

i-o

0

0-3

1-2

0-2

o-8

Haloptilus sp.

22-2

io-o

20-0

I I-O

14-0

4-4

Haloptilus occllatus

48

—

128

I20

164

Pyrostephos vanhoffeni

1 1 -2

48-4

16-1

80-3

10-5

27-1

Limacina helicina

20-5

28-7

36-3

5I-3

3i-9

37-2

Primno macropa

°'5

o-8

i'5

2-0

1-2

i-5

0400-1959

Throughout this paper hours are expressed according to the 24-hour notation. Bracketed figures
at the head of the columns represent the range in time of the observations, while the mean of the
4-hour period, which has been used in the construction of Figs. 18-20, is placed above.

96 DISCOVERY REPORTS

Among the various species there is a complete gradation from those which are
plentiful at night and completely vanish from the catches in daytime, to those which
show no diurnal variation at all, or which even increase during the day and diminish at
night. In Table I on p. 95 the various units are arranged roughly in order of the
magnitude of their diurnal variations, and the figures in the first six columns are the
average numbers per haul, upon which the curves in Figs. 18-20 are based. Thus
Pleiiromamma robusta averages 145 per haul between the hours of 2200 and 0159, and
sinks beyond the reach of the net during the daytime or at least the afternoon. On the
other hand such species as Pyrostephos vanhoffeni show no clear indication of any
diurnal variation and in Primno macropa the variation seems actually to be reversed.

It is now clear that observations taken in the middle of the day must be disregarded
in any consideration of the distribution of the more migratory species, and it must be
decided exactly which hauls can be admitted as an indication of the numbers of each
species present. Without a fuller knowledge of the details of vertical migration, as distinct
from diurnal variation, the best that can be done is to select, for each variable species,
a period of hours before and after midnight which will be appropriate to the diurnal
variations of that species. For Pleiiromamma, for instance, we might allow as valid only
the hauls between 9 p.m. and 3 a.m. (2100-0259), while we need not disregard any
hauls for such organisms as Pyrostephos. Column 7 in Table I shows the periods during
which a haul is regarded, for the purposes of the present paper, as an indication of the
presence or absence of a species. Those listed in the table below Rhincalanus are re-
garded as having insufficient diurnal variation to merit the exclusion of any of the day-
time hauls. It occasionally happens that a species, which on the average is plentiful
only at night, is caught in unexpectedly large numbers during the day. If such a catch
were comparable to the average haul taken at night about the same time and in the same
locality, it may be taken into consideration in the distribution of the species.

Hardy and Gunther (1934) have discussed the vertical migrations of certain species
which are included in Table I and conclude that a more or less marked migration is
shown by the following species : Calanus propinquus, C. simillimus, Metridia gerlachei,
Pareuchaeta antarctica, Parathemisto gaudichaudi, Vibilia antarctica, Cyllopus spp.,
Euphausia superba, E. frigida, E. triacantha, Thysanoessa spp., Limacina helicina, and
Salpa fasiformis. Table I shows that all except four of these species also show a clear
diurnal variation, and it is therefore mainly in agreement with Hardy and Gunther's
results. The exceptions are Calanus propinquns, Parathemisto gaudichaudi, Euphausia
superba, and Limacina helicina. The explanation of the fact that they show little diurnal
variation in the catches of the N 100 B is probably to be explained on the grounds that
they inhabit mainly the upper layers, and that their vertical migrations do not take the
bulk of them beyond the reach of the net at night. Hardy and Gunther find little or no
migration in Calanus acutus and Rhincalanus gigas, and this also is in agreement with
the results expressed in Table I.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

97

RELATIVE ABUNDANCE AND DISTRIBUTION
OF INDIVIDUAL SPECIES

In different parts of the Atlantic sector of the Antarctic the plankton varies considerably
in both abundance and constitution, and to find the actual relative quantities in which
the different species exist would be a very difficult matter, even if it was particularly
desirable to do so. However, it is necessary to give some indication as to which species
are abundant and which species are uncommon, and the following table shows the
average per haul (all hauls with the N ioo B in the Antarctic) and the largest single
catch for each species. A disproportionately large number of these hauls have been
made in the neighbourhood of South Georgia, but this is a region in which, from time
to time, representatives of the plankton typical of both the warmer and colder parts of
the Antarctic waters are found.

Table II. Relative abundance of species {expressed to two significant figures)

Species

Average
number
per haul

Largest
catch

Species

Average
number
per haul

Largest
catch

Abundant

Calanus acutus

1900-00

67,000

Thysanoessa spp.

420-00

1 5 ,000

Rhincalanus gigas

1200-00

20,000

Calanus propinquus

390-00

28,000

Euphausia superba*

1 ooo- 00

190,000

Chaetognatha

200-00

6,300

Limacina balea*

440-00

160,000

Numerous

Euphausia frigida

140-00

5,800 || Pareuchaeta sp.

52-00

5.500

Parathemisto gaudichaudi

no-oo

3,000

Pleuromamma robusta

40-00

6,200

Metridia gerlachei

87-00

6,100

Salpa fusiformis f . aspera*

29-00

6,700

Calanus simillimus

86-oo

41,000

Limacina helicina

21-00

960

Moderate

Antarctomysis sp.

9-80

1,800

Euchirella sp.

3-20

510

Pyrostephos vanhoffeni

9-00

1,200

Cleodora sulcata

3-10

55°

Euphausia valkntini

6-6o

1,800

Haloptilus sp.

2-40

IOO

Eucalanus sp.

5-5°

59°

Heterorhabdus sp.

2-20

260

Haloptilus ocellatus

S-40

520

Dimophyes arctica

1-90

180

Euphausia triacantha

4"5°

230

Candacia sp.

i-6o

IOO

Tomopteris sp. (small)

3-60

100

Clione antarctica

I-IO

40

Few

Vibilia antarctica

0-82

66

Auricularia antarctica

°-53

IOO

Primno macropa

0-77

90

Spongiobranchaea australis

0-40

27

Cyllopus spp.

o-6o

no

Solmundella sp.

0-29

32

Euphausia crystallorophias

o-6o

200

Eusirus antarcticus

0-15

9

Diphyes antarctica

o-59

23

Vanadis antarctica

0-08

3

Tomopteris sp. (large)

0-58

39

Sibogita borchgrevinki

0-05

2

* Species of which the majority have occurred in shoals or dense patches.

There are large differences in the uniformity of distribution of these species, and some
tend to form shoals, or large or small areas of dense concentration. Euphausia superba

98 DISCOVERY REPORTS

lives mainly in shoals, and Limacina balea and Salpa fusiformis, though normally found
in small numbers, sometimes occur in exceptionally dense masses which may extend
over many square miles. Other species such as Parathemislo gaudichaudi and Calanus
simillimus are found only occasionally in abnormally large numbers, the majority of
specimens being taken in hauls of normal size. Other species again seem to have no
tendency to form any definite aggregations. This point will be dealt with more fully
when each species is separately considered in the following pages.

The N ioo B analyses provide a large body of data on the distribution of the various
species, but separate species or groups will be dealt with in detail in subsequent pub-
lications, and the distribution of individual species will therefore be considered in the
briefest possible manner here. For the same reason, little reference will be made to
records of the occurrence of each species in the publications of other expeditions, for
if this were done the following notes would need to be greatly enlarged. Mention
should be made, however, of one or two papers which deal with certain aspects of the
distribution of the macroplankton in these waters. For example, Ruud (1932) discusses
in detail the general biology and distribution of the Antarctic Euphausiidae and the
connection between Enphaiisia superba and the distribution of whales. Further re-
ference will be made to this paper in a later section. Ottestad (1932) has treated the
principal species of Copepoda in a similar way. The substance of these papers does not
very much overlap that of the present paper, but, so far as their conclusions have any
bearing on it, they are mainly in agreement with the following notes. Rustad (1930)
has also published a paper on the identification and development of the southern
Euphausiidae, but deals only briefly with their distribution.

Except where otherwise stated, the following notes are derived entirely from the
samples on which this paper is based.

Diphyes antarctica. This species is confined to the colder parts of the Antarctic
waters, and its northern boundary in the area of the Falkland Islands Dependencies is
roughly a line running from the South Shetland Islands to South Georgia. It is usually
found in the vicinity of the pack-ice, or in water which has recently been covered by
the ice. It seems to be scattered very evenly through the regions which it inhabits,
appearing at each station in small numbers, and there have been no indications of any
tendency to form shoals or even to occur in definitely larger numbers than usual.

Dimophyes orctica. This species is known to occur at least in sub-Antarctic, as well
as Antarctic water, but it is only in the colder waters of the Bellingshausen and Weddell
Seas that it appears in any quantities. Here it is quite plentiful, but in warmer regions
it is rare. Its distribution is more "patchy" than that of Diphyes antarctica, but it
cannot be said to occur in shoals.

Pyrostephos vanhoffeni. Recorded from time to time everywhere from the Antarctic
convergence to the most southerly stations of the Bellingshausen and Weddell Seas,
but it is definitely commoner in the colder than in the warmer parts of the Antarctic
water. It is curious that it seems to be absent from the coastal regions of the South

DISTRIBUTION OF ANTARCTIC M ACROPLANKTON 99

Orkneys, South Shetlands, and the eastern Bellingshausen Sea. Of all the stations in
this area, including those in Bransfield Strait, it is recorded only at one or two stations
off Adelaide Island. The nectophores appear in very variable numbers in the samples,
but this is to be expected, as there are said to be about thirty nectophores to each
colony.

Sibogita borchgrevinki . Another species which is confined to the higher latitudes, or
at least to the colder waters. Its distribution resembles that of Diphyes antarctica, but
it appears to reach a little farther north. There is rarely more than a single specimen in
one sample, and it is probably distributed evenly through the waters to which it belongs.

SolmundeUa sp. This is a species which seems to havs no particular limits. It occurs in
small numbers, but has been found everywhere from the convergence to the cold waters
of the Bellingshausen and Weddell Seas. Browne (1910) mentions that S. mediterranea
is found from the tropics to the Antarctic. There are rarely more than ten in one sample,
and I have no record of more than thirty-two.

Tomopteris sp. (large). The large Tomopteris also is found everywhere from the con-
vergence to the higher latitudes, but it is distinctly commoner in the colder regions.
Like Pyrostephos, it seems to be absent from the Bransfield Strait, and the whole of the
coastal region from the South Orkneys to the eastern part of the Bellingshausen Sea.
It is an evenly distributed species and more than two or three in one sample are rarely
found. Only once have more than twelve been taken at a time, and that was an extra-
ordinary catch of thirty-nine at a station off South Georgia. There is an interesting
colour variety of Tomopteris carpenteri. Ordinary specimens are colourless and trans-
parent, but on certain occasions specimens have been taken which were traversed by
two broad bands of bright reddish brown which made the animal strikingly conspicuous.
These parti-coloured specimens have been found only in cold water in the neighbour-
hood of the pack-ice of the Weddell and Bellingshausen Seas.

Tomopteris sp. (small). Like the large Tomopteris it is found everywhere in the Ant-
arctic water, but is rather commoner in the colder regions. Unlike the other it is found
in the coastal region of the South Orkneys, South Shetlands, and eastern Bellingshausen
Sea. Its distribution is sometimes quite uniform and sometimes rather patchy, but
I have no record of anything which might be called a shoal.

Vanadis antarctica. Occurs in such small numbers that one cannot draw very certain
conclusions as to its distribution. It is almost entirely confined to the cold water of the
Weddell and Bellingshausen Seas, especially the latter, but it has also been found at one
or two stations near the Antarctic convergence in far wanner water than all the others
at which it is recorded. There seems to be no doubt about the identity of these speci-
mens, and their occurrence in such an unusual locality is difficult to account for. There
is rarely more than one in a single sample.

Auricidaria antarctica. Like Diphyes antarctica, this species is strictly confined to the
colder regions, but since it occurred at WS 198, WS 474 and St. 594 (see Figs. 10,
12 and 14), its northern limit must be placed a little beyond that of Diphyes. It is
found in the Bellingshausen Sea, the coastal waters of the South Shetlands and South

5-2

ioo DISCOVERY REPORTS

Orkneys, and the eastern Weddell Sea. It generally occurs in ones and twos, but catches
up to ioo were taken in the eastern Weddell area.

Calanus acutus. Numerically this is the most important species of all. It is dis-
tributed throughout the Antarctic surface waters, and there have been catches of iooo
or more everywhere from the convergence to the cold waters of the Bellingshausen and
Weddell Seas. The very big catches of 5000 and more have been taken only in the South
Georgia and South Sandwich areas, the western Bellingshausen and eastern Weddell
Seas. The species is on the whole evenly distributed and there is no tendency to form
shoals.

Calanus propinquus. This species also is found everywhere in both the warm and
cold parts of the Antarctic surface water, but the largest catches have been taken in the
neighbourhood of South Georgia and the South Sandwich Islands. It has no tendency
to form shoals, but is perhaps a little more localized than C. acutus. For instance at
Sts. 360-74 in February and March, 1929-30, it was taken in enormous numbers in the
eastern part of the Scotia Sea, and it was by far the most prominent copepod at
Sts. 453-72 between Bouvet Island and South Georgia in October 1930-1. Specimens
taken near the Antarctic convergence are mostly immature.

Calanus simillimus. Normally confined to the warmer waters of the Antarctic. It has
not been taken in the Bellingshausen or Weddell Seas and has not often appeared south
of a line running from the north side of the South Shetlands to South Georgia, though
around South Georgia it is common enough. In March of the warm 1929-30 season,
however, it was taken at Sts. 372-5, west of the South Sandwich Islands. This is normally
quite an evenly distributed species, and an abnormally large catch has been taken only
once, at WS 184, where over 40,000 were taken at a point close to the south coast of South
Georgia. The next largest catches were 5000 at WS 466 just on the north side of the
convergence, and 2200 at St. 321 off South Georgia.

Rhmcalanus gigas. This species has been caught in large numbers at practically every
station in the warmer parts of the Antarctic, and there is little doubt that it is abundantly
and evenly distributed in these waters at least during the greater part of the summer.
It is also taken in large numbers in the Bellingshausen Sea and in the Weddell Sea
water, but here it is not by any means always found in abundance. It occurs in small
numbers, though at most stations, in the coastal regions of the South Orkneys, South
Shetlands and eastern Bellingshausen Sea. It shows no tendency to form shoals, and
cannot be regarded as any more patchy than any of the other more numerous species.
Pleuromamma robusta. This is a warm-water species whose proper habitat in Ant-
arctic water appears to be the outer belt immediately south of the convergence. It is
of course also found north of the convergence. It is not uncommon in the Weddell Sea
water on both the west and east sides of the South Sandwich Islands, but it has occurred
at none of the stations in coastal regions of the South Orkneys and South Shetlands, nor
in the Bellingshausen or eastern Weddell Sea. As a rule it is as evenly distributed as
most of the other copepods, but a very exceptional catch of over 6000 was taken at
St. 452, north of Bouvet Island.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 101

Metridia gerlachei. On the whole a cold-water species. A peculiarity of its distribu-
tion is that it is found in much the largest quantities between the South Shetlands and
South Orkneys, along the Orkney-Sandwich ridge, and around the South Sandwich
Islands. It is one of the very few members of the macroplankton which are found in
any considerable numbers in this region. Possibly its method of reproduction is in some
way associated with the bottom or with the conditions found in shoal waters. However,
it is to be found in varying quantities everywhere from the convergence to the highest
latitudes. It is a species which is often distributed evenly over large areas but which
sometimes appears suddenly in exceptionally large numbers. It thus has some tendency
to form shoals, or at least to appear in local concentrations, and it must be regarded as
a patchy species.

Haloptilus ocellatus. This species is confined to the coldest water. It has occurred in
moderate numbers (ioo to 500 per haul) in the Bellingshausen Sea west of Peter 1st
Island and in the eastern Weddell Sea. Occasional specimens are recorded from the
Orkney- Shetland region and from the ice-edge near Bouvet Island. There is no evidence
to show that its distribution is at all patchy.

Haloptilus sp. Found everywhere from the convergence to the Bellingshausen Sea,
but it is not recorded from the Sandwich or eastern Weddell regions, and has appeared
only once in the Orkney- Shetland coastal region. However, it is among the rarer
Copepoda, and there is not enough material to form the basis of any definite conclusions.
It is perhaps commonest in the warmer parts of the Antarctic.

Parenchaeta sp. This genus also has a wide distribution, but it is commonest in the
warmer parts of the Antarctic. Like Rhincalanus it pervades the whole of the belt lying
immediately south of the convergence and is absent only from hauls taken during the
daytime. It appears sometimes in the Bellingshausen and Weddell Seas, and is met
with in small numbers in Bransfield Strait, in the coastal waters of the South Orkneys
and South Shetlands and in the eastern Bellingshausen Sea.

Heterorhabdus sp. Although this is an uncommon genus it clearly belongs to the
warmer parts of the Antarctic water; its distribution seems very similar to that of
Calanus simillimus.

Eucalanus sp. Also clearly typical of the warmer water, but its occurrence is curiously
spasmodic. At one time it may be found at every station of a line, say, from South
Georgia to the convergence, while at another time none will be caught along a similar
line. On the other hand its distribution does not appear to be in any way patchy.

Euchirella sp. This is one of the rarest of the large Antarctic copepod genera.
It has occurred here and there in the warmer water, off South Georgia, in the
Bellingshausen and Weddell Seas and in the outflowing Weddell water. We cannot say
more than that, as a genus, it is not confined either to the warmer or colder Antarctic
water.

Candacia sp. This is another scarce genus, but it has not been found in the
colder regions and is commonest close to the convergence. It is clearly a warm-
water form.

io2 DISCOVERY REPORTS

Parathemisto gaudichaudi. The big catches of this species have all been taken in the
warmer parts of the Antarctic water, including the South Georgia whaling grounds.
Moderate numbers have occurred in the South Sandwich region and small numbers in
the coastal region of the South Orkneys and South Sandwich Islands and in the Bellings-
hausen Sea. In the warmer waters there are practically no night stations at which it was
not present, but there is no doubt that it is a patchy species and it sometimes occurs
in dense local concentrations. During the 1927-31 period 2000-3000 have been taken
on several occasions, and some very big hauls were taken off South Georgia during the
first commission of the ' Discovery '. Sometimes, however, it is distributed quite evenly
over a large area.

Primno macropa. This species seems to have no particular limits to its distribution
in the Antarctic. It is perhaps slightly commoner in the warmer waters, but it is found
in small numbers almost everywhere. There is no evidence of any tendency to form
shoals or local concentrations.

Vibilia antarctica. Found everywhere from the convergence to the Bellingshausen
and Weddell Seas. It seems to be of very regular occurrence around the South Orkneys.
It never occurs in very large quantities.

Eusirns antarcticus. Another strictly cold-water species whose northern limit corre-
sponds closely with that of Diphyes antarctica. Few specimens have been taken and
there is no reason to suppose that its distribution is at all patchy.

Cyllopus spp. Quite evenly distributed from the convergence to the colder regions.
As a genus it seems to have no preference for either the warmer or colder parts of the
Antarctic, and there appears to be no tendency towards local concentrations.

Antarctomysis sp. This must be classed as a neritic genus. Among the samples upon
which the present paper is based it has been found only in the South Georgia and South
Sandwich region, but according to Rustad (1930, Mysidacea, p. 20) its distribution is
circumpolar. The greatest number have been taken from the shallow water on the south
side of South Georgia.

Euphausia snperba. In its habits and distribution the " krill ", as it is commonly called,
stands apart from all other macroplankton species. It will form the subject of separate
publications, and it will suffice here to say that it is seldom found north of a line running
from the north side of the South Shetland Islands north-eastwards past the western end
of South Georgia, and it has not so far been taken in any large numbers in the Bellings-
hausen and eastern Weddell Seas. It is one of the few species which are found in
occasionally large numbers in the coastal region of the South Orkneys and South
Shetlands. The average number for all hauls (see Table II, p. 97) shows that numeri-
cally it is among the most abundant of macroplankton species, and in mass of living
matter it might well be equal to all the rest of the Antarctic macroplankton combined.
In its distribution it is the most patchy and irregular of these species, and the great
majority of specimens have been taken from actual shoals.

Euphausia frigida. This species occurs mainly in the warm waters of the outer Ant-
arctic Zone and in the outflowing Weddell water in the South Sandwich neighbourhood.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 103

In the coastal region of the South Orkneys and South Shetlands and in the Bellings-
hausen Sea it occurs sometimes in small numbers. It was not taken at stations in the
eastern Weddell Sea. In general it occurs in varying quantities but appears never to
form shoals.

Euphausia crystallorophias. Like Antarclomysis this must be regarded as a neritic
species. It is found, however, only in the colder regions such as the Bransfield Strait
and the coastal waters of the eastern part of the Bellingshausen Sea.

Euphausia triacantha. This species occurs in a more strictly limited zone than perhaps
any other. To the north it is found only a little way beyond the convergence and to the
south it is bounded normally by a line from the South Shetlands running north-
eastwards and passing a little to the south of South Georgia. It is thus a typically
warm-water species, but it is curious that it does not normally extend far into the
sub-Antarctic Zone as do most of the other species typical of the warmer parts of the
Antarctic. This fact may perhaps be in some way due to the depth at which it lives.
There is no reason for supposing that it forms shoals or local concentrations.

Euphausia vallentini. This is reallv a sub-Antarctic species, but it sometimes strays
south of the convergence and has even been taken off South Georgia on one or two
occasions. Zimmer (1914) records a specimen from 580 29' S, 890 58' E, a very high
latitude for this species.

Thysanoessa spp. There have been very few samples which did not contain repre-
sentatives of this genus. It seems always to be scarce in the Bellingshausen and eastern
Weddell Seas, but in the warmer parts of the outer Antarctic, in the outflowing Weddell
water and around South Georgia it is often taken in thousands. As a genus it is very
variable and patchy, and this characteristic is probably shared by T. macrura and
T. vicina.

Cleodora sulcata. Distributed everywhere from the Bellingshausen Sea to the coldest
regions. In most places it is taken normally in ones and twos, but it is more plentiful
in the Weddell Sea water in the neighbourhood of the South Sandwich Islands. Here
there have been single hauls containing several hundred specimens, and such hauls
might be regarded as indicating at least a local concentration. It is distinctly a cold-
water species.

Limacina helicina. This is pre-eminently a cold-water species, but at times it is found
as far north as the convergence. There are no very definite limits to its distribution, but
its occurrence seems to vary according to the time of year. There is no indication that it
forms shoals or dense local aggregations such as we find in the case of L. balea.

Limacina balea. The distribution of this species is quite different from that of
L. helicina. The two are sometimes found together, but L. balea is more of a warm-
water species and seems to reach its greatest development near the convergence. No
specimens have been taken in the Bellingshausen Sea or in the eastern Weddell Sea,
though it occurs in the outflowing Weddell water. In the warmer parts of the Antarctic
water it has occasionally been found in enormous swarms. Some big catches have also
been taken off South Georgia and between South Georgia and Bouvet Island.

io4 DISCOVERY REPORTS

Spongiobranchaea anstralis. Found everywhere from the convergence to the coldest
regions. It is very evenly distributed, but is possibly a little commoner in the warmer
water than elsewhere. There are rarely more than one or two in a sample and there is
no evidence of any tendency to form shoals or local concentrations.

Clione antarctica. This species also is widely distributed, but it is rather commoner
in cold than in warm water. It is occasionally found as far north as the convergence.
On the whole it is very evenly distributed, occurring generally in very small numbers
and showing no tendency towards forming local aggregations.

Salpa fusiformis f . aspera. Found everywhere from the convergence to the Bellings-
hausen and Weddell Seas. Its distribution is not unlike that of Metridia, for it has been
caught in large numbers at all the night stations and at most of the day stations in the
vicinity of the South Orkney Islands. It seems that there is nearly everywhere a sprink-
ling of Salps, and here and there a dense patch. Catches in the South Sandwich region
have all been small, but big hauls have been taken off South Georgia and one large
aggregation has been met with in the Bellingshausen Sea. The shoaling tendency of
the Salps is of a different type from that of the "krill". The former seem occasionally
to be concentrated in large areas where the numbers reach a maximum in the centre.
A typical example is provided by seven successive stations (564-70), covering some 300
miles in the Bellingshausen Sea. At these the numbers of Salps taken were, respectively,
o, 4, 66, 550, 369, 57, o. Enphaiisia superba occurs in smaller, denser and more numerous
shoals.

Chaetognatha. This group, which is principally made up of Eukrohnia hamata, is
found in the greatest numbers in the outer, warmer part of the Antarctic water, and the
largest catches have, in general, been those closest to the convergence. Large numbers
are not likely to be taken south of a line running from the South Shetlands to South
Georgia, but some quite large catches have been taken in the western Bellingshausen
Sea. The distribution of the Chaetognatha appears on the whole to be very regular, and
it would appear that Eukrohnia hamata, at any rate, does not tend to form shoals. This
species, according to Bigelow and Leslie (1930, p. 564), "ranges from Arctic to the
Antarctic in the Atlantic, being confined to the bathyplankton in low and mid-
latitudes".

It will be seen from the foregoing notes that the macroplankton can be divided very
roughly into three groups : (i) those species which have a definite preference for, or
which are actually confined to, the cold water or high latitudes, (ii) those which are
definitely typical of the warmer water or lower latitudes, and (iii) those which may be
found everywhere and may be taken in maximum numbers in both the warmer and
colder regions.

It has been explained that the Antarctic surface water in the greater part of these
regions is drifting mainly in an easterly or north-easterly direction, and, as we
should expect, the isotherms are roughly parallel to the direction taken by the oceanic
currents (see Fig. 22). In Fig. 21 lines are drawn to show the distributional boundaries

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

i °S

of those species which show a clear limitation of distribution. For instance, no specimen
of Euphausia vallentini has been taken at any station on the south side of line No. 4, and
none of Diphyes antarctica to the north of line No. 1. Euphausia superba may be found
in large numbers anywhere south of line No. 1, but is very rare to the north of it. It
will be seen that the majority of these lines also run roughly parallel to the direction of
drift, and indeed it would be surprising if they did not. It is one thing, of course, to say

Fig. 21. Apparent distributional limits of
macroplankton species.

1, Northern limit of Diphyes antarctica and normal
northern limit of Euphausia superba.

2, Northern limit of Sibogita borchgrevinki.

3, Northern limit of pelagic Eusirus antarcticus.

4, Southern limit of Euphausia vallentini.
Northern limit of Auricularia antarctica.

6, Southern limit of Euphausia triacantha.

7, Southern limit of Candacia sp.

8, Southern limit of Limacina balea.

9, Southern limit of Calanus simillimus.

io, Southern limit of Pleuromamma robusta and northern limit of Haloptilus ocellatus.

that a species has not been found to the north or to the south of a particular line, and
another thing to say that it never occurs there. It must be emphasized therefore that
the lines shown on this chart are tentative. The utmost that can be inferred with
certainty is that there is a belt of change between South Georgia and the South Shetland
Islands. That is to say, an observer moving in a north-easterly direction in this part of
the Antarctic water would find few qualitative changes in the plankton, whereas if he

io6

DISCOVERY REPORTS

travels south-eastwards across the line of the isotherms he will meet with important
changes, especially while crossing a belt running approximately from the South Shetland
Islands to South Georgia. This zone of rapid qualitative change will be mentioned again
in a later section. The point to be emphasized at the moment is that the chart indicates
a connection between the distribution of species and the surface temperature of the
water within the Antarctic Zone. This temperature ranges from nearly - 2° C. to about
+ 50 C, and a quantitative comparison of the different species in respect of their

Fig. 22. Approximate positions of the mean summer isotherms.

occurrence in water of different temperatures will provide a clearer distinction between
the warm- and cold-water species than is given in the notes on general distribution.

It must be supposed that plankton organisms which have an optimum temperature
cannot make active movements to avoid changes in temperature caused by the seasons,
the weather, the approach and departure of ice, etc. We should therefore expect more
reliable results from the mean temperature of the locality than from the actual tempera-
ture of the water in which a species is taken. It would be desirable if possible to compare
the distribution of a species with the temperature of the water at the mean depth at
which it lives. But at a great many of the stations at which the N ioo B was used the

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

107

Table III. Distribution and temperature.
Average numbers per haul between the mean summer isotherms

— i-oo

o-oo

o-oo

1 00

2-00

3 '00

4-00

5-00

600

Isotherms: °C

to

to

to

to

to

to

to

to

to

- 'TO

- 0-99

099

199

299

3-99

499

5-99

699

Warm-water species

(a) Practically confined to water above 30 C.

Euphausia vallmtini °'°8

4-So

29-1

370

44- 1

(b) Typical warm-water species.

Eucalamis sp.

—

—

—

2-80

7-90

I7'3

14-4

498

969

Candacia sp.

—

—

2-30

—

3-80

138

5-00

20-8

18-7

Heterorhabdus sp.

—

—

I2-I

0-40

356

8-8o

5H

n-4

16-3

Euphausia triacantha

—

—

o-88

5-i8

21-8

IO-2

187

321

700

Calanus simillimus

—

—

5-80

28-3

190

344

820

871

293

Pleuromamma robusta

—

5-20

3'4°

73'i

201

400

146

104

252

(c) Warm-water species sometimes found in colder regions.

Chaetognatha

785

98-2

53'5

129

375

532

763

1550

2320

Limacina balea

117

813

470

97-8

22-6

250

394°

710
637

Pareuchaeta sp.

72-0

24-8

25-5

140

200

32-5

216

250

Parathemisto gaudichaudi

0-50

290

263

23-1

129

182

38-8

148

179

Euphausia frigida

—

7-00

581

iS°

34i

245

205

0-20

Widespread species

(d) Species found on all isotherms but with a slig

it preference for warmer water

Primno macropa

262

046

o-37

i-oo

1-50

3 94

9-54

5-50

Spongiobranchaea australis

0-20

0-16

0-40

0-63

0-41

2-37

0-36

0-44

i-oo

Thysanoessa spp.

48-5

147

513

1 100

1 1 20

1 1 60

954

1220
1060

1 220

Rhincalanus gigas

2460

1270

393

755

2470

1443

"37

1 1 10

(e) Neutral species.

Haloptilus sp.

51-2

002

°-95

0-19

2-68

042

400

6-15

Euchirella sp.

4-20

15-2

—

960

1-20

970

2-50

Solmundella sp.

—

0-48

096

0-36

0-62

0-22

1-14

i-33

4-57
067

Cyllopus spp.

0-50

1 04

0-82

1-40

110

i-oo

o-37

o-33

(/) Species found on all isotherms but with a slig

tit preference for colder water.

Calanus acutus

8830

1 100

1500

1110

1530

492

87-1

189

9-60

Vibilia antarctica

4' 1 6

2-58

1-40

067

2-98

0-40

0-25

0-14
265

Calanus propinquus

296

54°

1530

2510

811

327

577

20-0

Cold-water species

(g) Cold-water species which may occur in large numbers anywhere south of the

30 isothen

m.

Euphausia superba

I'OO

336

434

822

1700

95-4

i-8o

o-io

o-io

Cleodora sulcata

3"33

2-55

575

8i-8

069

010

o-37
4-60
o-iS

0-14

6-40

3'°°

Salpa fusiformis f. aspera
Tomopteris sp. (large)

166

1-27

S2'4

0-52

79-8
060

5°'4
1-17

24-8
0-98

1 1 -4

0-2I

4° '4
014

Tomopteris sp. (small)

18-0

36-1

5'7°

5-00

4-70

2-20

68-o

1 -oo
267
0-56

370

9-40

Metridia gerlachei

421

S43

289

538

"7

I 30

0-25

Clione antarctica

589

2-64

1 60

i-37

091

1-25

Limacina helicina

80-2

I9-S

25-5

41-6

7-20

7-60

5-50

3'°°

Pyrostephos vanhoffeni

22-6

32-0

S-30

i4'4

8-90

3'3°

1-40

3"5°

3'3°

(/i) Species typical of the coldest regions, which rarely or never approach the cor

vergence.

Sibogita borchgrevinki

0-27

o-io

—

0-08

012

0-07

0-20

1-40
067

Dimophyes arctica
Vanadis antarctica

14-4

7-27

7-80
0-19

o-8o

077

2-20
O-03

1*70
o-oS

—

—

0-42

Diphyes antarctica

2-09

1-48

0-67

i-6i

0-08

Eusirus antarcticus

1 -57

o-95

0-70

0-07

CIO

Auricularia antarctica

4-55

2-50

0-31

019

002

Haloptilus ocellatus

108

24-8

040

—

Neritic species

Antarctomysis sp.

—

—

—

38-0 j 48-0

—

~

Euphausia crystallorophias

—

1-06

0-07

— —

6-2

ioS DISCOVERY REPORTS

temperature was taken only at the surface. However, the mean surface isotherms bear
a sufficiently uniform relation to the mean isotherms at 50 or 100 m. and will serve our
purpose here. Since the great majority of stations have been taken in the summer
months a much more reliable chart can be constructed of the mean summer isotherms
than of the mean annual isotherms. Either would suffice, however, for we are seeking
a relative and not an absolute correlation. Fig. 22 shows the mean summer isotherms
and is a provisional chart derived from temperatures recorded in the three seasons
1928-31.

In Table III the species are arranged in order of their apparent preference for warm
or cold water. The figures in the table represent the average number of each species
per haul. Samples have been disregarded where the haul was made at a time of day when
the diurnal variations of any particular species precluded the taking of a representative
sample. Catches in the Bransfield Strait, where the plankton is disproportionately thin,
have usually been omitted, and for all but one or two of the rarer species, only one
station has been counted for each line of stations in each South Georgia survey. The
end station of the line is taken, or the outermost one permitted by the diurnal variations.
The actual figures are much influenced by variations in the general richness of the
plankton in different localities, or by patchiness of the plankton, and are therefore
individually unreliable. Thus the "— i°" column depends almost entirely on a few
stations in the Bellingshausen Sea, west of Peter 1st Island, where the plankton was
particularly rich, and the figures in this column are consequently deceptively high. The
actual figures are therefore used only as the principal basis of a general grouping of the
species according to the temperature belt they mainly inhabit.

The table shows the distinction between the warm-water species, the widespread
species, and the cold-water species. The neritic species are unimportant and are shown
in a separate category. There is of course no very hard and fast distinction between the
main groups or the subsidiary groups, and a larger body of data might bring about some
modification of the order in which the species are listed.

It is interesting to compare this table with the table of mean diurnal variations, for
it will be seen that the warm-water species are mostly those which have the most
marked diurnal variations and many of the cold-water species have no significant
variations. There are exceptions, for Parathemisto shows little diurnal variation, while
Salpa and Metridia vary considerably. Hardy and Gunther (1934), however, find a
distinct vertical migration in Parathemisto. The phenomenon is no doubt partly con-
nected with the reduction or absence of darkness in summer in the high latitudes in-
habited by the cold-water species, but the question will not be pursued any further here,
since it can be more properly dealt with when the samples from the 24-hour stations
have been finally worked out.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 109

DISTRIBUTION OF RICH AND POOR PLANKTON

It is well known that as a general rule the plankton of the surface waters of the tropics
is very thin, that in the temperate regions it is richer, and that in the Arctic and Antarctic
it is comparatively abundant. Murray and Hjort (1912) mention that " the closing nets
of the ' Michael Sars ', when hauled from 200 m. to the surface in the Sargasso Sea,
yielded on the average 3 c.c. of plankton, while in the Norwegian Sea from 85 to 225 c.c.
were obtained in numerous similar hauls". According to Jespersen (1923) the volume
of macroplankton in 50-600 N in the North Atlantic is 10-20 times as great as in 20-
300 N. The upwelling in the Antarctic Zone of water rich in nutrient salts results in a
luxuriant development of phytoplankton, and this in turn supports an abundant animal
plankton. Hart (1934) mentions the richness of the Antarctic phytoplankton, and
Hentschel (1928) and Hentschel and Wattenberg (1930) publish charts of the South
Atlantic which show the association of areas of rich plankton with the areas of upwelling
of cold water.

The richest plankton of all is found in the Antarctic surface water, but it is by no
means uniformly abundant over the whole of the Antarctic Zone. The quantity of
macroplankton, for instance, varies very much at different times and in different places,
and these variations so far as they can be ascertained from the present material will be
described in the following pages.

A chart showing the distribution of the total number of organisms per haul is roughly
representative of the distribution of the richness of plankton, though it does not neces-
sarily illustrate the distribution of the majority of species. It represents rather the half-
dozen commonest species. But the main difficulty in charting the richness of the plank-
ton is, perhaps, that there is no quite satisfactory means of allowing for diurnal variations,
since different samples have different proportions of the variable species. However, the
few species like Calanus acntus and Rhincalamis gigos, which make up the bulk of the
plankton, do not have a very marked diurnal variation and the time of day has a far smaller
effect on the amount of plankton in a sample than the actual distribution of quantity.

GENERAL DISTRIBUTION
Fig. 23 shows the numbers of organisms taken at all stations in January, February
and March in all four seasons (1927-31), with the exception of the intensive lines of
stations taken in the South Georgia and Bransfield Strait surveys. The figures stand for
the number of hundreds of organisms in each sample, but the shoaling species, Euphausia
superba, Limacina balea, and Salpa fusiformis are omitted. For instance at WS 551, the
last station on the line to the eastern Weddell Sea and marked 56, just over 5600
organisms were taken. Although the diurnal variations have a minor effect on the total
contents of a sample, the figures are shown in italics for hauls made in daytime between
the hours of 0600 and 1700. The shading is based purely on the figures shown on the
chart, and is intended merely to draw attention to the positions at which large, medium
or small hauls were taken.

no

DISCOVERY REPORTS

This chart shows that around the coasts of Graham Land, the South Orkneys, the
South Shetlands, Adelaide Island, Alexander ist Land, etc., the summer plankton is very
thin. Away from this region in all directions it seems that the richness of the plankton
increases, and reaches a maximum in the South Georgia and South Sandwich region,
and to the west of Peter ist Island in the Bellingshausen Sea.

Fig. 23. Distribution of macroplankton quantities in summer. Figures show the number of hundreds of
organisms in each sample, shoaling species being omitted. Those in italics are day hauls and others are night
hauls.

There can be no doubt of the contrast which normally exists between the South
Georgia and Graham Land areas, for these regions have been visited many times in a
succession of years. Nor can there be any doubt that the outer parts of the Antarctic
water between South Georgia and the Falkland Islands, harbour a quite rich plankton
during at least a large part of the summer. Cruises in the Bellingshausen and Weddell
Seas have been made on fewer occasions, but it is probable that there is always a rich
summer plankton in these regions.

DISTRIBUTION OF ANTARCTIC M ACROPLANKTON m

It may be mentioned that Hart (1934) finds that the Bransfield Strait is on the average
also poorer in phytoplankton than other areas, but the contrast is not so great as it is
with the macroplankton.

The reason for the scarcity of plankton around the South Shetlands and South
Orkneys cannot at present be stated with certainty, but this is an area in which there is
a considerable upwelling of water from the deeper layers, and these deeper layers are
of course comparatively poor in plankton. I am informed by our hydrologists that the
region of thin plankton shown in Fig. 23 coincides roughly with an area in which the
surface water has been found to have a low oxygen content. This suggests that the water
has recently risen from a deeper level where the consumption of oxygen has not been
balanced by exposure to air. It is curious that some species, such as Euphausia superba and
Salpa fusiformis, are nevertheless sometimes found in large numbers in parts of this area.

VARIATIONS IN ABUNDANCE
There is evidence to show that in the course of the year certain changes take place in
the distribution of the amount of plankton, and although these changes cannot be
followed with certainty at the present stage it is worth while to consider the material
in some detail. It has not been found practicable to make a series of repeated observa-
tions in one place throughout a year or part of a year, and it will be necessary to compare
the conditions in, for instance, December in one season with April in another season,
and so on. Conclusions drawn in this way must of course be formed with caution.

South Georgia

Figs. 24-30 represent the results of successive surveys of the South Georgia whaling
grounds in the four seasons 1927-31, and they are constructed on exactly the same lines
as Fig. 23.

The essential facts shown by these charts are as follows.

In the 1927-8 survey (February-March, Fig. 24), the plankton was very thin to the
north, east and west, but abundant on the south-west side. The increase here was not
due to a special development of one species but to a general increase of all the im-
portant species.

In the first 1928-9 survey (August-September, Fig. 25), there was a very thin
plankton to the west, north, east and south-east. The south and south-west sides were
not examined.

In the second 1928-9 survey (December-January, Fig. 26), the plankton was in most
places moderate or rather poor, but there was a patch of rich plankton to the south-east.
In general it might be called " patchy". Two extra lines (WS 417-26) worked in April
of the same season (Fig. 27) produced fairly heavy catches near the south end of the
island, but indicated a rather scarce plankton to the south-west.

In the 1929-30 survey (January-February, Fig. 28), the plankton was patchy in
places but in general very abundant. In the same season South Georgia was revisited
in May. Although the routine N 100 B was not used again, closing N 100 B's were

X 4tf JO 39^

34 ya"

)?/ 6 4 18 19 36

3 O

Jty 3-f

Fig. 24. Distribution of plankton quantities around South Georgia, 1927-8, Sts. WS 144-93.

30 +0* 30 39*

•/

I
•/

I
-•II"

•4

I
•2

I
•2

1

•/a

■o

■o

-&&-€

V

TSN

St

1 8 ^ / e i2 J

40' .«• 39* JO' 36* J£" 37*_

3i_ .'o 3+'

Fig. 25. Distribution of plankton quantities around South Georgia, 1928-9, Sts. WS 257-96.

DISTRIBUTION OF ANTARCTIC M ACROPLANKTON

113

towed at different levels at an uncompleted 24-hour station (393, see Station List). The
catches are not strictly comparable to the routine samples, but they strongly suggest
that off the north-east coast in May there was far less plankton than in January.

Fig. 26. Distribution of plankton quantities around South Georgia, 1928-9, Sts. WS 321-72.

Fig. 27. Distribution of plankton quantities around South Georgia, 1928-9, Sts. WS 417-26.

In the 1 930-1 survey (November, Fig. 29), a very rich plankton was again found on
the north side of the island, and a slightly less dense population to the south. The line
repeated in March in the same season (WS 565-75) showed a considerable reduction
in the plankton on the north side of the island (Fig. 30).

Fig. 28. Distribution of plankton quantities around South Georgia, 1929-30, Sts. 300-58.

Fig. 29. Distribution of plankton quantities around South Georgia, 1930-1, Sts. 478-525.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

115

The main plankton survey has usually been carried out near the middle of the season,
and in those years in which further hauls have been made later in the summer there has
been evidence of a reduction in the amount of plankton towards the autumn, at any
rate on the north side of the island.

Fig. 30. Distribution of plankton quantities around South Georgia, 1930-1, Sts. WS 565-75.

The following list includes all the N 100 B samples taken in the immediate vicinity
of South Georgia during the period 1927-31, and shows the average number of organ-
isms per haul for each group of stations.

Average number
Number of organisms

Date
1927-8 Feb .-March

April

1928-9 Aug.-Sept.
October
November
Dec-Jan.

April

1929-30 October
Jan -Feb.

May

1930-

November
March

Station

WS 144-93
WS 196

WS 257-96
WS311
WS313
WS 321-72
WS 417-26

WS464

3°°-58
393

475-525
WS 565-75

Position

South Georgia survey
South-west side of South Georgia

South Georgia survey
South-east of South Georgia
South-east of South Georgia
South Georgia survey
South side of South Georgia

North side of South Georgia
South Georgia survey
North side of South Georgia

South Georgia survey
North side of South Georgia

of hauls per haul

44
1

3.9*3

2,361

35
1

468
326

1

51

10

7.545
4.674
4.488

1

2,146

54

1

19,224
(Reduced
numbers)

47
7

9.485
2,062

In Fig. 31 all four seasons are taken together and the average number of organisms
per haul for each group of stations is plotted according to the month or mean date at
which the stations were taken. As would be expected the plankton is much richer in the
summer than in the winter, but the regularity of the curve is broken by the low figure
for the 1928-9 survey (December-January). It is difficult to account for this, but the
patchiness of the plankton may be seen from Figs. 24, 26 and 27, and might well result
in irregularities in the curve.

7-2

n6

DISCOVERY REPORTS

Little reliance can be placed on the single stations (WS 311, etc.) for reasons given
on p. 71, but it will be seen that they support the suggestion that the plankton increases
rapidly about the middle of November. Apart from this no reliance can be placed on the
curve, for the reason that the plotted points are derived from different seasons, and the
volume of plankton production in these seasons may have been different.

1X9-30

MONTHS

Fig. 31. Quantities of plankton at different times of year around South Georgia.
Figures on the vertical scale represent numbers of hundreds of organisms.

If we consider the charts of the four principal surveys in the order of the months in
which they were taken, i.e. November 1930-1, December-January 1928-9, January-
February 1929-30, and February-March 1927-8, we see that in the earliest (Fig. 29)
the richest plankton was mostly grouped to the north of South Georgia, while in the
latest (Fig. 24) it was concentrated to the south. In the January-February survey
(Fig. 28) it was very rich almost everywhere, but in the December-January survey
(Fig. 26) the only rich plankton was a patch to the south-east. It seems possible that
there is a tendency for the concentrated plankton to occupy the northern part of the
whaling area in the early part of the season, and later to shift down to the south side,
and that in 1928-9 the shift for some reason took place unusually early, resulting in the
reduction in the average number of organisms per haul in that survey. With the available
data we cannot be certain that this shift takes place : it can only be said that there is
some evidence for it.

Drake Passage and Scotia Sea

Figs. 32-9 are drawn on exactly the same principle as Fig. 23, but they show the
amount of plankton taken at all stations in the Scotia Sea, and each chart shows the
stations taken within a single short period. No certain conclusions can at present be
drawn from the figures shown, but it will be seen that, while at all times there is a scarce
plankton in the vicinity of the South Orkney and South Shetland Islands, the quantity
of plankton varies greatly in other places.

60°

50°

40°

*&iF

^- /

/S?>

/ 2J

^'

-^"^ \ *&/

■ ^'A Tti

55

X 25

\ •
\ •

r 7

55

\ ? ***

60°

•A *>

60

70°

eo"

50°

40° 30°

Fig. 32. Distribution of plankton quantities in the Scotia Sea, November 1929-30, Sts. WS 464-74. Figures
show the number of hundreds of organisms in each sample, shoaling species being omitted. Those in italics
are day hauls and others are night hauls.

Fig. 33. Distribution of plankton quantities in the Scotia Sea, December 1930-1, Sts. 528-41.

Fig. 34. Distribution of plankton quantities in the Scotia Sea, February 1928-9, Sts. WS 374-405.

Fig. 35. Distribution of plankton quantities in the Scotia Sea, February 1930-1, Sts. 613-29.

Fig. 36. Distribution of plankton quantities in the Scotia Sea, March 1929-30, Sts. 365-75 and WS 520-30.

Fig- 37

. Distribution of plankton quantities in the Scotia Sea, March 1930-1, Sts. 631-59 and WS 565-

60°

50°

40°

*•*

*w

/

rf*.«W° ls

/

%rj^\-^y^

.11

D

— l^W^

\

55

ffv \ 39'

55°

60

\ yS7? \

60

\ '''/ .^.SHETLAND

Is.

S.ORKNEY Is.

70° 60°

50°

W

30°

Fig. 38. Distribution of plankton quantities in the Scotia Sea, April 1929-30, Sts. 378-92.

Fig. 39. Distribution of plankton quantities in the Scotia Sea, April 1927-8, Sts. WS 196-205.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 121

These charts are shown primarily as a record for comparison with more recent data
or future material, but one interesting point may be mentioned here. In each of the
lines of stations between Cape Horn and the South Shetland Islands there is an abrupt
change from the rich plankton of the greater part of the Scotia Sea to the thin plankton
of the Shetland neighbourhood. The position of the change is indicated by a pecked
line in Figs. 32, 34, 37 and 38. If these four figures are compared it will be seen that
on the November and February lines (Figs. 32 and 34) the change comes at about 150
miles from the South Shetland Islands, that on the March line (Fig. 37) it is about 100
miles off, and on the April line (Fig. 38) it is only about 30 miles off. From this it seems
possible that there is normally a rich plankton in the central part of the Drake Strait
which spreads farther south towards the end of the summer. Here comparison is made
between different months in different years, so that the apparent shift may possibly be
the effect of a coincidence. The results of future work will no doubt settle the question.

If there is in fact such a southward trend of the plankton it would not of course imply
an actual transport of plankton towards the south. Such a thing would seem impossible
in the present state of our knowledge of the hydrology of these regions. All the evidence
goes to show that the Antarctic surface water in these latitudes moves towards the east
and north, and there seems no possibility of even a local or periodical deflection towards
the south. A more likely explanation might lie in the variation of the amount of upwelling
of the "new" water on the south side of the Drake Passage. It has been found that the
area occupied by this "new" water extends farthest to the north in midsummer and is
most contracted in winter. The boundaries of the " new " water and of the thin plankton
do not appear quite to coincide, but there is a similarity in the changes of position of
these boundaries. A rich plankton has been found in the Bellingshausen Sea in the
neighbourhood of Peter 1st Island, and it is to be supposed that this is carried by the
easterly drift towards the Drake Passage. A reduction in the upwelling of water and a
shrinkage of the area it occupies in the southern part of the Drake Passage might then
make way for the plankton from the Bellingshausen Sea and thus produce the effect of
a southerly trend of the rich plankton.

Little can be said of the variations in the quantity of plankton in the more eastern
parts of the Scotia Sea. Figs. 32-9 suggest that the plankton may perhaps be a little
more patchy in the later than in the earlier part of the summer.

DISTRIBUTION OF WARM- AND COLD-WATER SPECIES
It has been shown that among the Antarctic macroplankton species there are many
which prefer, or are confined to, either the warmer or the colder water, but just as there
are alterations in the position of the rich and poor plankton, so there are changes in the
distribution of the warm- and cold-water species.

SOUTH GEORGIA
Table IV shows, for each species, the average number of specimens per haul for each
group of stations off South Georgia. These are the same groups as those listed on p. 1 1 5,

DIX

I22 DISCOVERY REPORTS

but the single stations are omitted because they give no reliable indication of the
presence or absence of the less common species. The species are listed in order of their
preference for warm or cold water, but those which come under the heading of " Wide-
spread Species" in Table III, p. 107, are omitted. The groups of stations (vertical
columns) are arranged in order of the months in which they were taken.

Table IV. Warm- and cold-water species near South Georgia

South

South

South

!

South

North
side of
South

South
side of
South

South

Georgia

Georgia

Georgia

Georgia

Georgia

survey
1930-1

survey
1928-9

survey
1929-30

survey
1927-8

Georgia
1930-1

Georgia
1928-9

survey
1928-9

Nov.

Dec-
Jan.

Jan-
Feb.

Feb.-
Mar.

March

April

Sept.

Warm-water species

(a) Euphausia vallentini

—

—

—

0-02

—

—

i-33

(b) Eucalanus sp.

51'1

14-4

—

—

—

—

—

Candacia sp.

3I-7

—

4-17

4'00

■ —

—

1-50

Heterorhabdus sp.

571

—

—

17-0

—

—

2-56

Euphausia triacantha

2-14

1-08

52-9

I7-I

12-0

0-67

3-80

C alarms simillimus

IOI

188

372

34°

—

287

267

Pleurornamma robusta

727

119

387

106

1 6-o

21-3

112

(c) Chaetognatha

937

160

547

242

135

266

2-61

Limacina balea

3-95

12-4

63-0

1780

—

116

66-i

Pareuchaeta sp.

31-1

549

4*3

166

213

35-°

2-67

Parathemisto gaudichaudi

18-8

136

597

27-3

91-0

103

3-91

Euphausia frigida

128

208

1200

258

673

233

181

Cold-water species

(g) Euphausia superba

3260

29-8

1 1 60

720

1850

7-60

278

Cleodora sulcata

i-57

1-09

0-05

0-24

—

—

0-19

Salpa fusiformis f. aspera

7-14

500

0-65

6-53

90-0

0-20

i-ii

Tomopteris (large)

1-05

1-67

o-57

0-29

o-6o

—

0-14

Tomopteris (small)

7-28

2-85

0-46

0-46

0-40

—

0-51

Metridia gerlachei

372

IOI

12-5

106

194

8i-i

12-3

Clione antarctica

3-41

3.81

0-16

0-04

—

—

2-14

Limacina helicina

77-3

3!'5

—

0-09

—

—

1-29

Pyrostephos vanhoffeni

0-62

—

0-27

22-4

—

0-29

(h) Sibogita borchgrevinki

0-14

0-02

—

—

—

—

—

Dimophyes arctica

5-09

0-08

—

—

0-02

—

—

Vanadis antarctica

0-09

0-02

0-02

—

o-6o

—

—

Diphyes antarctica

0-37

—

—

—

—

—

Eusirus antarcticus

°-45

—

—

—

—

—

Auricularia antarctica

—

—

—

—

—

Haloptilus ocellatus

0-93

—

—

■ —

—

—

Neritic species

Antarctomysis sp.

54-2

36-1

175

227

375

6-57

It will be seen that the November survey of 1 930-1 differs from the others in that it
has representatives of all species but one of the very coldest groups (h), and it must be

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 123

remembered that these species never occur except in small numbers, and the mere
presence of one of them in a sample is of some significance. The moderately "cold"
species also (group (g)) are on the whole better represented in this survey than in the
others, Limacina helicina and Metridia being specially numerous. Of the warm-water
species Eucalamis, Candacia and the Chaetognatha are also strongly represented, but
Eaphausia triacantha, Calanus simillimus, Pleuromamma, Limacina balea, Pareuchaeta,
Parathemisto, and Euphansia frigida are all in comparatively small numbers. It can be
said in fact that in spite of the prominence of one or two warm-water species, this
November survey was characterized by a much "colder" plankton than any of the
other groups of stations.

During the survey of 1928-9 taken in December and January three of the very coldest
group were present and the moderately "cold" species were well represented, notably
Salpa, Tomopteris and Clione. Of the warm-water species Candacia and Heterorhabdus
were absent, Euphausia triacantha and the Chaetognatha were scarce ; but there was an
increase in the number of Calanus simillimus, Pleuromamma, Limacina balea, Parathemisto
and Euphausia frigida, and Pareuchaeta was more numerous than at any other time.

During the survey of 1929-30 in January and February only one of the coldest group
was taken, a single specimen of Vanadis at St. 336, and the other "cold" species were
very poorly represented. Of the warm-water species Euphausia triacantha, Calanus
simillimus, Pleuromamma, Parathemisto, and Euphausia frigida were more numerous than
at any other time (though it must be remembered that the plankton as a whole was very
abundant during this survey) and the Chaetognatha, Limacina balea, and Pareuchaeta
were also prominent.

The survey of 1927-8 in February and March was also characterized by a warm-water
plankton. None of the ' ' coldest ' ' group was taken, and with the exception of Metridia the
moderately " cold " species were all scarce. Among the warm-water species the presence
of the sub-Antarctic species, Euphausia vallentini, is specially significant and although
the plankton as a whole was not very rich Heterorhabdus and Limacina balea here
reached their maxima and Calanus simillimus and Euphausia frigida were both relatively
numerous.

At the stations taken in March 1 930-1 the plankton population was again of a colder
type, in which two of the very cold-water species are present and Pyrostephos, Metridia
and Salpa are strongly represented. There is support for the "warm" group, however,
in Pareuchaeta, Parathemisto and Euphausia frigida, whose numbers are large in pro-
portion to the total amount of plankton.

At the stations taken in April 1928-9, hardly any colder water species were re-
presented. Of the warm-water species Eucalanus, Candacia and Heterorhabdus were
absent, but Calanus simillimus, Limacina balea and Parathemisto were present in quite

large numbers.

Finally on the winter survey of September 1928, although the temperature of the
water was below o° C. at all stations and colder than at any other group of stations
around South Georgia, none of the " coldest " species was taken and all the warm-water

i24 DISCOVERY REPORTS

species except Eucalanus were present. Euphousia vallentini indeed was better re-
presented than during the 1927-8 survey.

The conclusion to be derived from these facts is that, if the stations taken in March
1 930-1 are excepted, the order of the groups of stations according to the months in
which they were taken is also the order of "coldness " of the plankton population. Thus
the November survey had much the "coldest" plankton. Next comes the December-
January survey with still a distinct cold-water element, then the January-February
survey with a warm-water plankton, and then the February-March survey with a
similar plankton but with the sub-Antarctic E. vallentini and no Vanadis. There is no
doubt about the order of " coldness " of these four surveys, and although we are dealing
with four successive years it is highly probable that as the summer advances the South
Georgia plankton becomes "warmer and warmer". The April stations and the Sep-
tember survey reveal perhaps the "warmest" plankton of all. The cold-water species
taken in March 1 930-1 can be explained by the fact that that season was an exceptionally
cold one in which the pack-ice remained for a long time far north of its usual limits.
Whether the November plankton around South Georgia normally has quite such a
strong element of cold-water species as it did in 1930-1 is doubtful, but at all events
the evidence leaves no reasonable doubt: (i) that in the South Georgia plankton the
cold-water species are most strongly represented in the spring, and that as the summer
advances they are reduced and the warm-water species gain ground, and that the warm-
water plankton continues right through the winter ; (ii) that an abnormally cold summer
results in a "colder" plankton which however still becomes "warmer" as the summer
passes.

There is evidence to show that changes of the same kind take place in other parts of
the Antarctic water, and it will be convenient first to take the eastern part of the area
covered by stations taken in 1927-31.

THE SOUTH SANDWICH AND WEDDELL SEA REGION

Table V shows, for various groups of stations in this region, the average number per
haul of the warm- and cold-water species. The species grouped under (d), (e), and (/) in
Table III, p. 107, and the neritic species Antarctomysis and Eiiphansia crystallorophias,
are omitted.

In the 1927-8 season no cruise was made to the east or south-east of South Georgia,
but in September and October 1928-9 the ' William Scoresby ' took a number of stations
along the edge of the pack-ice which lay about 100 miles to the east of the island (see
Fig. 11). Two of the stations (WS 287 and 288) were taken late in September before
the winter survey of the South Georgia area had been finished. The ship then returned
to South Georgia, completed the survey and then carried out the other ice-edge stations
early in October (WS 298-310). We have seen that during this winter survey the
plankton was of a clearly warm-water type (see Table IV, September 1928-9, p. 122).
The short journey to the ice, however, brought the ship into an entirely different
plankton. Column 1 in Table V shows that there was here a typically cold-water

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

125

plankton in which Sibogita, Dimophyes, Diphyes and Aiiricalaria were all present, while
the warm-water species were much reduced. The state of affairs is illustrated in Fig. 40.

Table V. Warm- and cold-water species in the South Sandwich
and Weddell regions

Season ...

1928-9

1929-30

1930-1

Month

Sept-
Oct.

Feb.

Mar.

Mar.

Dec.

Jan-
Feb.

Jan-
Feb.

Feb.

Feb.

WS

WS

287-8

536-43

WS

Station ...

and

WS

298-310

360-2

365-9

372-3

528-33

and

WS

555-6i

544-51

618-23

626—9

Number of samples ...

15

3

2

2

5

10

7

6

4

Column number

1

2

3

4

5

6

7

8

9

Warm-water species

(a) Euphausia vallentini

—

—

—

—

—

—

—

—

—

(b) Eucalanus sp.

—

—

—

—

—

—

—

—

—

Candacia sp.

—

—

—

25-0

—

0-71

—

—

—

Heterorhabdus sp.

1 00

200

—

—

—

—

—

—

—

Euphausia triacanlha

12-0

0-50

—

—

O-20

—

—

—

—

Calanus simillitnus

—

—

—

i°5

0-71

—

—

—

Pleuromamma robusta

—

200

—

—

—

239

—

7-00

5 00

(c) Chaetognatha

2-17

376

200

150

271

160

224

107

58-2

Limacina balea

5-83

25-0

—

—

—

393

—

—

3-33

Pareuchaeta sp.

—

800

—

400

240

17-1

55-o

290

250

Parathemisto gaudi-

046

467

—

51-0

9 00

28-3

—

59-6

14-2

chaudi

Euphausia frigida

30-0

1800

3'33

627

39-8

379

—

34'5

65-5

Cold-water species

(g) Euphausia superba

1580

4'33

1770

710

4-80

15.545

17-9

1243

35i8

Cleodora sulcata

85-0

i-oo

4-SO

7-50

9 00

246

975

5-00

4-00

Salpa fusiformis aspera

040

6 00

i-oo

4-5°

690

1-71

867

28-0

—

Tomopteris sp. (large)

o-S3

°-33

067

0-50

2 00

2'00

t-43

1-16

125

Tomopteris sp. (small)

6-53

—

0-50

—

6-8o

5IO

123

0-14

23-3

Metridia gerlachei

650

25-0

—

1652

I020

890

348

"55

Clione antarctica

i-33

—

—

—

2-60

6-io

i-oo

3-50

Limacina helicina

7-60

i-33

105

—

40-2

15-1

47-3

II-O

Pyrostephos vanhoffeni

—

—

22-0

—

46-8

218

12-0

—

(h) Sibogita borchgrevinhi

013

—

—

0-40

o-io

o-57

—

Dimophyes arctica

4-33

—

7-00

—

4-60

28-3

26-3

Vanadis antarctica

—

o-33

—

—

040

020

o-57

Diphyes antarctica

2-93

—

I-OO

—

360

6-oo

10-4

"

7-00

Eusirus antarcticus

—

—

0-50

—

i-6o

1-44

—

Auricularia antarctica

°-33

—

—

—

0-20

I-OO

28-3

—

Haloptilus ocellatus

—

—

—

_

2-00

306

In the 1929-30 season the ' Discovery II ' visited the South Sandwich Islands during
the end of February and the beginning of March, just after the South Georgia survey. We
have seen that this survey revealed a warm-water plankton, and at the first three stations
on the way to the South Sandwich Islands (Sts. 360-2, Fig. 12) the warm- water species
were still predominant (see Fig. 41 and column 2 in Table V). At St. 365, however,
Pyrostephos, Dimophyes and Eusirus make their appearance, and at St. 369 Pyrostephos
and Diphyes antarctica were taken. At St. 368, the only other N 100 B station taken at
this time, the sample could not be properly analysed owing to a large catch of Euphausia

40° 30°

55

0

55

1 o* !

1 •

1 ♦

1 *

OCTOBER 1928-9

1 *

WS STATIONS 2878c288
I & E98to3I0

1 1 '

40° 30°

Fig. 40. Distribution of warm- and cold-water plankton between South Georgia
and the South Sandwich Islands. The edge of the pack-ice is indicated.

30°

"\

WARM

PLANKTON
360 3GI|3e2

WARM PLANKTON
373 372

FEBRUARY- MARCH 1929-30

40°

30°

20°

Fig. 41. Distribution of warm- and cold-water plankton between South Georgia
and the South Sandwich Islands. No pack-ice was seen.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

127

superba. Column 3 in Table V, however, shows how greatly the warm-water species
were reduced at Sts. 365 and 369. On leaving the South Sandwich Islands the ship
started a line of stations to the west (Sts. 372, 373, etc.) and here we see the plankton
was again of the warm type. Column 4 in the table shows a very striking contrast
between Sts. 365 and 369, and Sts. 372 and 373. Fig. 41 shows the relative positions
of the "warm" and "cold" plankton.

In the 1 930-1 season a number of stations were taken to the east and south-east of
South Georgia (see Fig. 14, p. 82). In October the ' Discovery II ' sailed from Capetown
to South Georgia via Bouvet Island and the journey from Bouvet Island to South
Georgia was mainly along or through the outskirts of the pack-ice (see Fig. 42). As these
stations are in the form of an extended line they are not separated into groups and shown
in Table V. The stations at the east end of this line had not perhaps a very "cold"
plankton, but Diphyes antarctica was present at all of them, Dimophyes was taken at
St. 453, and Haloptilus ocellatus appeared at St. 460. At Sts. 462-9 the plankton was of

40

Fig. 42. Distribution of warm- and cold-water plankton between South Georgia and Bouvet Island. The

edge of the pack-ice is indicated.

a warmer type. There was a single specimen of Eusirus at St. 466 and of Auricularia
at St. 469, but at all of these stations there were large numbers of Limacina balea,
reaching a maximum at St. 466, and comparatively abundant Euphausia frigida (both
warm- water species). At St. 470 there were three specimens of Eusirus and at Sts. 471
and 472 Diphyes antarctica reappeared and the numbers of L. balea and E. frigida became
suddenly reduced. When the ship reached South Georgia the November survey was
begun, and as already noted revealed here a very cold-water plankton. It seems there-
fore that in the early summer of 1 930-1 there was a cold-water plankton in the neigh-
bourhood of Bouvet Island and South Georgia, but between the two a rather " warmer "
plankton.

When the ship reached South Georgia the pack-ice was lying close up to the island,
but at the end of November, when the survey was finished, the ice-edge had receded
some way to the south-east. In December the 'Discovery II' sailed in this direction,
and, on meeting the ice, followed its edge in a south-westerly direction, taking Sts. 528-
33 in the positions shown in Fig. 43. Here the constitution of the plankton was still
very " cold " (see column 5 in Table V, p. 1 25), even more so than around South Georgia

128

DISCOVERY REPORTS

in November. Of the warmest group only Euphausia triacantha was represented, and
all the coldest species were present except Haloptilus ocellatus.

Late in January the 'William Scoresby' also sailed south-eastwards from South
Georgia and found that the ice had retreated to the northern end of the Sandwich group.
This ice was skirted to the eastward and was found to fall away to the south and dis-
appear. The ship steamed southwards for some 600 miles in open water and finally
reached more ice in the eastern part of the Weddell Sea (Fig. 44). At the stations taken
off the ice around the South Sandwich Islands the same "cold" plankton was taken,
though some warm-water species, particularly Euphausia frigida, were well represented
(see Table V, column 6). Farther south (WS 544-51) the cold-water species increased

40°

30°

**w "C0LD" '

55*

Hk PLANKTON

55

GO

533^

^ 531A

X 532 ,X°

60°

DECEMBER 1930-31/

40° 30°

Fig. 43. Distribution of warm- and cold-water plankton in the eastern part of the Scotia Sea.

The edge of the pack-ice is indicated.

and the warm-water species almost vanished (see Table V, column 7). Haloptilus
ocellatus was taken here in exceptional numbers. These last stations were taken in an
area which is probably covered with pack-ice during the greater part of the year.

In February the 'Discovery II' returned to the South Sandwich region, working
stations along the ice-edge between the South Orkney Islands and the Sandwich group.
These stations (618-29) are shown in Fig. 45. The ice here had retreated very little since
December (Fig. 43), but the plankton had changed to a much warmer type. None of
the very " cold " group was present but such warm- water species as Pareuchaeta, Pleuro-
mamma, and Euphausia frigida were included in the catches. As the ship approached the
South Sandwich Islands, however, signs of a "colder" plankton appeared. Vanadis
occurred at St. 624, and Diphyes antarctica at Sts. 625, 626, 628 and 629. At the same
time smaller numbers of Pareuchaeta and Pleuromamma were taken (see Fig. 45 and

40°

30

^ou0" PLANKTON
W553G. WS537 "CO/ J
. O.VVS538

JANUARY- FEBRUARY
930-1

WS559' WS539 / %■

7/W554l^>
WS542'Z'

WS555

WS543

•WS544-
•WS545

WS547
VERY

"COLD'/__ WS548

30°

60

Fig. 44. Distribution of warm- and cold-water plankton near the South Sandwich Islands and in the eastern
part of the Weddell Sea. The observed and conjectured positions of the ice edge are indicated.

40

30°

40

Fig. 45. Distribution of warm- and cold-water plankton in the eastern part of the Scotia Sea. The edge of

the pack-ice is indicated.

9

i3o DISCOVERY REPORTS

Table V, column 8). The plankton here, however, although it contained an element of
cold-water species, was not nearly so "cold" as it was in the same region earlier in the
season. The conditions were in fact very similar to those which obtained about February
and March 1929-30 (Fig. 41).

It has been seen that the evidence strongly suggests that around South Georgia the
cold-water species are prominent in spring, but become reduced as the summer ad-
vances, and we now have clear evidence that in the 1 930-1 season, in the area roughly
between the South Sandwich and South Orkney Islands, a "cold" plankton in
December gave way to a comparatively "warm" plankton in February (compare Figs.
43 and 45). There is every reason to suppose that this is a normal process. 1 930-1 is
the only single season in which the plankton distribution can be compared in several
different months, but the conditions in October 1928-9 and February-March 1929-30
fall into place very well. The 1928-9 season was not so cold as the 1 930-1 season, the
ice did not reach quite so far north in October, and the area of the South Georgia survey
was not invaded by the "cold" plankton, which, however, came very near to it. The
1929-30 season was a mild one and there was no sign of ice round the South Sandwich
Islands, but the conditions in February-March were very similar to those in February
1 930-1, except that the cold-water species had retreated farther to the south-east.

It is difficult to decide exactly what connection exists between the pack-ice and the
presence of the very cold-water species (group (h) in Table III, p. 107). They are rarely
if ever found except close to the ice or in places in which there has recently been ice.
On the other hand the presence of ice does not necessarily entail the presence of cold-
water plankton.

Fig. 46 shows the changes in the position of the pack-ice during the 1930-1 season,
and its tendency to hang around the South Orkney and South Sandwich Islands while
farther to the east the sea becomes clear of pack-ice for hundreds of miles to the south-
ward. In Fig. 47 the boundaries between the warm- and cold-water plankton shown in
Figs. 40, 41, 42 and 45 are superimposed on one chart. This shows again how the cold-
water plankton retreats to the south-east of South Georgia. It will be seen that west-
ward from the South Sandwich Islands a warmer plankton is always found, and there
is evidence of a tendency towards a warmer plankton also to the east of the islands (see
Fig. 42). It seems in fact that around the South Sandwich Islands there is a tongue of
cold-water plankton reaching up from the south.

The cyclonic circulation in the Weddell Sea, and the cold water which flows out to-
wards the South Orkney and South Sandwich Islands is responsible for the persistent
ice and the cold-water plankton in this region, and it is possible that the ridge of the
South Sandwich chain temporarily deflects the current towards the north, carrying the
cold-water species up in the tongue mentioned above.

The presence of cold-water species near Bouvet Island suggests that there might be
another cyclonic system farther to the east, and Wiist (1928) has published a chart of
current systems in the Atlantic which does in fact show such a system. Its centre is
given as about 6o° S, 300 E, and its orbit just embraces the stations near Bouvet Island

Fig. 46. Changes in the position of the ice edge during the season 1930-1. Note the recession of the ice
as the season advances. The position of the ice in October 1928-9 is shown for comparison.

40°

30°

20°

s4

0

OCTOBERS — -

55

f& $

1930-1

;° y$

1 Q- 'a
OCTOBER <'j

5i'P '

1928-9 |o

t^lQ.

r , u

< lO

IUJ

£.fc| •

1

1

1

! '—1

i.U *
'UJ

'5 *

,'o *

ICO

1 i

FEB-MAR

\ — FEBRUARY

1929-30

0

\ 1930-1

eo

60

'%**

40° 30°

20°

Fig. 47. Relative positions of warm- and cold-water plankton at different times in the vicinity of South

Georgia and the South Sandwich Islands.

9-2

132

DISCOVERY REPORTS

at which the cold species were found. I understand, however, that in the light of more
recent work by the ' Discovery II ' it is to be doubted whether there is actually a cyclonic
movement here in any way comparable in importance to that of the Weddell Sea.

THE ORKNEY-SHETLAND REGION

Under this heading are included all the stations taken in the area of scarce plankton
around the South Orkney and South Shetland Islands and the lines of intensive stations
in the Bransfield Strait. In Table VI the separate groups of stations are arranged in
order of the months in which they were taken, as in Table IV on p. 122, but the
Bransfield Strait stations are dealt with separately. Many of the warm-water species
(groups (a) and (b) in Table III, p. 107) do not occur at all in this region and are there-
fore omitted from the table.

Table VI. Warm- and cold-water species in the Orkney-Shetland region

Bransfield Strait

South Orkneys to South Shetlands

WS

WS

WS

WS

WS

Station numbers ...

476-93

542-55

607-12

382-99

474

534-41

613-15

380,381

637-44

202

Number of samples

18

14

4

18

1

7

3

2

8

I

Year

1929-30

1930-1

1930-1

1928-9

1929-30

1 930-1

1930-1

1928-9

1 930- 1

I927-8

Month

Nov.

Dec.

Feb.

Feb.

Nov.

Dec.

Feb.

Feb.

Mar.

April

Column number ...

1

2

3

4

5

6

7

8

9

10

Warm-water species

(c) Chaetognatha

7'33

4-86

—

0-33

—

100

21-0

0-50

—

—

Limacina balea

i-io

—

—

—

—

—

—

—

—

—

Pareucliaeta sp.

—

f20

—

—

—

1-25

2-33

—

56-0

120

Parathemisto gaudichaudi

—

0-07

31-0

15-0

—

o-43

5-67

3-5°

u-4

35°

Euphausia frigida

—

2-20

32-0

277

54'°

14-2

23-0

16-0

14-0

18-0

Cold-water species

(g) Euphausia superba

170

I2'2

191

429

—

964

2-00

71-0

406

228

Cleodora sulcata

—

0-20

—

—

—

0-50

—

—

—

Salpa fusiformis f. aspera

—

o-8o

i-oo

—

—

420

191

437

177

3-oo

Tomopteris sp. (large)

—

—

—

8-oo

—

—

—

I -00

Tumopteris sp. (small)

0-17

1-07

0-25

—

—

o-43

—

029

—

Metridia gerlachei

4'7i

I4-2

—

13S

—

50-2

155

"•5

1580

3344

Clione antarctica

0-70

o-57

1 00

I-OO

0-40

°'33

—

—

—

Limacina helicina

578

3'43

—

0-I2

6-oo

7'29

0-50

0-14

—

Pyrostephos vanhoffeni

—

—

—

—

—

—

—

—

—

(h) Sibogita borchgrevinki

—

0-07

—

—

—

0-14

—

—

—

—

Dimophyes arctica

078

0-21

—

—

—

1-29

1-33

—

—

—

Vanadis antarctica

—

0-25

—

—

—

—

0-I2

—

Diphycs antarctica

0-28

0-29

—

—

—

o-43

067

—

0-12

—

Eusirus antarcticus

—

0-21

—

—

—

0-25

—

—

—

Auricularia antarctica

0-17

086

0-25

i-oo

—

°'33

—

0-14

—

Haloptilus ocellatus

—

—

—

i'43

3 00

—

—

—

The figures here emphasize the relative "coldness" of the 1 930-1 plankton rather
more clearly than the reduction of cold-water species as the summer goes on. However,
if the 1 930-1 season is considered separately it will be seen that in the Bransfield Strait
the cold-water species were much more strongly represented in December than in
February, and that in the latter month, although the Chaetognatha and Pareuchaeta

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 133

were not taken, Parathemisto and Euphausia frigida had increased very greatly (see
columns 2 and 3). The surveys in the Bransfield Strait in 1929-30 and 1928-9 show a
marked contrast. At the former, which was in November, plenty of " cold " species were
taken, and Limacina balea and some Chaetognatha were the only warm-water species.
At the latter, which was in February, none of the very "cold" species was present and
only one or two of the moderately "cold", while Parathemisto and Euphausia frigida
appeared in quite large numbers for this locality (see columns 1 and 4). It may be
mentioned here that the Chaetognatha, as a group, are not to be relied on as a warm
element in the plankton, especially in such places as this where Eukrohnia hamata is not
necessarily the dominant species.

In the region between the South Shetlands and the South Orkneys there was a large
element of cold-water species in December 1 930-1. In February in the same season
there was little, if any, diminution in the very cold-water species, but there were fewer
of the moderately "cold" group and more of the warm-water species. In March there
were definitely fewer of the "coldest " species, and the warm-water species on the whole
were stronger than in February (see columns 6, 7 and 9). Single stations are unreliable,
but the one taken in November 1929-30 (column 5) suggests a " colder" plankton than
was found in February 1928-9 or April 1927-8.

Table VI provides an interesting example of the unusual coldness of the 1 930-1
season. The catches in February of this season can be compared with those of February
1928-9 both in the Bransfield Strait and farther east (columns 3, 4, and 7, 8). It will
be seen that a much "colder" plankton was present in 1930-1 than in 1928-9.

THE BELLINGSHAUSEN SEA

The Bellingshausen Sea has been visited only twice, each time in the middle of the
summer, so that we have not the material for a comparison of the conditions in different
months, but there is some interest in a comparison between the plankton taken by the
' William Scoresby ' in 1929-30 and the ' Discovery II ' in 1930-1. Table VII shows the
average number of warm- and cold-water species taken during these two cruises. It has
been seen in a previous section that a much richer plankton was taken in the more
westerly part of the Bellingshausen Sea than in the eastern part, and these further
stations are therefore treated separately in the table.

The principal conclusion to be drawn from these figures is that, at any rate in the
western Bellingshausen Sea, the plankton of 1930-1 was not of a colder type than that
of 1929-30, although everywhere to the west of the South Shetlands exceptionally cold
conditions were met with in the former season, while in the latter season the conditions
were unusually mild. The dates on which the 'William Scoresby' took observations
in the western Bellingshausen Sea in 1929-30 were about three weeks later than those
of the corresponding stations of the ' Discovery II ' in 1930-1, and yet at the former the
very " cold " species were on the whole more numerous than at the latter. It is true that
in the eastern part of the Bellingshausen Sea a slightly "colder" plankton was taken in
1 930- 1 than in 1929-30, but the difference is not nearly so great here as it was for

134

DISCOVERY REPORTS

instance in the region between the South Shetland and South Sandwich Islands. The
figures, in fact, seem to suggest that if in any season exceptionally cold or mild con-
ditions are experienced on the east side of the Drake Passage, the same conditions do not
necessarily prevail on the west side. More evidence is needed, however, before the
point can be definitely settled.

Table VII. Warm- and cold-water species in the Bellingshausen Sea

Area

Eastern Bellingshausen

Western Bellingshausen

Season ...

1929-30

193O-I

1929-30

1930-1

Month

Jan -Feb.

Dec.-Jan.

Jan.

Jan.

WS 496-501

556-62

Station numbers

and
WS 508-17

and
580-603

WS 502-5

563-79

Number of samples

H

29

4

7

Warm-water species

(c) Chaetognatha

H'5

4-35

1223

182

Limacina balea

—

—

—

— ■

Pareuchaeta sp.

18-8

4'54

128

22-2

Parathemisto gaudichaudi

32-8

!'34

073

—

Euphausia frigida

—

075

—

4-00

Cold-water species

(g) Euphausia superba

I2-I

3-86

2-50

6-30

Cleodora sulcata

0-17

0-29

5-00

I-I2

Salpa fusiformis f. aspera

2-00

13-6

—

84-5

Totnopteris sp. (large)

0-07

0-31

4-00

I -00

Tomopteris sp. (small)

0-36

1 -83

45-2

16-6

Metridia gerlachei

150

477

400

7°"5

Clione antarctica

0-14

0-62

10-5

3-87

Limacina helicina

I2-I

7-93

5J-5

89-9

Pyrostephos vanhoffeni

20-2

2-34

55-2

2-94

(h) Sibogita borchgrevinki

—

—

0-25

0-12

Dimophyes arctica

—

0-03

4-00

15-2

Vanadis antarctica

—

0-03

I -00

0-41

Diphyes antarctica

—

0-03

1-50

0-94

Eusirus antarcticus

o-33

o-8i

0-50

I -oo

Auricularia antarctica

0-21

o-34

5-00

°-53

Haloptilus ocellatus

—

o-io

176

2-35

THE NORTHERN ANTARCTIC REGION

This heading refers to the belt of warmer water lying between the Antarctic
convergence and South Georgia and the South Shetlands. Most of the lines of stations
crossing this zone run between the Falkland Islands and South Georgia, some cross the
Drake Passage, and some more stations lie roughly between the South Orkney Islands
and the Falkland Islands. There is sufficient material for an independent monthly com-
parison in two of the seasons, and the four seasons are therefore taken separately in

<2

s

o

8

o
e

<3

,53

>

A

co

M

<*■

Cape
Horn

to

South

Georgia

m

3

3
3

Present

Present
Present

I2'0

607

3

3

U

cU

1

■*

*0

co

CD

cd ao

tH t/J

M Hi

-i-

OO '^ co mo

in n 6 o\ | 6 f^ 6 in I nooo 1 ! 6

m « -d- en ' 0 po

<0
1

co
rO

•"*■

■a "

~ x! >>

.2 ° b -

CO

1 1 S«^|^ Sll"" M i 1 1 h 1 1 1 M 1 1 1 1 1

o

m

1

0\

CI

co

1

OO

/^.
co

tt-

|9

m in

Q<2

(S

r-v 0 m

l I 1 aPI g I ssa s l?° I II I l l 11 1 1 1 1

\£) CO W H m

co
*5

m

ro

ti s -a 5 B

o-Ss^pS

.p c n ^ 2
co>3pq

H

u u •- *^ CO u

c/) 1 1

IT)

CO o 3 »-i

3 ■" o °

O

O O O CO

1 1 1 l1^ co&ctS'pH-1 1 ""M 1 1 1 1 1 1 1 1 1 II

O W M rv r\ _ J -.
M CO

>'

o
Z

CO

-1-

m

s o -gi2

3 *" ,2 «

"B tO jr

c^

rj txOcO O OOfO O

. ^ « - , ^ O fO CO O . . , O O O fO -* « CJ

O^J| |oco in 0 m « mwoo^-moo o

•-• u w "+ CO CJ w

>

o

ci

in

p-.2 ■«

— :- o rt

o o*p-;

rn ^ BJ

00

0 0 0000

. ? , 9 t 9 ? pptp « I y* y> 1 ? r 1 1 1 1 1 1 1 1

f hoo» * mo onh m \o 0 00 n in

1 *£) ' M vOyDcOQOOW ' O -*■ ' ' ' ' ' '
CO CI

1

GO

w

o
to <P

■*

3 » B
-j i- O rt

rv

c ° d c c ° °

coy P a> 53 03 0. .0 iiiiiiiiii

t4

-r

-<*■

CO

•O «

S p, « '5i

ro O p-3 I-

3 " o o
(i.

vo

^. |c.„. S „ ^. g^ j,„. g 1 1 „. 1 1 1 [ 1 1 1 1 1 1

U U O COCO Oih u
A Pl, C-l CN CN "^- pL,

CD

(I,

in r

-f

U0

03
U <fl

Q p-rt

m

6 |6co6o\wfl-co6y3 1^-1 ol

11 WOO « a h^O iio''t1-1'1

MM WW

d

V

a

CO

ro

-3 Ih O ™

tJ-

m rs. rv. o

9 C 9 iixP^,?^*riiiiiiii

P.. _,.^. d.n. Nd. ^wn-in- ^- O m ^- m -t r.,
O o\ m u r^i M CO M

M <N p_ CN

60
3
<

CO

CO

3 ,2 S

•a ^o

co

o ooo oo»^- OcotOO

r1 i??!f| 9999 P r> r r S° | V 1 | iiiiiii

m | M N m | \Q M'owcorv. 00|Om|m|

M CO

DO
1

(■>.
rl

a
<

o

o

C-t

^^ 1

i3 1) B
3 B O rt

o o o

'O O0>000 0>00 w^ o

N 6 1 ^t i" 0\W CO M O + N N 1 H | 1 + I

M 1 H OO « -t M C. h M II 1 1 0) 1 1 1 1 1 1 1 1 1 1

J3

H

CO

»

O O co O O
O o O O co O m t? O
i |I|lN*i,ocooi'i300 ..co.cnomSc-i iiiiiii

1 1 1 1 1 M eg £"££ M 1 * g IIIIIII

o

•Si
c3
<U

b

o

B

cj

.a

a

0

d
d
.2

55

in

3
"o

03
1

03
0

d

_o

3

o

fin

(0

s

3
B

a

3
Q

o

1 §

«- ^<fl T;f#, (J Mri trX~ « 42 epi

■3 1 . ~ S 5 ^ « § a ,.8E 8fssg§lgl

At Sl^i 3 % s t 25. ?t:s; => 1 1 ? is 83.5 II

U'^ txj O Hj fa} O A G'-J^^^ 2falOtoKhi^O»^A to Q tx Q fa] ^3 aj

136 DISCOVERY REPORTS

Table VIII. This table shows the average number of warm- and cold-water species per
haul for each line of stations. On some of the lines none of the stations was taken in the
middle of the night, and consequently the diurnal variations of some species prevent a
reliable estimate of their numbers. A query is inserted where a species is absent but
might have been taken if there had been a haul between the appropriate hours at night,
and "Present" is inserted where a few were actually taken but where a larger number
might be expected in a night haul.

In these waters the moderately "cold" species (group (g)) play a part corresponding
to that of the very "cold" group in the high latitudes. As in other regions, the figures
here show a diminution in the proportion of cold-water species as the summer advances.
A glance at columns 1 and 2 shows a much "warmer" plankton in April than in
February 1927-8. In the 1928-9 season the warm- water species were quite well
represented at the August stations, as they were in September off South Georgia, and
the "coldest" plankton is found in December, i.e. early in the summer. In February,
March and May there is a general reduction of cold-water species and an increase in
nearly all the warm-water species. In the 1929-30 season the "coldest" plankton, as
we should expect, is taken in November. The cold-water species mostly disappear in
March and April, though a specimen of Sibogita unaccountably appears in March. Of
the warm- water species Heterorhabdus sp., Euphausia triacantha, Pleuromamma, Pareu-
chaeta, Parathemisto and Euphausia frigida increase towards the end of the season.
Others are irregular and Eucalanus is more plentiful in November. In the 1 930-1
season stations in this area were taken only in March, but it will be seen that the catches
include slightly more cold-water species than those of March 1929-30 and 1928-9.
We should expect this in view of the fact that the 1930-1 season was a specially
cold one.

It is now clear that over the area covered by the investigations, wherever it is possible
to compare observations taken at different times of year, the plankton at a given point
has a larger element of cold-water species at the beginning of the summer (November
and December) than it has later on. It seems also certain that in unusually cold seasons,
such as 1930-1, the cold-water species have a tendency to persist in relatively low
latitudes for a longer period than they would in a normal season. The stations taken in
August and September 1928, suggest that the relatively "warm" plankton, found at the
end of the summer, persists throughout the winter. There was at that time a thin plankton
composed only of the warm-water species in proportions resembling those of the
previous season. It is highly probable that the plankton undergoes little change from
March or April or through the winter, remaining as a slowly diminishing population
until the spring, when there is an invasion of cold-water species and a development of
the rich summer plankton. There seems to be an increase in the proportion of warm-
water species as the summer goes on, but this is not quite so clearly defined as the dis-
appearance of the cold-water species.

The gradual change from a cold- towards a warm-water plankton does not of course
imply a southerly drift of the Antarctic surface water during the summer. It is quite

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 137

probable that the cold-water species generally develop earlier in the summer than the
warm-water species and become replaced by the latter as time goes on.

It is difficult to say what connection may exist between the plankton population and
the movements of the pack-ice. It is generally understood that the ice-edge reaches its
most northerly limit at the end of winter or in the spring. This is also the time at which
the plankton contains the maximum proportion of cold-water species. The distribution of
certain species, also, seems to be limited to regions within the range of the pack-ice, for
the northern limit of Diphyes antarctica, Easirus and some other species typical of the
coldest water coincides, at least approximately, with the northern limit of the pack-ice.

It is worth mentioning, as a matter of separate interest, that a curious red colouring
has been noticed in several species taken off the ice-edge in certain places, particularly
in the eastern part of the Weddell Sea and to the south-west of Bouvet Island. The
catches here included large numbers of Calanus propinquus in most of which the
antennae were of a bright red colour. The red-banded variety of Tomopteris carpenteri
mentioned on p. 99 also occurred in some of these samples, and a small red Amphipod,
at present unidentified, was not uncommon. The red antennae of the copepods often
imparted a striking reddish appearance to the whole sample, especially when combined
with other red species.

PLANKTON COMMUNITIES

Certain contrasts between the plankton of different regions have already been
established. It has been seen for instance that certain groups of species are typical of
the colder and others of the warmer waters. The chart shown in Fig. 21 (p. 105)
indicates that the northern limit of some species and the southern limit of others lie in
a belt running from the South Shetlands to South Georgia which coincides roughly with
the junction of the Bellingshausen and Weddell Sea water. It has further been seen that
in the coastal waters of the South Shetlands and South Orkney region the plankton has
a characteristic which appears to be constant, namely the persistent scarcity of nearly
all the species with which this paper is concerned (see Fig. 23, p. 1 10). On geographical
grounds three main water masses can be distinguished. It will be seen from Fig. 1 that
Graham Land and the South Shetlands divide the Weddell Sea from the Bellings-
hausen Sea in the higher latitudes, and leave a continuous outer belt of Antarctic water
in the lower latitudes, which flows from west to east immediately south of the con-
vergence.

In this section the plankton of the Weddell Sea, the Bellingshausen Sea, the outer belt
and the area of scarce plankton around the South Orkney and South Shetland Islands
are considered separately, but it will be remembered that between the plankton popula-
tions of the Weddell Sea and the outer belt there is a comparatively broad transition zone.
The populations of these four areas are compared and an attempt is made further to
subdivide them according to the nature of the populations. To do this it is necessary to
examine separately all the important lines of stations.

138

DISCOVERY REPORTS

THE WEDDELL SEA
A line of stations (WS 535-61) taken in the 1930-1 season gives a useful indication
of the plankton conditions in the Weddell Sea water east of the South Sandwich Islands
and, farther south, on the eastern side of the Weddell Sea itself (see Fig. 14). Table IX
shows, for each station, the numbers of the ten most important species taken on this
line, together with the total number of all species excluding those which tend to form
shoals. Most of the stations were taken on the outward journey, but some were taken
on the return, which was on very much the same course. The latter stations are
inserted in the table in their proper positions relative to the others.

Table IX. South Georgia to eastern Weddell Sea

Date

January- February 19;

1

North

Station

WS

WS

WS

WS

WS

WS

WS

WS

WS

WS

561

535

560

53°

537

538

559

539

557

541

Hour ...

21

22

21

22

01

13

21

22

08

12

Calanus acutus

39°°

?

>78o

7,400

—

?

18,000

35

3600

200

Rhincalanus gigas

1200

?

>26o

1,200

—

?

—

250

420

2000

Euphaitsia superba

1

190

000

9

21

13°

6000

2

9

1 1

114

Thysanoessa sp.

1200

?

94

200

73°

?

600

2100

1200

139

Calanus propinquus

40

?

34°

1,700

220

?

1,100

5°

180

2000

Euphausia jrigida

63

?

75

1,400

800

?

150

160

?

?

Parathemisto gaudichaudi

63

?

4

—

9

?

130

39

5

5

Metridia gerlachei

—

?

160

1,600

1200

?

4,000

160

So

?

Pleuromamma robusta

—

?

—

—

10

?

—

37

?

?

Haloptihis ocellatus

—

?

—

—

—

?

—

—

—

Total organisms (ex-

74°°

?

2000

13,600

3000

?

24,000

3000

5600

5800

cluding shoaling

species)

Date

Ja

nuary-Fe

aruary 19,

(I

South

Station

WS

WS

WS

WS

WS

WS

WS

WS

WS

WS

542

543

555

544

545

547

548

549

55°

55i

Hour ...

22

H

23

20

14

22

13

22

10

20

Calanus acutus

?

6,300

1300

2000

1200

57°°

3000

3800

4400

2600

Rhincalanus gigas

?

5,400

1 100

280

?

40

440

80

360

460

Euphausia superba

55°°

270

—

17

I

107

—

—

—

—

Thysanoessa sp.

?

120

100

>3&

20

86

36

41

23

15°

Calanus propinquus

?

960

120

1600

440

1 100

1400

520

840

660

Euphausia jrigida

?

45

13

—

?

—

?

—

?

—

Parathemisto gaudichaudi

?

2

—

—

—

—

—

—

—

Metridia gerlachei

?

400

100

2200

3100

360

320

480

440

480

Pleuromamma robusta

?

?

100

?

?

?

—

?

—

Haloptihis ocellatus

?

—

20

80

240

—

520

360

480

460

Total organisms (ex-

?

15,000

34°°

6800

5500

74°°

6100

57°°

7200

57°°

cluding shoaling

species)

Many of these stations were in high latitudes where there was little darkness at night
and where there seems to be little diurnal variation. The total numbers of organisms
are, however, shown above in italics where the hauls were made in daytime between the

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 139

hours of 0600 and 1700, and the figures for species with a marked diurnal variation in
lower latitudes are given in italics where the sample was taken in daytime. Where these
species are absent from such samples a query indicates that they might conceivably
have been present. Queries are also inserted where a sample is swamped by a large
catch of Euphansia super ba.

If the various stations in this table are compared it will be seen that at each of the
seven most southerly stations (WS 544-51) the total number of organisms lies round
about 6000-7000, while at the other stations the totals vary from 2000 to 24,000.
Euphausia frigida occurs at all of the former stations at which it can be expected to be
found, and at none of the latter. Parathemisto occurs at most of the northerly stations
and at none from WS 544 onwards. Euphausia superba appears only in small numbers
south of WS 555 and nowhere south of WS 547. Haloptilus ocellatus is taken only at
WS 555 and at stations farther south, and Pleuromamma does not occur south of WS 555.
The steadiness of the total numbers at WS 544-51 and the wide fluctuations at the more
northerly stations is reflected in the composition of the plankton population. At the
southerly stations each species is present in roughly similar numbers, while farther
north no two adjacent catches are alike. At WS 536 for instance Euphausia frigida and
certain copepods are all plentiful, while at WS 537 the same species, with the exception
of Metridia, are either absent or much reduced. At WS 559 there was an enormous catch
of Calanus acutus and Metridia was relatively abundant, while WS 539 differed from all
the other stations in producing large numbers of Thysanoessa, all other species being
scarce. Of all the samples from this line, from WS 561 to WS 555, the only two which
seem to contain a similar plankton are WS 561 and WS 557, and these are widely
separated from one another. In this line of stations, therefore, we can recognize two
distinct faunistic areas, the more southerly one characterized by a very uniform plankton,
and the more northerly one characterized, at least to the east of the South Sandwich
Islands, by a sharply fluctuating plankton. Each of the two areas also seems to contain
certain species which are absent from the other.

A distinction between these two areas is also found in the phytoplankton. Hart
(1934) gives a chart (Fig. 47, p. 102) on which the total quantities of phytoplankton are
shown for most of the stations included in Table IX. There were no phytoplankton
samples for WS 544 and WS 546, but at the stations near the South Sandwich Islands
the catches were variable and sometimes very large, while at the more southerly stations
the catches were more uniform, and, on the average, considerably smaller. The change
takes place, as with the macroplankton, between WS 543 and WS 545. There is also a
qualitative change in the phytoplankton from the South Sandwich area to the southern
stations, but this is not quite complete at WS 545.

No other samples are available from the eastern Weddell Sea, but Sts. 626-8 (Fig. 14)
were taken about a month later on the east side of the South Sandwich Islands. At St. 627
the sample was swamped with krill, but Sts. 626 and 628 revealed two quite different
types of plankton. Between South Georgia and the South Sandwich Islands we have
first the William Scoresby's stations off the ice-edge in October 1928-9 (WS 287-310,

140

DISCOVERY REPORTS

Fig. 1 1). Among these the plankton did not show very much fluctuation, but they cover
a small area and are not very illuminating. Sts. 360-2 (Fig. 12) are also close together,
but the proportions of the various copepods vary sharply, though some of the other
species are a little more uniform. The samples from Sts. 528 and 530 (Fig. 14) are quite
similar except that about eight times as many Calanus acutus were taken at the former
as at the latter. It seems then that on the occasions on which these regions have been
visited the most variable plankton existed to the east of the Sandwich group, while there
was a slightly less variable plankton between the Sandwich group and South Georgia.

The line of stations taken in October 1930-1 from Bouvet Island to South Georgia
(Sts. 459-72, Fig. 14 and inset B) does not appear to cross any faunistic boundary. It is
curious that although these stations are mostly in the "old" Weddell Sea water, which
has presumably drifted up from the region of fluctuating plankton, the samples at each
of them show a remarkable similarity. There are differences, described on p. 127,
in the occurrence of certain cold-water species, but the distribution of the numerically
important species along this line is very uniform. Thus Calamis propinquus was the
dominant copepod at every station, Euphausia frigida was well represented throughout,
Limacina balea was numerous at most stations, the occurrence of Thysanoessa was
uniform and other species occurred mostly in small numbers. Only Euphausia superba
had an irregular distribution.

The eastern part of the Scotia Sea, that is, the triangle formed by the South
Sandwich Islands, South Georgia and the South Orkneys, also contains water from the
Weddell Sea. Three lines of stations have been worked across this region: Sts. 372-5
in March 1929-30 (Fig. 12), 530-3 in December 1930-1 (Fig. 14), and 618-25 m
February 1930-1 (Fig. 14). The latter two lines followed the edge of the pack-ice. The
numbers of the numerically important species at these three lines of stations are shown
in Table X, which is arranged on the same plan as Table IX on p. 138.

Table X. Eastern Scotia Sea

Date

March 1930

December 1930

February 1931

Open sea

Ice-edge

Ice-edge

Station

West

375

374

373

East
372

West

533

53^

531

East
530

West
618

619

620

621

622

623

624

East
625

Hour

12

17

19

21

21

21

21

21

22

10

21

10

21

10

22

10

Calanus acutus
Rhincalamis gigas
Euphausia superba
Thysanoessa sp.
Calanus propituiuus
Chaetognatha
Euphausia frigida
Metridia gerlachei
Calanus shnillinius
Pareuchaeta sp.

10

?

60

IIOO

no
zyo

?

?

2Q0
?

100

6400

850

Si 00

10

50
?

50
?

20
10

57°
690
860

800
_?

210
10

35°
200
850
480
6400
30
45°

400

810
260

12
100
1 10

59
6100

80

990

120

1

92

248

23

56

220

990

1200

12

99
96

220

18

180

720

3000

8

41
240
940

16
280

40

?
?

54oo
?

?

?

?

?

?

?

29

19

136

33

19
»

?

?

?

1

22
140
920
100
140

64
3

2

1600
?
2

59
?

420
?
?
?

S

64

12

980

1 10

12

5
^6

22
10

3
22
10

1

?
?

400

3°
140

150
3°
15
29

no

640
?

2

33
?

1
1

?

?

?

Total organisms (ex-
cluding shoaling
species)

1 7 00

9200

2600

8500

7700

1800

2900

5400

?

100

920

2500

690

138

1500

970

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 141

The principal conclusion to be drawn from this table is that, while quite a different
type of plankton was taken on each of the three lines, the plankton at the individual
stations of each line was of the same type. In other words, the samples suggest that in
the eastern Scotia Sea the plankton is comparatively uniformly distributed, though its
constitution may differ at different times. Some species in these lines are more evenly
distributed than others. In the first line Calamus acatus is present at each station in
numbers which are very small for this species, and Thysanoessa increases steadily to the
westward. Calanus simillimns appears to do the same, while Pareuchaeta and Euphaiisia
frigida seem to become reduced, though allowance must be made for their diurnal
variations. The numbers of Calanus propinquus are more variable. The plankton re-
vealed by this line is thus not so uniform as it was in the eastern Weddell Sea (WS 544-
51), but it is more so than in the area to the east of the Sandwich group. In the second
line (WS 530-3) the numbers of Calanus acutus are very steady, Rhincalanus and the
Chaetognatha decrease to the westward, Euphaiisia superba, Thysanoessa and Euphausia
frigida are uniformly scarce, and the numbers of Calanus propinquus and Metridia show
no important fluctuations except for the large number of the latter taken at St. 533.
This, however, was something of the nature of a shoal. Thus the plankton taken on this
line was very uniform, and comparable to that of the eastern Weddell Sea. On the third
line also we have quite a uniform plankton. Allowance must be made for the alternate
day and night stations, and it will be seen that, apart from the exceptional catch of
Calanus acutus at St. 621, no important fluctuations occur in the numbers of each species.

THE BELLINGSHAUSEN SEA

In the Bellingshausen Sea two lines of stations have been worked westward from
Adelaide Island to a point beyond Peter 1 st Island. These were the ' Discovery II ' stations
in 1930-1 (Sts. 561-82, Fig. 14) and the 'William Scoresby' stations in 1929-30 (WS
502-8, Fig. 13). On both cruises the stations were taken at or near the ice-edge. There
have also been two shorter lines of stations running north-westwards from Adelaide
Island— that of the 'Discovery II' in 1930-1 (Sts. 583-97, Fig. 14) and that of the
'William Scoresby' in 1929-30 (WS 509-17, Fig. 13). Other stations in the Bellings-
hausen Sea lie up and down the coast of Adelaide Island, the Biscoe Islands and the
Palmer Archipelago.

During the cruise of the 'Discovery II ' to the western Bellingshausen Sea twelve
stations were taken on the outward journey (Sts. 561-72) and ten on the return (Sts.
573-82), and these are arranged in their relative positions on the line in Table XI, as in
Table IX.

Reference has already been made (p. 1 10) to the thin plankton of the eastern Bellings-
hausen Sea (near Adelaide Island, etc.), and this is clearly shown in Table XI at stations
east of St. 579. At the westerly stations the plankton tends to be rich, but there is great
variation in the size of the samples. However, the composition of the plankton is very
similar at these stations in spite of the fluctuations in abundance. Rhincalanus is usually
the dominant copepod, though Calanus acutus is more numerous at one or two stations,

142

DISCOVERY REPORTS

Table XI. Bellingshausen Sea

Date

January 1931

West

Station

572

57i

57°

573

569

574

568

575

576

567

577

Hour

20

09

22

10

1 1

22

22

10

21

10

10

Calamis acutus

560

880

14,000

360

400

36

20

4,600

12

4700

480

Rhincalanus gigas

1700

2100

4,700

4500

IIOO

210

1900

12,000

290

840

9,400

Euphausia superba

1

—

76

2

— .

—

Thysanoessa sp.

32

?

63

17

17

3

1

no

73

no

4

Calamis propinquus

60

?

160

So

8

14

—

80

68

?

?

Chaetognatha

290

I40

210

4S0

1S0

10

100

470

8

230

160

Euphausia frigida

—

6

—

11

?

—

16

?

?

?

Metridia gerlachei

300

60

160

?

40

—

—

?

—

?

?

Salpa fusiformis i. aspera

1

39

—

9

57

56

37°

?

—

550

21)0

Limacina helicina

23

4

73

12

22

120

280

9

36

82

no

Total organisms (ex-

3100

3200

19,000

5400

1800

410

2300

18,000

490

6100

10,000

cluding shoaling

species)

Date

January 193 1

East

Station

566

56s

578

564

563

579

580

5°2

581

582

561

Hour

22

10

22

22

1 1

10

22

22

20

10

10

Calamis acutus

160

760

900

35°

780

IIOO

99

76

360

55

5

Rhincalanus gigas

7200

5600

540

160

1200

470

16

I

13°

35

1

Euphausia superba

—

—

4

—

—

24

3

2

1

2

3

Thysanoessa sp.

22

6

IS

120

34

3S

250

4

96

2

52

Calamis propinquus

—

?

48

96

Q20

16

12

12

18

7

26

Chaetognatha

170

410

100

6

7

?

—

—

—

?

?

Euphausia frigida

16

?

—

—

?

?

—

—

?

?

Metridia gerlachei

—

?

104

—

?

16

2

—

34

2

?

Salpa fusiformis f. aspera

66

4

—

—

?

?

—

—

?

?

Limacina helicina

630

40

H

28

33

10

4

2

—

6

1

Total organisms (ex-

8300

6goo

1700

780

3000

1600

39°

100

650

no

85

cluding shoaling

species)

and it is in fact the variations in the numbers of these two species which are responsible
for the variations in the total numbers of organisms. The other species, which are not
very abundant, occur in quite uniform numbers.

Table XII shows the results of the ' William Scoresby ' line of stations.

WS 508 is in the region of thin plankton in the eastern Bellingshausen Sea, but heavy
catches were taken at the other four and the plankton becomes progressively richer
towards the west. Apart from this increasing abundance the plankton is very similar at
each of these stations. It resembles that of the ' Discovery II ' line in the scarcity of
Euphausians and in the numbers of Rhincalanus and Limacina, but differs from it in
the great abundance of Calanus acutus, and in the larger numbers of Calamis propinquus
and Chaetognatha.

These two lines of stations indicate that the Bellingshausen Sea may be divided, in

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

143

respect of its plankton population, into two regions : the eastern part which contains
a thin plankton, and the western part which normally has a rich plankton. The quantity
of plankton in the western part seems liable to vary very much from place to place, but
its composition seems to be quite uniform.

Table XII. Bellingshausen Sea

Date

January-February

1930

West

East

Station ...

WS503

WS502

WS504

WSSo5

WS508

Hour

r9

13

18

H

02

Calanus acutus

3 1 ,000

21,000

1 3 ,000

13,000

640

Rhincalanus gigas

5,900

2,200

2,000

?

—

Euphausia superba

—

—

—

10

1

Thysanoessa sp.

17

TOO

9

1

20

Calanus propinquus

1,000

510

64

450

170

Chaetognatha

2,600

1,500

709

120

5

Euphausia frigida

—

?

—

?

—

Metridia gerlachei

1,300

260

64

?

48

Salpa fusiformis f . aspera

?

?

?

?

10

Limacina helicina

31

94

81

17

Total organisms (excluding

43,000

26,000

16,000

13,500

910

shoaling species)

Various other stations have been worked in the eastern part of the Bellingshausen
Sea, the most important of which are those taken by the ' Discovery II' in a line running
north-westwards from Adelaide Island (Sts. 583-97, including stations on return
journey). The numbers of the principal species taken are shown in Table XIII.

Table XIII. Off Adelaide Island

Date

January I

931

North-west

South

-east

Station

592

59i

593

590

594

589

595

588

597

587

596

586

585

583

584

Hour

02

20

10

10

16

01

00

19

12

14

05

09

°5

22

01

Calanus acutus

28

8

1

6

12

3

II

4

6

15

5

3

78

66

29

Rhincalanus gigas

280

420

100

1

200

8

200

7

?

4

5

2

27

55

17

Euphausia superba

—

1

—

8

—

—

—

—

1

—

6

—

1

2

1

Thysanoessa sp.

2

2

IOO

2

6

20

5

2

22

1

1

TOO

28

3&

12

Calanus propinquus

12

2

?

1

p

5

13

5

3

1

4

25

55

9

Chaetognatha

37

13

4

?

20

2

7

1

?

5

I

Euphausia frigida

10

—

?

?

?

1

1

—

Parathemisto gaudichaudi

7

8

15

—

3

—

4

1

—

—

1

Metridia gerlachei

8

—

f

?

?

—

2

?

14

7

Limacina helicina

95

40

44

22

49

4

10

7

8

3

13

2

24

Total organisms (excluding

57o

500

2J0

37

330

51

280

21

30

30

22

120

200

220

no

shoaling species)

i44 DISCOVERY REPORTS

It should be mentioned that the stations at the south-eastern end of the line (near
Adelaide Island) were a little closer together than those to the north-west (see Fig. 14,
inset A).

The table shows the same thin plankton that was found at stations east of St. 579 in
Table XI. The poorest plankton of all is found in the middle, there is slightly more at
the inshore end (south-east) and the largest catches at the north-west end where indeed
one might expect to reach a region of richer plankton. The composition of the catches
must be regarded as quite uniform. Only Rhincalanus varies to some extent, but where
there are never more than a few hundred specimens in the samples, such fluctuations
cannot be of much significance.

During her cruise in the Bellingshausen Sea the 'William Scoresby' also worked a
line of stations north-westwards from Adelaide Island (WS 509-17, February 1929-30,
Fig. 13, inset). This was a short line, however, and the stations were close together, so
that it is of less importance here than the 'Discovery II' line. It may be mentioned
that the catches were quite similar in size and composition to those in the Discovery
samples, and the plankton seemed quite uniform along the line except at WS515,
where a surprisingly large catch was taken in which Calonus acutus, C. propinquiis,
Metridia, Parathemisto and Thysanoessa numbered three or four hundred each.

Other hauls taken in the Bellingshausen Sea include only some stations taken along
the coastal region between Adelaide Island and the South Shetlands. There were six
taken by the 'William Scoresby' at the beginning of January 1929-30 (WS 496-501,
Fig. 13), five taken by the 'Discovery II' at the end of December 1930-1 (Sts.
556-60, Fig. 14), and five more by the same ship late in January of the same season
(Sts. 598-603, Fig. 14). The samples taken at the ten stations of the 'Discovery II'
were all very small, averaging about 130 organisms per haul. By far the largest of these
was the catch taken at St. 598 which contained over 600 organisms of which Metridia
formed the majority. At three of the 'William Scoresby ' stations, however (WS 496,
497 and 501), quite large catches were taken, each containing several thousand
specimens of Calanus acirfus. None of the other species was very plentiful, but such
catches are very unusual in these coastal regions.

It seems then that the eastern or coastal region of the Bellingshausen Sea is charac-
terized by a thin plankton of fairly uniform distribution, but that here and there one or
more species may become concentrated, as a result, perhaps, of some local peculiarity
in the hydrological conditions.

THE ORKNEY-SHETLAND REGION

In this section those stations will be briefly considered which lie in the vicinity of
the South Shetlands and South Orkneys and in the Bransfield Strait, a region charac-
terized, as already noted, by a general scarcity of plankton.

In the Bransfield Strait three intensive surveys have been carried out, with lines of
stations at short intervals, and between the South Shetland Islands and the South

DISTRIBUTION OF ANTARCTIC M ACROPLANKTON

145

Orkney Islands various stations have been taken, many of which were at the edge of the
Weddell Sea pack-ice.

The three surveys of the Bransfield Strait were taken in February 1928-9, November
1929-30, and December 1930-1 (Figs. 15-17). It will not be necessary here to tabulate
the numbers of each species taken at the different stations. If the catches of Euphausia
superba are disregarded it may be said that every sample which has been taken from the
Bransfield Strait has been an extremely small one, and if one sample ( WS 393), which
contained the exceptional quantity of 520 Metridia, is also disregarded, the largest catch
from the three surveys contained only 300 organisms. It is known that the south side of
the Bransfield Strait is occupied by Weddell Sea water, and the north side by water from
the Bellingshausen Sea. The distribution of the macroplankton, however, does not
appear to differ very much in different parts of the Strait, and with such small samples,
many of them containing only twenty or thirty organisms, it is very difficult to establish
a reliable correlation between the local hydrological conditions and the distribution of
the macroplankton.

Between the South Shetland Islands and the South Orkney Islands there are no
straight lines of stations, and usually only two or three stations have been taken at a
time in this region, but in the 1930-1 season the 'Discovery II' worked eight con-
secutive stations in December along the ice-edge from near the South Orkneys towards
the Bransfield Strait (Sts. 534-41, Fig. 14), and eight more in roughly the same
position in March (Sts. 637-44, Fig- H)- Trie numbers of the principal species at these
stations are shown in Table XIV.

Table XIV. Shetland-Orkney region

Date

December 1930-1

March 1930-1

Station

West
54i

54°

539

538

537

536

535

East
534

West
644

643

642

641

640

639

East
638 | 637

Hour...

23

20

17

12

06

21

10

21

06

23

18

14

09

23

05

21

Calamis acutus
Rhincalanus gigas
Euphausia superba
Thysanoessa sp.
Calanus propinquus
Chaetognatha
Euphausia frigida
Parathemisto gaudichaudi
Metridia gerlachei
Salpa fusiformis f. aspera

76

2
10
23

6

4

60

15

4

6

4
360

37
25
68
26
3

H
?

1
?
?

60
12

27

508

?

14
?

?
?

?

?

7600

144

?

?
?
?
?
?

1800
40

7
8

3°
100

2
190

79

6

35

4

83
1

7
?

?
?

59

47

5

28

24

5

57

7
89

15

1

20

16

1

1

9

22
?

?

?

1800

49
42
?
220
56
IIOO

260

7
3
2

?
?

5

?

59°
26
?
?
?

?

JO

1
?

23

3

4
?

?

10
19

32

16

810

83
380

1

3
2800

55

1

1

15

19

?

7
1
S
1

1

75
72

41

22

380

300

Total organisms (ex-
cluding shoaling
species)

120

450

no

590

?

2200

130

280

65

1500

26

19

3400

42

590

It will be seen from Fig. 14 that in both these lines the more westerly stations are
closer together than the others.

This is still a definitely thin plankton, but it is richer than in the Bransfield Strait and
much more variable. There are considerable fluctuations both in the total number of

146

DISCOVERY REPORTS

organisms and among the individual species. Apart from Euphausia superba the most
variable species are Metridia gerlachei, Calanus aaitas and Salpa fusiformis.

The other stations in this area are WS 202 (Fig. 10) which was remarkable for a heavy
catch of Metridia (over 3000) and an exceptional number of Parathemisto (350), WS 380
and 381 (Fig. 1 1) at which the plankton was scarce except for a rather large number of
Salps, and Sts. 613-15 (Fig. 14) at which Salps were again quite plentiful and other
species scarce.

It appears, therefore, that the area between the South Orkneys and the South Shet-
lands is characterized, like the eastern part of the Bellingshausen Sea, by a plankton
which is generally scarce, but which may be found here and there in comparatively large
quantities, usually due to the concentration of one particular species.

THE NORTHERN ZONE
The lines of stations so far considered are those in waters which are covered by the
pack-ice during a large part of the year. Those next to be considered are mostly in
waters which are never reached by the ice. Numerous stations have been taken in this
area and we may consider first the lines crossing Drake Passage. There are three of these
lines (Figs. 11,12,14) and they are shown in Table XV, which gives for each station the
numbers of the ten most abundant species in this region.

Table XV. Drake Passage

Date

February 1928-9

April 1929-30

March 1 930-1

Station

South

WS
400

WS
401

WS
402

WS
4°3

North
WS

404

South
378

382

383

384

North
385

08

South
644

646

647

648

North
649

Hour...

10

00

13

01

13

12

01

10

20

06

17

00

13

23

Calanus acutus
Rhincalanus gigas
Limacina balea
Thysanoessa sp.
Calanus propinquus
Chaetognatha
Euphausia frigida
Parathemisto gaudichaudi
Calanus simillimus
Pareiichaeta sp.

?
?

33
?
?
?

8
?

?

17
8

9

92
9

84

20
iyo

?
180

?

16

?

73
2
?

32
3500

3700
320
380

320

35
160

35°

64

5100
10

2200
64
880
?

89
?

32

13

79
?

16
8
?
?

81

?

?

800
5200

5°

1900

660

33

92

400

800
1000

?

220
100
320

?

38
20
10

530

150
44°
37°
230
270

3i°

7500
40
IS
?

6300
1

31
20
?

15
1

16
1

1

9

I
?

I 1

64

?

13
I
?

810
1500

500

160

II

3°

5

160

40

64
1500
?

8
16
850
?

4

16

8

40

33°

430

290

88

1600

7
88

130

Total organisms (ex-
cluding shoaling
species)

43

290

460

9100

8400

200

9200

2500

2400

yooo

65

89

3200

2500

2800

The most northerly station in each of these three lines lies close to the Antarctic
convergence.

The first three stations of the 1928-9 line, the first station of the 1929-30 line and
the first two stations of the 1 930-1 line are in the region of thin plankton, and will be
regarded as belonging to the Orkney-Shetland region. The others are in the normal
plankton of the northern zone.

An inspection of Table XV will show that at WS 403 and WS 404, at Sts. 382-5,

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

i47

and at Sts. 647-9 there is a very uniform plankton. At all these stations Rhincalanus is
the dominant copepod, Limacina is found only near the convergence, the Chaetognatha
for the most part increase steadily towards the north, and Parenchaeta and Eiiphamia
frigida (except at St. 649) are present in moderate numbers at night stations. Calanus
acutus and C. propinquus are a little variable, but Parathemisto has a uniform distribution
in each line, except for a large catch at St. 384. Thysanoessa occurs in large numbers on
the first line, but in smaller numbers on the other two. In general it may be said that
there is a close similarity in the catches at consecutive stations across this part of the
northern zone, and if one or two species, such as Thysanoessa, are excepted, it can be
said that there was little difference between the plankton taken on the three lines.

North-east of Drake Passage there are several lines of stations lying roughly between
the Falkland Islands and the South Shetlands and South Orkneys (Figs. 10, 12, 14).
These stations and the principal species taken at them are shown in Table XVI.

Table

XVI.

Between Falkland Islands and South

Orkneys

Date ...

April

November

March

March

1927-8

1929-30

1929-30

1930-1

South

North

South North

South North

South

North

Station

WS

203

WS
204

WS
474

WS
473

WS
472

WS

47i

WS

470

WS

469

WS

527

WS

528

WS

529

637

636

635

634

633

Hour...

23

01

23

09

21

09

19

20

12

14

15

21

09

23

09

22

Calanus acutus

140

—

—

34°

270

700

530

48

—

—

—

5

140

120

600

Rhincalanus gigas

940

52

1

240

340

1300

1300

660

10

?

?

—

2

s8o

2200

3.300

Limacina balea

48

5400

—

?

16

?

12

—

10

?

?

—

?

—

180

2,800

Thysanoessa sp.

190

—

160

14

33

3

I

230

2J0

yoo

25

75

15

710

520

4.7O0

Calanus propinquus

260

80

IS

?

12

■>

32

790

18

bo

72

■?

790

80

1,500

Chaetognatha

1200

1 1 00

—

160

26

400

660

820

6yo

bo

10

—

?

740

2400

1,400

Euphausia frigida

87

?

54

?

6

?

—

?

?

?

41

?

62

?

—

Parathemisto gaudichaudi

10

25

—

2

2

—

21

200

10

10

210

22

25

13

53

60

Calanus simillimus

16

160

—

?

4

?

16

—

10

?

?

—

?

200

?

680

Parenchaeta sp.

96

92

—

?

?

?

—

?

?

?

—

?

120

40

880

Total organisms (ex-

3000

1600

240

990

75°

2500

2600

2100

1S00

800

3io

59°

.5°

3500

5400

1 5 ,000

cluding shoaling

species)

|

The samples taken at WS 203 and WS 204 resemble one another only in respect of
certain species. Thus the numbers of Chaetognatha, Parathemisto, and Parenchaeta are
similar, but Rhincalanus is the dominant copepod only at WS 203, and Thysanoessa
occurs only at that station. The suggestion of an increase in the number of Limacina
towards the north is in accordance with the conditions found in Drake Passage. The
November line (WS 469-74) shows again the uniform plankton indicated in Drake
Passage. At each station but one Rhincalanus is, as before, the dominant copepod, Calanus
acutus is uniformly distributed in modest numbers and the Chaetognatha steadily in-
crease towards the north. Limacina, Thysanoessa, Calanus propinquus, C. simillimus,
and Parenchaeta are scarce. WS 527-9 give curious results. The hauls were all taken in
the middle of the daytime and consequently are not very reliable, but the plankton is
not of the usual type found in the northern zone. The quantity is small, Calanus

148

DISCOVERY REPORTS

propinquus replaces Rhincalanus as the dominant copepod and the Chaetognatha
diminish towards the north instead of increasing. The distribution of the plankton,
however, seems to be quite uniform. In the 1930-1 line Sts. 637 and 636 are evidently
in the area of thin plankton. At Sts. 635, 634 and 633, however, Rhincalanus is again
dominant, Calanus acutus appears in moderate numbers, Limacina and the Chaeto-
gnatha increase towards the north and Parathemisto is evenly distributed in small
numbers. The other species mostly increase towards the convergence, but the increase
is in proportion to the increase in the total number of organisms, and the plankton here
can be regarded actually as very uniform.

Table XVI]

Between Falkland Islands and South

Georgia

Date

February

August

December

March

1927-8

1928-9

1

928-9

1928-9

NW

SE

NW

SE

NW

SE

NW

SE

Station ...

WS

WS

WS

WS

WS

WS

WS

WS

ws

WS

WS

WS

WS

140

141

142

254

255

°3

256

316

3i5

314

413

414

415

416

Hour

21

17

10

13

01

16

19

13

13

°5

17

16

Calanus acutus

260

1, 5°°

3

—

—

16

no

130

68

—

290

670

2800

Rhincalanus gigas

400

7,5°°

12

?

I

160

2200

2300

260

32

2700

640

IQOO

Limacina balea

300

160,000

300

?

26

—

120

—

?

280

53°

48

?

Thysanoessa sp.

180

2,700

52

2

130

560

6

260

2

230

15°

iboo

6

Calanus propinquus

3600

2,800

2

2

3

92

16

96

?

460

—

220

(>4

Chaetognatha

230

1,400

38

I

—

2

220

260

160

1000

44°

130

9

Euphausia frigida

68

1,200

?

?

63

55°

?

24

?

8

2

830

?

Parathemisto gaudichaudi

31

45°

28

2

6

—

36

2700

72

26

42

5°

260

Calanus simillimus

—

?

5

2

—

—

80

190

52

300

64

2bo

?

Pareuchaeta sp.

16

?

?

?

—

4

?

?

?

?

?

?

?

Total organisms (excluding

4900

18,000

150

9

210

15°°

2800

6200

640

2200

37°°

55°°

5000

shoaling species)

April

November

Febr

uary

March

1 928-9

1929-30

192?

-3°

1930-1

NW

SE

NW SE

NW

SE

NW SE

Station

WS

430

WS

429

WS

428

ws

427

WS
466

WS

465

WS

521

WS

522

WS
523

WS
524

ws

525

WS
526

656

657

658

°59

Hour

04

1 1

18

00

22

22

13

°3

14

10

23

14

10

09

°5

13

Calanus acutus

—

32

32

32

1500

224

—

°(

—

2,600

32

6

Rhincalanus gigas

760

3800

670

1100

1000

2500

830

1 90c

> 1200

800

83°

?

13S

?

—

2

Limacina balea

1200

120

40

?

54°

—

22,000

—

72

72

—

?

150

3

—

?

Thysanoessa sp.

250

130

140

16

48

800

330

380c

> 840

150

8,000

150

12

ib

82

34

Calanus propinquus

120

220

350

goo

—

—

f

32c

> 24

?

3,000

100

?

20

52

120

Chaetognatha

730

1100

840

2

1800

1600

790

30c

» 7s

.56

80

4

100

?

68

22

Euphausia frigida

—

?

350

33

—

32

?

1 40c

> ?

?

—

?

?

?

—

?

Parathemisto gaudi-

3

22

II

940

480

160

310

]

92

13°

140

250

8

10

18

12

chaudi

Calanus simillimus

,S6

32

45°

96

5100

35°

?

1 6c

> 64

ib

1,900

32

2

?

—

?

Pareuchaeta sp.

48

?

16

420

16

—

?

45<

> ?

?

—

?

?

?

?

?

Total organisms (ex-

2100

5400

2900

3800

11,000

6300

2,300

870c

> 2300

1200

17,000

630

260

5.5

2}0

190

cluding shoaling

species)

Finally, there are eight lines of stations lying between South Georgia and the Falk-
land Islands. These are shown in Table XVII and their positions in Figs. 10, 1 1, 12, 14.

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 149

Whereas other lines crossing the northern zone have at their southern or south-
eastern ends an area of scarce plankton, most of these have at their south-eastern ends
the rich plankton of the neighbourhood of South Georgia. The first line, however,
(WS 140-2) is an exception, for in February 1927-8 the plankton was unusually thin
around the north and west sides of South Georgia, and between WS 141 and WS 142
we find the familiar abrupt change from a rich to a poor plankton. It must be admitted
that at these three stations the plankton is patchy. Rhincalanus is the dominant
copepod at WS 141 and WS 142, but at WS 140 this species is far exceeded by the
young stages of Calanus propinquus. At WS 141 the net evidently passed through a shoal
of Limacina. Apart from these species, however, the differences between the three
stations lie in the quantity rather than in the constitution of the plankton.

The next line (WS 254-6) was the only winter line across the northern zone, and is
therefore hardly comparable with the others. The plankton of course was very thin,
but had moderate numbers of Euphausians towards South Georgia. The plankton at
WS 254, a midday station, was practically negligible.

At WS 314-16 there is again a rather sharp fall in the quantity of plankton from
WS 315 to WS 314. At all three stations however Rhincalanus dominates the whole
sample except for the exceptional catch of Parathemisto at WS 315. Other species are very
uniform, though there is a slightly disproportionate number of Thysanoessa at WS 315.

At WS 413-16 the Chaetognatha increase steadily towards the convergence and
Calanus acutus towards South Georgia. Parathemisto is fairly steady, but Rhincalanus
is the most plentiful copepod only at WS 414. At WS 415 it is equalled, and at WS 416
it is exceeded by Calanus acutus. Euphausians are inclined to be irregular.

In the line WS 427-30 Rhincalanus is the principal copepod at all four stations.
Limacina increases towards the convergence, and the Chaetognatha tend to do the same,
though they fall off at WS 430. Calanus propinquus and Thysanoessa are quite evenly
distributed, though the former increases a little towards South Georgia.

WS 465 is quite typical of the northern zone with its abundant Rhincalanus and
Chaetognatha, but WS 466, which is almost on the convergence, has unexpectedly large
numbers of Calanus acutus and C. simillimus.

In the line WS 521-6 there are more stations between the convergence and South
Georgia than in any other. The quantity of plankton varies considerably, but this is
largely due to the diurnal variations of Thysanoessa. At WS 522 and WS 525 the samples
were taken near the middle of the night and at the other stations near the middle of the
day, and at the former Thysatioessa was enormously abundant and at the latter relatively
scarce. At WS 521-4 Rhincalanus was much the most abundant copepod, but at WS 525
and WS 526 (nearer to South Georgia) it was replaced by Calanus acutus and C. pro-
pinquus. C. simillimus was evenly distributed except at WS 525, and Parathemisto except
at WS 522. There was the usual increase of Chaetognatha towards the convergence and
a shoal of Limacina at WS 521.

The last line (Sts. 656-9) differs from nearly all other lines across the northern zone
in the small amount of plankton taken at each station. Parathemisto and Thysanoessa

i5o DISCOVERY REPORTS

are very uniform, and the Chaetognatha increase towards the convergence, but the
commonest copepod at Sts. 657, 658 and 659 is Calanus propinqnus, Rhincalanus being
dominant only at St. 656. Apart from this, however, the plankton is similar in its make-
up to the plankton found on other lines in this area.

Enough has been said now to show that the northern zone contains, during the
summer, a plankton population sufficiently characteristic to form the basis of a separate
faunistic area. The most important features of this plankton may be summed up as
follows.

It is a population of homogeneously distributed typically warm-water species.
Rhincalanus is generally the most abundant species and almost always has a large
majority over other copepods. The Chaetognatha are nearly always present in large
numbers near the convergence and in steadily decreasing numbers to the south or
south-east of it. Calanus simillimus, Pareuchaeta sp. and Euphausia frigida are nearly
always present in small or moderate numbers in the night hauls. Owing to their marked
diurnal variations they are rarely taken in day hauls, but it is evident that they are
usually quite evenly distributed over the whole area. The occurrence of Calanus pro-
pinquns and Thysanoessa is a little irregular, but if they occur in large numbers at one
station in a line across the northern zone, it is probable that they will be taken also in
large numbers at other stations in that line.

In general, it can be said that, in the northern zone, the summer plankton is both
uniform and stable. That is to say that, compared with other parts of the area in-
vestigated, excepting perhaps the eastern Weddell Sea, a uniform type of plankton
community is to be found everywhere at a given time, and that the plankton does not
alter very much from month to month or from year to year. The northern zone seems
to be a little more uniform and stable in the Drake Passage than it is between the Falk-
land Islands and South Georgia. In the latter region, however, there is known to be an
eddy in the currents near the convergence, and this might be expected to cause some
irregularities in the distribution of the plankton.

FAUNISTIC DIVISIONS

We are now in a position to map out the Weddell sector of the Antarctic in respect
of the different types of plankton which have been found there, and it is possible, in
my opinion, to distinguish seven separate divisions. These are shown in Fig. 48 and they
may be defined as follows.

(1 ) The northern zone. This zone is bounded to the north by the Antarctic convergence,
and to the south by the region of scarce plankton in the vicinity of the South Shetlands,
and by the beginning of the transition zone in the vicinity of South Georgia. It is
characterized by a uniform and stable plankton, in which Rhincalanus gigas is generally
predominant ; the Chaetognatha and Limacina balea are plentiful, especially near the
convergence, and the typically warm-water species are fairly evenly distributed.

(2) The Graham Land area. This includes the whole of the region of scarce plankton

DISTRIBUTION OF ANTARCTIC MACROPLANKTON

151

in the neighbourhood of the South Shetlands and South Orkneys, the Bransfield Strait
and eastern part of the Bellingshausen Sea. The Shetland-Orkney region and the
eastern Bellingshausen Sea can be regarded as subdivisions. It is characterized normally
by a very thin plankton in which typically cold-water species are well represented. In
some places there may be considerable irregularity in the quantity of plankton. The

30°

Fig. 48. Chart showing the provisional boundaries of areas in which distinctive plankton communities have

been found.

limits of this area are ill-defined near the South Orkneys, and uncertain on the west
side of Drake Passage.

(3) The transition belt. Derived from Fig. 21, p. 105. This is the belt in which the
normal southern limits of the warmest water species and the northern limits of the coldest
water species are found. One would expect to find the cold-water species prominent
here in the early summer and the warm-water species later. This zone largely encloses
the line separating the waters derived from the Bellingshausen and Weddell Seas, and
may be taken to include South Georgia and the adjacent whaling grounds. Strictly

152 DISCOVERY REPORTS

speaking it overlaps a part of the Graham Land area, which is defined on a different
basis, for there is probably something of a transition zone everywhere to the south of
the northern zone.

(4) The eastern Scotia Sea. A rough triangle, having at its corners the South Sand-
wich Islands, the South Orkney Islands and South Georgia. The area is characterized
by a plankton which appears to be uniform but unstable. That is to say, a different type
of plankton has been taken on the different lines of stations crossing this area, but the
plankton was similar at consecutive stations in each line. The composition of the plankton
seems to depend on the time of year, and very likely varies from one year to another.

(5) The "old" Weddell water. It is difficult at present to give any general definition
of the plankton in this area, but it includes the unstable and heterogeneous plankton
population found on the east side of the South Sandwich Islands. Its southern limit,
at the time it was determined, lay between WS 544 and WS 555 (lat. 6o° 40' S, see p. 69),
and north of this it may be temporarily defined as the area occupied by water which has
flowed out of the Weddell Sea and passed the Scotia Arc in the vicinity of the South
Sandwich Islands.

(6) The eastern Weddell Sea. The limits of this area are not known, but it lies south
of WS 555, and, so far as can be judged from a single line of stations, the plankton is very
uniform and characterized by a high proportion of the very cold-water species. It
offers a marked contrast to the preceding division.

(7) The western Bellingshausen Sea. The exact limits of this area are also unknown,
but it may be regarded temporarily as that part of the Bellingshausen Sea in which the
plankton is rich in comparison with that found in the eastern part near Adelaide Island.
The plankton taken here was in many ways comparable to that found in the eastern
Weddell Sea, but it was rather more variable and the proportion of cold-water species
was not quite so high.

It must be specially emphasized that these divisions are intended to do no more than
represent the conditions as they were found at the time of investigation, and it would in
fact be surprising if certain modifications of the scheme were not found to be necessary
if the results of subsequent work are brought to bear on the question. It should also be
mentioned that the boundaries between the different divisions are not geographically
fixed lines. The Antarctic convergence may perhaps shift its position according to the
time of year, and it has been seen (p. 121) that the change from the scarce plankton of
the Graham Land area to the comparatively rich plankton of the northern zone has been
found in quite different positions at different times. The distribution of the plankton
should not, in any case, be thought of as a static pattern, but as a number of drifting
communities whose formation must depend partly on their origin and the methods of
propagation of the various organisms, and partly on the changing conditions through
which they are carried. In the northern zone the current runs steadily from west to
east with little interruption from islands, shoal waters, and varying ice conditions. This
probably explains the regularity in the distribution of its plankton. The water flowing

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 153

out of the Weddell Sea, on the other hand, impinges on the loop of the Scotia arc, and
although the hydrological conditions in this region have not yet been worked out in
detail, there can be little doubt that there are disturbances in the speed and direction
of the currents. The most variable and fluctuating macroplankton was found, as we
have seen, to the east of the South Sandwich Islands, and this is just where we should
expect to find disturbances and eddies in the water, like the eddies formed in a river on
the downstream side of a ford or the piers of a bridge. In the eastern Weddell Sea,
where the drift of the water is undisturbed, we find again a uniform plankton popula-
tion. Much of this area, as it is shown in Fig. 1, must contain water which is flowing
westwards in the counter-current which is known to exist in the high latitudes, but the
point at which the heterogeneous plankton of the old Weddell water changes to the
homogeneous plankton of the eastern Weddell Sea (i.e. between WS 543 and WS 544),
does not necessarily mark the division between the easterly and the westerly drift.

THE MACROPLANKTON AND THE
DISTRIBUTION OF WHALES

It need hardly be pointed out that no immediate solution of the problem of the dis-
tribution of whales can be sought in a study of the general distribution of the macro-
plankton of the waters in which they live. A more direct connection no doubt exists
between the movements of whales on the one hand and the distribution of " krill " and
the hydrological conditions on the other hand. What we can hope for in a study of the
associated plankton population is something in the nature of a symptom of the environ-
mental conditions which are most favourable to the concentration of whales. The
subject will not be pursued very far at the present stage, but it is worth while to mention
one or two facts which suggest a link between the behaviour of whales and the distri-
bution of the macroplankton in general.

The question is very difficult to approach because the movements and distribution
of whales are influenced by more than one factor of major importance. It is known that
they are migratory animals, visiting the warmer waters for purposes of breeding, and
seeking their food in the cold waters of the Antarctic. We might assume then that the
presence of a large concentration of whales in a particular place is to be explained on the
grounds either that the whales have found there the environment which they require,
or that they are on the way to, or looking for such an environment. The difficulty is to
know which.

Around South Georgia and the South Shetland and South Orkney Islands whaling
has been in progress for many years, and since the modern development of the factory
ship the areas within which whales are caught has been vastly extended. The positions
of the whaling grounds in the Atlantic sector of the Antarctic (as well as in other sectors)
has been defined in recent papers by Hjort, Lie and Ruud (1932 and 1933). The charts
published by these authors (1932, chart no. 1 ; 1933, pis. i, ii, v and vi), show that,
though important changes take place during the season, most of the whaling in the

IS4 DISCOVERY REPORTS

Atlantic sector is carried on at the ice-edge near the South Sandwich Islands, between
Bouvet Island and the South Sandwich Islands and between the South Sandwich and
South Orkney Islands. Ruud (1932, p. 76) also gives a useful chart showing the extent
of the whaling grounds in the season 1929-30. If the charts of these authors are com-
pared with Fig. 48, and it is remembered that South Georgia and the South Shetlands
are also whaling areas, it will be seen that whales are sufficiently plentiful to be hunted
in most of our faunistic areas, i.e. the Graham Land area (South Shetlands and South
Orkneys), the transition zone (South Georgia), the eastern Scotia Sea (ice-edge whaling),
and the "old" Weddell area (ice-edge whaling). During the cruises of the research
ships continuous observations on whales have been made as far as weather conditions
allowed, and our experience is that Blue and Fin whales may be met with anywhere
in the Graham Land area, the transition zone, the eastern Scotia Sea, the "old"
Weddell area, and the western Bellingshausen Sea. These are the same areas as those
mentioned above in which whaling is conducted, but with the addition of the western
Bellingshausen Sea in which whaling has not yet been developed. In the northern zone
it has been observed that whales are very much scarcer, and in the eastern Weddell area
practically none were seen. These are also the two areas (besides the Bellingshausen Sea)
in which whales are not regularly hunted. In the period 1927-31 only one cruise was
made in the eastern Weddell area, but during the recent second commission of the
' Discovery II ' this area was revisited and I am informed that, as before, there was a
notable scarcity of whales.

A comparison of Figs. 2 1 and 48 will show that the normal northern limit of the ' ' krill
(Euphausia superba) corresponds with the southern boundary of the northern zone, and
it has been seen that scarcely any ' ' krill ' ' was found in the eastern Weddell area (see Table
IX, p. 138, WS 544-51). This no doubt accounts for the scarcity of whales in these two
regions. It does occasionally happen, however, that a ship passing through the northern
zone comes across a large herd of whales. Thus on November 12, 1929, when the
'William Scoresby' was in 580 49' S, 570 50' W, near the position of WS 471 (see Fig.
12, p. 80), many whales were seen in various directions from the ship, and were recorded
as travelling in a southerly direction. There can be no question that these whales were
on their way to the southern feeding grounds, and it would be safe to assume that any
large body of whales met with in the northern zone is on its way to, or is returning from,
the environment it requires in the Antarctic. Even in the areas which they commonly
inhabit during the summer, the occurrence of the whales is, of course, very irregular.
There are some places, however, in which they are more often found in large numbers
than in other places. For instance in parts of the " old " Weddell area, especially to the
east of the South Sandwich Islands, and in the eastern part of the Graham Land area
around the South Orkneys and South Shetlands, enormous numbers of whales have
been seen from time to time, while in the central part of the eastern Scotia Sea area
large numbers of whales were rarely seen. The distribution of their food is no doubt the
most important factor controlling the local distribution of whales, but there are other
factors as well, for we must take into account the distinction between whales which are

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 155

travelling and those which are not, and it is almost certain that the whales are influenced
by the position of the pack-ice, the proximity of which is not necessary to the existence
of "krill". Ruud (1932), in studying the connection between the whaling grounds and
the distribution of Euphaiisia superba, points out that the "krill " is widely distributed
in Antarctic waters, but suggests that the principal feeding grounds of the whales are in
the "areas of convergence, backwaters, vortices of mixed layers, and in the centre of
areas with a cyclonic motion ", where a special concentration of ' ' krill " is brought about.
Since the present paper is concerned only with the distribution of the macroplankton
as a whole, these conclusions cannot be examined in detail, but there are certain aspects
of the distribution of the macroplankton which may have a bearing on them. We have
seen that in the northern zone and the eastern Weddell area there are few whales and
little or no "krill". These are also the two areas which apparently contain a particularly
homogeneous plankton population, a fact which should presumably be attributed to the
undisturbed flow of the ocean currents. In contrast to this an extremely variable plank-
ton was found to the east of the South Sandwich Islands, and here the ' ' krill "was abun-
dant and Fin whales were very numerous. The complex plankton distribution suggests
complex or disturbed hydrological conditions, which might be of the kind postulated
by Ruud as favourable to the concentration of ' ' krill ". Around the South Orkney Islands
also, and between the South Orkneys and the South Shetlands, the plankton, as we
have seen, tends to be variable and patchy, and here again large numbers of whales are
often seen. This apparent connection between a variable plankton and the occurrence
of whales does not always hold good, however, for in the Bellingshausen Sea, between
Sts. 565 and 566, and 577 and 578 (see Fig. 14) large numbers of whales were seen by
the ' Discovery II ', yet the plankton here could hardly be described as very variable
and patchy (see Table XII).

There is reason also to believe that in certain circumstances some connection exists
between the distribution of whales and the distribution of the total quantity of macro-
plankton. The evidence for this is inconclusive, but is worth mention. It has been shown
by Kemp and Bennett (1932, p. 178) that during the first half of the whaling season at
South Georgia, the whales are mostly to be found on the north-east side of the island, but
that in the later part of the season there is a tendency towards a greater concentration on
the south-west side. This tendency was noticeable in all four seasons of the period 1 927-3 1 .
We have already seen that there is evidence to suggest a similar shift of the more abundant
plankton from the north or north-east to the south or south-west side of South Georgia,
and it seems possible that this shift of both whales and plankton may be due to some
common cause. In particular the concentration of both whales and macroplankton on the
south side of the island in February-March 1927-8, is very striking (compare Fig. 24 with
Kemp and Bennett, pis. xi and xix). A further indication that concentrations of whales
are connected with concentrations of macroplankton is to be found in the cruises to the
Bellingshausen Sea by the 'William Scores by' in 1929-30 and by the 'Discovery II'
in 1930-1. The 'William Scoresby' took some plankton samples early in January in
the eastern part of the Bellingshausen Sea (WS 496-501). Here the plankton was

156

DISCOVERY REPORTS

unexpectedly rich for these coastal regions, and large numbers of whales were seen in the
vicinity. At the end of the month she travelled westwards from Adelaide Island to about
ioo° W, and taking stations on the return journey (WS 502-8), found a very rich plank-
ton. During this cruise numerous whales were seen south of Peter 1st Island. From
February 10-12 nine more stations were taken in the eastern Bellingshausen Sea off
Adelaide Island (WS 509-17) and the plankton here was now found to be much poorer
and no large numbers of whales were seen. During the next season (1 930-1) the
' Discovery II ' made a similar cruise, taking stations first in the eastern Bellingshausen
Sea (i.e. off the Biscoe Islands and Adelaide Island) at the end of December (Sts. 556-
60). Here the plankton was very thin and whales were sighted only occasionally. The
ship then proceeded westwards to about the same point as that reached by the ' William
Scoresby', and returned to Adelaide Island. During this part of the cruise (Sts. 561-
82) a fairly rich plankton was found and, as with the ' William Scoresby ' in the previous
year, large numbers of whales were seen during part of the cruise, this time some dis-
tance to the north-east of Peter 1st Island. About the middle of January more stations
were taken off Adelaide Island and the Biscoe Islands (Sts. 583-603) and again the
plankton was scarce and very few whales were seen. The results are summarized in the
following table in which the figures represent the average numbers of organisms per
haul with the N 100 B. The numbers of stations upon which the averages are based are
shown in brackets.

Table XVIII. Whales and quantity of plankton in the Bellingshausen Sea

1929-30
('William Scoresby')

1930-1
('Discovery II')

Oft Adelaide

Island and

Biscoe Island

Between

Adelaide Island

and 100° W

Off Adelaide

Island and

Biscoe Island

Between

Adelaide Island

and iooc W

Late December
Early January
Late January
Early January

Many whales

3300 (5 Sts.)
Many whales

360 (8 Sts.)
Several whales

20,000 (5 Sts.)
Many whales

102 (5 Sts.)
Several whales

113 (17 Sts.)
Few whales

4263 (22 Sts.)
Many whales

The important point here is that the coastal region near Adelaide Island was visited
four times. On three occasions the plankton was very scarce and whales were not
plentiful. On the fourth occasion the plankton was about ten times as rich, and large
numbers of whales were seen. This may be a coincidence, but it is suggestive, in view
of the similar evidence from South Georgia, of a connection between the whales and
the quantity of plankton. It is very improbable that the whales have any particular

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 157

inclination to follow a rich macroplankton unless there is also a rich development of
" krill", but it is quite possible that the hydrological conditions which give rise to a rich
macroplankton are also in certain circumstances the conditions which provide the
environment which the whales seek in the Antarctic.

It is obvious enough that the occurrence of whales does not vary regularly with the
quantity of macroplankton. The Bransfield Strait, for instance, is normally exceedingly
poor in plankton, and yet it is a region in which whales are known to be plentiful. We
have already seen, however, that there must be more than one factor controlling the
distribution of whales in the Antarctic, and consequently it cannot be expected that one
set of circumstances will determine that distribution at all times and places.

SUMMARY

This paper is concerned with the plankton samples taken in Antarctic surface waters
with the 1 m. oblique nets, and its object is to describe the horizontal distribution of the
macroplankton as a whole. It is based on the analysis of about 600 samples collected in
the period 1927-31.

The Antarctic convergence is the northern boundary of the area under consideration.
Some Antarctic species are as common in the sub-Antarctic water on the north side of
the convergence as they are in the Antarctic water on the south side, some occasionally
stray into the sub-Antarctic water, and others are entirely confined to the south side of
the convergence.

The amount of diurnal variation shown by each species has been estimated, and the
average number per haul at different times of day and night plotted on graphs. It is
found that there is every gradation between the species which can be caught in large
numbers in the middle of the night but which are entirely absent in the middle of the
day, and those which are taken equally at any time of day or night, or may even be
commoner during the day. A period of daytime hours is allotted to each of the more
variable species, and samples taken within that period are not regarded as valid in-
dications of the presence or absence of the species in question.

Brief notes are given on the individual distribution of each species, and it is shown
that some are typical of the colder water or high latitudes, some characterize the warmer
water or lower latitudes, and others again may be taken equally in warm or cold regions.

The richness of the plankton varies greatly in different places and at different times.
In the neighbourhood of the South Orkneys and South Shetlands, and in the eastern
part of the Bellingshausen Sea the plankton appears always to be very scarce, possibly
as a result of the upwelling of deep water. In other regions it is variable but generally
very rich except during the winter months. It is possible that the boundary between
the scarce plankton near the South Shetland Islands and the richer plankton of the
Drake Passage may shift southwards in the latter part of the summer.

The distribution of species typical of warm or cold water is specially interesting.
There is abundant evidence to show that in a given area the proportion of cold-water
species in the plankton population becomes reduced as the summer goes on. There is

i58 DISCOVERY REPORTS

a high proportion of cold-water species in spring and a low proportion in autumn, and
there is little doubt that the warm-water species remain dominant during the winter in
spite of the lower temperature of the water. The distribution of the cold-water species
appears to have some connection with the movements of the pack-ice, but in the lower
latitudes, which are not reached by the pack, the same reduction takes place as the
season advances. In a cold summer there is a higher proportion of cold-water species
which, however, become reduced in the same way. The material, so far as it goes,
suggests further that an exceptionally cold summer season in the Atlantic sector of the
Antarctic may not necessarily be accompanied by a correspondingly cold season on the
west side of the Drake Passage.

Different plankton communities can be distinguished in different water masses.
Fig. 48 shows roughly the different areas in which a characteristic plankton population
was found. These areas coincide to a large extent with different water masses. In the
outer or northern zone of the Antarctic there is a uniform plankton population which
undergoes little change in composition during the year. In other places, such as to the
east of the South Sandwich Islands and in parts of the Orkney-Shetland region the
plankton may fluctuate sharply from place to place and from time to time. It appears
that where the hydrological conditions are uniform and undisturbed there is a uniform
and stable plankton, and where the hydrological conditions are complex and variable
there is a variable and unstable plankton population.

In the last section the connection between the macroplankton and the distribution of
whales is discussed. It seems that whales are found in regions characterized by a
variable plankton rather than in those with a uniform and stable plankton, and there is
evidence which suggests that in certain circumstances the distribution of whales may
be correlated with the distribution of the quantity of plankton.

LIST OF LITERATURE

Barnard, K. H., 1932. Amphipoda. Discovery Reports, v, pp. 1-326, text-figs. 1-174, pi. i.
Benham, W. B., 1921. Polychaeta. Australian Antarctic Exped., ser. C, vi (3), pp. 1-128, pis. 5-10.
Bigelow, H. B., and Leslie, M., 1930. Reconnaissance of the Waters and Plankton of Monterey Bay, July,

1928. Bull. Mus. Comp. Zool., lxx, 5, pp. 427-581, text-figs. 1-43.
Bonnevie, K.R., 1913. Pteropoda. Rep. Sci. Results 'Michael Sars' North Atlantic Deep-Sea Exped., II,

Pt. 1, Zoology, pp. 1-15, text-figs. A and B, pis. i-v.
Browne, E. T., 1910. Coelenterata. V. Medusae. National Antarctic Exped., v, pp. 1-62, pis. i-vii.
Chun, C, 1897. Die Siphonophoren der Plankton-Expedition. Erg. d. Plankton-Expedition der Humboldt-

Stiftung, 11, K.b, pp. 1-126, text-figs. 1-2, pis. i-viii.
Deacon, G. E. R., 1933. A General Account of the Hydrology of the South Atlantic Ocean. Discovery

Reports, VII, pp. 171-238, text-figs. 1-24, pis. viii-x.
Eliot, C, 1907. Mollusca. VI. Pteropoda. National Antarctic Exped., in, pp. 1-15, pis. i-ii.
Hansen, H. J., 1913. Crustacea Schizopoda. Swedish Antarctic Exped. 1901-3, pp. 1-44, pis. i-vi.
Hardy, A. C, and Gunther, E. R., 1934. The Plankton of the South Georgia Whaling Grounds and

Adjacent Waters, 1926-7. Discovery Reports (in press).

DISTRIBUTION OF ANTARCTIC MACROPLANKTON 159

Hart, T. J., 1934. On the Phytoplankton of the South-west Atlantic and the Bellingshausen Sea, 1929-31.

Discovery Reports, vm, pp. 1-268, text-figs. 1-84.
Hentschel, E., 1928. Die Grundziige der Planktonverteilung an Sudatlantischen Ozean. Internat. Rev. d.

ges. Hydrobiol. u. Hydrogr., xxi, pp. 1-16, text-figs. 1-5.
Hentschel, E., andWATTENBERG, H., 1930. Plankton und Phosphat in der Oberflachenschicht des Siidatlantisches

Ozeans. Ann. d. Hydrogr. u. Marit. Meteor., lviii, Heft viii, pp. 273-7.
Herdman, H. F. P., 1932. Report on Soundings taken during the Discovery Investigations, 1926-32. Discovery

Reports, VI, pp. 205-36, text-figs. 1-5, pis. xlv-xlvii, charts 1-7.
Hjort, J., Lie, J., and Ruud, J. T., 1932. Norwegian Pelagic Whaling in the Antarctic, I. Hvalradets

Skrifter (Norske Vid.-Akad. Oslo), 3, pp. 1-37, charts 1-6.

T933- Norwegian Pelagic Whaling in the Antarctic, II. Hvalradets Skrifter (Norske Vid.-Akad. Oslo),

7, pp. 128-52, text-figs. 75-80, pis. i-vi.
Hjort, J., and Ruud, J. T., 1929. Whaling and Fishing in the North Atlantic. Rapp. et Proc. Verb., Conseil

Internat. pour l'Exploration de la Mer, lvi, pp. 1-123, text-figs. 1-59.
Holt, E. W. L., and Tattersall, W. M., 1906. Preliminary Notice of the Schizopoda collected by H.M.S.

'Discovery' in the Antarctic Region. Ann. Mag. Nat. Hist. (7), xvn, pp. 1-11.
Ihle, J. E. W., 1912. Salpae. I. Desmomyaries . Das Tierreich, pp. 1-66, text-figs. 1-68.
Jespersen, P., 1923. On the Quantity of Macroplankton in the Mediterranean and the Atlantic. Rep. Danish

Oceanogr. Expeds. No. 7, pp. 1-17, text-figs. 1-8.
Kemp, S., and Bennett, A. G., 1932. On the Distribution and Movements of Whales on the South Georgia

and South Shetland Whaling Grounds. Discovery Reports, vi, pp. 165-90, text-figs. 1-6, pis.

vii-xlii.
Kemp, S., and Hardy, A. C, 1929. Discovery Investigations: Objects, Equipment and Methods. Pt. II. The

Ships, their Equipment and the Methods used in Research. Discovery Reports, 1, pp. 151-222, text-
figs. 1-33, pis. vii-xvii.
MacBride, E. W., 1912. On a Collection of Young Holothuroids . National Antarctic Exped., vi, pp. 1-9,

pis. i-ii.

1920. Echinoderma (Part II) and Enteropneusta. British Antarctic ('Terra Nova') Exped., 1910. Zool.,

iv, 3, pp. 83-94, pis. i-ii.
Massy, A. L., 1920. Mollusca. Part III. Eupteropoda (Pteropoda Thecosomata) and Pterota (Pteropoda
Gymnosomata). British Antarctic (' Terra Nova ') Exped., 1910. Zool., 11, 9, pp. 203-32, text-figs. 1-9.

1932. Mollusca: Gastropoda, Thecosomata and Gymnosomata. Discovery Reports, in, pp. 267-96,

text-fig. i, pi. xxxix.
Moser, F., 1925. Die Siphonophoren der Deutsche Sudpolar-Expedition, 1901-3. Deutsche Sudpolar- Exped.,

17, Zool., pp. 1-541, text-figs. 1-61, pis. i-xxxvi.
Murray, J., and Hjort, J., 1912. The Depths of the Ocean. London, pp. 1-821, text-figs. 1-575, Pls- i_ix-
Ottestad, Per, 1932. On the Biology of some Southern Copepoda. Hvalradets Skrifter (Norske Vid.-Akad.

Oslo), 5, pp. 1-61, text-figs. 1-8.
Russell, F. S., 1925-31. The Vertical Distribution of the Macroplankton. Journ. Mar. Biol. Ass., xm-xvn.
Rustad, D., 1930. Euphausiacea, with Notes on their biography and development. Scientific Results of the

Norwegian Antarctic Expeditions, 1927-8 and 1928-9 (Norske Vid.-Akad. Oslo), No. 5, pp. 1-83,

text-figs. 1-51, pis. i-vii.
Rustad, D., 1930. Mysidacea. Scientific Results of the Norwegian Antarctic Expeditions, 1927-8 and

1928-9 (Norske Vid.-Akad. Oslo), No. 6, pp. 1-28, text-figs. 1-19, pis. i-iii.
Ruud, J. T., 1932. On the Biology of Southern Euphausiidae. Hvalradets Skrifter (Norske Vid.-Akad. Oslo),

No. 2, pp. 1-105, text-figs. 1-37.
Sars.G.O., 1903. Copepoda. Calanoidea. An Account of the Crustacea of Norway, iv, pp. 1-171, pis. i-cn.
Schmaus, P. H., and Lehnhofer, K., 1927. Copepoda. 4. Rhincalanus, Dana, 1852. Deutsche Tiefsee-

Exped., xxiii, pp. 355-400, text-figs. 1-29.
Stebbing, T. R. R., 1888. Report on the Amphipoda Collected by H.M.S. ' Challenger' during the years 1873-6.

Rep. Sci. Res. Voy. 'Challenger', xxix, text and plates.
- 1906. Amphipoda Gammaridea. Das Tierreich, Lief, xxi, pp. i-xxxix, 1-806, text-figs.
Tattersall, W. M., 1908. Schizopoda. National Antarctic Exped., IV, pt. vii, pp. 1-41, pis. i— viii.

l6o DISCOVERY REPORTS

Wolfenden, R. N., 1908. Crustacea. VIII. Copepoda. National Antarctic Exped. 1901-4, iv, pp. 1-46,

pis. i-viii.

. I9u. Die Marinen Copepoden der Deutschen Siidpolar-Expedition. Deutsche Sudpolar-Exped., XII,

Zool., iv, pp. 181-380, text-figs. 1-82, pis. xxii-xl.

WilsT, G., 1928. Der Ursprung der Atlantischen Tiefenwdsser . Zeitschr. Ges. Erdkunde Berlin, pp. 506-34,
text-fig. 32, pis. xxxiii-v.

Zimmer, C., 1914. Die Schisopoden der Deutschen Siidpolar-Expedition, 1901-1903. Deutsche Sudpolar-
Exped., xv, Zool., vii, pp. 377-445. Pls- xxiii-xxvi.

[Discovery Reports. Vol. IX, pp. 161-174, Plate I, June, 1934]

THE SUB-ANTARCTIC FORMS OF
THE GREAT SKUA

(CATHARACTA SKUA SKUA)

By
J. E. HAMILTON, M.Sc.

CONTENTS

Introduction Paie l63

Plumage l64

Measurements X"S

Geographical groups I7I

Conclusions *73

Note on the Distribution of McCormick's Skua 174

Literature J74

THE SUB-ANTARCTIC FORMS OF

THE GREAT SKUA

(CATHARACTA SKUA SKUA)

By J. E. Hamilton, m.sc.
(Plate I; text-fig. i)

INTRODUCTION

The component forms of the genus Catharacta may be readily divided into four
principal groups which are considered by Lowe and Kinnear (1930) to constitute
a single species, Catharacta skua (Briinnich), of which the Great Skua is the type.
The four groups are :

(1) The Great Skua, Catharacta skua skua (Briinnich), which extends over the
North Atlantic Islands to north-eastern America. Geographically it is completely
isolated from all the other groups and there can be no doubt as to its individuality.
There is usually, but not always, a strong chestnut cast in the plumage.

(2) The Chilian Skua, Catharacta skua chilensis (Bonaparte), occupying the east and
west coasts of South America from Rio de Janeiro to Callao. This is the characteristic
South American bird, but it extends to the Falkland Islands also. On the extreme eastern
side of that group I have seen it mixing freely and breeding with the form which is
numbered 4 below. The Chilian Skua is suffused with an even deeper chestnut colour
than is the Great Skua.

(3) McCormick's Skua, Catharacta skua maccormicki ( Saunders), is a strictly Antarctic
form distinguished by the entire absence at all stages of chestnut colouring, by the
vinous tinge of the lower surface, and usually by the presence of a conspicuous pale
collar of acuminate feathers many of which are bright gold in colour.

(4) The sub-Antarctic Skua, Catharacta skua (subspecies), which may conveniently
be termed the Brown Skua. This heading comprehends an assortment of birds which
are found on all the sub-Antarctic islands, on the Tristan da Cunha group and on Gough
Island. These birds may be described as brown in colour, often having on the neck
yellow acuminate feathers which are, however, not so brilliant as the corresponding
feathers of McCormick's Skua. Chestnut and pale markings are very frequently present,
the former on the dorsal and the latter on both dorsal and ventral surfaces. The presence
or absence of the chestnut patches on the feathers is quite fortuitous and sometimes
they are not bilaterally symmetrical ; a chestnut mark on one side may correspond to
a pale mark on the other, so that it appears probable that the pale marks are due to the
absence of chestnut pigment.

164

DISCOVERY REPORTS

Wilson (1907, p. 67) believed that the condition of paleness was due to weathering
of the plumage. In my opinion this is an error, due to the fact that the pale parts of the
feathers are much more susceptible to the effects of weathering than are those parts
which possess the darker pigments, brown and chestnut, as can easily be verified by
examination of skins.

The number of specimens of the Brown Skua available for examination has been
substantially increased since the publication of the Birds of Australia by G. M. Mathews
in 191 3, and the purpose of this paper is an examination of the validity of four subspecies
named in that work. According to Lowe and Kinnear (1930) the following is the correct
nomenclature :

Catharacta skua antarctica (Lesson), Falkland Islands, Tristan da Cunha and Gough
Island;

C. s. clarkei (Mathews), South Shetlands, South Orkneys and South Georgia;

C. s. intercedens (Mathews), Kerguelen and Crozets ; and

C. s. Idnnbergi (Mathews) of the New Zealand region.

These subspecies are supposed to be distinguishable by differences in plumage and
in the measurements of wing, tarsus and bill.

In order to facilitate an examination of the data I have adopted a division into seven
geographical groups:

Number of specimens

Locality

of which data are

available

New Zealand area ...

29

Kerguelen and Crozets

10

Tristan da Cunha and Gough Island

8

South Georgia

9

South Orkneys

13

South Shetlands

12

Falkland Islands

40

Total 121

PLUMAGE

As a preliminary it should be said that field observations must be accepted with
reserve ; it is extremely difficult and often impossible to say if a bird seen on the wing
is large or small, pale or dark. Alterations in appearance due to background, movement
or light are usually too numerous and sudden to permit of them being taken into account :
it may be added that at a short distance at sea all brown skuas look dark, sometimes
almost black.

The longest series of skins available for examination in the British Museum is that
from the Falkland Islands ; it contains thirty-seven specimens. Of these twenty-seven

PLUMAGE i6S

were collected by myself, but no attempt was made to pick out examples of any particular
plumage phase — the birds were killed quite at random.

The thirty-seven specimens were all collected between November and April and none
shows any sign of moult. Since the skuas leave the Falklands in April it follows that
the moult must take place during the southern winter when the birds are absent from
their breeding haunts, thus affording a contrast to the habit of McCormick's Skua which
moults after the breeding season (Wilson, 1907, p. 74).

The birds from the Falkland Islands undoubtedly belong to one race, and even a
superficial examination shows that it is very variable ; dark, light and intermediate forms
are all present. If the skins are arranged in order of size (as judged by wing length) it
will be found that there is no correlation between colour and measurement.

As has been said on p. 164 the differences between the subspecies of Brown Skua are
stated to be those of plumage and size, but in every subspecific group there is con-
siderable individual variation, and so far as colour is concerned the specimens of any
race may be matched by skins from the Falklands. There are, for example, two or three
skins which have the dusky plumage supposed to distinguish Mathew's lonnbergi and,
contrasting with them, are very light birds, plentifully splashed on the mantle with pale
marks, which are identical with South Shetland specimens of clarkei. In view of so
much agreement I do not consider that plumage differences can be regarded as valid
subspecific (racial) characters.

MEASUREMENTS

The subject of differences based on dimensions may now be considered. The four
standard measurements are those used by Lowe and Kinnear and are the following:

(1) Length of wing from the carpal joint to the tip of the longest primary, the manus
having been straightened out as much as possible. This measurement must be considered
as only approximately correct; complete accuracy is impossible since the amount of
bending of the dried manus varies from one specimen to another and it is never possible
completely to straighten it out.

(2) Length of exposed culmen.

(3) Depth of bill at base of exposed cidmen.

(4) Length of tarsus.

Since it is obvious that the greater the variability of a character the less is its value in
the separation of subspecies, I have attempted to determine the range of variation of
the four measurements in each of my geographical groups. The measurements are shown
in Table I and Fig. 1. Table II gives for each group firstly the average, minimum and
maximum measurements, and secondly the divergence of the minimum and maximum
from the mean, expressed as percentages of the latter, and range of variation also ex-
pressed as the percentage of the mean.

It will be observed that the lengths of the wing and tarsus are the least variable of
the measurements, and that the depth of the bill shows greater range than the length.

1 66

DISCOVERY REPORTS

Table I. Measurements of Catharacta skua

Length

Length

Depth

Length

Locality and reference

Sex

of wing

of bill

of bill

of tarsus

mm.

mm.

mm.

mm.

New Zealand

Tring Museum

0*

442

56

24

77

B.M. 05.12.30. 172

c?

424

53

22-5

76

B.M. 05. 12.30.167

(J

423

51

23

95

B.M. 05.12.30.168

0*

421

55

24

78-5

B.M. 05.12.30.177

9

447

55

26-5

79

Tring Museum

9

437

57

24

80

> > >>

9

437

56

25

81

B.M. 05.12.30.178

$

435

56

25

79-5

Tring Museum

$

434

56

23

80

B.M. 97. 12. 6. 41

$

430

54

24

79

B.M.43.7.11.38

$

425

48

23

74

Tring Museum

?

424

58

25

82

yy >»

9

418

57

23

78

>» yy

9

410

55

21

78

)» »»

9

407

57

23

77

B.M. 93.6.24.1

—

443

5i-5

22

75

B.M.01.1.7.55

—

433

51

24

80

B.M. 51.7.18.34

—

43 1

81

B.M. 91. 6. 16. 44

—

43 !

—

23

84

B.M. 92.4. 15. 1

—

429

52-5

23-5

80

Tring Museum

—

421

59

24

78

>> »»

—

420

55

25

79

B.M. 91. 12. 16. 51

—

419

53

24

78

Tring Museum

—

4J5

56

23

76

>3 )1

—

411

57

22

78

>> )»

—

407

57

23

82

B.M. 45.7.6.68

—

4°5

5i

22

81*

B.M. 91. 6. 11. 15

—

395

53

24

84

B.M. 03.3.20.3

—

—

56

23

83

Kerguelen and Crozets

Tring Museum

0"

398

56

22

75

B.M. 09. 11. 16. 9

9

422

57-5

22

77-5

Tring Museum

9

414

57

22

76

B.M. 41.768

—

410

52

23

78

B.M. 76.4.26.22

—

406

57

24

78

B.M. 09. 1 1. 16. 8

—

4°3

49-5

20-5

7i-5

B.M. 80. 11. 18. 734

—

402

54

22-5

74

B.M. 41.767

—

401

53"5

22

74-5

B.M. 91. 6. 16. 8

—

395

54

22

76

B.M. 80. 11. 18. 734

—

395

5°

20

73-5

South Georgia

B.M. 14.3.8.48

3

414

49

22-5

72

B.M. 14.3.8.49

a

4°3

54

22

76

Tring Museum

s

401

54

22

75

B.M. 22.12.6.32

s

39°

5°

22

74

B.M. 14.3.8.50

d*

395

52

23

71

B.M. 22.12.6.36

9

422

52-5

22-5

77

Tring Museum

9

400

55

22

75

>» >»

—

421

58

23

77

>» )>

—

397

56

24

76

From the Cape of Good Hope.

MEASUREMENTS
Table I (cont.)

167

Length

Length

Depth

Length

Locality and reference

Sex

of wing

of bill

of bill

of tarsus

mm.

mm.

mm.

mm.

South Orkneys

Royal Scottish Museum

—

422

56

—

78

) ») t>

—

419

52

—

73

» J) »»

—

418

49

—

79

» )» ))

—

415

49

—

78

1 )J >»

—

412

46

—

72

» I) )>

—

412

52

—

77

» JJ ))

—

411

53

—

72

) »> )»

—

411

52

24

76

» »l >»

—

410

44

—

77

» »» >»

—

406

52

—

74

1 i> >J

—

406

51

—

71

y a >j

—

4°3

53

—

76

» »» )J

—

401

50

—

72

South Shetlands

B.M. 24.5.8.85

6"

4J5

52

23

75

B.M. 23. 9. 10. 1

6*

414

48

21

72

B.M. 24.5.8.86

cJ

410

5°-5

22

75

Paris Museum

9

427

50-5

23

72

B.M. 24.5.8.8

?

392

49

22

73

B.M. 24.5.8.83

■

400

5°

22

72

Paris Museum

<J

406

53

23

68

B.M. 24.5.8.9

cJ

375

47

18-5

65

B.M. 24.5.8.10

¥

393

44

J9

64

B.M. 24.5.8.7

9

381

46

!9

68

B.M. 24.5.8.4

—

392

46-5

20

66

B.M. 23.9.10.2

—

387

5°

J9

63

Falkland Islands

B.M. 15.xii.30

0"

39°

46

!9

67

B.M. 4. i. 31

<J

390

46

20

64

B.M. 19.1.32

0"

385

48

20-5

67

B.M. 17.xii.30

0*

381

46

*9"5

66

B.M. 9.1.31

cJ

380

47

!9-5

65

B.M. 18.xii.30

d

379

45

20

67

B.M. 4.1.31

6*

374

45

!9'5

63

Tring Museum

(J

372

5i

20

69

B.M. 4. i. 31

(J

371

45

i9-5

63

Tring Museum

cf

37°

49

21

68

B.M. i.iii.32

c?

370

44

19

64

B.M. 17.xii.30

0*

360

47

18-5

64

B.M. 17.xii.30

6"

360

47

!9'5

68

B.M. 13.1.30

9

400

47

J9

67-5

B.M. 26.1.31

9

396

5i-5

J9

70

B.M. 4. i. 31

9

394

48

19-5

69

Tring Museum

9

389

50

21

68

B.M. 9. i. 31

9

389

49

21-5

66-5

Tring Museum

9

388

52

21

69

B.M. 9. i. 31

9

387

50

22-5

67

B.M. 28. i. 32

9

387

49

21-5

70

B.M. 4. i. 31

9

383

46

20-5

69

1 68

DISCOVERY REPORTS

Table I (cont.)

Length

Length

Depth

Length

Locality and reference

Sex

of wing

of bill

of bill

of tarsus

mm.

mm.

mm.

mm.

Falkland Islands (contd.)

B.M. 26.xi.30

$

383

45

21

67

B.M. 19.1.32

9

38i

46

20-5

64

B.M. 17.xii.30

9

380

48

!9-5

"7-5

B.M. 18.xii.30

9

380

47

I9-5

69

B.M. 20.xi.31

?

380

47

19-5

65

B.M. 91. 5. 22. 61

$

380

48

20

69

B.M. 26.xi.31

9

379

46

21-5

64

B.M.

2

377

47

20-5

69

Tring Museum

9

377

5°

21

69

B.M. 25.1.32

9

374

47

19-5

65

B.M. 80. 11. 18. 735

9

368

47

20-5

69

B.M. 1894. 10. 28. 3

—

394

49

20

67

Tring Museum

—

385

51

20

67

B.M. xi.30

—

385

46

21

72

B.M. xi.30

—

383

47

20-5

69

Royal Scottish Museum

—

382

46

70

B.M. 91.5.22.60

—

370

47

20-5

66-5

B.M. i. 31

—

368

48

20-5

69

B.M. 1894. 10.28.2

—

36S

48

20

68

Tristan da Cunha Group

and Gough Island

Tring Museum

6"

385

53

21

70

B.M. 22.12.6.33

S

385

5°-5

20-5

73

B.M. 22.12.6.35

6*

38i

52

22

73-5

Royal Scottish Museum

3

373

51

—

69

B.M. 22.12.6.34

9

39°

53-5

24

75-5

Tring Museum

9

39°

55

22

72

)» >»

9

380

54

22

71

Royal Scottish Museum

—

389

54

75

The majority of the above measurements have been extracted from British Antarctic {'Terra Nova')
Expedition, 1910, Lowe and Kinnear (1930).

The specimens referred to "Tring Museum" are now in America.

It is of particular interest that the two longest series of specimens have the greatest
variability in length of wing and tarsus, thus suggesting that a short series may well fail
to exhibit the degree of variation to which a group is liable and is therefore the worse
basis for subspecific differentiation — a point which has apparently been overlooked
occasionally in the past.

MEASUREMENTS

169

S

nv>

S

^>

~.
a

to

<3

^
«

IT!

—
ca

+-»

as

U

rt T3

Tristan
a Cunh
roup ar
Gough
Island

O

On to

ON tO ~+" ti-j

~~5
0^

M ON O

to O to

Tt-NO O

r^ co 0

H

•*■ conO

N O to

M O ■*■ N ON tO

CO N NO

■*■ ^00

O OnnO

n*- ■*■ on

S

CO bO>

10 10>0

N N N t^NO t~~

w

en co co

"0 oc

1-

£ T3

0

10

N to to N

^0

CO CO NO

-4- to ON

^t- "H-00

N m CO

J3 c

3

000

■■*■ 4- to

6 00 N K CO N

0^

to to 0

t^ OnnO

00 M ON

NO t^ CO

3= —

g

CO NO O

n tn ^t

N i-h N \C NO !"■*

HH

1H

M M

1-1

fcn ^

outh

tlands

irsus

nder

mm.)

R

O

ON tONO

f-» ■<(• CO

t^ to t^

Onoo co io coco

"~5
0^

NO ^" p

co ^-co

CO M ON
tNMCC

T1 C^°?

no NO N

IH tONO

g

00 r^ 0

-*- tJ- to

M M (N) NO NO NO

M 1H

r-« .3 & 3 O

Ed ^

nn +

3Uth

tlands
irsus
ver
mm.)

i^

0

N CM

NsP

CO N to

OOO

ThNO O

no 10 «

P

on n r^

O 00 N

N « CO CO N to

CTN

Th -+00

■*■ "*-co

to CO On

" N •*

£

O ON N

to ^t- to

cni cn! n t^ r^ r~

CD ^

■*• CO -*■

-3 3 1)

CO

ON

O to n+

NsP

W ON O

0 -*■ -t-

on to ■*■

N 1-1 CO

Sout

hetla

(all

ecim

H

00 ■*■ CO
•* -r to

hX CO ON CO IO

C>^

NO NO CO

0 co 00

1-1 On ih

onoo 1-^

g

ON J-^ N
CO CO tJ-

N H N NO no r--

M

H

CD °-i
^ to

to

3 3

R

N

M M N

6 tj-no

to -^- to

*-* *-* ■'-' 0

•O "O "O to »-i on

N N to

n to r^

CO O CO

1 1 1

CO CONO
to to 6

O •£

g

MOM

OOO t-»^^

M 1-1 N

Mo

-+■ -+■ "+■

ctj

-S '5b

**■

•*■

no 00

NvP

no 11 r^

N no CO

r^ N on

M ON O

3 £

h

IO to N

CO ONOO

N N ■■*- + " N

O^

N ^J-NO

00 CO NO

N NO 00

IO N 00

O g
CD0

g

O O M

■ 0 ■<*■ to

n n n r^ r^ t~~

•*■ co Th

3

erguele

and
Crozets

no

O to to

0 ■*■ to

N^P

■+nr~

co toco

H w N

N -d-NO

H

t)- to N

■*■ on r~»

N O ^t- to m CO

c>-

N t}-nO

00 NO "*-

ON ON00

to cooo

g

O ON N
Tt- CO ■+

to th to

N N N t^. t~» r^-

M

T3

fj

ii-00 o>

to to 00

CO M NO ON -^- tO

■nP

0^

►1 tJ- to

NO tO W

N 0> w
N (nO

NO 00 ^t-
6 N CO

co O co
i^. ONNO

N

g

**■ CO ■+

vo + 10

(Nj N N t^ t^ ON

z

0

C 3 C

3 3 3

age

egg

rt « O

C3 « O

3 3

9 1

3 2

g E E E
3 3 ^ 0

si

E E.s

E £.2

E S.2

£ £ .2

3 I.i

0 .3 ca

3 g.i

ttt 'c X

ean

inim

axim

ean

inim

axim

0 > ^

00

g > ^
O 0 ^

1 > ^

£ £ «

s •*" >

-2 ° JI

DEPARTURES
D RANGE OF

"53 -9 ^

13 -3 ^s

13 -o ^

§2§ s

«<^r,

m <! °

bo

be

00

i; sp

to

H

2

si

a

IS

Depth of bill:
Length of tarsus

3

M

=3 OS

3

a
- o3

^3

2 3
n

i

a
«i

<

0

3
V

h-5

0

a

3
U

0 2

t4-l

0

J3
3

s

tw

O

So

3

V

Ct-I

0

-3
0,

V

Q

O

-3

cjo
3
u

GEOGRAPHICAL GROUPS 171

GEOGRAPHICAL GROUPS

Tables I and II contain the numerical data for each of the seven groups, which are
arranged in order from east to west.

New Zealand and Falkland Islands. The differences in average measurements are
greatest between the New Zealand and the Falkland Islands groups, which are also the
farthest apart, and if these, the extremes, are compared first the positions of the inter-
mediate groups will be easier to define.

Measurements from forty-one Falkland birds are available, and of them thirty-four
are in the British Museum, while the New Zealand specimens are twenty-nine in
number with sixteen in the British Museum.

One New Zealand specimen has a tarsal length which is altogether exceptional, 95 mm.,
the next highest being 84 mm. ; the isolation of the larger measurement suggests that it
may be the result of development which is not quite normal. Between the two series
the wing measurements do not overlap ; but the difference is only 7 mm., less than 2 per
cent of the maximum for the Falkland Islands (400 mm.), and this 7 mm. is quite
possibly within the error of measurement (p. 165). There is a noticeable overlapping
in bill measurements, and in tarsal length there is a difference of only 1 mm. (Fig. 1).

In spite of this close approximation the New Zealand birds are on the whole larger
than those from the Falklands and they often have rather duller plumage and a less
developed yellow collar. But, since there are no real plumage differences and no abrupt
hiatus between the series of measurements, the two forms cannot be regarded as being
more than subspecies of Catharacta skua, as is suggested by Lowe and Kinnear, and the
names used by these authors should be retained, namely C. s. antarctica (Lesson) for the
Falkland Islands form and C. s. lonnbergi (Mathews) for that of the New Zealand region.

Kerguelen and Crozets. The other groups of Brown Skua may now be considered
in accordance with their distribution.

Nearest in space to lonnbergi is the race from the Crozets and Kerguelen : it has been
distinguished as C. s. intercedens by Mathews. An examination of the eight specimens
in the British Museum shows that there is no difference between them and the New
Zealand birds so far as plumage is concerned.

It will further be observed that the measurements of eight out of the ten specimens
from which the data were taken fall within the limits of the dimensions of lonnbergi, in
spite of the fact that the average of the wing lengths is 20 mm. less than the corresponding
figure for that race (Fig. 1). In the other measurements there is a difference of 1 mm.
in the depth of the bill, of 1 mm. in the length of the tarsus of one specimen, and of
2-5 mm. in the length of the tarsus of another. It appears, then, that the distinction
between these two races rests on the possession of a wing which may be, but is by no
means always, shorter in the Kerguelen bird, and in the occasional occurrence of a
slightly shorter tarsus. Since it is known that the range of measurements in the Brown
Skuas is considerable it cannot be considered that such slight differences are grounds
for separating the Kerguelen group as a subspecies.

i72 DISCOVERY REPORTS

Support is given to this statement by a specimen from a position as far west as the
Cape of Good Hope, where it would be reasonable to expect a representative of the
Kerguelen race if it really had a separate existence ; but on grounds of measurement the
African bird should be placed in C. s. lonnbergi.

Another specimen, from Travancore, in the possession of Mr H. Whistler, should
also be included in the lonnbergi subspecies. C. s. intercedens should be abandoned and
the birds from Kerguelen and the Crozets included in C. s. lonnbergi.

South Georgia. The statements with regard to the Kerguelen birds apply also to
those from South Georgia; there are no plumage differences of importance and the
measurements of the nine specimens of which data are available fall almost without
exception within the range of the dimensions of C. s. lonnbergi (Fig. i). It follows that
the section of C. s. clarkei (Mathews), which has been stated to occupy South Georgia,
should be included in C. s. lonnbergi.

South Orkneys. In the matter of plumage Eagle Clarke (191 5) states that in a series
from the South Orkneys there are two types which, however, do not differ in size. One
is a dark form which has few and sometimes no pale marks on the mantle, while the
other is lighter with many pale marks on the back and sometimes on the breast. The
pale form had "the yellow streaks on the neck much more numerous and pronounced
than in the darker birds ; and they agreed with the form described by Saunders ... as
inhabiting the Falklands except that they are not smaller than the ordinary dark form"
(of the South Orkneys). " One of these light birds was observed to be mating with one
of the dark examples." This occurrence of two roughly distinguishable types — a light
and a dark — is common among the Brown Skuas, but there is no real division between
the two phases, which grade into one another (p. 165).

There are no specimens from the South Orkneys in the British Museum ; the measure-
ments of Lowe and Kinnear are taken from the series in the Royal Scottish Museum
and these are summarized in Table II. The average wing measurement is rather greater
than that for South Georgia, and so is the tarsal length, but the bill length is rather less.
That is to say in wing and tarsal measurement the South Orkney birds are nearer to the
New Zealand group than are those from South Georgia.

Within the limits of the South Orkney group of specimens the bill varies as much as
11 mm. and the tarsus 8 mm., whereas the greatest difference from the New Zealand
specimens is 4 mm. for the bill and 3 mm. for the tarsus. I do not consider that the
South Orkney birds can be separated as part of C. s. clarkei on the basis of small and
inconstant differences in measurements. Like those from South Georgia, therefore, the
South Orkney skuas must be included in C. s. lonnbergi because the differences in
measurements are so small and no separation is possible on the basis of plumage
differentiation.

South Shetlands. In the South Shetlands the position is one of some interest.
Details of twelve specimens are available, ten being in the British Museum and two in
Paris. In the paper by Lowe and Kinnear (1930, p. 118), it will be observed that the
list of C. s. clarkei includes only ten from the South Shetlands — those in the British

GEOGRAPHICAL GROUPS 173

Museum; but under C. s. maccormicki (p. 122) a male and a female are recorded as
having been taken by Charcot's expedition at Deception Island on December 2, 1909.
In the report of Charcot's second expedition the list of maccormicki contains no corre-
sponding data, but two examples, one of each sex, of Megalestris antarctica are included
as having been killed at Deception on the date quoted (Gain, 1915, pp. 109-10). Since
it is obvious that these two birds were quite accidentally added to the list of maccormicki
which was sent to Lowe and Kinnear from Paris, I include them in my list of Brown
Skuas from the South Shetlands.

Since there are only twelve specimens caution must be observed, but on reference
to the figures it will be noticed that there appears to be a natural tendency for the birds
to fall into two groups of six each, one of which has a tarsal length of over 70 mm. while
in the other it is under 70 mm. The two groups are separated in the Table (p. 167).

The interesting point is that the measurements of the larger group, that is, with the
longer tarsus, incline towards those of lonnbergi, while those of the smaller birds are
nearer to the dimensions of the Falkland Islands subspecies, C. s. antarctica.

Plumage distinctions between the two South Shetland groups are non-existent, they
do not differ materially from any other series of skins : one of the smaller birds, it may
be remarked, has a particularly bright collar such as is found in some Falkland specimens.

CONCLUSIONS

There is found in the Falklands a smaller form of Brown Skua which frequently has
a fairly bright collar of yellow acuminate feathers on the neck ; the wing seldom attains
a length of 400 mm. and the tarsus is almost always less than 70 mm. long. In the New
Zealand area there is a larger bird which often lacks the yellow collar and has a tendency
towards a duskier plumage, the wing almost always exceeds 400 mm. and the tarsus does
not usually fall below 74 mm., but the differences in measurements and still more in
plumage are by no means sharply defined between the two groups.

The specimens from the Crozets, Kerguelen, South Georgia and the South Orkneys
are slightly smaller than the New Zealand birds, but the majority of the measurements
overlap markedly and on grounds of plumage differences they cannot be separated at all.

Birds from the South Shetlands appear to fall into two groups, each of which contains
darker and lighter forms ; in their measurements the smaller birds resemble the Falkland
Islands form, while the larger belong rather to the same group as the New Zealand
specimens.

The skuas of Tristan da Cunha and Gough Island, except for the slightly shorter
wing, are comparable to those of South Georgia and should be included with them.

At the risk of redundancy I would emphasize these two points :

(1) Plumage differences alone show themselves to be unreliable in separating the
four subspecies of Brown Skua recognized by Lowe and Kinnear.

(2) With the exception of the birds from the Falkland Islands, some from the South
Shetlands and one or two odd specimens besides, the measurements of all Brown Skuas

174 DISCOVERY REPORTS

fall within the limits of variation of the New Zealand race, although rather in the lower
part of the range of size.

There is no advantage in dividing up the Brown Skuas into four subspecies of which
two are so ill defined that they cannot be identified except by reference to the localities
on the data labels; the validity of subspecies should depend on the possession of
intrinsic characters. It follows therefore that groups which cannot be differentiated
from the New Zealand race should be included with it in one subspecies.

It is my opinion that the Brown Skua should be divided into two subspecies, a larger
and a smaller, which have a common meeting place in the South Shetlands. The names
and distribution are : C. s. antarctica (Lesson), of the Falkland Islands and the South
Shetlands, and C. s. lonnbergi (Mathews), of New Zealand, the Crozets, Kerguelen,
South Georgia, the South Orkneys, Tristan da Cunha, Gough Island and the South
Shetlands. C. s. clarkei (Mathews) and C. s. intercedens (Mathews) thus cease to exist,
part of the former and the whole of the latter being absorbed in C. s. lonnbergi and the
remainder of clarkei being included in C. s. antarctica.

NOTE ON THE DISTRIBUTION OF McCORMICK'S SKUA
L. Gain (191 5) has stated that he found McCormick's Skua on Deception Island,
South Shetlands, where it was mixing with the Brown Skuas but was more numerous.
He found maccormicki nesting as far north as Admiralty Bay, 620 6' S, and mentions
it as nesting at Port Lockroy in Wiencke Island, 64 ° 49J' S. In the course of a number
of visits to the South Shetlands by Mr A. G. Bennett and myself no specimen of
maccormicki has been collected, and in my observation the whole area as far south as 650
of latitude is occupied by the darker bird, i.e. C. s. antarctica.

The brown form is abundant and nests at Deception, and while Gain records having
seen it as far south as Wiencke Island at the south end of the Belgica Straits he states
that it did not nest in the Straits. In 1922, however, I found that Port Lockroy had
become a nesting site.

It may be suggested that the Brown Skua, which is slightly larger than maccormicki,
favoured by the abundant food supply derived from the extensive whaling operations
subsequent to 1909, had succeeded in ousting maccormicki from its former breeding
places in the South Shetlands, at least as far south as Port Lockroy. It is well known
that the skuas are most combative birds and even given to cannibalism at times.

LITERATURE

Clarke, W. Eagle, 1915. On the Birds of the South Orkney Islands. Scottish National Antarctic Expedition,
IV, pp. 219-47. Edinburgh.

Gain,L., 1915. Oiseaux Antarctiques . Deuxieme Expedition Antarctique Francaise, pp. 1-200, 15 pis. Paris.

Lowe, P. R. and Kinnear, N. B., 1930. Birds. British Antarctic ('Terra Nova') Expedition. Nat. Hist., iv,
pt. 5, pp. 103-93, 16 pis. London.

Mathews, G. M., 1913. Birds of Australia, 11, pt. 4. London.

Wilson, E. A., 1907. Aves. National Antarctic Expedition. Nat. Hist., 11, pt. 2, pp. 1-121, 13 pis., 46 text-
figs. London.

PLATE I

Fig. i. Catharacta skua lonnbergi (Mathews). Adults, Cumberland
Bay, South Georgia.

Fig. 2. Catharacta skua lonnbergi (Mathews). Chick about ten days
old, South Georgia.

Fig. 3. Catharacta skua, either lonnbergi (Mathews) or antarctica
(Lesson). Young chick, Port Lockroy, South Shetlands.

DISCOVERY REPORTS, VOL. IX

PLATE I

£*>*# * -l^'>

*- > •*"

CATHARACTA SKUA

[Discovery Reports, Vol. IX, pp. 175-206, Plates II-XIV, August, 1934.]

THE MARINE DEPOSITS OF THE
PATAGONIAN CONTINENTAL SHELF

By
L. HARRISON MATTHEWS, MA.

CONTENTS

Introduction page 177

Topography of the sea-bottom of the region J77

Distribution of sampling !79

Collection of samples J79

Analysis of the samples J8o

Coarser materials J8o

Apparatus 180

Method of use 180

Finer materials 181

Apparatus 181

Method of use 182

Distribution of the grades 184

Types of deposit 189

Description of the types 190

Distribution of the types 193

Distribution of the deposits by the texture of the samples as a whole . . 196

The silt grounds 197

Coarser grounds within the silt grounds 197

The fine sand grounds 198

Coarser grounds within the fine sand grounds 198

Discussion 200

Summary 201

References • 202

Tabulated data of the samples examined 203

Plates I I-XIV following page 206

THE MARINE DEPOSITS OF THE
PATAGONIAN CONTINENTAL SHELF

By L. Harrison Matthews, m.a.
(Plates II-XIV; text-figs. 1-3)

INTRODUCTION

This report is a description of the nature of the sea-bottom of the continental shelf
lying off the coasts of Patagonia and Tierra del Fuego south of lat. 430 S. This
region includes the Falkland Islands, and the Burdwood Bank which lies south of them.
It has been the subject of fishery surveys made by the R.R.S. 'William Scoresby',
when the region was explored by means of the large otter trawl, in addition to the usual
oceanographic gear, to ascertain whether it could sustain a commercial trawl fishery.
At many of the stations bottom samples were obtained with the conical dredge, and
these form the subject of this report. From the biological standpoint the type and
texture of the bottom deposits are of great importance when considered in relation to
the animals and plants found living on or near the bottom, or burrowing into it. For
this reason the deposits are here classified and charted into grounds showing the type
and size of the particles that form them, providing data of the habitats of the organisms
living on or near the bottom in the region examined.

TOPOGRAPHY OF THE SEA-BOTTOM OF THE REGION

The region from which bottom samples have been examined comprises the Pata-
gonian continental shelf south of lat. 430 S, and part of the Burdwood Bank. The
Patagonian continental shelf includes the Falkland Islands. Plate III shows the bathy-
metric configuration of the region. It is constructed from the soundings taken by the
R.R.S. 'William Scoresby', with additions from the Admiralty charts of the region.

The limit of the continental shelf may be taken as the 200 m. contour. This contour
roughly follows the sixtieth meridian from lat. 43 ° to 460 S, then bends westward to the
neighbourhood of long. 61 ° W and follows this meridian to 490 S. Here it turns south-
east to lat. 510 S and then skirts the outline of the Falkland Islands, following the east,
south and west coasts to lat. 510 S in the neighbourhood of long. 620 W. At this point
it turns south-west and runs towards the mouth of the Strait of Magellan as far as
640 30' W, where it again turns southerly and runs down to the eastward end of Staten
Island.

South of the Falkland Islands, between lat. 540 and 550 S and long. 560 and 620 W,
lies the Burdwood Bank, which shoals to less than 50 m. The bank is separated from the
Falkland Islands and the Patagonian continental shelf by deep-water channels which run

i78

DISCOVERY REPORTS

in from the south and west (sections E, F, G, Fig. i), and join to form a bay cutting into
the continental shelf between the Falkland Islands and the Patagonian coast.

C SAN DIEGO

C.ORFORO

1000m

500u

Fig. i. Sections of the Patagonian continental shelf. A. Section in lat. 430 S. The continental shelf
lies chiefly between 50 m. and 100 m. The bottom slopes less gently from 100 m. to 200 m. and then
descends steeply. B. Section in lat. 470 30' S. The continental shelf lies chiefly between 100 m. and
200 m. The bottom slopes less gently from 200 m. and 500 m. and then descends steeply. C. Section
in lat. 500 S. The continental shelf lies chiefly between 100 in. and 200 m. It slopes less gently to 400 m.
and then descends steeply. D. Section in lat. 520 S. The section passes through the Falkland Islands
and shows the proximity of the islands to the edge of the continental shelf. West of the Falkland Islands
is seen the northern end of the bay of deeper water formed by the meeting of the channels separating the
Burdwood Bank from the Falkland Islands and Tierra del Fuego. E. Section in lat. 540 10' S. The
section passes through the Burdwood Bank and the coast of Tierra del Fuego. It shows a small bank
lying in the deeper water channel separating the Burdwood Bank from the continental shelf. F. The
section is taken on the sixty-fifth meridian between lats. 490 55' S and 540 45' S and shows the Burdwood
Bank separated by deep water from the continental shelf south of the Falkland Islands. G. Section from
Cape Orford, West Falkland Islands, to Cape San Diego, Tierra del Fuego. The section runs approxi-
mately NE to SW, passing NW of the west end of the Burdwood Bank, and crosses the bay of deep
water that runs northwards from the channels separating the Burdwood Bank from the Falkland Islands
and Tierra del Fuego.

The 100 m. contour lies at a distance of 60-100 miles from the coasts of Tierra del
Fuego and Patagonia as far north as lat. 450 S. At this latitude it trends away from the
coast in a north-easterly direction and approaches closely to the 200 m. contour east of
long. 6o° W. The northern part of the continental shelf is thus mostly under 100 m.

MARINE DEPOSITS OF PATAGONIA i79

deep (section A, Fig. i), while the southern part is mostly 100-200 m. deep (sections B,
C, Fig. 1). The 100 m. contour round the Falkland Islands lies close to the coast so
that the shores of the islands descend comparatively steeply to the level of the continental
shelf, which on the east, south and west coasts of the islands is narrow.

Outside the limit of the continental shelf at 200 m. the bottom north of lat. 460 20' S
runs steeply down to depths of 2000 m. and more (section A, Fig. 1). From lat. 460 20'
to 500 S the 200 m. contour lies farther to the west and the seaward slope of the shelf
is more gentle to 500 m., below which it dips steeply (sections B, C, Fig. 1). South of
lat. 500 S the Falkland Islands lie close to the edge of the continental shelf and the steep
slope from 300 m. downwards is near the coast (section D, Fig. 1).

The gradient of the continental shelf from the coast to the continental slope is very
gentle: in lat. 430 S it is only 1 in 2690 for a distance of 160 miles from the coast out to
the 100 m. line, and in lat. 500 S it is 1 in 3150 for a distance of 340 miles from the coast
out to the 200 m. line.

DISTRIBUTION OF SAMPLING

The area of the region from which samples have been examined amounts to about
185,000 square miles, excluding the 6500 square miles of the Falkland Islands. The
samples number 112 and are fairly evenly distributed, so that each sample represents
on the chart about 1650 square miles of the sea-bottom. This distribution of sampling
cannot give any detailed picture of the sea-bottom, but as the region is characterized by
great uniformity over large areas the outlines of the nature and configuration of the
bottom as sketched in this report may be taken as being approximately correct. Plate II
shows the positions of all the stations from which bottom samples were received.

COLLECTION OF SAMPLES

The samples were collected with the conical dredge, which has a mouth 18 in. in
diameter and a canvas bag. The dredge has a heavy metal lip which cuts into the bottom
so that a large sample is collected from some 6 in. to 1 ft. below the surface of the
ground. The dredge is described in detail by Borley (1923).

When the dredge is emptied on deck the sample preserved is taken from the middle
of the mass of material turned out, so that the portion of the dredging that was upper-
most during hauling, and consequently may have had some of the finer deposits washed
out, is rejected. The samples were placed in wide-mouthed screw-capped jars and
preserved in spirit.

A haul with the conical dredge was taken as a routine at nearly every station made
by the R.R.S. 'William Scoresby' during her survey of the region surrounding the
Falkland Islands, and the samples then obtained are the collection forming the subject
of this investigation.

i8o

DISCOVERY REPORTS

ANALYSIS OF THE SAMPLES

For the sake of uniformity, and to afford comparison with other similar investigations,
the standard grades of texture used by Allen (1899) and Borley (1923) for the analysis
of bottom samples were adopted.

The coarser materials were passed through sieves, and the finer materials were
levigated, so that they were divided into grades as follows :

Material over 15 mm. in diameter.

,, 10 mm. and under 15 mm. in diameter.

5 - IO

(I)

Large fragments

(II)

Very coarse gravel

(III)

Coarse gravel

(IV)

Medium gravel . . .

(V)

Fine gravel

(VI)

Coarse sand

(VII)

Medium sand . . .

(VIII)

Fine sand

(IX)

Silt

2-5

i-o

°-5
o-i

5
2-5

i-o
°-5

Material under o- 1 mm. in diameter.

Borley 's (1923) method of separating the grades consisted in twice washing the
sample to free it from salt, and drying it. A portion of this was then weighed, sifted,
usually in water, and the grades produced dried and weighed. A portion of the material
passing through the finest sieve was then weighed, moistened and levigated. The re-
sulting grades, except the finest, were dried and weighed. The finest material was
rejected and its weight arrived at by subtraction.

In this investigation a technique that avoids the repeated dryings and weighings was
devised. The sifting and levigating is combined with the washing, and the grades are
weighed once only, and by addition give the total weight of the sample examined.

COARSER MATERIALS
Apparatus
A series of sieves made of perforated zinc with circular holes 15, 10, 5, 2-5, 1-5, i-o,
and 0-5 mm. in diameter were used for separating grades I-VII from each other and from
the finer materials. The diameter of the sieve plates was 15 cm.; the sides were 7 cm.
high and tapered slightly so that the mouths of the sieves were 16 cm. in diameter. The
sieve plates were 1 cm. from the bottom of the sides, which thus rose 6 cm. above them.
Round the mouth of the sieves was fitted an india-rubber ring with a groove in it so
that it slipped on to the rim. The sieves could be stacked one above the other, each
slightly projecting into the one below, the india-rubber rings on the rims making a
watertight joint between each (Fig. 2).

Method of use
The sample to be analysed was thoroughly stirred and a portion was placed in the
upper one of the stack of sieves. A strong stream of water was then directed through the

MARINE DEPOSITS OF PATAGONIA

181

Fig. 2. Sieves stacked one above the other.
a, rubber ring; b, sieve plate.

sieves from above while they were vigorously shaken. The material was thus quickly

sorted into its components of various diameters down to 0-5 mm. in diameter, and at

the same time it was thoroughly

washed free from salt. The material

finer than 0-5 mm. in diameter

passed through the finest sieve at

the bottom of the stack and was

discarded. The sieves were then

separated, and their contents dried

and weighed. The figures obtained,

in conjunction with those obtained

from the levigator, were combined

to show the percentage of the

grades in the total sample.

FINER MATERIALS

A portion of the untreated
sample was placed in a sieve with
holes 0-5 mm. in diameter. This
was then thoroughly agitated in a
vessel of water so that all the material finer than 0-5 mm. in diameter passed through
into the vessel, and the coarser material of all grades over 0-5 mm. in diameter remained
in the sieve. The latter material was then well washed by a stream of water over the
vessel, dried and weighed. The finer material that passed through the sieve was then
levigated and thus separated into the two finest grades as described below.

Borley ( 1 923) found that in North Sea deposits the separation of grades VIII (fine sand)
and IX (silt) could not be readily effected by decantation and so made use of a levigator.
The method of levigation was adopted in this investigation as being more accurate and
reliable than other methods. Schone (quoted by Borley) estimated that a vertical cur-
rent of water travelling at 7 mm. per second would separate spherical quartz particles
over o-i mm. in diameter from those under o-i mm. in diameter. Levigation by a
vertical current of this speed was found by Borley to give a very good degree of accuracy
(5 per cent) with North Sea deposits, and is the method adopted in this investigation.
The levigator used by Borley separated the material analysed into four grades, 0-2 mm.
in diameter and over, o-2-o-i mm. in diameter, 0-1-0-05 mm. in diameter, and 0-05 mm.
in diameter and under. The first two of these grades taken together formed grade " fine
sand" and the last two the grade "silt". The levigator used in this investigation separ
ated the material only into the two grades fine sand and silt, containing particles over
o-i mm. and below o-i mm. in diameter respectively.

Apparatus
The levigator devised for use in this investigation is a development of that used by
Borley (1923), which in turn was modified from that of Schone.

182

DISCOVERY REPORTS

The apparatus (Fig. 3) consists of a tank a into which three tubes, b, c and d, pass
from below. The tube c passes into an inner receptacle, the walls of which are not as
high as those of the tank. Tube b is the water-supply pipe, tube c the waste pipe, and
tube d the delivery tube. The delivery tube d is
connected to a down tube/. The down tube /is
connected to a T-piece which in turn is con-
nected to a tube which passes through a hole in
bung j of receptacle i by a short length of tube
and two rubber junctions. The upper rubber
junction carries a screw clip e with a large knurled
head, and the lower one carries a spring clip h.
The side limb of the T-piece is joined to the gauge
tube g, which is bent through a right angle to
bring it vertical. The bung/ of receptacle i has
two holes, to one of which is connected down-
tube/, as described above, while to the other is
connected the levigator tube n by means of a
short length of tube and a rubber junction. At
its upper end the levigator tube n is joined to the
levigator funnel / by means of a rubber junction
bearing a spring clip. The rubber junctions at
the upper and lower ends of the levigator tube
are of the same internal diameter as the tube.
The levigator funnel / has a very short tube, of
the same diameter as the levigator tube, and
a wide tube 0 passes from the side of the upper
part of it. A wide-mouthed funnel m is fixed
below the outlet of the side tube. The apparatus
is levelled so that the tubes are vertical.

-g

f-

Method of use
When the water is turned on at the supply
pipe b the water rises in the tank to the level of
the top of the walls of the inner receptacle. The
excess water overflows into the inner receptacle
and is conducted away by waste pipe c. A con-
stant head of water free from the irregularities
of pressure of the supply is thus maintained above the delivery tube d. The water
flows down the tube/ into the receptacle i, rises up tubes g and n and fills the funnel /,
overflowing through tube o into funnel m. By means of the screw clip e the amount of
water passing through the apparatus is regulated until the rate of flow in the levigator
tube n is 7 mm. per second. The tubing used in this investigation was 8 mm. in

Fig. 3. Levigator. a, tank; b, supply tube;
c, waste tube; d, delivery tube; e, screw clip;
/, down tube; §, gauge tube; h, spring clip;
/, lower receptacle; /, bung; k, spring clip;
/, levigator funnel; m, wide-mouthed funnel;
n, levigator tube; o, side tube; s, removable
septum.

MARINE DEPOSITS OF PATAGONIA 183

internal diameter and the rate of flow was adjusted until 253-4 rnillilitres were
delivered at tube 0 in 12 min., this being the amount delivered at a rate of flow of 7 mm.
per second. The height of the water in the gauge tube g was then marked and the
apparatus was thus calibrated for the required rate of flow of water. Owing to the con-
striction caused by the screw clip e just beyond the junction of gauge tube g with the
down tube/, the water-level in the gauge tube varies, on the principle of the Venturi
water meter, with the velocity of the water flowing through the apparatus. If the
apparatus is to be used for rates of flow of very small amount the gauge tube g may
be placed at an oblique angle, instead of vertically, so that small changes in level cause
comparatively large movements of the meniscus along the gauge tube.

In using the apparatus a weighed filter paper was placed in funnel m, the clips // and
k opened and the water turned on, the correct level in the gauge tube being verified.
The fine material in the vessel of water, separated from the coarser grades as described
above, was well stirred and poured into the levigator funnel / and the vessel well washed
down into it. The material and water is poured in gently on the side of the movable
septum s, away from tube o, so that swirling does not carry over coarse particles through
this tube. Separation begins immediately, the larger particles passing down the levi-
gator tube n and accumulating at the bottom of receptacle i, while the finer material
remains in the levigator funnel /, or is carried over into funnel m where it is caught on
the weighed filter paper.

The amount of material placed in funnel / must be such that the filter m does not
become clogged and overflow before separation is complete. In practice this occurred
only on a few occasions with very muddy samples, when the levigation had to be
repeated with a smaller quantity of material.

When separation is judged to be complete clips k and h are closed, when the material
in tube n settles to the bottom of receptacle i. Receptacle i is removed and its contents
turned out into a funnel containing a weighed filter paper, and well washed down.
Levigator funnel /, with its rubber junction and clip, is removed and emptied into
funnel m and well washed down. The filter papers and their contents from funnel m and
from the funnel containing the material from receptacle i are then washed, dried and
weighed. This gives the weights of the grades fine sand and silt, and these weights,
together with that of the material from which they were separated, which has also been
dried and weighed, give the total weight of the sample analysed. From this the per-
centages of the grades fine sand and silt and the percentage of all the other grades taken
together are found. The weights of the other grades, as found from the sieved sample,
when added together are therefore equal to the latter percentage of the total. From this

. . , /percentage of grades together x weight of grade\ , _/,„4.„„/5 „r

can be calculated v- 2 § s _^_ _ the percentage ol

\ sum or grades /

each grade in the complete sample.

1 84

DISCOVERY REPORTS

DISTRIBUTION OF THE GRADES

The results obtained from the quantitative analysis are tabulated and charted to show
the distribution of the various grades of material present in the deposits. Each grade is
first considered separately and charted according to the percentage in which it is
present. Contours are then drawn round areas of similar percentages, thus defining the
grounds covered by each grade. Thereafter the texture of each bottom sample as a
whole is charted by means of an index number obtained from the proportions of the
various grades present in each sample. The grounds covered by deposits of similar
texture are defined by contours.

(I) LARGE FRAGMENTS (Plate IV)
(Material 15 mm. in diameter and over)

Most of the region under consideration is free from large fragments, which occur
mainly in isolated patches and not in extensive grounds. In the Gulf of San Jorge a
greenish grey Tuff (Campbell Smith, W., and Rayner, G., 1934) was picked up in some
quantity by the trawl, while off the coast to the south of Cape Tres Puntas and Point
Deseado lies a ground in which large fragments are present in quantity varying from
5 to 59 per cent of the deposits. Farther to the east lies a smaller area with 8-10 per cent
of large fragments. In lat. 450 13' S and long. 590 56' 30" W a patch containing 7 per
cent of large fragments lies on the edge of deep water.

Off the southern shores of the Falkland Islands large fragments form up to 48 per cent
of the deposits in the shallower water, but the proportion rapidly falls off to 12-15 per
cent between the depths of 200 and 300 m., and the grade is absent below 300 m. The
stations on the Burdwood Bank show up to 55 per cent of large fragments on the shal-
lower part with the percentage falling to 1 5 per cent by 200 m. and absent below. The
occurrence of outcrops of rock on the Burdwood Bank is regarded by Macfadyen (1933)
as probable, from the occurrence of loose fossil Foraminifera in the deposits.

Between the Falkland Islands and the entrance of the Strait of Magellan lies an ex-
tensive patch carrying 35-51 per cent of large fragments, while another smaller one lies
off the coast of Tierra del Fuego to the south of the entrance to the Strait.

On the line joining the Jason Islands to the south end of the large-fragment grounds
south of Point Deseado lie two areas of large fragments. The larger and easternmost
one, containing 16-60 per cent of large fragments, lies in the middle of the continental
shelf between lat. 500 and 510 S; while the smaller and more westerly one, carrying
1-6 per cent of large fragments, lies between lat. 490 40' and 500 S, to the south of the
Point Deseado ground.

The remainder of the stations from which large fragments were obtained are few in
number and widely scattered over the continental shelf south of lat. 500 S. At St. WS
783 in lat. 500 02' 45" S and long. 6o° 10' W, the sample obtained shows only 1 per cent
of large fragments, but two unsuccessful hauls were made at this station before the
sample was obtained. In the first two hauls the dredge was bent by rock. This may

MARINE DEPOSITS OF PATAGONIA 185

indicate the presence of large boulders or the outcrop of rock through the deposits.
St. WS 825 in lat. 500 50' S and long. 570 15' 15" W, with 35 per cent large fragments,
lies on the edge of the continental shelf to the north-east of the Falkland Islands.

(II) VERY COARSE GRAVEL (Plate V)
(Material under 15 mm. and over 10 mm. in diameter)

The distribution of very coarse gravel follows closely that of the large fragments, but
is rather more widely spread. The grade does not occur in the Gulf of San Jorge, but to
the north a small patch lies close to the coast. The large-fragment ground off Cape
Tres Puntas and Point Deseado carries up to 9 per cent of very coarse gravel, and the
similar ground between it and the Jason Islands carries up to 13 per cent. Two southerly
extensions of the latter occur, one carrying up to 8 per cent of the grade, the other
1 per cent.

The south coasts of the Falkland Islands are bordered by an area carrying a small
proportion of very coarse gravel, and an extension of the area into deeper water occurs
towards the south-west. On the edge of the continental shelf to the north-east of the
islands a patch occurs bearing up to 6 per cent of very coarse gravel : this patch coincides
with the area of large fragments in the same place. The Burdwood Bank shows an area
of this grade extending into deep water towards the north-west, but the grade is absent
from the northern slope.

Two small patches of very coarse gravel occur between the Falkland Islands and the
entrance of the Strait of Magellan, whilst another occurs off the coast of Tierra del
Fuego south of the entrance of the Strait.

Very coarse gravel occurs elsewhere in the region only in isolated patches, par-
ticularly towards the edge of the continental shelf north of the Falkland Islands.

(Ill) COARSE GRAVEL (Plate VI)
(Material under 10 mm. and over 5 mm. in diameter)

This grade occurs over a large part of the region under consideration. A line drawn
from Cape San Diego to a point in lat. 500 S long. 6o° W and thence to Cape Tres
Puntas approximately encloses to the westward the main area of distribution of coarse
gravel. A bay in the northern part of this area is free from the grade, as is also a coastal
strip extending from Desvelos Bay to Point Gallegos. This large area of coarse gravel
does not extend as far east as the Falkland Islands, but west of the islands it carries a
higher proportion, up to 18 per cent of the grade, than over the rest of the area where it
does not carry more than 5 per cent. North of the area of coarse gravel just described
the grade occurs only as isolated patches, small ones in the middle of the continental
shelf and at the extreme north, a larger one at the edge of deeper water, while the largest
is a coastal strip north of the Gulf of San Jorge.

Coarse gravel forms up to 8 per cent of the deposits on the rough patch at the edge
of deep water north-east of the Falkland Islands. It also occurs in small quantity on
the rough ground, south of the Falkland Islands, which extends in a south-westerly

,86 DISCOVERY REPORTS

direction towards deeper water. The grade forms up to 7 per cent of the deposits off the
southern entrance of Falkland Sound.

On the Burdwood Bank coarse gravel forms up to 19 per cent of the deposits on the
top of the bank, but the proportion drops to only 2 per cent as deeper water is ap-
proached to the north.

(IV) MEDIUM GRAVEL (Plate VII)
(Particles over 2-5 mm. and under 5 mm. in diameter)
This grade is widely distributed over the region examined. It is absent only from (1) a
belt about 60 miles wide running south-east from the Gulf of San Jorge to lat. 400 S
and then turning eastwardly to deep water, (ii) from a coastal strip extending from Point
Deseado to Point Gallegos and (iii) from the deeper water below the edge of the con-
tinental shelf to the north-east of the Falkland Islands.

The grade occurs over most of the region in small amounts, up to 5 per cent, but a
wide tongue in which the proportion is up to 24 per cent extends from the coast of
Tierra del Fuego in a north-easterly direction, narrowing to a point short of the Jason
Islands. A patch bearing up to 10 per cent of medium gravel lies south of the belt free
from the grade that runs south-westerly and westerly from the Gulf of San Jorge.

Deposits off the southern end of Falkland Sound carry 9 per cent of medium gravel,
as do those of a patch to the north of West Falkland Island.

The rough patch to the north-east of the Falkland Islands, on the edge of deep water,
bears 7 per cent of the grade, whilst the shallower part of the Burdwood Bank shows
16 per cent decreasing to 10 per cent on the slope towards deeper water to the north.

Four isolated patches bearing from 5 to 10 per cent medium gravel occur on the
continental shelf south of the Gulf of San Jorge.

The highest proportion of this grade in any of the deposits is only 24 per cent, and if
the contour separating the percentages up to 5 per cent from the percentages 6-10 per
cent is removed, the wide and even distribution in small proportion of medium gravel
over nearly the whole of the region is emphasized.

(V) FINE GRAVEL (Plate VIII)
(Particles over 1-5 mm. and under 2-5 mm. in diameter)

This grade occurs in small proportions over the whole region examined with the
exception of the Gulf of San Jorge, a coastal strip from Point Deseado to Point Gallegos,
and three isolated patches, one about the middle of the continental shelf east of Point
Deseado, another on the edge of deep water farther to the east, and the third in the
deep water lying between the Burdwood Bank and Tierra del Fuego.

Fine gravel forms up to 9 per cent of the deposits in coastal belts to the north and
south of West Falkland Island, and in two large grounds lying between the Falkland
Islands and Patagonia and Tierra del Fuego. It also forms up to 7 per cent of the
rough patch to the north-east of the islands. On the Burdwood Bank it is present up to
4 per cent on the shallower parts and increases to 12 per cent towards the deeper water
to the north. Three small areas carrying up to 10 per cent of this giade lie on the

MARINE DEPOSITS OF PATAGONIA 187

continental shelf westerly of Point Deseado, while another adjoins the area free from
the grade off Point Gallegos.

Fine gravel is nowhere present in greater amount than 12 per cent of the deposits
and is both widely and evenly distributed over most of the continental shelf.

(VI) COARSE SAND (Plate IX)
(Particles over 1 mm. and under 1-5 mm. in diameter)

This grade occurs in small amount over nearly the whole region examined, forming
not more than 6 per cent of the deposits practically everywhere. It is absent from the
Gulf of San Jorge and the area eastward of it, from the coastal belt stretching from Point
Deseado to Point Gallegos, from two areas on the continental slope, one north-east of
the Falkland Islands and the other east of Point Deseado, and from two small patches
near the outer edge of the continental shelf east of the Gulf of San Jorge.

Coarse sand occurs in larger amount, up to 12 per cent of the deposits, off the southern
entrance of Falkland Sound, off the west of West Falkland Island, and on the coarse
patch north-east of the Falkland Islands. It also occurs on the northern slope of the
Burdwood Bank, and in two isolated patches, one about 80 miles east of the entrance of
the Strait of Magellan and the other about 40 miles east of Point Gallegos.

(VII) MEDIUM SAND (Plate X)
(Particles over 0-5 mm. and under 1 mm. in diameter)

Medium sand occurs in much larger amount than any of the preceding grades in
nearly every part of the region examined. It is absent only from the Gulf of San Jorge
and the coastal belt running from Point Deseado to Point Gallegos, and from a patch
on the continental slope in lat. 490 42' S, long. 540 14' 30" W. Over most of the re-
mainder of the region it occurs in amounts forming up to 10 per cent of the deposits.
A large ground extending eastward nearly to long. 640 W from the coast between Point
Gallegos and the Strait of Magellan carries a proportion up to 37 per cent towards its
eastern end, while a smaller area to the south-east bears up to 15 per cent. A belt round
the western shores of West Falkland Island carries amounts up to 36 per cent, while a
patch south of East Falkland Island carries 28 per cent of the grade. The rough patch
to the north-east of the Falkland Islands on the continental slope carries 22 per cent,
while the northern slope of the Burdwood Bank carries 29 per cent of this grade. Five
small patches scattered to the north-west of the Falkland Islands, between them and
Point Deseado, carry amounts up to 77 per cent, and are the only places in the northern
part of the region with more than 10 per cent of medium sand in the deposits.

(VIII) FINE SAND (Plate XI)
(Particles over o-i mm. and under 0-5 mm. in diameter)

Fine sand is the characteristic component of the bottom deposits of the whole region
and occurs everywhere, usually in large amounts, forming 76-90 per cent of the
deposits over more than half the region. It forms 76-90 per cent of the deposits of the
continental shelf north of lat. 500 S, with the exception of (i) that part west of long.

!88 DISCOVERY REPORTS

65° W off Point Deseado and Cape Tres Puntas, (ii) the Gulf of San Jorge, and an area
lying north and west of it as far as long. 640 W, and (iii) a belt following a sinuous
course almost half-way across the continental shelf in a north-westerly direction towards
Cape Tres Puntas from the edge of the continental slope in lat. 500 S.

East, south and south-west of the Falkland Islands fine sand forms 76-98 per cent of
the deposits, excepting a coastal patch south of East Falkland Island and the summit of
the Burdwood Bank, while a wedge-shaped area carrying the same amounts extends
eastwards from the southern part of the Patagonian coast to long. 640 W.

A belt in which the amount of fine sand is smaller, from 5 1 to 75 per cent, crosses the
continental shelf from the north and north-west parts of the Falkland Islands to the
neighbourhood of Point Santa Cruz on the Patagonian coast. In the middle of the con-
tinental shelf this belt bears a smaller amount, 14-20 per cent, of fine sand on its
northern edge, while closer to the coast there is a patch carrying 22 per cent, abutting
on to the coastal region.

The coastal part of the northern section of the region which carries less than 76-98
per cent of fine sand in the deposits carries 5 1-75 per cent of the grade in the outer zone
off the Gulf of San Jorge and to the northwards ; while in the Gulf itself and off Cape
Tres Puntas and Point Deseado the amount drops to less than 25 per cent. A smaller
inner ground off Cape dos Bahias carries 47 per cent, and at the southern extremity of
the Point Deseado ground there is a patch carrying 57-62 per cent.

The sinuous tongue which encroaches on to the continental shelf from the east in
lat. 490 to 500 S carries 51-75 per cent of fine sand in its outer part, but this amount
decreases to 46 and 19 per cent in the central and narrower part and then increases to
56-68 per cent again at its termination half-way across the continental shelf.

The rough patch off the north-east of the Falkland Islands on the edge of the con-
tinental slope carries 34 per cent of fine sand in its outer and deeper part: the amount
diminishes to 24 per cent in its inner and shallower part.

A broad belt of irregular outline and sinuous course runs from the west of the
Falkland Islands, turning in a southerly direction to the coast of Tierra del Fuego, and
carries only 26-50 per cent of fine sand. At about its centre in long. 650 W and lat. 510
to 520 S there are two patches in which the amount rises to 58 and 86 per cent respec-
tively, separated by another in which it diminishes to 8 per cent only.

South of East Falkland Island a coastal belt runs east as far as long. 590 W, carrying
30 per cent of fine sand off the southern end of Falkland Sound ; the amount drops to
17 per cent in long. 590 W and rises again to 51 per cent as the belt trends away from
the coast into deeper water in an easterly direction.

The Burdwood Bank shows only 1 per cent of fine sand in the deposits on the summit,
but the amount increases to 19 per cent on the northerly slope.

A small area off Point Gallegos carries only 1 1 per cent of fine sand, but the amount
increases to 35 per cent farther to the east where it meets the wedge-shaped ground
carrying a high proportion of fine sand that extends eastwards towards the Falkland
Islands.

MARINE DEPOSITS OF PATAGONIA 189

(IX) SILT (Plate XII)
(Particles under ci mm. in diameter)

Silt occurs in the bottom deposits everywhere in the region under investigation, but
over most of it is present in small amounts. South of lat. 460 30' S it forms up to 5 per
cent of the deposits of the continental shelf, but north of this latitude lies a wide belt
where it is present in larger amounts, as it is also on the eastward part of the continental
shelf north of the Falkland Islands.

The Gulf of San Jorge contains a very high proportion of silt, 94 per cent, as does a
coastal strip off Point Gallegos with 89 per cent. East and north of the Gulf of San
Jorge there lies a wide area stretching as far east as long. 620 30' W which carries up to
45 per cent of silt. From the eastward edge of this area and extending in a southerly
direction to the edge of the continental shelf lies an area carrying up to 10 per cent of
silt. Towards the southern end of the latter area and stretching in from the continental
slope is a large patch carrying 17 per cent in its northern part and 83 per cent in its
southern part. North of the former area and eastwards on the continental slope the
proportion of silt in the deposits drops again to 3-5 per cent.

North of the Falkland Islands an area bearing up to 33 per cent of silt extends up the
continental slope and on to the continental shelf as far west as long. 620 50' W. Towards
its north-west extremity the amount of silt in this area drops to 6 per cent, while on its
south-eastern border lies a patch in which the amount drops to 10 per cent. A belt
containing 12 per cent of silt runs in from this area towards the north coast of West
Falkland Island.

In the bay of deep water north-west of the Burdwood Bank lies an area in which the
amount of silt in the deposits rises to 23 per cent, while on each side of the entrance of
the Strait of Magellan there is a patch containing silt up to 14 per cent in the northern
and up to 6 per cent in the southern one.

Silt up to 13 per cent of the deposits occurs in a small patch to the north-west of the
area extending in westwards from the continental slope north of the Falkland Islands.

Amounts higher than 5 per cent of silt in the deposits occur elsewhere only as a line
of patches extending north-west from the Jason Islands towards Point San Julian.
A coastal patch off the Jason Islands carries up to 25 per cent of the grade; farther to
the north-west two patches lie near the middle of the continental shelf and carry 6 and
7 per cent of silt respectively, while off Point San Julian a patch carries 6 per cent of silt
in the deposits.

Consideration of the distribution of the deposits based on the texture of each sample
as a whole is deferred until a description of the types of deposit found in the region has
been given.

TYPES OF DEPOSIT

Dried samples from all the stations were assembled and arranged into twenty-nine
types, which fall into six well-defined groups. The groups are here designated by an

,9o DISCOVERY REPORTS

initial letter, and the types within the groups by the addition of a numeral to the initial
(e.g. A4).

The types are described by their composition and colour and to a less degree by their
texture, but identification of the minerals composing them has not been attempted. The
description does not in all cases correspond with that given in the station list and noted
at the time that the station sounding was made. This is largely due to the fact that the
samples change colour, many of them considerably, in drying. Samples that are light
greenish grey for example when dry appear dark grey, almost black, when wet ; others
change similarly.

The groups are sharply defined from each other and the allotment of any given sample
to its group is obvious. The types within the groups, however, are -not in all cases so
easily separated and tend to merge into one another through intermediate types. A
strong family resemblance running through all the groups with the exception of group A,
and occurring in nearly all samples, is the green colour of the silt. This varies from
brownish and greyish green, through dark and light greens to yellowish green. It
appears, in many instances at least, to be due to a glauconitic substance: glauconite has
been identified by Macfadyen (1933) in some samples of bottom deposits from the
Burdwood Bank.

Groups recognized:

Group A. Chief components, shell and coral fragments. Contains five types:
samples from eight stations.

Group B. Chief component, white or yellowish-white sand. Contains four types.
Samples from ten stations.

Group C. Chief component, brown sand. Contains seven types. Samples from
twenty-six stations.

Group D. Chief component, greyish-brown sand. Contains five types. Samples
from twenty-six stations.

Group E. Chief component, grey sand and silt. Contains four types. Samples from
fifteen stations.

Group F. Chief components, green or greyish-green sand and silt. Contains four
types. Samples from twenty-six stations.

DESCRIPTION OF THE TYPES
GROUP A. SHELL AND CORAL FRAGMENTS

Type A 1 . Medium grades consist of fragments of coral (Turbinolidae) with a smaller proportion
of clear angular sand grains, shell fragments and grey gravel. Silt yellow, appearing to consist of
coral detritus. Two samples, WS 93, WS 802.

Type A 2. Medium grades consist of coral and shell fragments with dark grey gravel. Large
dark grey pebbles and echinoderm spines in the coarser grades. Fine sand and silt yellow, con-
sisting of coral and shell detritus. One sample, WS 86.

Type A3. Medium grades consist of coral and shell fragments; a large proportion of whole and
broken bivalve shells (Pecten and clam types), dark grey pebbles and gravel in the coarser grades.

MARINE DEPOSITS OF PATAGONIA 191

Some white sand grains in the finer grades. Silt yellow to yellowish green. Two samples, WS 83,
WS 246 (192 m.).

Type A 4. Medium grades consist of shell fragments ; grey stones, pebbles and gravel, with large
fragments of shell (clam, Pecten and brachiopod), especially in WS 825, in the coarser grades. A large
admixture of white and clear sand grains in the finer grades, which are similar to those of group B.
Silt buffish grey. This type consists of almost equal parts of types A 1 and B 1. Three samples, WS 84,
WS 228, WS 825.

Type A 5. Finer grades consist of shell, coral, and calcareous polyzoan fragments; brown and
grey stones much encrusted with calcareous Polyzoa, small bivalve shells, large fragments of
Polyzoa, grey and brown gravel in the coarser grades. Silt yellowish buff. One sample, WS 88.

GROUP B. WHITE AND YELLOWISH-WHITE SANDS

Type B 1 . A fine yellowish-white silver sand with very few dark grey grains. Silt yellow. One
sample, WS 823.

Type B2. A fine yellowish-white silver sand with a greenish tint owing to the green or greyish-
green colour of the silt. Some small shell fragments and dark and light grey sand grains. In WS 230
some small transparent green angular fragments. Six samples, WS 227, WS 229, WS 230, WS 781,
WS 782, WS 824.

Type B3. Fine yellowish-white silver sand with a greenish tint owing to the green or greyish-
green colour of the silt, differing from type B 2 in the larger proportion of large shell fragments and
dark grey stones and pebbles. In the finer grades a large proportion of transparent greenish grains.
One sample, WS 248.

Type B4. Fine yellowish-white silver sand conspicuously speckled with an admixture of dark
grey and black grains, and with a small proportion of transparent green grains. At WS 246 (267 m.),
pinkish grey pebbles and some brown and yellow gravel. Silt yellowish brown to greenish brown.
Two samples, WS 246 (267 m.), WS 250.

GROUP C. BROWN SANDS

Type Ci. Light brown speckled sands, with a few shell fragments in the coarser grades, and an
admixture of white, black and grey grains in the finer grades. Silt brown to brownish green. Six
samples, WS 219, WS 220, WS 787, WS 796, WS 808, WS 809.

Type C2. Darker brown sands of less speckled appearance and finer texture than type Ci.
A few small bivalve shells and worm tubes in the coarse grades of some samples, and a small ad-
mixture of black and white grains in the finer grades. A few Foraminifera in some samples. Silt
yellowish to brownish green. Six samples, WS 235, WS 765, WS 771, WS 774, WS 792, WS 801.

Type C3. Light brown sands with a large proportion of clear and white grains and a small pro-
portion of dark ones. Shell fragments and brown pebbles in the coarser grades. Silt brown. Three
samples, WS 96, WS 222, WS 772.

Type C4. Darker brown speckled sands with a large proportion of black grains as well as white
ones. Texture coarser than the preceding types of group C. Some samples with shell fragments, or
worm tubes made of sand, in the coarser grades. Silt brown to brownish green. Four samples,
WS 94, WS 226, WS 240, WS 806.

Type C5. Darker brown sands with a large proportion of black grains and fewer white ones.
Texture coarser than type C4., with black, grey, brown and yellow gravel. In the coarser grades dark
grey stones, coral, and calcareous polyzoan fragments, worm tubes both calcareous and made of
sand or gravel, small shells and echinoderm spines. Four samples, WS 95, WS 799, WS 807, WS 837.

Type C6. Brown sands with a large admixture of calcareous matter producing a conspicuous
speckling of white. Some brown and grey pebbles in the coarser grades, with coral and calcareous

i92 DISCOVERY REPORTS

polyzoan fragments, echinoderm spines, shell fragments, and worm tubes made of sand. A large
proportion of black grains in the finer grades, and many white foraminiferan shells which give the
characteristic speckling to the type. Silt light brownish green. Two samples, WS 804, WS 838.

Type C7. Light brown sand with an admixture of clear transparent angular grains, and a small
proportion of pink grains in the finer grades. The coarser grades, which form about half of the type,
consist of irregular and angular light brown stones, large fragments of shell, calcareous worm tubes,
with some echinoderm spines and coral fragments. Silt grey with a slight green tint. One sample,
WS 221.

GROUP D. GREYISH-BROWN SANDS

Type D i. Fine greyish brown sands, of even rather than speckled appearance, though composed
of a mixture of brown, black or dark grey, and white grains, with a slight greenish tint owing to the
colour of the silt. Small bivalve shells or worm tubes made of fine sand in the medium grades of
some samples ; some Foraminifera in the finer grades of others. Silt greyish and brownish green to
green. Eight samples, WS 78, WS 79, WS 217, WS 227, WS 764, WS 785, WS 786, WS 793.

Type D 2. Darker and coarser greyish-brown sands composed of black, dark grey, brown and
yellow grains. Large worm tubes made of medium and fine gravel, calcareous worm tubes and coral
in the coarse grades of WS 849. The coarser grades dark grey, light grey and brown gravels. Silt
yellowish to greyish green. Three samples, WS 80, WS 243, WS 849.

Type D3. Dark greyish-brown sands with marked speckling. Greyer and less brown than the
preceding types of group D. Dark grey, brown, and yellow or white grains are the chief components,
with a few Foraminifera in the finer grades. Worm tubes made of sand and gravel, calcareous worm
tubes, calcareous polyzoan fragments, gastropod shells and light brown pebbles are present in the
coarser grades of WS 811. Silt yellowish brown, through brown to greyish green. Five samples,
WS 77, WS 242, WS 811, WS 814, WS 817.

Type D4. Dark greyish-brown sands of more even colour than type D 3, with a higher proportion
of coarser grades consisting of dark grey and brown stones and pebbles, coral and broken bivalve
shell. Some angular fragments of brown stone in WS 848, and large worm tubes made of gravel in
WS 800 : Foraminifera are conspicuous in the finer grades of WS 91 . Silt yellowish to greyish brown.
Four samples, WS 91, WS 92, WS 800, WS 848.

Type D5. In this type the gravels and large fragments preponderate in quantity over the finer
grades. The finer grades are dark greyish-brown sands of rather even tint, the coarser grades con-
sisting mainly of dark and light grey, brown, and yellow stones and gravels. Shell fragments, coral,
calcareous encrusting Polyzoa and worm tubes, and worm tubes made of gravel, occur in the coarser
grades. Silt greenish yellow to greenish brown. Six samples, WS 225, WS 798, WS 803, WS 805,
WS 816, WS 850.

GROUP E. GREY SANDS AND SILTS

Type E i. Very fine sandy muds of even, light grey colour. No grades above fine sand occur, and
silt preponderates except in WS 776 where it forms about half of the sample. WS 777 is a stiff grey
clay which dries to solid lumps and not to powder, while the silt of the other examples of the type
dries into hard cakes, grey in colour, which bind the fine sand. Three samples, WS 776, WS 777,
WS 812.

Type E2. Fine muddy sands of light speckled grey colour. The sand grains are mostly black or
dark grey, and white, with a small admixture of brown grains not sufficient to give their tint to the
type. There are a few grey and yellow pebbles in the coarser grades. Silt greyish green to yellowish
green. Three samples, WS 90, WS 218, WS 814.

Type E3. Fine muddy sands of even dark grey colour. Dark and light grey, brown and yellow
stones and gravel occur in the coarser grades of some samples ; coral, both Turbinolid and an ar-
borescent form, shell fragments, papery worm tubes and echinoderm spines in others. The finer

MARINE DEPOSITS OF PATAGONIA 193

grades consist of dark and light grey grains with some brown, yellow, or white grains. Foraminifera
are present in some numbers in WS 833, while fine fragments of coral form a conspicuous part of
the finest grades of WS 766. Silt grey to greyish green. Eight samples, WS 89, WS 238, WS 245,
WS 762, WS 766, WS 788, WS 833, WS 834.

Type E4. Light grey gravel composed of irregular but slightly rounded fragments of soft grey
stone or very hard grey clay, with some greyish-brown pebbles and gravel in the coarsest grades.
Bivalve, gastropod, and brachiopod shells and fragments are numerous in the medium grades. Silt
yellowish brown. This type is quite distinct from all the other types of group E. One sample, WS 841.

GROUP F. GREYISH-GREEN SAND AND SILT

Type F i . Medium to dark coloured greyish-green muddy sands of fine texture, consisting almost
entirely of grades fine sand and silt. A few samples have small shell fragments and grey or brown
gravel in small proportions in the coarser grades. The finer grades are grey to greyish green, with
some admixture of white grains in some cases. Silt greyish green through green to light yellowish
green. Thirteen samples, WS 99, WS 211, WS 212, WS 214, WS 232, WS 236, WS 244, WS 756,
WS 773, WS 790, WS 791, WS 821, WS 839.

Type F2. Greyish-green muddy sands lighter than those of type Fi by reason of the larger ad-
mixture of yellow and white grains, as well as lighter grey and greyish-green grains. Fine in texture,
with a small proportion of shell and coral fragments, dark and light grey, white, and yellow gravel
in the coarser grades of some samples. Foraminifera occur in small numbers in some samples. Silt
dark greyish green through green to yellowish green. Nine samples, WS 76, WS 210, WS 239,
WS 775, WS 783, WS 784, WS 810, WS 819, WS 820.

Type F3. Darker greyish-green sands of coarser texture than type F2 with coral (Turbinolid)
forming a large proportion of the coarser grades in which there are also some grey pebbles. The finer
grades are mixed with coral detritus. Silt dark greyish green. Two samples, WS 215, WS 818.

Type F4. A light greyish-green type of coarse texture with some grey and brown stones and a
high proportion of gravels composed of grey, greyish-green, brown, and yellow pebbles. Small
fragments of shell and echinoderm spines occur, while Foraminifera are conspicuous in the finer
grades. Silt green. One sample, WS 97.

DISTRIBUTION OF THE TYPES

The types described above are distributed in well-defined areas mostly covering wide
expanses of the sea-bottom. Plate XIII shows the distribution of the types by group,
and the occurrence of the various types within the groups is described below.

GROUP A

The distribution of the types assigned to group A is confined to the southern part
of the region examined. They occur as a coastal belt off the west and south shores of
the Falkland Islands from a little north of the Jason Islands to long. 580 30' S on
the south coast, but not in Falkland Sound. A patch extends in a north-easterly
direction from the north coast of East Falkland Island to the edge of the continental
slope. In the northern part of the coastal belt the types are A 1, passing through A3 off
the south coast of West Falkland Island and Falkland Sound to A4 off the south coast
of East Falkland Island. The patch to the north-east of the islands is entirely A 4.

On the Burdwood Bank there is a patch on the northern slope consisting of type A 2.
Farther to the west, north of the western extremity of Tierra del Fuego, lies another
patch of deposits assigned to this group and consisting of type A 5.

3-2

i94 DISCOVERY REPORTS

GROUP B
The distribution of this group, like that of group A, is confined to the southern part
of the region, and occurs only to the north-east, east and south of the Falkland Islands.
A line extending in a north-easterly direction from long. 590 30' W off the north coast
of West Falkland Island forms the north-eastern boundary of the area, while a line
extending in a south-easterly direction from the edge of the coastal belt of deposits of
group A, south of the western extremity of West Falkland Island, forms the south-
eastern boundary. Within these boundaries deposits of group B cover the continental
slope and the coastal shelf, except for the coastal belt of group A deposits off the south
coasts of the Falkland Islands and the patch of group A deposits lying on the continental
shelf to the north-east of East Falkland Island. They also occur in the southern end of
Falkland Sound. The northern part of the area as far south as lat. 510 30' S consists
entirely of type B 2, which appears also to form the deposits on the continental shelf east
of East Falkland Island. On the continental slope south and east of the Falkland
Islands type B4 predominates, while a patch of type B3 occurs just below the 200 m.
contour west of Beauchene Island. The deposits in the eastern end of Falkland Sound

consist of type B 1 .

GROUP C

Deposits of group C cover a large area of the continental shelf. At the north of the
region examined they cover the continental shelf to the edge of the continental slope
nearly as far south as lat. 450 S, with the exception of a broad coastal belt south of
Delgada Point. These deposits are of the type C2, with the type C3 occurring at the
south-east point of the area.

A large area stretching west and north from the coastal belt of group F deposits
between Point Deseado and Point Santa Cruz covers about half of the continental shelf
in its southern part, and about three-quarters of it in its northern part. At about its
centre it reaches as far east as the edge of the continental slope. Types C 1 and C 2
occupy all the northern half of the area and the area on the western or coastal part of
the southern half. The eastern or outer part of the southern half consists of types of
higher index number, type C5 occurring to the north and C4 to the south. Type C6
occurs at the extreme south-east point of the area, while type C7 intrudes to the east
on to the type C 1 and C 2 portions of the area in lat. 48° 20' S, long. 650 20' W. North
of this lies a small area of type C3.

A third area bearing deposits of group C stretches north-east from the coast of Tierra
del Fuego as far as lat. 510 40' S, long. 650 W, separated from the entrance of the
Strait of Magellan by a coastal belt of group E deposits. The index number of the
types increases from type C4 in the northern part of the area through type C5 in the
central part, to type C 6 in the southern part.

GROUP D
The main area covered by deposits of group D lies in the southern part of the region
under consideration. Between Point San Julian and the entrance of the Strait of

MARINE DEPOSITS OF PATAGONIA 195

Magellan a belt of groups E and F lines the coast: outside this belt an extensive area of
deposits of group D stretches eastward as far as long. 650 W. Here it sends a wide limb
in a southerly direction towards Staten Island, outside the areas covered by deposits of
groups C and A, bounded on the east approximately by the 200 m. contour at the edge
of the bay of deeper water extending in a northerly direction between the Falkland
Islands and South America. North of this southerly-directed limb the area extends to
the east to about long. 6i° 40' W, where it approaches closely the deposits of group A
off the Jason Islands. Here a narrower neck joins it to an area occupying the centre of
the continental shelf between lat. 480 20' and 50 S. The northern part of this latter
area consists of type D4, the remainder of type D 1. The narrow neck joining it to the
rest of the area covered by deposits of this group consists of type D 5 which also occurs
in a northerly directed extension of the main area and along the adjacent eastern border.
Type D4 forms most of the southerly directed limb, with type D3 appearing on the
east and type D5 on the west of its northern half.

Types D 1 and D2 occupy the central part of the area, type D 1 occurring again near
the coast between Point Gallegos and Point Santa Cruz; the intermediate portion is
covered by deposits of types D3 and D4.

Another area covered by deposits of group D lies on the eastern third of the con-
tinental shelf between lat. 440 30' and 460 40' S. Between 460 and 460 40' S it bends
towards the south-east and extends out to the edge of the continental slope. The de-
posits of this area consist entirely of type D 1 . A further small area covered by deposits
of group D lies east of Cape Tres Puntas and stretches over the continental shelf as far
as the area covered by deposits of group F. It consists of type D5.

GROUP E

Deposits of group E occur in scattered areas, of smaller extent than those of most of
the other groups, in all parts of the region. The largest area covered by deposits of this
group lies off the coast between Delgada Point and Cape Tres Puntas, including the
Gulf of San Jorge. The northern part of this area consists of deposits of type E3, while
the Gulf of San Jorge and the area to the east of it consist of deposits of type E 1. East
of the Gulf of San Jorge at the edge of the continental shelf an area of deposits of
group E occurs; on the continental shelf the type is E3, whilst on the continental slope
it is E2.

A coastal belt running from Point Santa Cruz to San Sebastian Bay in Tierra del
Fuego consists of deposits of type Ei as far south as Cape Virgins, where type E2
appears, while the remainder off the entrance of the Strait of Magellan consists of
type E3.

On the Burdwood Bank deposits of the type E4 cover the western part of the summit
and adjoin deposits of type E2 lying on the western slope.

Two isolated patches of deposits of type E occur, one at the edge of the slope to
deeper water between the Falkland Islands and Tierra del Fuego in lat. 520 40' S, long.

196 DISCOVERY REPORTS

630 40' W, the other towards the outer part of the continental shelf in lat. 480 30' S,
long. 61 ° 50' W, both of type E3.

GROUP F

An extensive area of deposits of this group covers the deep-water region between the
Burdwood Bank and the Falkland Islands, and the Burdwood Bank and Tierra del
Fuego. It covers the floor of the bay of deep water stretching north between the Falk-
land Islands and South America, and becomes constricted to a narrow neck west of
the Jason Islands. To the north it opens out widely again and covers the continental
shelf north of the Falkland Islands east of long. 620 W as far north as lat. 480 20' S.
From this point the area stretches south to the north coast of West Falkland Island.
To the north-east deposits of group F cover the continental slope as far north as
lat. 470 S, where they again extend up on to the continental shelf as far west as long.
6i° 30' W.

In the portion of the area north of the Falkland Islands, the eastern parts on the con-
tinental slope and on the continental shelf north of West Falkland Island are covered
by deposits of type Fi, except in lat. 470 40' S where a patch of type F3 lies on the
continental slope. The remainder of this area on the continental shelf is covered by
deposits of type F2, while a patch of type F4 occurs at its north-west corner in lat. 490 S.

The narrow part of the group F area west of the Jason Islands consists of deposits of
type F2, as does the central and deeper part of the area between the Falkland Islands
and Tierra del Fuego. The northern part of the deep water bay and its eastern and
western slopes are covered by deposits of type F 1, while a patch of type F3 lies on the
western slope adjacent to the patch of type E3.

A wedge-shaped area of deposits of group F lies in the middle of the continental shelf
between latitudes 440 and 470 S, the point of the wedge being directed southwards. The
deposits are of type F 1 except at the pointed southern end of the area where they are
of type F2.

The third area covered by deposits of this group is a coastal belt running from Point
Deseado to Point Santa Cruz, consisting of type F2.

DISTRIBUTION OF THE DEPOSITS BY THE TEXTURE
OF THE SAMPLES AS A WHOLE

Having detailed the distribution of the various component textures and types of the
bottom deposits, the division of the region into grounds based on the texture of each
sample as a whole is now considered. The results are shown in Plate XIV.

In order to arrive at a value for the texture of the deposits which may be plotted com-
paratively the following procedure is adopted. In each sample the percentage of each
grade of material is multiplied by the minimum diameter of the particles occurring in the
grade, with the exception of the silt grade. The silt grade is omitted as the minimum
diameter of the particles composing it is infinitely low. The figures for each grade thus
obtained are added together and divided by 100, the resulting index number being the

MARINE DEPOSITS OF PATAGONIA 197

" representative number " of the sample. Thus a sample containing only large fragments
would have a representative number of 15, while one containing only silt would have a
representative number of o. All other proportions of the grades are represented by the
numbers between o and 15, the representative number increasing in value as the
coarseness of the texture increases. The samples are then referred to the grade con-
taining particles whose diameter in millimetres corresponds with the representative
number. Thus a sample with representative number 0-31 is referred to grade fine sand,
which contains particles 0-1-0-5 mm. in diameter, whilst a sample with representative
number 7-61 is referred to grade coarse gravel, which contains particles 5-10 mm. in
diameter. Samples with representative numbers below o-i are classed as silt. The
deposits are then plotted on the chart according to the grades to which they are
assigned.

The finest two grades, fine sand and silt, cover the greater part of the region con-
sidered and form extensive grounds, while the coarser grades occur in smaller grounds
and patches. A line running east from Cape Tres Puntas to the edge of the continental
shelf, and then turning southerly and following the edge of the shelf to the north-east of
the Falkland Islands, where it turns eastward again, divides the two main grounds of
the region, the silt grounds lying to the north and east, and the fine sand grounds to
the south and west of the line.

THE SILT GROUNDS

The silt grounds, about 60,250 square miles in area, are composed of deposits of the
groups C, D, E and F. Group E deposits occur in the western part, those of group F
on the continental slope and in the middle of the continental shelf. Deposits of group C
cover the northern part and a wedge-shaped area at the centre of the continental shelf,
while those of group D lie to the east of this latter area. North-east of the Falkland
Islands between lat. 500 and 510 S the silt ground extends over the edge of the con-
tinental slope on to the continental shelf and reaches in towards the north coast of the
Falkland Islands at the northern end of Falkland Sound. This part, both on the con-
tinental slope and shelf consists of deposits of group B.

Smaller coastal silt grounds about 800 square miles in area lie off Point San Julian
and Point Gallegos, the former with deposits of group F, the latter of group E. Another
small area of about 850 square miles lies on the continental slope west of the deep water
lying north-east of the Burdwood Bank. Its deposits belong to group F. Grounds of
coarser texture lie in smaller areas within the main silt ground as detailed below.

COARSER GROUNDS WITHIN THE SILT GROUNDS

There is a coastal belt of fine sand 2800 square miles in area, with deposits of group E,
between Cape dos Bahias and the Valdes Peninsula, while at the edge of the con-
tinental slope between lat. 450 and 470 S lies a ground of fine sand of about 1200 square
miles area. Its deposits are of group E to the north, and of group D at the south, with

198 DISCOVERY REPORTS

a patch of 750 square miles of medium sand of group E midway between the northern
and southern extremities.

THE FINE SAND GROUNDS
The fine sand grounds of the southern and western parts of the region examined
cover an area of about 106,500 square miles and carry deposits of all the groups
recognized. Deposits of group A occur off the west of West Falkland Island, of group F
north of this and in deeper water west and south of the Falkland Islands and the
Burdwood Bank, and off the coast north of Point Santa Cruz. Between lat. 470 30' and
48° 30' S deposits of type C occupy the continental shelf between long. 6i° and 640 40' W.
This group appears again a little to the south-west on the continental shelf nearer the
coast between long. 630 and 670 W and lat. 490 and 500 10' S. A further area covered by
deposits of group C lies on the continental shelf east of the Strait of Magellan and Tierra
del Fuego east of lat. 65° W. Its eastern boundary coincides with that of the grounds
of coarser texture on the edge of deeper water as described below. Deposits of group D
occupy a central area on the continental shelf between lat. 480 30' and 500 S, and also
the area of fine sand grounds between lat. 500 10' and 520 10' S and long. 630 20' and
68° W. Deposits of group E occur in a coastal belt between Point Santa Cruz and the
entrance of the Strait of Magellan, and on the western slope of the Burdwood Bank.
The area to the west of the Falkland Islands and the continental slope to the south of
them is covered by deposits of group B.

COARSER GROUNDS WITHIN THE FINE SAND GROUNDS
The find sand grounds are interrupted by many coarser grounds. North-east of the
Falkland Islands they are separated from the silt grounds by a ground of coarse gravel
about 1360 square miles in area on the continental shelf, from which a ground of
medium sand about 400 square miles in area extends down the continental slope into
the fine sand ground. The deposits of both these grounds are of group A.

A coarser ground about 5600 square miles in area, extending eastwards from Cape
Tres Puntas as far as long. 640 40' W and southwards as far as lat. 490 S, consists of
various grades of gravel. Off Cape Tres Puntas an area of about 1100 square miles
consists of medium gravel of group D. It adjoins an area of 950 square miles of fine
gravel off Point Deseado. Off Nodales Bay lies an area of about 850 square miles of
very coarse gravel, south of which lies an area of about 2300 square miles of coarse
gravel. The coastal portions of the last three grounds consist of deposits of group F,
while in the outer parts the deposits belong to group C, as do those of an area of about
470 square miles of medium gravel lying at the south-eastern extremity of these grounds
of coarser grade.

A little more than half-way between the last grounds and the edge of the continental
shelf lies a ground of finer texture. It extends in approximately a north-westerly and
south-easterly direction between lat. 480 and 490 10' S, and is about 2250 square miles
in area. The northern fourth of it, 470 square miles in area, consists of medium sand of
group C, the second fourth of about 560 square miles of fine gravel of group D, the

MARINE DEPOSITS OF PATAGONIA 199

third fourth of about 660 square miles of coarse sand of group E, and the southern
fourth of about 560 square miles of medium gravel of group F.

To the north-east of this ground and lying on the continental slope at the junction of
the silt and fine sand grounds lies a patch of about 400 square miles of very coarse
gravel of group F.

An extensive ground about 12,000 square miles in area and of coarse texture lies
between the Falkland Islands and the coasts of Patagonia and Tierra del Fuego. It runs
from the neighbourhood of Cape San Diego in a north-easterly and north-north-
easterly direction as far as lat. 500 S. The middle part of it is a comparatively narrow
belt, while the northern and southern ends expand to form wider areas. Most of the
northern area, of about 2400 square miles, consists of coarse gravel, of group D at the
eastern and western extremities and group C in the centre. The northern half of the
belt, about 2000 square miles in area, is of medium gravel of group D. The southern half
of the belt and most of the southern area, about 4700 square miles, consist of coarse
gravel, of group D on the continental shelf, while on the continental slope deposits of
group E occur in the northern part and of group F in the southern. An area of about
900 square miles of medium gravel of group A occurs at the southern extremity, and a
smaller area of about 200 square miles of coarse sand of group D lies at the western
border of the belt in lat. 520 S.

To the west of the northern area of the preceding ground lies another gravel ground,
about 2250 square miles in area, between lat. 500 and 510 S and long. 650 and 66° W.
It is wedge-shaped, with its point directed westwards, while its base towards the east is
deeply indented. The western part, 650 square miles in area, consists of coarse gravel
of group D, the eastern part of 1600 square miles of fine gravel, of group D in the south
and group C in the north.

A small ground of medium gravel, about 350 square miles in area, lies about 40 miles
east of Point Gallegos: its deposits belong to group D. A ground of about 450 square
miles of very coarse gravel of group E lies off the coast between the entrance of the
Strait of Magellan and San Sebastian Bay. About 60 miles east of this ground, half-way
between it and the extensive gravel grounds west of the Falkland Islands, lies a small
patch of about 650 square miles of coarse sand of group C.

A gravel ground lies off the south coasts of the Falkland Islands and extends down
the continental slope at its eastern and western ends. Its area is about 3250 square
miles. The western part of the ground, 980 square miles in area, off West Falkland
Island consists of medium gravel and extends over the continental slope. The eastern
part, about 1850 square miles in area, off the southern end of Falkland Sound and East
Falkland Island consists of coarse gravel, but does not extend down the continental
slope nor into Falkland Sound where the ground is fine sand of group B. All the portion
of this ground on the continental shelf carries deposits of group A, but where it extends
over to the continental slope the deposits are of group B. The eastern extremity of the
ground extends down the continental slope south of East Falkland Island and consists
of fine gravel of group B.

200 DISCOVERY REPORTS

The Burdwood Bank shows a ground exceeding iooo square miles in area of very
coarse gravel of group E on its summit, and one of about 850 square miles of medium
gravel of group A on its northern slope.

The main features of the texture of the bottom deposits shown by charting the
deposits according to the grade are the two chief grounds, silt to the north and east, fine
sand to the south and west. Extensive but smaller grounds of various coarser grades of
sand and gravel occupy large areas of the continental shelf west of long. 6o° W and south
of lat. 470 S. Though the areas covered by the various types of deposit do not coincide
with the grounds grouped according to texture, it is noticeable that the contours
delimiting deposit types frequently coincide in part of their course with those de-
limiting the grounds.

DISCUSSION

The sparseness of sampling in the region examined is such as to preclude any detailed
discussion of the causes of the distribution of the bottom deposits. The entire region is
characterized by the presence of enormous masses of fine sand and silt. A line running
first in a northerly and then in a north-westerly direction from the Falkland Islands to
the South American coast separates the silt grounds on the north and east from the
fine-sand grounds to the south and west. The currents and prevailing winds of the part
of the South Atlantic comprised in the region examined travel approximately from
south-west to north-east ; to these agencies and to the heavy swells which prevail the
segregation of the silt grounds to the north and east of the region is no doubt due. A
chain of coarser grounds also extends in unbroken line for nearly 300 miles from the
eastern extremity of Tierra del Fuego in a north-north-easterly direction as far as
lat. 500 S. Thereafter the line is interrupted, but patches of coarse material lie on the
course of the line for another 300 miles to lat. 45° S. Grounds of coarser texture also
extend in a north-easterly direction from the north-east corner of the Falkland Islands.
It is possible that the position of these grounds, extending in a south-westerly to north-
easterly direction, is correlated with the currents and prevailing winds of the region
which travel in the same direction. In support of this it is noticeable that the contours
enclosing the areas covered by the different types of deposit (see Plate XIII) in many
cases run in a south-westerly to north-easterly direction. This is particularly marked
in the area covered by deposits of group F which extends across the region from south-
west to north-east west of the Falkland Islands. On the other hand, the coarse grounds
stretching across the area in a north-easterly direction may also be correlated with rock
outcrops occurring there.

It appears probable that some at least of the deposits are being formed in situ on the
sea bottom and are not derived from the land or other regions. On the Burdwood Bank,
Macfadyen (1933) has deduced the occurrence of rock outcrops from the presence of
fossil Foraminifera in the deposits. The same deposits also contained portions of the
rock from which the fossils had been washed out. The occurrence of glauconite in the

MARINE DEPOSITS OF PATAGONIA 201

Burdwood Bank deposits is traced by the same author to the disintegrating rock, the
glauconite being of fossil and not of recent origin. An area similar to the Burdwood
Bank in this respect, lying in the south-west of the region considered, in which outcrops
of rock are being disintegrated by attrition and otherwise, would act as a reservoir from
which materials would be transported by currents and drifts to the north-east and
segregated into finer and finer grades the farther they travelled from their point of
origin. The glauconitic silts of the northern part of the region may therefore owe their
presence to some such combination of circumstances.

No very large rivers discharge into the region under consideration, but the deposits
in the coastal belts are doubtless in large part derived from the land. In the Gulf of
San Jorge the deposits have been shown by Campbell Smith and Rayner (1934) to be
in part of volcanic origin. The deposits contain a high proportion of finely divided
volcanic glass and feldspar derived from the volcanoes of the Andes and, together with
the clay of non-volcanic origin, the materials are in process of forming a sedimentary
volcanic tuff. The white sands occurring off the Falkland Islands would appear to be of
terrigenous origin and to be derived from the quartzite rocks of the islands. * A deposit of
similar type north of the extremity of Tierra del Fuego may be of much the same origin,
while the deposits of like type on the north slope of the Burdwood Bank would indicate
an outcrop of quartzite rock there, similar to the outcrops of shales found on the bank
farther to the east by Macfadyen.

SUMMARY

This report describes the marine deposits of the Patagonian continental shelf from
the standpoint of providing data regarding the habitats of the bottom-living fauna of
the region.

The topography of the sea-bottom of the region is described. Methods of separating
the deposits into their component grades of texture, and an improved form of levigator
for separating fine materials are described. The distribution of the various grades of
material in the deposits is given in detail.

Twenty-nine types of deposit, falling into six groups, were found : their characters
and distribution are described.

Finally the region is divided into grounds according to the texture of the deposits.
The distribution of the grounds, and the types of deposit found on them, are described.

Though fuller data are required for a detailed discussion of the causes of the present
distribution of the deposits, the probable effect of the currents of the region in segre-
gating finer materials towards the north is pointed out.

1 The quartzite rocks of the Devono-Carboniferous are shown by Baker (1924) to be characteristic of
West Falkland Island and the north part of East Falkland Island.

4-2

DISCOVERY REPORTS

REFERENCES

Allen, E. J., 1899. ®n tne Fauna and Bottom Deposits near the 30-fathom line from the Eddystone Grounds

to Start Point. Journ. Marine Biol. Ass., v, pp. 365-542.
Baker, H. A., 1924. Final report on Geological Investigations in the Falkland Islands, 1920-1922. Port

Stanley, Falkland Islands.
Borley, J. O., 1923. The Marine Deposits of the Southern North Sea. Fishery Investigations (Ministry of

Agriculture and Fisheries), series 11, IV, no. 6.
Campbell Smith, W., and Rayner, G., 1934. A Recent Sedimentary Volcanic Tuff. Nature, cxxxm,

p. 216.
Macfadyen, W. A., 1933. Fossil Foraminifera from the Burdwood Bank and their Geological Significance.

Discovery Reports, VII, pp. 1-16.

MARINE DEPOSITS OF PATAGONIA

203

"<3
S
I
H

<3

SO

-5:

a

C.ik O
^3

•g u

OS

55555s b >>« 5 sr5 5>>>x55£x555x5~5S~

V

u r 5

N h riOWO m \C fON O ■+■-$• ^t- n£> O* -<t- m r*« co co-O O O >-* *-> O >-< tJ-CO CO ■+

r* r*1 r* r1 r* r ^ ?°p r* v° 00 o *t MC-OMO'-'-'oor^.'-i-i^HHorir^M

ooooofi r^ co rf- b 6 vb^-<6 ONNfibboHobobbbbbbb

■luui 1.0 japun

<:»TTS XI

c> m m m o n

■uiiu £.o— 1.0
'puES auij -IIIA

•uiiu 6.0-S.0
'puES umipajAj -jIA

? 2 ??n° *** OTfrJ so sOOO in ON "n\0 f> O ro >-* -J- *t-co mo 1^ o >-< wi r*
i^so C*CO t^cn m hhc^m r^ co»ri*J- 0>O t^ ■* I-- f^OO «*>O0 0000 OiCiO>h00vO m

PO ~ PI N M ,_, _ PO - - MH M

M

c

2

s

V

•UIUI t.I-O.I
'pUES 3SJE03 'IA

_f_rn_^ri^co p*> H.r)i N n +& _i_ p» 4.

'+ + +" + +

n N tj- m

•iuui f.z-S.i

'[3AEJ§ 3UIJ 'A

■miu S.t-S.z
'[3aejS lunipaj^ -Aj

M PI Pi pi

r- _, o I ~ 1 i-l

I "+"++++++""

+ +

pi pi 0*

+ 2 IT' **" *t " ^ N I *"~ M °° I + I *+ "+ + + I *"'*"

•iuui 6-S

'|3AEj3 3SJE03 'III

+ +

•iuui £1-01

'J3AEJ8 3SJE0D iiaA -JJ

Ml I N I M

J3AO pUE -UIUI Si

'sju9uiSbjj gSjEq -j

r^ 00 i*i pi Tt-

ri rj- .h pt o

■s s

at!

£- O m el el r^ tn wcor*: N mu-jc^ o O^O v© m h h noo CvvO •-« u-^oo vD O -1- ri

(S-. MM^ MM MMM MM MNMMMNNMC1MM MMM

(2

bo

c

O O O O O1-1
o m en m co 4,

"- V V v ^ tJC

n m -rf a t^ t-r
o ro o 10 ro o

o o o o o ,h «

vO O \D \0 sO n, _,

o-2

J J

c
o

J

CO

ra^j

•^-rt-t-M OOOO O OOOCJ N if 1 in 10 N

co c/}pi3

0" ^ "Li "^ *L- w
M M M f-v flj

0000 m^

00000 c

M W M M O C

»fl If) Ifj If , 10 ^.

"aj o

"<*

co O t«

E

c

>i

o b

m o ° •
rt- roCC>

- - C/3^
COCO
in irj to

a % % %

10 u% m O

m m m ro

OOOOO
o "-. o o

00000

o co co ooo o 000 r^^o f< r>. t^co oo o o

NNhh r^co cocococooo^oococooococo

MNNNMrir.IrIMrjMMMMNMC)N

.£ > > >

N CO CO tJ- ^f-

^o r~» r^co o\ o^ o^ o m m h

mmmmmNNCOCOCO

N CO cn -4-00 O O

voce o o m r» co ^ u-i\o 1^.00 "t n •+ »n r-co o*o h n tin

COCOCO O OOOv cS0000-*MfMMMMMMriricidN

MNMNNMNNCIMCtMM

^f2^^^^ M CTJCOCfj CO COCOCO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO

££££££ £ £££ £ £££ ^«:jg^^^^^^^S:^?:^^^^^

204

DISCOVERY REPORTS

OS

^N*NN«N«mN^"N»com*m*~t<>«N<^Nm-NN>-«NNNN«'-.-< !•> «

<u

01 .&

<" £ 5

f^ml^M rl f^H-oooOOOO^O O^O O* O ^-00 r^OOOOM^OOriwOO~00<£«-1-0 n *o a

_, _ r^. _ _ ^ >-t o riMtHMMXhnnHt^HOroHHpH h o p pt1 c*s N h « c*i m p

bbbbbbboNbbobbtN^fiPibbbooMbbbbbbbbobooo'flOo

■luui i.o Japun nf.Mr) ^y^ r*> i-^ m in. n -t^nnM idn n o*o O^d m « no >on + 0 ^-o ^- N m h *£ n

'ins -xi "M ~+r' °°" *'*" ""

■UIUI t.O-I-O

■* m n r^ r~- « rift o o w » co oo ■(■niflM m ooo » ff» -oo r^o rro o co -f w o> wo f; i> w
oo »«ioiof- ooo I- w ■+■ o r^oo r»> ■*- r-» woo >o*»»oti coimsi o>»i oo cm~ r~ o> o» w *i-o

'LUUI 6.0-S-O

'puEs uinipaj^ "ii^v

r^ -1- r^o o *t *t- i-*o OMnoNMN

m m ^ _i_ m

+ ">+ +

nNni-MNMi'iM

•UIUI f.I-O-I
'pUES aSJEO^ 'l\

N+r-+HH+|w«oxMmM«Tj-M«+m« + +M+| |++| | +

• + +

•uiui f.z-S-i

'|3AEjS 3UUJ *y\

■uiui S.t"-£.z
'[3Aej3 uinip3j/\[ -^j

^+(i+l »+++nM««Hn+NH + +H+| | + +

+ +

-+ +

N I I OHnH|(HN»H++«lt(++«+| | + + | I + I " + + + " " +

•uiui 6-S

'|3AEJ§ 3SJECQ "J J J

•UIUI f'l-OI
'|3AEj8 3SJEO0 AJ3^\ 'JJ

J3AO pUE UIIU Si

's}u3ui3bxj aSiEq -j

+ +

r-l - i ■+ I INI IN

| N +

Q S

4n n « o a> m i> * o> * h •♦wo ao « w « o o « +o>o »nM - o ♦ ♦ w i^ w i-". o *^ o

HHroNNi-.MMNt-tMH-M"HrIfOMNNNM i- *« w rorlMt-ii-. *, M « *-< W M -.

o

be

C
O

~0 "o oo *b Vl "w"o "o "o "o *0 ~0 ~0 'o ~0 ~0 "o ~0 ~0 CO v- CO ~OvC - O m rl m N -^ O m O >o J >n * + O
^1 i>,\>i S4-i-«^W««mr^-f*O0t-)Oi^OinN"imOinOr'>ONU1N>HOC0i«NW^-

°n i> o Voo V- °o °" °n wo Vn V« °m So VcVh °o Vo °o °o °n Vm So So V- °n nV^n

O WW WW So O O so vO O O O O vO O WW WO O O WO O O O O O O w wo o o o o o o

CO

"o oo "o Wo V>^d *i#i"o No vm^o to "o o VVVaV'o °S.V*W%!IJ?2 ^ 1- ^^ ^ n ^^j J

pi o itj^m ri ir,tnnH u-,o O O f^fi N ^ i- ^ w n O « '+1-' NO i-n inr^pj o ^- N <

o o o o o o oo o ooo oo o o oo ooo o ooo oo o oo_o o_°_.0_ 0_.°_.0,
O « O O O a I>0 GC^M-.^NNNNM-0r2^:!^!^l^^ ^s£.^£. i2 Jl 9, & £ 2

Q

oooocooooooococooooocooooocooococoooco ,_, M M M w M ^ M _, ^

Tl^^NNNNNNNNNNNNNNNN^^^^^^^^^^^^^

n w fi w n f^ M f^

d n

>>>>>>

>>>?■>>■>>>•

s-gxxx««x *

nxx^^xxx

d « r^ wo o oo" wo t^ r^oo oddo>dd°"2^r CT^2 « ° m - n m mo ■* w wo ri t^- r-, 4

m „„__ — — — ^'in mm

CO

o r^oo o o N wo oo os o n f> ■+ wo o oo o o n ■*■ wo m N m •£ wo t-. « n m * wo r--oo o
n n n n £ A mrnm n + + t + + + + + "i wo ooo i^r^t^i^r^r^ r^oo oococoooooooco a-
n n n n "nnmnnnnnnnnnnn r-i^t^t^i^r^r-r-t^r^t^r^i^r-r-r-r-r-r-t^r-

MARINE DEPOSITS OF PATAGONIA

205

Type

of

deposit

-0 V

OS

Repre-
sentative
number

_MMMrr3f>.^M frj^j S£)r^Mr^'-'iHOrlO(strO'-iNi-'i-'HMriu^Mr^.'~«r')Oi-'OOI^ro

OOOO-I-O'hOOO pOsO Oi-'OOO^-OOOOOr^OOOOOOO-'i-'OOOOt-'-'fn

"3
0

^-*
0
o>

00

c

V

0
u

<u

a

3

s

h

•uiui 1.0 ispun
''IIS XT

m*0 00 r i PO i MMm 1-4 ri i M _i >h vO N O^i-'i^j^i-ir^dr^Tj-fOi-J-r^'-t n C* M " i N m
m T n T CO ~T "T _ PI T „ h

•uiiu io-1.0
'puEsauij -iIIA

"ON MM rJ-O N ©) r^.-^-— O p") i \D +Tt-inH ONOO sD Nh^'J-'l'M -J-00 ■^-t^t^'-' (*)m N ~ 00
GO O O C7- •-» mO OvO — c*) N OO On On On PO m r-%, 0O PO0O 0O r- OnOO N 1^ fOOO 00 On N !"■* N

•uiui 6.0-S.0
'puES umipaj^ -jia

N N m N m\D On i 00 O- t^ O -tOO PI m 1 in 1 OO N n'J-'J-M h nN N^O C*>nO On _i_ *t — O 00 0O

1" M N M | M | Hi H HH N ~T

•UiUI tM-O.I
'pUES 3SJE03 "JA

_J__[_ 1 1 PI "1+ 1 \£l M\fi H 1 MM H 1 f>.j MM_|_MI-l_|_|_|_MU^IHNNl_(__|_M«Nt^

'UIIU t'.E-S.I
'[3AEJ3 3UUJ -^

_[__!_ | 1 mor"*_i_mpi'i-mim_j |_| ©n| »-'nOin«-<_] i i h in i nx h i i^ + n"/!

•uiui S.t-S-Z
'|3aej8 uinipgjAj '^j

+ + 1 1 Z™ ^+M *° ?+=*+ + I 2 | - ^+*+ + " | +^+": M | 1 1™ mt

•uiui 6-S

'[3AEj3 SSJBOQ U[

_l_| 1 Hi-«,| loiniOiml |^|x«30>|i| |j.^|ul+x jOwiMin

•IUUI ^I-OI
'pAEig 3SJEO0 Al3\ "JJ

i i i i *+i i i -^«-« i i i - 1 i "i «i + i i r <>\ M i +"~M

J3AO pUE 'UIUI Si

's;u3ui§bjj aSjEq -j

1 1 1 1 ^ I °° 1 I S?S £ I «• " I 1 ?, 1 1 4 1 S 1 1 1 1 1 S 1 & 1 1 1 IK*0"*

■S 3

r*- pi o no On-h omN + ox O mo I"- moo mh o « nn « n o*o moo t^-oo oo moo On in t*^ r*-

on O m o "1- -+■ p*vsO m (Mn i- ro pi n o On On in h i/)0^i-*h mox -trnr^N 0> -J- O n© 0 "■ f^m

c

0
a,

oi

c
0

m ^O I^-Vl O 0 0 O m «1 m m 0 m

ino n n noo o>oo 0 m on o> i mn onoo « n OnO oniaon- m « r^tnooooo o onon>-h -toe o
inN-i-rt-o -+0 m no tN n o -t-N o O in tJ- m ii n mmino riio 0 NO n *tN pi inn

ooooooooooooooooooooooooooooooooooooooo
NNMp^inNNiip|ptr^-tin un\D t^-oo oo yO ■+ ■+■ po n i O ooo r^-oo oo^o inr^i OnD -i-^
nD nO nO nO ^O 'O nO ^O \C> ^O ">£> ^ *0 O ^O nO nO O nO *0 nO nO nO O nO nO nO •n ino vO ^ n© -O nO nO \0 nO ^O

en

in o f* in in in mininO O "mn mininm inpimmom inininmoinoinm

rf-po po m Th w -t -t m n m pi h +m m + hHini-i-m"- -tHTj-HprtnTj-Tj-

oo onoo On pi *f injD o p^pi o poo Ooof^ i^-nO ma^H w pnin-^-ONO O i^onm o n i r^-Noo
n^ mmmo "-> N *^-popi i O in-^-Pi i N m >»i-o n nfnmH pi mnm't ^ mini ninw

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

inmmN t-^00 XMOOOOO-^^aNHHMPiPlNNriNPlONNNn^ntOO-
i,ttttt"t,t-ininininu"i,tTl"Tt'tin|ninininininininininininininin"nininininin

2
Q

POPOPOroromPOPON NPINN PINNNNNNNNNN nnnnn^nnnnnnnnn

;i;nnnnnnnrnrnf*irnnnnnnnfnnnn«^^^n^MMn'*)

'x,^x,x'x,i<TS'H'H'H'-*-'"'H'M'H''"'"'H'"'"'"'"'rt'H*H,H,H:::::;:::':"::::::=:=:::

NHHHfltlNN

c
0

1

73

i n povO x a>o " n PTt mo i^-oo o- o m n -4-0 r^-oo ono "-> p*iTi-inpnTt- r^oo On 0 — 00 ON O

OnOnOnOnOnQnO 0 O O OO O O O O -i mh m m -HfcH m NNN NN P^P'jPnPOPO'rh'i-'^-'^-in

r* r>. r^ r^- r^ r^oo oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

DISCOVERY REPORTS, VOL. IX

PLATE II

STATIONS

MADE BY THE R.5.5.

WILLIAM SCORESBY'

MARCH 1927

TO

FEBRUARY 1932

* Cape Horn

-i ' ' ' ' ' i- i

45

-228
;227

!.FaUdaiid I.
/

J

56

55

65c

60c

DISTRIBUTION OF BOTTOM SAMPLES OBTAINED ON THE
PATAGONIAN CONTINENTAL SHELF

DISCOVERY REPORTS, VOL. IX

PLATE III

TOPOGRAPHY OF THE SEA BOTTOM

DISCOVERY REPORTS, VOL. IX

PLATE IV

DISTRIBUTION OF LARGE FRAGMENTS, 15 mm. AND OVER

DISCOVERY REPORTS, VOL. IX

PLATE V

DISTRIBUTION OF VERY COARSE GRAVEL, 10-14 mm.

DISCOVERY REPORTS, VOL. IX

PLATE VI

DISTRIBUTION OF COARSE GRAVEL, 5-9 mm.

DISCOVERY REPORTS, VOL. IX

PLATE VII

45

ABSENT
UP TO 5%

6%- 10%
1 1 %— 30%

LIMIT OF 0B5N-5

i i i i '

i'""F

I I'''

65°

GO°

DISTRIBUTION OF MEDIUM GRAVEL, 2-5-4-5 mm.

DISCOVERY REPORTS, VOL. IX

PLATE VIII

i i i i 1 1 i i i i i

ABSENT
UP TO 3%
4% -10%
11% -14%

LIMIT OF DBS N-5

65°

I > i i i i i |-<

45

5d

60c

i ' ' ' ' ■ i i i '

55

DISTRIBUTION OF FINE GRAVEL, 1-5-2-4 mm.

DISCOVERY REPORTS, VOL. IX

PLATE IX

DISTRIBUTION OF COARSE SAND, 1-0-1-4 mm.

DISCOVERY REPORTS, VOL. IX

PLATE X

60°

AB5ENT
UP TO 10%

ll%-20%

LIMIT OF OB5N-5

c^KVca

-i ' ' ' ' ' i i i

45

50

55

65°

60c

-1 ' ' ' ' ' 1-

-\ I I I I I fH

DISTRIBUTION OF MEDIUM SAND, 0-5-0-9 mm.

DISCOVERY REPORTS, VOL. IX

PLATE XI

60°

45

up to 25%
26%- 50%
51% "75%
76%-38°/o

LIMIT OF OB5"

-i i ■ ■ ■ i t-> ■ i " i

i | i i i i i i-t-i 1 1 i i

65°

60^

DISTRIBUTION OF FINE SAND, 0-1-0-4 mm.

45

50

-1 ' ' ' ' ' 1 1

55

DISCOVERY REPORTS, VOL. IX

PLATE XII

UP TO 5%

b%~ 10*
IIS-50S

51% -95%

LIMIT OF 0BSN-s

-j i-i I rxT[j irqiTi^t

65°

''''I i ' ' ' ' ' i =p

60°

-i i i ' ' "i

DISTRIBUTION OF SILT, UNDER o-i mm.

DISCOVERY REPORTS, VOL. IX

PLATE XIII

TYPES OF BOTTOM DEPOSIT

DISCOVERY REPORTS, VOL. IX

PLATE XIV

SILT

FINE SAND
MEDIUM SAND
COARSE SAND
FINE GRAVEL
MEDIUM GRAVEL
COARSE GRAVEL
VERY COARSEGRAVEL ■
LIMIT OF OBSERVATIONS

-i ■ ■ ' ■ ■ F j i i i i i ^

-1 i i i i ' (-■-■ ■ ■ ■ | i o=e=

I t ■ ' ' ' ' M

45

SO

55

65°

60°

DISTRIBUTION OF DEPOSITS BY THE TEXTURE OF THE
SAMPLE AS A WHOLE

[Discovery Reports. Vol. IX, pp. 207-214, September, 1934]

THE DEVELOPMENT OF RHINCALANUS

By
ROBERT GURNEY

THE DEVELOPMENT OF RHINCALANUS

By Robert Gurney

(Text-figs. 1-7)

The study of the development of the marine Calanoida presents peculiar difficulties,
inasmuch as very few species carry their eggs, and it is consequently rarely possible
to identify the nauplius by hatching from the egg. This has only been done in the case
of Calanus finmarchicus (Lebour, 1916; Gibbons, 1933).1 In other cases the series of
stages has been laboriously traced out by examination of plankton material, and it is
therefore not surprising that, of all the marine Calanoids, we know only the development
of the following :

Calanidae. Calanus finmarchicus.

Paracalanidae. Paracalanus parvus (Oberg, 1905).

Pseudocalanidae. Pseudocalattus elongatus (Oberg).

Centropagidae. Centropages hamatus (Oberg); C. kroyeri (Grandori, 1912); C.
typicus (Grandori, 1925).

Temoridae. Temora longicornis (Oberg).

Acartiidae. Acartia bifilosa, A. clausi (Oberg); A. clausi (Grandori, 1912).

There are, it is true, partial descriptions, particularly of copepodid stages, of a number
of other forms, but of complete life histories no more are known. I feel, therefore, that
no apology is needed for offering a description of the larval stages of Rhincalanus, a
genus of the Eucalanidae.

The material was derived mainly from a plankton sample taken by the R.R.S. 'Dis-
covery' at St. 278, off Port Gentil, French Congo, on August 8, 1927. This sample
contained numerous nauplii and early copepodid stages of R. comutus, but no adults or
copepodids older than stage IV. The later stages have been obtained from other samples
taken off the African coast, but in none of them have nauplii or early copepodids been
found. The nauplii of R. comutus are all in the last three stages, but a few nauplii of
R. nasutus were found in material from St. 93, off Saldanha Bay, south-west Africa,
which represent five of the probable six stages. These two series can conveniently be
taken together, as the differences between them are very small.

The nauplius of R. nasutus was identified by Giesbrecht (1893) by observation of the
moult to the copepodid stage, and he made use of this elongated nauplius to disprove
Claus' statement, which had gained general acceptance, that maxilla and maxillipede of
Copepoda are parts of one appendage. It is curious that Claus, who in 1866 had figured
an elongated nauplius of the same general form as that of Rhincalanus, did not himself

1 Since this was written, Mr A. G. Nicholls (1934) has published an account of the development of
Euc/taeta norvegica from the egg.

2IO DISCOVERY REPORTS

observe the quite independent origin of the two appendages. The nauplius figured by
Claus was attributed by him to Calanella {Eucalanus), but according to Giesbrecht it
was wrongly named, the nauplius of Eucalanus being like that of Rhincalanus . Similar
elongated nauplii are said by Giesbrecht to be found among Pontellidae.

With (1915, p. 46) has described copepodid IV and V of R. nasutus, and has figured
leg 5 of the male in these two stages. Schmaus and Lehnhofer (1927) have dealt with
the later copepodid stages of all three species of Rhincalanus and particularly with the
development of leg 5 of the male. They showed also that there are two forms of
R. cornutus (forma typica and forma atlantica) distinguished by the form of leg 5, and
that similar differences can be seen in copepodid stages IV and V.

THE NAUPLIUS STAGES

Five stages only are recognizable in the material, but it is almost certain that the
youngest is actually stage II.

Stage II, R. cornutus. Length 0-45-0-47 mm. (Fig. 1.)

C

Fig. 1.

D

Fig. 2.

Fig.

Rhincalanus nasutus, nauplius stages.

Stage II. Fig. 2. A, antennule, stage II; B, antennule, stage III; C, furcal region, stage III;

D, furcal region, stage IV.

THE DEVELOPMENT OF RHINCALANUS 2II

Body pear-shaped, tapering gradually backwards, with a dorsal line of segmentation
m the maxillary region. Furcal region with one pair of unequal setae, that of the right
side being the longer, and crossing dorsally over the left one. Labrum very large, more
or less rectangular, and fringed with fine hairs.

F'g- 3-

Fig- 4-
Rhincahinus cornutus, nauplius stages.

Fig- 5-

Fig. 3. Stage V, lateral view. Fig. 4. Stage VI, lateral view. Fig. 5. Stage VI, ventral view. 1,2,3
indicate lines of division between somites dorsally. a, position of anus.

Antennules (Fig. 2 A) with four apical setae, without aesthete; segments 1 and 2 not
clearly jointed, each with a small seta at end. The seta of segment 1 is very small, and
neither at this, nor in any later stage, has a proximal seta been seen on segment 2, although
two setae on this segment is the rule among Copepoda.

Antenna with large molar process on coxa ; exopod with seven setae.

DISCOVERY REPORTS

Ant.2
Mand-

Mandible segment i with a strong spine, but without molar process ; exopod with five
setae; endopod an undivided plate, with six setae or spines.

Stage III, R. cornutus. Length 0-65-0-70, average o-66 mm.

Body more elongated, the length about 3 J times greatest width. Furcal region
(Fig. 2 C) cleft, each branch bearing a short spine on ventral surface, a terminal spine,
and an inner terminal soft seta. The right spine is longer than the left, but only about
one-tenth length of body.

Antennule with six terminal setae, but no aesthete. (Fig. 2B.)

Antenna, exopod with eight setae.
There is no trace of the maxillule.

/\ Stage IV. Length: R. nasutus, 0-82-
0-89, average 0-85 mm. ; R. cornutus,
i-o mm.

Body now very slender, five times as
long as wide. Furcal region with two
additional pairs of lateral spines. In
R. nasutus the furcal spines have scarcely
increased in length, so that the right
spine is now shorter in proportion to
the body, and not greatly longer than
the left one (62 : 49). In R. cornutus it
is nearly one-third of the length of the
body (60 : 195) and about 2 1 times as
long as the left spine (57 : 22).

Mandible with strong molar process
on coxa.

Antennule with eight setae and aes-
thete, arranged thus: 5-A-3.1 In some
specimens of R. cornutus the maxillule

Fig. 6. Rhincalanus cornutus. Nauplius, stages V and VI. « represented by a slight fold with a
A, upper lip, stage VI, showing arrangement of hairs on small projecting spinule, but I have
body wall; B, mandible, stage V. seen no trace of it in R. nasutus.

Stage V. Length: R. nasutus, 1 mm.; R. cornutus, 1-2 mm.

Antennule with 7-A-4 terminal setae.

Maxillule distinct, as a small fold bearing a small seta.

Maxilla traceable as a faint fold with a minute spine.

Body distinctly divided into an anterior region including the somite of the maxilla,

1 The arrangement of the setae on segment 3 may be conveniently expressed in this way: the first
number being the setae behind the aesthete (A), and the second those in front of it.

THE DEVELOPMENT OF RHINCALANUS

213

and a posterior region in which three additional somites can be seen under the skin just
before the moult.

Stage VI. Length: R. nasutus, 1-16 mm. ;
R. cormitus, 1-33 mm.

I have seen only one specimen of R. nasutus
at this stage, which is probably below the
average length.

Antennule with setae 9-A-6.

Maxillule and maxilla projecting as small
lobes, the former with a few slender setae,
and the latter with two small spines. Maxilli-
pede represented by a small fold bearing
a spine and a seta, and borne upon the same
somite as the rudiment of leg 1 . Legs 1 and

2 both represented by bilobed rudiments
bearing small spines. The general form of
the copepodid appendages can generally be
traced under the skin, the rudiment of leg

3 being distinctly visible.

COPEPODID

St ageI, R. cormitus. Length i^i-i-jmm.

This stage differs from later stages in
having the head scarcely produced, and
without rostral processes. The somite of
leg 4 is separated, with a pair of dorsal
spines, and the somite of leg 5 can be seen
under the skin of the terminal somite. The
furcal rami are not separated from the somite,
and bear five setae. Of the three terminal
setae one is inserted rather dorsally and is
soft, presumably corresponding to the soft
seta of the nauplius, and the inner seta of
the left side, which is much the longest in
later stages and in the adult, is of the same
length as its fellow of the right side. It is a remarkable fact that it is the right spine
which is longest in the nauplius.

Fig. 7. Rhincalanus cormitus. Copepodid, stage I.
A, lateral view; B, furcal rami, dorsal view.

2i4 DISCOVERY REPORTS

LITERATURE

Gibbons, S. G., 1933. A study of the biology of Calanus finmarchicus in the north-western North Sea.

Fisheries, Scotland. Sci. Invest., 1933, no. 1, 24 pp., 3 pis., 2 text-figs.
Giesbrecht, W., 1893. Mittheilungen iiber Copepoden. VI. Zur Morphologie der Maxillipeden. Mitth. zool.

Stat. Neapel, xi, 83-102, 1 pi.
Grandori, R., 1912. Studi sullo sviluppo larvale dei Copepodi pelagici. Redia, Firenze, VIII, 360-457, 6 pis.
Grandori, R., 1925. Sullo sviluppo larvale di Centropages typicus, Kr. Riv. biol. Milane, vn, 137-145,

10 figs.
Lebour, M. V., 1916. Stages in the life history of Calanus finmarchicus experimentally reared by Mr L. R.

Crawshay in the Plymouth Laboratory. Journ. Mar. Biol. Ass. Plymouth, N.S., XI, 1-17, 5 pis.
Nicholls, A. G., 1934. The developmental stages of Euchaeta norvegica, Boeck. Proc. Roy. Soc. Edinb.,

liv, 31-50, 8 figs.
Oberg, M., 1905. Die Metamorphose der Planktoncopepoden der Kieler Bucht. Wiss. Meeresunt. Abt. Kiel,

ix, 37-103, 7 pis.
Schmaus, P. H., and Lehnhofer, K., 1927. Copepoda 4; Rhincalanus, Dana 1852 der deutschen Tiefsee-
Expedition. Systematik und Verbreitung der Gattung. Deutsche Tiefsee Exped., xxin, Heft 8,

PP- 355-400, 29 figs.
With, C, 1915. Copepoda I. Calanoida Amphascandria. Danish Ingolf Expedition, HI, 4, 260 pp., b pis.,

78 text-figs.

[Discovery Reports. Vol. IX, pp. 215-294, Plates XV, XVI, November, [934.]

NEMERTEANS

FROM THE SOUTH ATLANTIC AND

SOUTHERN OCEANS

By

J. F. G. WHEELER, D.Sc.

Bermuda Biological Station for Research, Inc.

CONTENTS

Introduction page 217

Methods 217

List of stations at which Nemerteans were collected, with the species

obtained 219

Systematic account 225

Part I. Nemerteans from Saldanha Bay, South Africa 225

Part II. Nemerteans from the Falkland Islands, South Georgia and the

islands and banks of the western South Atlantic Ocean 247

Part III. The pelagic Nemerteans 280

Notes on the distribution of the southern Nemerteans 288

List of literature 290

Index 293

Plates XV, XVI following page 294

NEMERTEANS

FROM THE SOUTH ATLANTIC AND
SOUTHERN OCEANS

By J. F. G. Wheeler, d.Sc.

Bermuda Biological Station for Research, Inc.

(Plates XV, XVI ; text-figs. 1-66)

INTRODUCTION

A large number of the Nemerteans described in this report were examined alive
directly after capture, and whenever opportunity occurred they were sketched to
show form and colour. This applies particularly to the littoral species from Saldanha
Bay, South Africa, and King Edward Cove, South Georgia, where I was engaged on
whaling investigations as a member of the Discovery Committee's scientific staff at
different periods from 1925 to 193 1. Many other specimens were collected in the nets
of the R.R.S. 'Discovery', R.R.S. 'Discovery II', and R.S.S. 'William Scoresby' in the
course of the investigations. Other members of the staff found time to make sketches
of unusual forms, although opportunities were far less frequent at sea, and on most
occasions a note of the colour and markings had to suffice.

I wish to express my thanks to Dr Kemp, Director of Research, for his interest and
for allowing me to work on the group, to my former colleagues who took especial care
in the collection of specimens, and to Professor D. M. S. Watson, who very kindly
placed a room and the facilities of his department at my disposal for the completion of
the work. That this has been accomplished I owe to my wife who prepared the hundreds
of sections — those necessary for comparison with the observations of other workers and
those to complete our knowledge of particular species and to make the fullest possible
use of the collection.

METHODS

The frequent collections of Nemerteans at South Georgia and from the whaling
station at Saldanha Bay were not recorded under station numbers. Most of them were
shore collections made in the interval between other work. The dates on which they
were made are included in the systematic account. At South Georgia a kelp grapnel
made of three shark hooks lashed together was used for tearing kelp roots from the
bottom. This was found to be a more efficacious method of capturing undamaged
specimens than the dredge or small trawl. The worms collected at sea were taken in a
variety of nets to which a key is given below.

The animals sketched and noted in life were preserved and numbered N 1, N 2, N 3

218 DISCOVERY REPORTS

and So on, for work later, since there was little time for section cutting and no oppor-
tunity for consulting the literature of the group. The ultimate preserving fluid was
75 per cent alcohol for nearly all specimens, but I found that fixation in Da Fano i
(cobalt nitrate and formalin — see Lee, 1928, p. 348) and subsequent preservation in
5 per cent formalin worked well with Linens corrugatus, especially the large red-brown
form, which does not bleach as readily with this treatment as it does in spirit. Chloral
hydrate was the usual anaesthetic. The crystals were added to the sea water in the Petri
dish containing the specimen. I found that small forms could be dealt with by sucking
them into a glass tube rather smaller in bore than their diameter and holding the tube
under hot water running from a tap. When the worms were blown out of the tube into
the fixing fluid they contracted very little and not only kept fairly straight but often left
their protective mucous coat in the tube and thus facilitated the subsequent examination
in cedar oil for deeply embedded eyespots. Bouin and Bouin Duboscq were frequently
used as fixing reagents. Corrosive sublimate both hot and cold was tried but was not
successful. On one occasion large specimens of Linens corrugatus were immobilized by
the natural freezing of the surface layer in the basin left overnight outside the station at
South Georgia. The animals were by no means dead although they were at first com-
pletely insensitive. As mentioned later in the notes on this species, the slackened mus-
culature of the semi-frozen animals threw light on differences that had been noticed in
the body form of preserved specimens.

When the material from the ships was being worked over the numbered series was
continued, since very few of the specimens could be identified at sight ; and it was used
throughout in the numbering of the serial sections. The slides were marked with a
diamond N 1, N 2, N 3 and so on, in addition to the serial number in each particular
series, and the number was, of course, added to the label replaced with the remainder
of the specimen.

The pelagic forms were fixed and preserved in 5 per cent formalin. Before sections
were cut the preserved specimens were measured and examined for form, colour and
markings. Eyespots were sought for by clearing in cedar or anilin oil. In the armed
species the armature was especially looked for, and variations in the number of stylets
noted.

Identification with previously described species has, however, not been an easy
matter, for variations occur in certain characters that have been used for specific de-
termination, for instance, in the number of nerves in the proboscis and the size of the
armature. The state of contraction of the body affects the relative thickness of the body
layers and the course of the lateral nerves when they leave the brain. Even the shape of
the brain and the position of the organs in relation to it can vary from this cause, though
I do not know to what extent. Body form and colour vary considerably in some forms,
the outstanding example being L. corrugatus, which appears to be the L. ruber of the
south, judging by the colour differences between individuals that it exhibits when alive
and the changes in body form that take place on fixation. If there are difficulties of
identification with live animals there are greater difficulties with preserved specimens

NEMERTEANS 219

whose life form and colour are unknown. Only by combining colour sketches and ob-
servation in life with anatomical work can the identification of the Nemerteans be made
with any degree of certainty.

The complete discontinuity of the African and Falkland sector Nemertean faunas is
reflected in this report by treatment in separate sections. It is indeed a striking fact that
this discontinuity should be so complete, for the predominant form of the south —
L. corrugotus — was first collected at Kerguelen, and it might naturally be expected to
appear in the fauna of the southern extremity of Africa. Instead there has been found
an extension of the Mediterranean fauna to the southern hemisphere as far as Saldanha
Bay. This was foreshadowed by the capture of Carinella annulata in Simon's Bay near
Cape Town, reported by Stimpson in 1856, and it suggests that the littoral Nemerteans
depend upon the continuity of land rather than upon ocean currents for their dispersal.

The pelagic forms fall naturally into a separate group on account of their structural
peculiarities. Some of these, such as their general transparency and expanded leaf-like
form, can be considered as adaptations to their mode of life ; but the anomalous position
of the male generative organs and the sexual dimorphism exhibited by forms like Necto-
nemertes must be due to deeper causes. The work of Brinkmann on the pelagic forms
shows how close is their relationship with the Drepanophoridae and it is curious that
this small family, alone among Nemerteans, should have taken to the pelagic habit.

The synonymy of most of the known forms has been thoroughly worked out by Burger
in his magnificent Naples monograph (1895) and in his section (Nemertini, 1904) of
Das Tierreich. Synonymies are only given in the following report where fresh data
warrant a revised opinion on the relationship of previously described species. I have
followed Coe (1905) in uniting the Protonemertini and Mesonemertini of Burger under
the order Paleonemertea, and in retaining Hubrecht's order Hoplonemertea for the
armed species instead of Metanemertini. The names are less unwieldy than those pro-
posed by Poche (1926). The further division of the Hoplonemertea into Monostilifera
and Polystilifera (Brinkmann, 1917) is adopted.

LIST OF STATIONS AT WHICH NEMERTEANS WERE
COLLECTED, WITH THE SPECIES OBTAINED

In the list of stations the following symbols represent the various kinds of gear used :

B Oblique.

BTS Small beam trawl. Beam 8 ft. in length (2-45 m.): mesh at cod-end J in. (12-5 mm.).

DL Large dredge. Light pattern, 4 ft. in length (1-2 m.).

DS Large dredge. Heavy pattern, 4 ft. in length (1-2 m.).

N 4-T Small dredge.

H Horizontal.

OI H I

Nets with mesh of 4 or 7 mm. (0-16 or 0-28 in.) attached to back of trawl.

N7-TJ

N 70 70 cm. tow-net. Mouth circular, 70 cm. in diameter (27-5 in.): mesh graded, at cod-end
74 to the linear inch.

22o DISCOVERY REPORTS

N ioo i m. tow-net. Mouth circular, i m. in diameter (3-3 ft.) : mesh graded. Cod-end of stramin

with 11-12 meshes to the linear inch.
N 200 2 m. tow-net. Mouth circular, 2 m. in diameter (6-6 ft.): mesh graded, at cod-end 4 mm.

(0-16 in.).
N450 4I m. tow-net. Mouth circular, 4! m. in diameter (14-8 ft.): mesh graded, at cod-end

7 mm. (0-28 in.).
NCS-T Tow-net of coarse silk, with 16 meshes to the linear inch, attached to trawl.
NH Hand net.

NRL Large rectangular net. Frame 8 ft. long and z\ ft. wide (2-45 m. x 07 m.) with bag of

J in. mesh (12-5 mm.).
OTC Commercial otter trawl. Head rope 80 ft. long (24-5 m.): mesh at cod-end i| in. (3-8 cm.).
OTL Large otter trawl. Head rope 40 ft. long (12-2 m.): mesh at cod-end ij in. (3-2 cm.).
OTM Medium otter trawl. Head rope 30 ft. long (9-14 m.) : mesh at cod-end i\ in. (3-2 cm.).
RM Mussel rake

Sh. coll. Shore collection.
TYF Young-fish trawl. Mouth about 20 ft. in circumference (6 m.) : bag of stramin with 11-12

meshes to the linear inch.
V Vertical.

R.R.S. 'DISCOVERY' AND R.R.S. 'DISCOVERY IT

2. xi. 25. 6° 55' N, 150 54' W. N 200, 0-800 m.
Pelagonemertes rollestoni, Moseley. Nectonemertes kempt, n.sp.

Crassonemertes robusta, Brinkmann.

St. 4. 30. i. 26. Tristan da Cunha, 360 55' S, 120 12' W. DL, 40-46 m.
Cerebratulus fuscus, Mcintosh.

St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, 3-3 miles S 440 E of Jason Light.
DL, no m.
Amphiporus lecointei, Burger.

St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, from 8 cables S 81 ° W of Merton Rock
to 1-3 miles N 70 E of Macmahon Rock. OTL, 179-235 m.
Amphiporus spiuosus, Burger. Linens corrugatus, Mcintosh.

N4-T, 179-235 m.
Tetrastemma esbenseni, n.sp.

St. 42. 1. iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 890 E of Jason
Light to 4 miles N 390 E of Jason Light. OTL, 120-204 m.
Tetrastemma longistriatum, n.sp.

St. 45. 6. iv. 26. 27 miles S 850 E of Jason Light, South Georgia. OTL, 238-270 m.
Amphiporus spinosus, Burger. Lineus corrugatus, Mcintosh.

St. 51. 4. v. 26. Off Eddystone Rock, East Falkland Island, from 7 miles N 500 E to 7-6 miles
N 630 E of Eddystone Rock. OTL, 105-115 m.
Tetrastemma georgiamim, Burger. Lineus corrugatus, Mcintosh.

St. 53. 12. v. 26. Port Stanley, East Falkland Island, Hulk of ' Great Britain '. RM, 0-2 m.
Tetrastemma hand, Burger. Lineus corrugatus, Mcintosh.

NEMERTEANS 221

St. 71. 30. v. 26. 430 20' S, 460 02' W. N 70 V, 1000-750 m. TYF, 2000 (-0) in.
Pelagonemertes rollestoni, Moseley.

St. 72. i.vi. 26. 41 ° 43' 20* S, 4a0 20' 40" W. N 450, 2000 (-0) m.
Pelagonemertes rollestoni, Moseley.

St. 76. 5. vi. 26. 390 50' 30" S, 360 23'W. N 450, 1500 (-o)m.
Pelagonemertes rollestoni, Moseley.

St. 78. 12. vi. 26. 350 18' S, 190 01' 10" W. TYF, 1000 (-0) m.
Pelagonemertes rollestoni, Moseley.

St. 79. 13. vi. 26. 340 48' S, 1 6° 36' W. N 450, 1000-0 m.
Pelagonemertes rollestoni, Moseley.

St. 85. 23. vi. 26. 330 07' 40" S, 40 30' 20" E. N 450, 2000 (-0) m.
Bathynemertes hubrechti, Brinkmann. Pelagonemertes rollestoni, Moseley.

St. 86. 24. vi. 26. 330 25' S, 6° 31' E. N 450, 1000 (-0) m.
Bathynemertes hardyi, n.sp. Pelagonemertes rollestoni, Moseley.

St. 87. 25. vi. 26. 330 53' 45" S, 90 26' 30" E. TYF, 1000 (-0) m.
Nectonemertes tnirabilis, Verrill.

St. 89. 28. vi. 26. 34° 05' 15" S, i6° 00' 45" E. TYF, 1000 (-0) m.
Bathynemertes hubrechti, Brinkmann. Probalaenanemertes irenae, n.sp.

Pelagonemertes rollestoni, Moseley.

St. ioor. 4. x. 26. 33° 20' to 330 46' S, 150 18' to 150 08' E. TYF, 2500 (-0) m.
Bathynemertes hubrechti, Brinkmann.

St. 101. 14. x. 26. 330 50' to 34° 13' S, 160 04' to 150 49' E. N 450, 1310-1410 m.
Bathynemertes hubrechti, Brinkmann.

St. 107. 4. xi. 26. 450 03' S, 170 03' E. N 450, 850-950 m.
Pelagonemertes rollestoni, Moseley.

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 540 E of
Larsen Point, to 1-2 miles S 620 W of Merton Rock. OTL, 230-250 m.
Amphiporus moseleyi, Hubrecht. T. georgianum, Burger.

A. spinosus, Burger. Parapolia grytvikenensis, n.sp.

Tetrastemma esbenseni, n.sp. Lineus corrugatus, Mcintosh.

St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, from 540 02' S,
360 38' W to 540 1 1' 30" S, 360 29' W. OTL, 122-136 m.

Amphiporus lecointei, Burger. Tetrastemma georgianum, Burger.

A. spinosus, Burger. Cerebratulus larscni, n.sp.

St. 141. 29. xii. 26. East Cumberland Bay, South Georgia, 200 yards from shore, under Mount
Duse. BTS, 17-27 m.
Tetrastemma longistr latum, n.sp.

St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, from 540 04' S, 360 27' W to
530 58' S, 360 26' W. NCS-T, 155-178 m.
Tetrastemma gulliveri, Burger.

222 DISCOVERY REPORTS

St. 156. 20. i. 27. 53°5i'S, 36°2i'3o" W. DLH, 200-236 m.
Amphiporus lecointei, Burger.

St. 15S. 21.1.27. 53° 48' 3°" S, 35° 57' W. DLH, 401 m.
Amphiporus lecointei, Burger.

St. 159. 21.1.27. 53°52'3°"s> 36° 08' W. DLH, 160 m.
Amphiporus lecointei, Burger.

St. 160. 7. ii. 27. Near Shag Rocks, 530 43' 40" S, 400 57' W. DLH, 177 m.
Tetrastemma iveddelli, n.sp. Linens corrugatus, Mcintosh.

St. 163. 17. ii. 27. Paul Harbour, Signy Island, South Orkneys. BTS, 18-27 m-
Tetrastemma longistriatum, n.sp. Linens corrugatus, Mcintosh.

St. 164. 18. ii. 27. East end of Normanna Strait, South Orkneys, near Cape Hansen, Coronation
Island. BTS, 24-36 m.
Linens corrugatus, Mcintosh.

St. 167. 20. ii. 27. Off Signy Island, South Orkneys, 6o° 50' 30" S, 460 15' W. N 4-T and
N 7-T, 244-344 m.
Linens longifusus, Hubrecht. L. corrugatus, Mcintosh.

St. 173. 28. ii. 27. Port Foster, Deception Island, South Shetlands, close to SE shore, near Lake
Point. BTS, 5-60 m.

Linens corrugatus, Mcintosh.

St. 175. 2. iii. 27. Bransfield Strait, South Shetlands, 630 17' 20" S, 59° 48' 15" W. DLH,
200 m.

Tetrastemma validum, Burger.

St. 179. 10. iii. 27. Melchior Island, Schollaert Channel, Palmer Archipelago, in creek to S of
SW anchorage. DS, 4-10 m.

Tetrastemma validum, Burger.

St. 182. 14. iii. 27. Schollaert Channel, Palmer Archipelago, 640 21' S, 620 58' W. N 7-T,
278-500 m.
Amphiporus schollaerti, n.sp. Baseodiscus antarcticus, Baylis.

St. 186. 16. iii. 27. Fournier Bay, Anvers Island, Palmer Archipelago, 64° 25' 30" S, 630 02' W.
DLH, 295 m.
Linens corrugatus, Mcintosh.

St. 195. 30. iii. 27. Admiralty Bay, King George Island, South Shetlands, 620 07' S, 580 28' 30"
W. OTM, 391 m.
Amphiporus lecointei, Burger. Linens corrugatus, Mcintosh.

St. 256. 23. vi. 27. 350 14' S, 6° 49' E. TYF, S50-1 100 (-0) m.
Pelagonemertes rollestoni, Moseley.

St. 283. 14. viii. 27. Off Annobon, Gulf of Guinea, 075 to 1 mile N 12° E of Pyramid Rock,
Annobon. DLH, 18-30 m.
Linens geniculatus (Chiaje).

NEMERTEANS 223

St. 395. 13. v. 30. 48° 26f ' S, 220 10' W to 480 26J' S, 220 o6£' W. N 450 H, 1500-1600 m.
Pelagonemertes rollestoni, Moseley.

St. 405. 4. vi. 30. 330 50-!-' S, 150 46' E to 340 16' S, 150 02' E. TYFB, 2200-0 m.
Pelagonemertes rollestoni, Moseley.

R.R.S. 'WILLIAM SCORESBY'

St. WS 4. 30. ix. 26. 320 45' S, 180 10' E. DL, 45-47 m.
Nemertopsis tenuis, Burger.

St. WS 25. 17. xii. 26. Undine Harbour (North), South Georgia. BTS, 18-27 m-
Amphiporus lecointei, Burger. Linens corrugatus, Mcintosh.

St. WS 56. 14. i. 27. Larsen Harbour, Drygalski Fjord, South Georgia. NH, 2 m.
Amphiporus spinosus, Burger. Lineus corrugatus, Mcintosh.

St. WS 62. 19. i. 27. Wilson Harbour, South Georgia. BTS, 26-83 m.
Amphiporus spinosus, Burger. Lineus corrugatus, Mcintosh.

St. WS 65. 22. i. 27. Undine Harbour (North), South Georgia. Sh. coll.
Amphiporus spinosus. Burger.

St. WS 73. 6. iii. 27. 510 01' S, 580 54' W, from 510 02' S, 580 55' W to 510 00' S, 580 53' W.
OTC, 121-130 m.

Baseodiscus antarcticus, Baylis Lineus corrugatus, Mcintosh.

Amphiporus spinosus, Burger. Cerebratulus malvini, n.sp.

St. WS 77. 12. iii. 27. 51° 01' S, 66° 31' 30" W, from 51° 00' S, 66° 30' W to 510 02' S, 66° 33'W.
OTC, 1 10-113 m-
Lineus corrugatus, Mcintosh.

St. WS79. 13. iii. 27. 5i°oi'3o"S, 64°59'3o"W, from 5i0oo'S, 650 00' W to 5i°03'S,
64°59'W. OTC, 132-131 m.
Lineus corrugatus, Mcintosh.

St. WS 80. 14. iii. 27. 500 57' S, 630 37' 30" W, from 50° 58' S, 630 39' W to 500 55' 30" S,
63°36'W. OTC, 152-156 m.
Lineus corrugatus, Mcintosh.

St. WS 84. 24. iii. 27. 7^ miles S 90 W of Sea Lion Island, East Falkland Island, from 520 33' S,
590 08' W to 520 34' 30" S,"59° 11' W. OTC, 75-74 m.
Lineus corrugatus, Mcintosh. Amphiporus falklandicus, n.sp.

St. WS 88. 6. iv. 27. 540 00' S, 640 57' 30" W, from 540 00' S, 650 00' W to 540 00' S, 640 55' W.
OTC, 118 m.
Lineus corrugatus, Mcintosh.

St. WS 93. 9. iv. 27. 7 miles S 8o° W of Beaver Island, West Falkland Island, from 51° 51' S,
6i° 30' W to 510 54' S, 6i° 30' W. N 7-T, 133-130 m.
Amphiporus lecointei, Biirger.

St. WS 97. 18. iv. 27. 490 00' 30" S, 6i° 58' W, from 490 00' S, 620 00' W to 490 01' S, 61° 56'
W. OTC, 146-145 m.
Amphiporus falklandicus, n.sp.

224 DISCOVERY REPORTS

St. WS219. 3. vi. 28. 400 06' S, 620 12' W. NCS-T, 116-11401.
Amphiporas moseleyi, Hubrecht.

St. WS 225. 9. vi. 28. 50° 20' S, 62° 30' W. OTC, 162-161 m.
Amphiporus falklandicus, n.sp. Linens corrugates, Mcintosh.

A. gerlachei, Burger.

St. WS 228. 30. vi. 28. 50° 50' S, 560 58' W. OTC, 229-236 m.
Amphiporus falklandicus, n.sp Cerebratulus malvini, n.sp.

Linens eorrugatus, Mcintosh.

NCS-T, 229-236 m.
Linens eorrugatus, Mcintosh.

N 4-T, 229-236 m.
Linens eorrugatus, Mcintosh.

St. WS 231. 4. vii. 28. 500 10' S, 580 42' W. NCS-T, 167-159 m.
Amphiporus inexpectatus, n.sp.

St. WS 237. 7. vii. 28. 460 00' S, 6o° 05' W. NCS-T, 148-256 m.
Linens eorrugatus, Mcintosh.

St. WS 239. 15. vii. 28. 510 10' S, 620 10' W. OTC, 196-193 m.
Cerebratulus malvini, n.sp.

St. WS246. 19. vii. 28. 52°25'S, 6i°oo'W. OTC, 267-208 m.
Amphiporus falklandicus, n.sp. Cerebratulus malvini, n.sp.

A. gerlachei, Burger. Linens eorrugatus, Mcintosh.

St. WS 248. 20. vii. 28. 520 40' S, 580 30' W. OTC, 210-242 m.
Amphiporus falklandicus, n.sp. Linens eorrugatus, Mcintosh.

St. WS249. 20. vii. 28. 520 10' S, 570 30' W. DLH, 166 m.
Amphiporus gerlachei, Burger. Cerebratulus malvini, n.sp.

Linens eorrugatus, Mcintosh.

St. WS302. 6.x. 28. 540 57' 20" S, 3 1° 49' 35" W. N70B, 98-om.
Amphiporus scoresbyi, n.sp.

St. WS548. 31. i. 31. 64°o7' S, i5°38' W. N 100 B, 106-0 m.
Amphiporus scoresbyi, n.sp.

St. WS550. i.ii. 31. 66°5i|'S, i5°24' W. N 70 B, 121-om.
Amphiporus scoresbyi, n.sp.

MARINE BIOLOGICAL STATION, SOUTH GEORGIA

St. MS 67. 28. ii. 26. East Cumberland Bay, 3 cables NE of Hobart Rock to \ cable W of Hope
Point. BTS, 38 m.
Tetrastemma gidliveri, Burger. • Imphiporus moseleyi, Hubrecht.

St. MS 68. 2. iii. 26. East Cumberland Bay, 17 miles S | E to 81 cables SE x E of Sappho Point.
NRL, 220-247 m.
Amphiporus spinosus, Burger. Linens eorrugatus, Mcintosh.

NEMERTEANS 225

St. MS 70. 9. iii. 26. Maiviken, West Cumberland Bay. Sh. coll.
Tetrastemma esbenseni, n.sp. T. maivikenensis, n.sp.

St MS 71. 9. iii. 26. East Cumberland Bay, 9J- cables E x S to 1-2 miles E x S of Sappho Point.
BTS, 110-60 m.
Amphiporus spinosus, Burger.

NCS-T, 110-60 m.
Amphiporus lecointei, Burger.

SYSTEMATIC ACCOUNT

PART I. NEMERTEANS FROM SALDANHA BAY,

SOUTH AFRICA

Saldanha Bay lies on the west coast of Africa about sixty miles north of Cape Town.
The bay is a shallow sandy-bottomed inlet fifteen miles long. From the narrow en-
trance the bay turns south nearly parallel to the outer coast from which it is separated
by a ridge rising in some places into considerable hills. Dredging inside the bay was
very unproductive. The Nemerteans described here were collected mainly from the
roots of kelp torn from the boulders at low tide. A few were found beneath stones and
others in roots and kelp tangles washed up on the beaches. Although the conditions
near the two whaling stations were different from those on the outer coast there was no
apparent difference in the kelp root faunas except in size. The largest specimens came
from the outer coast.

Twelve species were found, eight of which were already known from the Mediter-
ranean and coasts of northern Europe. Tubulanus nothus and Cerebratulus fuscus were
the commonest species, but the yellow Emplectonema ophiocephala was also frequently
captured. Limits geniculatus, another Mediterranean form, was collected at Annobon
Island in the Gulf of Guinea close on the equator. Three new species — Zygonemertes
capensis, Tetrastemma nigrolineatum and Oerstedia maadata — are described in the fol-
lowing section.

The extension of range into the southern hemisphere appeared so important that
every effort has been made to verify the specific identity of the animals with the previous
descriptions. It is for this reason that the colour sketches of the worms in life are re-
produced (PI. XV).

Order PALEONEMERTEA

Genus Tubulanus, Renier
Tubulanus nothus, Burger (Plate XV, figs. 1,2; Figs. 1-4).
Carinella nothus, Burger, 1895, p. 527, pi. 1, fig. 12.

Twenty-two specimens of this species were taken from kelp roots from the rocks be-
tween tide marks. The lengths ranged from 10 to 20 cm. The corresponding breadths
were o-6 mm. (head i-o) and i-omm. (head 1-5). One very large fragment of body

226 DISCOVERY REPORTS

2-5 mm. broad was found. The worm to which this belonged must have been 30-40 cm.
long. A thin transparent tube was sometimes secreted when the worms had been in
captivity for a few hours.

Form and colour in life. The body is round in section, but it tapers and is a little
flattened from above down towards the tail. The head is round in outline from above,
broader than it is long, and about one and a half times as broad as the body. The pro-
boscis pore is just ventral to the tip of the head and the mouth is a small longitudinal
slit immediately behind the junction of the head and neck. There is a large laterally
spread group of black pigment specks at the edge of the head on each side of the pro-
boscis pore. The tail is somewhat bulbous ; its margin is nearly transparent, and the gut,
which shows through, is not swollen.

The general colour is brownish red, fading gradually to yellow at the tail. The under-
side is paler than the back, more so posteriorly. The shades of red vary from dusky
crimson to light yellow red. The colour fades abruptly near the tip of the head and there
is here a transverse white mark nearly complete ventrally to form a white ring. The first
body ring occurs about the breadth of the head from its posterior end. This ring is a
white chevron with its apex pointing back. It is incomplete ventrally. The rest of the
body is marked by a series of white annulations also usually incomplete ventrally. The
number varies from seventeen to eighty. Some of the earlier rings are complete and the
larger ones are formed of two narrow bands placed close together. The interval between
the third and fourth body rings is more than double the interval between any other
consecutive rings. Traces of lateral longitudinal white lines are found, and sometimes
there is a thin mid-dorsal line traceable after the first few rings behind the head and dis-
appearing again at about the fortieth ring.

Form and colour of preserved specimens. On fixing shrinkage takes place, but the body

retains its shape to a great extent. The head is flattened but very round in outline from

above. There is a fold between head and body in which are hidden the

openings of the canals of the cerebral organs (Fig. 1). Side organs are

present. They appear as white marks one on each side of the body in

the dark brown region behind the third white ring. The body is often

somewhat swollen here. The colour of the body anteriorly is light brown

on the back, paler beneath. There is a sharp transition after the second

white ring to dark dirty brown which is similar dorsally and ventrally. ', D..

° J J J nothus, Burger.

This gradually fades and at the eleventh ring the colour is again light Head of preserved
brown. Lateral white lines can just be seen. The patches of eyespecks specimen, ventral
are sometimes visible as a greyish blur. surface.

Internal structure. The epithelium is thick, especially in the oesophageal region, and
contains numerous eosinophile gland cells. Near the tip of the head the distribution of
these cells is unequal, there being many ventrally and few dorsally while with the pig-
ment cells the opposite is the case. Head glands and frontal organs are not present. The
mouth is evident in transverse section before the brain has disappeared (Fig. 2 A). The
gut is unbranched.

NEMERTEANS

227

The vascular system is first seen in serial sections as a single relatively large lacuna,
with strands of tissue passing across it vertically, lying above the commencement of the
rhynchodaeum at the tip of the head. Farther back the intrusion of the rhynchodaeum
causes this lacuna to be divided into two lateral and a median lacuna. The latter has a

Fig. 2. Tubulanus nothus, Burger. A, transverse section of head at anterior end of mouth;
B, transverse section of body at nephridial canal.

narrow lumen and lies above the rhynchodaeum. Where the pharyngeal nerves leave
the brain the lateral blood spaces are joined beneath the rhynchodaeum by a ventral
connecting lacuna and at the same place the laterals are again united with the median
dorsal. The latter shortly afterwards disappears. Immediately
behind the ventral connecting blood space a pair of small
vessels separate off" from the laterals. These remain close to
the angle of the pharynx, but at the posterior end of the
mouth they rejoin the laterals. There are subsequently one
or two downward extensions of the laterals forming small
loops (Fig. 3). I could not trace the close connection of the
vascular system with the excretory tubules described by
Oudemans (1885) in Carinella annulata. The excretory
tubules occur above the lateral vessels and open to the ex-
terior by a single duct high up on the dorsal surface
(Fig. 2B).

The nervous system lies between the basement membrane
and the circular muscles. The brain consists of two lateral
curved plates of tissue with no observed demarcation be-
tween cellular and fibrous tracts. The plates are united by grapbhic reconstruction showin
a stout ventral commissure, and a little farther back by a the vascular system (dorsal).
slender dorsal commissure. There is little differentiation

into dorsal and ventral ganglia until the posterior end of the brain. Anteriorly the brain
breaks up into many nerves to serve the tip of the head. Just in advance of the ventral
commissure a pair of nerves is given off for the proboscis. Another pair originates from
the ventral ganglia after the commissure and innervates the pharynx. A mid-dorsal nerve
intersects the dorsal commissure. It passes down the body between the basement mem-

Fig- 3-
Burger.

Tubulanus nothus,
Diagram from a

228 DISCOVERY REPORTS

brane and the muscles, and it has been traced forwards for some distance into the head.
The lateral nerve stems are united by frequent commissural strands. From the trans-
verse furrows between head and body the cerebral canals pass in on each side directly
towards the brain. At the basement membrane ^

they meet and become embedded within a nerve
from the dorsal ganglion (Fig. 4). The canal just
penetrates the basement membrane but cannot be
traced farther.

These worms have been identified with Tubu-
lamis nothus, described by Burger from the
Mediterranean, from the form of the head, colour
of the body and the arrangement of white rings,
the pigmented patches of eyespecks, the side
organs, and some anatomical features such as the p.g ^ Tubulanus )l0thlis> Biirger. Section
thickness of the epithelium and the position of showing the cerebral canal entering the
the nephridial canals. The penetration of the nerve from the dorsal ganglion, e, epi-

cerebral canals beyond the basement membrane thelium; cc> cerebral canal; #. dorsal

, . . .- T „ , ,. ganglion; Im, longitudinal muscles; vg, ven-

is not considered of significance. In 1 . banyulensis, *ral lion

whose anatomy according to Burger (1895) is

similar to that of T. nothus, the cerebral organ is a finger-like pit which reaches

the basement membrane and is innervated from the posterior angle of the dorsal

ganglion (loc. cit., p. 526).

It is curious that T. annulatus was not taken at Saldanha Bay, for it was described by
Stimpson (1856) from Simon's Bay near Cape Town.

Order HETERONEMERTEA
Genus Lineus, Sowerby

Lineus bilineatus, Renier, 1804 (Plate XV, fig. 10).

One specimen of this characteristically marked species was taken in August from a
kelp root attached to a granite boulder on the outer shore near Eland Point — the end of
the southern arm of the bay. The length was between 15 and 20 cm., but the specimen
was damaged. The breadth of the head was o-6 mm. Eggs were present.

Form and colour in life. The body is soft, somewhat flattened and slim. The head is not
distinctly marked off from the body ; it is flat, obtusely pointed and wider at half length
than the anterior part of the body. The tip is slightly notched. No eyes are visible.

The colour is light brown with a double white line down the back from the tip of the
head to the tip of the tail.

Form and colour of preserved specimen. In spirit the worm is white and no trace can be
seen of the white lines. The cephalic slits are very long. The mouth is a small round
aperture opposite the posterior ends of the cephalic slits.

No eyes could be seen after clearing in anilin oil.

NEMERTEANS

229

Lineus geniculates (Chiaje, 1828) (Fig. 5).

Two specimens (N 69) were taken at St. 283 off Annobon in the Gulf of Guinea
This was in August 1927. They were practically identical
in size: length 27-5 cm.; breadth 0-55 cm.; thickness
0-15 cm.

Form and colour of preserved specimens. The body is very
flattened, the head blunt in outline from above and the tail
pointed. The mouth is a longitudinal slit 7 mm. long

(Fig. 5)-

The colour is brownish dorsally and ventrally with com-
plete white annulations — fifty-three in one specimen, fifty-
six in the other. I could find no trace of longitudinal
markings and no sign of eyes : but of the latter eighty-eight
were afterwards counted in the sections embedded in the
tissue of the head at the edges of the cephalic slits. They
were most numerous at the tip of the head but occurred as
far back as the brain.

The head glands stain markedly with haematoxylin. They A B

are scattered in two symmetrical areas on each side of the Fig. 5. Lineus geniculatus
rhynchocoel, passing away on each side above and below (Chiaje). A, dorsal; B, ventral
the cephalic slits. They disappear from the sections before surface of the Preserved sPed"

men x 1
the brain appears. The cephalic organs are somewhat small.

A full description of this species is given by Burger (1895), and to this the specimens

from Annobon exactly conform.

Lineus ruber (O. F. Muller), 1771 (Plate XV, fig. 4; Fig. 6).

Twenty-three specimens were taken between July and September from the shore be-
tween tide-marks. Most of the worms were found near the whaling stations under
stones on the mud, but some occurred in kelp roots from rocks on the shore outside the
Bay. The largest worm was 14-0 cm. long, with a breadth of i-o mm. ; the smallest was
5-0 cm. long, breadth 0-5 mm.

Form and colour in life. The body is round and tapers gradually at the tail. The head
is rather flat and broader than the succeeding part of the body. The snout is broad and
mobile. The cephalic slits are long and pale. The mouth is a small elongated slit with
pale lips far back on the ventral surface of the head. The colour in life is brown, deepest
anteriorly. The tail is very pale. Behind the head both dorsally and ventrally there is
a large undefined reddish mark. A series of pale rings is visible on the body in some
specimens. In an 11 cm. worm seventeen were counted. The tip of the snout is pale.
The eyes are variable, but not more than ten were found on each side. Usually there
were five in a line at the edge of the pigment.

In one specimen there was a row of minute pores showing pale on each side of the
body nearer the dorsal than the ventral side. The row, marking the openings of the

23o DISCOVERY REPORTS

genital sacs, commenced about one-third of the length of the body from the anterior

end.

Form and colour of preserved specimens. The colour is bleached in spirit to a greyish
white. No eyes are visible. The proboscis pore is terminal and the anterior end of the
mouth is at the level of the pos-
terior ends of the cephalic slits
(Fig. 6A).

Internal structure. Frontal
organs were not seen although
the tip of the head was sectioned.
The muscle layers are arranged
typically (Fig. 6 B).

There is a single vascular
lacuna dorsal to the rhyn-
chodaeum in the head. This
divides and the two lacunae thus
formed split up after the region of

. . Fig. 6. Lmeus ruber, O. F. Muller. A, head of spirit specimen,

the ganglia into numerous vessels ventro.laterai . B> transverse section of body at the passing of the
round the gut. The dorsal vessel excretory tubules to the exterior,
protrudes into the rhynchocoel.

The excretory tubules bulge into the large blood vessels on each side of the rhyn-
chocoel. They occur a considerable distance behind the brain. Only one duct was found
leading to the exterior on each side and it opened high up on the dorsal surface (Fig.
6B). From the persistence of the tubule in the last sections cut in series of this worm
it is probable that further ducts were present. The number given by Punnett (i 901) is
from six to twelve on each side.

The cephalic slits continue nearly to the tip of the snout and they disappear pos-
teriorly as soon as the canal has entered the head. The fibrous tissue of the dorsal
ganglion divides posteriorly into an upper branch which quickly disappears, and a lower
which becomes invested with the cerebral organ and forms the posterior lobe of the
brain. This lies almost free in a blood sinus at its hinder end. When the ventral ganglia
merge into the lateral nerves they shift laterally from their position directly beneath the

dorsal ganglia.

The genital sacs (in the female) open high up laterally.

The green form of L. ruber was described by Stimpson (1856 and 1857-8) from speci-
mens collected in Simon's Bay under the name Cerebratulus oleaginus (Meckelia olivacea
in the earlier paper). The identification with Lineus ruber was suggested by Burger
(1895). L. ruber has been recorded from Madeira by Langerhans (1880).

NEMERTEANS

23'

Genus Cerebratulus, Renier

Cerebratulus aerugatus, Burger, 1892 (Plate XV, fig. 5; Figs. 7, 8).

The anterior end of a Lineid worm was found in a kelp root torn from the rocks be-
tween the whaling stations. The length was 5-5 cm., the breadth of the head about
i-omm. A distinct shallow groove was
noticed in the mid-dorsal line of the head.
The diverticula of the gut were very regularly
placed opposite one another.

Form and colour in life. The body is soft
and a little flattened. The head is lance-
shaped, flat and broader at the level of the
ends of the cephalic slits than the succeeding
part of the body. The cephalic slits are very
long. There are no eyes. The mouth is
small, placed mid-ventrally behind the ends
of the cephalic slits. The body is reddish
brown towards the anterior end, fading to
yellow posteriorly. The greater part of the
head is white. A vague red patch is visible
dorsally and ventrally where the body colour
fades at the back of the head. In spirit the
colour is brownish.

Internal structure. The tip of the snout was unfortunately
missed in sectioning so that the frontal organs are not known.
The head glands are thin and scattered. They extend back as
far as the brain. No trace of eyes could be seen. In the
stomach region the epithelium is about as thick as that part
of the longitudinal muscles into which the cuticular glands
penetrate. The outer longitudinal muscle layer is from three
to four times as thick as the circular layer, while the inner
longitudinal layer is thinner than the circular. There is no
diagonal layer. The circular muscles of the rhynchocoel are
about as thick as those of the circular layer of the body

(Fig- 7)-

There is a vascular loop in the head. Posterior to the brain
the dorsal vessel protrudes into the rhynchocoel and a number
of blood spaces of varying size surround the stomach (Fig. 7).
The excretory system is not known.

The relations of the brain and cerebral organs are shown in
Fig. 8. The fibrous tissue of the dorsal ganglia is about twice
as extensive as that of the ventral and the brain is large

Fig. y. Cerebratulus aerugatus, Burger. Transverse
section through the body to show the relative
thickness of the muscle layers.

Fig. 8. Cerebratulus aerugatus ,
Burger. Diagram from a
graphic reconstruction of the
brain and cerebral organs

232

DISCOVERY REPORTS

for the size of the head. The ventral commissure is wider than the dorsal. The
dorsal ganglia divide posteriorly into upper and lower branches. The small upper
branches taper away rapidly while the lower become included in the cerebral organs
which swell to form substantial lobes lying almost free at their hinder ends in blood
sinuses formed from the lateral and median vessels. The ventral ganglia become the
lateral nerves beneath the cerebral organs and leave their positions on either side
of the rhynchocoel to pass down the body. Neurochord cells were observed in the
cellular tissue of the inner face of the ventral ganglia just posterior to the ventral
commissure.

The cephalic slits are apparent very near the tip of the head. They do not extend
posteriorly beyond the cerebral canals, which pass in directly opposite the lower cornua
of the dorsal ganglia.

As far as I am able to judge from the incomplete specimen the identification of this
worm with C. aerugatus is justified.

Cerebratulus fuscus, Mcintosh, 1873 (Plate XV, fig. 8; Figs. 9, 10).

Eighteen specimens were taken during August and September. This species was
fairly common in the kelp roots between tide-marks both inside and outside the Bay.
The lengths varied between 2-3 and 3-5 cm. with breadths of o-8 and 1-4 mm. Speci-
mens with eggs were collected on August 14.

Form and colour in life. The body is a little flattened from above down. The head is
flat, not distinctly marked off from the body, and it tapers to a blunt snout. The cephalic
slits are very long and the mouth is
small, placed at the end of the head
behind the brain. The small eyes vary
in number and are deeply embedded.
There are usually ten visible on each
side ; five together near the tip of the
snout, the others in a line farther back.
A caudal appendage is present. The

range of colour is from pale buff to a d

pink or light red on the back with Fig> 9. Cerebratulus fuscus, Mcintosh
scattered reddish brown splashes preserved specimen, x 4
mainly near the middle line and less show the eyespots.
definite and paler at the posterior end of the body. The underside is pale flesh colour.

Form and colour of spirit specimens. An outline sketch of a preserved specimen is
shown in Fig. 9A. Neither eyes nor markings are visible. On clearing in anilin oil the
eyes can be seen (Fig. 9B). The colour is completely bleached.

Internal structure. Frontal organs are present. The head glands are fairly well de-
veloped. They are more numerous dorsally than ventrally and they can be seen dorsally
in transverse sections almost to the posterior end of the brain. They stain deeply with
haematoxylin but not so deeply as the subepithelial gland cells. Eosinophile cells are

A, sketch of the
lateral view of the head to

NEMERTEANS

233

numerous in the epithelium. Farther back on the body many of the subepithelial cells
stain with eosin.

The rhynchocoel is long, but the proboscis is attached in the first half of the
body.

Two vascular lacunae close to the rhynchodaeum can be traced back to the anterior
insertion of the proboscis. They reappear as a median lacuna ventral to the rhynchocoel ;
and from this lacuna arises the dorsal vessel and a dorsal lacuna which later divides into
two lateral lacunae in which lie the cerebral organs. Posterior to the brain the vascular
system has not been traced.

The brain is very large (Fig. 10). There is a strong upper branch to the dorsal ganglion
posteriorly. This upper horn disappears before the lower, which is invested with the
cerebral organ and forms the posterior lobe of the brain. A dorsal nerve arises from the
dorsal commissure. On the intrusion of
the mouth the ventral ganglia separate and
from this point they may be called the lateral
nerves. Previous to this the ganglia are
close together and directly beneath the dorsal
ganglia. They curve outwards sharply to
take up their lateral positions. A pair of
nerves is given off before the separation for
the innervation of the pharyngeal muscles.

The differences between these specimens

and the worms described by Mcintosh

(1873) are so slight that I have no hesita-

v ' J/ .... _, „, Fig. 10. Cerebratulus fuscus, Mcintosh. Irans-

tion in identifying them as C. fuscus. 1 he yerse section of the body at the extreme anterior
pigmentation is more distinct, but worms limit of the mouth,
of the same species even more distinctly

marked have been described by Joubin (1894). Burger (1895) gives the body form as
characteristic— the tail being very wide. Evidently this is a variable character for
Mcintosh's description is "slightly tapered towards either extremity". The internal
structure bears out the identification, although I have not been able to find neurochord
cells in the ganglia.

An autotomized specimen (N 129) of a small heteronemertean with a caudal appendage
was collected at Tristan da Cunha (St. 4). No colour note was made. The approximate
length was 30-0 mm.

The body is round in section. The head is flat with long cephalic slits (3-0 mm.) and
a straight slit-like mouth 1-5 mm. long. The colour is uniformly greyish. The long pro-
boscis is protruded. The anatomical details of this form are interesting in that they are
definitely against the inclusion of this form with the Cerebratulus of the Falklands
(C. malvini). The peculiarly wide cephalic slits are present and the eyes, which are
absent in C. malvini, are here evident, confined however near the tip of the head. Well-
marked frontal organs are present. Head glands are very thin and scattered. The brain

3-2

234 DISCOVERY REPORTS

is large and is similar to that of C.fuscus. There is a strong dorsal branch to the massive
dorsal ganglion. In the absence of notes on the form and colour during life I hesitate
to separate this form from C. fuscus, which anatomically it closely resembles.

Order HOPLONEMERTEA

Sub-order MONOSTILIFERA
Genus Emplectonema, Stimpson

Emplectonema ophiocephala (Schmarda), 1859 (Plate XV, fig. 14; Figs. 11, 12).

Thirty-eight specimens were captured between June and September in kelp roots
from rocks between tide-marks inside and outside the Bay. The usual length was about
25 cm., breadth 1-1-5 mm. The largest worm was 40 cm., greatest breadth 2-0 mm.
The majority were mature and in consequence rather swollen. In one specimen the gut
was full of grey mud. Damaged worms were frequent as the body is very soft.

Form and colour during life. The body is long and soft. A distinction can be drawn
between head and body, as the posterior end of the flattened lanceolate head is slightly
broader than the succeeding part of the body. Anteriorly the head tapers to an acute
snout. Neither mouth nor cephalic slits can be seen, but occasionally a pair of faintly
marked sloping lateral grooves are apparent behind the head. Deep transverse wrinkles
appear when the body contracts. The colour is generally yellow, but may vary from
lemon yellow or yellow-brown to reddish or orange. The head and tail are paler than the
rest of the body. There are elongated groups of small eyespots on each side of the head —
about twenty in each group— spreading out somewhat posteriorly. In small worms
(6-7 cm. long) there are from four to twelve eyes visible on each side. The brain shows
as a pinkish, brownish or greyish bilobed structure through the skin just behind the
groups of eyespots (Fig. 11 A). The genital sacs show white through the skin.

Form and colour of preserved specimens. The worms frequently contract violently and
break up during preservation. In spirit they are bleached, and the form is often dis-
torted by bulges and knot-like swellings. The eyes and cephalic slits are not visible. The
proboscis pore is just beneath the tip of the head. On clearing in anilin oil the brown or
black cup-shaped eyes can be seen. They vary greatly in number from seven or eight on
each side to thirty. When seen partly from the side (Fig. 1 1 B) or from above, each group
can be divided into two rows. The eyes of the outer row open forwards, those of the
inner row backwards and upwards. There are usually more eyes in the inner row, but
they are smaller than those of the outer.

Internal structure. The oesophagus opens into the rhynchodaeum. The gut is
capacious. Behind the rhynchocoel it occupies the entire body cavity. The anterior
caecum does not extend far forwards (it does not appear in either of my two series of
sections). Frontal organs are absent. The head glands consist of a thin solid strand,
staining with eosin, dorsal to the rhynchodaeum, extending back, thinning out and dis-
appearing before the separation of the oesophagus and rhynchodaeum. Near the tip of

NEMERTEANS

235

the head there are small round eosinophile cells scattered among the muscle strands.
These apparently open into the rhynchodaeum. The tissues of the body cavity stain
deeply with haematoxylin or carmine, giving the impression of a thick mucus in which
deeply staining granules are embedded.

The epithelium is very thick. The basement membrane is a little thicker than the
circular muscle layer. It is homogeneous in appearance and stains with eosin. The
longitudinal muscle layer is somewhat thicker than the epithelium and the fibres are
packed in conspicuous bundles. The ganglia are enveloped in a coat of longitudinal
fibres. In the body behind the brain this muscle layer splits into bundles which pass
gradually outwards and rejoin the longitudinal layer of the body. The eyes are just inside
the muscles, embedded in the connective tissue.

A

Fig. 11. Emplectonema ophiocephala (Schmarda). A, sketch of head cleared in cedar oil; B, head cleared
in cedar oil, dorso-lateral view to show the eyespots; C, armature, cleared in cedar oil.

The rhynchocoel does not extend beyond the anterior third of the body. The pro-
boscis is short and thin. Both rhynchocoel and proboscis are greatly restricted by the
closely opposed ganglia. The armature appears to vary. Two accessory stylet reservoirs
are always present, but the number of stylets differs. In two worms of nearly the same
size taken at the same time, the accessory stylets numbered one and two in one, and
eight and nine in the other (Fig. 11C). A third specimen possessed two and three
stylets. There is a stout main stylet mounted on a slightly hour-glass-shaped base
shorter than the stylet itself.

Vascular system. There are two lateral vessels in the head. They have not been traced
at the ganglia, but the blood is responsible for the colour of the brain in the living animal,
so the vessels are broken up into fine branches. Directly behind the brain there are two

236

DISCOVERY REPORTS

vessels at the outside upper corners of the stomach and two lower than these near the
lateral nerves. Farther back in the body there are two laterals near the nerve trunks and
a dorsal median vessel above the gut.

Excretory system. The convoluted tubules lie beside the stomach and gut. They ex-
tend forwards to just behind the brain. The posterior limit of the tubules was not seen
in either series of sections, but two pairs of efferent ducts — the posterior pair almost
2 mm. behind the anterior — were observed. The ducts pass over the lateral nerves, then
downwards and outwards through the body wall.

Nervous system. The brain is peculiarly compact (Figs. 11 A, 12). The ganglia are
very close together, so that the rhynchocoel and proboscis are much constricted where
they pass through the brain. There is little dis-
tinction between dorsal and ventral ganglion. The
ventral commissure is extraordinarily deep and
broad. In the specimen examined the brain was
0-4 mm. long and 0-25 mm. thick. The ventral
commissure was 0-27 mm. long and 0-125 mm.
thick — half the thickness of the brain and nearly
three-quarters of the length. The dorsal com-
missure was 0-08 mm. long. The lateral nerves
leave the brain at right angles to the long axis.
They make a sharp turn back to pass laterally
down the body inside the longitudinal muscles.
A marked swelling of the nerves occurs after
they leave the brain and before they turn.

The cerebral organs are very small. They are sac-like organs lying some distance in
front of the brain and are connected with the dorsal ganglion of each side by a stout
nerve which penetrates the muscular sheath and reaches the ganglion about one-third
of the length of the brain from the anterior end. The cerebral canals open to the exterior
near the tip of the snout (Fig. 1 1 A).

Ommatoplea ophiocephala , Schmarda (1859), *s considered to have been a Eupolia by
Burger (1895, p. 27). Schmarda refers the genus Ommatoplea to Ehrenberg and this is
synonymous with Eunemertes (Burger, 1895, p. 13). Schmarda 's account of the worms
found under stones and in sand in Table Bay, South Africa, corresponds fairly well with
the description given above. His specimens were larger. He mentions the length as
1 m., the greatest breadth 10 mm., the colour as lemon yellow or golden brown, and
the (eight) eyes in two lines on both sides of the head. The "small egg-shaped sub-
terminal" mouth is evidently the opening of the rhynchodaeum, and the terminal pore
may have been the opening of the head gland.

In many ways these specimens are similar to Eunemertes antonina, Quatrefages. The
peculiarities of the brain which characterize this Mediterranean species, the swelling and
course of the lateral nerves, the position and size of the cerebral organs and canals are
alike in both. Even the darkly-stained connective tissue of the cavity of the body, re-

Fig. 12.
(Schmarda).
the brain.

Emplectonema ophiocephala
Transverse section through

NEMERTEANS 237

marked upon by Burger (1895, p. 547), appears in these Saldanha Bay forms. The chief
difference lies in the uniform rose red or carmine tint of E. antonina, which is considered
characteristic of the species by Hubrecht (1879, p. 231). The colour is noted also by
Joubin (1894, p. 206). The body of E. antonina appears to be considerably more slender
than this South African worm, and the main stylet in the armature of the latter does not
bear the proportion 3 : 1 to its base, although it is certainly longer than the base in the
specimens I have examined. Taking these facts into consideration I do not consider the
identification of these worms with E. antonina is justified, but there seem to be good
grounds for considering that the same forms were described by Schmarda under the
specific name ophiocephala.

Genus Nemertopsis, Burger

Nemertopsis tenuis, Burger, 1895 (Plate XV, fig. 7).

Eight specimens were found at different dates in attached kelp roots from the granite
boulders between tide-marks on the outer coast. The lengths ranged from 10 to 41 mm.,
the corresponding breadths being 0-3 and 0-4-0-5 mm.

Form and colour in life. The body is very thin, almost Nematode-like. The head is not
distinct from the body, and the tail tapers acutely. The four small eyes occur in close-set
pairs, the posterior pair being very far behind the anterior. The colour is yellowish
brown and the gut shows pale through the skin. Yellow-red blood vessels, especially the
dorsal vessel in the posterior half of the body, are evident in some specimens.

Form and colour of preserved specimens. Considerable shrinkage takes place after
preservation. The colour is bleached. On clearing in anilin oil the brown cup-like eyes
can be faintly seen.

Internal structure. The oesophagus opens into the rhynchodaeum just posterior to
the cerebral organs and in front of the brain. The contents of the cells of the stomach
wall stain deeply with haematoxylin. The anterior caecum has no forward branches.

The epithelium is as thick as the longitudinal muscle layer, and the basement mem-
brane is about half as thick as the circular muscle layer. The longitudinal muscles are
not conspicuously divided into bundles. A compact head gland is present, opening
dorsally just above the proboscis pore. The head gland reaches the brain.

The rhynchocoel extends about one-third of the body length. The proboscis has ten
nerves. The armature is not known.

There are two definite blood vessels in the head forming a loop beneath the head
gland. At the ganglia these vessels become difficult to trace, but posterior to the ex-
cretory tubules there are two vessels lateral to the rhynchocoel and a median vessel
above the gut. This median vessel is formed originally of branches from the laterals.

The convoluted excretory tubule is packed closely behind the dorsal ganglion and
opens to the exterior by a single duct above the lateral nerve.

The brain is not peculiar. The ventral commissure is thicker than the dorsal, and the
lateral nerves leave the brain sharply and turn back as sharply after undergoing a knot-

238 DISCOVERY REPORTS

like enlargement. The eyes are embedded within the longitudinal muscle layer. The
second pair occurs immediately before the brain. The cerebral organs are very small.
They lie in front of the brain and their canals open behind the first pair of eyes into a
transverse lateral groove that occurs on each side of the head.

Nearly ripe eggs were present in some specimens. The eggs are shed just above the

lateral nerves.

In two particulars — the yellow rather than rose colour of the body and the presence
of a head gland — this description differs from that of Burger.

Another specimen (N 109) was identified from sections. It was 10 mm. long, 0-3 mm.
broad, bleached and coiled. This was taken at St. WS 4 in 40 m.

Genus Amphiporus, Ehrenberg

Amphiporus pulcher, Johnston, 1837 (Plate XV, fig. 13).

Two specimens (N 36) were taken from attached and washed-up kelp roots in Sep-
tember 1926 on the southern point of Saldanha Bay. The lengths and breadths were
3-5 cm., 1-5 mm., breadth of head i-omm.; 4-2 cm., 17 mm. (swollen with eggs).

Form and colour in life. The body is round anteriorly, somewhat flatter and wider
posteriorly. The head is a little flattened, almost semicircular in outline from above, but
has a slight snout. The mouth, proboscis pore and cephalic slits are not visible, but a
chevron groove can sometimes be seen at the back of the head. The colour is pinkish
yellow or buff, lighter anteriorly and deepest on the back. The ganglia show red through
the skin. About ten eyes are visible on each side in no definite order, but there is usually
a row of four in a line nearly parallel to the edge of the head from the tip to the widest

part.

Form and colour of preserved specimens. In spirit the worms are white and contracted.
Cephalic grooves appear as vertical furrows at the sides behind the head, curving for-
wards ventrally. The proboscis pore is ventral to the tip of the snout in a furrow that
passes vertically round the head.

Internal structure. Well developed head glands are present but they do not reach the
brain. There is a strand close above the rhynchodaeum which stretches back beneath
the vascular loop almost to the ganglia, and more diffuse glands among the muscles of
the head which become restricted near the ganglia to the sides of the body cavity inside
the longitudinal muscles. The duct of the head gland opens just ventral to the tip of the
snout. The eyes are embedded deeply in the tissues of the head.

The epithelium is thinner than the longitudinal muscles and about three times as
thick as the basement membrane and circular muscles together. The basement membrane
is twice as thick as the circular layer and appears to contain fibres. The epithelium con-
tains a large number of eosinophile cells.

The oesophagus opens into the rhynchodaeum in front of the brain. The stomach
walls are not much folded and most of the cell contents stain deeply with haematoxylin.
The anterior caecum has forwardly directed branches, two of which extend beyond the

NEMERTEANS 239

others on each side of the rhynchocoel. They do not approach the brain. The proboscis
is attached about the half length of the body. The rhynchocoel extends the whole length .
The accessory armature consists of two reservoirs each with five or six stylets. There are
eleven nerve strands in the proboscis.

Two blood vessels, one on each side of the rhynchodaeum, form a loop in the head.
They are lost at the ganglia, but reappear as two lateral vessels above the nerves. A dorsal
vessel above the gut is connected by a branch on each side with the lateral.

The excretory tubules are packed close behind the ganglia. From them single ducts
on each side pass back above the lateral nerves and vessels and turn outwards some dis-
tance behind the brain.

The dorsal commissure is longer and thinner than the ventral. The ventral ganglia,
becoming the lateral nerves, shift very gradually outwards. The dorsal ganglia taper
posteriorly and the cerebral canals pass in beneath them and widen into the cerebral
organs which are thus wedged between dorsal and ventral ganglia behind the ventral
commissure. The organs swell as the dorsal ganglia diminish in size, and when the latter
join them they are somewhat flattened bodies larger in cross-section than the lateral
nerves at the same point. They are rounded posteriorly, but extend back some distance
from the point of fusion with the fibres of the dorsal ganglia.

There seems little doubt about this identification. Graphic reconstruction of the head
of the specimen sectioned gives a plan of the vascular and nervous systems identical
with that figured by Mcintosh (1873), and in all particulars the description corresponds
with those of other workers.

Another specimen (N 127) was collected by Mr E. R. Gunther from a sponge brought
to the surface from 292-402 m. by a trawler from Cape Town on July 8, 1927. This was
noted in life as "flesh-coloured".

The preserved specimen was 27 mm. in length, 7 mm. broad and 3-5 mm. thick.
The body was stout and flattened from above down. It was sectioned and identified
with this species.

Genus Zygonemertes, Montgomery

Zygonemertes capensis, n.sp. (Plate XV, figs. 3, 6, 12; Figs. 13-16).

Variations in colour and form were responsible for five descriptions and sketches of
this worm (N 26, N 29, N 33, N 38, N 39). These five have been reduced to three
fairly constant colour variations which are described and figured separately.

(1) Green form. Twenty-seven specimens were taken from attached and washed-up
kelp roots inside and outside the south arm of the Bay. The largest worms were
80 mm. long, 1-5-2-0 mm. in breadth; the smallest 14 mm. long, 0-5 mm. in breadth.

In life the body is slightly flattened and soft. The head is flat, broader than the suc-
ceeding part of the body and somewhat diamond-shaped in outline from above. The
snout is blunt. The tail is pointed and a little bulbous. Neither mouth nor cephalic
slits can be seen. The colour on the back is green, greyish green, or light brown tinged

24o DISCOVERY REPORTS

with green. The underside is buff or yellowish. The head appears lighter than the body
but is, in fact, more distinctly green, and there shows faintly upon it a pale mid-dorsal
longitudinal streak. The posterior end of the body is yellowish. Under low magnifica-
tion black specks can be seen scattered thickly over the head and body. The eyes are
many and small. They occur in four sectors over the head, leaving three narrow paths
devoid of eyes diverging from one another from the tip of the snout (Plate XV, fig. 3).
As the eyes are embedded in the muscles of the head the number seen in life — about
twelve in the inner and eighteen in the outer group on each side — is not only variable
but is nothing like the total number, because the posterior groups only become visible
on clearing in anilin or clove oil.

A specimen similar to this form, 50 mm. long and 1-2 mm. broad (breadth of head
0-75 mm.), was collected at the end of July. A transverse groove encircled the body a
short distance behind the head. The eyes were very indistinct. The ganglia showed
brown through the skin and the swollen body was tinged with orange. This specimen
was a ripe male.

(2) Brown form. Three specimens from kelp roots from the outer rocks were similar
in size and shape to the green form but were of a light brown colour very distinctly
tinged with mauve. A slight speckling of reddish brown occurred, especially anteriorly.
The ganglia showed green through the skin. The eyes were irregular in size and arrange-
ment (Plate XV, fig. 12).

(3) Colourless form. Twenty-four specimens were taken. The largest was 34 mm.
long and 0-5 mm. broad. They were almost transparent at the edges of the body and
white or very faint yellow when seen against a dark background. The eyes appeared to
be even more irregular than in the brown form. They were collected from inside and
outside the Bay (Plate XV, fig. 6).

The fifth colour variant was similar to the colourless forms, but there was a pale
brown collar in the region of the ganglia. One specimen only was taken.

Form and colour of preserved specimens. Shrinkage takes place but the body form is
retained. In about half the specimens the proboscis is partially protruded. Both the
coloured forms retain a hard bright blue-green colour on the back, and this colour
resists spirit for at least two years. The underside of preserved specimens is pale yellow,
and the colourless forms are white. No eyes can be seen. The proboscis pore is situated
just ventral to the tip of the head. Cephalic furrows can be seen curving round dorsally
as they pass back from the snout.

On clearing in anilin oil three groups of eyespots can be seen on each side of the head.
Two of these groups are the ones already mentioned. The third group are dorso-lateral
on the body immediately behind the head. The eyes are small, variable sized and cup-
shaped. Those of the outer anterior group — about fifty — open forwards; those of the
inner anterior group — forty to forty-five — open rather backwards ; and the posterior set
— thirty or so — open mainly laterally (Fig. 13). In the uncoloured forms the eyespots
themselves are bright green.

Internal structure. No frontal organs could be recognized in sections. The epithelium

NEMERTEANS

241

is almost as thick as the longitudinal layer of muscles and between two and four times as
thick as the basement membrane and circular muscles together. These last are about
equal in thickness. There are many large gland cells in the epithelium ; they may be the
black specks of the living animal. The sickle-shaped bodies similar to those first de-
scribed by Marion (1872) are found in the epithelium of all forms ; they appear to resist
acid and are yellowish brown in colour (Fig. 14). The fibre-containing basement mem-
brane stains deeply with haematoxylin, and numerous minute ducts can be seen pene-
trating it, especially on the head. Often the glands are themselves in the membrane but
sometimes they are more deeply placed in the longitudinal muscles. Farther down the
body they are much less frequent.

-{£=■- cm

Fig. 13. Zygonemertes capensis, n.sp. Head, dorsal Fig. 14. Zygonemertes capensis, n.sp. Part of a

surface, cleared in anilin oil to show the three transverse section highly magnified, e, epithelium

groups of eyespots. containing sickle-shaped bodies; bm, basement

membrane; cm, circular muscle layer; /?«, longi-
tudinal muscle layer.

The oesophagus opens into the rhynchodaeum before the brain. In the region of the
ganglia the oesophagus widens. Farther back it opens into the stomach with dark-
stained granules in the cells of the folded walls and on each side appears a branch of the
anterior caecum. These two diverticula do not reach the brain.

There is a vascular loop in the tip of the head. In the brain region the lateral vessels
are widened, their walls become definite and they are connected with the dorsal vessel
above the gut. The three vessels pass down the body and join again above the gut just
before the insertion of the rhynchocoel into the body wall. The excretory vessels he
above the lateral nerves behind the brain. A single duct on each side opens to the ex-
terior opposite the nerves. The ganglia are not peculiar, but the brain is rather large
( Fig . 1 5 A) . The lateral nerves give off branches which pass round the rhynchocoel , and at
the posterior end of the body they join above the gut anterior to the anus. The cerebral
canals open ventro-laterally a little way behind the opening of the oesophagus into the

4-2

242

DISCOVERY REPORTS

rhynchodaeum. The organs are small and do not reach the ganglia, but a stout nerve has
been traced to them from the brain.

A - B

Fig. 15. Zygonemertes capensis, n.sp. A, transverse section across the head at the level of the ventral com-
missure ; B, transverse section across the body at about half its length to show the opening of the genital sacs ;
C, transverse section 0-025 mm- from tip of tail.

In some specimens the protruded proboscis is nearly as long as the body, in others it
is about a quarter of the body length. These differences are probably due to different
states of contraction and extension of the proboscis and body on fixing. There are thir-
teen nerves. That part of the proboscis posterior to the armature
is shorter than the anterior part and is inserted in the rhynchocoel
at about one-third or half the length of the body. The rhynchocoel
extends to the posterior end of the body (Figs. 15 B and C). The
armature (Fig. 16) consists of a main stylet on a base of unusual
length, and two accessory reservoirs each with two, three or four
stylets. Often one of the accessory stylets is incomplete. The base
of the main stylet varies between two and a half to five times the
length of the stylet. Measurements of three armatures from green
and brown forms are given below :

o-i33

0-130

Main stylet (mm.) ... 0-115 °-I56
Base of stylet (mm.) ... 0-518 0-714
Accessory stylet (mm.) 0-164 0-182

In the uncoloured forms the difference in length between base
and main stylet is not so marked :

Main stylet (mm.) 0-078 0-050

Base of stylet (mm.) ... 0-117 0-084

Fig. 16. Zygonemertes
capensis, n.sp. Arma-
ture of the green form
(cleared in anilin oil).

NEMERTEANS 243

Both sexes of each form were examined. Ova are shed by separate ducts forming
when required through the dorsal muscles and epithelium. Spermatozoa are shed just
above the lateral nerves (Fig. 15B). Ripe eggs are present in August.

The examination of sections showed that the original identification of these worms
with Emplectonema echinoderma, Marion, was inadmissable on account of the extent of
the rhynchocoel. It is evident that there is a close relationship with Zygonemertes
virescens (Verrill) Montgomery, but there are also differences— the larger size, the absence
of a terminal sense organ, the relatively larger base to the main stylet, the larger number
of eyes, the greater number of nerves in the proboscis— which justify the establishment
of a new species.

Genus Tetrastemma, Ehrenberg
Tetrastemma candidum, O. F. Muller, 1774 (Plate XV, fig. 15).
T. incisum, Stimpson, 1856.

Thirty-five specimens were captured in attached and washed-up kelp roots from the
rocks at the entrance of the Bay. The lengths were usually about 10 mm., breadth
0-5 mm., but sometimes they reached 16 mm., breadth 075 mm.

Form and colour in life. The body is round in section, tapering posteriorly to the tail.
The head is slightly broader than the body. The mouth is not visible. On each side of
the head there is a nearly vertical groove and occasionally a second pair of grooves can
be seen immediately behind the head.

The colour is pale buff or light brown. The edges of the body appear transparent, and
the gut is visible through the skin. There are four eyespots forming the corners of an
almost perfect square. The posterior pair is just behind the anterior grooves.

Form and colour of preserved specimens. In spirit the ratio of length to breadth is
altered— an 11 mm. worm being 1-5 mm. broad, and a 9-5 mm. worm 1-2 mm. broad.
The colour is bleached, and the eyespots cannot be seen either before or after clearing in
clove or anilin oil.

Internal structure. The oesophagus and rhynchodaeum coincide and the common
opening is immediately ventral to the tip of the head. The head glands are thin and
small. They just reach the brain. There are lateral and dorsal strands, the former joining
together near the tip of the head to form a ventral strand. They open at the extreme end
of the snout.

In the region of the anterior caecum the epithelium is about two and a half times as
thick as the circular muscle layer and the basement membrane together, while the
circular muscle layer is a little thicker than the basement membrane. The longitudinal
muscle is thin. The eyes are visible in sections deeply sunk in the muscles of the head.
Branches from the anterior caecum extend as far forward as the brain.
The cephalic slits are deep grooves almost completely ventral near the tip of the head.
They pass back, upwards and outwards and become canals sinking into the longitudinal

244

DISCOVERY REPORTS

muscles. Glandular tissue encloses the canals and a nerve from each dorsal ganglion
passes into the organ.

The dorsal ganglia are smaller than the ventral. There are ten nerves in the proboscis.

No excretory vessels or ducts could be traced.

The armature consists of a main stylet on a base somewhat longer than itself and two
accessory reservoirs. In one worm three accessory stylets were present in each reservoir
but in others there were only two. The lengths were as follows:

Main stylet 0-067 mm-> accessory stylet 0-058 mm., base 0-077 mm-

This species appears to correspond with Tetrastemma incisum, Stimpson (1856).
I can see no reason for separating it from T. candidum, Muller, since in size and shape
of the body and head, and in the absence of markings it agrees closely with a colour
variant of this species.

Tetrastemma nigrolineatum, n.sp. (Plate XV, fig. 9; Fig. 17).

A single specimen was taken in July from a kelp root inside the Bay. The length was
25 mm., breadth about 0-3 mm.

This slender worm appears to be rectangular in section, and the head is wedge-shaped
when seen from the side. There are no eyespots and neither mouth nor cephalic slits
are visible. The colour is whitish green, with two parallel
black lines passing down the back from tip of snout to tip
of tail. At the head these lines are thinner than they are
on the body.

In spirit traces of the double dark line can be seen.

Internal structure. Head glands are absent. The epi-
thelium is about as thick as the longitudinal muscle layer.
The basement membrane and circular muscles are thin.
The former is thinner than the latter and stains with
haematoxylin. Brown pigment granules are present at the
base of the epithelial cells where the dark lines can be
seen in life. There are no traces of eyespots.

The oesophagus opens into the rhynchodaeum, and
the common pore is ventral to the tip of the head. The
narrow oesophagus opens into a folded stomach with
deeply staining walls just posterior to the dorsal ganglia.
The unbranched anterior caecum ends a long way behind
the brain.

The proboscis extends well into the posterior half of
the body, and the rhynchocoel into the posterior third.
The armature consists of a single main stylet on a reddish
brown base and two reservoirs with seven or eight accessory stylets. Some of these
are incomplete.

Fig. 17. Tetrastemma nigro-
lineatum, n.sp. Diagram from a
graphic reconstruction of the
brain, cerebral organs, excretory
and alimentary systems.

NEMERTEANS 245

The two blood vessels in the head form a loop. They lie one on each side of the
rhynchocoel in the brain region and subsequently dorsal to the lateral nerves, after con-
necting up with the dorsal vessel in the ventral wall of the rhynchocoel. Just before the
excretory ducts pass to the exterior the lateral vessels shift round inside the lateral
nerves and continue down the body beneath them. The excretory system consists of the
usual convoluted tubules close above the lateral nerves. In front of the anterior caecum
a duct leads to the exterior on each side above the nerve (on one side of this worm there
were two ducts).

The brain is large for the size of the head, but it is compact and well defined. There is
a large proportion of fibrous tissue in both ganglia. The dorsal ganglia are smaller than
the ventral and the dorsal commissure is thinner and in advance of the ventral. The
cerebral organs are small sacs opening near the proboscis pore on the ventral surface.
They just reach the brain. The relations of brain, cerebral organs, excretory and ali-
mentary systems are shown in Fig. 17.

This worm is very similar to Nemertopsis bivittata, Chiaje, with the exception of the
four eyes, but its internal structure shows that its affinities are with Tetrastemma rather
than Nemertopsis.

Genus Oerstedia, Quatrefages

Oerstedia maculata, n.sp. (Plate XV, fig. 11; Fig. 18).

Five specimens (N 40) were collected from kelp roots from the outer beaches in
September. The lengths were 6, 7 and 8 mm. with breadths of 0-25-0-4 mm. The
largest specimen was 12 mm., breadth 0-4 mm. This species exhibited a semi-rigid
form like that of O. dorsalis, Abildg., and when in movement the head was often held
upon one side. In one of the worms this feature seemed permanent. A double eyespot
was observed in one specimen and ripe eggs were present in one.

Form and colour in life. The body is round in section and short, with little distinction
between head and tail. The snout is blunt. Four large eyespots are apparent, but, as
shown later, the eyespots are double. They are placed in pairs one behind the other at
a greater distance than in Tetrastemma candidum. The colour is pale buff with one or two
indefinite and irregular small brown spots on the back.

In spirit specimens the eyespots and markings are not visible, but the eyes on clearing
can be seen faintly as brownish marks.

Internal structure. The epithelium in the stomach region is somewhat thicker than
basement membrane and the two muscle layers together. The basement membrane is
homogeneous and thicker than the circular muscles. The longitudinal muscle layer is
not divided into bundles. The oesophagus opens into the rhynchodaeum near the
opening of the latter to the exterior. A large unbranched anterior caecum is present,
reaching almost as far forwards as the brain. The proboscis is long and the rhynchocoel
extends almost the whole length of the body. The head glands form a compact mass,
spreading posteriorly and just reaching back to the ganglia. They open above the pro-
boscis pore.

246 DISCOVERY REPORTS

The armature consists of a main stylet and two accessory stylet reservoirs each with
two stylets.

The vascular system is of the usual type— two lateral vessels and a dorsal vessel above
the gut. In the brain region the laterals are above the lateral nerves, but just behind the
brain they pass below and continue beside the gut (Fig. 18A).

There is a convoluted excretory tubule on each side close behind the ganglia opening
above the lateral nerves.

The ganglia are large in proportion to the size of the animal. Both commissures are
thin. The fibres of the dorsal ganglia (lower posterior angle) continue into the lateral

A

Fig. 18. Oerstedia metadata, n.sp. A, transverse section of the body in the region of the stomach and
anterior caecum; B, part of a transverse section of the head, highly magnified, to show the double eyes.

nerves and retain their individuality down the body. They are separated from the
ventral fibres by cellular tissue. The lateral nerves unite before the anus above the gut.
There are ten nerves in the proboscis.

The cerebral organs are small sacs lying a long way in front of the brain, each with a
short narrow canal leading to the exterior on the ventral surface near the proboscis pore.
There is no external furrow.

The eyespots are double (Fig. 18B). The anterior pair is close behind the cerebral
organs and the posterior pair directly in front of the dorsal ganglia.

The specimen sectioned was a male with ripe sperm.

NEMRRTEANS

247

PART II. NEMERTEANS FROM THE FALKLAND ISLANDS,
SOUTH GEORGIA AND THE ISLANDS AND BANKS
OF THE WESTERN SOUTH ATLANTIC OCEAN

Of the twenty-five species here described only the littoral forms from South Georgia
and the Falklands were sketched in life. The remainder were taken in nets at sea and
only occasional colour notes were made. Twelve species had been previously described
and six of these are here figured for the first time in colour. The South Georgia speci-
mens were mainly collected from the roots of kelp off the Point, King Edward Cove,
close to the whaling station of the Cia Argentina de Pesca. The water here was from
4 to 6 m. deep and the bottom was very muddy and greasy from the years of whaling
activity at the head of the cove. Throughout the area Linens corrugatus, Mcintosh, was
by far the most common species.

On the Burdwood Bank, off the Falklands, South Shetlands and South Orkneys,
specimens were brought up by the ships while dredging or trawling. The deepest cap-
ture was Amp/iiporus lecointei, Burger, from 401 m. at St. 158.

No representatives of the Paleonemertea were taken. The Hoplonemertea and the
Heteronemertea are strongly represented, the former by nine species each of Tetra-
stemma and Amphiporns , the latter by one or more species of the genera Baseodiscus,
Parapolia, Linens and Cerebratulns .

Order HETERONEMERTEA
Genus Baseodiscus, Diesing

Baseodiscus antarcticus, Baylis, 1915 (Figs. 19, 20).

One specimen (N 59) was taken at St. 182 in 278-500 m. No record exists of its
appearance in life. Two large specimens taken at St. WS 73 were found among the
collection and identified with the previous specimen. These
were accompanied by a note: " Larger specimen 28 cm. long
and 0-5 cm. wide. Both of a pinkish colour, deeper along the
mid-dorsal and mid-ventral lines". The specimen from
St. 182 is a small nearly cylindrical worm, blunt at both
ends, 25 mm. long and 2 mm. in diameter. The colour is very
pale brown with no trace of markings. The larger specimen
from St. WS 73 is 300 mm. long and 6-5 mm. wide, the
smaller 130 mm. long and 3-5 mm. wide. Both are pale
yellow-brown with no markings.

The following notes may be added to the description given preserved specimen, ventral

bv Bavlis (iQicV surface, to show the annular

mi ij- r^iri^ 1 1,,^, furrow and the mouth (m).

ihe body is soft and flattened except at the head. Both

head and tail are blunt ; in fact, when the worms have been preserved it is difficult to

identify the head, it resembles so much a broken end of the body. The back is very

Fig. 19. Baseodiscus ant-
arcticus, Baylis. Head of

248

DISCOVERY REPORTS

wrinkled, the small deep wrinkles giving almost a matted appearance. Down the centre
of the back is a raised ridge with a longi-
tudinal groove on each side of it. Ventrally
the wrinkling is mainly in the form of longi-
tudinal furrows of which the median ones are
the deepest. There is a furrow at the back of
the head and the mouth is just behind this
in the middle line (Fig. 19).

Internal structure. There appear to be
frontal organs as shallow pits at the corners
of a depression at the tip of the head. Head
glands are not evident in the early sections,
but about midway between the proboscis
pore and the brain they appear, stained
palely with haematoxylin, scattered through
the musculature, especially ventrolaterally.
Dorsally they are thin, also mid-ventrally,
and they disappear altogether before the
brain.

The position of the brain, cerebral organs,
proboscis pore and mouth can be seen in
the graphic reconstruction (Fig. 20). The
proboscis is thin and also its sheath. The
mouth is small and rounded. The epithelial
layer is thin and the cutis deep. The
gelatinous tissue developed between the
bundles of the outer longitudinal muscle
layer (Baylis) is extraordinarily well-marked
in the larger specimens and is responsible for
the unusual degree of wrinkling of the skin.

Fig. 20. Baseodiscus antarcticus, Baylis. Diagram
of the organs at the anterior end of the body and
head, co, cerebral organ; dg, dorsal ganglion;
fo, frontal organ; hg, head gland; m, mouth;
wr, deep wrinkles on the dorsal surface of the
head. (From a graphic reconstruction.)

Genus Parapolia, Coe

Parapolia grytvikenensis, n.sp. (Plate XVI, fig. 14; Figs. 21, 22).

One specimen (N 43) was collected at St. 123 from 230-250 m. The contracted length
in life was 4 cm., breadth 0-27 cm.

Form and colour in life. The body is stout and cylindrical in the passive and con-
tracted condition in which the single specimen was found. The head is acutely pointed,
flattened, and is separated from the body by a slight "neck". The tail is blunt. The
mouth can be seen as a small longitudinal slit some way back from the tip of the head.
Neither cephalic slits nor eyespots can be seen. The colour is pinkish brown, darkening
towards the tail.

NEMERTEANS

249

B

Form and colour of spirit specimen. The length is 2-8 cm., the greatest breadth 0-35
cm. The body is round in section anteriorly, flattened from above down posteriorly.
The head is small and pointed
and is separated from the m
body by a shallow circular
depression. The mouth is
mid-ventral — a small slit with
definite lips — situated in the
depression and on each side
are small vertical furrows
(Fig. 21 A).

Anatomy. Frontal organs
are present but head glands
appear to be completely miss-
ing. The proboscis pore is a
narrow slit placed ventrally
not far from the tip of the
head. Before its appearance

in transverse section the vas- Fig 2I Parapolia grytvikenensis, n.sp. A, sketch of preserved
cular lacunae are present. specimen; m, mouth; eg, cephalic slit. B, outline of the head (from

The body layers can be seen above) from a graphic reconstruction showing the brain, cerebral
in Fig. 22. The glands in the organs, and the position of the mouth (m).
cutis are few and incon-
spicuous. The epithelium is
thin. The slender proboscis
has definitely the Lineid ar-
rangement of muscle layers.

The brain is small and /r^^v ',, ^^S^^^^.?- ^■'.^vd?'

ill-defined, the fibrous part
mingling with the muscles.
The cerebral organs are also
small. They lie close above
the ventral ganglia where these
are passing laterally into the
lateral nerves. I could trace no
direct connection between the '*/ *fc>

cerebral organs and the dorsal pig. 22. Parapolia grytvikenensis, n.sp. Part of a transverse section
ganglia. The organs open to of the head at the posterior end of the brain, cc, cerebral canal;
the exterior by fine lateral dS> dorsal ganglion ;p, proboscis; vg, ventral ganglion.

canals (Fig. 21 B).

The anterior end of the mouth follows directly the disappearance of the cerebral
organs from transverse sections. There are no eyes.

5-2

250 DISCOVERY REPORTS

The identification of this specimen with the genus Parapolia rests on the Lineid
structure of the proboscis and the absence of head slits. The specific name is taken from
Grytviken, the name of the original headquarters of the whaling industry of the South
Atlantic, in Cumberland Bay, South Georgia.

Genus Lineus, Sowerby

Lineus corrugatus, Mcintosh, 1887 (Plate XVI, figs. 16, 19, 20, 21 ; Figs. 23-27).

Cerebratithts corrugatus (Mcintosh), Hubrecht, 1887; C. steinini, C. validus, C. subti/is, Burger,
1893 ; C. magelhaensicus, Burger, 1895 ; C. Charcoti, Joubin, 1908 ; Lineus autrani, Joubin, 1908.

Some hundreds of Lineid worms conforming in life to the type L. corrugatus,
Mcintosh, were taken from South Georgia and others were collected at stations made
off the Falklands and the South Shetland Islands. A great number were examined alive
and notes and sketches made of them at the time. A number of methods were used in
fixing and preserving them. Identification with the descriptions given by earlier workers
has proved a difficult task, and the conclusion I have reached is that there is one species
present, widely distributed and very common in this part of the Southern Ocean and as
variable in form, colour and size as L. ruber of European waters.

The following description covers the external appearance of all the specimens in life.

The length is from a few centimetres to fifty or more at South Georgia. Considerably
larger specimens were collected at the South Shetlands. When in motion a 52 cm. worm
was 3 mm. across the broadest part of the head; a 17-5 cm. worm was 1-5 mm. across
the body.

The colour varies from light fawn to greyish black through all shades of fawn, light
reddish brown, greenish brown, dark red-brown and brownish black. The ventral side
is nearly always paler than the dorsal. Two white tags from near the posterior ends of the
cephalic slits pass upwards. The incomplete band thus formed is occasionally complete
and very rarely double. Sometimes the tags are faintly marked and sometimes they are
absent. The tip of the snout is white, and the cephalic slits are lined with white. The
white lining may extend the whole length or it may stop at the tags. The body colour
usually fades towards the tail. The colour of the head may be slightly greenish and some-
times a reddish patch may be apparent on the body just posterior to the head. Occasion-
ally there is a trace of a light median line on the snout. The cephalic slits have a white
granular appearance inside and a trace of red at the hinder end. Irregular light trans-
verse wrinkles are present especially towards the posterior end.

The body is slightly flattened. The head is distinctly flat and somewhat wedge-shaped.
A "neck" is visible when the animal is moving. The mouth is large and no traces can
be seen of pigmented eyespots. The darker colours are more general in the smaller
worms, and the white tags, besides being very much more evident by contrast, appear to
be invariable.

The specimens in spirit or formalin can be divided into three types : (i) Large pale
specimens contracted — often into a spiral with the tail inside — and wrinkled both longi-

NEMERTEANS

251

tudinally and transversely to a greater or less extent at different parts of the body. The
colour that remains is usually stronger dorsally than ventrally (Fig. 23 A). These may have
a long slit-like mouth and firm unprotruded lips or the mouth may be a small slit with
pursed lips (Figs. 23 B, D, E). (ii) Elongated uncontracted forms of very pale uniform
colour, much thinner than type (i) . The mouth is very large and the lips are thin, protruded
and distorted (Fig. 23 F). (iii) Small specimens coiled spirally with small mouths and
pursed lips. The colour is dark and often the same dorsally and ventrally (Fig. 23 C).

A ■**

o

B

E 1- a

Fig. 23. F'S- 24.

Fig. 23 . Lineus cormgatus, Mcintosh. A, preserved specimen of the large type (i) ; B, E, specimens of type (1)
with the mouth pursed, the former from Port Stanley, Falkland Islands; D, specimen of type (i) with the
mouth unpursed; C, type (iii); F, the elongated type (ii).

Fig. 24. Lineus cormgatus, Mcintosh. A, transverse section of the rhynchodaeum near the proboscis pore;
B, section of the rhynchodaeum farther back than A, showing the outpushings (op), x 175.

That these differences were not discernible in life can be judged from the fact that the
different types were sometimes preserved together from the same haul which suggests
individual reactions to the fixing fluids. In April 1927 many brownish, reddish and
black specimens were dredged from red algae and stones in King Edward Cove. There
was no apparent difference in shape, although it was noticed that the brown forms were
perhaps rather stouter in build than the others. The roots and worms were left outside
the Biological Station for the night, during which the surface water in the pan froze
after being mixed with snow. On the following day some of the worms had relaxed and
taken on the appearance of type (ii), but on killing with hot water they contracted to the
type (i) like the remainder of the catch.

252

DISCOVERY REPORTS

B

Fig. 25. Lineus corrugatus, Mcintosh. A series of diagrammatic half-sections at the level of the extreme
anterior end of the mouth. A, type (i) ; B, type (i) corresponding to B of Fig. 23 ; C, type (iii) ; D , type (i) corre-
sponding to D and E of Fig. 23; E, type (ii). cc, cerebral canal; co, cerebral organ; dg, dorsal ganglion;
In, lateral nerve ;pn, pharyngeal nerve; vg, ventral ganglion.

NEMERTEANS

253

An investigation was made of the length of the mouth and of the cephalic slits relative
to the body length in the preserved specimens with the interesting result that the ap-
parently very large mouth of type (ii) is actually smaller for the length of the animal
than it is in either of the other types.

Internal anatomy . Frontal organs are present. The head glands are thin and scattered.
They stain with haematoxylin and near the tip of the head are grouped into three areas.
They do not reach the brain. The vascular and nervous systems have been well described
by previous authors. In the rhynchodaeum there are outpushings noted by Joubin and
considered by him to connect the blood lacunae with the exterior. Although they pro-

.Iki

A

mi

olm $g£&9i#

B

D

Fig. 26. Litieus corrugatus, Mcintosh. The dorso-lateral epidermal layers in transverse section. Reference
letters as in Fig. 25. bm, basement membrane; c, cutis; e, epithelium;/?, fibrous layer; olm, outer longi-
tudinal muscles.

trude into the lacunae I have not been able to find any gap in the intervening tissue that
would suggest free communication. These outpushings vary in number but have been
seen in all types (Fig. 24).

The divergences in the anatomy of the three types are connected with the relative
position of the organs (Fig. 25) and in the gland cells of the cutis and the cutis itself.
Fig. 26 shows the forms of the dorso-lateral cuticular layers at the level of the anterior
end of the mouth. Type (i) possesses the thick muscle-free basement membrane of
Cerebratulus charcoti, and of C. corrugatus as described by Burger (1904, p. 96); type
(iii) the thinner layer with circular fibres of C. steinini. A second series of type (i) shows

254

DISCOVERY REPORTS

the structure described in C. validus, Burger. During the examination of the material
by means of hand sections cleared in anilin oil it was generally found that the basement
membrane conformed to type, but sometimes it did not. On cutting diagonal and longi-
tudinal sections an explanation, based on the contraction of the muscles of the body,
was found to cover the differences. When the body is contracted the cutis is thrown into
circular wrinkles between which the basement membrane
is compressed, giving the appearance of type (iii), while
where the cutis bulges the type (i) cross-section occurs.

No eyespots could be seen on clearing in anilin oil. In
the sections, however, a series of curious organs can be
observed close to the cephalic slits (Fig. 27). These are
spherical bodies that produced the granular appearance of
the inside of the slits in the living animal. They may
possibly have visual function although they appear very
similar in structure to fibrous nerve tissue.

I give below a list of the stations at which specimens were Intosh. Part of a transverse
captured with the serial numbers of the specimens of which sect[on of the head at the base

^ . of the cephalic slit, eg, to show

special investigation was made and sections cut. As 1 have the spherical granuiar bodies, eo.
remarked this species is common in the area and a large

number of specimens not included in the list below were taken from King Edward
Cove and also from Port Stanley Harbour, Falkland Islands, under stones at low tide.
List of stations at which Lineus corrugatus was taken. The dates, positions and depths
of these stations will be found in the list of stations on pp. 220-5.

Fig. 27. Lineus corrugatus , Mc-

Station

Gear

No. of
specimens

Serial
number

Station

Gear

No. of
specimens

Serial
number

St. 39
St. 45

OTL

18

St. WS 73

OTC

6

N128

OTL

4

N 134

St. WS 77

OTC

2

N 88, N 89

St. 51
St. 53

OTL

16

N 126

St. WS 79

OTC

1

N 124

RM

8

N 132

St. WS 80

OTC

1

—

St. 123

OTL

1

—

St. WS 84

OTC

4

—

St. 160

DLH

1

N71

St. WS 88

OTC

2

N 133

St. 163
St. 164

BTS

1

—

St. WS 225

OTC

2

N61, N95

BTS

1

— ,

St. WS 228

OTC

1

N55 '

St. 167

N7-T
N4-T

3

5

N 101

NCS-T1
N4-T )

2

N 113, N86

St. 173
St. 186

BTS

2

—

St. WS 237

NCS-T

2

N74, N51

DLH

1

—

St. WS 246

OTC

1

—

St. 195

St. WS 25

OTM

1

—

St. WS 248

OTC

1

—

BTS

42

—

St. WS 249

DLH

2

N 66, N 67

St. WS 56

NH

3

—

St. MS 68

NRL

1

—

St. WS 62

BTS

6

_

NEMERTEANS

255

Lineus longifissus (Hubrecht).

Cerebratulus longifissus, Hubrecht, 1887.

One hundred and thirty-four fragments including eleven heads and four tails were
preserved at St. 167 (N 81, N 81 a). Fourteen complete worms were examined later.
They had been caught in another net at the same station.

Two worms were reconstructed from reddish backed fragments giving approximate
lengths of 7-0 cm. The breadth and depth were 07 and 0-35 cm. respectively.

The colour of the spirit specimens is reddish or greyish on the back and pale grey or
white beneath. The body is flattened, anterior end more cylindrical than posterior, and
the tail is pointed. The surface of the skin shows slight circular wrinkling especially
anteriorly. As remarked by Hubrecht the mouth is small and the cephalic slits very
long. They become shallower gradually, but appear much more sharply cut than in other
Lineids. The following measurements were made of complete worms and fragmented
heads.

Length of frag-
ment from tip
of head
cm.

Greatest

breadth

cm.

Length of

cephalic

slits

cm.

Length of

mouth

cm.

Distance from

tip of head to

anterior end of

mouth

cm.

39

3'3

2-55

3-5

3-5

4'1

Complete worm
5-8
5-25

o-45
0-625

°-5

o-6

o-5
o-4S

o-43
°-45

2-1

27

1-9

2-05

i-8

1-65

i-6
i-6

0-225

0-20
0-05
0-I5
0-05

0-06

o-i
0-07

o-5S
0-65

°-4S
07

o-45
o-6

0-4
0'5

I can add the following notes to the account given by Hubrecht. Frontal organs are
present. Head glands are diffuse, stain with haematoxylin and do not reach the anterior
end of the brain. The small cephalic canals pass from the fissures just after the level of
the ventral commissure. Before they penetrate the brain a dorsal branch is given off
by the dorsal ganglion. This branch is extremely short. The posterior lobes of the brain
lie in a blood sinus.

One of the characters of the species is the very marked power of autotomy.

Lineus roseocephalus, n.sp. (Plate XVI, fig. 24).

With the dark brown L. corrugatns collected under stones in the harbour of Port
Stanley, Falkland Islands, was this light red form represented by a single specimen
(N 22) 45-0 mm. long and about i-o mm. in diameter. In addition to the colour, dif-
ferences from L. corrugatus were readily perceptible in the tapering shape of the body
and the rounded anterior end. No mouth was visible but the cephalic slits were very

256 DISCOVERY REPORTS

long. The tip of the head was brown and the anterior end of the body was of a more
crimson tint than the body farther back.

Anatomically the worm is very closely allied to L. corrugatus. The "eyes" are not
present, however, and the mouth is very small.

Genus Cerebratulus, Renier

Cerebratulus larseni, n.sp. (Plate XVI, fig. 8; Fig. 28).

One specimen, somewhat damaged at the posterior end, was taken at St. 140 in 122-
136 m. The length was 2-3 cm., breadth 0-14 cm.

The body is round in section but the head is flat. In outline from above it takes the
form of an elongated lozenge. The mouth is a small longitudinal slit with swollen lips
just behind the pinkish blotch on the head. The
colour is pale yellow with a bright pink vague
patch on the head and a pink line showing
down each side of the body. A caudal ap-
pendage is present.

In spirit the specimen had broken up. The
colour was bleached.

The head is rectangular in the early sections.
The vascular and nervous systems conform to
type but the upper branch of the dorsal ganglion Fig. 2g. Cerebratulus larseni, n.sp. Tr

is Stout and short. It is separated at its distal sectionof the head in the posterior brain region,
end from the rest of the ganglion (Fig. 28). cc> cerebral canal ;dg, dorsal ganglion ; «/>, upper

m, , • j- 1 11 r a -l j branch of the dorsal ganglion; vp, ventral

lhe longitudinal muscle layer or the body is ,. " b

.... ganglion.

strongly developed. The cutis is thin in com-
parison with the same layer in Linens corrugatus. The basement layer is as thick as the
subepithelial muscle layer.

I have named this species after Captain C. A. Larsen, the pioneer of whaling in the
South Atlantic.

Cerebratulus malvini, n.sp. (Figs. 29, 30).

Certain of the Heteronemerteans from the area around the Falkland Islands possessed
a caudal appendage. These worms were not sketched or noted in life so that the colour
remains unknown. As far as one can judge from specimens in alcohol (N 52, N 65, N 72,
N 97, N 100) the colour is very dark brown, darker on the back than the underside.
No trace of markings remain. The lengths and breadths of four specimens are as follows:
36-0, 3-5 mm.; n-o, i-o mm.; 55-0, 2-0 mm.; 45-0, 3-5 mm.

The cephalic slits are long and the mouth is very small. In one specimen (N 52) the
caudal appendage was double (Fig. 29B). Apart from the absence of "eyes" there
appears to be no feature by which one can distinguish the sections of these worms from

ransverse

NEMERTEANS

257

Linens corrugatus. The cephalic slits are wider and larger altogether (Fig. 30), and this,
with the absence of " eyes " and presence of a caudal appendage characterize the species.

Fig. 29. Cerebratulus malvini, n.sp. A, B, outline
sketches of preserved specimens.

Fig. 30. Cerebratulus malvini, n.sp. Transverse
section of the head near the tip showing the wide
cephalic slits.

The specimens were taken at the following WS stations: 73 (N 52), 228 (N 100), 239
(N 72), 246 (N 97), 249 (N 65).

Order HOPLONEMERTEA
Genus Amphiporus, Ehrenberg
Amphiporus falklandicus, n.sp. (Fig. 36 D).

This species is represented by a number of specimens. It is closely related anatomi-
cally to A. gerlachei, Burger and A. lecointei, Burger. I have separated them on the
following grounds: the colour in life from a single colour note; the size and shape of
the preserved specimens and the position of the junction of the oesophagus and stomach
with relation to the brain.

The station numbers and details of gear and depth, with the length, breadth and serial
numbers of the specimens are given in the table below :

St.

Date

Gear

Depth

Length

Breadth

Serial

m.

mm.

mm.

no.

WS84

24. iii. 27

OTC

75-74

WS97

18. iv. 27

OTC

146-145

55-o

3-3

N 122

WS225

9. vi. 28

OTC

162-161

48-0
36-0

2-5
4-0

N62
N96

WS228

30. vi. 28

OTC

229-236

47-0
40-0
36-0

4-8
4-0
2-5

N84
N85

WS246

19. vii. 28

OTC

267-208

11-0

2-0

N98

WS248

20. vii. 28

OTC

210-242

—

—

6-2

258

DISCOVERY REPORTS

The body is more elongated and slender than that of A. gerlachei, though the posterior
end of the body is flattened as in this species (Fig. 36 D). The anatomy is similar except
that the oesophagus does not open into the stomach until after the brain. This is similar
to A. lecointei, but from this form A.falklandicus differs in colour and shape. There are
twelve nerves in the proboscis and an accessory armature of two reservoirs with from two
to six stylets. Head glands and cerebral subepithelial glands are present and the cerebral
organs persist behind the dorsal ganglia.

Male and female specimens were included in the collection and the ova of the latter
contained the "paranucleus" remarked upon by Hubrecht in A. marioni.

Amphiporus gerlachei, Burger, 1904a (Figs. 31, 36 C).

This species appears to be fairly common, though it was not captured in King
Edward Cove and no sketch was made of the living animal. From a colour note and
three series of sections I have identified it with
A. gerlachei, Burger. Seventeen specimens
taken from the base of a large hollow sponge at
St. WS 225 were noted as " bright pink " in life
(N94).

The lengths and breadths of preserved speci-
mens were : 35-0, 6-o mm. (N 63) ; 35-0, 5-5 mm.
(N64); 30-0, 7-0 mm. (N94); 40-0, 7-0 mm.;
50-0, 7-0 mm.

The anterior end is cylindrical, the posterior

very flat (Fig. 36C). The oesophagus opens Fig. 31. Amphiporus gerlachei, Burger. Trans-
into the stomach in front of the brain (Fig. 31). verse section of the head in front of the brain.
In other ways the anatomy corresponds to that "s' cerebral subePithelial Slands; «> eyesPot =

r /i 7 • ■ mi , A?> head glands; p, proboscis; s, stomach.

ot A. lecointei. 1 here are twelve nerves in the

proboscis, and the accessory armature consists of two reservoirs with from two to five
stylets. The sex was determined in one specimen only — a male.
A. gerlachei was taken at the following WS stations:

WS 225. 9. vi. 28. OTC, 162-161 m. (N 94).

WS 246. 19. vii. 28. OTC, 267-208 m.

WS 249. 20. vii. 28. DLH, 166 m. (N 63, N 64).

Amphiporus inexpectatus, n.sp. (Fig. 32).

This species is represented by serial sections of a single specimen (N 108) taken at
St. WS 231 off the Falkland Islands and not noted in life. The preserved specimen was
16-0 mm. long and 2-0 mm. in diameter, round in section and bleached of all colour. The
proboscis was protruded but no armature could be made out after clearing.

Anatomy. The head glands open at the tip of the head above the proboscis pore.
They form a thick strand close to the rhynchocoel and do not stain with haematoxylin
(Fig. 32). They disappear before the brain. The epithelium at the level of the posterior

NEMERTEANS

259

end of the stomach is about twice as thick as the basement layer which is considerably
thicker than the circular muscle layer. The longitudinal
muscles are well developed.

There appears to be no distinction between oeso-
phagus and stomach. A large folded tube passes through
the region of the brain and opens by a frilled mouth
under the proboscis pore. The tube enlarges slightly
posteriorly. Branches of the anterior caecum reach the
brain.

The proboscis possesses thirteen nerves and is stout
and muscular. There is no dorsal strand in the lateral
nerves. The cerebral organs are small and only just reach
the brain. Their short canals open ventro-laterally. There
are about fifteen eyespots on each side. The specimen
was a male with immature testes.

I have separated this worm from the other Amphi-
porids on the nerves in the proboscis and the position and size of the cerebral organs.

Fig. 32. Amphiporus inexpectatus ,
n.sp. Transverse section at the
extreme tip of the head with the
protruded proboscis, p. hg, head
glands.

Amphiporus lecointei, Burger, 1904 a (Plate XVI, fig. 9; Figs. 33-35, 36B)-

This species was taken at the following stations :

St. 27. 15. iii. 26. DL, nom. 1 (N 114).

St. 140. 23. xii. 26. OTL, 122-136 m. 4 (N 45).

St. 156. 20. i. 27. DLH, 200-236 m. 2.

St. 158. 21. i. 27. DLH, 401 m. 2.

St. 159. 21. i. 27. DLH, 160 m. 4.

St. 195. 30. iii. 27. OTM, 391 m. 2.

St. WS 25. 17. xii. 26. BTS, 18-27 m- 2 (N I20)-

St. WS 93. 9. iv. 27. N 7-T, 133-130 m. 6 (N 125).

St. MS 71. 9. iii. 26. NCS-T, 1 10-60 m. 1 (N 115).

One specimen (N 45) was examined and sketched in life and afterwards sectioned.
This, though similar in many respects to A. michaelseni, Burger, has been separated
from it by reason of the cerebral organs, armature and the number of nerves in the pro-
boscis. It has been identified with A. lecointei, Burger, on the shape, head glands, brain
and cerebral organs. Much should have been added to make the original description
adequate, for there is great similarity in the anatomy of the closely related species
lecointei, gerlachei, falklandicus and marioni.

The length was 20 mm. and breadth 2-3 mm. A specimen sketched by Mr D. D.
John at St. 156 was 23 mm. long.

Form and colour in life. The size and shape is remarkably uniform. The body is stout,
almost circular in cross-section, bluntly pointed at the head and tapering to the tad. The
head is marked off from the body by a transverse groove deep ventrally and incomplete
dorsally. At each side the groove takes the shape of a backwardly directed V, and from

a6o

DISCOVERY REPORTS

Fig. 33. Amphiporus lecointei, Burger. A, dorsal, and
B, ventral sides of the head.

it ventrally there are several furrows passing forward upon the head (Fig. 33 B). The
opening of the head gland can sometimes be seen at the tip of the head : the larger
opening of the rhynchodaeum is subterminal.

There is a semi-lunar group of about twenty deeply embedded eyespots showing
palely through the skin passing outwards on each side from the tip of the head along the
margin and turning medially. Follow-
ing these eyespots there is a deep
closely-set posterior group of eyes
behind the cephalic furrow of each
side (Fig. 33 A).

The distinctive marking of the
species is a broad brownish red band
down the back extending on to the
head. The edges of the body and the
underside are uncoloured.

Form and colour of preserved speci-
mens. The stout body does not twist
much. Often the ventral side is more
convex than the dorsal and there is a

tendency for the body to curl with the dorsal surface inside. Occasionally the ventral
surface is flat or concave while the dorsal is humped. The proboscis is usually pro-
truded. It is nearly the same length as the body. The colour can in some specimens
be traced as a grey band. The eyespots are large. On clearing, the anterior group can
be seen to consist of 10-20 on each side opening forward, the posterior of 16-18 opening
laterally or posteriorly.

Anatomy. The basement membrane stains somewhat with haematoxylin and is
nearly as thick as the epithelium, which is itself
very thick. Each of these layers is four to five
times as thick as the circular muscles. The longi-
tudinal muscles are thick and show a marked
pennate arrangement of the bundles (cf. A.
marioni, Hubrecht). There are subepithelial
glands in the head confined to lateral tracts from
the tip to just beyond the cerebral canals. These
I propose to call cerebral subepithelial glands.
They differ in appearance and staining reaction
from the head glands and their ducts can be
seen traversing the body layers direct to the
exterior (Fig. 34). The head glands just reach the
brain. They are compact strands opening by a
median pore at the tip of the head.

The proboscis is thin and is attached at about

Fig. 34. Amphiporus lecointei, Burger. Trans-
verse section of the tip of the head, bm, base-
ment layer; cc, cerebral canal; cm, circular
muscle layer; csg, cerebral subepithelial
glands; e, eye; ep, epithelium; Im, longi-
tudinal muscle layer; rd, rhynchodaeum.

NEMERTEANS

261

half the length of the body, but
the rhynchocoel extends to the
tail. The proboscis has twelve
nerves. The armature consists of
a main stylet on a base not longer
than itself and two reservoirs, each
with from four to seven stylets.

The relative positions of the
stomach, excretory organs and
branches of the anterior caecum
are believed to vary with the
degree of contraction of the body.
Thus in one series of sections the
stomach wall is visible some
distance before the brain, showing
that the alimentary canal has the
power of independent movement
and that not much reliance can
be placed in distinguishing charac-
ters based on position. The anterior
caecum has branches which are
first seen close behind the excre-
tory tubules (Fig. 35).

The vascular system consists of
two lateral vessels forming a head
loop and passing one on each side
of the rhynchocoel through the
brain region. They unite with a
dorsal vessel by a transverse con-
nection and pass down the body
ventral to the nerves.

There is a convoluted excretory
tubule on each side close behind
the cerebral organ, lying above
the nerve and opening to the ex-
terior by a single dorso-lateral
duct.

The brain is of good size. The
dorsal ganglia are larger than the
ventral and lie immediately over
them. At the posterior end of the
ventral ganglia there maybe a con-

Fig. 35. Amphiporus lecointei, Biirger. Diagram from a graphic
reconstruction of the head and anterior end of the body to show
the relations of the brain, cerebral organs, excretory and
alimentary systems, ac, anterior caecum; co, cerebral organ;
csg, cerebral subepithelial glands; dg, dorsal ganglion; e, eye-
spots; ex, excretory tubule; hg, head glands; In, lateral nerve;
phg, opening of the head gland; rd, rhynchodaeum.

Fig. 36. Outline drawings (x 3 approx.) of Amphiporus
marioni (A), A. lecointei (B), A. gerlachei (C), and A. falk-
landicus (D).

262

DISCOVERY REPORTS

traction twist when they become the lateral nerves. No dorsal strand could be detected
in the lateral nerves. The cerebral canals are short. Their openings are ventro-lateral.
The organs themselves are large and at first closely applied to the sides of the dorsal
ganglia. Fibres from the ganglia pass into them and form the only connection. They
shift ventrally, wedging themselves between the dorsal and ventral ganglia. On the dis-
appearance of the dorsal ganglia the organs also diminish and end. Fig. 35 shows the
relative positions of the various organs in the head from a graphic reconstruction.
Reference to Fig. 37 will show how closely the main anatomical features of this species
resemble those of A. marioni, allowing for the greater degree of contraction in the latter
and the fact that the proboscis is considerably thicker. In Fig. 36 outline drawings of
spirit specimens of the four species are given illustrating the differences in shape of the
body. A. lecointei should be easily recognized in life by its distinctive form and colour.

Amphiporus marioni, Hubrecht, 1887 (Figs. 36 A, 37, 38).

One specimen (N 5) was taken from a kelp root from King Edward Cove. It was
not sketched until the following day when it was sorted from A . moseleyi with which it
had been fixed. The colour and general appearance are probably very similar. The
length was 9-3 mm., the breadth 1-5 mm. (in spirit).

The small stoutly built body is flattened posteriorly. No eyes can be seen (Fig. 36A).

Anatomy. In the region of the stomach the longitudinal muscle layer is three or four
times as thick as the epithelium. Farther back down the body it is less than twice as
thick and in some positions — mid-ventrally under the gut and mid-dorsally— it is
thinner than the epithelium. The basement membrane is very thick and appears fibrous.

The head glands are large and compact. One
strand, dorsal to the rhynchocoel, opens by a pore
ventral to the tip of the snout. This strand does not
reach the dorsal commissure. There are two lateral
strands, one on each side of the rhynchocoel, which
extend back ventral to the brain as far as the ventral
commissure and opening of the cerebral canals
(Fig. 37). There are cerebral subepithelial glands on
each side from the tip of the head to the cerebral
organs. About fifteen eyespots with very light brown
pigment are present on each side.

The oesophagus opens into the rhynchodaeum

before the brain and at the level of the cerebral

organs it becomes the stomach. The anterior caecum

is much branched and the branches extend forwards

and approach the posterior end of the brain. The

lateral diverticula of the gut extend above the lateral „. „ , . . . „ , , „

fig. 37. Amphiporus marioni, Hubrecht.

nerves (Fig. 38). The proboscis is thick. There are Graphic reconstruction of the anterior
fifteen nerves. The armature consists of a main stylet end of the body.

NEMERTEANS

263

and two accessory stylet reservoirs with one or two complete stylets and a number
of fragments. The rhynchocoel extends to the posterior end of the body. The vascular
and excretory systems are similar to those of A. lecointei.

The brain is small. The dorsal ganglia are considerably larger than the ventral
and lie immediately over them. At the posterior end of the ventral ganglia the
lateral nerves are given off sharply at right angles to the long axis of the body
(Fig- 37)- They appear to double upon themselves again just after turning down
the body. The twists are vertical, instead
of lateral as described in A. moseleyi, and
are plainly the effects of contraction. The
cerebral canals are short. Their openings
are ventral and lateral into a deep circular
groove near the tip of the head; The
organs themselves are large and closely
applied to the sides of the dorsal ganglia,
but posteriorly they push between dorsal
and ventral ganglia and extend back
beyond the former.

Fig. 38. Amphiporus marioni, Hubrecht. Transverse
section of the body at about half its length. In, lateral
nerve; ov, gonad.

The generative sacs (female) are both dorsal and ventral to the gut branches and the
nuclei of the ova contain a deeply staining nucleolus. The "paranuclei", as described
by Hubrecht (1887), are present.

The features that have caused a distinction to be made between this species and
A. gerlachei, Burger, which it closely resembles when preserved, are the number of
nerves in the proboscis and the colour. A reddish worm would undoubtedly have been
separated from the specimens of A. moseleyi on capture.

Amphiporus moseleyi, Hubrecht, 1887 (Plate XVI, figs. 3, 18; Fig. 39).

Amphiporus Racovitzai, Burger, 1904 a.

This species (N 4) was very common in King Edward Cove and round the coast of
South Georgia. Almost every kelp root sheltered one or more specimens, and a con-
siderable range in colour and size was observed. The largest specimen was 10-2 cm.
long and 12-0 mm. broad. Specimens were also taken at the following stations:

St. 123. 15. xii. 26. OTL, 230-250 m.

St. WS219. 3.V1. 28. NCS-T, ii6-ii4m. (N 79).

St. MS 67. 28. ii. 26. BTS, 38 m.

Form and colour in life. The usual length is 5-6 cm. and the breadth 7-8 mm. The
body is stoutly built and considerably flattened. The ventral surface is flat and forms a
broad "sole". The dorsal surface is convex, especially anteriorly where a longitudinal
hump marks the presence of the muscular proboscis. There is no distinction of head
from body, but the head end is rather less blunt than the tail. In outline the body
resembles a willow leaf. No cephalic slits or eyespots are visible. The opening of the

J 64

DISCOVERY REPORTS

proboscis pore is just ventral to the tip of the snout. The colour is blue-green, yellow-
green, pale buff or light brown on the back, while the underside is always pale buff. The
colour is deepest in individual worms on the hump caused by the proboscis. Occasionally
a reddish tinge marks the position of the ganglia and a narrow whitish stripe at the margin
of the body anteriorly the lateral glands.

Form and colour of preserved specimens. After anaesthetization in chloral hydrate very
little contraction appears to take place on fixing. The green colour is retained in spirit
specimens for many months. The body does not change in shape but contraction causes
two grooves to appear, one near the tip of the head, the other a little farther back. When
only slightly marked these grooves take the form of two pairs of short lateral vertical
furrows. Two irregular groups of very small eyespots — up to sixty — can be seen when
the head has been cleared in anilin oil. The eversible part of the proboscis is apple green
in colour.

The anatomy of this species has been described by Hubrecht (1887) and Burger
(1904 a, 1907). Variations occur in the following details. The eyespots, though always
very small, vary considerably in number,
and the pigment granules which they con-
tain are of a deep green-blue colour.

The lateral glands (very strongly de-
veloped cerebral subepithelial gland cells)
stain deeply with haematoxylin ; the head
glands less deeply. The number of nerves
in the proboscis may be from eleven to six-
teen though the usual number is fourteen.
The base of the main stylet may be brown
or green and there is a belt of brown gland
cells round the armature. The base of the
main stylet is less than twice as long as the
stylet itself. The brain and cerebral organs
are shown in Fig. 39. As noted by Burger
(1907 (1912), p. 173) the dorsal ganglia lie
laterally rather than dorsally to the ventral.
They are, as it were, pressed forwards and
outwards so that they are in almost the same
plane as the ventral ganglia. In consequence
the dorsal commissure is very long. The
double twist in the lateral nerves is probably
the result of contraction. The cerebral organs, consisting of a narrow canal ensheathed
by eosinophile tissue on each side, extend a short way beneath the dorsal ganglia but
do not penetrate them.

Fig. 39. Amphipoms moseleyi, Hubrecht. Diagram
of the anterior end of the body showing the brain,
cerebral organs, cerebral subepithelial glands and
the position of the opening of the rhynchodaeum
and the opening of the oesophagus into it.

NEMERTEANS

265

£>>

A B

Fig. 40. Amphiporus schollaerti, n.sp.
A, ventral surface of head of preserved
specimen; B, dorsal surface.

Amphiporus schollaerti, n.sp. (Figs. 40, 41).

A single specimen (N 58) of a large worm was taken at St. 182 in the Schollaert
Channel, Palmer Archipelago. The colour as noted in life was pale buff. The preserved
specimen was 97 mm. long, and 5-0 mm. broad.
It was almost round in section from one end to the
other except the head which was blunt and flat
(Fig. 40). There was a pore visible at the tip and
a larger aperture just ventral to this. Lateral vertical
grooves were also present, joining ventrally. The
body was slightly dusky at the sides but otherwise
colourless.

The anatomy is similar to A. lecointei with the
exception of the proboscis which has fourteen
nerves. Head glands, cerebral subepithelial glands
are present and the cerebral organs occupy a position
between the ganglia and extend behind the dorsal
ganglia (Fig. 41).

The excretory tubules are large and the anterior
caecum has forward branches but they are far be-
hind the brain.

I have separated this worm from the other
Amphiporids on its size, colour and innervation
of the proboscis. The armature was not seen. The
specimen was male.

Amphiporus scoresbyi, n.sp. (Figs. 42, 43).

Three specimens of this small distinctively
shaped species were obtained at St. WS 302 (N 77),
the colour note being "orange-yellow to orange-
red dorsally and ventrally with very pale yellow
periphery and proboscis. Darker pigment in mid-
line posteriorly". Two specimens were taken at
St. WS 548 and one at St. WS 550. These were of
the same size and shape. They had been preserved
in formalin and all colour was bleached.

Form and colour of preserved specimens. The
body is fusiform, the length 5-5 mm., diameter of
the mid-body 2-0 mm. Its surface is thrown into small rounded eminences which
at the anterior end form three or four rows of short fimbriae (Fig. 42). The colour
is uniformly yellowish with no sign of eyes or markings.

Anatomy. The head glands open at the tip of the head just before the proboscis pore.
The glands stain deeply with haematoxylin. The thin strands disappear before the

Fig. 41.
Diagram

Amphiporus schollaerti, n.sp.

showing the organs at the
anterior end of the body. From a graphic
reconstruction, co, cerebral organ; csg,
cerebral subepithelial glands ; dg, dorsal
ganglion; e, eye; hg, head gland.

7-2

266

DISCOVERY REPORTS

brain, but anteriorly they fill the head. The oesophagus joins the rhynchodaeum very
close to the external opening and is from this point lined with the deeply staining cells
characteristic of the stomach. Its walls are folded (Fig. 43). I could see no eyespots in
the cleared specimens nor were any evident in the sections. There are no cerebral sub-
epithelial glands, but there are cutis glands in the longitudinal muscles.

The basement layer is very thick and stains lightly. The epithelium in all the speci-
mens is much wrinkled but apparently it is not much thicker than the basement mem-
brane. The circular muscles are less than half as thick. The longitudinal layer is well
developed but does not show the pennate arrangement of bundles so noticeable in

Fig. 42. Amphiporus
scoresbyi, n.sp. Sketch
of a preserved speci-
men, x 8 approx.

Fig. 43. Amphiporus scoresbyi, n.sp. Section, slightly
oblique, across the brain region, ac, lateral branch of
the anterior caecum; bm, basement layer; cm, circular
muscle layer; co, cerebral organ; ep, epithelium.

A. marioni and other species. The proboscis is stout. It has twelve nerves and the
accessory armature consists of two reservoirs each with two or three stylets. The main
stylet could not be seen.

Two lateral branches of the anterior caecum extend forward and end above the brain
(Fig. 43). The excretory tubules, as usual, lie between the cerebral organs and the
branches of the anterior caecum. The efferent duct is continued back above the lateral
nerve and opens laterally much nearer the tail than the head.

The brain is of fair size, both ganglia being nearly equal. The cerebral organs open
by a fine canal laterally and run in towards the brain obliquely back. The organs wedge
themselves between the ganglia and protrude posteriorly beyond the dorsal ganglia with
which they have nervous connections. The lateral nerves are not much flattened in the
body. They join above the anus.

The sex of the sectioned specimen could not be determined.

Amphiporus spinosus, Burger, 1893 (Plate XVI, fig. 22; Figs. 44, 45).

A. spinosissimus, Burger, 1893; A. cruciatus, Burger, 1893; A. multihastatus, Joubin, 1914.

This species could nearly always be found in kelp roots. Three types were originally
described under N 3, N 16, N 20, the different sizes, number of eyespots, colour and

NEMERTEANS

267

states of contraction leading to confusion. The length and breadth vary from 25 and
o-8 mm. to 180 and 2-8 mm., though the majority are shorter and broader than the
latter.

The body is round, tapering at the posterior end. The snout is broadly acute. There
may be no distinction between head and body, or the head may be broader than the
broadest part of the body (in the smaller specimens) and two partial lateral furrows may
be present, almost vertical in direction, with a chevron groove behind the head. This
groove is complete ventrally. Sometimes the aperture of the rhynchodaeum can be seen.
The colour is pale pink, pinkish red, brick red, orange or light orange red, usually
deeper on the back and anteriorly. Sometimes the underside of the body is much paler
than the back. There are two groups of eyespots
on each side, often appearing as vague blackish
patches.

Form and colour of preserved specimens. The
lengths and breadths of a number of specimens
are as follows: 51-0, 5-0 mm.; 31-0, 3-2 mm.;
6o-o, 4-5 mm.; 85-0, 4-5 mm.; 14-0, i-8mm.;
32-0, i-8 mm.; 130-0, 3-5 mm.

The body is round in section, the head blunt
and the tail acute. The head is usually marked
off from the body by a distinct and almost
complete annular groove ; dorsally this groove
forms a wide V with the apex pointing back
(Fig. 44). Anterior to this there are often lateral grooves, as in Fig. 44, and sometimes
a median vertical groove in which can be seen the openings of the head gland and
rhynchodaeum. The eyes are not all visible unless the specimen is cleared. On clearing
two groups of brown cup-shaped eyespots can be determined on each side. The eyes
vary from fifty or so in each group to small numbers like seven in the anterior and four
in the posterior group of one side. They appear to increase with the size of the worm.
The anterior eyes open forwards, the posterior more to the side, and they vary in size.
The colour is usually bleached, but sometimes a faint general pinkish tinge is observed.

Anatomy. The head glands open at the tip of the head. There is a main compact
dorsal strand which is joined by a smaller ventral strand under the vascular loop just
anterior to the opening of the rhynchodaeum (Fig. 45 A). The dorsal strand becomes
thin and scattered posteriorly and does not extend to the brain. The ventral strand
forms an investment to the rhynchodaeum on each side and continues back past the
junction with the oesophagus. It, too, disappears before the brain.

The epithelium is very thick. In places it is thicker than basement membrane, circular
muscles and longitudinal muscles put together. The basement membrane itself is thick
and stains strongly. In the head, at about the level of the cerebral organs, there are
eosinophile glands in the longitudinal layer whose ducts pass through the circular layer
and basement membrane. These are not seen farther down the body.

Fig. 44. Amphiporus spinosus, Burger.
A, ventral ; and B, dorsal surface of head.

268

DISCOVERY REPORTS

The oesophagus is at first thin and small. It expands in the region of the brain into
the stomach. Far back there is an anterior caecum with two short forward branches.
The vascular system is normal and the excretory tubules occur just posterior to the
brain.

The proboscis as far as the armature is almost half the length of the body. The
accessory armature and proboscidial nerves vary greatly, though there is always a main
stylet mounted on a brownish pear-shaped base (Fig. 45 B). The following range of
variation has been found :

Serial

Length of

Accessory

Number of stylets

Number of

No.

worm, mm.

reservoirs

in each reservoir

nerves

N16

6o-o

18

1

N 16
N 20

90-0
14-0

18

7

1(2)
1 (2) (3)

14

N 20

32-0

2

3

—

N 20

I TO

6

1

n

N 20

30-0

8

2(3)

—

N3

130-0
65-0
60 -o

38-0

30-0

12

13

9
8

2(3)
2(3)
2(1)
2(1)
2(3)

17

N3
N3
N3

62-0

7
18
12

2(1)
2(1)

17

IS

is

N3

—

11

2

15

N3
N 16

92-0

16
22

2(1)
1(2)

17
18

N 20

—

5

—

—

N3

—

—

16

N no

75'°

J3

2(1)

r9

N 121

45'°

22

1

—

—

27-0

10

2(1)

17

N 131

Ni35
N 136

50-0

95'°
27-0

13

—

*7

26

16

It seems evident that no specific value can be given either to the armature or to the
proboscidial nerves. There is a possibility of increase with size or age.

The brain is fairly large but does not show any peculiarity. In one series of sections
the lateral nerves left the brain by a sharp twist outwards, but this is merely the effect
of contraction. The cerebral organs open by two very small pores ventro-laterally behind
the opening of the rhynchodaeum. The organs are small and do not reach the brain.

Some of the specimens were males, some females. The gonads are shed laterally both
above and below the nerves (Fig. 45 D). Eggs were ripe in November and December.

In view of the extreme variation in the accessory armature and proboscis nerves the
distinctions that have been drawn on these characters cannot hold. I therefore feel
justified in bringing together the species described by Burger and Joubin under
A. spinosus, Burger.

NEMERTEANS

269

In addition to the collection from King Edward Cove this species was taken at the
following stations :

St. 39. 25. iii. 26. OTL, 179-235 m.

St. 45. 6. iv. 26. OTL, 238-270 m.

St. 123. 15. xii. 26. OTL, 230-250 m.

St. 140. 23. xii. 26. OTL, 122-136 m.

St. WS 56. 14. i. 27. NH, 2 m.

St. WS 62. 19. i. 27. BTS, 26-83 m.

St. WS 65. 22. i. 27. Sh. coll.

St. WS 73. 6. iii. 27. OTC, 121-130 m.

St. MS 68. 2. iii. 26. NRL, 220-247 m.

St. MS 71. 9. iii. 26. BTS, no-6om.

Fig. 45. Amphiporus spinosas, Burger. A, section across the tip of the head; B, armature; C, accessory
stylet; D, transverse section of the body, bm, basement membrane; cm, circular muscle layer; e, eye;
ep, epithelium; hg, head gland; ///;, longitudinal muscle layer; In, lateral nerve; ov, gonad; />, proboscis;
rd, rhynchodaeum.

270

DISCOVERY REPORTS

Genus Tetrastemma, Ehrenberg

Tetrastemma esbenseni, n.sp. (Plate XVI, figs. 4, 23 ; Figs. 46, 47).

This species is not a common one. It occurred as follows:

7. iii. 26. King Edward Cove, South Georgia, 1 (N 18).

9. iii. 26. Maiviken, under stones, 7 (N 18).

St. 39. 25. iii. 26. OTL, 179-235 m. 2 (N 44) ; N 4-T, 20 (N 23, N 93, N 1 12).

St. 123. 15. xii. 26. OTL, 230-250 m. 1 (N 18).

The lengths and corresponding breadths were: 30-0, 07 mm.; 90-0, o-8mm.;
22-0, 07 mm.; 53-0, 1-3 mm.

The body is thin, soft and round in section, tapering at the head and tail. The head is
pointed and there is no neck between it and the body. There is a chevron groove be-
hind the head, complete ventrally. The eyespots are arranged in four groups at the
corners of a rectangle, the number in each group varying from one to six, the anterior
groups usually containing more than the posterior. The colour is light yellow or yellowish
red. The gut can usually be seen through the body wall. In the specimens which were
separated as N 44 the vascular system was more engorged than usual and no eyes were
seen (Plate XVI, fig. 23), but they were identified with the other forms from the
sections.

Form and colour of spirit specimens. The body is contracted and cylindrical (lengths
18 and 15 mm. with breadths of 1-4 and i-omm.). All colour is bleached and no
eyes are visible. On clearing in cedar oil one
specimen possessed the usual two pairs of eyes
imperfectly formed, another had four large eyes
and several small perfect brown cup-shaped eyes
close to them. Two more variations are shown in
Fig. 46. The proboscis is always small and thin.
The armature consists of a main stylet 0-065 mm-
long on a base 0-083 mm- and a curiously irregular
accessory armature. One specimen had two ac-
cessory reservoirs, one with five stylets and the
other, which appeared to be double, with five and
six stylets. Another specimen had three stylets in each of two reservoirs and yet another
had five reservoirs crowded together on one side of the main stylet each with four or
five. I could not determine the nerves in the proboscis.

Anatomy. The head glands open at a median ventral pit. They are very large, filling
the head and persisting dorsally to the posterior end of the brain, ventrally farther back
still.

The basement membrane is thick, nearly as thick as the epithelium. It does not stain.
The circular muscle layer is about half as thick as the basement membrane and the longi-
tudinal layer about as thick as the epithelium but thicker ventrally.

The vascular system is well developed on the usual plan, but the vessels are much

Fig. 46. Tetrastemma esbenseni, n.sp. Heads
of two specimens cleared in cedar oil showing
variation in eyespots.

NEMERTEANS

271

dilated or very capacious and full of nucleated corpuscles. This appears to be a constant
character of the species, though it does not always appear in life in the marked form
shown in Plate XVI, fig. 23.

The cerebral organs are very small and far in front
of the brain. The relative positions of the brain, cerebral f-jpr
organs, eyes and head glands can be seen in the graphic
reconstruction (Fig. 47). The anterior caecum has no
forward branches and lies far behind the brain. The
brain is large, the dorsal ganglia being about equal in
size to the ventral and continuing into the lateral nerves
down the body.

I have not been able to identify this form with any
known species although it approaches T. belgicae,
Burger, 1904a.

Tetrastemma georgianum, Burger, 1893 (Plate XVI,
figs. 2, 13; Fig. 48).

This species was first noted from King Edward
Cove. Two specimens were taken in September and
four in October 1925, from kelp roots. A single speci-
men occurred at St. WS 65 in Undine Harbour. The
smallest worm was 3-5 mm. long, but the usual length
was from 10 to 15 mm., breadth 0-7 mm. The original
sketch and description were made from six specimens
(N 7). In 1929 a single specimen was collected in the Fi8- 47- Tetrastemma esbenseni, n.sp.
, 1 • it- t^j j r> i_ • 1 -j sn j Graphic reconstruction of the head and

dredge in King bdward Cove which was identified . . c , , , , ,

0 ° anterior end 01 the body, co, cerebral

later with a larger form described under N 15, no organ;*, anterior eye -hg, head glands,
differences being found in sections. The size attained

may be 25-0 mm. Other captures were made — three specimens at St. 123, one at St. 140
and one at St. 51 (identified from sections, N 126).

Form and colour in life. The head is somewhat round in outline from above, but the
snout has a median vertical cleft. The body is broadest at about half its length and is
tubular and soft. The colour is brown, brownish red or mauve on the back, much paler
beneath. With a low magnification (x 10) the pigment appears granular. On the head
the pigment band narrows and becomes defined so that the head appears to be bordered
with white. On each side a white tag or strip passes up, while a little farther back the
body colour is separated from that of the head by a fine white line which forms a wide
V with the apex pointing back. I have never seen the anterior tags joined as a band.
Four eyespots are present in two pairs. The first pair is situated at the anterior edge of
the head pigment ; the second lies within the white tags. A pair of vertical furrows which
may be tinged with red can be seen in the tags.

Form and colour of preserved specimens. Length 9 and 13 mm., breadth o-8 and

272 DISCOVERY REPORTS

1*3 mm. respectively. The head is blunter than the tail and is marked off from the body
by lateral opposite furrows and a complete circular chevron groove whose apex points
back ventrally. No eyes are visible. The colour is usually bleached but may be yellowish.

Anatomy. The epithelium is thick. In one series of sections it was at the head almost
twice as thick as the basement membrane, circular and longitudinal muscles together.
Farther back on the body its relative thickness was less.
I have some doubt of the value of comparative measure-
ments of the epithelium and muscle layers, for another
series shows a thick basement membrane not much
stained, and a clear layer of circular muscles, the two
together being about equal in thickness to the epi-
thelium; it seems probable that bad fixation may be
responsible for these differences.

The head gland is compact. It opens by a median
pore at the tip of the head. There are three strands, one
median dorsal and two lying laterally along the rhyn-
chocoel. The former nearly reaches the dorsal com-
missure. The oesophagus opens into the rhynchodaeum
near the snout and enlarges to the stomach just in
advance of the brain. The anterior caecum sends forward
two diverticula above the lateral nerves which overlap
the posterior ends of the dorsal ganglia (Fig. 48),
although the relative position of the branches and ganglia
is altered by the state of contraction. The single ex-
cretory ducts open before the junction of the anterior bocjy
caecum with the gut.

The proboscis is stout. The armature consists of a main stylet on a pear-shaped base
larger than itself and two accessory reservoirs each with one or two stylets. There are
ten nerves.

The ventral ganglia are a little larger than the dorsal. The cerebral canals appear as
furrows on the ventral surface just in advance of the anterior pair of eyespots. They
deepen and sink in. The organs join the dorsal ganglia from a ventro-lateral direction at
the level of the ventral commissure (Fig. 48).

Fig. 48. Tetrastemma georgiatmm,
Burger. Graphic reconstruction of
the head and anterior end of the

Tetrastemma gulliveri, Burger, 1893 (Plate XVI, figs. 11, 17; Fig. 49).

This species was originally described and sketched at different times under the desig-
nations N 10 and N 14, but intermediate forms showed later that the differences lay in
the size of the body and the distinctness of its markings. Subsequent investigation of
the armature proved the identity. Thirty-four animals were examined in life, and from
sections the following single specimen was sorted from the collection: St. 144, 5. i. 27,
NCS-T, 155 178 m. (N 123). The lengths varied from 4 to 17 mm., the breadths
from o-6 to i-6 mm.

NEMERTEANS

273

Form and colour in life. The body is somewhat fusiform and in movement resembles
Oerstedia in its semi-rigidity. The tail is more pointed than the head. Eyespots are
visible in the smaller specimens. The colour is yellowish brown with a narrow median
ventral pale "sole". The larger worms are darker and show less of the mottling — light
irregular patches and dark spots — that is apparent in the smaller. There is a light ring
circling the body at the back of the head, usually behind the second pair of eyespots but
sometimes coincident with them. This pale annulation is characteristic of the species in
life.

Form and colour of spirit specimens. Lengths 10 and 16 mm., breadths 07 and
1-4 mm. The body is nearly cylindrical, with the head thicker and blunter than the tail.
There is sometimes a complete annular groove at the back of
the head. The eyespots are not visible. The colour is usually
completely bleached, but occasionally sufficient brown re-
mains to display the white neck band. When cleared in anilin
oil the pigment shows as black specks. The eyes, which are
also rendered visible, are two pairs of brown cups, the con-
cavity of the anterior pair being antero-lateral, that of the
posterior pair postero-lateral. The eversible part of the pro-
boscis is less than half the length of the body.

Anatomy. The epithelium is very thick. The relative thick-
ness of the body layers can be seen in Fig. 49. The head
glands fill the head completely and extend back on all sides
of the brain. Posteriorly they end dorsally with the dorsal
ganglia, but they continue beneath the stomach and anterior
caecum until the two forward diverticula of the latter join it.
The diverticula almost reach the brain. The proboscis is well
developed and possesses ten nerves. The armature consists of
a slim main stylet on a rounded base and two reservoirs each
with four, five or six stylets.

The dorsal ganglia are small for the size of the brain and
the commissure is thin. The lateral nerves pass outwards and
up from the ventral ganglia and they each carry down the
body a strand of fibres from the dorsal ganglion. The cerebral
organs are very small and thin. They open laterally not far
from the head-gland pore and consist of a tube sheathed with large gland cells
extending a short distance behind the first pair of eyes. The excretory ducts open far
after the brain.

The specimens taken in King Edward Cove in 1929 and 1930 came from a small red
alga. Three of these were found in mucous tubes attached to the weed. In 1929 four
brown worms within mucous tubes had been found attached to the rootlets of kelp and
described as N 50. On sectioning they proved to be Tetrastemma gulliveri. The epi-
thelium was torn off and the internal organs degenerated, especially the brain. The

8-2

Fig. 49. Tetrastemma gulli-
veri, Burger. Diagram from
a graphic reconstruction of
the head showing the eyes,
brain, cerebral organs, and
head glands.

274

DISCOVERY REPORTS

cerebral organs could be seen and the eyes and enormously developed head glands,
though the latter did not stain as they had in the free specimens. The lateral nerves had
degenerated considerably and the body consisted of a thin investment of basement mem-
brane and muscles and a large amount of internal structureless substance in which the
gut walls could still be recognized. No genital products could be seen. A specimen of
T. gitlliveri taken from red algae off the Point in King Edward Cove in March 1930
was full of eggs, so that there is evidence of a form of hibernation in this species.

Tetrastemma hansi, Burger, 1893 (Plate XVI, figs. 6, 15; Fig. 50).

This species was first collected from the kelp roots in King Edward Cove in October
1925. Thirteen specimens were taken, sketched and noted (N 12). In March 1926 seven
specimens differing in colour and body form were collected and described under N 17.
These were later proved from their anatomy to be identical with the earlier forms. In
April 1927 many specimens were collected of intermediate size and covering the range
of colour. In November 1927 a specimen was taken that contained nearly ripe gonads
visible through the ventral surface of the body.

Several small immature specimens were collected at St. 53 (Falkland Islands) and
identified from sections (N 106).

Form and colour in life. The length and breadth rarely exceed io-o and i-o mm.,
though the breadth may be less for the same length. The body is round ; the head just
visibly marked from the body by its greater width. There are two lateral grooves on each
side of the head and two pairs of eyespots. The head
tapers to an acute snout. The colour varies from very
light brown through shades of yellow-brown to light
red and pink. There are sometimes traces of a pale
median line down the back, not visible near the head
and fading before the tail. Ventrally the animals are
pale.

Form and colour of preserved specimens. The body
is round in section, about 8 mm. long and 0-5 mm.
broad. The head is blunter than the tail. The colour
is completely bleached and the eyes are only visible
faintly in the smallest specimens. On clearing the
small proboscis can be seen not reaching to half the
length of the body. The eyespots are very small.

Anatomy. The epithelium is very thick; the base-
ment membrane and circular muscles very thin and
equal in thickness. There are eosinophile gland cells
beneath the longitudinal muscle layer ventral to the
brain. The head glands are extremely developed though
they do not stain with haematoxylin. They fill the head Graphic reconstruction of the head,
and stretch back beyond the brain and the two forward hg, head glands.

NEMERTEANS 275

branches of the anterior caecum (Fig. 50). The nephridial tubules open to the exterior
laterally at the level of the branches of the anterior caecum. The proboscis has ten
nerves and is armed with the usual main stylet and two reservoirs each with two, three
or four stylets. The brain is fairly large. The dorsal ganglia are smaller than the
ventral and there is no strand of fibres from the former to the lateral nerves. The
cerebral organs open ventro-laterally just in advance of the brain and pass inwards,
increasing in size and becoming applied to the brain between the dorsal and ventral
ganglia.

One of my sectioned specimens was an immature male, the other an immature

female.

The identification of this form with T. hand, Burger, is based on the head glands and
the cerebral organs. Both of my series of sections show two short lateral diverticula
from the caecum.

Tetrastemma longistriatum, n.sp. (Plate XVI, fig. 7; Fig. 51).

This species occurred at the following stations. It was sketched and described under

N48:

St. 42. 1. iv. 26. OTL, 120-204 m. 2 (N 57).

St. 141. 29. xii. 26. BTS, 17-27 m. 2 (N 48).

St. 163. 17. ii. 27. BTS, 18-27 m- J (N 91)-

11. iv. 27. Kelp roots, King Edward Cove, 2 (N 48).

The lengths and breadths of some of the specimens were : io-o, 1 -o mm. ; 5-0, 07 mm. ;
5-0, o-6 mm.; 2-0, 0-2 mm.

Form and colour in life. The body is round in section, tapering to the tail which is
much more acute than the head. There is a pair of deeply embedded eyespots just in
front of the head markings and another pair at the anterior end of the body pigment.
The ground colour of the body is pale yellow. On the back there are two longitudinal
brown bands, fading laterally, which leave a broad sharply defined streak between them.
At the head the streaks end, but there is a pair of very deep reddish brown patches on
the head in the form of two elongated right-angled triangles placed transversely, the
right angles being the inner anterior angles.

Form and colour of preserved specimens. The length is about 5-0 mm., the breadth
0-35 mm. The body is cylindrical, the head blunter than the tail. The eyes are not
visible, but on clearing, the anterior pair are larger than the posterior. The colour and
markings are faintly seen.

Anatomy. The most noticeable feature of the species is the extreme development of
the epithelium, which is at least twice the thickness of the remainder of the body wall.
It bears within its cells large masses which stain with haematoxylin. The head glands
are compact and solid masses opening just beneath the tip of the head. One strand
spreads over the rhynchocoel, just reaching the brain. Another forms a thick U-shaped
investment to the oesophagus but diminishes and disappears before the dorsal strand.
The lateral branches of the anterior caecum overlap the dorsal ganglia (Fig. 51).

276

DISCOVERY REPORTS

The proboscis is stout. I could not make out more than nine nerves in one of my
series of sections, but there were ten in another. The arma-
ture consists of a main stylet with two reservoirs each with
three stylets.

The brain is large, the dorsal ganglia being somewhat
smaller than the ventral. There is no dorsal strand in the
lateral nerves. The cerebral organs are small. They open
laterally and only just reach the brain (Fig. 51).

Two small specimens from St. 42 were examined among
the preserved material. One was sectioned and proved to
be a female with large eggs. In all characters this specimen
agreed with the above description.

Tetrastemma maivikenensis, n.sp. (Plate XVI, fig. 10).

Only the external characters of a single specimen (N 19)
of this worm are known, and it bears a resemblance (except
in size) to T. vermiculus, Quatrefages. The length in life
was 40 mm., the breadth 075 mm. The body was soft and
round, tapering to the tail. There were two pairs of eyes
and the brain could be seen as a bilobed pinkish mass
through the body wall. The colour was pale green except
at the head which was yellowish. The distinctive marking Fig. 51. Tetrastemma hngistria-
consisted of a streak of brown pigment between the eyes of *». n-sP- GraPhic ^construe-

tion of the head and section
each Side. across the body to show the

It takes its name from Maiviken, South Georgia, where extreme development of the
the specimen was found (St. MS 70). epithelium.

Tetrastemma stanleyi, n.sp. (Plate XVI, fig. 12).

Three specimens of this form were collected in Port Stanley harbour under stones at
low tide on April 29, 1926 (N 21). The lengths ranged about 40-0 mm., the breadths

1-2 mm.

The body is round in section, the head slightly constricted from it, bluntly pointed
and somewhat flattened. The tail tapers but ends acutely. There are from one to four
eyespots in each group of the two pairs. The general colour effect is reddish brown near
the head, paler towards the tail. One animal was olive green. The tip of the snout is
conspicuously dark brown. This pigment is continued as a median line to the level of the
posterior eyes between which it broadens and terminates . In spirit the colour is bleached .
One specimen was of a dull bluish grey colour, evidently caused by the contents of the
gut. No eyes could be seen even when cleared.

Anatomy. The head glands form a compact layer dorsal to the rhynchodaeum and
disappear before the brain.

The epithelium is not very thick but it is equal to the three other body layers. Of

NEMERTEANS 277

these the basement membrane and circular muscles are equal in thickness and are to-
gether about one-quarter of the thickness of the epithelium. The oesophagus opens into
the stomach after the brain. There is no anterior caecum and the stomach is very long.

The proboscis has fourteen nerves. One specimen had an accessory armature of two
reservoirs each with two stylets; another had two reservoirs each with four.

The brain is small and the dorsal ganglia do not send fibres into the lateral nerves.
The cerebral organs are small and just reach the underside of the brain. The canals open
ventro-laterally.

The specimen sectioned was an immature male.

Tetrastemma validum, Burger, 1893 (Plate XVI, fig. 1 ; Figs. 52, 53).

In colour and especially in form this species is the most easily recognized of the
southern Tetrastemma. Single specimens were taken from kelp roots in King Edward
Cove in 1927 (N 49) and 1929. In March 1930 thirty-eight specimens were caught by
the dredge with a mass of red algae. Other captures were at St. 175 (N 130) and St. 179
(eleven specimens — N 99). The lengths ranged from 8-5 and io-omm. to 35-0 and
40-0 mm. with corresponding breadths of o-8, 1-4, 2-5 and 2-3 mm.

The body is stoutly built, flattened from above and fusiform. The head is pointed,
distinct from the body by a slight "neck". Cephalic grooves join under the head in a
wide V with its apex forward and from this junction another groove runs forward in the
mid-ventral line. There is a groove behind these at the shallow depression of the " neck ",
but this is incomplete dorsally. Sometimes the posterior part of the head is round and
broad, the anterior end being drawn out into a kind of beak. It has a distinctly shark-
like appearance, especially pronounced from the side. The colour on the back is dark
reddish brown, yellow-brown or purple-brown. The underside is white. On each side
of the head a very definite almost rectangular white tag shows up strongly against the
dark pigment. Just before this tag there is an encroachment of white on the pigment of
the head, and in this pigmentless patch lies the first pair of eyes. The second pair is
situated behind on the anterior edges of the white patches.

Form and colour of preserved specimens. The body retains its shape to a great extent.
The anterior end is blunter than the tail. The back is convex, the belly flat or concave.
Very little colour remains in the specimens, but usually there is sufficient duskiness to
make the white patches faintly visible.

Anatomy (Fig. 52). The head glands open near the tip of the head. There are at first
three strands towards the dorsal side but they soon completely fill the head. At the
anterior end of the brain the dorsal glands have coalesced, while ventrally there are still
small packets. The glands do not stain with haematoxylin. After the brain they persist
ventral to the branches of the anterior caecum. The epithelium is thicker than the sum
thickness of the other layers.

The rhynchocoel extends to the end of the body and the nerves join just behind its
attachment over the gut. The proboscis possesses ten nerves (in one specimen twelve)
and has two accessory stylet reservoirs each with two stylets similar to the main stylet

278 DISCOVERY REPORTS

which is 0-074 mm- l°ng mounted on a pear-shaped base o-ioi mm. long. Branches of
the anterior caecum do not reach the brain.

The brain is small. There is a dorsal strand in the lateral nerve. The cerebral organs
are fairly large. They open on the under-surface of the head but do not reach the brain.

On two fronds of the red alga dredged with the large haul of this species a semi-

Fig. 52. Tetrastemma validum, Burger. Graphic
reconstruction of the head to show the develop-
ment of the head glands and the position of the
brain, cerebral organs, eyes and branches of the
anterior caecum.

Fig. 53. Tetrastemma validum, Burger. Mem-
branous pouch, open at both ends, attached to
a frond of red alga.

translucent sheath was found attached. The length was 28 mm., breadth 5-0 mm. The
pouch was open at both ends and from one the animal crawled while under observation

(Fig. 53)-

In several characters this species is similar to Amphiporus michaelseni, Burger, as de-
scribed by Joubin (1908), but the four eyes and the internal structure show that its
affinities are with Tetrastemma.

Tetrastemma weddelli, n.sp. (Figs. 54, 55).

One specimen (N 70) was collected at St. 160 between South Georgia and the South
Orkney Islands. No note was made of the colour or form, but its anatomy is so dis-
tinctive that it should be easily recognized. The body was cylindrical, 11 mm. long
and 075 mm. broad. The proboscis was protruded to a length of 5-5 mm. and just be-
low it was the frilled opening of the mouth. The colour was bleached (Fig. 54 A).

NEMERTEANS

279

Anatomy. A feature of the sections throughout the body is the thickness of the base-
ment membrane (Fig. 55) which is rather thicker than the epithelium. Both circular and
longitudinal muscles are strongly developed. I could not be certain of the presence of
head glands owing to the peculiar staining of the head, but there appears to be a thick
strand staining deeply with haematoxylin dorsal to the rhynchocoel. The proboscis has
fourteen nerves and is armed with a main stylet on an elongated pear-shaped base
0-098 mm. long and two reservoirs each with two stylets. The frilled mouth is a con-
tinuation of the stomach into which the narrow folded tube expands after the brain.

Fig. 54. Tetrastemma weddelli, n.sp. A, outline
sketch of the preserved specimen, x 3 approx.;
B, graphic reconstruction of the anterior end.

Fig. 55. Tetrastemma weddelli, n.sp. Transverse
section across the body, bm, basement membrane;
cm, circular muscles; ep, epithelium; hn, longi-
tudinal muscles;/), proboscis.

There is a single forward diverticulum of the gut which is lateral to the stomach and
does not reach the brain. The excretory tubules open dorso-laterally above the lateral

nerves.

There are two pairs of eyes. In this specimen there is a double eye on one side. The
brain is large and the dorsal ganglia are far larger than the ventral. The ganglia of each
side are close together, especially the ventral ganglia at the anterior end (Fig. 54B), so
that the ventral commissure is very short. There is a dorsal strand in the lateral nerves.
The cerebral organs are large and lie along the dorsal ganglia. They open laterally from
their posterior ends, though whether this is connected with the pushing forward of the
alimentary canal or the normal position in life I am unable to say.

The specimen was a male with ripe sperm.

28o

DISCOVERY REPORTS

-4

-4

i

PART III. THE PELAGIC NEMERTEANS

Forty-five pelagic Nemerteans were included in the collection. Of these, thirty-five
proved to be Pelagonemertes rollestoni, one of which, taken from a haul of seventeen at
St. 107, was sketched in life. Most of the specimens were noted when captured but only
one other colour sketch was made. This is reproduced in Plate XVI, fig. 5, and is a new
species, Bathynemertes hardyi.

I have followed Brinkmann (1917) in his classification of the suborder Polystihfera,

tribe Pelagica.

Genus Bathynemertes Brinkmann

Bathynemertes hardyi, n.sp. (Plate XVI, fig. 5; Figs. 56, 57).

This interesting form, curiously substantial for a pelagic worm, was sketched and

noted in life by Mr (now Prof.) A. C. Hardy. It was captured at St. 86 (330 25' 00" S,

6° 31' 00" E) in the \\ m. net at 1000 (-0) m.

The body was scarlet with black irregular

markings; the proboscis lighter than the body.

The body was almost round in section, tapering

a little to the head and considerably at the tail

which was not expanded into a fin.

Form and colour of preserved specimen. The

body is no mm. long, with a maximum breadth

and thickness of 25 and 13 mm. It is some-
what flattened and is faintly marked by narrow

annular wrinkles. The tail is more definitely

flattened. It has, however, neither fins nor lateral
lappets. The body has a tendency to curl up at
the ends due perhaps to the greater shrinkage of
the proboscis and rhynchocoel wall towards the
dorsal side of the body cavity (Fig. 56).

The mouth is not coincident with the rhyn-
chodaeum and has a frilled edge. The proboscis
is very strong. It is 8 mm. in diameter and
covered with papillae. The rhynchocoel extends
into the tail.

The body colour is light brown-orange with irregular patches of dark brown pigment
especially on the under side. A lighter streak can be seen down the mid-dorsal and mid-
ventral lines, and on each side there is a pale slightly raised lateral line breaking the pig-
ment patches.

Anatomy (Fig. 57 A). The epithelium is almost entirely absent; what there is being
ventral (Fig. 57 B). The basement layer is very thick and the circular and longitudinal
muscles are fairly developed, little thinner laterally than dorsally or ventrally. Strong
strands of muscle pass from the dorsal to the ventral body wall through the body.

J

Fig. 56. Bathynemertes hardyi, n.sp. A, ven-
tral, B, dorsal aspect of the preserved speci-
men, x f approx.

NEMERTEANS

281

The alimentary canal opens at the frilled mouth ; there is no oesophagus. The branches
of the gut are small but numerous anteriorly, and they reach forward to the brain.

The proboscis has twenty-four nerves and is armed by the usual curved rod , although no
stylets could be seen . The rhynchocoel wall consists of interlaced longitudinal and circular

Fig. 57. Bathynemertes hardyi, n.sp. A, transverse section towards the anterior end of the body, bm, base-
ment membrane; cm, circular muscle layer; dv, dorsal vessel; ep, epithelium; Im, longitudinal muscle layer;
In, lateral nerve; ov, gonad; re, rhynchocoel. B, section of the body wall, from A, magnified.

fibres. The brain is not large. Its structure, with the strands of nerve fibres passing
from the dorsal ganglia into the lateral nerves, corresponds with the description given by
Brinkmann for B. hubrechti. I could find no dorsal nerve between basement membrane
and circular muscles such as Burger, 1907 (1912), describes in DrepanopJiorus pelagicus.
The specimen was a female with small eggs. The gonads were only seen towards the
ventral side close to the lateral nerves.

Bathynemertes hubrechti, Brinkmann, 1917 (Figs. 58, 59).

Four specimens of fair size were collected at the stations given below

St. 85. 23. vi. 26. 330 07' 40" S, 40 30' 20" E. N 450 H, 2000 (-0) m. "Scarlet" (N 168).

St. 89. 29. vi. 26. 340 05' 15" S, 1 6° 00' 45" E. TYF, 1000 (-0) m. "Dull orange" (N 160).

St. 100c. 4.x. 26. 330 20' to 33°46'S, 150 18' to i5°o8'E. TYF, 2500 (-0). "Bright brick
red" (N 167).

St. 101. 15.X.26. 33°5o' to 340 13'S, i6°04' to i5°49'E. N 450, 1310-1410 m. "Orange
red" (N 164).

The sizes in spirit were :

Serial

Length

Greatest breadth

Depth

No.

mm.

mm.

mm.

N 168

20-0

5-o

3'°

N 160

7-0

*'3

1-2

N 167

30-0

6-o

4-0

N 164

21-0

5'0

4-0

9-2

282

DISCOVERY REPORTS

The form is flattened and bluntly pointed at both ends. Some degree of translucency
may be observed but there is no colour (Fig. 58).

Fig. 58. Baihynemertes hubrechli, Brinkmann. Outline sketches of the preserved specimens.
A, N 164 anterior end above; B, N 160 from the side; C, N 168 from the side.

Fig. 59. Baihynemertes hubrechti, Brinkmann. A, transverse section through the body
near the head (N 164); B, transverse section farther back (N 160).

Anatomy (Fig. 59). In all specimens the epithelium is missing. The basement mem-
brane is high and serrated but not much stained. The musculature is reduced laterally
and is thin dorsally and ventrally although it is thicker in the small specimen (N 160)
than in the larger. There is no oesophagus. The gut branches fill the body cavity and the

NEMERTEANS

283

lateral nerves are included among them and not pressed against the body wall. The
rhynchocoel extends to the posterior end of the body. Its wall consists of interlaced
fibres. The proboscis is very stout and the number of nerve strands varies between
22 (N 160), 24 (N 168), 25 (N 164) and 26 (N 167). The brain is of fair size and con-
forms to the description given by Brinkmann (1917).

N 160 and 164 were female. Small eggs could be seen close beside the lateral nerves
(Fig. 59B). In the other specimens I could find no trace of gonads.

Genus Crassonemertes, Brinkmann

Crassonemertes robusta, Brinkmann, 1917 (Fig. 60).

No colour note was made of this specimen (N 170), taken with Nectonemertes kempt
in a 2-metre net at lat. 6° 55' N, 150 54' W, 0-800 m.

Fig. 60. Crassonemertes robusta, Brinkmann. A, dorsal surface, x 3 ; B, transverse section at the anterior end.

In spirit the body is white (it had been fixed in corrosive sublimate) and completely
opaque. It is broad and flattened and has a distinct tail. A sharp end protruded from the
tail; this was afterwards found to be the proboscis (Fig. 60A). The length is 15-0 mm.,
greatest breadth 7-0 mm., thickness 5-0 mm.

No epithelium is present. The basement layer is deeply stained and appears to have
suffered from shrinkage or drying. The muscle layers are reduced, except dorsally
(Fig. 60A). There is no oesophagus. The muscular stomach opens into the long pylorus
at the brain. The gut branches are wide and very numerous. They fill the body cavity
so that no sign of separate diverticula can be seen in the posterior part of the body when
cleared with anilin oil.

The proboscis is very stout and long. It has twenty-one or twenty-two nerves. The
wall of the rhynchocoel is composed of interlacing fibres. The brain is small. There does
not appear to be a dorsal strand in the lateral nerves. No gonads could be made out.

284

DISCOVERY REPORTS

Genus Nectonemertes, Verrill (part)

Nectonemertes mirabilis, Verrill, 1892 (Fig. 61).

A note was made that the colour in life of this specimen was "pinky red". It was

taken at St. 87.

The length of the preserved specimen (N 80) was
14 mm., breadth 2-1 mm., thickness 1 mm. The
colour was yellow and semi-transparent. The anatomy
has been thoroughly described by previous workers.
This specimen was a female with large eggs. The
proboscis was missing. An outline sketch is given in
Fig. 61 to show the form of the body and the position
of the gonads.

Fig. 61. Nectonemertes mirabilis, Verrill.
Outline sketch showing the form of
the body and the position of the
gonads, x 7.

Fig. 62. Nectonemertes kempi, n.sp. A, transverse section
through the brain. B, transverse section through the body.
dv, dorsal vessel; In, lateral nerve; re, rhynchocoel.

Nectonemertes kempi, n.sp. (Fig. 62).

A small worm was collected with Crassonemertes robust a from 0-800 m. at
6°55'N, i5°54'W. It was fixed in corrosive sublimate. No note was made of
the colour and form in life. The preserved specimen (N 169) was 7 mm. long and
resembled the post-larval stage of a fish. It was white, somewhat flattened and had

a tail fin .

The sections show a very large brain (Fig. 62 A) completely filling the head. It seems
evident, however, that the specimen has suffered shrinkage from the fixative and this has
probably affected the basement layer and muscles. The most striking feature apart from
the size of the brain is the thickness and size of the rhynchocoel and its wall. The latter

NEMERTEANS 285

is composed of an inner longitudinal layer definitely divided into bundles and an outer
thick layer composed of circular fibres with some longitudinal fibres among them.
Before the brain the muscle layers are reversed, i.e. the inner layer is circular and the
outer longitudinal. The proboscis is missing.

There is no oesophagus. The branches of the gut reach forward beside the brain

posteriorly.

The lateral nerves are pressed against the ventral body wall (Fig. 62 B) and contain an

evident dorsal strand.

The specimen was a female with eggs developing singly close beside the larval

nerves.

That the animal belongs to the genus Nectonemertes is certain from its form, the
muscles of the rhynchocoel wall and the position of the lateral nerve cords. I have
separated it from Brinkmann's species— primitiva and minima— on the ground of its
form and the size of the rhynchocoel and brain.

Genus Pelagonemertes

Pelagonemertes rollestoni, Moseley, 1875 (Fig. 63).
The following captures were made of this species :

St.
No.

Date

Gear

Depth

No. of
specimens

Sex

Serial
No.

2. ii. 25

N 200

0-800

9

N 139

71

30. v. 26

N70V

1000-750

—

N145

TYF

2000 (-0)

9

N149

72

1. vi. 26

N450

2000 (-0)

?

N 141

76

5. vi. 26

N450H

1 500-0

6"

N147

78

12. vi. 26

TYF

1000 (-0)

99?

N143

79

13. vi. 26

N450H

1 000-0

9

N 140

8S

23. vi. 26

N450H

2000 (-0)

6*

N154

86

24. vi. 26

N450H

1000 (-0)

?

N 146

89

28. vi. 26

TYF

1000 (-0)

2

$

N158

$

N 163

107

4. xi. 26

N 450

850-950

17

99,3o"

9

N 138

N42

256

23. vi. 27

TYF

850-1100 (-0)

1

9

N151

395

13- v- 3°

N450H

1 500- 1 600

3

3S

4°5

4. vi. 30

TYFB

2200-0

1

9

A specimen from the large haul at St. 107 was sketched directly after capture. An out-
line drawing of this sketch is given in Fig. 63 to which the imagination can readily add
the transparency of the body, the opacity of the brain, lateral nerves, gonads, proboscis
and its sheath, and the startlingly vivid colouring of yellow on red of the branches of the
gut. This specimen measured in life 45 mm., and was 23 mm. broad. Another was
21 and 12 mm.

After fixation the gut branches appear considerably thicker than they are in life. The

2 86

DISCOVERY REPORTS

variation in number of branches and in the size of the animal can be gauged from the
following :

Length

Breadth

Thickness

Gut

mm.

mm.

mm.

branches

33'0

22-0

7-0

15: is

21-0

15-0

—

16 : 14

3o-o

21'0

6-o

17: 17

20-0

13-0

—

i5: ll

3o-o

20-0

—

17 : 16

35'°

1 8-o

—

H: J5

22-0

18-0

—

13 : 13

i6-o

8-o

—

H : x5

22-0

1 6-o

6-o

13: H

13-0

12-0

16: 16

The anatomy of this species has been worked out very thoroughly by Burger, 1907
(1912), and Brinkmann 1917. Fig. 63 B shows the position of the testes and gives an

Fig. 63. Pelagonemertes rollestoni, Moseley. Outline sketches of entire animals. A, female, from a
colour sketch of a living animal; B, male, from a spirit specimen, ov, ovaries; t, testes, x 2.

idea, when compared with Fig. 63 A, of the degree of enlargement of the gut branches
after fixation. The number of nerves in the proboscis was twenty-two in two of my
specimens, but they are difficult to count because of the irregular thickening of the nerve

NEMERTEANS

287

net. Brinkmann gives sixteen nerves. I have examined my series of sections for the
rudimentary eyes and in two males several small organs corresponding to the description
and figures (Brinkmann, 1917) were found on each side of the lappet of the body
ventrally very close beside the rhynchodaeum, but I could not trace them in the female.

Genus Probalaenanemertes, Brinkmann

Probalaenanemertes irenae, n.sp. (Figs. 64, 65).

Two specimens of this form were collected at St. 89. The colour notes were " scarlet
with orange spots down sides" (N 161); "pale orange with deep rose pink spots down
sides" (N 162). In spirit the sizes were: length 9-0 and 17-0 mm.; greatest breadth
3-0, 3-5 mm.; thickness 17, 2-0 mm. Both were somewhat flattened, bluntly pointed
at the thicker end of the body and with a distinctly flat spade-shaped tail (Fig. 64). In

A B

Fig. 64. Probalaenanemertes irenae, n.sp. Outline sketches of the preserved specimens.
A, N 161, dorsal; B, lateral; C, N 162, lateral, x 4 approx.

one five brownish gonads could be seen on each side and the lateral nerves were faintly
visible outside them ; in the other there were fourteen gonads on one side, twelve on the
other. The colour was pale brown and the bodies were translucent.

Anatomy (Fig. 65). The epithelium is present in patches at the anterior end and is a
deep layer into which the serrated basement layer penetrates deeply (Fig. 65 A). The
circular and longitudinal muscles are very much reduced, especially the former. At the
tail there are dorsal and ventral median strengthening strands of longitudinal muscle
(Fig. 65 C). There is no oesophagus. The folded tube shown in Fig. 65 A passes back
unbranched through the brain region and beyond. Meanwhile two gut branches have
appeared above and lateral to the brain and farther back more appear which invade the

288 DISCOVERY REPORTS

body cavity ventrally and finally open into a short median anterior caecum ending
blindly forwards.

The rhynchocoel is spacious but the wall is thin. It is composed of inner longitudinal
and outer circular muscles and extends nearly to the end of the body. The proboscis is
thin and the muscles are poorly developed. It possesses seventeen nerves.

Fig. 65. Probalaenanemertes irenae, n.sp. Transverse sections. A, the region of the brain;
B, mid-body; C, tail. In, lateral nerve; ov, egg.

The brain is fairly large, dorsal ganglia rather larger than ventral. The lateral nerves,
containing a distinct mass of fibres from the dorsal ganglia, are pressed ventrally until
they rest against the muscles of the body wall and close to them towards the middle line
the gonads begin development. The eggs are immense (Fig. 65 B).

NOTES ON THE DISTRIBUTION OF THE
SOUTHERN NEMERTEANS

The extension of the Mediterranean fauna to the south western extremity of Africa
has already been indicated. Certain species, of course, can be reckoned as cosmopolitan
and their occurrence leads to no comment; for instance, Linens ruber, Tetrastemma
candidum and Amphiporus pidcher, which have been found widely distributed in the
temperate seas in the northern hemisphere: but the inclusion of Tabidamis nothus,
Nemertopsis tennis, Linens bilineotus and Cerebratidus fuscus — all easily identified by
striking colour, markings or body form apart from their anatomy — points to a com-
mon origin with the fauna of the Mediterranean even if there is discontinuity across the
equatorial region where at present the littoral fauna is not known. That more records
are not available from southern waters is probably an expression of the relatively little
systematic work that has been done south of the equator in comparison with the
northern coast lines. Amphiporus pidcher has been recorded from Chili (Isler), Tubulanus

NEMERTEANS 289

annulatus from the Cape of Good Hope (Stimpson) with Tetrastemma candidum,
although the latter was described under the name T. incisum.

Dredging off the west coast of Africa was unfortunately not undertaken outside
Saldanha Bay. I should expect Baseodiscus delineatus to appear in the collection not-
withstanding its absence from the littoral fauna.

The material from South Georgia and the Falkland Islands consists of sub-Antarctic
and Antarctic species peculiar to these regions ; with representatives of the genera
Linens, Cerebratulus, Amphiporus and Tetrastemma in great diversity and numbers of
individuals. The Paleonemertea are apparently not represented in the far south. Most
of the forms taken are widespread or occur in the catches as isolated specimens from
which no deduction as to distribution can be made. Linens longifissus is an exception.
This species was collected by H.M.S. 'Challenger' from a depth of 126 metres off
Marion Island and described by Hubrecht from one complete and one fragmentary
specimen. At St. 167 the R.R.S. 'Discovery' took this species again from a depth of
244-344 m- °ff Signy Island in the South Orkney group. This was perhaps the most
curious haul made, for at this one station two nets attached to the back of the trawl
produced between them twenty-five specimens. The same gear was used off South
Georgia, the South Shetlands and the Falkland Islands, but no other specimen was
captured. The only parallel instance that occurred among the Nemerteans was the haul
of seventeen Pelagonemertes rollestoni in one haul of the 4 \ m. net at St. 107. Pelagone-
tnertes, however, is a purely pelagic form, and swarming or "patchiness" in pelagic
animals is a recognized phenomenon, not necessarily connected with breeding and some-
times on a large scale (as in Enphansia superba), so that the two hauls are not really com-
parable. What is interesting is the connection of Marion Island with the Antarctic by
the occurrence of this species and of Amphiporus marioni, Hubrecht, now recorded from
South Georgia, just as Kerguelen is linked by Linens corrugatns, Mcintosh, and
Amphiporus moseleyi, Hubrecht, and both are separated from the northern continent by
the dissimilarity in their Nemertean faunas. Coe (1905, p. 77) suggested that ocean
currents with limiting climatic conditions are a factor, if not the factor, in the dispersal
of Nemerteans in the Bering Sea. Applied to the data given above, however, dispersal
by ocean currents does not seem a possible hypothesis, for it is difficult to understand
how the current which passes up the coast of West Africa could have influenced the dis-
persal of the northern littoral forms to the southward, or, supposing dispersal took place
in the opposite direction, why Linens corrugatus has not become established on the
West African coast with the help of the cold Benguela current. The facts indicate, I
think, a creeping dispersal along an unbroken coastline, or at any rate a line of shallows
with no extensive deep-water masses in the path of dispersal; and they suggest, in
addition, that both Marion Island and Kerguelen were once more intimately connected
with Antarctica than they now are.

The stations at which pelagic forms were captured are shown in Fig. 66. The most
frequent capture was Pelagonemertes rollestoni, widely distributed in the South Atlantic
and Indian Oceans and also ranging north of the Equator. P. rollestoni has always been

2 go

DISCOVERY REPORTS

taken in deep-water nets, yet the depth at which it lives has not been definitely settled.
We have now an observation with a closing net which indicates 850-950 m. between

50-

60

i_

50

b»

90

60

40'

_l_

30°

—I

50UTH

760o

72 o ATLANTIC

039S

"" *~1 ■ 1 —

50 40 30

— FT

20

85000 «&

OCEAN

10

10°

■

E0°

— |—

30

— I 1-

40 50

60° 7Q°

-10

-10

20

■30

-40

-50

Fig. 66. Station positions at which pelagic Nemerteans were captured.

i and 4 p.m.; another, at which only one specimen was captured, of 750-1000 m. be-
tween 10 and 11 a.m.; and a third of 1500-1600 m. between noon and 3.30 p.m.

LIST OF LITERATURE

Baylis, H. A., 1915. Nemertinea. British Antarctic ('Terra Nova') Exped. 1910. Zool. 11, pp. 113-34,

2 pis., 4 figs.
IqI6. Some Nemertinea, free-living Nematoda and Oligochaeta from the Falklands. Ann. Mag. Nat.

Hist. (8), xvii, pp. 288-98, 4 figs.
Burger, O., 1892. Zur Systematik der Nemertinenfauna des Golfes von Neapel. Nachr. Ges. Wiss. Gottingen,

pp. 137-78.
^93. Siidgeorgische und andere exotische Nemertinen. Zool. Jahrb., Abth. Syst., VII, pp. 207-40, 2 pis.

1895. Die Nemertinen. Fauna und Flora des Golfes von Neapel, Monog. 22, pp. 1-743, 31 pis.

1899. Nemertinen. Hamburger Magalhaensische Sammelreise, iv, pp. 1-13, Hamburg.

■ 1904. Nemertini. Das Tierreich, xx, pp. 1-151, 15 figs., Berlin.

1904 a. Nemertinen. Exped. antarct., 1897-1899, pp. 1-10, pis. 1 and 2.

1897-1907. Nemertini (Schnurwiirmer). Bronn's Klassen, 4, Suppl., pp. 1-545, 22 P's-

■ 1907 (1912). Die Nemertinen. Deutsche Tiefsee-Exped., 1898-1899, xvi, 2 Heft, pp. 170-221, 13 pis.

Brinkmann, A., 1917. Die Pelagischen Nemertinen. Bergens Mus. Skrift., in, No. 1, pp. 1-194, 16 pis.

1921. Die Pelagischen Nemertinen der Deutschen Siidpolar-Exped., 1901-1903. Deutsche Sudpolar-

Exped., xvi (Zool., vm), pp. 279-98, 3 pis.

NEMERTEANS 291

Coe, W. R., 1905. Nemerteans of the west and northwest coasts of America. Bull. Mus. Comp. Zool. Harvard,

XLVII, pp. I-318, 25 pis.

Hubrecht, A. A. W., 1879. The genera of European Nemerteans critically revised, with descriptions of several

new species. Notes Leyden Mus., I, pp. 193-232.
1887. Report on the Nemertea collected by H.M.S. Challenger during the years 1873-76. Rep. Sci. Res.

'Challenger', xix, pp. 1-151, 16 pis.
Isler, E., 1902. Die Nemertinen der Sammlung Plate: Fauna Chilensis. Zool. Jahrb., Suppl.-Bd. v,

pp. 273-80.
Joubin, L., 1890. Recherches sur les Turbellariis des cdtes de France (Nemertes). Arch. Zool. exp. gen.
(2), vm, pp. 461-602, pis. xxv-xxxi.

1894. Les Nimertiens. Faune Francaise, pp. 1-235, Pls- i_iv> Paris-

1908. Nimertiens. Exped. antarct. francaise, 1903-5, pp. 1-16, 9 figs.

1906. Description des Nimertiens bathypilagiques capture's au corns des dernier s Campagnes du Prince de

Monaco. Bull. Mus. Oceanogr. Monaco, lxxviii, pp. 1-24.

■ 1914. Nhnertiens. Deuxieme Exped. antarct. francaise, 1908-10, pp. 1-33, 13 pis.

Langerhans, P., 1880. Die Wurmfauna von Madeira, III. Zeit. wiss. Zool., xxxiv, pp. 136-40, 9 figs.

1884. Die Wurmfauna von Madeira, IV. Zeit. wiss. Zool., XL, p. 283.

Lee, A. B., 1928. The Microtomist's Vade-Mecum, 9th Ed.

Marion, A. F., 1872. Recherches sur les animaux inferieurs du Golfe de Marseille. Ann. Sci. nat. (5),

xvii, Art. 6, pp. 1-23, pi. 17.
McIntosh.W. C, 1873. A Monograph of the British Annelids. Parti. The Nemerteans. 218 pp., 23 pis.,

Ray Soc. Pub., London.
Oudemans, A. C, 1885. The Circulatory and Nephridial Apparatus of the Nemertea. Quart. Journ. Microsc.

Sci., xxv, suppl., pp. 1-80, pis. 1-3.
Poche, F., 1926. Das System der Platodaria. Arch. Naturgesch., xci, pp. 388-96.
Punnett, R. C, 1901. Lineus. Liverpool Mar. Biol. Comm. Mem. pp. 1-37, pis. 1-4.
Quatrefages, A. de, 1846. Etudes sur les types inferieurs. Memoire sur la famille des Nemertiens. Ann. Sci.

nat. (3), vi, pp. 173-303, pis. 8-14.
Schmarda, K., 1859. Neue wirbellose Thiere beobachtet und gesammelt auf einer Reise urn die Lrde, 1853-57,

pp. 38-46, 3 pis. Leipzig.
Stimpson, W., 1856. Descriptions of some of the new Marine Invertebrata from the Chinese and Japanese Seas.

Proc. Acad. Nat. Sci. Philadelph., vil, pp. 375"84> 385"94-
1857-8. Prodromus descript. animalium evertebratorum, quae in Exped. ad Ocean. Pacificum. . .observavit

et descrip. Proc. Acad. Nat. Sci. Philadelph., pp. 159-65-
Woodworth, W. McM., 1899. Preliminary account of Planktonemertes agassizii, a new pelagic nemertean.

Bull. Mus. Comp. Zool. Harvard, xxxv, 1, pp. 1-3, 1 pi.

INDEX

(Synonyms are indicated by italics)

aerugatus, Cerebratulus, 231
Amphiporus cruciatus, 266

falklandicus, 223, 224, 257, 259

gerlachei, 224, 257, 258, 259, 263

inexpectatus, 224, 258

lecointei, 220, 221, 222, 223, 225, 257, 258,
263, 265

marioni, 259, 260, 262, 266, 289

michaelseni, 259, 278

moseleyi, 221, 224, 262, 263, 289

multihastatus, 266

pulcher, 238, 288

racovitzai, 263

schollaerti, 222, 265

scoresbyi, 224, 265

spinosissimus , 266

spinosus, 220, 221, 223, 224, 225, 266
amiulata, Carinella, 219, 227
annulatus, Tubulanus, 228, 288
antarcticus, Baseodiscus, 222, 223, 247
antonina, Eunemertes, 236, 237
autrani, Lineus, 250

banyulensis, Tubulanus, 228
Baseodiscus antarcticus, 222, 223, 247

delineatus, 289
Bathynemertes hardyi, 221, 280

hubrecbti, 221, 281
belgicae, Tetrastemma, 271
bilineatus, Lineus, 228, 288
bivittata, Nemertopsis, 245

Crassonemertes robusta, 220, 283, 284
cruciatus, Amphiporus, 266

delineatus, Baseodiscus, 289
dorsalis, Oerstedia, 245
259, Drepanophorus pelagicus, 281

echinoderma, Emplectonema, 243
Emplectonema echinoderma, 243

ophiocephala, 225, 234
esbenseni, Tetrastemma, 220, 221, 225, 270
Eunemertes antonina, 236, 237

falklandicus, Amphiporus, 223, 224, 257, 259
fuscus, Cerebratulus, 220, 225, 232, 288

geniculatus, Lineus, 222, 225
georgianum, Tetrastemma, 220, 221, 271
gerlachei, Amphiporus, 224, 257, 258, 259, 263
grytvikenensis, Parapolia, 221, 248
gulliveri, Tetrastemma, 221, 224, 272

hansi, Tetrastemma, 220, 274
hardyi, Bathynemertes, 221, 280
hubrechti, Bathynemertes, 221, 281

incisum, Tetrastemma, 243, 244, 289
inexpectatus, Amphiporus, 224, 258
irenae, Probalaenanemertes, 221, 287

candidum, Tetrastemma, 243, 245, 288, 289
capensis, Zygonemertes, 225, 239
Carinella annulata, 219, 227

nothus, 225
Cerebratulus aerugatus, 231

charcoti, 250, 253

corrugatus, 250

fuscus, 220, 225, 232, 288

larseni, 221, 256

longifissus, 255

magelhaensicus, 250

malvini, 223, 224, 233, 256

oleaginus, 230

steinini, 250, 253

subtilis, 250

validus, 250, 253
charcoti, Cerebratulus, 250, 253
corrugatus, Cerebratulus, 250

corrugatus, Lineus, 218, 219, 220, 221, 222, 223,
224, 250, 255, 256, 257, 289

kempi, Nectonemertes, 220, 283, 284

larseni, Cerebratulus, 221, 256

lecointei, Amphiporus, 220, 221, 222, 223, 225, 257,

258,259,263,265
Lineus autrani, 250

bilineatus, 228

corrugatus, 218, 219, 220, 221, 222, 223, 224, 250,

255. 2S6> 257. 289

geniculatus, 222, 225, 229

longifissus, 222, 255, 289

roseocephalus, 255

ruber, 218, 229, 230, 288
longifissus, Cerebratulus, 255
longifissus, Lineus, 222, 255, 289
longistriatum, Tetrastemma, 220, 221, 222, 275

maculata, Oerstedia, 225, 245
magelhaensicus, Cerebratulus, 250
maivikenensis, Tetrastemma, 225, 276
malvini, Cerebratulus, 223, 224, 233, 256
marioni, Amphiporus, 259, 260, 262, 266, 289

294

Meckelia olivacea, 230

michaelseni, Amphiporus, 259, 278

mirabilis, Nectonemertes, 221, 284

moseleyi, Amphiporus, 221, 224, 262, 263, 289

multihastatus, Amphiporus, 266

Nectonemertes kempi, 220, 283, 284

mirabilis, 221, 284
Nemertopsis bivittata, 245

tenuis, 223, 237, 288
nigrolineatum, Tetrastemma, 225, 244
nothus, Carinella, 225, 288
nothus, Tubulanus, 225, 228

Oerstedia dorsalis, 245

maculata, 225, 245
oleaginus, Cerebratulus, 230
olivacea, Meckelia, 230
Ommatoplea ophiocephala, 236
ophiocephala, Emplectonema, 225, 234
ophiocephala, Ommatoplea, 236

Parapolia grytvikenensis, 221, 248
pelagicus, Drepanophorus, 281
Pelagonemertes rollestoni, 220, 221, 222, 223,

285, 289
Probalaenanemertes irenae, 221, 287
pulcher, Amphiporus, 238, 288

racovitzai, Amphiporus, 263

robusta, Crassonemertes, 220, 283, 284

rollestoni, Pelagonemertes, 220, 221, 222, 223,

285, 289
roseocephalus, Lineus, 255
ruber, Lineus, 218, 229, 230, 288

INDEX

schollaerti, Amphiporus, 222, 265

scoresbyi, Amphiporus, 224, 265

spinosissimus , Amphiporus, 266

spinosus, Amphiporus, 220, 221, 223, 224, 225, 266

stanleyi, Tetrastemma, 276

steinini, Cerebratulus, 250, 253

subtilis, Cerebratulus, 250

tenuis, Nemertopsis, 223, 237, 288

Tetrastemma belgicae, 271

candidum, 243, 245, 288, 289

esbenseni, 220, 221, 225, 270

georgianum, 220, 221, 271

gulliveri, 221, 224, 272

hansi, 220, 274

incisum, 243, 244, 289

longistriatum, 220, 221, 222, 275

maivikenensis, 225, 276

nigrolineatum, 225, 244

stanleyi, 276

validum, 222, 277

vermiculus, 276

weddelli, 222, 278
Tubulanus annulatus, 228, 288

banyulensis, 228
280, nothus, 225, 228, 288

validum, Tetrastemma, 222, 277
validus, Cerebratulus, 250, 253
vermiculus, Tetrastemma, 276
virescens, Zygonemertes, 243

280, weddelli, Tetrastemma, 222, 278

Zygonemertes capensis, 225, 239
virescens, 243

PLATE XV

Fig. i. Tubulanus nothus, Burger, x 2.

Fig. 2. Tubulanus nothus. Head from above.

Fig. 3- Zygonemertes capensis, n.sp., green variant, x 5.

Fig. 4. Lineus ruber, O. F. Miiller. Head, x 6.

Fig. 5. Cerebratulus aerugatus, Burger. Head, x 6.

Fig. 6. Zygonemertes capensis, n.sp. Head of colourless variant, x 5.

Fig. 7. Nemertopsis tenuis, Burger, x 1 £.

Fig. 8. Cerebratulus fuscus, Mcintosh, x 5.

Fig. 9. Tetrastemma nigrolineatum, n.sp. Head, x 10.

Fig. 10. L?Wi« bilineatus, Renier. x 10.

Fig. 11. Oerstedia maculata, n.sp. x 10.

Fig. 12. Zygonemertes capensis, n.sp. Brown variant, x 5.

Fig. 13. Amphiporus pulcher, Johnston, x 5.

Fig. 14. Emplectonema ophiocephala (Schmarda). x 5.

Fig. 15. Tetrastemma candidum, O. F. Miiller. Head, x 5.

DISCOVERY REPORTS, VOL. IX

PLATE XV

■r. i ™ L widim

NEMERTEANS

'

PLATE XVI

Fig. i. Tetrastemma validum, Burger, x 3.

Fig. 2. Tetrastemma georgianum, Burger, x 5.

Fig. 3. Amphiporus moseleyi, Hubrecht. x 1.

Fig. 4. Tetrastemma esbenseni, n.sp. x 5.

Fig. 5. Bathynemertes hardyi, n.sp. x 1.

Fig. 6. Tetrastemma hansi, Burger, x 10.

Fig. 7. Tetrastemma longistriatum, n.sp. x 10.

Fig. 8. Cerebratulas larseni, n.sp. Head, x 5.

Fig. 9. Amphiporus lecointei, Burger, x 4.

Fig. 10. Tetrastemma maivikenensis , n.sp. Head, x 5.

Fig. 11. Tetrastemma gulliveri, Burger. Head, x 5.

Fig. 12. Tetrastemma stanleyi,n.sp. x 3.

Fig. 13. Tetrastemma georgianum, Burger, x 6.

Fig. 14. Parapolia grytvikenensis , n.sp. x 6.

Fig. 15. Tetrastemma hansi, Burger, x 10.

Fig. 16. Lineus corrugatus, Mcintosh, x 1.

Fig. 17. Tetrastemma gulliveri, Burger, x 5.

Fig. 18. Amphiporus moseleyi, Hubrecht. x 1.

Fig. 19. Lineus corrugatus, Mcintosh. Head, x 3.

Fig. 20. Lineus corrugatus, Mcintosh. Head, x 3.

Fig. 21. Lineus corrugatus, Mcintosh, x 1.

Fig. 22. Amphiporus spinosus, Burger, x 3.

Fig. 23. Tetrastemma esbenseni, n.sp. x 5.

Fig. 24. Lineus roseocephalus, n.sp. x 3.

DISCOVERY REPORTS, VOL. IX

PLATE XVI

NEMERTEANS

[Discovery Reports. Vol. IX, pp. 295-350, Plates XVU-XXII, December 1934]

THE SEA-FLOOR DEPOSITS

I. GENERAL CHARACTERS AND DISTRIBUTION

By
E. NEAVERSON, D.Sc, F.G.S.

4-

ON

WOODS
MOLE,

CONTENTS

Introduction

Classification of the Deposits . • •
General Distribution of the Deposits .

South Atlantic Region

The West and South Coasts of Africa

The Falkland Islands

South Georgia

The South Shetlands

The Bellingshausen Sea ....

The Western Coast of South America
Descriptions of the Samples . . . •
Plates

page 297
. 298

300

300

3°4

• 3°5
306

. 3°7

. 3°9

310

. . ■ • 312
following page 349

THE SEA-FLOOR DEPOSITS

I. GENERAL CHARACTERS AND DISTRIBUTION

By E. Neaverson, D.Sc, F.G.S.

(Plates XVII-XXII)

INTRODUCTION

IN 1929 142 samples of sea-floor deposits, collected during voyages of R.R.S.
'Discovery' and R.R.S. 'William Scoresby', were sent to Professor P. G. H. Boswell
for examination by himself and his colleagues on the staff of the Department of Geology,
University of Liverpool. Before the writer's part of the task was completed additional
samples collected on later voyages became available, and descriptions of this material
are now incorporated in this report. Most of the samples are numbered according to
the observation stations on the voyages, thus giving a chronological arrangement, which
is retained in the detailed descriptions of the samples. In an account of the distribution
of the sediments a regional classification is more convenient. Such is shown in the
accompanying table, the regions being defined on the basis of maps which are appended
to the official Station Lists. Sixteen of the earlier samples are not numbered, but the
latitude and longitude of the localities are recorded. These samples are now indicated
by the letters A to P, and are placed in their appropriate position in the geographical
classification.

The purpose of this report is to record the general characters and distribution of the
sea-floor deposits before the samples are used for a detailed mineralogical analysis. The
station numbers are those given in the official Station Lists; the localities of the
'William Scoresby' are distinguished by prefixing the letters WS to the number.
Precise determination of organic species is not attempted here ; such detail is not con-
sidered to lie within the scope of a report on general characters of the sea-floor deposits.
In many cases, however, genera of diatoms, Foraminifera and Radiolaria are noted
where they are abundant or otherwise conspicuous. Similarly the specific identification
of minerals is left over for the future report on the mineralogical characters of the
deposits. Casual reference, however, is made to certain minerals when they are
especially significant or abundant. Such references are, for example, to the abundance
of well-formed crystals of hypersthene at stations around the Falkland Islands (p. 305) ;
the plentiful occurrence of volcanic glass in deposits around the South Shetland Islands
(p. 308) ; the abundance of glaucophane in deposits from the western part of Bransfield
Strait (p. 309); and the sporadic occurrence of phillipsite at several localities in the
southern oceans.

298

DISCOVERY REPORTS

Table I

Region

Mid and South Atlantic

Plate

XVII

2. West and South coasts of
Africa

3. Falkland Islands

4. South Georgia

South Shetlands

XVII

XVIII

XIX

XX

6. Bellingshausen Sea

7. West coast of South
America

XXI

XXII

Locality numbers in Station List

R.R.S. 'Discovery'

8, 9, 10, 11, 12, 63, 64,
71.74.77.78,83,84.89.
117, 160, 162, 167, 169,
212, 235, 236, 237, 363,
366, 1165, A, B, C, D,
E, F, G, J

93. 94. 95- 96> 97, 98,
101, 102, 263, 264, 265,
271, 283, 425, K, L, M

48, 51, 228, 230

13, 14, 15, 16, 19, 20,
23, 28, 29, 30, 31, 41,
42, 123, 129, 131, 136,

145, ISI> I57, H. !>
MS 68

171, 172, 175, 180, 185,
187, 191, 192, 194, 195.
196, 197, 198, 199, 200,

201, 202, 203, 204, 206,
209, 211, N, O

R.S.S. 'William Scoresby'

128, 129, 201, 202, 203, 204,

205, 255, 314, 317. 3r9, 374,
377, 381,403,406,428,429,

433,468,469,470,471,472.
474,518,520,521,522,524,

525, 526, 768

7i,76, 77, 78, 79, 8o> 83.84.
86,87,88,89,90,91,92,93,

94. 95. 9°. 97. 98. 99, Io8,
109

18, 20, 26, 28, 32, 33, 37, 39,
40, 41,42,43,45,46. 47. 48,
49, 5°, 52> 63. p

382, 383, 384. 385, 386, 387.
388, 389, 391, 392. 393. 394,
395, 396, 397, 399, 4°°, 475.
476, 477, 479, 480. 481, 482,
483, 484, 485, 486, 487, 488,
489, 490, 493, 494A, 494B

495, 496, 497, 498, 499, 5°i,
502, 503, 505, 506, 507, 508,
509, 510, 511, 512, 513, 514.
515, 5l6> 5*7

591.596, 597. 598, 599. 6o2'
604, 605, 610, 616, 617, 619,
622, 623, 638, 639, 640, 641,
642, 647, 648, 649, 650, 651,
652, 655, 658, 663, 664, 665,
666, 667, 669, 671, 676, 677,
678, 680, 686, 689, 692, 694,
696, 697, 700, 701, 702, 703,
705,708,711,717

CLASSIFICATION OF THE DEPOSITS

In this report the well-known classification of marine deposits, devised by Murray

and Retard,* is followed in the main. Most of the samples are already named on this

basis in the official Station List. But as the investigation proceeded it became evident

1 Murray, J. and Renard, A. F., 1891. The Deep Sea Deposits. Challenger Reports.

SEA-FLOOR DEPOSITS. PART I 299

that some of these terms (presumably based on a necessarily casual examination on
board ship) are not sufficiently precise for a formal description. Moreover it is thought
desirable to modify the classification of Murray and Renard to a slight extent.

It has long been recognized that the dividing line between pelagic and terrigenous
deposits is indefinite, and that deep-sea oozes may contain a slight proportion of
detrital material transported from areas of denudation. But many of the deposits named
"diatom ooze" in the official Station List have a predominance of detrital grains,
although diatoms may occur in considerable abundance. Here, such deposits are classi-
fied as " diatomaceous mud", a term which indicates the importance of both organic
and mineral material in their constitution. Some samples labelled "green mud" are
also placed in this new category.

A constant feature of the diatomaceous muds is the abundance of flocculent material
which encloses and buoys up quantities of detrital mineral grains. This flocculent sub-
stance is apparently formed largely of perishable organic debris (such as algae and
animal plankton in various stages of decomposition), which is capable of drifting some
considerable distance. The flocculent masses must form an important factor in the
transport of terrigenous material which is sometimes found in surprising abundance at
considerable depth. As far as the writer is aware this factor has only been noted casually
in published works, but is evidently of wide application. The same phenomenon has
been observed in black muds (fetid with sulphuretted hydrogen and containing sporadic
oil globules) from the margins of the Wash, and from submerged banks in the estuary
of the River Mersey in England. Here, the organic debris clings round the detrital
quartz grains, and masses of the combined material float in moving water ; when move-
ment ceases they sink rapidly and soon become agglutinated.1 The whole question seems
to call for detailed observation over a wider field.

" Green mud ", as the term is usually defined, is distinguished by the abundance of the
mineral glauconite, which is sufficient to impart the characteristic green colour to the
deposit. In many of the so-called "green muds " in the Discovery collection, no grains
of glauconite have been noted. In these instances the green colour appears to be due to
the presence of chlorophyll, the green colouring matter which occurs in the chromato-
phores of diatoms and other algae, though often masked by other pigments. Much
organic debris occurs in such samples; in some the chlorophyll grains themselves are
visible, and in others the preservative liquid has become green by solution of the pig-
ment. It may be added that chlorophyll granules become brownish on decomposition
and oxidation, and thus may give rise to grey or brown muds. Hence mere colour is no
criterion of origin and should not be used as a guide to classification of the deposits.
Most of these "green muds" are more adequately described as diatomaceous muds.
The samples which do contain glauconite are here termed glauconitic muds in order to
indicate their characteristic quality more precisely. The quantity of glauconite is often
insufficient to impart a green colour to the sediment, but the presence of the mineral is
indicative of peculiar conditions in the sea water.

1 Neaverson, E., 1928. Stratigraphical Palaeontology , p. 100, 8vo, London.

3oo

DISCOVERY REPORTS
The classification adopted for the Discovery samples may be tabulated as follows:

Table II

Shallow-water deposits, be-
tween low water mark and
ioo fathoms deep

Deep-sea deposits, from
depths greater than ioo
fathoms

Gravels
Sands
Muds

Glauconitic mud
Diatomaceous mud
Diatom ooze
Globigerina ooze
Radiolarian ooze

Terrigenous deposits, con-
V taining detrital material
derived from land masses

Pelagic deposits, of organic
origin, formed outside the
range of detritus-bearing
currents

GENERAL DISTRIBUTION OF THE DEPOSITS

SOUTH ATLANTIC REGION (Plate XVII)

This region is represented by fifty-nine samples from numbered stations as indicated
in Table III. In addition, eight samples from other localities (indicated by letters) are
included, of which five (A, B, C, D, E) are from stations situated north of the area
covered by Plate XVII. The sixty-seven samples may be classified thus:

Table III

Globigerina ooze

Radiolarian ooze
Diatom ooze
Diatomaceous mud

Glauconitic mud
Terrigenous mud
Sand

Locality numbers in Station List

R.R.S. 'Discovery'

A,B,C,D,E,8,9,64,78,83,
84,89

71.74.77, il65
F, G, 10, 11, 12
63, 117, 162, 167, 169, 212

J

160, 235, 236, 237, 363, 366

R.S.S. 'William Scoresby'

128, 129, 520, 521, 522

255- 374. 377. 429

201, 202, 203, 204, 205, 403,
428, 468, 469, 470, 471, 472,
474. S24. 526, 768

433. Si8

314,317,319,381,406,525

Globigerina ooze is the most widespread deposit accumulating in the vast region
covered by the South Atlantic Ocean, and all the localities here recorded fall within the
area shown for Globigerina ooze on the map appended to the Challenger Report. The
majority lie in the main area to the north of lat. 460 S, but four of them appear somewhat
isolated in the Discovery records. The most westerly occurrence is at St. 64, to the north-
east of the Falkland Islands; the deposit is hardly typical Globigerina ooze, but the
locality is near the line drawn by Murray west of the South American coast, to separate

SEA-FLOOR DEPOSITS. PART I 301

the ooze from terrigenous deposits ; moreover the sample contains grains of glauconite.
Three stations, namely, WS 520, WS 521, and WS 522, which lie between the Falkland
Islands and South Georgia, between lat. 520 and 53 ° S, represent the most southerly
localities from which Globigerina ooze was collected by the Discovery Expeditions.
A noteworthy feature at St. WS 522 is the presence of a fair proportion of diatoms
among the smaller organisms ; but an extensive area occupied by diatom ooze occurs at
no great distance to the east. The deposit at St. WS 522 also contains the zeolite mineral
phillipsite along with coarse sand grains and occasional pebbles, the latter up to 1 cm.
in diameter. Another occurrence of pebbles at an even greater depth in the same area
(St. WS 317) is mentioned on p. 304 of this report. In general, Globigerina ooze is
confined to stations of considerable depth (2000-4600 m. in the present investigation)
situated far from land. The occurrence of pebbles in such deposits is therefore an
unusual feature; in this instance it seems to be connected with the unusually varied
character of the submarine topography.1

Radiolarian ooze. Hitherto, no radiolarian ooze has been recorded from the South
Atlantic Ocean, but the Discovery Expedition obtained samples of this deposit in the
western part of the region from Sts. 71, 74 and 77, at depths of 5460, 5446 and 5186 m.
respectively. In an examination of the deposits themselves, Radiolaria are not much in
evidence, being obscured by the finer material, but the organisms are very plentiful in
coarse washings from the sediments ; diatoms and sponge spicules occur in subordinate
quantity. Reference to Plate XVI I shows that radiolarian ooze must occupy a considerable
area in this part of the South Atlantic which was included by Murray and also by Pirie
in the belt of Globigerina ooze. It may be noted that St. 71 is near the middle of the
Argentine Basin.2 Another sample of radiolarian ooze was obtained at St. 11 65 (long.
90 25-5' E, lat. 400 547' S) at a depth of 4642 m.

Diatom ooze. Five samples of diatom ooze were obtained by R.R.S. 'Discovery'
along a north-east-south-west line, north-east of South Georgia. The samples from
Sts. 10 and 11 are recorded in the Station List as "radiolarian ooze", but Radiolaria
are found to be subordinate to diatoms in these deposits. Four stations made by the
'William Scoresby' south-west and west of South Georgia also yield typical diatom
ooze. Hence a wide stretch of ocean in this region is marked by the abundance of
diatoms and the absence of Foraminifera. The occurrence of diatomaceous deposits here
does not conform to Murray's map which includes the area in the belt of Globigerina
ooze, as does the later map of Pirie.3 The most westerly of these stations, namely,
WS 377 and WS 429, are in approximately 450 W long. This appears to be near the
western limit of the diatom ooze, for St. WS 522, in 470 W long, and about the same
latitude as St. WS 429, yields Globigerina ooze. These stations lie along the line of a
submarine ridge which can be traced almost continuously between the Burdwood Bank,

1 See Herdman, H. F. P., 1932. Report on Soundings, etc. Discovery Reports, VI, p. 219 and plate xlv.

2 See Wust, G., 1933. Wiss. Ergebn. Deutschen Atlant. Exped. 'Meteor', VI, i, plate viii.

3 Pirie, J. H. H., 1913. Deep-Sea Deposits of the Scottish National Antarctic Expedition. Trans. Roy.
Soc. Edin., xlix, pp. 645-86.

302 DISCOVERY REPORTS

south of the Falkland Islands, and the Shag Rocks, west of South Georgia. As Herdman
has shown, there appears to be a gap of about 60 miles (long. 47-490 W) where the sound-
ings show greater depths than usual. This depression is occupied by Globigerina ooze
(Sts. WS 521, WS 522), while the diatom ooze under discussion is accumulating on the
eastern slope at a depth of 2549 m. (St. 429). This is about an average depth for the
occurrence of diatom ooze, though the deposit is recorded from 5000 m. at St. 11,
north-east of South Georgia. The deposit is remarkably uniform in character over the
whole region. The genera which occur most constantly are Coscinodisciis, Cocconeis,
Fragilaria, Thalassiothrix and Rhizosolenia, while other forms, such as Tholassiosira,
Biddulphia and Achnanthes are occasionally present. There is a minimum of terrigenous
material and an entire absence of the perishable organic debris which is so characteristic
of diatomaceous muds. Nearer to South Georgia the diatomaceous deposits contain a
considerable proportion of terrigenous material evidently transported from the land
area; these deposits are assigned to the category next to be described.

Diatomaceous mud. The great majority of the twenty-six stations which yield
diatomaceous mud lie in the Scotia Sea, north of the South Shetland and South Orkney
Islands. The samples from these stations conform very closely to those from the regions
of South Georgia and the South Shetlands, but there is some variation in general con-
stitution. In this connection, the sample from St. WS 204 may be quoted as illustrating
gradation from diatomaceous mud to diatom ooze, for it is almost of sufficient purity to
be included in the latter category; moreover the perishable part of the organic debris
has almost disappeared. The diatomaceous muds occur at greatly varying depth; St.
WS 768 yields diatomaceous mud at a depth of 108 m., and a similar deposit occurs at
St. WS 201 at a depth of 4134 m. In both deposits some of the mineral grains reach a
diameter of 0-5 mm., a size which is surprising at the latter depth. It must be inferred
that depth in itself is unimportant compared with distance from land and the influence
of detritus-bearing currents. Among the inorganic constituents, quartz and green horn-
blende are the most abundant and widespread minerals. The presence of a few other
minerals, noted incidentally during the general examination of the deposits, may be
recorded here, though the significance of their occurrence can only be determined when
the mineralogical analysis of all the samples is completed. Well-formed crystals of
hypersthene are noted in the samples from Sts. WS 400 and WS 471 ; this mineral also
occurs in samples from the west of the Falkland Islands (p. 305). The presence of
glaucophane at St. WS 403 doubtless has some connection with its widespread occur-
rence at stations in the western part of the Bransfield Strait further south (p. 309). The
interesting zeolite-mineral phillipsite recorded from St. WS 470 is of sporadic occur-
rence and is probably connected with the chemical and physical conditions which de-
termine the decomposition of volcanic rocks on the sea floor. The organic constituents
may include Radiolaria and sponge spicules, but diatoms are preponderant. Coscino-
disciis, Tholassiosira, Fragilaria, Cocconeis, Rhizosolenia and Corethron are the genera
most commonly seen, though from time to time other forms such as Triceratium,
Biddulphia and Asteromphalus are noted. Only occasionally are calcareous shells present

SEA-FLOOR DEPOSITS. PART I 303

in these deposits. In samples from Sts. WS 468 and WS 469 tests of Globigerina occur,
but their broken and worn appearance suggests that they are drifted material. The
characters of a sample from Drake Strait (St. WS 403) are less easy to interpret. The
coarser fraction contains many tests of Globigerina and rotaline Foraminifera together
with Radiolaria, but mingled with these are abundant angular sand grains up to o-i mm.
across ; the abundance of the latter prevents classification as an ooze. On the other hand,
the finer fraction is mainly of diatomaceous origin ; this and the terrigenous matter seem
to determine the deposit as a diatomaceous mud, despite the presence of the Fora-
minifera. St. WS 403 lies within the limits of a belt through Drake Strait considered by
Pirie to consist of diatom ooze.

With regard to the sample from Bouvet Island, St. 117 lies well within the belt of
diatom ooze as indicated on the maps of Murray and Pirie. Both these observers would
doubtless classify the sample as diatom ooze, but they constantly remark that the de-
posits of this area are not typical diatom ooze.

Glauconitic mud. Most of the glauconitic deep-sea deposits in the Discovery
collection are from the South African region (p. 304), but two stations, far removed
from that area, must be considered here. About 80 miles east of the Falkland Islands
(Sts. WS 433 and WS 518) glauconitic mud was dredged from depths of 1035 and
1258 m. In each sample grains of glauconite are seen infilling the broken tests of
Foraminifera, and the shape of other grains is consistent with formation within such
shells. The localities are near the south-western margin of the Globigerina ooze belt, and
they may be within the overlapping limits of the cold north-flowing Antarctic Drift and
the warm south-flowing extension of the Brazil current. The occurrence of glauconite
here is therefore in accordance with prevalent theories regarding the conditions under
which the mineral is formed.

Terrigenous deposits. It is only to be expected that sediments of detrital origin do
not occur to any great extent in this oceanic region. Twelve stations, however, have
yielded samples of terrigenous deposits ; eleven of these are classed as sands and one as
terrigenous mud. The latter is a rather sandy grey mud obtained at St. J in a depth of
892 m. St. J is about 150 miles north-east of the Falkland Islands, which lie on an
extension of the Patagonian continental shelf; the station appears to be situated on the
somewhat steep gradient which slopes rapidly to oceanic depths. The deposit is there-
fore consistent with a situation within the influence of detritus-bearing currents. Even
so, the paucity of organic remains is remarkable.

The sands are from widely separated localities, each of which may be discussed
separately. Sts. 235, 236 and 237 to the north-east of the Falkland Islands have yielded
sands from depths of 600, 612, and 904 m. respectively. Evidently these stations lie on
an easy gradient which continues northwards to about lat. 470, where it begins to slope
more rapidly to oceanic depths. The occurrence of these sands is consistent with north-
flowing currents bearing detritus from the Falkland Islands area.

The samples from St. WS 317 and WS 319 appear to be incomplete; indeed the record
of the latter sample in the Station List bears a note, " finer material from bottom sample

304 DISCOVERY REPORTS

washed out". The presence of small pebbles, however, at depths of 3369 and 1602 m.
is unexpected, especially as the former locality is within the belt of Globigerina ooze.
In view of the uncertainty as to the constitution of these samples, the deposits are not
shown on the accompanying map.

St. WS 406, on the northern side of Drake Strait, yields a foraminiferal sand from a
depth of 1234 m. The proportion of sand grains forbids reference to this deposit as an
ooze, though Pirie postulates a belt of Globigerina ooze through Drake Strait. It may be
recalled that the same difficulty occurs with regard to a sample from St. WS 403 farther
south, which is tentatively classified as diatomaceous mud (p. 303).

A clean coarse sand from St. 381, just west of Elephant Island in the South Shetlands,
at a depth of 425 m., is noteworthy for the presence of garnet among the detrital
minerals.

Two samples of sand from the South Sandwich Islands (Sts. 363 and 366) are
remarkable for the abundant fragments of vesicular volcanic glass, the angularity and
freshness of which proclaim their local origin. These stations, with depths of 329 and
340 m., lie on the eastern part of an arcuate submarine ridge which connects South
Georgia with the South Orkney Islands, and includes the South Sandwich group. The
latter islands are recorded1 as volcanic centres, and doubtless the sands are formed by
the denudation of the eruptive material.

The two remaining sands are from Sts. 160 and WS 314, in the neighbourhood of the
Shag Rocks, west of South Georgia, at depths of 177 and 137 m. respectively. The most
interesting feature is the presence at St. WS 314 of colourless flakes some of which
show hexagonal outlines; the radial texture of many suggests that they are zeolitic
aggregates formed by the decomposition of volcanic rock.

THE WEST AND SOUTH COASTS OF AFRICA (Plate XVII inset)

Seventeen stations, namely Sts. K, L, M, 89, 93, 94, 95, 96, 97, 98, 101, 102, 263,
264, 265, 283 and 425, are grouped in this area. The most northerly station (283) off
Annobon in the Gulf of Guinea is outside the limits of Plate XVII. The sample is a fairly
coarse foraminiferal sand, containing very few mineral grains; the depth is recorded as
18-30 m.

The remaining samples were collected off the south and south-west coasts of Africa.
The stations are shown on Plate XVII (inset), with the exception of the isolated St. 425,
south-east of Port Elizabeth, which is outside the eastern limit of the map. The samples
from the stations nearest land (Sts. 93, 94, 95, 96, 97, 263, 264, 265) contain the mineral
glauconite, though this constituent is not abundant in any of them. It generally occurs
in rounded grains, some of which are compound or lobate. In some samples (e.g.
St. 96) the glauconite is seen infilling the chambers of broken Globigerina shells. These
foraminiferal tests are so abundant at Sts. 95, 96 and 97 that the samples might almost

1 Douglas, G. W., and Campbell Smith, W., 1930. Zavodovski Island, and notes on Rock Fragments
dredged in the Weddell Sea. Quest Report, p. 63. Kemp, S., and Nelson, A. L., 1931. The South
Sandwich Islands. Discovery Reports, III, p. 150.

SEA-FLOOR DEPOSITS. PART I 305

be described as Globigerina ooze; but they contain a fair proportion of mineral grains,
and the presence of glauconite gives a distinct character to the deposit. Glauconitic
mud is accumulating at depths of 165 to nearly 1000 m., to a distance of some 60 miles
from the coast, while the Agulhas Bank, long famed for glauconitic deposits, lies at no
great distance to the south-east. Farther west, at Sts. 89, 98, 101 and 102, typical
Globigerina ooze with only occasional mineral grains occurs at depths of 3926, 3640,
3734 and 1800 m. respectively.

THE FALKLAND ISLANDS (Plate XVIII)

The twenty-eight samples from this region are of very uniform character. Most of
them are sands composed essentially of terrigenous material. The depth is slight,
varying in different parts of the area between 23 and 251 m.; and the region is uni-
versally accepted as forming part of the South American continental shelf.

Recently, these sands have been classified by L. H. Matthews1 on the basis of
mechanical analysis, with the object of providing data concerning the habitats of the
organisms which live on them. Geologically, the chief interest of these deposits lies in
their mineral content, but as this will be fully discussed in a future report, only a brief
reference to minerals of common occurrence need be given here. Like the majority of
sands, these contain a preponderance of quartz grains, and this feature may be corre-
lated with the wide distribution of sandstones and quartzitic rocks in West Falkland
Island and the north part of East Falkland Island.'2 The outstanding point of interest is
the abundance, in most of the samples, of well-formed (but worn) crystals of hypers-
thene, a mineral which is characteristic of the andesitic volcanic rocks of South America.
The sample from St. 48 yields rounded grains of red garnet and brown tourmaline in
addition to hypersthene. Foraminifera occur plentifully in some of the samples, and
diatoms are also present, either alone or associated with the Foraminifera.

South of the Falkland Islands, between lat. 540 and 550 S and long. 560 and 620 W,
is the Burdwood Bank, which is separated from the Falklands by comparatively deep
water. Two samples (Sts. WS 86 and WS 87) are available from the Burdwood Bank at
depths of 151 and 96 m. respectively. They consist of sandy deposits with a varied
assortment of shelly material including an abundance of Foraminifera. A sample from
the eastern portion of this bank, obtained by the Scotia Expedition (1903), has been
described by Pirie,3 and the station was included by him in the belt of Globigerina ooze.
This, however, is hardly justified by his description of the deposit, as the Foraminifera
are said to be of shallow-water Antarctic types. The abundance of terrigenous material
in the deposits, considered in conjunction with the soundings,4 shows that the Burdwood

1 Matthews, L. H., 1934. The Marine Deposits of the Patagonian Continental Shelf. Discovery Reports,
ix, pp. 175-206.

2 Baker, H. A., 1923. Final Report on Geological Investigations in the Falkland Islands (1920-1922),
Fol. London.

3 Pirie, J. H. H., 1913. Deep-Sea Deposits of the Scottish National Antarctic Expedition. Tram. Roy.
Soc. Edin., xlix, pp. 645-86.

4 Herdman, H. F. P., 1932. Report on Soundings, etc. Discovery Reports, VI, pp. 205-36.

3o6 DISCOVERY REPORTS

Bank belongs to the shallow seas of the continental shelf. It may be noted that frag-
ments of shale from St. WS 87 have been examined and described by W. A. Macfayden,1
who deduces a Cretaceous age for the shale from the evidence of Foraminifera contained
therein. He also suggests, from the abundance of shale pebbles in the dredgings, that
the beds must outcrop on the sea-bottom at or close to the stations. Some of the shale
fragments are highly glauconitic, grains of this mineral being prominently exposed on
the worn surfaces. It follows that some (at least) of the glauconite grains present in the
modern deposit are "derived" by disintegration of the shale.

Two stations (228 and 230) are situated in the depression between the Falkland
Islands and Burdwood Bank, where depths of 660 and 675 m. are recorded. The first-
named station yields a diatomaceous mud which is similar in general constitution to the
great majority of such deposits from the Scotia Sea. The deposit from St. 230 is classed
as terrigenous mud because of the abundance of detrital mineral grains and the absence
of recognizable organic remains. This procedure, though unavoidable, is somewhat
unsatisfactory, for it leaves out of account the considerable proportion of flocculent
material which contributes largely to the bulk of the deposit. It is undoubtedly of
organic origin but is so indefinite that its precise nature cannot be determined. But
from the geological standpoint the omission is not so serious ; for when thoroughly dried,
the flocculent material shrinks to a mere film round the detrital grains, and the deposit
eventually approaches the type exemplified by the marine silts of the English Fenland.

SOUTH GEORGIA (Plate XIX)

From this region the 'Discovery' collected twenty-three samples, while a further
twenty-one were obtained by the 'William Scoresby'. The majority of the forty-five
samples are of very uniform character, falling within the group classified in this report
as diatomaceous mud.

Judging by the records of soundings, South Georgia is surrounded by a belt of shallow
sea, varying from 20 to 300 m. in depth. Outside this area, usually at about 30 or 40
miles from the island, the depth increases rapidly to 1000 m. and more. This variation
in depth appears to have little (if any) influence on the character of the deposits, for
diatomaceous mud from Sts. 129 (depth 1001 m.), 151 (depth 3200 m.), WS 26 (depth
1 180 m.) and WS 63 (depth 1752 m.), are essentially similar to diatomaceous mud from
stations close inshore. The finer fraction consists mainly of diatoms, with sponge
spicules and Radiolaria in subordinate quantity. Mineral grains are present in con-
spicuous proportion, and the size of the grains appears not to be related to depth alone.
For, at St. 129 the largest grains noted are about o-i mm. across, whereas at St. 151
grains reaching 0-2 mm. in diameter were noted, though the depth is three times that
of the former locality. Green hornblende occurs in several of the samples, particularly
those from the more southerly stations. The relative abundance of this mineral in
deposits from the mouth of Drygalski Fjord, at the south-east of the island, is probably

1 Macfayden, W. S., 1933. Fossil Foraminifera from the Burdwood Bank and their geological significance.
Discovery Reports, vn, pp. 1-16.

SEA-FLOOR DEPOSITS. PART I

307

to be explained by the recorded1 occurrence of igneous rocks such as gabbro and
diorite on the adjacent mainland. At the more northerly stations the mineral seems to
be less plentiful. Of the diatoms, Coscinodiscus, Thalassiosira and Fragilaria are the
genera most frequently seen ; Rhizosolenia, so abundant in samples from the Scotia Sea,
is relatively scarce.

The sample from St. 136 differs from all the others in containing deep green glau-
conite among the abundant mineral grains. Some of this green material is also seen
infilling the central canal of sponge spicules. The finer constituent of the deposit con-
sists almost entirely of diatom frustules, along with some sponge spicules. The apparent
absence of calcareous material is noteworthy in view of the presence of glauconite, for
in other instances calcareous shelly material is constantly seen in association with this
mineral.

The remaining three samples, from Sts. 28, 29 and 30 in West Cumberland Bay, are
classified as terrigenous mud. The material is exceedingly fine-grained, and recognizable
organic debris is restricted to occasional centric diatoms, though oil globules are com-
monly seen. The deposit, apart from the diatoms, is strikingly similar to sedimentary
material from the estuary of the River Mersey, and from the shores of the bay known
as the Wash, in England.

THE SOUTH SHETL AND S (Plate XX)
This region is conveniently subdivided into two, (a) the Bransfield Strait, (b) the
Palmer Archipelago, represented by forty-two and fifteen samples respectively. The
stations are as follows:

Table IV

Bransfield Strait

Locality numbers in Station List

R.R.S. 'Discovery'

R.S.S. "William Scoresby'

171, 172, 175, 194, 195,
196, 197, 198, 199, 200,
201, 202, 203, 204, 206,
209, N

382, 383, 384, 385, 386,
387,388,389,391,392.
393. 394. 475. 476, 477,
479, 480, 481, 482, 483,
484, 485, 486, 487, 493

Palmer Archipelago

180, 185, 187, 191, 192,
O

395. 396, 397. 399. 488,
489, 490, 494 a, 494 B

Two outlying stations (211, WS 400) to the north-west, also fall within the area of
the chart.

0) Bransfield Strait, between the South Shetlands and the northern part of Graham
Land, has its greatest recorded depth (about 2000 m.) near its north-eastern outlet.
From this end of the strait, between King George Island and Trinity Peninsula,

1 Tyrrell, G. W., 1930. The Petrography and Geology of South Georgia. Quest Report, pp. 28-34.

3o8 DISCOVERY REPORTS

twenty-one samples are described, of which fifteen are diatomaceous muds, four are
terrigenous muds, and two are classed as sands. One of the latter (St. 195) is from a
depth of 391 m. in Admiralty Bay, King George Island, and the other (St. WS 481) is
from a depth of 453 m. on the Graham Land shelf, east of Astrolabe Island. The samples
of terrigenous mud are from Sts. 196, WS 312, WS 476 and WS 477, south of Martin's
Head, King George Island, at depths which vary from 425 to 1892 m. Most of the
samples of diatomaceous mud are from the centre of the depression, where the depth
varies between 446 m. towards the south of the traverse and 2085 m. south of King
George Island. While the distribution of terrigenous and diatomaceous muds appears to
accord generally with depth, there are records which invite further discussion. South
of King George Island terrigenous mud occurs at unusual depths for this deposit, but
the location is at no great distance from the land surface, which slopes down under the
sea with a steep gradient. On the southern side of the Strait the sea-floor falls away
quite gently from the land, so that the depth contour of 250 m. lies some 25 miles from
the coast of Trinity Peninsula. Thus the occurrence of diatomaceous mud at the slight
depth of 345 m. in this part of the strait is explained. But the presence of a similar
deposit at a depth of 152 m. close to the northern coast of Trinity Peninsula (St.
WS 482) seems to demand conditions, perhaps peculiar to the locality, which are not
apparent from the records.

A further series of twenty-one samples is available from the western part of Bransfield
Strait. Here the depression is more shallow and the deposits contain a larger proportion
of detrital material. Fourteen samples of diatomaceous mud are from the deeper parts
of the sea-floor (800-1000 m.), but towards the south the gentle slope of the bottom
allows accumulation of such material at depths of 200-300 m. Terrigenous deposits,
containing numerous fragments of volcanic glass, are forming at Sts. 172, 209, WS 394
and WS 493 near Deception Island at depths of 168-500 m. The abundance of volcanic
detritus here is only to be expected from Andersson's description of the active volcano
which forms Deception Island,1 "a large crater island, 15-19 km. in diameter. The
crater itself forms a basin, 9-10 km. in width and connected with the ocean by a very
narrow entrance, only about 200 m. wide". Farther south, at St. 175, a sandy deposit
at a depth of 200 m. supports Herdman's2 suggestion that "there appears to be a ridge
between Deception and Tower Islands ". Again, sandy material appears at St. WS 389,
west of Astrolabe Island at a depth of 130 m. ; this station is on the coastal shelf of
Graham Land.

(b) The Palmer Archipelago is separated from Graham Land by the narrow Gerlache
Strait, with an average depth of 600-700 m., which continues westwards by the shal-
lower Bismarck Strait to the Bellingshausen Sea. Five samples of the sea-floor along
this line consist of diatomaceous mud, while terrigenous mud occurs at a depth of
300 m. in Schollaert Channel between Anvers Island and Brabant Island. North of the
last-named island, between it and Low Island, is a wide and shallow depression from

1 Andersson, J. G., 1906. The Geology of Graham Land. Bull. Geol. Inst. Upsala, vn, p. 49.

2 Herdman, H. F. P., 1932. Report on Soundings, etc. Discovery Reports, vi, p. 231.

SEA-FLOOR DEPOSITS. PART I 3o9

which six samples are available. Three stations in the centre of the depression, namely
Sts. WS 395, WS 396 and WS 489, yield samples of diatomaceous mud at depths of
297, 318, and 308 m. respectively. On the southern slope of the depression (St. WS488)
terrigenous mud occurs at a depth of 220 m., while the samples from the northern slope
(Sts. WS 397 and WS 490) are sands from depths of 150 and 262 m. Incidentally,
another sandy deposit occurs at St. WS 399 at a depth of 738 m. on the submarine
extension of the South Shetland ridge. Farther east, between Deception Island and the
Palmer Archipelago (Sts. WS 494 A and WS 494 b) sands occur at depths of 1081 and
505 m. respectively. These stations are on either side of a channel between Deception
Island and Trinity Island, which connects with the open ocean to the north-west. Herdman
states that southward-flowing warm water enters this channel between Smith and Snow
Islands and passes south of Deception Island into the larger basin of Bransfield Strait.
The transport of detrital material by such a current would account for the accumulation
of sandy material in this area, especially if the existence of a ridge south of Deception
Island be substantiated.

An outstanding feature of these deposits from the western part of the South Shetland
region is the occurrence of glaucophane among the detrital minerals. It is most abundant
in samples from Sts. WS 396 and WS 489, and it occurs in less quantity in samples from
Sts. WS 395 and WS 488 to the south, Sts. WS 494 a and WS 494 b to the east, and
Sts. WS 397, WS 490 and WS 399 to the north, while it has also been detected in a
sample from the Scotia Sea much farther north. The provenance of this mineral has yet
to be determined, but the size and distribution of the grains suggests that there may be
a current flowing eastward between Low Island and Brabant Island in addition to the
current already suggested by Herdman between Smith and Snow Islands.

The diatomaceous muds of the South Shetlands region partake of the general cha-
racters described on p. 299. They are generally greenish muds with a considerable pro-
portion of flocculent material enclosing the detrital sand grains. The genera Coscino-
discus, Fragilaria and Cocconeis are most constantly present, as they occur in nearly every
sample examined. At some stations the predominance of these genera is rivalled by
Corethron and Rhizosolenia, which are especially well developed and abundant in the
sample from St. WS 384, for example. Other genera noted in many samples include
Tholassiosiro, Triceratium, Biddidphia and Thalassiothrix, and probably a critical ex-
amination of the diatom assemblages would reveal the presence of others.

THE BELLINGSHAUSEN SEA (Plate XXI)

This region is represented by samples from twenty-one stations. The most northerly
locality (St. WS 501), in Bismarck Strait south of Anvers Island, yields diatomaceous
mud at a depth of 583 m. The deposit is very similar to the diatomaceous mud which is
widely distributed in the Scotia Sea. In the shallower waters around the Biscoe Islands
(Sts. WS 498, WS 499) terrigenous material predominates, though diatom debris still
contributes largely to the bulk of the sediments.

Farther to the south-west a series of eleven soundings taken along a north-west

3I0 DISCOVERY REPORTS

traverse from Adelaide Island shows that the continental shelf extends some 80 miles
from the coast, and that the edge of the shelf is very steep, the depth increasing from
500 to 3000 m. in 15 miles. A profile section (adapted from Herdman1) is shown on
Plate XXI. On the shelf the deposits from Sts. WS 509-5 1 5 have all the characteristics of
diatomaceous mud, and the assemblage of diatoms is similar to that seen in samples
from the Scotia Sea. Near the edge of the shelf (St. WS 516), at a depth of 261 1 m.,
the bulk of the deposit is a coarse sand containing angular grains of quartz and opaque
fragments of rock. Farther to the north-west (St. WS 517) the detrital grains are of
smaller size, and the organic remains become more important so that the deposit must
be classified as diatomaceous mud. The presence of coarse sand at so great a depth on
the continental slope, and at so great a distance from land, raises interesting questions
as to the cause of its occurrence. But speculation must remain in abeyance for the
present, as no other sample from a similar situation is available and generalization is
hardly possible on a single example.

The remaining seven samples are from stations farther south and west, as far as lat.
700 S and long. ioo° W. St. WS 495, north-west of Alexander I Island, yields diatom-
aceous mud of the usual type at a depth of 2582 m. At St. WS 508 the deposit (taken
from a depth of 309 m.) is a coarse sand or fine gravel which may indicate a bottom of
bare rock. The samples from Sts. WS 506 and WS 507 at a depth of about 580 m. can only
be classified as terrigenous mud ; for though flocculent material appears as conspicuous
as the small mineral grains enclosed therein, there are few recognizable organisms. The
same brown unctuous mud occurs farther west at greater depths, namely at Sts. WS 505
(1500 m.) and WS 503 (4073 m.); it is probably typical of the sea-floor in these high
latitudes, for Herdman states that " the average depth of the ocean bed in the Bellings-
hausen Sea varies very little", the average depth being approximately between 3900
and 4400 m. A brown mud from a depth of 4334 m. at St. WS 502, however, is classi-
fied as diatomaceous mud, because recognizable diatoms are present in some quantity,
and their occurrence seems to account for the (presumably organic) flocculent material.
But the abundance of terrigenous material at these great depths and at so great a
distance from land is surprising, and seems to demand some special means of transport.
It may be that much detrital material has been contributed by the melting of icebergs,
as Pirie2 has suggested for the muddy deposits of the Weddell Sea.

THE WESTERN COAST OF SOUTH AMERICA (Plate XXII)
Samples of sea-floor deposits are available from fifty-one stations off the western
coast of South America, between the equator and lat. 400 S, along the course of the
Humboldt current. The majority of the samples are remarkably uniform in character,
thirty-six being classified as diatomaceous mud, eight as terrigenous mud and eight as
sand. The first-named deposit has much the same general constitution as its equivalent

1 Herdman, 1932. Report on Soundings, etc. Discovery Reports, VI, p. 228.

2 Pirie, J. H. H., 1914. Deep-Sea Deposits of the Scottish National Antarctic Expedition. Trans. Roy.
Soc. Edin., xlix, p. 677.

SEA-FLOOR DEPOSITS. PART I 311

in the Scotia Sea, containing a considerable proportion both of mineral grains and of
flocculent diatom debris ; the latter is usually sufficient in quantity to impart a green
colour to the mud. In the South American samples, however, the flocculent aggregates
constantly enclose numerous brown resting spores which presumably belong to various
algae; moreover similar spores are often seen in position within the frustules of
Coscinodiscus and other diatoms. While this peculiarity is undoubtedly of a seasonal
character, it tends to support the previously formed conclusion that the flocculent
material is largely formed of the disorganized cell contents of algae. It is worthy of
note, too, that the assemblage of diatoms in the deposits under discussion differs con-
siderably from that found in the samples from the Scotia Sea. Coscinodiscus is the pre-
dominant genus in both regions, and it is particularly large and abundant in most
samples from South American waters. Here it is constantly accompanied by Actino-
ptychus and Navicula, while Aster ompholns, Bacteriastrum, Entopyla, Grammatophora,
Pleurosigma, Licmophora, Choetoceros and Triceratium are more sporadic in occurrence.
Many of these genera have not been observed in the deposits from the Scotia Sea during
the present investigation. There (it may be recalled) Thalassiosira, Fragilaria, Rhizoso-
lenia and Corethron are (after Coscinodiscus) most consistent in occurrence, but these
genera are virtually absent from the South American assemblages. Another feature of
minor interest is the presence of the widespread silicoflagellate Dictyocha fibida, Ehr.,
in several of the South American samples.

As in other regions, the accumulation of diatomaceous mud shows little relation to
depth, for such deposits are found at the slight depth of 29 m. (St. WS 676), while they
are also present at depths of over 4000 m. (Sts. WS 686, WS 703-705). This bathy-
metrical range is similar to that shown by the terrigenous mud which occurs in a depth
of 369 m. at St. WS 396, and at St. WS 617 in 4864 m. It must be noted, however,
that some of these deposits differ from diatomaceous mud only in the proportion of
flocculent matter and in the paucity of recognizable organic remains. In any case, the
great vertical range of terrigenous material seems to be due to the circumstance that
the continental shelf is very narrow along the western coast of South America, with the
consequence that areas of considerable depth are within the range of currents which
transport detritus from the land surface. The irregular distribution of diatoms and
other organisms is probably due to local differences in chemical and physical conditions.

The mineral content of the diatomaceous muds is, in general, similar to the detrital
material which makes up the bulk of the terrigenous deposits. Quartz is the most
abundant mineral and accounts for the bulk of the detrital grains in all the deposits.
Green hornblende also is of widespread distribution, having been noted in thirty-five
of the fifty-two samples, but brown hornblende is apparently rarer and of sporadic
occurrence. Hypersthene is noted only occasionally in these South American deposits
at three stations (WS 591, WS 596, WS 604) between 320 and 360 S lat. though volcanic
glass occurs in about half the samples. The presence of plagioclase felspar at three
stations probably indicates that the mineral has not always been distinguished by the
writer from other colourless grains ; complete mineralogical analysis of the samples will

3i2 DISCOVERY REPORTS

probably reveal a wider distribution than is shown in the present record. Among other
minerals, white mica is noted at five stations (WS 604, WS 615, WS 617, WS 692 and
WS 697), tourmaline at one locality (St. WS 697) and a yellow mineral which appears
to be staurolite at Sts. WS 604 and WS 605 ; the last record, however, needs confirma-
tion. Finally, the authigenic mineral phillipsite is recorded from two stations (WS 602
and WS 650) which are separated by a considerable distance, at depths of 75 and 143 m.
respectively. The occurrence of phillipsite at such slight depth is unusual, judging by
Murray's statement that "it always occurs in the deeper deposits", but the mineral
appears to have a longer bathymetrical range than that recorded in the literature.

Grateful acknowledgment is made of assistance during the course of this investigation.
The writer has had much helpful discussion with Professor P. G. H. Boswell of the
Royal School of Mines, London, under whose direction the work has been done.
Conversation with colleagues in the University of Liverpool has served to elucidate
many obscure details, while Dr Stanley Kemp and his staff have always been ready to
give information in amplification of the published records. The charts have been drawn
by Miss E. C. Humphreys.

DESCRIPTIONS OF THE SAMPLES

R.R.S. 'DISCOVERY'

Station A. 29. ix. 25. Lat. 470 34' N. Long. 8° 20' W. 4287 m.

Globigerina ooze. A greyish white mud composed of Globigerina fragments ; unbroken tests are
rather scarce. The finer fraction is extremely fine-grained and contains coccoliths in abundance.
Only occasional mineral grains are seen.

Station B. 16. x. 25. Lat. 290 56' 50" N. Long. 150 03' 10" W. 3400 m.
Globigerina ooze. Similar in all respects to sample from Station A.

Station C. 26. xi. 25. Lat. 220 32' 15" S. Long. 160 40' 10" W. 4330 m.

Globigerina ooze. Similar to foregoing samples, but coccoliths and rhabdoliths in the finer
fraction are perhaps even more in eyidence. No mineral grains seen.

Station D. 30. ix. 25. Lat. 260 07' 40" S. Long. 140 36' 20" W. 3195 m. •
Globigerina ooze. A light brown, coherent, granular deposit. Coarse washings consist prin-
cipally of foraminiferal shells of the genera Orbulina, Globigerina and Pulvinulina, with a few frag-
ments of echinid spines. Fine washings contain mainly minute particles, which include coccoliths
and rhabdoliths.

Station E. 30. xi. 25. Lat. 260 17' 40" S. Long. 140 36' 20" W. 3170 m.

Globigerina ooze. Greyish white in colour: foraminiferal tests mostly fragmentary. Coccoliths
and rhabdoliths both plentiful in the finer material.

Station F. 19. ii. 26. Lat. 530 00' S. Long. 340 22' 30" W. 2472 m. (Plate XVII.)

Diatom ooze. A typical and pure diatom ooze, with occasional Radiolaria and a few detrital

mineral grains. The genera Coscinodiscus, Thalassiosira, Fragilaria, Achnanthes, Rhizosolenia and

Thalassiothrix are noted.

SEA-FLOOR DEPOSITS. PART I 313

Station G. 19. ii. 26. Lat. 520 54' S. Long. 34° 05' W. 3469 m. (Plate XVII.)

Diatom ooze. A whitish, coherent, powdery sediment, made up entirely of diatom frustules,
among which large centric forms such as Coscinodiscus and Thalassiosira are most conspicuous;
other forms include navicular and elongate genera, e.g. Fragilaria, Achnanthes, Thalassiothrix and
Rhizosolenia.

Station H. 2. iii. 26. East Cumberland Bay, South Georgia. Lat. 540 15' 15" S. Long. 360 32' W.
229-256 m. (Plate XIX, inset.)

Diatomaceous mud. A grey fine-grained mud, mainly composed of angular to subangular mineral
grains of extremely small size. Some fragments of quartz, however, are more than 01 mm. across.
Some centric diatoms {Coscinodiscus) are present; they are usually rather small but some reach
o-i mm. in diameter.

Station I. April 1926. Off Cumberland Bay, South Georgia. 183 m.

Diatomaceous mud. The deposit contains abundant mineral fragments (quartz and green horn-
blende) ; a few reach 0-2 mm. in diameter but the majority are smaller. Fine-grained flocculent material
contains centric diatoms (chiefly Coscinodiscus), many of which retain their greenish protoplasmic
contents; finer particles may be comminuted diatom tests. The greenish colour of the deposit is
apparently due to chlorophyll, for the preservative liquid is coloured green.

Station J. 21. v. 26. Lat. 490 32' 30" S. Long. 550 28' 30" W. 892 m. (Plate XVII.)
Terrigenous mud. This rather sandy grey mud consists mainly of mineral grains among which
some reach a diameter of 0-2 mm., but most are below diameter of 0-05 mm. Grains of quartz and
green hornblende are abundant. There is some admixture of diatoms and sponge spicules.
Station K. 25. ix. 26. Lat. 330 05' S. Long. 1 6° 18' E. 3260 m. (Plate XVII, inset.)
Globigerina ooze of same type as foregoing samples. The finer material consists largely of com-
minuted foraminiferal shells, coccoliths and rhabdoliths.

Station L. 28. ix. 26. Lat. 330 19' S. Long. 170 40' E. 235 m. (Plate XVII, inset.)
Glauconitic mud. This is a detrital sediment consisting mainly of clean quartz grains, apparently
evenly graded at a diameter of about 0-05 mm. Some glauconite is present as dark green grains while
dark opaque grains, slightly larger, occur in some abundance. Flocculent matter also occurs.
Station M. 29. ix. 26. Lat. 330 21' S. Long. 160 47' E. 2020 m. (Plate XVII, inset.)
Globigerina ooze. Mainly broken tests of Globigerina. Little detrital matter if any. Shields of
calcareous flagellates (especially coccoliths) are plentiful among the finer particles.
Station N. 26. ii. 27. Lat. 620 57' S. Long. 6o° 21' 30" W. 967 m. (Plate XX.)
Diatomaceous mud. The sample consists mainly of greenish flocculent material which is ap-
parently diatomaceous in origin. Besides frustules of Rhizosolenia, Fragilaria and Coscinodiscus it
contains small sand grains with a diameter of usually less than 0-05 mm., among which quartz and
volcanic glass are noted.

Station O. 19. iii. 27. Lat. 640 56' S. Long. 640 43' W. 435 m. (Plate XX.)

Diatomaceous mud. A light grey mud with abundant grains of quartz and green hornblende ;
some of the grains are 0-5 mm. across, but the majority are less than 0-05 mm. in diameter. Organic
remains form a subordinate proportion of the sample ; they are mainly broken diatom tests (but some
whole frustules of Coscinodiscus) and sponge spicules.

Station P. 21. xii. 26. Drygalski Fjord, South Georgia; about i-ii miles from glacier at end of
fjord. 178-3 m. (Plate XIX.)

Diatomaceous mud. A greyish mud, containing roughly equal proportions of small sand grains
and flocculent material. The sand grains, generally less than 0-05 mm. in diameter, are mainly
fragments of quartz, but green hornblende is also plentiful. The flocculent matter is largely diatom-
aceous ; there are some unbroken centric frustules.

3-2

3i4 DISCOVERY REPORTS

Station 8. 9. ii. 26. Lat. 420 36' 300 S. Long. 18° 19' 30" W. 3450 m. (Plate XVII.)
Globigerina ooze. Light grey, coherent and granular. The larger constituents are mainly shells

of Globigerina, up to 0-2 mm. in diameter. Finer particles extremely small, chiefly coccoliths, some

of which occur in groups ; rhabdoliths seem to be scarcer.

Station 9. 11. ii. 26. Lat. 460 09' S. Long. 220 26' W. 2226 m. (Plate XVII.)

Globigerina ooze. Brownish, coherent and granular. Coarser material consists of broken

Globigerina tests, with a few mineral grains. The finer particles consist largely of coccoliths in some

variety; a few more or less complete coccospheres are noted.

Station 10. 13. ii. 26. Lat. 460 35' S. Long. 240 15' 30" W. 4402 m. (Plate XVII.)
Diatom ooze. A grey-brown, coherent mud. The bulk of the deposit is formed of diatom frustules
among which large centric forms (Coscinodiscus) are conspicuous; navicular and elongate forms,
including Fragilaria, Achnanthes, Rhizosolenia and Thalassiothrix, are also present. There is much
comminuted material, together with some detrital minerals and occasional nasselarian Radiolaria.
The presence of the last-named probably accounts for the description of this sample as " radiolarian
ooze " in the official Station List, but the deposit is undoubtedly diatom ooze.

Station 11. 16. ii. 26. Lat. 500 26' S. Long. 300 27' W. 5000 m. (Plate XVII.)
Diatom ooze. A grey-brown coherent sediment, becoming light grey on drying. The bulk of the
sample consists of the usual diatom assemblage in which frustules of Coscinodiscus, Biddulphia,
Achnanthes, Fragilaria, Rhizosolenia and Thalassiothrix are prominent, with a few nasselarian
Radiolaria. Some small, angular quartz grains (up to o-i mm. diameter) and fragments of green
hornblende are present, but the sample is too small for mineral analysis.

Station 12. 18. ii. 26. Lat. 510 55' 05" S. Long. n° 44' W. 2744 m. (Plate XVII.)

Diatom ooze. A grey, powdery sediment containing a variety of diatoms, including Coscinodiscus,
Fragilaria, Rhizosolenia, Achnanthes, Thalassiothrix and Biddulphia, with some sponge spicules and
small mineral grains.

Station 13. 3. iii. 26. 57 miles N 49^° E of Jason Light, South Georgia. 143 m. (Plate XIX,
inset.)

Diatomaceous mud. A greenish grey " buttery " clay. The bulk of the deposit consists of mineral
grains, usually very small, but quartz grains occasionally reach a diameter of 0-2 mm., and prisms of
green hornblende a length of 0-05 mm. Much comminuted diatom material is present as well as
recognizable centric and navicular diatoms and sponge spicules.

Station 14. 3. iii. 26. 15-4 miles N 44J-0 E of Jason Light, South Georgia. 260 m. (Plate XIX,
inset.)

Diatomaceous mud. The deposit has a "buttery" consistency and is mainly composed of finely
comminuted diatom material. Some frustules of Coscinodiscus are large, but most are less than
o-i mm. across; tests of Fragilaria and Thalassiothrix, as well as sponge spicules are present. Some
rounded quartz grains reach a diameter of 0-2 mm., but the average is below 0-05 mm. Grains of
brown hornblende are about 0-05 mm. long. There is apparently no mineral to account for the
greenish hue of the mud, which may be due to a vegetable pigment such as chlorophyll.

Station 15. 3. iii. 26. 25 miles N45|°E of Jason Light, South Georgia. 191m. (Plate XIX, inset.)

Diatomaceous mud. Similar to samples from Sts. 13 and 14; mineral grains reach 01 mm. in
diameter.

Station 16. 3. iii. 26. 36-5 miles N 460 E of Jason Light, South Georgia. 727 m. (Plate XIX.)

Diatomaceous mud. The sediment has separated into layers during storage, dark grey below,
light brown above. The finer particles are mainly diatoms, Coscinodiscus and Fragilaria being con-
spicuous forms; sponge spicules and fragments of Radiolaria are also present. Quartz grains are
abundant and large, many being over 01 mm. in diameter; they are angular to subangular in shape.
Grains of green and brown hornblende (o-i mm.) and occasional splinters of volcanic glass are noted.

SEA-FLOOR DEPOSITS. PART I 315

Station 19. 4. iii. 26. 10 miles N 390 E of Cape Saunders, South Georgia. 200 m. (Plate XIX,
inset.)

Diatomaceous mud. This dark, greenish sandy sediment approaches a diatom ooze in the
abundance of diatoms. Coscinodiscus is the most conspicuous genus, but valves of Fragilaria are
also present with occasional sponge spicules and radiolarian fragments. Quartz grains reach 0-2 mm.
in diameter, and prisms of green hornblende 0-15 mm. in length, but most of the grains are smaller.
Some amount of greenish flocculent material is present.

Station 20. 4. iii. 26. 14-6 miles N 41° E of Cape Saunders, South Georgia. 210 m. (Plate XIX,
inset.)

Diatomaceous mud. This sediment is fine-grained and greenish grey in colour. The bulk of the
sample consists of diatoms {Coscinodiscus, Fragilaria and other forms) and exceedingly fine particles
of comminuted diatoms. Mineral grains include quartz (up to 0-2 mm., abundant) and green horn-
blende (o-i mm.). Sponge spicules also occur. The greenish colour of the sediment is probably due
to chlorophyll or other pigment of the diatoms, for the preservative liquid has acquired a green
colour.

Station 23. 14. iii. 26. 5-3 miles N 440 E of Merton Rock, South Georgia. 228 m. (Plate XIX,
inset.)

Diatomaceous mud. This greenish grey "buttery" deposit contains much detrital material,
chiefly subangular quartz grains which range down from a diameter of about 0-05 mm. to exceedingly
small particles. Flakes of white mica, however, attain a diameter up to o-i mm. The organic material
is chiefly diatomaceous. Frustules of Coscinodiscus are large and abundant; they are accompanied by
Fragilaria, Thalassiosira and other forms. Some sponge spicules are noted. The greenish hue is
thought to be due to vegetable pigments.

Station 28. 16. iii. 26. West Cumberland Bay, South Georgia. 3-3 miles S 450 W of Jason Light.
Two samples from depths of 65 and 168 m. respectively. (Plate XIX, inset.)

Terrigenous mud. (a) The sample from 65 m. is a grey mud formed of exceedingly fine mineral
particles which react to polarized light. There is a fair quantity of subangular detrital grains and
centric diatoms which reach a diameter of o-i mm.

(6) The sample from 168 m. is a black mud, microscopically similar to sample (a), but yellow oil
globules are commonly seen. Apart from the diatoms the material is remarkably like some of the
black muds of the Mersey estuary in England.

Station 29. 16. iii. 26. West Cumberland Bay, South Georgia. 5-9 miles Ssi°W of Jason
Light. 23 m. (Plate XIX, inset.)

Terrigenous mud. A sediment of "buttery" consistency. The top layers of the sample have
oxidized slightly, giving a yellowish tinge. Though some particles reach a diameter of o-i mm., the
bulk of the material is extremely fine-grained, most of the particles being below o-oi mm. in diameter.
Some of the fine particles react to polarized light, others appear not to do so. No organic material
was recognized definitely.

Station 30. 16. iii. 26. West Cumberland Bay, South Georgia. 2-8 miles S 240 W of Jason
Light. 251 m. (Plate XIX, inset.)

Terrigenous mud. Similar to sample from St. 29 in general constitution, but perhaps more of the
coarser particles. Some small frustules of Coscinodiscus from o-oi to 0-05 mm. in diameter are noted.

Station 31. 17. iii. 26. 13-5 miles N 890 E of Jason Light, South Georgia. 220 m. (Plate XIX,
inset.)

Diatomaceous mud. The bulk of the sample consists of diatomaceous remains, mostly com-
minuted, but containing some frustules of Coscinodiscus, Fragilaria and Thalassiothrix. The centric
forms are often fairly large, but much of the comminuted material is below o-oi mm. in diameter.
Sponge spicules and green filamentous algae are also present. Many quartz grains reach a diameter

3i6 DISCOVERY REPORTS

of o- 1 mm. ; green hornblende (005 mm.) is also noted. The greenish tinge of the sediment, and the
acquisition of a green colour by the preserving liquid is probably caused by the presence of vegetable
pigments.

Station 41. 28. iii. 26. i6J miles N 39° E of Banff Point, South Georgia. 272m. (Plate XIX, inset.)

Diatomaceotjs mud. A slightly greenish mud, the bulk of which consists of diatom frustules,
mostly fragmentary but with some entire frustules of Coscinodiscus, Fragilaria and other forms ; the
centric forms are often especially well-developed. Some grains of quartz and green hornblende are
noted. A quantity of greenish flocculent material is present.

Station 42. 1. iv. 26. Off mouth of Cumberland Bay, South Georgia. From 6-3 miles N 890 E
to 4 miles N 390 E of Jason Light. 204 m. (Plate XIX, inset.)

Diatomaceous mud. This sample consists mostly of finely comminuted diatomaceous material.
Some of the centric forms are large, but the majority are below o-oi mm. in diameter. The navicular
and elongate diatoms are small. Sponge spicules are present. The quartz grains, 0-1-0-05 mm. in
diameter, are mainly angular in form.

Station 48. 3. v. 26. 8-3 miles N 530 E of William Point Beacon, Port William, Falkland Islands.
Two samples from 105 m. (shoot) and 115 m. (haul). (Plate XVIII.)

Shelly sand. Both samples are similar in constitution. They are speckled grey and white sands,
consisting mainly of clean white quartz in subangular to rounded grains, with occasional rounded
crystals of hypersthene, red garnet, and brown tourmaline. The organic remains include many
Foraminifera (Polystomella and other rotalids) up to 1 mm. in diameter, echinid spines (mostly
broken and some rolled), a few young gastropods and a quantity of broken molluscan shells.

Station 51. 4. v. 26. Off Eddystone Rock, East Falkland Island. 7 miles N 500 E, to 7-6 miles
N 630 E of Eddystone Rock. 115 m. (Plate XVIII.)

Fine sand. The bulk of the sand grains are less than 0-5 mm. in diameter but some grains are
larger; they are mainly subangular fragments of quartz, with occasional rounded grains of hypers-
thene. There are numerous shells and small pebbles, while among the finer constituents are Fora-
minifera, echinid spines and shell fragments.

Station 63. 22. v. 26. Lat. 480 50' S. Long. 530 56' W. Depth not recorded. (Plate XVII.)
Diatomaceous mud. The coarser fraction of this greenish mud consists of sand grains up to
about o-i mm. in diameter; they are subangular to rounded in shape. The finer material also con-
tains some mineral grains (perhaps more angular than the coarser grains), but diatomaceous debris
forms the greater bulk. Diatoms are present in variety and include the genera Coscinodiscus,
Thalassiosira, Cocconeis, Fragilaria, Thalassiothrix and Rhizosolenia. A considerable proportion of
the mud consists of flocculent aggregates formed of disintegrated diatom frustules and their cell
contents.

Station 64. 22. v. 26. Lat. 480 34' S. Long. 530 34' 30" W. 4136 m. (Plate XVII.)
Globigerina ooze. The very small sample of sandy deposit contains entire examples of Globi-
gerina and rotalines, but also much comminuted material. Many glauconite grains occur, some dark
green but the majority show a yellowish green colour. Mineral grains, mostly rounded quartz frag-
ments, reach a maximum diameter of 0-5 mm. This is obviously not a typical Globigerina ooze, but
a mixed deposit which occurs at the margin of the Globigerina belt.

Station 71. 30. v. 26. Lat. 430 20' S. Long. 46° 02' W. 5460 m. (Plate XVII.)

Radiolarian ooze. A brown unctuous mud. A large part of the material consists of extremely

small particles, which form a flocculent "matrix". Quartz grains are dispersed through the deposit,

and some reach a diameter of o-i mm. The coarse residue after washing shows plentiful Radiolaria.

Centric and navicular diatoms are also present. The sample is similar to the deposits from Sts. 74

and 77 which are discussed in greater detail.

SEA-FLOOR DEPOSITS. PART I 317

Station 74. 3. vi. 26. Lat. 400 31' 40" S. Long. 380 14' 50" W. 5446 m. (Plate XVII.) Three
samples: upper 3 cm., middle 3 cm., and bottom 3 cm., from a core 24 cm. in length.

Radiolarian ooze. An unctuous brown mud, mainly composed of extremely small particles
which together form a flocculent mass. This contains some centric and navicular diatoms, sponge
spicules and, more abundantly, Radiolaria. Tests of the latter group are plentiful in coarser washings
from the top section of the sample. They are mainly spheroid and discoid forms, such as Ceno-
sphaera, Hexastylus, Carposp/iaera, Heliodiscus, Porodiscus, and Rhopalastrtem. Nasselarian forms are
rare and small in size. Radiolaria are not abundant in the meagre washings yielded by the middle
and lower sections of the sample.

Mineral grains are common in the upper section, with diameters up to about o-i mm. They are
mainly angular grains of quartz, but some coloured minerals occur also. On the whole, mineral
grains appear to decrease in quantity downwards in the core.

Station 77. 6. vi. 26. Lat. 390 19' 30" S. Long. 350 27' 40" W. 5186 m. (Plate XVII.)

Radiolarian ooze. The sample consists of a core 47 cm. long. In general constitution the sample
is a brown mud which becomes pale on drying. A considerable part of the deposit is composed of
flocculent material ; this consists of extremely small particles most of which appear to be isotropic
in polarized light, though others are certainly anisotropic. Mingled with this material are mineral
grains, Radiolaria and a few other organic remains (such as diatoms), which vary proportionately in
different parts of the core.

The top part of the core is extremely rich in Radiolaria, which show a great variety of form.
Speaking generally, the spumellarian forms are abundant, but nasselarians are comparatively rare in
the residue examined. Of the spumellarians, the spheroid genera Cenosphaera, Carposphaera, and
Hexastylus, and the discoid genera Heliodiscus and Porodiscus are specially abundant and well de-
veloped, while Rhopalastrum and Hymeniastrum are both rare and smaller. Among the few nasse-
larians, the genera Sethopyramis, Lychnocanium and Clathrocyclas are noted. Only a few small
angular mineral grains appear in the residue.

The middle section yields a much smaller residue consisting mainly of Radiolaria. The spheroid
and discoid genera are predominant, and similar to those mentioned above; likewise, nasselarian
forms are rare.

The lower section contains a larger proportion of mineral grains, which vary in diameter up to
o-i mm., but some are even larger. Radiolaria are present, especially spheroid and discoid forms,
but they are less plentiful than in the higher sections of the core.

Station 78. 12. vi. 26. Lat. 350 18' S. Long. 190 01' 10" W. 3410 m. (Plate XVII.)
Globigerina ooze consisting mainly of extremely fine particles, many of which react to polarized
light; among these, coccoliths and rhabdoliths are plentiful. Large tests (up to 0-5 mm.) of Globi-
gerina occur, but no mineral grains were seen.

Station 83. 21. vi. 26. Lat. 320 31' 50" S. Long. i° 23' 30" W. 4308 m. (Plate XVII.)
Globigerina ooze. A pale ooze formed mainly of comminuted Globigerina tests, but some whole

shells reach 0-5 mm. in diameter. Some broken echinid spines occur, but no mineral grains are noted.

Coccoliths and rhabdoliths occur plentifully in the finer fraction.

Station 84. 22. vi. 26. Lat. 320 52' S. Long. i° 55' E. 2233 m. (Plate XVII.)
Globigerina ooze. Large tests of Globigerina with much comminuted material, coccoliths and
rhabdoliths in the finer fraction. No mineral grains were seen.

Station 89. 28. vi. 26. Lat. 340 05' 15" S. Long. 160 00' 45" E. 3926 m. (Plate XVII, inset.)

Globigerina ooze. The sample consists of a core, 30 cm. long. The deposit is closely similar to

the samples from Sts. 83 and 84, but there is a slight proportion of small mineral grains (up to

0-05 mm. diameter). Otherwise, there are the usual Globigerina tests up to 0-2 mm. across and a

considerable proportion of fine material in which coccoliths, rhabdoliths, sponge spicules and

3IS DISCOVERY REPORTS

foraminiferal fragments can be recognized. No appreciable differences can be detected in slides
made from various parts of the core.

Station 93. 23. ix. 26. Lat. 330 08' S. Long. 170 50' E. 165 m. (Plate XVII, inset.)
Glauconitic mud. The sample is largely composed of indefinite flocculent material which encloses
particles of foraminiferal shells, coccoliths and fragments of sponge spicules. Held in this flocculent
mass are various Foraminifera (including Globigerina with a diameter up to 0-3 mm.) and mineral
grains. The latter are chiefly colourless angular quartz fragments reaching a diameter of 0-2 mm.,
but occasional grains of greenish material appear to be glauconite.

Station 94. 23. ix. 26. Lat. 330 18' S. Long, 170 40' E. 281m. (Plate XVII, inset.)
Glauconitic mud. This sample is mainly composed of rounded and subangular mineral grains
averaging about 0-05 mm. in diameter. The grains are chiefly quartz, but some coloured minerals are
present. Deep green glauconite occurs in rounded grains. The organic material consists mainly of
echinid spines and sponge spicules. The only Foraminifera appear to be a few small (dwarf or im-
mature) specimens of Globigerina. There is little flocculent material.

Station 95. 23-24. ix. 26. Lat. 33°3i'S. Long. 17° 29' E. 440 m. (Plate XVII, inset.)
Glauconitic mud. The sample is much paler in colour than the two preceding (Sts. 93 and 94).
It consists mainly of foraminiferal tests (up to 0-2 mm. diameter), most of which are broken and
apparently abraded. Sponge spicules also occur. Mineral grains include subangular fragments of
quartz (o-i mm.) and deep green glauconite (0-5 mm.). There is some quantity of finely divided
flocculent material in which coccoliths are plentiful.

This is hardly a typical glauconitic mud, as the foraminiferal tests occur in larger proportion than
usual, but they have the appearance of drifted material.

Station 96. 24. ix. 26. Lat. 330 06' S. Long. 170 01' E. 620 m. (Plate XVII, inset.)
Glauconitic mud. In the official Station List this sample is described as muddy Globigerina ooze.
It is essentially a Globigerina ooze in which many of the tests are infilled with deep green glauconite.
Some grains are seen in situ, others are rounded and compound, clearly internal moulds of Globi-
gerina shells. The finer fraction includes coccoliths among comminuted shells. The deposit is here
classified on the occurrence of glauconite.

Station 97. 24. ix. 26. Lat. 33°n'S. Long. 160 55' 30" E. 995 m. (Plate XVII, inset.)
Glauconitic mud. A pale deposit consisting largely of Globigerina and other Foraminifera; the
tests are often broken. The finely divided material contains numbers of tiny calcite rods which seem
to be derived by comminution from foraminiferal tests. Sponge spicules and coccoliths also occur.
The sample is described as Globigerina ooze in the Station List, but the quantity of mineral grains
is greater than one would expect in an ooze. Angular grains of quartz reach a diameter of o-i mm.,
and some glauconite grains have a similar size.

In common with samples from Sts. 95 and 96, this deposit is typical neither of glauconitic mud
nor of Globigerina ooze. Reference to the map shows that the three stations lie on the border-line
between the great region of oceanic Globigerina ooze to the west, and the area of terrigenous deposits
on the landward side. The mixed character of the sediments is thus explained.

Station 98. 25. ix. 26. Lat. 330 23' S. Long. 150 50' E. 3640 m. (Plate XVII, inset.)
Globigerina ooze. A white, extremely fine-grained ooze, consisting almost entirely of frag-
mentary foraminiferal tests with some quantity of small unbroken shells of Globigerina and other
genera. Among the finest comminuted material, coccoliths and rhabdoliths are abundant.

Station 101. 14. x. 26. Lat. 330 50' to 340 13' S. Long. 160 04' to 150 49' E. 3734 m.
(Plate XVII, inset.)

Globigerina ooze. A small sample consisting of large tests of Globigerina (up to 0-3 mm. diameter)
with comminuted material, some of which is exceedingly fine grained. There are a few fragmentary

SEA-FLOOR.DEPOSITS. PARTI 319

sponge spicules and occasional mineral grains, while coccoliths and rhabdoliths occur in the finer
material.

Station 102. 28. x. 26. Lat. 350 29' 20" S. Long. 1 8° 33' 40" E. 1800 m. (Plate XVII, inset.)
Globigerina ooze. A grey mud, rather pale when dry. The deposit consists mainly of minute
particles (including coccoliths) which form flocculent patches on the glass slip. Foraminiferal shells
are fairly abundant, mostly those of Globigerina. Sponge spicules also occur. There is a fair propor-
tion of mineral grains, mostly rounded or subangular, up to o-i mm. in diameter. While most of the
mineral grains are colourless, some rounded green grains, which show aggregate polarization, may
be glauconite. Though classified as a Globigerina ooze, this deposit is not typical of that group ; the
proportion of mineral grains is too high and that of Foraminifera is correspondingly small. The
station is situated on the margin of a region marked by the occurrence of glauconitic deposits.

Station 117. 17. xi. 26. Lat. 520 20' 40" S. Long. 30 48' 45" E. 1723 m. (Plate XVII.)

Diatomaceous mud. A small sample, brownish grey when wet, from "about 5 miles N 720 E of
Bouvet Island ". The deposit consists mainly of diatom tests, among which Coscinodiscus, Fragilaria
and Thalassiothrix are most abundant, the first named attaining a diameter of 0-25 mm. There is a
fair quantity of mineral grains, chiefly angular and rounded quartz grains up to 0-2 mm. in diameter ;
prismatic fragments of green hornblende, up to 0-25 mm. in length, are also present. The proportion
of mineral material determines the classification of this deposit as diatomaceous mud rather than
diatom ooze.

Station 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia. From 4-1 miles N54°E
of Larsen Point to 1-2 miles S 620 W of Merton Rock. 250 m. (Plate XIX, inset.)

Diatomaceous mud. An extremely fine-grained mud, with entire tests of centric and navicular
diatoms, the diameter of which is mainly below o-i mm. The fine material seems to consist in great
part of comminuted diatoms. Quartz grains are present in fair quantity, but are mainly less than
0-05 mm. in diameter; the occasional grains of green hornblende are also small.

Station 129. 19. xii. 26. Lat. 530 28' 30" S. Long. 370 08' W. 1001 m. (Plate XIX.)
Diatomaceous mud. A very small sample which has settled out into black and light-coloured
layers, representing the approximate separation of detrital and organic constituents. The organic
material contains centric and navicular diatoms, the former reaching a diameter of 0-2 mm. Elongate
diatoms and sponge spicules are present in some abundance and occasional Radiolaria are noted.
Much light flocculent material consists entirely of fragmentary diatom tests. Mineral grains, reaching
a diameter of o-i mm., are present in fair quantity. They are mostly rounded and subangular frag-
ments of quartz, but blue-green prismatic grains of hornblende and some opaque grains occur.
No stones are present in the sample, though these are recorded in the Station List.

Station 131. 20. xii. 26. Lat. 530 59' 30" S. Long. 360 11' W. 240 m. (Plate XIX, inset.)
Diatomaceous mud. The material consists largely of diatoms in some variety, among which
Fragilaria and Coscinodiscus are predominant, the latter reaching 0-2 mm. in diameter. Sponge
spicules are also present. There is some quantity of mineral grains which rarely attain o-i mm. in
diameter; they are mainly rounded colourless quartz fragments, but prismatic green hornblende is
noted. Stones are recorded in the Station List but there are none in the sample.
Station 136. 21. xii. 26. Lat. 540 22' S. Long. 350 21' W. 246 m. (Plate XIX.)
Glauconitic mud. The sample is a light grey mud when dry, but is greenish when wet. No
stones are present. The bulk of the material consists of mineral grains reaching a diameter of o-2inm.,
though the average size is much less. There is probably a fair variety of minerals, among which
green hornblende is conspicuous, but the large majority of grains are subangular fragments of quartz.
Deep green glauconite is present, sometimes filling the canals of sponge spicules. The finer material
includes diatom frustules, mainly fragmentary, but whole tests of Coscinodiscus and Fragilaria are
present. The apparent absence of calcareous tests is noteworthy, in view of their constant association
with glauconite elsewhere.

32o DISCOVERY REPORTS

Station 145. 7. i. 27. Stromness Harbour, South Georgia; between Grass Island and Tonsberg
Point. 26-35 m. (Plate XIX, inset.)

Diatomaceous mud. The coarser constituents include a variety of diatoms, chiefly Coscinodiscus
(up to 0-2 mm. in diameter), Fragilaria, Cocconeis and Licmophora with occasional examples of
Pleurostgma. Simple and branched sponge spicules, amphipod fragments and sea-weeds were also
noted. Mineral grains, mainly quartz fragments, reach 0-2 mm. in diameter, but the average size is
much smaller. The finest material consists of mineral particles together with comminuted diatoms,
and other organic debris.

Station 151. 16. i. 27. Lat. 530 25' S. Long. 350 15' W. 3200 m. (Plate XIX.)
Diatomaceous mud. The organic constituent is almost entirely diatomaceous, frustules of
Coscinodiscus, Fragilaria and Thalassiothrix being abundantly represented. Occasionally Radiolaria
and sponge spicules are seen. Mineral grains up to o-2 mm. in diameter occur in some quantity ; the
size of the grains is larger than would be expected at this depth and locality.

Station 157. 20. i. 27. Lat. 530 51' S. Long. 360 11' 15" W. 970 m. (Plate XIX, inset.)
Diatomaceous mud. Stones are reported for this deposit in the Station List but none is present
in the small sample. The organic constituents consist of the usual diatomaceous assemblage
(Coscinodiscus, Fragilaria and Thalassiothrix) together with sponge spicules. The numerous mineral
grains are rounded and subangular in shape, and reach 0-2 mm. in diameter.

Station 160. 7. ii. 27. Lat. 53°43'4o" S. Long. 400 57' W, near Shag Rocks. 177 m. (Plate XVII.)
Sand. This is a very " mixed " deposit, described in the official Station List as " Grey mud with
stones and rock ". One stone, 8 mm. in diameter, is present in the small sample available. The de-
posit consists of mineral grains and rock fragments up to 1 mm. in diameter, Globigerina and other
Foraminifera, together with exceedingly fine-grained, supernatant material which includes the
diatoms, Coscinodiscus, T/ialassiosira, Fragilaria, Rkizosolenia.

Station 162. 17. ii. 27. Lat. 6o° 48' S. Long. 460 08' W. 320 m. (Plate XVII.)
Diatomaceous mud. A sample obtained off Signy Island, South Orkneys, and described as "green
mud " in the official Station List. The sediment is composed largely of fine-grained material which
appears to be comminuted diatoms, and there are many whole frustules of Coscinodiscus, Fragilaria
and Rhizosolenia. Mineral grains are generally less than 0-05 mm. in diameter and most of them are
angular fragments of quartz, but prismatic grains of green hornblende and flakes of white mica also
occur.

Station 167. 20. ii. 27. Lat. 6o° 50' 30" S. Long. 460 15' W. 244-344 m. (Plate XVII.)
Diatomaceous mud. A fine-grained sediment consisting largely of mineral grains, only a few of
which attain a diameter of 0-2 mm. Quartz, hornblende and white mica are noted. Some sponge
spicules are present, and also frustules of Coscinodiscus and Fragilaria. The finer fraction of the sedi-
ment appears to consist of comminuted diatom frustules.

Station 169. 22. ii. 27. Lat. 6o° 48' 50" S. Long. 5 1° 00' 20" W. 2514 m. (Plate XVII.)
Diatomaceous mud. The sample consists of two sections of core, in all about 40 cm. long. Various
parts of the core have been examined microscopically, but no appreciable differences have been
detected throughout its length. This dark, greenish grey mud consists essentially of diatom frustules
in considerable variety. Centric forms, especially Coscinodiscus and Thalassiosira, are conspicuous
by their size and abundance, while Biddulp/iia is less numerous ; navicular genera are represented by
Fragilaria and Achnanthes; and among elongate forms, Rhizosolenia is both abundant and well-
developed, while Corethron is smaller and less plentiful. There is a quantity of comminuted frustules,
the particles of which adhere and form flocculent groups. Mineral grains are present in fair quantity,
and reach a diameter of about 01 mm. They include grains of quartz and green hornblende and
flakes of white mica. The deposit might almost be called a diatom ooze, but the proportion of mineral
grains, though not great, is a real distinction between this sample and that from St. 12, for instance.

SEA-FLOOR DEPOSITS. PART I 321

Station 171. 25. ii. 27. Lat. 620 07' S. Long. 570 03' W. 16 miles off Cape Melville, King
George Island, South Shetlands. 1542 m. (Plate XX.)

Diatomaceous mud. This brownish grey mud contains abundant mineral grains, many ex-
ceedingly small and the great majority less than o- 1 mm. in diameter. Besides the quartz grains,
prismatic fragments of green hornblende and splinters of brown volcanic glass are noted. There is a
large proportion of indeterminate flocculent material which contains small diatoms (whole and
fragmentary frustules of Coscinodiscus, Fragilaria and Rhizosolenia) together with sponge spicules.

Station 172. 26. ii. 27. Lat. 620 59' S. Long. 6o° 28'W. Off Deception Island, South Shetlands.

525 m. (Plate XX.)

Gravel. The sample consists of black pebbles, presumably of volcanic rock, on some of which are

adherent polyzoa.

Station 175. 2. iii. 27. Lat. 630 17' 20" S. Long. 590 48' 15" W. Bransfield Strait, South Shet-
lands. 200 m. (Plate XX.)

Sand. The sample consists of a small quantity of sand. The particles are mainly coarse, the larger
grains reaching a diameter of 0-5 mm., but some of the material is finer and shows a variety of
minerals. Quartz is most abundant, while hornblende and volcanic glass are noted among the coloured
grains. No organic remains were seen.

Station 180. 11. iii. 27. 17 miles west of northern point of Gand Island, Palmer Archipelago.
160 m. (Plate XX.)

Terrigenous mud. This pale grey mud consists almost entirely of detrital mineral grains, chiefly
quartz, but green hornblende (up to o-2 mm. diameter) and occasional splinters of volcanic glass are
conspicuous. Only occasional diatoms and sponge spicules are seen.

Station 185. 16. iii. 27. Gerlache Strait, Palmer Archipelago. 3-5 miles S 1190 E of Cape van
Wycks, Anvers Island. 598 m. (Plate XX.)

Diatomaceous mud. Roughly half the bulk of this brownish grey mud consists of fairly angular
detrital grains which reach a diameter of o-i mm. Occasionally frustules of Coscinodiscus are
larger than this. The other half consists of flocculent material which includes tiny mineral grains, small
diatoms (Fragilaria, etc.), fragments of sponge spicules and indeterminate debris.

Station 187. 18. iii. 27. Lat. 640 48' 30" S. Long. 630 31' 30" W. Neumayr Channel, Palmer
Archipelago. Two samples from depths of 200 and 259 m. respectively. (Plate XX.)

Diatomaceous mud. The sample from 200 m. is a brownish mud consisting of flocculent material
with some mineral grains whose diameter is usually less than 0-05 mm. Occasional navicular diatoms
and sponge spicules are present.

The sample from 259 m. is a grey mud, largely of detrital origin, with mineral grains reaching
more than 02 mm. in diameter and consisting chiefly of quartz and green hornblende, together with
fragments of volcanic glass. Centric diatoms (about 0-05 mm. diameter), occasional Foraminifera
and sponge spicules are noted.

Station 191. 25. iii. 27. Gerlache Strait, Palmer Archipelago. 2-5 miles Nii4°E of Cape
Astrup, Wiencke Island. 310 m. (Plate XX.)

Diatomaceous mud. A light grey mud with abundant detrital minerals, some of which are
0-2 mm. in diameter, including grains of green hornblende, some of which are idiomorphic. These
are enclosed in flocculent material, apparently organic debris, which also contains small diatoms,
both centric and navicular forms, but these are not abundant.

Station 192. 27. iii. 27. Lat. 640 14' S. Long. 6i° 49' W. Off Cape Kaiser, Brabant Island,
Palmer Archipelago. 800 m. (Plate XX.)

Diatomaceous mud. This brownish grey mud is mainly a mass of flocculent material, apparently
organic debris, which encloses small centric and navicular diatoms (0-02 mm.) together with com-
minuted frustules. Some of the material may be mineral grains, but determination of these tiny

322 DISCOVERY REPORTS

fragments is difficult. Occasionally larger grains (o-i mm. diameter) of quartz and volcanic glass
are seen.

Station 194. 28. iii. 27. Lat. 620 57' 30" S. Long. 6o° 22' W., z\ miles east of Deception Island,
South Shetlands. 812 m. (Plate XX.)

Diatomaceous mud. A brownish grey mud containing numerous mineral grains, some of which
reach a diameter of 0-5 mm., though the average size is much less. The largest are fragments of brown
volcanic glass (0-5 mm.), but angular and subangular grains of quartz and green hornblende are
also noted. Organic remains are not conspicuous, though frustules of Coscinodiscus and Fragilaria,
together with sponge spicules, do occur. The amount of flocculent material is subordinate.

Station 195. 30. iii. 27. Lat. 620 07' S. Long. 580 28' 30" W. Admiralty Bay, King George
Island, South Shetlands. 391 m. (Plate XX.)

Sand. This sample is a non-graded sand, of which the largest grains have a diameter of about
0-25 mm. ; they are chiefly rounded and subangular grains of quartz, felspar and green hornblende.
The only muddy material has formed a coherent film on the surface in the jar ; it shows the same
abundance of detrital minerals in smaller fragments. Organic remains are scarce and comprise a
few sponge spicules and some small diatom frustules (Coscinodiscus and Fragilaria).

Station 196. 3. iv. 27. Lat. 620 17' 30" S. Long. 58° 21' W. Bransfield Strait, South Shetlands.
Two samples from depths of 720 and ion m. respectively. (Plate XX.)

Terrigenous mud. Both samples are brownish grey muds, consisting of homogeneous flocculent
material in which small mineral grains are embedded. Occasionally, larger grains (0-2 mm.) of
quartz and green hornblende are seen. A few diatoms {Coscinodiscus, Fragilaria and Rhizosolenia)
and sponge spicules are the only organic forms recognized.

Station 197. 3.^.27. Lat. 620 27' S. Long. 580 11' 30" W. Bransfield Strait, South Shetlands.
1974 m. (Plate XX.)

Diatomaceous mud. A brownish grey mud of homogeneous constitution. It consists mainly of
flocculent organic debris, some flocks having a greenish tinge. Diatom fragments form the bulk of
the material, together with unbroken frustules of Thalassiosira, Coscinodiscus, Fragilaria, Rhizosolenia
and Corethron. Some of these diatoms retain their protoplasmic content. One textularian Fora-
minifer is noted together with the usual sponge spicules. Mineral grains, quartz and green horn-
blende, occur in small quantity ; they are mostly less than 0-05 mm. in diameter, and few reach
o-i mm.

Station 198. 3-4. iv. 27. Lat. 620 38' S. Long. 58° 04' W. Bransfield Strait, South Shetlands.
1600 m. (Plate XX.)

Diatomaceous mud. A brownish grey mud similar to that from St. 197. It shows the same
characters under the microscope, but no Foraminifera were seen.

Station 199. 4. iv. 27. Lat. 620 49' S. Long. 570 56' 30" W. Bransfield Strait, South Shetlands.
735 m. (Plate XX.)

Diatomaceous mud. This sample of brownish mud contains one pebble roughly 1 cm. in dia-
meter ; it is more sandy than the preceding sample and does not coagulate on drying. The average
size of the abundant quartz grains is larger than in samples at Sts. 197 and 198, the maximum dia-
meter being about 0-2 mm. Green hornblende occurs also. Centric diatoms, Coscinodiscus and
Thalassiosira, are plentiful and reach o- 1 mm. across ; the smaller forms include Cocconeis, Fragilaria.
Sponge spicules are noted. Flocculent material of the usual character forms a "matrix", in which
the diatoms and mineral grains are enclosed.

Station 200. 4. iv. 27. Lat. 690 59' 30" S. Long. 570 49' W. Bransfield Strait, South Shetlands.
345 m. (Plate XX.)

Diatomaceous mud. A very small sample of brownish mud, which differs from the preceding in
size of the mineral grains and the slight degree of rounding. Quartz grains reach a diameter of

SEA-FLOOR DEPOSITS. PART I 323

0-25 mm., and the average size is estimated at about 005 mm. Centric diatoms, Coscinodiscas and
Thalassiosira, reach o-i mm. in diameter, while Cocconeis and Fragilaria are also well developed.
These forms occur with sponge spicules in the flocculent material which, however, appears to be less
in proportionate bulk than in the preceding samples (Sts. 197, 198 and 199).

Station 201. 5. iv. 27. Lat. 630 00' 30" S. Long. 590 06' 30" W. Bransfield Strait, South Shet-
lands. 343 m. (Plate XX.)

Diatomaceous mud. This small sample contains numerous grains of quartz and green horn-
blende, the average diameter of which is probably between o-i and 0-05 mm. Angular fragments of
vesicular volcanic glass are also conspicuous. The grains are held in flocculent material composed
of diatom tests (whole and fragmentary) and organic debris, of which some at least consists of the
protoplasmic constituent of the diatoms.

Station 202. 5. iv. 27. Lat. 620 48' S. Long. 6o°05'W. Bransfield Strait, South Shetlands.
909 m. (Plate XX.)

Diatomaceous mud. A considerable proportion of this dark grey mud consists of detrital grains,
angular to rounded in shape; some of the grains are more than o-i mm. in diameter, but also there
are many more whose diameter is less than 0-02 mm. Fragments of volcanic glass are sharply angular
while quartz grains are usually more rounded. Flocculent material is composed largely of tiny
mineral grains or rods, with some fragmentary diatom frustules. The genera Coscinodiscus, Thalas-
siosira, Cocconeis, Fragilaria and Rhizosolenia are present but not abundant.

Station 203. 5-6. iv. 27. Lat. 62° 56' S. Long. 590 50' W. Bransfield Strait, South Shetlands.
949 m. (Plate XX.)

Diatomaceous mud. A brownish mud with a conspicuous proportion of detrital grains, some of
which reach a diameter of 0-25 mm. Green hornblende and brown volcanic glass are noted in
addition to the predominant quartz. The fine-grained fraction, of the usual flocculent character,
includes tiny mineral fragments, sponge spicules, comminuted diatom frustules, and small diatoms
about 0-05 mm. in diameter. The genera Thalassiosira, Coscinodiscus, Cocconeis and Fragilaria are
represented.

Station 204. 6. iv. 27. Lat. 630 05' S. Long. 590 42' W. Bransfield Strait, South Shetlands.
943 m. (Plate XX.)

Diatomaceous mud. Similar in general character to the sample from St. 203. Some of this sedi-
ment was spread on a glass slip in a thick film of water; movement of the cover-glass caused groups
of sand grains to be rafted along on pieces of flocculent material which preserved their continuity.

Station 206. 6. iv. 27. Lat. 630 26' S. Long. 590 28' W. Bransfield Strait, South Shetlands.
310 m. (Plate XX.)

Diatomaceous mud. This is essentially similar to the sample from St. 204, consisting mainly of
flocculent material which encloses diatom frustules and mineral grains. The latter are usually very
small, but there are larger grains (up to o-i mm. diameter) of quartz, volcanic glass and green
hornblende. The diatoms include representatives of the genera Coscinodiscus, Thalassiosira, Cocconeis,
Fragilaria.

Station 209. 14. iv. 27. Port Foster, Deception Island, South Shetlands. 168 m. (Plate XX.)

Terrigenous mud. The mineral fraction consists mainly of angular fragments of brown, vesicular,
volcanic glass which vary in size between a diameter of about 0-2 mm. and exceedingly small di-
mensions. Some flocculent matter is present, with the diatoms Thalassiosira, Cocconeis, Licmophora
and Fragilaria, but unbroken frustules are not abundant. The flocculent material causes agglutination
so that the deposit forms a coherent mass on drying.

Station 211. 15. iv. 27. Lat. 620 35' S. Long. 63 ° 20' W. 2865 m. (Plate XX.)
Diatomaceous mud. A brown-grey mud containing abundant mineral grains which are usually
less than 0-05 mm. in diameter. The grains are mainly of quartz, but occasional fragments of green

324 DISCOVERY REPORTS

hornblende and glaucophane occur. The bulk of the deposit consists of flocculent material which
surrounds the sand grains and diatoms. Among the latter are species of Fragilaria, Rhizosolenia,
Thalassiosira and Coscinodiscus. Some of the diatoms retain their protoplasmic contents, and much
of the flocculent matter is doubtless of organic origin.

Station 212. 16. iv. 27. Lat. 6i° 15' S. Long. 640 42' 50" W. 3350 m. (Plate XVII.)
Diatomaceous mud. A brown mud, essentially similar to that from St. 211, but the mineral grains
attain a larger size, about o-i mm. in diameter. The occurrence of quartz, green hornblende and
white mica is noted. The organic material is chiefly diatomaceous but some arthropod remains and
eroded tests of Globigerina occur.

Station 228. 2. v. 27. Lat. 530 33' S. Long. 61 ° 49' 30" W. 660 m. (Plate XVIII.)
Diatomaceous mud. A grey mud with abundant detrital grains up to o-i mm. diameter. Quartz
grains preponderate but occasional grains of hypersthene are seen. Large (up to o-i mm.) centric
diatoms (e.g. Coscinodiscus), navicular forms (especially Fragilaria) and sponge spicules are present.
Flocculent material is conspicuous, but not in such large proportion as in the foregoing samples.

Station 230. 5. v. 27. Lat. 530 17' S. Long. 6o° 25' W. 675 m. (Plate XVIII.)

Terrigenous mud. This grey mud has an abundance of very small mineral grains (quartz and

occasionally hypersthene) in a matrix of flocculent material which is more granular in appearance

than usual. Recognizable organic remains are virtually absent.

Station 235. 29. v. 27. Lat. 500 45' S. Long. 560 18' 30" W. 600 m. (Plate XVII.)
Sand. A fine-grained sand or silt, some of the constituent grains (chiefly quartz) reaching a
diameter of 0-25 mm. Fragments of green hornblende are smaller and less numerous. A few Fora-
minifera and diatoms occur, but organic remains other than sponge spicules are scarce. The pre-
servative liquid has acquired a slight greenish tinge.

Station 236. 29-30. v. 27. Lat. 500 35' 30" S. Long. 550 59' 15" W. 612 m. (Plate XVII.)

Station 237. 30. v. 27. Lat. 500 17' 40" S. Long. 550 31' 30" W. 904 m. (Plate XVII.)
Sand. The samples from Sts. 236 and 237 are so similar that they may be described together. The
sand grains, mostly quartz, reach a diameter of 0-5 mm. ; some green hornblende is present. Large
opaque grains, which form a dark layer in the sediment, together with smaller yellow-green grains
show the characters of glauconite. Flocculent matter is present in small amount. The organic
fraction includes some Foraminifera (especially Globigerina), centric diatoms and sponge spicules.
Some of the latter are infilled with yellow-green " glauconitic " material which shows aggregate
polarization between crossed nicols.

Station 263. 19. vii. 27. Lat. 330 06' S. Long. 170 08' W. 461 m. (Plate XVII, inset.)
Glauconitic mud. The material has settled into layers during storage. The upper part consists
of homogeneous, extremely fine-grained, flocculent material with a few sponge spicules and occa-
sional mineral grains. The lower layer contains abundant Globigerina and some rounded grains of
glauconite, up to o- 1 5 mm. diameter. The fine-grained constituent apparently consists of comminuted
foraminiferal tests and coccoliths with a considerable proportion of protoplasmic debris, the latter
having a greenish tinge. Some of the Globigerina shells are infilled with glauconite.

Station 264. 19-20. vii. 27. Lat. 330 06' S. Long. 160 55' E. 645 m. (Plate XVII, inset.)
Glauconitic mud. A "mixed" deposit consisting of muddy flocculent material with a con-
siderable proportion of mineral grains and foraminiferal tests. The mineral fragments (chiefly
quartz) reach a diameter of o- 1 mm. Some bright green glauconite grains, rounded to angular in
shape, have about the same average size, but one grain measured 0-25 mm. in diameter. The
Foraminifera often retain their protoplasmic contents ; Globigerina and Frondicularia were noted as
reaching a diameter of 0-2 mm. Coccoliths are frequently seen in the finer material.

SEA-FLOOR DEPOSITS. PART I 325

Station 265. 20. vii. 27. Lat. 330 06' 30" S. Long. 160 32' E. 1620 m. (Plate XVII, inset.)
Glauconitic mud. A pale, fine-grained sediment, mainly flocculent material in which are scattered
Foraminifera and a few sand grains. The flocculent material seems to be in part comminuted fora-
miniferal shells together with coccoliths and rhabdoliths, in part decayed organic matter. The
minerals are chiefly quartz and glauconite, but both are scarce. The Foraminifera are mainly
Glokigerina.

This and the preceding sample (St. 264) are classified in the official Station List as Globigerina
ooze. But as Foraminifera by no means form the bulk of the deposit, the sediments are here regarded
as glauconitic mud, with a subordinate proportion of foraminiferal debris.

Station 283. 13. viii. 27. Off Annobon, Gulf of Guinea. 075 to 1 mile N 120 E of Pyramid
Rock, Annobon. 18-30 m.

Shell sand. The sand consists of foraminiferal and small molluscan shells which average about
1 mm. in diameter. Some of the Foraminifera (e.g. Cristellaria) are larger than this, but the shells
of Globigerina are smaller ; large textularians are also conspicuous. Few mineral grains occur.

Station 363. 26. ii. 30. 2-5 miles S 8o° E of south-east point of Zavodovski Island, South
Sandwich Islands. 329 m. (Plate XVII.)

Fine sand. The bulk of the sample is a dark brown sand, the grains reaching a diameter of 0-5 mm.
and ranging downwards in size, the average being about 025 mm. Much of the material consists of
brown volcanic glass in sharply angular fragments, and many of the grains are vesicular: the sand
owes its dark colour to this material. Quartz grains (also up to 0-5 mm. diameter) appear to be more

rounded.

The fine-grained material, which has formed a film on the sand in the storage jar, is a mixture of
tiny mineral particles and diatomaceous debris. Fragilaria and Thalassiothrix appear to be the most
abundant genera, while centric forms {Coscinodiscus and Thalassiosira) are not especially common in
occurrence.

Station 366. 6. iii. 30. 4 cables south of Cook Island, South Sandwich Islands. 340m. (Plate XVII.)
Fine sand. A dark brown, nearly black, sand composed almost entirely of fragments of quartz and

volcanic glass. In size the grains range from a diameter of about o-i mm. to exceedingly small particles.

Most of the grains are markedly angular in shape, though some have suffered a slight degree of

rounding. A subordinate quantity of flocculent material gives a green colour to the preserving liquid.

The presence of sponge spicules and frustules of Coscinodiscus, Cocconeis and Fragilaria is noted.

The deposit is similar to that from St. 363, but is rather finer in grain.

Station 425. 4. ix. 30. Lat. 340 50' S to 340 53! ' S. Long. 260 \i\' E to 260 30V E. 4107 m.

Globigerina ooze. The sample is perhaps more sandy than usual in a typical ooze. Tests of
Globigerina and Orbulina are abundant and well developed. Diatoms and sponge spicules also occur,
but the former are not numerous. Coccoliths are abundant in the finer fraction, while rhabdoliths
(as usual) are less in evidence.

Station 1165. 4. iii. 33. Lat. 400 547' S. Long. 9° 25-5' E. 4642 m. (Plate XVII.)
Radiolarian ooze. This is a brownish mud composed largely of organic debris, with some ad-
mixture of sand grains which are generally less than 0-05 mm. in diameter. The coarser fraction
consists mainly of radiolarian tests, among which sphaeroid forms are dominant, discoid forms are of
common occurrence, while nasselarians are comparatively rare. The genera recognized are Ceno-
sphaera, Carposphaera, Xipkosphaera, Hexastylus, Heliodiscus, Porodiscus, Hymeniastrum, Rhopa-
lastrum and Lithostrobus. Some Foraminifera {Globigerina), diatoms (Coscinodiscus), and anchor-shaped
spines of the holothuroidean Synapta are noted.

326 DISCOVERY REPORTS

R.R.S. 'WILLIAM SCORESBY'

Station WS 18. 26. xi. 26. Lat. 54° 07' S. Long. 360 23' W. 113 m. (Plate XIX.)
Diatomaceous mud. A small sample containing much detrital mineral material. Large mineral
grains, up to o-i mm. across, are mingled with exceedingly fine sediment. The latter includes com-
minuted diatoms and possibly other organic debris. Centric diatoms, some more than o-i mm. in
diameter, and sponge spicules are present.

Station WS 20. 28. xi. 26. Lat. 530 52' 30" S. Long. 360 00' W. 535 m.

A small sample in a glass tube which is enclosed with a label in a bottle. The label states — " Brought
up in a net lowered to 500 m. and hauled to 250, then closed. Lucas Sounder gave 535 m., rock".

The tube contains a small quantity of sand containing grains up to 0-25 mm. in diameter. There is
no muddy material. Some tests of Globigerina occur.

As this sample is probably incomplete it is not classified with the others in this report and the
station is not entered on Plate XIX.

Station WS 26. 18. xii. 26. Lat. 530 33' 15" S. Long. 37° 45' 15" W. 1180 m. (Plate XIX.)

Diatomaceous mud. A greenish grey mud ; the preservative liquid has acquired a greenish tinge.
This sediment might be classed as diatom ooze except for the presence of green flocculent material
and the proportion of mineral grains which reach a diameter of o-i mm., though the majority are
below 0-05 mm. Quartz and green hornblende are the only conspicuous minerals. The sample as a
whole is extremely fine grained and the diatoms are small. Coscinodiscus, Rhizosolenia and Fragilaria
are the chief genera.

Station WS 28. 19. xii. 26. Lat. 530 48' 15" S. Long. 380 13' W. Two samples from depth of
346 and 150 m. respectively. (Plate XIX.)

Diatomaceous mud. The sample from 346 m. consists of abundant mineral grains (a few up to
0-4 mm. diameter) enclosed in a "matrix" of flocculent material. The latter consists of diatoms
{Coscinodiscus and Fragilaria), tiny mineral grains and indeterminable debris, some of which appears
to be of diatomaceous origin.

A label in the bottle containing the second sample states, " N 70 V touched bottom at 145 m." The
sample is a clean sand with many dark and opaque grains which appear to be rock fragments ; many
grains reach a diameter of 0-25 mm. and more. This is probably an incomplete sample, the finer
constituents having been lost.

Station WS 32. 21. xii. 26. Mouth of Drygalski Fjord, South Georgia. 225 m. (Plate XIX.)

Diatomaceous mud. A grey sediment of the usual diatomaceous character. Fragilaria, Coscino-
discus and Thalassiosira are the chief diatoms. The sand grains are small, being less than 005 mm.
in diameter. Green hornblende, as well as quartz, is abundantly represented among the grains.

Station WS 33. 21. xii. 26. Lat. 540 59' S. Long. 350 24' W. 135 m. (Plate XIX.)
Diatomaceous mud. A grey mud containing a fair amount of detrital grains, some reaching o- 2 mm.
in diameter, many up to o-i mm. Green hornblende is of common occurrence. There is much
flocculent material containing diatom frustules (Coscinodiscus, Fragilaria, etc.) and sponge spicules;
the masses of this material often show a green colour which is sufficient to give a greenish tinge to the
whole sample when wet.

Station WS 37. 22. xii. 26. Lat. 540 45' S. Long. 350 11' W. 318 m. (Plate XIX.)
Diatomaceous mud. A grey mud similar to that from St. WS 33. There are abundant detrital
grains, some 0-2 mm. in diameter, among which quartz, felspar and green hornblende are noted.
Diatoms are abundant, whole tests are more conspicuous than in the foregoing sample and perhaps
in greater variety; centric forms (Coscinodiscus and Thalassiosira) reach a diameter of o-i mm. The
flocculent material contains green masses, some of which firmly enwrap sand grains. It may be
suggested that the enclosure of detrital grains in the buoyant flocculent material is an important
factor in the transport of the mineral grains far from their original source.

SEA-FLOOR DEPOSITS. PART I 327

Station WS 39. 23. xii. 26. Lat. 540 08' S. Long. 35° 43' W. 237 m. (Plate XIX.)
Diatomaceous mud. A grey mud which differs from the preceding sample in the somewhat smaller
average size of the sand grains, and the greater proportionate bulk of the flocculent constituent, but
otherwise very similar. The greenish tinge of the sediment is due to the presence of green flocculent
masses.

Station WS 40. 7. i. 27. Lat. 550 09' S. Long. 350 58' W. 183 m. (Plate XIX.)
Diatomaceous mud.

Station WS 41. 7. i. 27. Lat. 54° 32' 45" S. Long. 360 43' 45" W. 140 m. (Plate XIX.)
Diatomaceous mud.

Station WS 42. 7.1.27. Lat. 54° 41' 45" S. Long. 360 47' W. 175 m. (Plate XIX.)
Diatomaceous mud.

Station WS 43. 7-8. i. 27. Lat. 540 54' S. Long. 360 50' W. 200 m. (Plate XIX.)
Diatomaceous mud. These four samples are very similar, and are typical of diatomaceous muds
in general. The mineral grains (quartz and green hornblende) are usually small, but some attain
diameters up to 0-2 mm. The detrital grains are enclosed in greenish masses of flocculent material
which contains small diatoms.

Station WS 45. 8. i. 27. Lat. 540 38' 30" S. Long. 370 30' 55" W. 180 m. (Plate XIX.)
Diatomaceous mud. A grey mud enclosing some hard black pebbles, 1 cm. in diameter. There
are abundant detrital grains including green hornblende up to 0-2 mm. in diameter, while a few reach
0-5 mm. across. Centric diatoms are conspicuous, and sponge spicules are frequent in occurrence.
The finer material includes diatom debris and indeterminable flocculent matter.

Station WS 46. 8. i. 27. Lat. 540 20' 15" S. Long. 370 32' 30" W. 194 m. (Plate XIX.)
Diatomaceous mud. A grey mud with small black pebbles, essentially like the preceding deposit.
A fair proportion of mineral grains is enclosed in fine-grained flocculent material which contains the
usual diatoms. Some grains of quartz and green hornblende reach a diameter of o-i mm.

Station WS 47. 9. i. 27. Lat. 54° 22' S. Long. 370 50' W. 160 m. (Plate XIX.)
Diatomaceous mud. A grey mud, paler in colour than the preceding sample, but of essentially
similar character. Detrital grains, especially quartz and (more rarely) green hornblende, form a
considerable bulk in the deposit ; they are mostly below 0-05 mm. in diameter, but some reach 0-25 mm.
Centric diatoms are conspicuous in the flocculent "matrix".

Station WS 48. 9. i. 27. Lat. 540 24' S. Long. 380 09' W. 224 m. (Plate XIX.)
Diatomaceous mud. A grey mud when dry, greenish grey when wet. This sediment is almost a
diatom ooze with centric (up to 0-2 mm. in diameter), navicular and elongate diatoms, together with
some sponge spicules, but there is a considerable quantity of green flocculent matter. There are few
mineral grains and those are mostly of small size. A fragment of rock occurs in the sample, measuring
2-5 by i"3 cm.; in thin section this is seen to be a tuff containing fragments of felspar and green
hornblende.

Station WS 49. 9. i. 27. Lat. 540 28' S. Long. 380 22' 15" W. 223 m. (Plate XIX.)
Diatomaceous mud. This sample is darker in colour than the preceding, and the preservative
liquid has acquired a green tinge. There are abundant detrital grains, mostly less than 0-05 mm. in
diameter, but some (including green hornblende) greater than o-i mm. These and large centric
diatoms are embedded in flocculent material of the usual character.

Station WS 50. 9. i. 27. Lat. 540 30' 30" S. Long. 380 40' 30" W. 230 m. (Plate XIX.)

Diatomaceous mud. The sample has dried to a light grey colour. The pale colour appears to be

due to the relative scarcity of dense flocculent masses ; the fine material is abundant, but consists

almost entirely of comminuted diatom tests and mineral grains. Frustules of Coscinodiscus and

Fragilaria are abundant, Rhizosolenia less plentiful and sponge spicules are present in small quantity.

328 DISCOVERY REPORTS

The numerous detrital grains are mostly below 005 mm. in diameter, but a few reach o-i mm. ; they
include small crystals of green hornblende.

Station WS 52. 10. i. 27. Lat. 540 03' 30" S. Long. 3 8° 35' W. 184 m. (Plate XIX.)
Diatomaceous mud. A grey mud with abundant detrital grains mostly below 0-05 mm. in
diameter, but some reaching 0-25 mm. The flocculent material encloses frustules of Coscinodisais
(large), Fragilaria, Cocconeis and Rhizosolenia, together with sponge spicules and much comminuted
material.

Station WS 63. 20-21. i. 27. Lat. 540 36' S. Long. 390 14' W. 1752 m. (Plate XIX.)
Diatomaceous mud. No stones are present in this sample of grey mud, though such material is
reported in the Station List. The detrital grains, including green hornblende, attain a diameter of
o-i mm. The finer fraction consists largely of diatom frustules; the larger centric tests are broken,
but the smaller ones (mostly navicular forms) are whole. Occasional Radiolaria and sponge spicules
are noted.

Station WS 71. 23. ii. 27. Lat. 510 38' S. Long. 570 32' 30" W. 6 miles N 6o° E of Cape Pem-
broke Light, East Falkland Island. 80 m. (Plate XVIII.)

Sand. The label states: "Bottom sample taken from contents of trawl". The mineral grains
reach a diameter of 0-5 mm., and there is apparently little material of medium grade, though some
very small grains are present. There is a considerable proportion of calcareous matter, such as
Foraminifera, echinid spines and small molluscan shells. Many of the shells are undamaged but there
is also much broken material, some exceedingly small. The Foraminifera include miliolines (Bilo-
culina and Miliolina) and rotalines {Globigerina and Anomalina).

Station WS 76. n.iii. 27. Lat. 5i°oo'S. Long. 620 02' 30" W. 207 m. (Plate XVIII.)
Fine dark sand. A fine-grained sand, the average diameter of the grains being less than o-i mm.,
though some fragments are larger. The minerals are mainly quartz and hypersthene. There is much
flocculent matter which encloses organic remains such as sponge spicules and, more rarely, centric
diatoms.

Station WS 77. 12. hi. 27. Lat. 51° 01' S. Long. 66° 31' 30" W. no m. (Plate XVIII.)
Coarse dark sand. The sample consists entirely of mineral grains up to 1 mm. in diameter. There
is apparently a fair variety of minerals in rounded and subangular grains ; quartz (as usual) is most
abundant, but hypersthene is plentiful as worn crystals up to 0-5 mm. in length.

Station WS 78. 13. Hi. 27. Lat. 51° 01' S. Long. 68° 04' 30" W. 95 m. (Plate XVIII.)
Fine dark sand. Consists of rounded mineral grains up to 0-3 mm. in diameter, but the average is
much less; the smaller grains seem to be more angular. Supernatant flocculent material.

Station WS 79. 13. Hi. 27. Lat. 51° 01' 30" S. Long. 640 59' 30" W. 132 m. (Plate XVIII.)
Fine dark sand. The rounded sand grains, which average about 0-2 mm. in diameter, are mostly

of quartz, but hypersthene is also abundant. There is much flocculent material consisting of tiny

mineral grains, navicular diatoms and sponge spicules.

Station WS 80. 14. Hi. 27. Lat. 500 57' S. Long. 630 37' 30" W. 152 m. (Plate XVIII.)
Fine dark sand. The deposit consists mainly of rounded mineral grains, about 0-3 mm. in dia-
meter, but includes also many small angular grains. There is much muddy material of the usual
flocculent character, which includes tiny mineral grains, sponge spicules, diatoms (centric and navi-
cular forms) a few small Foraminifera, and probably other organic debris.

Station WS 83. 24. iii. 27. Lat. 520 30' S. Long. 6o° 08' W (approximately). 14 miles S 640 W
of George Island, East Falkland Island. 137 m. (Plate XVIII.)

Sand. This sample is described in the official Station List as " fine green sand with shells ". The
mineral grains attain a diameter of about 0-5 mm., but most of them are much smaller; they are
chiefly angular fragments of quartz, but grains of hypersthene are occasionally seen. Some opaque

SEA-FLOOR DEPOSITS. PART I 329

green grains (probably glauconite) are present, but the colour of the deposit is not entirely due to
these grains ; moreover the preserving liquid is coloured greenish yellow. Shelly material is present
in about the same proportion as the mineral constituent. It consists mainly of microscopic lamelli-
branch shells mostly in a fragmentary condition ; there is a variety of Foraminifera, which includes
Globigerina, Polystomella and other genera. The flocculent material contains an abundance of green
algal cells with mucilaginous walls ; these algae apparently contribute largely to the green colour of the
deposit.

Station WS 84. 24. iii. 27. Lat. 52° 34' S. Long. 590 10' W (approximately). 7J miles S 900 W
of Sea Lion Island, East Falkland Island. 75 m. (Plate XVIII.)

Coarse sand. A clean speckled sand with shells. The sand consists mostly of clear quartz grains
about 0-5 mm. in diameter; there are some dark green grains which may be glauconite. Much shelly
material is present ; mainly fragments of molluscan shells, it includes also small lamellibranchs (fry),
Foraminifera and echinid spines.

Station WS 86. 3. iv. 27. Lat. 530 53' 30" S. Long. 60° 34' 30" W. 151 m. (Plate XVIII.)
Sand. A comparatively fine speckled sand with much shelly material, consisting of rounded or
subangular grains of clear quartz and other minerals, the average diameter being about 0-2 mm.
Occasional grains of hypersthene are noted. The shelly material includes fragments of lamelli-
branchs (many worn and rolled), echinid spines and Foraminifera. The small amount of muddy
material in the sample contains centric and compound diatoms, as well as flocculent aggregates.

Station WS 87. 3. iv. 27. Lat. 540 07' 30" S. Long. 580 16' 80" W. 96 m. (Plate XVIII.)
Sand with shells and stones. A speckled sand with fragments of shale and large shells, especially
Pec ten. The sand itself is largely composed of shelly material, including remains of lamellibranchs,
Serpulae, Polyzoa, echinoids (spines), Foraminifera. Some detrital sand grains are present as well as
the shale fragments, some of which are highly glauconitic. Supernatant flocculent material contains
diatoms and possibly other small organisms.

Station WS 88. 6. iv. 27. Lat. 54° 00' S. Long. 640 57' 30" W. 118 m. (Plate XVIII.)
Sand. A clean speckled sand, largely composed of rounded quartz grains less than 0-5 mm. in
diameter; rounded crystals of hypersthene up to 0-4 mm. in diameter are fairly common ; other dark
grains may be glauconite. The organic fraction includes much shelly detritus, rolled fragments of
echinoid spines, and Foraminifera, many of which are not waterworn. Polyzoa also occur, encrusting
small pebbles.

Station WS 89. 7. iv. 27. Lat. 530 00' S. Long. 68° 06' W. 9 miles N 210 E of Arenas Point
Light, Terra del Fuego. 23 m. (Plate XVIII.)

Gravel. A non-graded deposit. The largest constituents are rounded pebbles, up to 4 cm. in
diameter, encrusted with Polyzoa. There are smaller pebbles from 1 cm. diameter downwards, some
of quartz, others of coloured material. The sand grade contains a variety of heavy minerals. The
remaining constituent is a grey slimy mud composed of flocculent aggregates with sponge spicules
and tiny mineral grains.

Station WS 90. 7. iv. 27. Lat. 520 18' S. Long. 68° 00' W. 13 miles N 830 E of Cape Virgins
Light, Argentine Republic. 82 m. (Plate XVIII.)

Sand. The sample contains a large proportion of clean, rounded quartz grains, with an average
diameter of about 0-3 mm. ; grains of hypersthene are also fairly common. There is apparently no
shelly material. The finer fraction is composed of the usual flocculent material with small quartz
grains, diatoms and spicules.

Station WS 91. 8. iv. 27. Lat. 520 53' 45" S. Long. 640 37' 30" W. 191 m. (Plate XVIII.)
Sand. A clean sand with a variety of mineral grains, up to about 0-5 mm. in diameter, rounded and
subangular in shape. Quartz and hypersthene are the most conspicuous minerals. Shelly material

5-2

330 DISCOVERY REPORTS

includes worn fragments of large scallop shells ; a damaged simple coral is also present. There is a
subordinate proportion of flocculent material.

Station WS 92. 8. iv. 27. Lat. 510 58' 30" S. Long. 650 01' W. 145 m. (Plate XVIII.)
Sand. A clean sand with rounded mineral grains, the majority of which are about 0-2 mm. in
diameter, though some reach 0-5 mm. Besides the predominant quartz, which is often stained, there
are many opaque grains, some of which may be glauconite. Grains of hypersthene are plentiful and
some worn crystals (beautifully pleochroic) reach a length of 0-5 mm. Green hornblende also is seen,
though rarely. Occasional Foraminifera are noted. The finer fraction is flocculent when examined in
quantity and contains extremely small mineral particles, spicules, Foraminifera and diatoms.

Station WS93. 9. iv. 27. Lat. 51° 52' 30" S. Long. 6i°3o'W. 7 miles S 8o° W of Beaver
Island, West Falkland Island. 133 m. (Plate XVIII.)

Sand. A clean sand with some muddy material. The sample is rather coarser than that from the
preceding station (WS 92), the rounded mineral grains reaching a diameter of 0-5 mm. or more. The
sand is chiefly composed of quartz grains but also includes worn crystals of hypersthene, rounded
shell fragments, Foraminifera in some abundance, and broken echinoid spines. There is a fair
amount of flocculent material containing tetrad sponge spicules and diatoms.

Station WS 94. 16. iv. 27. Lat. 500 00' 15" S. Long. 640 57' 45" W. no m. (Plate XVIII.)
Sand. A clean sand with rounded grains, many coloured ones mingled with the predominant
quartz. Grains of hypersthene are plentiful and some crystals (up to 0-5 mm. in length) are only
slightly rounded. No shelly matter was seen and there is little flocculent material.

Station WS 95. 17. iv. 27. Lat. 480 58' 15" S. Long. 640 45' W. 109 m. (Plate XVIII.)
Sand. The sand is clean and composed of rounded grains of quartz and coloured minerals, among
which hypersthene is prominent, with an average diameter of about o-2 mm. There are many
small pebbles, but virtually no shelly matter as sand. A layer of flocculent material has separated out
at the top of the sample ; this is of the usual type, but centric diatoms (especially Cosci/iodiscus) are
conspicuous in it.

Station WS 96. 17. iv. 27. Lat. 480 00' 45" S. Long. 640 58' W. 96 m. (Plate XVIII.)
Sand. A dark, evenly graded sand, the rounded and often stained quartz grains having an average
diameter of about 0-3 mm. Occasional grains of hypersthene are noted. There is no shelly material.

Station WS 97. 18. iv. 27. Lat. 490 00' 30" S. Long. 6i° 58' W. 146 m. (Plate XVIII.)
Sand. A brownish fine-grained sand with a fair proportion of finer material. The mineral grains
reach a diameter of 0-5 mm. though most are smaller; they are usually rounded and stained. Besides
the preponderant quartz, grains of hypersthene are plentiful. The shelly material includes Fora-
minifera, broken molluscan shells and worn echinoid spines. The flocculent material consists of
extremely small particles, mostly indeterminate but including some mineral grains and fragments of
foraminiferal tests, also sponge spicules and elongate diatoms.

Station WS 98. 18. iv. 27. Lat. 490 54' 15" S. Long. 6o° 35' 30" W. 173 m. (Plate XVIII.)
Sand. A fine non-graded, dark sand with much flocculent matter containing occasional diatoms.
Mineral grains reach a diameter of 0-5 mm. and grade downwards to very small dimensions. Glau-
conite seems to be present as internal moulds of foraminiferal shells of which one species is fairly
abundant.

Station WS 99. 19. iv. 27. Lat. 490 42' S. Long. 590 14' 30" W. 251m. (Plate XVIII.)
Sand. A fine even-graded sand composed mainly of rounded quartz grains, between 0-2 and o-i mm.
in diameter; the dark-coloured grains include some glauconite. This green material is also seen in-
filling the canals of broken sponge spicules. There are few foraminiferal or other shells. Some amount
of flocculent material is present.

SEA-FLOOR DEPOSITS. PART I 331

Station WS 108. 25. iv. 27. Lat. 480 30' 45" S. Long. 630 33' 45" W. 118 m. (Plate XVIII.)
Sand. A clean, evenly graded quartz sand with plentiful hypersthene and some opaque grains ; the

mineral fragments have an average diameter of about 0-3 mm., and are generally rounded in shape.

There is a small amount of flocculent matter, but apparently no shelly material.

Station WS 109. 26. iv. 27. Lat. 500 18' 48" S. Long. 580 28' 30" W. 145 m. (Plate XVIII.)
Sand. The sample consists of clean rounded quartz grains with some sponge spicules and a little
muddy material. The latter consists of extremely small flocculent particles, the aggregates of which
apparently give the dark colour to the sand. Many of the mineral grains have a diameter between
0-25 and 0-5 mm. Besides quartz grains, rounded crystals of hypersthene, up to 0-5 mm. in length,
are occasionally seen.

Station WS 128-129. 10-u.vi.27. Between 40° 19' S, io°04' Wand 400 10' 30" S, 9° 40' 45" W.
Between 2000 and 3000 m. (Plate XVII.)

Globigerina ooze. This is a typical Globigerina ooze with the usual preponderance of Globigerina
shells up to 0-5 mm. across. Some rotalines are also present. There is a small proportion of angular
sand grains, some of which are 0-2 mm. in diameter. In the finer fraction, coccoliths are extremely
abundant, but rhabdoliths are less plentiful.

Station WS 201. 22. iv. 28. Lat. 590 57' S. Long. 500 12' W. 4134 m. (Plate XVII.)
Diatomaceous mud. A small proportion of large, dark-coloured mineral grains includes volcanic
glass (up to 0-5 mm. diameter), while quartz (0-2 mm.), green hornblende and a few small prismatic
grains of glaucophane are noted. The larger grains are mostly well rounded. The fine-grained material
which makes up the bulk of the deposit consists partly of angular mineral grains, but largely of
diatom debris among which whole frustules of Coscinodiscus, Cocconeis and Fragilaria are abundantly
represented. The mud is dark brown when wet but becomes much lighter in colour on drying, and in
the almost colourless quality of the fine-grained material the deposit approaches a diatom ooze.

Station WS 202. 23. iv. 28. Lat. 6o° 23' S. Long. 520 52' W. 3987 m. (Plate XVII.)
Diatomaceous mud. The coarser fraction consists largely of mineral grains up to 0-3 mm. in
diameter; it includes angular fragments of quartz, green hornblende and brown volcanic glass. The
finer material is an admixture of tiny mineral grains and diatom debris among which large frustules of
Coscinodiscus up to 0-25 mm. in diameter are conspicuous. Triceratium , Biddu/p/iia and other centric
forms, as well as Rhizosolenia and Fragilaria, are also present. Like the sample from St. WS 201,
this sample approaches a diatom ooze in its light colour and in the quality of its fine-grained material.

Station WS 203. 25. iv. 28. Lat. 570 42' S. Long. 530 12' W. 4259 m. (Plate XVII.)
Diatomaceous mud. This sample is similar in constitution to the preceding. Some of the mineral
grains are 0-3 mm. in diameter though the average is much lower ; they are mostly angular in shape but
some are fairly rounded. Broken diatoms provide much of the finer material, in which whole frustules
of Coscinodiscus, Thalassiosira and Fragilaria are conspicuous. Broken tests of Globigerina are occa-
sionally seen. The sample is lighter in colour than most examples of diatomaceous mud.

Station WS 204. 26. iv. 28. Lat. 560 27' S. Long. 54° 22' W. 3328 m. (Plate XVII.)
Diatomaceous mud. This light brown mud has a considerable proportion of mineral grains which
range in size from exceedingly small to about 0-1 mm. in diameter. Angular grains of quartz and
green hornblende are conspicuous. The fine material consists in great part of diatom debris among
which Coscinodiscus, Thalassiosira and Fragilaria are of common occurrence. A few broken and worn
tests of Globigerina and Radiolaria are noted.

Station WS 205. 27. iv. 28. Lat. 550 49' S. Long. 560 18' W. 4207 m. (Plate XVII.)
Diatomaceous mud. This sample is essentially similar to that from St. WS 204 in general con-
stitution. The angular mineral grains (including quartz and green hornblende) are generally small.
Light-coloured flocculent aggregates which enclose the smallest mineral particles, contain broken

332 DISCOVERY REPORTS

tests of diatoms while some whole frustules of Coscinodiscus and Fragilaria are present, along with
sponge spicules.

This, like the preceding three samples, is lighter in colour than most samples of diatomaceous mud,
and apart from the considerable proportion of mineral grains, has almost the appearance of diatom
ooze.

Station WS 255. 22-23. viii- 2§- Lat- 53° 23' s- Long- 44° I0' w- 3003 m. (Plate XVII.)
Diatom ooze. The bulk of this light-coloured mud is composed of diatom frustules. In the
considerable variety of forms, the genera Coscinodiscus, Thalassiosira, Cocconeis, Fragilaria, Thalassio-
thrix and Rhizosolenia are noted. Sponge spicules and Radiolaria are represented, the latter by
occasional nasselarian forms. There is a small proportion of quartz grains, mostly below 0-05 mm.
in diameter.

Station WS 314. i.xii. 28. Lat. 530 36' S. Long. 410 05' W. 137 m. (Plate XVII.)
Sand. The most interesting feature of this extremely small sample of sand is the large proportion
of colourless flakes up to 0-2 mm. in diameter, some of which show a sharp, unworn, hexagonal
outline; some show a radial texture and are evidently zeolitic aggregates. A few diatoms are present,
among which Coscinodiscus and Fragilaria are noted. There is some evidence of ferruginous
cementation.

Station WS 317. 4. xii. 28. Lat. 52° 41' S. Long. 490 39' 30" W. 3369 m. (Plate XVII.)
Gravel. The sample consists of a few small black pebbles. The label states, "It is probable that
the finer constituents of the sample were washed out, and the coarse only retained ".

Station WS 319. 5. xii. 28. Lat. 52° 01' S. Long. 540 52' W. 1602 m.

Gravel and coarse sand. A small sample concerning which the Station List notes "finer material
from bottom sample washed out". It contains a few tests of Globigerina, some small pebbles,
and a little coarse sand. The latter includes some dark opaque grains (up to 1 mm. diameter),
rounded and lobate in shape ; when crushed they show the optical characters of glauconite. As the
sample is incomplete the station has not been entered on Plate XVII.

Station WS 374. 6-7. ii. 29. Lat. 550 09' S. Long. 400 00' W. 3226 m. (Plate XVII.)
Diatom ooze. A brownish deposit with some admixture of mineral grains, but a fairly typical
diatom ooze. The diatoms include species of Coscinodiscus, Fragilaria, Thalassiothrix and Rhizo-
solenia. A few tests of Globigerina and Radiolaria are present.

Station WS 377. 9. ii. 29. Lat. 580 34' S. Long. 440 47' W. 2552 m. (Plate XVII.)
Diatom ooze. The mineral constituent in this brown mud is very small, consisting of angular
grains which are usually much smaller in diameter than the maximum of o-i mm. Diatoms make up
the bulk of the deposit, occurring in some variety as whole frustules and also as broken material in all
stages of disintegration. Coscinodiscus, Fragilaria, Rhizosolenia and Thalassiothrix are most abundant,
while Thalassiosira and Asteromphalus occur more rarely. Occasional fragments of Radiolaria and
Foraminifera are noted.

Station WS 381. 14. ii. 29. Lat. 6i° 26' S. Long. 560 19' W. 425 m. (Plate XVII.)
Coarse sand. This small sample contains mineral grains up to 1 mm. in diameter, all fairly angular
in shape. Quartz, green hornblende and volcanic glass are represented, as well as smaller grains of
phillipsite which often occurs as cross-twin or flat, hexagonal crystals. The sample is almost free of
organic debris, only a few diatoms being noted.

Station WS 382. 15. ii. 29. Lat. 620 15' 35" S. Long. 580 18' 30" W. 425 m. (Plate XX.)
Terrigenous mud. The coarse fraction consists almost entirely of mineral grains, the size of
which ranges downwards from about 0-2 mm. in diameter. The finer material comprises exceedingly
small mineral grains together with some indefinite flocculent aggregates. Recognizable organic
remains are scanty, but include diatoms (Coscinodiscus and Fragilaria) and sponge spicules.

SEA-FLOOR DEPOSITS. PART I 333

Station WS 383. 15. ii. 29. Lat. 620 20' 40" S. Long. 580 13' W. 2085 m. (Plate XX.)
Diatomaceous MUD. This dark, greenish brown mud is finer in grain than the last, and contains
more diatom frustules. Among the latter Coscinodisciis, Thalassiosira, Cocconeis, Fragilaria, Corethron
and Rhizosolenia are noted. Some retain their protoplasmic content which is similar in appearance to
the fine-grained flocculent material. The mineral grains (chiefly subangular fragments of quartz)
range downwards from a diameter of about o-i mm.

Station WS 384. 15. ii. 29. Lat. 620 25' 40" S. Long. 580 06' 10" W. 1957 m. (Plate XX.)
Diatomaceous mud. This sample shows close resemblance in general characters to that from
St. WS 383. The assemblage of diatoms contains the same genera, but the great abundance of large
and well-developed frustules of Corethron and Rhizosolenia is worthy of special mention.

Station WS 385. 16. ii. 29. Lat. 620 32' S. Long. 570 55' W. 1838 m. (Plate XX.)
Diatomaceous mud. This sample is typical of its kind, and differs in no essential particular from
those last described. It contains the same assemblage of diatoms, namely Coschwdiscus, Thalassiosira,
Cocconeis, Fragilaria, Corethron and Rhizosolenia in quantity. The protoplasmic contents of some
diatoms are seen within the valves, and appear to contribute to the flocculent aggregates. The
preserving liquid is tinted green. The average size of the sand grains is perhaps smaller than in the
preceding samples.

Station WS 386. 16. ii. 29. Lat. 620 41' S. Long. 570 44' W. 1392 m. (Plate XX.)
Diatomaceous mud. The largest sand grains in this sample attain a diameter of 0-3 mm. and grade
downwards in size. The proportion of large grains, however, is not sufficient materially to affect the
general constitution of the deposit which is closely similar to those from preceding stations. The
diatoms include species of Coscinodisciis, Thalassiosira, Cocconeis, Fragilaria, Rhizosolenia and
Corethron.

Station WS 387. 16. ii. 29. Lat. 620 49' S. Long. 57° 40' W. 640 m. (Plate XX.)
Diatomaceous mud. This sample is much coarser in grain than the preceding, many of the
angular mineral grains reaching a diameter of 0-5 mm., though the average is much less. Occasional
grains of a glauconitic mineral are noted. The mineral content is perhaps larger in proportion than in
the former samples, but the diatomaceous material is still important. The genera Coscinodisciis,
Thalassiosira, Cocconeis, Fragilaria, Corethron and Rhizosolenia are present in quantity, along with
sponge spicules. The preserving liquid is coloured green and many of the diatoms retain their
organic content.

Station WS 388. 16. ii. 29. Lat. 620 55' 30" S. Long. 570 40' W. 446 m. (Plate XX.)
Diatomaceous mud(?). This sample consists only of a small quantity of mud and a few small
pebbles. As green mud is also recorded in the Station List, the sample seems to be incomplete, and it
can only be said that the small quantity of fine material available resembles a diatomaceous mud.

Station WS 389. 16. ii. 29. Lat. 630 17' S. Long. 580 51' 05" W. 130 m. (Plate XX.)
Sand. Many of the mineral grains in this sample reach a diameter of 0-5 mm. and the deposit is
not coherent. As a rule the grains of volcanic glass are somewhat angular in shape while those of
quartz are somewhat rounded. Diatom frustules are present in the fine-grained flocculent material,
but mainly in a broken condition; some undamaged valves of Coscinodisciis, Cocconeis and Biddulphia
are noted.

Station WS 391. 17. ii. 29. Lat. 630 02' S. Long. 590 12' W. 877 m. (Plate XX.)
Diatomaceous mud. The bulk of the deposit consists of diatom frustules and small mineral grains,
in a "matrix" of flocculent material. The chief genera of diatoms are: Coscinodisciis, Thalassiosira,
Triceratium , Cocconeis, Fragilaria, Thalassiothrix, Corethron and Rhizosolenia. While the great
majority of the mineral grains are less than o-i mm. across, some reach a diameter of about 0-5 mm.
The largest are angular fragments of brown volcanic glass; quartz grains are smaller and more
rounded.

334 DISCOVERY REPORTS

Station WS 392. 17. ii. 29. Lat. 620 52' S. Long. 590 26' W. 591 m. (Plate XX.)

Dark sand. This small sample consists of sand, dark-coloured from the abundance of opaque

grains of volcanic glass which reach 0-25 mm. in diameter. Grains of quartz occur but sparingly.

Zeolitic aggregates are also noted.

Station WS 393. 17-18. ii. 29. Lat. 62°4i'S. Long. 590 41' W. Three samples from depths of
900-1000, 1051 and 1 138 m. (Plate XX.)

Diatomaceous mud. The three samples are very similar in character and are typical of diatom-
aceous mud. The usual flocculent material encloses angular sand grains (mostly below O'Ojmm. in
diameter) and a variety of diatoms. The genera Cosdnodiscus, Thalassiosira, Cocconeis, Fragilaria,
Corethron and Rhizosolenia are the most conspicuous forms. In detail the samples show noteworthy
differences. The largest mineral grains are noted in the sample from 1138 m., in which grains with a
diameter of 0-5 mm. are noted, while mineral fragments up to 0-2 mm. are common at the inter-
mediate depth. Nearly all the grains are sharply angular, and many of them are splinters of volcanic
glass.

Station WS 394. 18. ii. 29. Lat. 620 51' S. Long. 6o° 40' W. 274 m. (Plate XX.)
Terrigenous mud. The bulk of this sample consists of mineral grains, mostly angular in shape,
and ranging downwards in size from a diameter of about 0-25 mm., the average being roughly
0-05 mm. Among the mineral material are grains of quartz and vesicular volcanic glass, and occa-
sional small grains of glaucophane. There is a considerable amount of fine, flocculent matter which
contains some frustules of Cosdnodiscus and Fragilaria. The diatoms are not plentiful, hence this
deposit is classed as terrigenous mud.

Station WS 395. 19. ii. 29. Lat. 630 48' 30" S. Long. 620 26' W. 297 m. (Plate XX.)
Diatomaceous mud. This sample is a typical diatomaceous mud. The mineral grains are small,
mostly less than 0-05 mm. in diameter, though some are o-i mm. across. Green hornblende, hypers-
thene (0.1 mm.) and glaucophane (o-i mm.) are present, as well as fragments of volcanic glass.
Diatoms are present in some abundance, the largest being species of Cosdnodiscus and Triceratium
with a diameter of about 0-2 mm. The genera Thalassiosira, Cocconeis, Fragilaria, Corethron and
Rhizosolenia are also represented. A single example of a sphaeroid radiolarian is noted.

Station WS 396. 19. ii. 29. Lat. 630 38' 30" S. Long. 620 28' 30" W. 318 m. (Plate XX.)
Diatomaceous mud. The coarser fraction contains angular mineral grains, some of them more
than 0-25 mm. in diameter, but the great majority are much smaller. Grains of quartz, green horn-
blende and glaucophane are fairly common in occurrence together with splinters of volcanic glass.
Diatom frustules are plentiful in the finer fraction, Cosdnodiscus, Thalassiosira, Asteromphalus,
Cocconeis and Fragilaria being the most abundant genera. Rotaline Foraminifera are occasionally
seen. The flocculent material is of the usual type.

Station WS 397. 19. ii. 29. Lat. 630 29' 25" S. Long. 620 37' W. 150 m. (Plate XX.)
Sand. This sample is a small quantity of clean sand of which the grains are mainly between 0-25
and 0-05 mm. in diameter. Subangular to rounded in shape, the majority are quartz grains, but some
are opaque. Several grains of glaucophane are noted, but the sample is too small for any judgment to
be pronounced on its relative abundance in the deposit.

Station WS 399. 20. ii. 29. Lat. 620 50' S. Long. 61° 58' 30" W. 738 m. (Plate XX.)
Sand. Two fractions are somewhat sharply differentiated in this small sample. The coarser con-
stituents are mineral grains varying between 0-5 and o-i mm. in diameter. Some angular grains are
fragments of brown volcanic glass, which is sometimes vesicular in character. There are also occa-
sional grains of glaucophane. The finer fraction consists largely of small angular fragments of quartz
mingled with flocculent material in which some diatoms are present. Cosdnodiscus, Biddulphia,
Fragilaria and Rhizosolenia are the genera noted.

SEA-FLOOR DEPOSITS. PART I 335

Station WS 400. 21. ii. 29. Lat. 620 07' S. Long. 620 33' W. 4517 m. (Plate XX.)
Diatomaceous mud. This brownish grey mud contains a considerable proportion of inorganic
material, unworn and angular mineral grains varying from o-i mm. in diameter to exceedingly small
dimensions. Besides quartz, grains of green hornblende, glaucophane and hypersthene are noted,
some of the latter showing crystal faces. There is much flocculent material which consists largely
of broken frustules of diatoms in all stages of disintegration. Whole frustules of Coscinodiscus,
Thalassiosira, Cocconeis and Fragilaria are of common occurrence.

Station WS 403. 22-23. ii. 29. Lat. 59° 41' S. Long. 640 35' W. 3721m. (Plate XVII.)
Diatomaceous mud. The coarser fraction contains mineral grains up to o-i mm., mostly angular
in condition ; quartz, green hornblende and occasional grains of glaucophane are among the minerals
present. There is also some quantity of large tests of Globigerina and rotaline Foraminifera, together
with radiolarian tests. The finer fraction consists of flocculent material which includes whole and
broken frustules of diatoms, Coscinodiscus, Fragilaria, Cocconeis, Thalassiothrix being prominent
genera.

This sample is not easy to classify. The coarse washing might almost be a Globigerina ooze, but
the proportion of mineral grains is too large. The finer material, which forms the greater proportion
of the deposit, has the typical constitution of diatomaceous mud. The deposit is therefore classified
as such, despite the unusual abundance of Foraminifera.

Station WS 406. 24. ii. 29. Lat. 560 50' 30" S. Long. 670 03' W. 1234 m. (Plate XVII.)
Foraminiferal sand. The sample is a small quantity of sand which consists of quartz grains and
tests of Foraminifera. Among the latter Globigerina together with rotalines and textularians are
plentiful.

Station WS 428. 29. vi. 29. Lat. 530 07' S. Long. 420 30' W. 1966 m. (Plate XVII.)
Diatomaceous mud. There is a considerable proportion of diatomaceous debris containing entire
frustules of Coscinodiscus, Cocconeis, Fragilaria and Thalassiothrix, together with sponge spicules and
occasional radiolarian tests. The finer particles form flocculent aggregates. The grains of the
mineral constituents vary in diameter from 0-25 mm. down to exceedingly small dimensions, and are
generally angular or subangular in shape. Most of the colourless fragments are of quartz, and some
of the opaque grains appear to be glauconite.

Station WS 429. 30. iv. 29. Lat. 530 02' 30" S. Long. 450 28' W. 2549 m. (Plate XVII.)
Diatom ooze. This is a typical diatom ooze, formed almost entirely of frustules among which
species of Coscinodiscus, Thalassiosira, Cocconeis, Achnanthes, Rhizosolenia and Thalassiothrix are
noted. Fragments of radiolarian tests are also present. Mineral grains occur in small quantity; they
are mostly very small but occasionally grains of quartz and prisms of hornblende reach a length of
o-i mm.

Station WS 433. 5-6. v. 29. Lat. 5i°44'S. Long. 560 23' W. 1035 m. (Plate XVII.)
Glauconitic mud. The sediment has formed layers in the storage bottle, dark sand below, light
brown mud above, and the liquid is coloured green. The coarser fraction consists of mineral grains
up to 0-25 mm. across. Many of these are rounded grains of glauconite, some of them infilling broken
foraminiferal tests; there are, however, many shells which do not contain glauconite. Grains of
quartz and hornblende are also noted. Foraminifera are represented by Globigerina and rotalines.
The finer fraction consists of small mineral grains, broken sponge spicules and flocculent matter.

Station WS 468. 9-10. xi. 29. Lat. 550 52' S. Long. 560 53' W. 4344 m. (Plate XVII.)
Diatomaceous mud. This deposit contains a large proportion of mineral grains, some of which
reach a diameter of 0-2 mm. Quartz and green hornblende are conspicuous in angular grains, but
some fragments are well rounded. Occasional grains of garnet are noted. The finer material consists of
organic debris, sponge spicules and diatom frustules; most of the latter are fragmentary, but some
whole frustules of Coscinodiscus occur. Broken tests of Globigerina are present.

336 DISCOVERY REPORTS

Station WS 469. 10. xi. 29. Lat. 560 42' S. Long. 570 00' W. 3959 m. (Plate XVII.)
Diatomaceous mud. Essentially similar to the foregoing sample but perhaps containing more
broken tests of Globigerina.

Station WS 470. 11. xi. 29. Lat. 570 50' S. Long. 570 27' W. 3572 m. (Plate XVII.)
Diatomaceous mud. The sample consists of a few black pebbles with a small quantity of sediment.
The latter is composed mainly of quartz grains which average about 0-05 mm. in diameter, but some
reach 0-2 mm. and more, together with occasional grains of green hornblende. A colourless mineral,
in cross-shaped twin crystals about 0-05 mm. in length, with low refractive index (about 1-50) and
low birefringence is referred to phillipsite. Sponge spicules and the diatoms Coscinodiscus and
Fragilaria are comparatively plentiful.

Station WS 471. 12. xi. 29. Lat. 580 53' S. Long. 57" 54' W. 3762 m. (Plate XVII.)
Diatomaceous mud. This deposit, fine grained and brown in colour, is typical of its class. Mineral
grains are usually small (less than 0-05 mm.), though some reach a diameter of 0-2 mm. ; quartz,
green hornblende and hypersthene are noted. Diatoms are present in fair variety, including species of
Coscinodiscus, Cocconeis, Fragilaria and Thalassiothrix. Comminuted frustules form part of the
finest material, the particles of which often cohere in flocculent masses. Radiolarian and foraminiferal
tests are seen occasionally.

Station WS 472. 12. xi. 29. Lat. 590 42V S. Long. 580 01' W. 3580 m. (Plate XVII.)
Diatomaceous mud. There is a fair amount of flocculent material composed of diatom frustules
and other organic debris; the chief genera are Coscinodiscus and Fragilaria. Mineral grains, chiefly
quartz and green hornblende, vary in size from o-i mm. to less than o-oi mm. in diameter.

Station WS 474. 13. xi. 29. Lat. 6i° 03' S. Long. 56° 42' W. 2813 m. (Plate XVII.)
Diatomaceous mud. The coarser fraction consists of mineral grains, ranging from 0-5 to less than

001 mm. in diameter. Angular to rounded grains of quartz and green hornblende are prominent.

Among the diatoms, Coscinodiscus is large (up to 0-2 mm.) and abundant. Other genera include

Thalassiosira and Fragilaria. Sponge spicules and broken frustules of diatoms contribute to the

flocculent aggregates of the finer material.

Station WS 475. 14. xi. 29. Lat. 6i° 48' S. Long. 550 51' W. 1047 m. (Plate XX.)
Diatomaceous mud. The chief feature of this dark grey mud is the fine-grained fraction, which
consists almost entirely of diatom remains — chiefly fragmentary frustules; but many entire tests
occur, some of which retain their greenish cell contents. Species of Coscinodiscus (up to 0-2 mm.
diameter), Cocconeis, Fragilaria, Rhizosolenia are the most abundant forms. Mineral grains up to
0-25 mm. in diameter form a subordinate fraction of the deposit; the quartz grains are more or less
rounded, but splinters of vesicular volcanic glass are sharply angular; there are also flakes of white
mica, 0-25 mm. across.

Station WS 476. 14. xi. 29. Lat. 620 16' S. Long. 580 18' W. 542 in. (Plate XX.)
Terrigenous mud. In this dark grey mud organic remains are scanty, and chiefly consist of small
diatom frustules in the finer material. The coarser fraction consists of subangular mineral grains the
size of which ranges downwards from a diameter of about 0-2 mm. Quartz, felspar, green hornblende
and volcanic glass are noted.

Station WS 477. 14. xi. 29. Lat. 620 2o£' S. Long. 580 14' W. 1892 m. (Plate XX.)
Terrigenous mud. The bulk of this sample is formed of small angular mineral grains whose

diameter is less than o-oi mm., but there is a fair proportion of larger grains up to about o-i mm.

across. Hence the deposit is classed as a terrigenous mud although diatoms are by no means rare.

The genera Coscinodiscus, Cocconeis, Fragilaria, Rhizosolenia and Corethron are represented. The

mineral constituent includes grains of quartz, green hornblende and volcanic glass.

SEA-FLOOR DEPOSITS. PART I 337

Station WS 479. 15. xi. 29. Lat. 620 32}' S. Long. 57° 55' W. 1523 m. (Plate XX.)
Diatomaceous MUD. This dark grey mud consists mainly of diatom frustules together with
flocculent aggregates of organic debris; many of the frustules retain their protoplasmic content.
Coscinodiscus, Thalassiosira, Cocconeis, Fragilaria, Corethron and Rhizosolenia are the chief genera.
Mineral grains are comparatively rare in occurrence and small in size, being mostly less than o-oi mm.
in diameter, though some are o- 1 mm. across ; most of them are angular fragments of quartz, but green
hornblende also is noted.

Station WS 480. 16. xi. 29. Lat. 620 51 i' S. Long. 570 47J' W. 740 m. (Plate XX.)
Diatomaceous mud. The sample of greenish mud contains well- developed diatom frustules in fair
abundance and variety. Coscinodiscus, which preponderates among the centric forms, is associated
with Thalassiosira, Biddulphia and Cocconeis, while Rhizosolenia, Corethron and Fragilaria represent
the elongate and pennate forms repectively. Mineral grains occur in fair quantity, mostly angular or
subangular in shape, and usually less than o- 1 mm. in diameter ; quartz, green hornblende and volcanic
glass are noted.

Station WS 481. 16. xi. 29. Lat. 620 59' S. Long. 57° 28' W. 453 m. (Plate XX.)
Sand. The bulk of this sample consists of mineral grains which range from a diameter of about
0-5 mm. to exceedingly small dimensions. Quartz, felspar, green hornblende, white mica and vol-
canic glass are the most prominent constituents. Among the organic constituents are Polyzoa,
Foraminifera and sponge spicules. The supernatant material forms flocculent aggregates which
enclose and buoy up diatoms and small mineral grains. The diatom genera include Coscinodiscus,
Triceratium, Thalassiosira and Grammatophora.
Station WS 482. 16. xi. 29. Lat. 630 10' S. Long. 570 i6|' W. 152 m. (Plate XX.)
Diatomaceous mud. The deposit has settled in layers during storage ; the top layer is a fine-grained
greenish mud, the lower portion is darker in colour (almost black) and coarser in grain. The two
fractions are approximately equal in bulk. The finer material consists largely of diatom debris, often
forming flocculent aggregates in which small angular mineral grains (mostly less than 001 mm. in
diameter) are enclosed. A considerable number of entire diatom frustules occur, among which
specimens of Triceratiutn are conspicuous by their size (0-5 mm. across); they are associated with
species of Coscinodiscus, Thalassiosira, Cocconeis, Fragilaria and Grammatophora. Some of the
frustules still have their cell contents, and the preserving liquid has a green tinge. Monaxid and
tetraxid sponge spicules are present. In the coarser material, mineral grains reach a diameter of
0-5 mm. but the average size is about 0-2 mm. ; the grains are chiefly angular to subangular. The
chief constituents are quartz, green hornblende and volcanic glass. While the proportion of sandy
material is considerable, character is given to the deposit by the diatoms ; hence it is classed as a
diatomaceous mud.

Station WS 483. 21. xi. 29. Lat. 620 46! ' S. Long. 590 37-^ W. 1420 m. (Plate XX.)
Diatomaceous mud. This grey-green mud consists largely of flocculent aggregates formed of
diatom debris and enclosing a quantity of mineral grains which are usually less than 0-05 mm. in
diameter. Entire frustules of Corethron, Coscinodiscus, Cocconeis, Fragilaria, Biddulphia, Rhizoso-
lenia and Thalassiosira are noted, the first named being particularly abundant. The chief mineral
constituents are quartz, green hornblende and volcanic glass, splinters of the latter being especially
prominent.

Station WS 484. 21. xi. 29. Lat. 620 54' S. Long. 590 28' W. 1008 m. (Plate XX.)
Diatomaceous mud. This green mud is eminently typical of its class. Frustules of Coscinodiscus,
Cocconeis, Fragilaria Rhizosolenia and Corethron occur in an undamaged condition, and also in all
stages of fragmentation. The flocculent aggregates which compose the bulk of the deposit enclose
angular mineral grains which are usually less than 0-05 mm. in diameter, but the size ranges up to
about o-i mm. Angular fragments of volcanic glass are conspicuous, along with grains of quartz,
green hornblende and white mica, in the mineral fraction.

6-2

338 DISCOVERY REPORTS

Station WS 485. 21. xi. 29. Lat. 630 02|' S. Long. 590 17' W. 805 m. (Plate XX.)
Diatomaceous mud. This greenish deposit agrees so closely with that from St. WS 484 that the
same description applies.

Station WS 486. 21. xi. 29. Lat. 630 ui' S. Long. 590 13' W. 7S7 m. (Plate XX.)
Diatomaceous mud. Closely similar to the foregoing.

Station WS 487. 22. xi. 29. Lat. 630 17' S. Long. 59° 20' W. 790 m. (Plate XX.)
Diatomaceous mud. This also agrees with the last three samples.

Station WS 488. 22. xi. 29. Lat. 63° 51A' S. Long. 620 31' W. 220 m. (Plate XX.)
Terrigenous mud. The character of this deposit is given by its inorganic constituent. The
mineral grains range in size from about 0-5 mm. downwards to exceedingly small dimensions. The
larger grains are fairly well rounded, but a large proportion of the smaller grains are angular. Apart
from quartz, green hornblende seems to be the most abundant mineral ; hypersthene and glaucophane
also occur, but the last named only rarely. Organic remains, present only in subordinate quantity,
include sponge spicules and diatoms; among the latter Coscinodiscus, Cocconeis, Thalassiosira and
Fragilaria are noted.

Station WS 489. 22. xi. 29. Lat. 630 38' S. Long. 620 32' W. 308 m. (Plate XX.)
Diatomaceous mud. This is a greenish grey mud which contains the usual admixture of diatom
debris and mineral grains. The latter are mainly angular in shape, and apart from the ubiquitous
quartz (up to o-i mm. diameter) include green hornblende (0-05 mm.) and rich blue glaucophane
(o-i mm.) in some abundance; the last-named mineral shows the characteristic pleochroism (violet
to blue). The diatoms include Coscinodiscus (up to 0-25 mm. across) Thalassiosira, Cocconeis, Fragi-
laria, Rhizosolenia. Textularian Foraminifera and sponge spicules are also noted.

Station WS 490. 22. xi. 29. Lat. 630 24I' S. Long. 620 35 J' W. 262 m. (Plate XX.)
Sand. This greyish deposit is best classed as a sand, though it contains a fair quantity of fine-
grained flocculent material. The bulk of the sample consists of quartz grains ranging downwards in
size from a diameter of about 0-5 mm. Occasional pleochroic prisms of glaucophane (o-i mm. long)
and green hornblende (o-2 mm. maximum) are noted and doubtless other minerals are present.
Sponge spicules and diatoms occur, but are not sufficiently abundant to give character to the deposit.

Station WS 493. 23. xi. 29. Lat. 620 51' S. Long. 6o° 34' W. 220 m. (Plate XX.)
Sand. This is a dark-coloured sand, with grains from 0-5 to 0-05 mm. in diameter, the average
being about 0-25 mm. Most of the grains are brown in colour by transmitted light and isotropic
between crossed nicols ; they are sharply angular fragments of glassy and vesicular volcanic material.
There is only a small proportion of rounded quartz grains.

Station WS 494a. 28. xi. 29. Lat. 630 15' S. Long. 6i° 05' W. 1035 m. (Plate XX.)
Sand. The character of the detrital material outweighs in importance the organic constituent, and
the size of the grains determines the classification of this sample as a sand. The mineral grains include
angular and subangular fragments of quartz (up to 0-3 mm. diameter), sharply angular splinters of
volcanic glass (up to 0-5 mm.), occasional grains of green hornblende (0-15 mm.) and glaucophane
(0-2 mm.). The fine-grained flocculent material encloses frustules of Coscinodiscus, Cocconeis,
Franlaria and Rhizosolenia.

"6*

Station WS 494b. 28. xi. 29. Lat. 630 37J' S. Long. 6i° 16' W. 505 m. (Plate XX.)
Sand. This is a greenish sandy deposit which owes its colour to the fine-grained flocculent con-
stituent, which is largely diatomaceous in character. The bulk of the deposit, however, consists of
mineral grains up to 0-5 mm. in diameter. Angular and rounded grains of quartz, sharp splinters of
volcanic glass, and prismatic grains of green hornblende (0-25 mm. long) are abundant, while occa-
sional small grains of glaucophane occur.

SEA-FLOOR DEPOSITS. PART I 339

Station WS 495. 22. xii. 29. Lat. 670 47' S. Long. 730 51' W. 2582 m. (Plate XXI.)
Diatomaceous mud. This is a brownish mud formed largely of flocculent material which encloses
diatom frustules and mineral grains. The latter are mostly below 0-05 mm. in diameter, though some
reach 0-2 mm. Rounded and subangular grains of quartz are most abundant, while green hornblende
and white mica are also noted. The chief diatom genera are Coscinodiscus and Fragilaria. Some tests
of Radiolaria include Sethoconus and fragments of other genera.

Station WS 496. 30. xii. 29. Lat. 670 14' S. Long. 700 12' W. 631m. (Plate XXI.)
Diatomaceous mud. An unctuous green-grey mud, composed largely of diatom debris, among
which undamaged frustules of Cosctnodiscus, Thalassiosira, Fragilaria, Cocconeis, Corethron (often
with chloroplasts) and Biddulphia are plentiful. A few Radiolaria and sponge spicules are noted.
There is a fair proportion of mineral grains, mostly subangular in shape, up to o-i mm. in diameter,
including quartz, white mica and green hornblende.

Station WS 497. 1. i. 30. Lat. 670 05' S. Long. 700 40' W. 534 m. (Plate XXI.)
Diatomaceous mud. A grey unctuous mud with the same general characters as the preceding
sample. The diatoms include Coscinodiscus, Thalassiosira, Fragilaria, Rhizosoletiia, Cocconeis and
Biddulphia. A few Radiolaria and rotaline Foraminifera are seen. Mineral grains reach a diameter of
about 0-15 mm. and include subangular fragments of quartz with some prismatic grains of green
hornblende.

Station WS 498. 2-3. i. 30. Lat. 66° 21' S. Long. 690 01' W. 398 m. (Plate XXI.)
Sand. The mineral grains in the coarser fraction reach a diameter of about 0-5 mm. Angular
grains of quartz predominate over prisms of green hornblende, grains of hypersthene and fragments
of volcanic glass. Flocculent material is almost equal in bulk and includes diatom remains ; but only
frustules of Fragilaria are identified.

Station WS 499. 3. i. 30. Lat. 650 45' S. Long. 67° 18' W. 179 m. (Plate XXI.)
Sand. The mineral grains range downwards in size from about 0-25 mm. in diameter. They in-
clude subangular grains of quartz, green hornblende and pink garnet. The fine-grained flocculent
material is largely diatomaceous in origin ; the chief genera are : Coscinodiscus, Fragilaria, Biddulphia,
Cocconeis, Thalassiosira and Corethron.

Station WS 501. 3. i. 30. Lat. 640 52' S. Long. 630 58' W. 583 m. (Plate XXI.)
Diatomaceous mud. This greenish mud consists largely of flocculent aggregates formed of
diatoms in all stages of disintegration along with other organic debris ; the liquid also is coloured
green. Whole frustules of the genera Triceratium (0-5 mm. across), Coscinodiscus, Thalassiosira,
Fragilaria, Cocconeis, Corethron are plentiful. Mineral grains form a subordinate proportion of the
deposit ; they are chiefly of quartz, but occasional grains of green hornblende and possibly of chlorite
also occur. The grains are mainly below 0-05 mm. in diameter, but some fragments are 0-2 mm.
across.

Station WS 502. 30.1.30. Lat. 69°43'S. Long. 990 38' W. 4224 m. (Plate XXI.)

Diatomaceous mud. This sample, brown in colour, is composed mainly of flocculent material of

the usual appearance. It contains whole frustules of Fragilaria and Coscinodiscus, together with

fragments of Radiolaria. Mineral grains of varying size (o-2 mm. to less than o-oi mm. in diameter)

consist chiefly of angular and rounded fragments of quartz with occasional grains of green hornblende.

Station WS 503. 30. i. 30. Lat. 700 03$' S. Long. ioo° 39' W. 4072 m. (Plate XXI.)
Terrigenous mud. The sample is grey and coherent when dry. It consists largely of angular or
subangular mineral grains up to about 0-2 mm. in diameter. Quartz is most abundant, and some
grains show undulose extinction in polarized light. Fragments of green hornblende are not un-
common. There is some amount of flocculent material which forms aggregates enclosing mineral
grains, but the sample is singularly deficient in recognizable organic remains, except for a few sponge
spicules and frustules of Coscinodiscus which have an eroded appearance.

34o DISCOVERY REPORTS

Station WS 505. 4. ii. 30. Lat. 700 10J' S. Long. 870 46' W. 1500 m. (Plate XXI.)
Terrigenous mud. A brown mud composed of flocculent matter which encloses mineral grains.
The grains grade downwards in size from a diameter of about o-i mm. ; they are fairly angular in
shape and consist chiefly of quartz with occasional fragments of green hornblende. There are few
recognizable organisms, only a few tests of Cristellaria and fragments of Globigerina being noted.

Station WS 506. 7. ii. 30. Lat. 70° 31' S. Long. 81° 36' W. 584 m. (Plate XXI.)
Terrigenous mud. A brown unctuous mud with the same general characters as the preceding
sample.

Station WS 507a. 8. ii. 30. Lat. 700 32J' S. Long. 8i° 42' W. 572 m. (Plate XXI.)
Terrigenous mud. This sample resembles those from Sts. WS 505 and WS 506. The bulk of the
mineral grains are less than o-oi mm. in diameter, but some larger grains (o-i mm.) are of quartz
and green hornblende. The flocculent material encloses diatoms (frustules of Coscinodiscus), but
they are not plentiful.

Station WS 507b. 8. ii. 30. Lat. 700 34' S. Long. 8i° 55' W. 580 m. (Plate XXI.)
Terrigenous mud. This deposit is closely similar to the last in general constitution. The larger

mineral grains (o- 1 mm. diameter) include quartz and green hornblende. Frustules of Coscinodiscus,

enclosed in flocculent aggregates, still retain their cell contents.

Station WS 508. 10. ii. 30. Lat. 690 04' S. Long. 770 40' W. 309 m. (Plate XXI.)

Fine gravel. The sample consists of a few small rounded pebbles, about 5 mm. in diameter.

Station WS 509. 11. ii. 30. Lat. 670 18' S. Long. 690 28' W. 445 m. (Plate XXI.)
Diatomaceous mud. The flocculent material in this sample is definitely diatomaceous. It contains
many frustules of Coscinodiscus, Cocconeis, Thalassiosira and Fragilaria which are rather small, but
also larger ones (including Triceratium) which retain their cell contents. The mineral grains, ranging
downwards in size from 0-2 mm. in diameter, include angular fragments of quartz and green horn-
blende.

Station WS 510. 11. ii. 30. Lat. 670 11' S. Long. 690 46' W. 505 m. (Plate XXI.)
Diatomaceous mud. This is a green-grey mud in which flocculent diatomaceous material is
dominant, and many of the frustules still enclose greenish cell contents. The chief genera are
Coscinodiscus, Fragilaria, Thalassiosira and Cocconeis. The mineral constituent consists mainly of
subangular grains of quartz (up to o-i mm. diameter) with occasional prismatic grains of green
hornblende.

Station WS 511. n. ii. 30. Lat. 670 04' S. Long. 700 04' W. 635 m. (Plate XXI.)
Diatomaceous mud. A greenish grey mud which is mainly flocculent diatom debris. Whole
frustules of Coscinodiscus (0-2 mm.), Thalassiosira, Fragilaria, Cocconeis and Corethron occur along
with sponge spicules. Mineral grains (quartz and green hornblende) are usually angular and very
small, but some reach a diameter of 0-05 mm.

Station WS 512. 11. ii. 30. Lat. 66° 57' S. Long. 700 22' W. 652 m. (Plate XXI.)
Diatomaceous mud. Flocculent diatom debris makes up the bulk of this greenish grey mud, in

which whole frustules of Coscinodiscus, Cocconeis, Fragilaria, Rhizosolenia and Thalassiosira are noted.

There is a subordinate proportion of subangular mineral grains, mostly less than 0-05 mm. in

diameter, among which quartz and green hornblende are recognized.

Station WS 513. 11. ii. 30. Lat. 66° 49^ S. Long. 700 40J' W. 560. m. (Plate XXI.)
Diatomaceous mud. A greenish mud of the same type as the preceding sample. Frustules of

Coscinodiscus, Thalassiosira, Cocconeis, Biddulphia, Fragilaria, Corethron and Rhizosolenia are noted.

Quartz grains form the greater part of the mineral fraction, while green hornblende also is present.

SEA-FLOOR DEPOSITS. PART I 34»

Station WS 514. ii.ii. 30. Lat. 66° 40!' S. Long. 71° oi'W. 531m. (Plate XXI.)
Diatomaceous mud. The sample contains flocculent material of the same type as the foregoing
sample but less abundantly. It encloses frustules of Coscinodiscus, Thalassiosira, Biddulphia and
Fragilaria. The mineral grains attain a larger size, angular grains of quartz and green hornblende
reaching o-2 mm., and dark opaque grains 0-25 mm. in diameter.

Station WS 515. ii.ii. 30. Lat. 66° 3^' S. Long. 71° aoi'W. 512 m. (Plate XXI.)
Diatomaceous mud. The deposit is of the same type and contains the same diatoms as the pre-
ceding samples. Mineral grains are perhaps more abundant ; angular grains of quartz reach 02 mm.
and green hornblende o-i mm. in diameter.

Station WS 516. 12. ii. 30. Lat. 66° 25!-' S. Long. 71° 38J' W. 261 1 m. (Plate XXI.)
Sand. The sample has separated into layers during storage, brown mud above, and dark-coloured
sand below.

The former consists of tiny mineral grains usually less than o-oi mm. in diameter, together with
plentiful tests of Radiolaria, occasional whole frustules of diatoms, and a fair proportion of frag-
mentary tests. The coarser fraction contains mineral grains up to 1 mm. in diameter, chiefly angular
grains of quartz and opaque rock fragments. These give character to the deposit which must be
classed as a sand.

Station WS 517. 12. ii. 30. Lat. 66° 17J' S. Long. 71° 57' W. 2770 m. (Plate XXI.)
Diatomaceous mud. This brown mud consists largely of detrital mineral grains, which are mostly
below 0-05 mm. in diameter though occasional grains are 0-25 mm. across. Angular fragments of
quartz (up to 0-25 mm.) and prismatic grains of green hornblende (o-i mm. long) are the most
abundant minerals. The finer fraction contains a large proportion of tiny angular grains, mingled
with flocculent material which consists partly of diatom debris. Recognizable organic remains are not
abundant, but they include the diatoms Fragilaria, Cocconeis and Coscinodiscus, along with fragments
of radiolarian tests. The diatomaceous material is sufficiently important to be taken into account
with the mineral content in classifying this deposit.

Station 518. 27. ii. 30. Lat. 51° 55V S. Long. 55° 35' W. 1258 m. (Plate XVII.)
Glauconitic mud. This is a greyish deposit with dark-coloured sand grains. The coarser fraction
consists of foraminiferal shells, fragments of quartz and grains of glauconite. The latter reach a
diameter of about 0-5 mm., and they retain the shape of foraminiferal tests which they once infilled ;
Globigerina and textularian forms are represented, and portions of the shell still adhere to some of the
grains. The quartz grains are usually smaller. Unbroken tests of Globigerina, Orbulina and rotalines
are plentiful. The finer fraction consists of comminuted foraminiferal shells together with tiny
quartz grains, broken sponge spicules and occasional Radiolaria, the whole often forming flocculent
aggregates.

Station WS 520. 27-28. ii. 30. Lat. 52° 25' S. Long. 51° 20' W. 3128 m. (Plate XVII.)
Globigerina ooze. A small quantity of light-coloured deposit with dark grains. The bulk of the
sample consists of well-developed tests of Orbulina and Globigerina, up to o-i mm. in diameter, many
of which are undamaged. The dark grains, with the same average diameter, are glauconitic moulds of
Globigerina from which the shell has dissolved ; the septa of the test remain, however, so that the
mould can be identified. There are also angular grains of quartz which are usually smaller than the
Foraminifera. The small proportion of flocculent matter consists of coccoliths, rhabdoliths, and tiny
fragments of disintegrated Globigerina shells, the aggregates often enclosing small mineral grains.

Station WS 521. 28. ii. 30. Lat. 52° 41' S. Long. 49° 14' W. 3780 m. (Plate XVII.)
Globigerina ooze. This is similar in general characters to the preceding sample, but the propor-
tion of mineral grains is larger. Many of the grains reach a diameter of o-i mm., while a few are
0-4 mm. across. The proportion of fine flocculent material is perhaps greater than in the former
sample and a few diatoms are present.

342 DISCOVERY REPORTS

Station WS 522. 28. ii-i. iii. 30. Lat. 52° 56' S. Long. 470 14' W. 2550 m. (Plate XVII.)
Globigerina ooze. The coarser fraction, which constitutes the bulk of the deposit, has the charac-
ters of a Globigerina ooze, with some admixture of terrigenous material. Some sand grains are
0-5 mm. in diameter; the minerals represented include angular fragments of quartz and red garnet,
and some opaque grains of glauconite with rounded and lobate outlines. There are also a few small
pebbles, 1 cm. across. The finer material has a fair proportion of diatom frustules (Coscinodiscus,
Fragilaria, Rhizosolenia), with coccoliths and occasional fragments of Radiolaria. A few cross-twins
of the mineral phillipsite are noted.

The main character of the deposit is given by the comparatively large and abundant calcareous
organisms, and hence it is classed as Globigerina ooze.

Station WS 524. 2. iii. 30. Lat. 530 36' S. Long. 430 00' W. 1697 m. (Plate XVII.)
Diatomaceous mud. The coarser fraction contains most of the mineral grains, while the finer
material is almost entirely composed of diatom frustules. The latter are present in some variety,
Coscinodiscus, Arachnoidiscus, Thalassiosira, Fragilaria, Rhizosolenia and Thalassiothrix being the
chief genera represented. The mineral grains are subangular to rounded in shape and the great
majority are about o-i mm. in diameter; quartz (o-2 mm.), green hornblende and felspar are noted.
A few tests of Globigerina also occur in the coarser material.

Station WS 525. 2. iii. 30. Lat. 530 38J-' S. Long. 41 ° 09' W. 162 m. (Plate XVII.)
Gravel (?). The sample consists only of a few small pebbles and grains of quartz.

Station WS 526. 3. iii. 30. Lat. 530 51' S. Long. 390 45' W. 1545 m. (Plate XVII.)
Diatomaceous mud. This greenish mud owes its colour to the presence of green flocculent matter
of organic origin ; the preserving liquid has acquired a green tinge. There is a variety of diatom frustules
including species of Coscinodiscus, Fragilaria, Biddulphia, Rhizosolenia and Thalassiothrix. A few
nasselarian Radiolaria are noted. The coarser fraction consists mainly of subangular quartz grains
up to about o-i mm. diameter.

Station WS 591. 18. v. 31. Lat. 350 47' S. Long. 720 39' W. 27 m. (Plate XXII.)
Sand. This is a "Bottom sample from armed lead", consisting of a small quantity of sand, the
grains of which average about 0-3 mm. in diameter. The material includes grains of hypersthene
(0-3 mm.), besides quartz and opaque grains ; these are mostly angular or only slightly rounded. There
is no fine-grained material.

Station WS 596. 18. v. 31. Lat. 350 35' 36" S. Long. 730 07' 30" W. 369 m. (Plate XXII.)
Terrigenous mud. This dark green deposit has layered out during storage. The coarser fraction
consists of mineral grains which average about 0-2 mm. in diameter; they are mainly angular frag-
ments of quartz, but some are grains of hypersthene and green hornblende. The finer fraction is a
flocculent material of rather granular appearance which remains suspended in the green liquid for
some considerable time after disturbance. No organisms are recognized except sponge spicules, but
the flocculent material generally resembles organic debris; it encloses tiny mineral grains. The
detrital mineral constituent gives the main character to the deposit.

Station WS 597. 19. v. 31. Lat. 350 39' 42" S. Long. 730 19' 30" W. 1593 m. (Plate XXII.)
Terrigenous mud. A dark green mud, consisting of somewhat granular flocculent matter, in clots
of which small sand grains are embedded. The latter are mostly less than 0-05 mm., but occasionally
angular grains of quartz and green hornblende reach a diameter of o-i mm. There is a general
paucity of recognizable organic remains but occasional diatoms (Coscinodiscus), rotaline Foraminitera,
sphaeroid Radiolaria and sponge spicules are seen.

Station WS 598. 19. v. 31. Lat. 350 43' S. Long. 730 32' W. 2307 m. (Plate XXII.)
Diatomaceous mud. This is a green mud formed of flocculent material enclosing small quartz
grains which are generally less than 0-05 mm. in diameter. The organic debris is granular in

SEA-FLOOR DEPOSITS. PART I 343

appearance and contains cells of green algae and diatoms, among which frustules of Coscinodiscus,
Rhizosolenia and Chaetoceros are noted.

Station WS 599. 19. v. 31. Lat. 35° 41' 30" S. Long. 730 43' W. 3265 m. (Plate XXII.)
Diatomaceous mud. This is essentially similar in general constitution to the preceding sample.

Some fragments of green hornblende are noted among the more abundant quartz grains. The

diatoms include frustules of Coscinodiscus and Chaetoceros.

Station WS 602. 28. v. 31. Lat. 320 04' 45" S. Long. 710 34' W. 75 m. (Plate XXII.)
Shelly sand. This sample consists largely of calcareous material which includes the brachiopod
Magellania flavescens (one entire shell), echinid spines, sponge spicules, fragments of gastropod
shells, Polyzoa, pteropods and Foraminifera. Mineral grains, chiefly angular fragments of quartz, reach
a diameter of 0-25 mm. Crystals of phillipsite (o-i mm.) are plentiful among the finer material.

Station WS 604. 28. v. 31. Lat. 320 05' S. Long. 710 45' 30" W. 687 m. (Plate XXII.)
Terrigenous mud. A dark green mud, formed mainly of angular mineral grains of which a large
proportion are about o-i mm. in diameter. The minerals include quartz, white mica, green horn-
blende (some fibrous), staurolite (?) and volcanic glass. There is a small proportion of flocculent
material which appears to be organic debris though no recognizable remains are noted.

Station WS 605. 28. v. 31. Lat. 32°os'S. Long. 71° 50' W. 1296 m. (Plate XXII.)
Terrigenous mud. A brownish mud, consisting mainly of mineral grains, but with some flocculent
matter. The great majority of the grains are less than o-i mm. in diameter. The larger grains in-
clude angular fragments of quartz (0-25 mm.), prismatic grains of green and brown hornblende
(o-i mm.), flakes of white mica (0-2 mm.) showing strain shadows, grains of staurolite and volcanic
glass. Rotaline and textularian Foraminifera, together with sponge spicules are sparsely distributed
through the deposit.

Station WS 610. 30. v. 31. Lat. 310 45' 30" S. Long. 720 01' 30" W. 2500 m. (approx.).
(Plate XXII.)

Diatomaceous mud. This is a light brown mud in which the mineral constituent is subordinate in
proportion to the flocculent material. The latter is of the usual indefinite character, but it includes
frustules of Coscinodiscus together with occasional fragments of Radiolaria, milioline Foraminifera and
sponge spicules. The mineral grains are mostly below 0-05 mm. in diameter, and include fragments
of quartz and green hornblende besides splinters of volcanic glass.

Station WS 616. 5. vi. 31. Lat. 270 08' S. Long. 710 10' W. 1768 m. (Plate XXII.)
Terrigenous mud. This deposit has a large proportion of terrigenous material. Subangular grains
of quartz up to 0-2 mm. in diameter are preponderant while prismatic grains of brown and green
hornblende up to o- 1 5 mm. in length are often seen. The granular flocculent material encloses minute
mineral grains and has a greenish hue. Sponge spicules are plentiful, and occasional rotaline Fora-
minifera, diatoms (Coscinodiscus) and filaments of green algae are noted.

Station WS 617. 5. vi. 31. Lat. 270 09' 30" S. Long. 710 15' 42" W. 3031m. (Plate XXII.)
Terrigenous mud. Of similar constitution to the deposit from St. WS 616. Detrital grains include
angular fragments of quartz (0-25 mm.), prismatic green hornblende, flakes of white mica, and splinters
of volcanic glass. A few Radiolaria, sponge spicules, diatoms (Coscinodiscus) and rotaline Foramini-
fera occur, but not in sufficient quantity to give character to the deposit.

Station WS 619. 6. vi. 31. Lat. 270 03' 30" S. Long. 710 30' W. 4864 m. (Plate XXII.)
Terrigenous mud. A black fetid mud which has oxidized to a brown colour round the margin of
the sample. The black colour appears to be due to dense flocculent aggregates of exceedingly small
particles whose precise nature is uncertain. These aggregates enclose a quantity of small, angular
mineral grains. The bulk of the sample, however, consists of sand grains, the largest of which are
about 0-15 mm. in diameter. The condition of the minerals varies from well-rounded grains to sharp,

DISCOVERY REPORTS
344

j u-- ™;™ anrlm-een hornblende are noted. Apart from abundant

Radiolaria.

o ^ „(j> c T ntw 70° 38' ^o" W. 1 48 m. (Plate XXII.)

are sponge spicules and grains . quam -f^,1^*^^ ^2 1. There Je a few

ST^t!^rSfr^^ Fo^nif.ra, «es,s are pWH Ending

textularian, cristellarian and rotahne forms.

t * 0°,.,'™"^ Innn 7i° 06' W. 2277 m. (Plate XXII.)
££££!£ A o^t^/on^face of ,he sa'm'pie doring -**
UIATOM, , , „„ „f _.,.-:,». flncculent aggregates which contains occasional angular

^o'^iit^^^^^^ **■* *-■ Diatoms are notr

SEf £££ unTmagfd frustules of Coscinodiscus are noted, together with sponge spicules.

Station WS 639 10. vi. 31. Lat. 18° 44' S. Long. 70° 48' W. 1485 m. (Plate XXII.)
SgZjsmud A greenish brown mud, formed mainly of terrigenous material The mmera
erals nlde ngular fragments of quartz (o- , 5 mm.), green hornblende (o- , mm.) and volcanic ^
fn mm There is some quantity of granular flocculent matter, sometimes in dense aggregates,

of Coscinodiscus and Entopyla are present.

Station WS 641. 19.vi.31. Lat. 1 8° 27' 30" S. Long. 70° 26' 30" W. 105 m. (Plate XXII.)
DuioJSoxis mud9 This is a green mud very similar to the preceding sample. The terrigenous
materTmcludes angular grains of quartz and green hornblende, up to about 0-2 mm. in diameter
Xret a at proportion of flocculent matter, often in dense aggregates enclosing small mineral
££L Coscinodiscls again is conspicuous by the size of the frustules, and other diatoms include
Navicula, Fragilaria, Actinoptychus and Amphora.

Station WS 642. 19.vi.31. Lat. 18° 28' 24" S. Long. 7o° 32' 12" W. 148 m. (Plate XXII.)
D "omachoos uJ. A irk green mud composed mainly of dense, green, *^W*£*
which are granular in appearance. This material contains brown resting spores and plentiful diatoms
Tf he "ZTcoscinoaZs, Actinoptychus, Thalassiothrix and Fragilaria. One example of Bact^as-
trumi seen to contain brown resting spores similar to, but smaller than, the loose spores n the
nTJent material. The mineral constituent consists mainly of angular grains of quartz and volcanic
glass, embedded in the flocculent "matrix".

SEA-FLOOR DEPOSITS. PART I 345

Station WS 647. 22. vi. 31. Lat. 150 19' 12" S. Long. 750 11' 30" W. 65 m. (Plate XXII.)
Fine sand. The sample consists almost entirely of detrital mineral grains, but there is a small
amount of flocculent material which, however, is sufficient to impart a greenish tinge to the sand.
The grains are mainly angular and include quartz (0-25 mm.), plagioclase felspar (0-2 mm.), green
hornblende (o- 1 mm.) and volcanic glass (0-2 mm.). Among diatoms, large frustules of Coscinodiscus are
abundant and some contain brown resting spores. Actinoptychus commonly occurs, while Navicula,
Synedra, Grammotophora, Achnanihes, Pleurosigma and Entopyla are less numerous. Sponge spicules
and Foraminifera (cristellarians and textularians) are also present in the deposit.

Station WS 648. 22. vi. 31. Lat. 15° 19' 30" S. Long. 75° 13' W. mm. (Plate XXII.)
Diatomaceous mud. A green mud, consisting mainly of flocculent diatomaceous material with
only a small proportion of detrital mineral grains. The latter are chiefly angular grains of quartz and
green hornblende which do not average more than 0-05 mm. in diameter. Though diatoms are
plentiful, the variety of forms is not great. The largest and most abundant frustules are those of
Coscinodiscus, many of which reach a diameter of nearly 0-5 mm. ; some have greenish cell contents,
others contain brown resting spores. Other forms which are occasionally seen are Actinoptychus,
Achnanihes and Grammatophora. Brown resting spores also occur loose in the green flocculent
material.

Station WS 649. 22. vi. 31. Lat. i502o'S. Long. 750 16' 30" W. 137 m. (Plate XXII.)
Diatomaceous mud. This sample differs from the preceding chiefly in the greater maximum size
of the sand grains; in the present sample some angular quartz grains are 0-2 mm. across, and prisms
of green hornblende are nearly as long. But the mineral constituent is relatively smaller in amount
than the flocculent matter which is closely similar to that from St. WS 648. There is also the same
preponderance of large frustules of Coscinodiscus, and the associated diatoms include Actinoptychus,
Achnanthes and Navicula.

Station WS 650. 22. vi. 31. Lat. 150 22' 30" S. Long. 750 22' W. 143 m. (Plate XXII.)
Coarse sand. The major fraction consists of colourless and opaque sand grains, more or less rounded
in shape, up to 2 mm. in diameter. They are often coated with green flocculent matter. In the finer
material a large species of Coscinodiscus commonly occurs with textularian and rotaline Foraminifera.
The granular flocculent material also encloses abundant crystals of phillipsite.

Station WS 651. 22. vi. 31. Lat. i503i'S. Long. 750 37' 30" W. 1264 m. (Plate XXII.)
Diatomaceous mud. This deposit is composed mainly of green flocculent matter in which
broken tests of diatoms, and brown resting spores are abundant. A large species of Coscinodiscus
occurs plentifully, while frustules of Entopyla, Achnanthes and Actinoptychus are less numerous.
Some of the frustules enclose brown resting spores. The sand grains embedded in the flocculent
mass are mostly less than 0-05 m. in diameter, though some reach o-i mm. Angular grains of quartz
and prismatic grains of green hornblende represent the chief minerals.

Station WS 652. 23. vi. 31. Lat. 160 21' 30" S. Long. 760 30' 12" W. 3840 m. (Plate XXII.)
Diatomaceous mud. An unctuous grey-brown mud composed almost entirely of flocculent
matter, in which diatom frustules occur in all stages of disintegration. Coscinodiscus is represented
in abundance by unbroken frustules, and Radiolaria are present in small numbers. The minor
proportion of mineral grains includes quartz and green hornblende in fragments which are usually
less than 0-05 mm. in diameter.

Station WS 655. 24. vi. 31. Lat. i6°o8'S. Long. 760 22' W. 3315 m. (Plate XXII.)
Diatomaceous mud. This brown mud is so similar to the preceding that the same description
applies.

Station WS 658. 25. vi. 31. Lat. 130 45' 30" S. Long. 760 20' W. 16 m. (Plate XXII.)
Diatomaceous mud. A dark green, rather sandy, mud of which the greater proportion consists of
mineral grains, mostly angular in form and ranging from a diameter of about 025 mm. to exceedingly

346 DISCOVERY REPORTS

small dimensions. Quartz, volcanic glass and green hornblende are the chief constituents. The
flocculent material contains diatoms, among which Coscinodiscus is large and abundant, Actinoptychus
of common occurrence, while navicular forms are less plentiful.

Station WS 663. i.vii. 31. Lat. 120 09' 36" S. Long. 770 15' W. 62 m. (Plate XXII.)
Diatomaceous mud. A dark green mud of similar constitution to the preceding. The minerals are
represented by quartz, felspar, green hornblende and volcanic glass in angular grains up to o-2 mm.
in diameter. There is a fair amount of flocculent material containing diatom frustules and brown
resting spores. The diatoms noted are Coscinodiscus (abundant), Chaetoceros, Navicula.

Station WS 664. i.vii. 31. Lat. 120 11' 30" S. Long. 770 17' W. 100 m. (Plate XXII.)
Diatomaceous mud. This green mud differs from the last two samples in the greater proportion of
flocculent material. The presence of fragments of algal thalli containing resting spores and other cell
contents leaves little doubt as to the nature of the flocculent aggregates. Coscinodiscus and Actinop-
tychus are the principal diatoms; the former is particularly large and abundant, and many frustules
contain brown resting spores. Mineral grains, including quartz and green hornblende, are mostly
angular in shape, and occur up to 0-2 mm. in diameter.

Station WS 665. i.vii. 31. Lat. 120 13' 18" S. Long. 770 21' 48" W. 132 m. (Plate XXII.)
Diatomaceous mud. The coarser fraction of this dark green mud consists of detrital material
grains with a maximum diameter of about 0-3 mm. ; the average size, however, is much less. Angular
fragments of quartz, green hornblende, felspar and volcanic glass are noted. There is a considerable
amount of green flocculent material which includes pieces of algal thalli and diatoms, among which
Coscinodiscus, Actinoptychus, Navicula and Triceratium are prominent ; some frustules contain brown
resting spores. Occasional textularian and rotaline Foraminifera, a radiolarian test and the silico-
flagellate Dictyocha fibula are noted.

Station WS 666. i.vii. 31. Lat. 120 18' 30" S. Long. 770 30' 30" W. 198 m. (Plate XXII.)
Diatomaceous mud. Dark green mud of which flocculent matter is the chief constituent. This
green and granular aggregate contains frustules of Coscinodiscus, Actinoptychus, Cocconeis and Am-
phora, together with a small proportion of mineral grains. The latter are mainly angular fragments of
quartz, green hornblende and volcanic glass, the size of which is less than a maximum of 0-2 mm.
in diameter.

Station WS 667. i.vii. 31. Lat. 120 23' 12" S. Long. 770 39' 30" W. 474 m. (Plate XXII.)
Sand and gravel. This sample consists of a few angular, black pebbles up to 5 mm. across, and a
small quantity of sand. The sand grains are mainly angular fragments of dark brown volcanic glass
(0-2 mm.), some shell fragments and a few grains of green hornblende. There are also some colourless
radial aggregates of a zeolite mineral, probably phillipsite, some of which reach a diameter of about
0-2 mm.

Station WS 669. 2. vii. 31. Lat. 120 33' 30" S. Long. 780 21' 36" W. 3840 m. (Plate XXII.)
Diatomaceous mud. This sample consists mainly of the usual flocculent aggregates with diatoms

and sand grains. Whole frustules of Coscinodiscus are abundant, and many contain resting spores.

The silicoflagellate Dictyocha fibula is noted as of occasional occurrence. Mineral grains are generally

angular and of small size, but occasional grains of quartz, green hornblende and volcanic glass reach a

diameter of about 0-2 mm.

Station WS 671. 3. vii. 31. Lat. 120 10' 48" S. Long. 770 59' 12" W. 1934 m. (Plate XXII.)
Diatomaceous mud. The main constituent is granular flocculent material containing brown resting
spores and diatoms. The most prominent genera are Coscinodiscus (large and abundant), Triceratium,
Actinoptychus, Synedra and Navicula. Occasional Radiolaria (Eucyrtidium and Lithostrobus) and the
silicoflagellate Dictyocha fibula are noted. Mineral grains, sparsely distributed in the deposit, include
quartz, green hornblende and volcanic glass, in grains up to o-i mm. in diameter.

SEA-FLOOR DEPOSITS. PART I 347

Station WS 676. 10. vii. 31. Lat. 080 17' S. Long. 790 01' 30" W. 29 m. (Plate XXII.)
Diatomaceous mud. A black mud in which the flocculent aggregates are in excess of the mineral
constituent. The former contain resting spores and diatom debris, besides whole frustules of
Coscinodiscus (large and abundant), Actinoptychus and Naviada. The spores and the flocculent matter
are almost black, especially the denser portions of the latter, a feature which is presumably due to
local peculiarities of decomposition. Textularian Foraminifera and fragments of Radiolaria also occur.
Most of the mineral grains are extremely small, but some range up to a diameter of about o-i mm. ;
they include angular grains of quartz and volcanic glass.

Station WS 677. 10. vii. 31. Lat. 080 19' 30" S. Long. 790 05' 45" W. 45 m. (Plate XXII.)
Diatomaceous mud. The general constitution of this deposit is very similar to that last described,
but the flocculent matter is green and the spores remain brown. Among diatoms, Coscinodiscus is
large and abundant, Actinoptychus and Navicula are less plentiful. Textularian and rotaline Fora-
minifera are of frequent occurrence. Mineral grains, present in subordinate quantity, are mostly very
small, but quartz grains up to o-i mm. and splinters of volcanic glass up to 0-2 mm. are noted.
Station WS 678. 10. vii. 31. Lat. 080 35' 30" S. Long. 780 57' W. 54 m. (Plate XXII.)
Diatomaceous mud. This sample presents no essential difference from the last, though some of the
quartz grains attain a greater size, namely, about 0-25 mm. in diameter. Otherwise there is the same
excess of flocculent matter over the mineral constituents and the same assemblage of diatoms.

Station WS 680. 10. vii. 31. Lat. 080 44' S. Long. 790 15' W. 91m. (Plate XXII.)
Fine sand. A green sandy deposit with many mineral grains of about 0-25 mm. in diameter. They
include angular grains of quartz (up to 0-4 mm. diameter), prismatic grains of green hornblende
(0-25 mm. long), brown hornblende of the same habit and size, and splinters of volcanic glass (o-i mm.
across). The larger grains are in excess of the finer material. There is a small amount of green, granular,
flocculent matter which gives colour to the sand and the liquid. Resting spores and fragments of
Coscinodiscus are the only organic remains recognized.

Station WS 686. 17. vii. 31. Lat. 090 25' 30" S. Long. 8o° 22' W. 4206 m. (Plate XXII.)
Diatomaceous mud. A grey-green mud in which the flocculent material is in excess compared
with the mineral constituent. The mineral grains include quartz (about 005 mm. diameter), green
hornblende (prismatic grains up to o-i mm. in length) and splinters of volcanic glass. The principal
diatoms in the flocculent material are Coscinodiscus, Actinoptychus and Achnanthes.

Station WS 689. 18. vii. 31. Lat. o7°oi'S. Long. 8i° 09' W. 2093 m. (Plate XXII.)
Diatomaceous mud. A green mud very similar to the last, but with perhaps a larger proportion of
mineral grains which include quartz (o-i mm.), green hornblende (o-i mm.) and volcanic glass
(0-05 mm.). The flocculent material is largely diatomaceous in character; it contains whole frustules
of Coscinodiscus (large and abundant) together with small rotaline Foraminifera.

Station WS 692. 19. vii. 31. Lat. 060 29' 15" S. Long. 8o° 33' W. 25 m. (Plate XXII.)
Medium sand. The sample is a small quantity of sand with shell fragments. The minerals include
quartz in angular grains up to 0-4 mm. in diameter, flakes of white mica 0-5 mm. across, and opaque
grains with about the same diameter. Large frustules of Coscinodiscus are present.

Station WS 694. 19. vii. 31. Lat. 060 38' S. Long. 8o° 49' 54" W. 1216 m. (Plate XXII.)
Diatomaceous mud. The green flocculent material contains brown resting spores and diatom
debris, among which many whole frustules of Coscinodiscus, Actinoptychus, Synedra and Achnanthes
are noted. The mineral grains are usually very small, but some angular grains of quartz, brown horn-
blende and volcanic glass reach a diameter of about o- 1 mm.

Station WS 696. 20. vii. 31. Lat. 060 54' 48" S. Long. 8i° 02' W. 1901m. (Plate XXII.)
Diatomaceous mud. This deposit has the same general characters as the foregoing sample. The
flocculent material is in excess over the mineral constituent and contains whole frustules of Coscino-
discus, Triceratium and Navicula, the first named being most numerous. The silicoflagellate Dictyocha

348 DISCOVERY REPORTS

fibula is occasionally seen. The mineral grains include quartz (up to 0-15 mm. diameter) and green
hornblende (0-05 mm. diameter).

Station WS 697. 21. vii. 31. Lat. 050 55' 30" S. Long. 8i° 09' W. 75 m. (Plate XXII.)
Diatomaceous mud. A dark sandy mud in which mineral grains form a greater bulk than flocculent
matter. The mineral grains are mostly angular fragments of quartz, brown and green hornblende,
volcanic glass, with a maximum diameter of about 0-2 mm., but flakes of white mica are sometimes
0-5 mm. across; a small crystal of tourmaline (0-05 mm.) is also noted. The flocculent material is
present in considerable quantity, and whole frustules of Coscinodiscus, Biddulphia, Actinoptychus and
Navicula are plentiful. A few tests of the silicoflagellate Dictyocha fibula are noted.

Station WS 700. 21. vii. 31. Lat. 050 52' S. Long. 81° 15' 30" W. 310 m. (Plate XXII.)
Diatomaceous mud. The sample consists mainly of green, granular, flocculent material which
contains entire frustules of Coscinodiscus, Actinoptychus, Achnanthes and Navicula, the first-named
being particularly large and abundant. The mineral fraction contains small angular grains of quartz,
volcanic glass and brown hornblende, which are mostly less than 0-05 mm. in diameter.

Station WS 701. 21. vii. 31. Lat. 050 48' S. Long. 8i° 22' 30" W. 1083 m. (Plate XXII.)
Diatomaceous mud. This consists largely of flocculent material of the usual type, in which frus-
tules of Coscinodiscus are prominent; the silicoflagellate Dictyocha fibula and textularian Foraminifera
are occasionally seen. Mineral grains, usually below 0-05 mm. in diameter, are mainly of quartz and
volcanic glass.

Station WS 702. 21. vii. 31. Lat. 050 38' S. Long. 8i° 40' W. 3102 m. (Plate XXII.)
Diatomaceous mud. Large frustules of Coscinodiscus are plentiful together with smaller tests of
Actinoptychus and Synedra in the flocculent material which is of the usual type. In the small amount
of mineral grains, quartz, volcanic glass and green hornblende occur in fragments which are usually
less than 0-05 mm. in diameter.

Station WS 703. 22. vii. 31. Lat. 050 34' S. Long. 8i° 11' 30" W. 4742 m. (Plate XXII.)
Diatomaceous mud. The deposit is composed largely of granular flocculent material in which
large frustules of Coscinodiscus are plentifully distributed. Apart from this, recognizable diatoms are
not numerous but the occurrence of Aster omphalus is noted. Angular quartz grains (0-05 mm.) are
sparsely scattered through the flocculent mass.

Station WS 705. 23. vii. 31. Lat. 05° 35' 30" S. Long. 830 41' 45" W. 4026 m. (Plate XXII.)
Diatomaceous mud. An unctuous brown mud with a small proportion of tiny mineral grains (less
than o-o 1 mm. diameter), chiefly angular particles of quartz and volcanic glass. The flocculent
aggregates enclose abundant frustules of Coscinodiscus together with some of Synedra and Thalassio-
thrix, also fragments of Radiolaria.

Station WS 708. 25. vii. 31. Lat. 040 18' S. Long. 820 05' W. 4314 m. (Plate XXII).

Diatomaceous mud. Composed of green flocculent material of the usual type with a very small
amount of mineral grains, mostly less than o-oi mm. in diameter. Coscinodiscus is the most prominent
of the diatoms and there are numerous fragments of Radiolaria.

Station WS 711. 27. vii. 31. Lat. 040 19' 30" S. Long. 8i° 27' W. 1885 m. (Plate XXII.)
Terrigenous mud. This sample is formed largely of flocculent material which is responsible for
the green colour of the deposit. A few diatoms (Coscinodiscus) are present, usually retaining their
cell contents. The flocculent matter is more abundant than usual in a typical terrigenous deposit, but
there is a general paucity of recognizable organisms. The mineral grains are mainly subangular in
shape and include quartz (up to 0-25 mm.) and green hornblende.

Station WS 717. 31. vii. 31. Lat. 02° 02' S. Long. 81 ° 18' 30" W. 1649 m. (Plate XXII.)
Diatomaceous mud. This sample has a fair proportion of sandy material, the grains of which are
mainly about o-i mm. in diameter. The quartz grains reach a maximum of 0-25 mm., while fragments

SEA-FLOOR DEPOSITS. PART I 349

of green hornblende are usually much smaller (0-05 mm.). The flocculent material forms dense green
aggregates which impart their colour to the deposit; it contains entire frustules of Coscinodisais and
occasional fragments of Radiolaria.

Station WS 768. 20. x. 31. Lat. 450 31' S. Long. 630 23' W. 108 m. (Plate XVII.)
Diatomaceous mud. The deposit contains a considerable proportion of mineral grains. Some
subangular and rounded quartz grains reach 0-5 mm. in diameter, while prismatic grains of green
hornblende and brown volcanic glass are 0-2 mm. in length. The worn appearance of these larger
grains is in contrast with the fresh and angular quality of many grains between o-i and o-oi mm.
in diameter. Though flocculent material forms a conspicuous proportion of the deposit, recognizable
diatoms are not very plentiful; they include Thalassiosira, Coscmodiscus, Pleurosigma, and a chain
form which is not identified. A few foraminiferal tests are present.

MARINE BIOLOGICAL STATION

Station MS 68. 2. iii. 25. East Cumberland Bay, South Georgia; 17 miles S J° E to 8| cables
south-east by east of Sappho Point. 220-247 m.

Diatomaceous mud. A fine-grained deposit, mainly debris of diatom tests, but some entire centric
diatoms. The mineral grains are extremely small, though some of the larger detrital grains are more
than o-i mm. in diameter.

SOUTH ATLANTIC AND SOUTHERN OCEAI>

PLATE XVIII

»,)

P

WS78

65°

60°

>— i |— i t=t-

\20Om.
\

\

o

r> WS96

o

WSI08

O WS95

QWS97

=P3=

M n

"I

\

WS94-

0

QWS77 0WS79 0WS80

QwS?8

~Q

WS99

©WS109

^s

Q WS76

ZOOm.

I

. WS93°~

/

©230

228

/..•'•

O

WS86

SOOm
-•■ -fOIOM

//

Burdwood Bank

\

^:-,.-::'^'

0WS87 I

GRAVELS and SANDS

TERRIGENOUS MUDS (

DIATOMACEOUS MUD

±3Z

ZE^Z

E

I — ■ »— * ■ — t-U^r-

50

55

60°

FALKLAND ISLANDS

PLATE XIX

39"

30'

38°

30

ZT

30'

36°

30

35

3o'

3000m
\

30'

3000m

I

TERRIGENOUS MUDS
GLAUCONITIC MUD

DIATOMACEOU5 MUD

' 2000m

V ./'

53'

30

53

30-

3000m-

2000m

IOOOm
500 m

r

^"^ ©129 ^""^
'©WS26-V ''

r—^ZSOM.

5 151

7" r.> f"^

V v.

^ V _

-30

-v^

16/ \., \

3000m

30'-,

54

3d

55

30

- 56

SOUTH GEORGIA

•A
W

<

w

w

en

<

o

w

m

PLATE XXII

WEST COAST OF SOUTH AMERICA

[Discovery Reports. Vol. IX, pp. 351-372, December, 1934.]

ON THE STOCK OF WHALES AT
SOUTH GEORGIA

By

J. F. G. WHEELER, D.Sc.
Bermuda Biological Station for Research, Inc.

ON THE STOCK OF WHALES AT
SOUTH GEORGIA

By J. F. G. Wheeler, D.Sc.
(Text-figs. 1-3)

The application of the method of age determination described in a previous paper
has brought into prominence the complex nature of the whale population which
year after year since 1904 has been attacked by the catchers of the companies operating
from South Georgia. Study of the age composition of the population strongly suggests
reduction of the stock, and some advance has been made towards ascertaining the rate
at which reduction has proceeded. Work on these lines, when considered in relation to
other investigations, such as the movements of stock, affords a fair prospect of reaching,
in this area at least, a scientific basis for the attainment and maintenance of economic
security.

Several points of interest arose as the results of age determinations were studied. The
calculated rate of reduction is not without interest and suggestiveness. Wholly depen-
dent as it is upon the unit character of the recurring whale population and carrying
within itself evidence that sections, at least, of the population are not present in the area
in their true proportion, it provokes consideration of the real nature of the catch and
enquiry into the population from which the catch was drawn.

This paper is an attempt to give the steps by which the calculation of the rate of
reduction was made possible and a brief review of the problems underlying the study of
whale stocks.

The treatment of the data had of necessity to be undertaken from a statistical point
of view and for this part of the work I am indebted to Mr T. Edser of the Ministry of
Agriculture and Fisheries, whose expert advice has been invaluable, and to Mr G. E. R.
Deacon of the Discovery staff. I am also grateful to Dr Stanley Kemp, and to Sir Sidney
Harmer and Mr J. O. Borley of the Discovery Committee, for much constructive and
willing assistance.

Since 1925 members of the scientific staff of the Discovery Committee have been
engaged at South Georgia in an intensive enquiry into the biology of whales, during
the course of which a large proportion of the whales brought in for commercial purposes
by the Compania Argentina de Pesca has been examined. Following out an indication
of a method of age determination given on pp. 450-2 of Southern Blue and Fin Whales
(Mackintosh and Wheeler, 1929), a definite correlation was established in season
1929-30 between the number of corpora lutea in the ovaries of female Fin whales and
physical maturity as indicated by the ankylosis of the vertebral epiphyses to their centra,
and further light was thrown on the theory that age can be determined in the females of
this species from the number of corpora lutea. These results were put forward in a later

354

DISCOVERY REPORTS

paper (Wheeler, 1930), in which it is shown that physical maturity is correlated with the
presence of fifteen corpora lutea in the ovaries and reasons are given for considering
that females with from one to four corpora lutea are in their first or second year from
sexual maturity; those with from five to nine corpora lutea are in their third or fourth
year from sexual maturity ; those with from ten to fourteen are in their fifth or sixth
year, and so on.

Such a method of estimation is, of course, applicable only to females. No way is yet
known by which the age of the mature male whale can be determined. Also, as Blue
whales have not figured in the catches to the same extent as Fins, our conclusions are
confined to the latter species.

Observations made by the Discovery staff cover seven seasons during five of which
almost the entire catch of the company was examined. Some months of fishing were
unavoidably missed in each of the remaining seasons, and, as will be shown later, this
is a matter of importance when assessing the representative nature of the catch.

The total number of female Fin whales recorded was 879, but of these 277 were
shown to be sexually immature by examination of the internal genitalia, and 130 were
so badly decomposed, or so badly damaged internally, that the genitalia could not be
examined. An approximate division of this last group into sexually mature and immature
can, of course, be made using length as criterion (Mackintosh and Wheeler, p. 417),
and the figures then read 281 immature (32 per cent of the total) and 598 mature whales.
In 472 of the latter full ovarian records are available.

The analysis of the 472 sexually mature whales into two-year groups commencing at
sexual maturity and dependent upon the number of corpora lutea is as follows :

Table I
Female Fin whales

Number of corpora lutea

1-4

5-9

1
10-141 15-19

20-24

25-29

30-34

35-39

40-44

45-49

Years from sexual maturity

1-2

3-4

5-6

7-8

9-10

11-12

13-H

15-16

17-18

19-20

1924-25
1925-26
1926-27
1927-28
1928-29
1929-30

16

19

4

1

32
52
23

6
26

3

2

14
42
10

6

l5
6

1

H

21
16

2
7
4
1
11
24
7

3
4

1
2

24
7

2

3
1

2

16

6

1

2
6
1

1

2
1

1

1

Totals ...

i47

103

79

56

41

3°

10

4

1

1

The taking of an annual unselected sample from an ageing stock to which annual
additions are made at the lower age limit (which is, in effect, what the whaling com-
panies have been doing) results inevitably in a heightening of the progressive diminution
among the older members of the catch naturally caused by ordinary mortality. The

THE STOCK OF WHALES AT SOUTH GEORGIA 355

figures given, although not impressively numerous, demonstrate this rapid diminution.
If the catch of a single season were sufficiently large, or if examination could have been
made of all the female Fin whales brought in to all the stations during a single season
the sample would have been nearer the ideal. In the circumstances the figures for several
seasons have been taken together and considered as a unit, although it is realized that
changes may have occurred in the size of the stock and the intensity of fishing from
season to season during the period of investigation to make the calculated rate of reduc-
tion only an approximation. It is, for instance, likely that the last two seasons were
adversely affected by the pelagic fishery to the south and east of the island ; and as in
these seasons a greater percentage of the whales brought in was examined, the rate of
reduction will be unduly weighted by them. The proportion of the catch we examined,
however, was a random sample of the catch of the Pesca company, since, when whales
were missed, no selection was made. Our sample should be therefore just as representa-
tive of the local population as the slightly larger Pesca catch. The over- weighting
of the later seasons will certainly affect the rate of reduction if the stock is attacked
elsewhere, but that is part of the disability of treating several seasons together, and
if the damage has been done, then the calculated rate of reduction is more closely
applicable to present day conditions than it would be if the earlier seasons alone were
considered.

The missing of whales here and there during the season cannot affect the sample
when no selection is made, but the absence of observations for a considerable part of
the season is a different matter. Partial catches are not admissible in the analysis of age-
group reduction since fluctuations are known to occur during the season not only be-
tween the numbers of each species present on the grounds but also between the numbers
of each sex and stage of maturity of each species. Thus in Southern Blue and Fin Whales
(p. 460) it is remarked that "when the first and second halves of a season at South
Georgia are compared it is found that in the first half the catch is composed of a majority
of mature whales, while in the second there is an influx of immature whales (and per-
haps a withdrawal of adults) which causes a sharp reduction in the average lengths " and
again "Apart from the fact that immature whales have occurred in relatively greater
numbers in the latter part of the season, there is an indication that of the adult whales
themselves, those taken early in the season are mostly older than those taken later".
This is shown very clearly in Fig. 1 in which physical maturity has been used as a dis-
tinguishing character contrasted with early sexual maturity and sexual immaturity.
Over five complete seasons the majority of physically mature females appeared in
December, the majority of sexually mature not physically mature females in January
and the peak of the influx of immatures was in February.

In seasons 1924-5 and 1927-8 observations were not commenced until well into the
second half of the season, consequently the figures for these seasons must be neglected
in the consideration of stock reduction since in both the proportion of the younger
section of the population will be abnormal. The number of whales is now reduced by
41 to 431.

356

DISCOVERY REPORTS

The division of the catch into age groups may now be examined. Following the
attainment of sexual maturity whales are considered to give birth normally at recurrent
intervals of two years. Each two-year period is occupied in gestation (eleven and a half
months) and lactation (six months) after which is a resting period when the activity of
the reproductive organs subsides until the onset of the next sexual season. Each age
group should therefore consist of ovulating, pregnant, lactating and resting whales
covering the two-year period. It should be capable, moreover, of further analysis into

OCT
IMMATURE — -O--

NOV DEC JAN

SEXUALLY MATURE —

FEB MAR APR

PHYSICALLY MATURE

Fig. i . Fin whales, females. Numbers of immature, sexually mature but not physically mature,
and physically mature whales in five complete seasons at South Georgia.

a one-year grouping, the lactating and resting whales being a year older than those
ovulating or pregnant. At South Georgia, however, ovulating whales are rare and it has
been shown that breeding is, in all likelihood, restricted to northern waters. Ovulation
does occur exceptionally in the south but this does not necessarily involve copulation,
and may indeed be due to some pathological condition of the individual whale. Omitting
ovulation, then, the grouping of the South Georgia catch is limited to the pregnant and
lactating and resting whales, of which the full analysis for five seasons is shown in
Table II.

THE STOCK OF WHALES AT SOUTH GEORGIA

Table II
Female Fin whales

357

Number of corpora lutea

1-4

5-9

10-14

15-19

20-24

Years from sexual maturity

1-2

3-4

5-6

7-8

9-10

Pregnant (/>), lactating (/),
resting (r)

P

I

r

P

I

r

P

I

r

/>

/

r

P

I

r

1925-26
1926-27
1928-29
1929-30
1 930-3 I

5
1

20

40

l9

2

4
2

H
3

10
8
2

15

2
10

27
6

3

1
8

1

8
1
3
7
3

6

5
8

11

4

1
4

5

1

5
3
5

3
2
8

J7
6

1

1

3

4
1

3

1

3

3

1
12

5

1

8

1

1

4
1

Totals

85

8

37

60

13

22

44

9

l9

36

10

7

21

10

6

Totals in i-year groups ...

85

45

60

35

44

28

36

17

21

16

Totals in 2-year groups . . .

130

95

72

53

37

Number of corpora lutea

25-29

3°-34

35-39

40-44

45-49

Years from sexual maturity

11-12

I3-H

15-16

17-18

19-20

Pregnant (p), lactating (/),
resting (r)

P

/

r

P

/

r

P

/

r

P

/

r

P

I

r

1925-26
1926-27
1928-29
1929-30
1930-31

3

2

12

6

—

4

1

5
1

—

1

1
1

1

—

1

1
1

1

—

—

—

—

1

Totals

23

—

5

7

—

3

1

—

3

1

—

—

—

—

1

Totals in i-year groups ...

23

5

7

3

1 3

1

0

0

1

Totals in 2-year groups . . .

28

10

4

1

1

The totals under each two-year period show a progressive reduction, as was expected,
and in this there seemed to be the promise of a formula expressing the rate of reduction.
All the stations at South Georgia dip into the same stock since the catchers from all the
stations work the same grounds; it is safe therefore to assume that the catches at all
stations are similar in composition and show the same rate of reduction. If the total
catch from the area could be analysed it would consist of the same age groups repre-
sented by larger numbers of individuals, but the numbers would bear the same relation
to one another, i.e. the rate would be the same. I wish to make clear the fact that this

3S8 DISCOVERY REPORTS

rate of reduction is not a local phenomenon; it is not confined to the catch of the Pesca
company, neither is it increased when the activities of the other stations are taken into
consideration. It is the rate at which the stock has been reduced, at South Georgia or
elsewhere, by hunting, by natural death, disease or any other cause of elimination.

The validity of the rate of reduction, granting the acceptance of the method of age
determination, depends upon two postulates, (i) the annual recurrence of the same
stock to these waters, and (ii) the representativeness of the sample considered. Both of
these difficult questions are discussed later on. It will be noticed that up to the present
they have been taken as facts. As to the sample, i.e. the catch, one source of error has
been pointed out and eliminated by the omission of the figures for the partial seasons
1924-5 and 1927-8. A further possible difficulty has been brought to light by the
analysis into one-year groups. Throughout these figures a serious shortage of lactating
and resting whales is evident. As regards the former an artificial check on the numbers
captured lies in the protection by law of female whales running with calves, though I
think that the whaling community would agree that unless a calf is present it is not
possible to distinguish a lactating whale at sea and certainly a number are brought in to
the stations. With females in the resting condition there is no such explanation and the
fact can only mean that a certain proportion of whales in the post-pregnant condition
do not accompany the main migratory schools, or, at all events, they do not appear on
the South Georgia grounds during the fishing season. Is this, however, a serious draw-
back to the calculation of a rate of reduction? Fewer lactating whales may mean that the
fishery falls more heavily upon the pregnant whales but it would appear from the figures
that the proportion of absent females of each age is approximately constant. In the
first five age groups the percentage pregnant is 65, 63, 61, 68, and 57. Pregnant
whales can be considered as a two-year series, apart from lactating and resting whales,
and these latter if fully represented should make another series. The numbers in the
former series are not large, and as has been said, those in the latter are reduced by the
operation of the regulations of whaling to still smaller numbers. Probably owing to the
paucity of data, neither series exhibits a constant rate of reduction. Since, however, the
total number in each of the first five age groups of Table II is made up of pregnant
whales on the one hand and lactating and resting whales on the other in approximately
constant proportions, it seems legitimate to treat the data in the age groups of that
table as representative, notwithstanding the fact that the condition of the whales in each
group is not homogeneous. Fig. 2 illustrates the reduction of stock as unanalysed and
in the light of the conditions found within the various groups.

We now come to consideration of the series of numbers, 130, 95, 72, 53, 37, 28, 10,
4, 1, 1 which represents the number of females taken during five seasons which could
be allocated into successive two-year age groups.

Edser's conclusions and comments may be summarized as follows: The series of
numbers suggests the terms of a geometric progression of the form

or, ar(i - r), ar (1 - r)2,. . .
where a represents the initial stock of mature females and r the rate of reduction.

THE STOCK OF WHALES AT SOUTH GEORGIA 359

Dividing each term by its predecessor the factor 1 — r is left, and thus between the
first six age classes (3-4 years to 13-14 years) the rate of reduction r has an average
value of 26 per cent, the actual figures being 27, 24, 26, 30 and 24 per cent. Beyond the
13-14 year class the rate of reduction is greater — 64, 60 and 75 per cent. For the first
six age classes the reduction is therefore regular, as if it were due to one source of loss,
but beyond 13-14 years the high reduction rates suggest that there is some other cause
which removes the older whales from the South Georgia population, or, perhaps more

1 r

3 A-

1 ■

5

1
6

1

7

1

6

1 1—

9 IO

YEARS

"I-
II

OF

II'

12 13

AGE

1

14

1

15

1

16

~i r

20 21

TWO-YEAR GROUPING O PREGNANT WHALES---* LACTATING OR RESTING WHALES - -C

Fig. 2. Fin whales, females. Numbers of each age group in the total catch of five seasons:

mature whales only.

likely, that the old whales do not linger in these waters. They may even avoid the locality
altogether on their way farther south.

Edser further calculates from these figures the number of immature females that must
become mature in order to keep the stock at a constant level, and also the number of
young females that can, in the most favourable circumstances, be produced by the
mature stock.

By equating the sum of the first six terms of the algebraic series to the sum of the
first six catches1 a has been calculated, and with a equal to 497 and r to 26 per cent the
theoretical catches based on the series and the stocks of whales in the corresponding

/i - (1 - o-26)6\
o-20fl — — =415, i.e. the sum of the first six terms of the actual catch : from which a=4Q7.

\i - (1 -0-26)/ * J +v/

360

DISCOVERY REPORTS

age classes have been determined. These are given in Table III, in which it will be
noticed that the only actual figures are those in column 2, those in columns 3-8 being
theoretical. It will be seen how closely the theoretical catches in column 3 agree with
the actual catches in column 2. By extending the series backwards for one term the
theoretical catch and stock of immature whales have been determined.

Table III
Female Fin whales

I

2

3

4

5

6

7 8

Figures for five years

Average
figures for two years

Age, years
from birth

Actual
catch

Theoretical
catch

Theoretical
stock

Surviving
stock

Stock ab-
sent from
South
Georgia

Stock at

South

Georgia

Total

stock*

0-2

3-4
5-6
7-8
9-10
11-12

13-H

15-16

17-18
19-20
21-22

130
95
72
53
37
28
10

4

1
1

(175)
129

96

71
52

39
29

(672)

497

368

272

201

149

1 10

38

15

4

4

(497)

368

272

201

149

1 10

81

28

11

3

3

43
13

7
0

(269)

147

109

80

60

44

15

6

2
1

(269)

147

109

80

60

44
33
24
18

13

10

Totals for three years and upwards... 464 538

* Figures based on the assumption that the normal rate of reduction (26 per cent) is continued with the
older whales.

Column 5 shows the numbers of whales in each age class that are not caught (stock
minus catch). As far as the 13-14 year class the calculated stock in each age group
is the same as the number of whales surviving in the previous age group. Beyond the
13-14 year class, where the series no longer holds, the stock has been calculated from
the individual figures of the catch on the assumption that the catch is 26 per cent of the
stock, and the figures obtained, by subtraction from the surviving stock of the previous
age group, lead directly to the numbers of whales that have disappeared. It should be
remembered that the whales missing from one age group do not appear in the stock of
the next group. The numbers of missing whales are shown in column 6 (column 5
minus the next line of column 4). The absence of about half the older stock (52, 48 and
46 per cent are the figures) is a curious phenomenon whose implications have already
been remarked upon.

THE STOCK OF WHALES AT SOUTH GEORGIA 361

All these figures are based on the catch for five seasons and it is necessary to find out
how many whales can be produced by the remaining stock in two years, since each
female gives birth normally to one whale in this period. This number is readily obtained
by taking two-fifths of the mature whales which survive catching and are not absent
from the catching area for other reasons, i.e. two-fifths of column 5 minus column 6.
The numbers for each age class are given in column 7 and the sum — 464 — represents
the number of whales of both sexes that can, in the most favourable circumstances, be
produced by the mature stock. Roughly 232 of these will be females.

Now the theoretical stock of immature females for five seasons has already been
worked out and two-fifths of this number (269) gives the initial immature stock for two
years. It follows, therefore, since 269 immature females are necessary to keep the stock
at its present level and only 232 females can be produced, that considerable damage is
being done.

If, however, the older whales that are absent from the South Georgia population are
not eliminated, but are still adding to the immature stock and are subject elsewhere to
the same rate of reduction, the total of surviving mature females will be considerably
greater. In column 8 the numbers of surviving stock of the different age classes are
given supposing that no whales are missing; and the number of immature females
produced — 269, half the sum of this column — will just balance the losses and the stock
will remain stable. It should be remembered that the figures for the reproduction of the
stock are optimum values and that probably the conditions necessary to attain them
are never reached.

While again stressing the indicative rather than the exact nature of the results it
appears probable that on these lines some definition can be given to the effect of the
fishery upon the whale population in this area.

A more difficult question is that of the killing of immature whales. The number of
immatures in the South Georgia catches has been lately remarked on by Hjort, Lie and
Ruud (1932) — "Why is it that in certain areas, especially where whaling has gone on
for 25 years (from the old land stations) there are larger numbers of young animals than
elsewhere? We cannot infer that more young animals are caught on the old grounds
because the stock has decreased ... it appears more likely that the young animals have a
migration route or area of distribution of their own, which does not quite coincide with
that of the adult whales".

The sexually immature females formed 32 per cent of the total catch examined by
us, which is, on the face of it, an alarming proportion, if, as we have supposed, all the
immatures are between the ages of one and two years. Up to the present we have had
no method of age determination among the immatures except that afforded by deduction
from length frequency. By utilizing such evidence as was available in 1928 the life
history up to the time of sexual maturity was suggested by us as follows (Mackintosh
and Wheeler, 1929, p. 444). Birth takes place at about 6-5 metres. Weaning at about
12 metres some six or seven months after birth in early summer when presumably
mother and calf have migrated southward during the spring. On the analogy of the

362

DISCOVERY REPORTS

rather more complete evidence for the Blue whale there follows another migration
northward and southward, at the end of which the whales are verging upon sexual
maturity and again migrate to the north for breeding, i.e. two age groups should be
present in southern waters, one of whales just weaned, the other of whales "verging on
maturity". Representatives of the first group should not appear at all in the catches for
mothers and calves are protected. Therefore all the immatures taken should belong to

one age group only.

S b Table IV

Length and scar ages of female Fin whales immature by
examination or less than 20 metres

12-90
16-70
13-90

H-65

Number of scar ages

17-10
19-60/)

i9-75/>
1 6-45
19-65
19-10

17-55
16-90

i8-45
17-70

18-30
18-00
19-30
19-80
14-10
17-75

i8-35

18-85
16-60
18-40
18-65
1 6-oo

17-75
19-80
19-40
17-80

i9-95/>
16-93

16-15
18-50
16-70
19-05
17-95
I7-95
16-25
14-60
17-10
18-10

19-90/)

19-30

18-90

18-75/.

19-70

19-65/)

19-30/.

19-00

19-50

I9-55

19-90/)

19-80/)

19-30

19-23

17-00

19-85/.

19-25

18-20

19-60

16-50

iS-55
18-25
19-50
19-87
19-85
18-25
18-20
19-00

16-30

I9-55*-
21-40
19-20
19-65

18-16

Note : /» = pregnant ; r = resting.

THE STOCK OF WHALES AT SOUTH GEORGIA 363

Some data collected during season 1929-30 on the presence of different ages of white
scars can be adduced in support of migratory movements, though the light they throw
upon the age problem is at best uncertain. Leaving aside the mode of formation of the
pits and scars (Mackintosh and Wheeler, 1929, pp. 373-9) there remains the fact that
whales caught off South Georgia are without exception scarred to a greater or less extent,
while those caught in northern waters show the earlier stages of open or crescent pits
(Olsen, 1913 ; Risting, 1928 ; Mackintosh and Wheeler, 1929). Every northern migration,
then, adds a new batch of open pits, and the whales taken during the southern migration
show a fresh series of white scars which can often be distinguished from the older scars
by their whiteness, the depth of the central depression and by the fact that a new scar
sometimes covers an older one in which the pigment has partly returned and the de-
pression filled out. Naturally the more series that are superimposed one upon another
the more difficult it is to estimate the number. Evidence of this kind is unsatisfactory in
character since it depends largely upon observations of an indefinite quality ; but the
difficulties of obtaining any sort of evidence regarding age warrant its inclusion. There
are listed in Table IV the lengths of individual female Fin whales, immature by
examination or less than 20 metres in length (the critical length of sexual maturity)
taken during season 1929-30, with the recognized ages of scars present.

The fact that the scarred condition of all the whales caught at South Georgia is proof
of a north to south migration has been pointed out by Harmer (1931, p. 107). The
further inference that the whales have migrated back and forth is not really proved since
there are no data concerning the possibility of successive ages of scars made during a
prolonged sojourn in northern waters. In this list there is confirmation of the first
migratory movement in the presence of newly weaned or still lactating calves, which
have come from the north as their scars bear witness.

If the connection between scar ages and migration is allowed the whales with two
series of scars should have been twice in northern waters and twice south, i.e. they are
the "immatures verging upon sexual maturity". What then of the group with three
ages of scars, of which many are still immature? This suggestive point can be carried
further by plotting the length frequencies of all the whales and differentiating between
the age series of scars. This has been done in Fig. 3, which brings together all the data
collected in 1929-30 relative to the ages of scars. Doubtful readings are omitted; but
where the recorded ages were "three or more" the whales have been added to the
fourth graph (four recognized age series and over). Two hundred and fifty-one female
Fin whales are represented out of the catch of two hundred and seventy-two.

The group of whales with four and more age series of scars has been at least four
times in northern waters and four times south if the correlation between scar series and
migration is allowed. With one exception all are mature according to expectation. Of
the whales that have apparently made three migrations south the majority are mature,
but there are sufficient immatures to justify the statement that, on this evidence, by no
means all female Fin whales become mature two years from birth. Therefore the group
of "immatures verging upon maturity", considered up to now as a single age group,

364

DISCOVERY REPORTS

may be really composed of two groups, and the younger of these groups is from the
appearance of the graph composed of somewhat smaller whales than the older. This to
some extent corroborates the inference that two groups are present. The mature whales
with two age series of scars can be explained on the assumption that they missed a north-
ward migration, or alternatively, that they made a prolonged stay in the north so that

in

L±J
_J
<

X

15-

in

tt 10
w

m

D
2

U.

o

5-

25

U

z

W

D
3

uj 2Q
a.

10-

5-

ONE AGE-5ERIE5

TWO AGE- 5ERIE5

THREE AGE-SERIES

FOUR AGE-SERIES and OVER

13

14

15

~r

16

~T

— r-

17

~r

SEXUALLY

LENGTH
IMMATURE O—

— 1—
IB

IN

T

T

— 1 —

20

T

~~ r

21

~r

19
METRE5

SEXUALLY MATURE

— 1 —
22

— 1 —
23

E4

Fig. 3. Fin whales, females. South Georgia, 1929-30. Length frequency in half-metre length groups and

number of age series of scars.

it is difficult to tell where an older scar series left off and a newer one began. It is, in
fact, possible to find a feasible explanation for a number less than the theoretically
correct number of superimposed age series : it is more difficult to explain the capture
of immature whales with a greater number of age series than their physiological con-
dition warrants, unless they really are of greater age.

While therefore the foregoing cannot be advanced as proof of the contention, it does
suggest that the apparent disproportion between the catch of sexually mature and im-

THE STOCK OF WHALES AT SOUTH GEORGIA 365

mature females may be capable of a natural explanation on the ground of mixed ages
rather than on the hypothesis of excessive fishing of a single age group comprising the
immature whales.

If some female whales become mature at two years from birth and some at three the
subsequent ages will also vary by one year and the determination of age after maturity
by the corpora lutea is involved in this complication. Previously the addition of two
years to the number of years from sexual maturity gave the age from birth, but on the
new evidence whales pregnant with from one to four corpora lutea may be either two
or three years old, those lactating with the same number of corpora lutea either three or
four years old, instead of two and three years from birth respectively. Similarly with the
later groups in the form of a double series. The rate of reduction is unaffected, but it
treats two years as one, in that pregnant whales may be actually of the same age from
birth as lactating whales and lactating whales as pregnant whales if they differ in series.
Physiologically the age is the same, as also is the actual age starting from sexual maturity.
The rate of reduction then is artificial in that its starting point is physiological and it
does not take into account the years preceding sexual maturity.

The uniformity of the curve of reduction is the most striking argument in its favour,
but it applies only to mature whales. The immatures cannot be added to the graph of
reduction nor can the figures be considered with those of the mature whales. The reasons
are not far to seek. There are 126 mature whales whose age is not known, which, because
they formed part of the catch, should be included in the figures, as against only four
whales believed to be immature on account of their length ; also there is uncertainty in
the number of ages represented by the 281 immatures as has just been deduced from
the evidence of the scars.

Considered broadly our figures suggest a population unduly weighted with im-
matures and with pregnant whales, and we have now reached the point when general
consideration of the catch and stock becomes all important : How far does the season's
catch represent the stock and what evidence have we concerning the recurrence of the
same stock at South Georgia?

Considerable changes have taken place in the South Georgia fishery during the
twenty-seven years of its existence. Until 1907 the company founded by Larsen worked
alone and made large catches of Humpbacks and Southern Right whales. Matthews
(1931) says — " He (Larsen) arrived at Grytviken in December, 1904, and at once started
work. Whales were plentiful and undisturbed, for months his catcher did not leave
Cumberland Bay, as she was able to get all the whales that could be dealt with close
inshore". These were the conditions that attracted other ventures in 1907, '08 and '09,
with the result that in 1909-10 seventeen catchers were operating from a series of stations
on the north side of the island forming a base line to the fishery roughly eighty miles in
length. There were no restrictions. All species were taken when opportunity occurred,
but few Sperms and Sei whales appeared in the catches and these species played no part
in the later development of the industry. It is said that Right whales were abundant at
the beginning of the fishery, but their number since that time has been small.

366 DISCOVERY REPORTS

Humpbacks did not long figure in the catches to the practical exclusion of other
species, for in 191 2-1 3 they had become scarce and Blue and Fin whales were being
hunted in their stead. This scarcity of Humpbacks caused considerable anxiety to the
whalers, whose attitude is recorded in the following passage from the Magistrate's
report on season 1913-14, "The question has often been debated by the local whale-
men as to the real cause of this continual scarcity of the Humpback whale. Is it the
continual killing that has thinned them down and frightened the remainder off? Or in
the course of their ocean migrations have they merely changed their course for the time
being to come back again? The general feeling is hopeful and inclined to take the latter

view".

But the Humpbacks did not return and the companies swiftly adapted themselves to
the new conditions. That there was a definite change of attitude is illustrated by the
following extract from Barrett-Hamilton's general notes quoted by Hinton (1925,
p. 155): "The species most hunted at South Georgia are Humpbacks, which are pre-
ferred to the Finners and Blue whales. A few Sperm and Right whales are caught, the
former in November and December, the latter when amongst the other whales, but
(according to Mr Henriksen, the manager of the South Georgia Company) usually they
keep to themselves, north-west of the island, and are not worth hunting specially there.
Though a Right whale is three times as valuable as a Humpback, the latter are preferred
where abundant, because their size is convenient for handling. Similarly, Finners and
Blue whales are not killed if Humpbacks can be obtained, being hard to kill and only
manageable in fine weather. Blue whales are a bit too large for the tackle if adult.
Therefore, South Georgia is primarily a Humpback fishery (Henriksen)". This was in

I9I3~I4- . , , -.

Blue and Fin whales then became the mainstay of the catch and since the beginning

of this fishery the companies have been fortunate in that any improvements they intro-
duced in the catchers and plant to meet the needs of the fishery with regard to one species
met also its needs with regard to the other, and by drawing their catch from two species
they were much less at the mercy of fluctuations in the supply of either. The failure of
the Humpbacks, however, is a fact that cannot be ignored. There is no evidence at all
that the course of their ocean migration was changed. It seems far more likely that the
decline was due to overfishing, and if so it follows that it was the same stock that was
affected each year.

The catches of Blue and Fin whales now claim our attention, and, limited as our
knowledge is by lack of definite evidence of migrations, certain features of the supply
of these whales can be outlined. It is evident from study of the catches that extensive
movements of large herds of whales take place both into and out of the area. At one
time fishing was continuous throughout the year but the catches made during the winter
were far from profitable and it was realized as a fact of importance to the industry that
of the large numbers of whales accessible during the summer only a small remnant re-
mained round the island during the winter. The movements of the herds as deduced
from monthly catch statistics have been studied by Risting (1928), and in more detail

THE STOCK OF WHALES AT SOUTH GEORGIA 367

by Harmer (193 1). It is shown that definite incursions take place but that the time of
their advent is variable and usually divergent in the two species. In certain seasons the
incursions are indicated in the graph of the catch as a single peak, showing that the
supply increased for a time, reached a maximum and then declined ; in other seasons
there are two maxima separated by an interval of several weeks. As indicated by Harmer
(1931, p. no) certain localities are characterized by one or other type of season-graph,
the study of which suggests that there are definite migration routes, probably not the
same for both Blue and Fin whales, on which under certain conditions, again specific,
the moving herds are delayed and the concentrations so profitable to the whalers
occur.

The difference in the relative abundance of Blue and Fin whales is sometimes very
sharply marked, and gives rise to the terms " Blue " season and " Fin " season, according
to whether there is a great preponderance of one species or the other in the season's
catch. These differences do not always occur, of course; sometimes both species are
plentiful, sometimes both are scarce. The cause of these fluctuations is not thoroughly
known, nor is it certain that the predominance of one species in the catch always indi-
cates the true state of affairs in the catching area. If one species concentrates nearer the
island than the other, the former will be the predominant one although the latter may
be superior in numbers, because the catchers are, after all, concerned only with taking
as many whales as they can as quickly as possible. The concentrations and movements
of Fin and Blue whales have lately been the subject of a report by Kemp and Bennett
(1932). The results obtained are based on returns made by the catchers, and, as the
authors remark (p. 169) : " When whales are plentiful in inshore waters the catchers will
naturally not go farther afield, and we have evidently no means of knowing how
abundant whales may be in the unexplored parts ". It is shown that the main concentra-
tions of Fins and Blues as plotted from the combined data of eight seasons (loc. cit.,
pi. xxiii) are in remarkable agreement and that this agreement is correlated with an
abundant food supply; but that in different individual seasons the positions of the
centres of the concentrations differ greatly, some part of which is attributed by the
authors to irregularities in the time of arrival of different schools of whales.

For many years the whalers have noticed that some form of correlation can be traced
between the presence of Blue whales and the ice conditions. Risting (1930, pp. 56, 93
and 97) definitely connects the action of warm and cold currents upon the drifting pack
ice with the growth of plankton and the occurrence of whales. Harmer (1931) discusses
this question fully and correlates the September mean air temperature at South Georgia
(closely connected with the events leading to the melting of the ice) with the order in
which the two species reach their maximum in the catches of the season immediately
following (p. 131); and by an examination of the records he shows that the species which
is first in excess nearly always maintains its superiority in both halves of the season and
in the total catch (p. 146).

We have then evidence from the catches that concentrations of one or other or both
species are formed off the island, which vary both in position and in time, such varia-

368 DISCOVERY REPORTS

tions being probably controlled by hydrographical and meteorological conditions. There
is nothing here inimical to the idea of a closed stock of whales, i.e. a body of whales
breeding in a certain area, migrating on a certain line and ending their food migration
in South Georgian waters or passing through these waters on their way farther south.
This idea indeed underlies the work on variation of seasonal concentrations just as it
is implied in the following remark from Kemp and Bennett (1932, p. 179): "The
considerable extension of the grounds during the recent four-year period, and the
fact that with the same number of whale-catchers fewer whales have been taken,
lends support to the generally held opinion that whales are now less abundant than
formerly".

Drawing conclusions from the diminution in number of sexually mature individuals
taken off South Georgia and the South Shetlands, Harmer (193 1 , p. 100) says : " Another
conclusion which it seems legitimate to draw from these figures is that the whales do not
wander at random throughout the Antarctic area, but are to some extent separated into
assemblages which have preference for particular localities. If this were not the case,
it would be difficult to account for the fact that the percentage of sexually mature in-
dividuals appears to be correlated with the length of the period during which whales
have been hunted in a locality".

The same author (loc. cit., p. 108), after discussing the evidence of movements of
immature and mature whales, fat and lean whales and whales covered with diatom film
and scars, draws the conclusion that " The various maxima noticed in the season-graphs
are thus not necessarily due to the continued migration of a simple stock of whales.
There is good reason for believing that they are, or may be, the resultant of a com-
plex series of movements, partly of whales, often immature, from the north, and
partly of well fed individuals moving northwards from their Antarctic feeding
grounds".

Now the arrival of batches of whales at different stages of maturity has already been
touched upon (p. 355), and a sequence of physically mature, sexually mature not
physically mature, and immature in December, January and February has been demon-
strated. This sequence may be simply an expression of the different rates of progress
physically possible to the different sections of the community ; for physically mature
whales are, on the average, one metre longer than the rest of the sexually mature whales1
and the latter are, of course, considerably longer than the immature. These differences
coming into play on a long migration would be an efficient cause of segregation. There
is, however, no evidence concerning this migration other than that already indicated by
the heavily scarred condition of the larger whales. Against this there is the undoubted
fact that batches of whales appear in the South Georgia catches covered with thick
diatom film (Bennett, 1920), which suggests strongly a stay of some length in Antarctic
waters.

There is evidence that in the early part of the season, when the influx would be

1 The average length of 329 female Finners with less than fifteen corpora lutea is 21-41 m.; that of the
remaining 143 mature whales is 22-41 m.

THE STOCK OF WHALES AT SOUTH GEORGIA 369

expected from the north, Fin whales are more often than not moving in a northerly
direction (Kemp and Bennett, 1932, p. 181).

Collective differences between the sexes have also been recognized. Sometimes for
a week or more the catch is made up almost entirely of male whales as it was in the
early part of January 1926 (Mackintosh and Wheeler, 1929, p. 461) ; sometimes females
are predominant; and it has been shown by Risting (1928) that males generally out-
number females in the total catches.

Segregation can be explained on the ground of physical differences between the sexes
during the course of a long migration. It is not often possible to say which sex precedes
the other on the grounds, but when the large numbers of immature male whales ap-
peared in 1926, there was a suggestion, in the later arrival of immature females, that the
former might have outstripped the latter in a migration from the same area.

There are two illuminating features regarding the prevalence of males in the catches.
From the statistics of the British Museum (Nat. Hist.) it is recorded that males were
slightly in excess of females at the South African stations as well as those of the De-
pendencies of the Falkland Islands (Mackintosh and Wheeler, 1929, p. 322). From the
same statistics there is evidence that males outnumber females considerably in "Fin"
seasons; while in seasons when Fins are scarce, generally, though not invariably,
females slightly outnumber males. These indications point to somewhat different areas
or methods of concentration of the sexes. The females may concentrate rather farther
from the island or in a more diffuse form than the males. Probably also the females are
more timid and easily scared by the hunters. In any event there is no reason to suppose
that the catch of females is less representative of the ages present than the catch of
males, unless deficiency of lactating and resting females is due to some local distri-
butional factor at present unsuspected.

The apparent undue prominence of immatures in the catches applies to males as well
as females and to Blues as well as Fins. The fact was commented on by Hinton (1925,
p. 162) who, in dealing with the earlier years of the Blue and Fin fishery suggested that
from his data, it appeared that the attacks of the catchers were being directed at
adolescent Finners and sexually immature Blue whales rather than the adults, since
they were easier to kill and certainly easier to handle with tackle designed for the Hump-
back. With the change of objective of the fishery came changes in design and power of
the catchers and improvements in the tackle, so that Hinton's second reason for the
excess of immatures is no longer applicable. There is no doubt, however, that immatures
are easier to kill since they are not as easily scared as the older whales. Also they tend
to run nearer to the coast and are thus more easily accessible. Both these reasons play
their part in increasing the proportion of small whales in the catch.

In recapitulation let us investigate the catch of Finners taken by the Pesca company
during a single season, using all the evidence at our command. In 1929-30 (a "Fin"
season) the company captured 750 whales. Fifteen whales were taken before our arrival
in South Georgia and nine were missed during the season. Our sample of 726 is there-
fore 96-8 per cent of the company's catch.

3-2

370 DISCOVERY REPORTS

We examined 570 Fins of which 272 were female, and 70 of these females were
sexually immature. From the mature females 189 ovarian records were obtained. The
remaining 13 were decomposed to such an extent that the number of corpora lutea could
not be determined with certainty. They were sexually mature, however, and we have
to consider the immatures in relation to our full sample of 272 ; while the age series
determined by the corpora lutea represents the distribution of age groups among 189
sexually mature whales. In Table V the percentages have been calculated as though
the entire sample had been examined.

The percentage of immatures in our sample of the population is 257. Among them
are five of small size (see Table IV), three of which had evidently not long been weaned.
They were less than one year old. The other two had two series of scars, suggesting
from this together with their size that they were very late calves of the previous season
and were thus in their second year when killed. From the two series of scars marking
33 immatures, with one whale of 167 m. with one scar series, and the two small ones
already noted, we have 36 whales more than one but less than two years old. Whales of
nearly three years include 26 immatures as well as ten mature whales with two series of
scars and 53 with three. The mature whales are, however, included in the estimations
determined by the corpora lutea and we have already seen that 40 are nearly three or
nearly four years old and twelve are nearly four or five. The apparent disporportion of
immatures is now explained by the three successive year groups that are represented in
the catch. The earliest of these can for practical purposes be neglected, for not only is
the appearance of its representatives in the catch subject to penalty and only due to
accident, but it is almost certain that the figure obtained conveys a wrong impression of
the population. It is probable that large numbers of the newly weaned whales never
make the migration and this explanation covers the shortage of lactating and resting
whales of all ages which is a feature of the South Georgia catch. The large catches of
immature whales made at Saldanha Bay, South Africa, suggest that many of the early
calves are weaned and left in the coastal waters of the breeding areas, while late calves
are weaned during migration or even after arrival in southern waters. In support of this
I can point out that, although lactating whales are few, they are certainly more numerous
in the catches at the end of the South Georgia season than at the beginning. This must
not be taken to imply necessarily the identification of the South African whales with
those of South Georgia. It is supposed, however, that similar conditions prevail
wherever the breeding area may be.

The original group of "immatures verging on maturity", i.e. the immatures more
than one but less than two years old, now form but 13 per cent of the sample catch.
On returning to the north, many of this group will commence to breed and they will
be accompanied on the next southern migration by the rest of the group (9I per cent
of the sample catch) which will not become mature until their return to northern waters
in the following winter when they are just three years old.

With the aid of the estimations of age from Table II we can now sketch the relative
age composition of our sample as follows :

THE STOCK OF WHALES AT SOUTH GEORGIA

Table V

Female Fin zvhales. Estimated age composition of the
observed catch in season 1929-30

371

Number

Percentage of sample
(272)

Sexually immature females

Less than one year from birth

3

1

More than one and less than two years from birth

36 1,
26 j62

13

More than two and less than three years from birth

9*

Sexually mature females

More than three and less than five years from birth

4°

14!

More than four and less than six years from birth

12

4*

More than five and less than seven years from birth

27

10

More than six and less than eight years from birth

15

5l

More than seven and less than nine years from birth

H

5

More than eight and less than ten years from birth

7

2i

Physically and sexually mature females

More than nine and less than eleven years from birth

17

6

More than ten and less than twelve years from birth

7

2i

More than eleven and less than thirteen years from birth

12

4l

More than twelve and less than fourteen years from birth

12

4*

More than thirteen and less than fifteen years from birth

12

4!

More than fourteen and less than sixteen years from birth

4

1*

More than fifteen and less than seventeen years from birth

5

2

More than sixteen and less than eighteen years from birth

1

Less than J

Remainder of the sample catch

4

4

Total number of sexually mature females = 189

The greatest damage inflicted on the population now appears to be spread over the
period between the second and seventh year of life. There is, however, another way in
which these results bear upon the question of depopulation. If it be allowed that
sexually mature females produce a calf every two years and that every other calf is
female, then in the distribution of the population shown in Table V the increment
per two years of female calves more than one and less than three years from birth
would be one quarter of the mature females, i.e. 47. But the rate of destruction is
shown to be 62, so that on this evidence, in so far as Table V is representative of
the stock, severe damage is being done. Shortage of old whales does not appear to
have been a feature of season 1929-30, but the absence of lactating and resting whales
is quite strongly marked. The numbers taken are too small to respond satisfactorily to
mathematical treatment, and indeed, the figures that have been used in computing the
rate of reduction must not be accepted too literally. Apart from the errors introduced
by taking five seasons together there are probably more serious faults, due to the over-
lapping of the age groups, which are impossible to estimate. The conclusions must
therefore be regarded as indicative of the state of affairs only.

The crux of the problem is the identity of the whales visiting the area in successive

372 DISCOVERY REPORTS

seasons. Only proof of identity can give value to the rate of reduction and it does not
seem possible to obtain this proof indirectly. The failure of the Humpback fishery has
been adduced as evidence to show the results of continual attack upon a closed stock,
and although some of the data concerning Fins and Blues, especially as regards their
movements and the presence of diatom film, appear antagonistic to the general theory
of a recurrent population, there is, at any rate, a possibility that this population exists as
an entity, and there the matter must be left until marking of whales on a large scale is
successfully carried out.

SUMMARY

The whale population of the South Georgia area has been investigated in the light
of evidence from the corpora lutea and scars on the epidermis.

The catch has been analysed showing that maxima of physically mature females occur
in December, sexually mature not physically mature females in January, and sexually
immature females in February.

Analysis of the age groups of mature whales determined by the corpora lutea over
five seasons shows a rate of reduction of 26 per cent and a suggestion that the remaining
mature whales, after this percentage is withdrawn, might keep the population at its
present level notwithstanding a deficiency in numbers among lactating and resting
whales. A shortage of the older age groups is demonstrated, which, should the missing
whales be eliminated elsewhere, must lead to an adverse balance.

On evidence obtained from scars formed in northern waters it has been possible to
divide the immature catch into three age groups.

The importance of definite evidence regarding the recurrence of the same stock in
these waters is pointed out and the main features of the lines of research into this
problem are put forward.

LIST OF LITERATURE

Bennett, A. G., 1920. On the occurrence of Diatoms on the skin of whales. Proc. Roy. Soc. Lond., B, xci,

P-352-
Harmer, S. F., 193 1. Southern Whaling. Proc. Linn. Soc. London.
Hinton, M. A. C, 1925. Report on the papers left by the late Major Barrett-Hamilton, relating to the whales

of South Georgia. Crown agents for the Colonies. London.
Hjort, J., Lie, J. and Ruud, J. T., 1932. Norwegian pelagic whaling in the Antarctic. I. Whaling grounds in

1929-30 and 1930-31. Hvalradets skrifter, Nr. 3.
Kemp, S. and Bennett, A. G., 1932. On the distribution and movements of whales on the South Georgia and

South Shetland whaling grounds. Discovery Reports, vi, pp. 165-90.
Mackintosh, N. A. and Wheeler, J. F. G., 1929. Southern Blue and Fin Whales. Discovery Reports, 1,

PP- 257-54C-
Matthews, L. H., 1931. A History of South Georgia.
Olsen, 0., 1913. On the External Characters and Biology of Bryde's Whale (Balaenoptera brydei) a new

Rorqual from the Coast of South Africa. Proc. Zool. Soc. Lond., pp. 1073-90.
Risting, S., 1928. Whales and whale foetuses. Statistics of catch and measurement collected from the Norwegian

Whalers' Association, 1922-5. Rapp. Cons. Expl. Mer., l, pp. 1-122.
Risting, S., 1930. Norsk Hvalfangst-tidende.
Wheeler, J. F. G., 1930. The age of Fin whales at physical maturity. Discovery Reports, 11, pp. 403-34.

DISCOVERY
REPORTS

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands

Vol. IX, pp. i-vi

TITLE-PAGE AND LIST OF CONTENTS

CAMBRIDGE

AT THE UNIVERSITY PRESS
1934

Price ninepence net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo
Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

WALTER Wy.?toy LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. IX, pp. 1-64

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of tfie Falkland Islands

HYDROLOGY OF THE BRANSFIELD STRAIT

by
A. J. Clowes, M.Sc, A.R.C.S.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1934
Price ten shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. IX, pp. 65-160

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands

DISTRIBUTION OF THE MACROPLANKTON
IN THE ATLANTIC SECTOR OF THE

ANTARCTIC

by
N. A. Mackintosh, D.Sc.

CAMBRIDGE

AT THE UNIVERSITY PRESS

1934
Price fifteen shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

PRINTED

1 N GREAT BRITAIN

BY

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. IX, pp. 161-174, plate I

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands

THE SUB^ANTARCTIC FORMS OF
THE GREAT SKUA

fCATHARACTA SKUA SKUA)

by
J. E. Hamilton, M.Sc.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1934
Price three shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo
Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

WALTBR

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. IX, pp. 175-206, plates II-X1V

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands

THE MARINE DEPOSITS OF THE
PATAGONIAN CONTINENTAL SHELF

by
L. Harrison Matthews, M.A.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1934
Price nine shillings net

Cambridge University Pre9s
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

PRINTED

I N GREAT BRITAIN

BY

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. IX, pp. 207-214

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovemment of the Dependencies of the Falkland Islands

THE DEVELOPMENT OF RH1NCALANUS

by
Robert Gurney

^tsn

CAMBRIDGE

AT THE UNIVERSITY PRESS

1934
Price two shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BV

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. IX, pp. 215-294, plates XV, XVI

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands

NEMERTEANS

FROM THE SOUTH ATLANTIC AND

SOUTHERN OCEANS

by
J. F. G. Wheeler, D.Sc. *

CAMBRIDGE
AT THE UNIVERSITY PRESS

193*
Price sixteen shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo
Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. IX, pp. 295-350, plates XVII-XXII

Issued by the Discovery Committee, Colonial Office, London
on behalf of the (government of the Dependencies of the Falkland Islands

THE SEA^FLOOR DEPOSITS

L GENERAL CHARACTERS AND DISTRIBUTION

by
E. Neaverson, D.Sc, F.G.S.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1934
Price fourteen shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo
Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

WALTER

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. IX, pp. 351-372

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands

ON THE STOCK OF WHALES AT
SOUTH GEORGIA

by
J. R G. Wheeler, D.Sc.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1934
Price four shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

MacmiUan

Tokyo

Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

$F«

v., «{.-,

wtis

tin

Wi<

**£&&&.

W^

m

£ -V-r^

■«M- 'S

if*

*£**.-

SB*

I .*.

Il»

£

4

V^Vsih

i

JV

BTV

rr .*•

^Ti '

lEir

I

3WS