m
DISCOVERY REPORTS
VOLUME XXIII
-^ /N ^
/V7
DISCOVERY REPORTS
Issued by the Discovery Committee
Colonial Office, London
on behalf of the Government of the Dependencies
of the Falkland Islands
VOLUME XXIII
<^
W^ 0.
/
CAMBRIDGE
AT THE UNIVERSITY PRESS
1947
Printed in Great Britain at the University Press, Cambridge
{Brooke Crutchley, University Printer)
and published by the Cambridge University Press
(Cambridge, and Bentley House, London)
Agents for U.S.A., Canada, and India: Macmillan
WOODS
HOLE,
CONTENTS x^^ASs.
THE GUT OF NEBALIACEA (published 17th November, 1943)
By Helen G. Q. Rowett page 3
ON A SPECIMEN OF THE SOUTHERN BOTTLENOSED WHALE, HYPEROODON
PLANIFRONS (published 23rd March, 1945)
By F. C. Eraser, D.Sc page 21
REPORT ON ROCKS FROM WEST ANTARCTICA AND THE SCOTIA ARC (published
27th June, 1945)
By G. W. Tyrrell, A.R.C.Sc, D.Sc, F.G.S., F.R.S.E.
Foreword, by J. M. Wordie, M.A page 39
I. Petrography of the South Shetland Islands, West Antarctica 41
II. Petrography of Rocks From the Graham Land Peninsula and Adelaide Island, West
Antarctica 66
HI. Petrography of Rocks from the Elephant and Clarence Group 76
IV. Petrography of Stones dredged from the .Vicinity of the Shag Rocks ... 89
V. Petrography of the South Sandwich Islands 92
THE DEVELOPMENT AND LIFE-HISTORY OF ADOLESCENT AND ADULT
KRILL, EUPHAUSIA superb a (published 20th July, 1945)
By Helene E. Bargmann, Ph.D.
Introduction page 105
Development 106
Average Growth Rate 120
Factors Influencing Growth Rate 1 28
Conclusions 13°
Bibliography 131
Appendix 132
THE ANTARCTIC CONVERGENCE AND THE DISTRIBUTION OF SURFACE
TEMPERATURES IN ANTARCTIC WATERS (published 28th January, 1946)
By N. A. Mackintosh, D.Sc.
Part I. The Antarctic Convergence page 179
Part II. The Distribution of Surface Temperature in Antarctic Waters . . . -194
References 204
Appendix. Table 9 205
Notes on the Plates 211
Plates I-XIV following page 212
61172
vi CONTENTS
NEBALIOPSIS TYPICA (published 21st October, 1946)
By H. Graham Cannon, Sc.D., F.R.S page 21s
REPORT ON TRAWLING SURVEYS ON THE PATAGONIAN CONTINENTAL
SHELF (pubhshed 20th December, 1946)
By T. John Hart, D.Sc.
Foreword, by N. A. Mackintosh page 226
Introduction ^^7
General Account OF THE Fish Fauna 251
Distribution and General Notes on the Species 259
Features of General Biological Interest 3^2
Prospects of Commercial Development 3^7
References 392
Appendices 39°
Plate following page 408
[Discovery Reports. Vol. XXIII, pp. 1-17, Octobei- 1943 .J
THE GUT OF NEBALIACEA
By
HELEN G. Q. ROWETT
CONTENTS
Introduction 3
Methods 3
The structure of the gut of Nebalia bipes (Fabricius) .... 3
The structure of the gut of Nebaliella extrema (f. Thiele) ... 3
I. Fore-gut 3
II. Mid- and hind-gut 6
Musculature 7
The structure of the gut of Nebaliopsis typica (Sars) .... 8
Fore-gut 8
Mid- and hind-gut 8
Musculature 9
The structure of the gut of Paranebalia longipes (Wilemoes Suhm) 1 1
Mode of functioning of the gut of Nebalia bipes 12
Mode of functioning of the gut of Nebaliella extrema . . . . 13
Mode of functioning of the gut of Nebaliopsis typica .... 14
Conclusions 15
Bibliography 17
WOOD'
THE GUT OF NEBALIACEA V ^J9if'
MASS
By Helen G. Q. Rowett, Grisedale Scholar, Manchester University
INTRODUCTION
MUCH attention has been paid by Cannon, Manton, Lowndes and others to the ' feeding mechanisms '
of Crustacea, but no attempt has so far been made to correlate changes in the structure of the gut
with the type of food available and the condition in which it is passed into the mouth. For this purpose
it is necessary to compare members of one group which have different habits and habitats rather than
isolated examples from different groups. A survey of the Nebaliacea has therefore been made with
the object of discovering how far the structure of the gut shows group resemblances and how far it
may be associated with the environment and habits of the species concerned.
METHODS
With the exception of Nebalia bipes, material for this investigation was limited to Discovery specimens
of Nebaliopsis typica and Nebaliella extrema kindly made available by Professor Cannon.
Reconstructions were made using transverse sections of Nebaliopsis typica specimen E {Discovery
Reports, 1931, vol. in) and sagittal sections of half of specimen F2 (op. cit.) and the unsectioned
other half of this specimen. From these reconstructions Figs. 4, 5, 6 C and 7 A were made.
The single specimen of Nebaliella extrema was sectioned transversely, and the reconstructions
shown in Figs. 2, 3, 6 B and 7 C were obtained.
Besides sectioned material the cast skins of Nebalia bipes were examined and living specimens were
watched in a jar with sea water and some of the mud from their usual habitat in Rum Bay, Plymouth,
and also isolated in dishes under a microscope. Carmine was fed to some and, using strong illumina-
tion, the passage of the red particles through the gut was easily seen through the semitransparent body.
THE STRUCTURE OF THE GUT OF NEBALIA BIPES (Fabricius)
The structure of the gut of Nebalia was described in great detail by Claus (1889) and later by Jordan
(1909, 19 1 2) in papers comparing the pyloric section with that of Idothea, Ganimanis and Astacus.
A detailed description need not, therefore, be given here, but for the sake of clarity in making com-
parisons with other types Figs, i, 6 A and 7 B have been made which show the various parts and
associated musculature.
In one important respect, however (which is not mentioned by Jordan), Claus's description is
definitely incorrect. The structures which he describes as chitinous pads with thickened striations
are actually rows of very strong evenly set setae (^.^.1 in Fig. i) which with the spines (g.s.2) form a
tube in which grinding takes place.
THE STRUCTURE OF THE GUT OF NEBALIELLA EXTREMA {i. Thide)
I. FORE-GUT
The structure of the oesophagus and cardiac region of the stomach of A^. extrema is, as Thiele (1905)
says, very like that of Nebalia. Certain important differences may, however, be noted.
There are many fewer setae throughout. The anterior median projection {a.m.p.) is much reduced
and the lateral fanlike plate of setae which is found on the right side only in Nebalia {l.p. Fig. i) is
absent. The spatial relations of the homologous parts are so altered that there is no grinding tube
such as is seen in Nebalia. As shown in Figs. 2 and 3 A, the spines {g.s.o) are ventral instead of dorsal
^
s
a
QiO to 3^fc:
Ta'tioiiii
o
O
e
C
o
OJ CA
XI SO
(1)
a<j
rt
• •.
1)
tfl
CO
<3)
U
O
E
n
o.
s
o
o
V)
•o
— DO
h a-
o _
0-, lu
o ^
'5b -Oh
u 3
o
V
-a
V — I
« o
C3
o o
=3 «
i-H a-
ii .2
cd be
— ' o
J3 o
.2 S
S -a
c "«
(L)
> ■-
C rt ■
':! "^
a-. ,
-c; ti3 >-
^ J2
O 1-
C .5
P m
t* It;
-T3 2
^*H "Ts -^ rt
° .. •' a:;
j3 0*
bc e
■ c
o
tx >,
o
o
1) rt
o o
tie :r
o ^
T3 .
■a 5
■ - o-S.
o S c
CO O >
THE GUT OF NEBALIACEA S
to the regular row of setae (gs.i). Of equivalent functional significance is the grinding organ formed
by the setae alone (g.s.i), which slope diagonally backwards and outwards almost parallel to, and
rubbing against, a slightly setose horizontal shelf of each lateral wall. The more anterior of these
grinding setae are longer than the posterior ones, their tips curve upwards and a pair of small ridges
(d.r.') lie medial to the basal thickening on which they are inserted.
a.m
oes
mo.
Fig. 2. Diagram of the right half of the fore-gut of Nebaliella extrema. A, region at which Fig. 3 A was cut; B, region at
which Fig. 3 B was cut. a.m.p. anterior median projection; d.ca. dorsal caecum; d.gl. digestive glands; d.gl.o. opening of the
digestive glands; d.p. dorsal process; d.r. dorsal ridge; d.r.' small ridge on the dorsal ridge; ^.^.j strong setae; ^.y., short stiff
spines; i. intestine; l.p. lateral pad; m. mandible; mo. mouth; yn.p. median pad; oes. oesophagus;/), long projections; v.p.ch.
ventral pyloric chamber with opening of the digestive glands ; outUne of the lumen of the gut laterally.
Ventral to the row of spines {g.s.2) is a slight ridge on each lateral wall behind which is a strong
contractor muscle. This ridge marks the division between the oesophagus and the stomach.
There is no distinct division into cardiac and pyloric regions, but, posteriorly, where the spines
ig.s.i) and setae {g.s.i) cease, the lateral walls approach one another more closely and their surfaces
6 DISCOVERY REPORTS
are soft and irregularly corrugated. In this region the dorsal glandular caeca [d.ca.) open into the
dorso-lateral angles of the lumen. Slightly posterior to these openings the dorsal ridge [d.r.) becomes
free from the dorsal wall and projects a short distance as a soft pad {d.p.). Similarly swellings of the
lateral and ventral walls split off together from the gut walls and form a trilobed chitin-covered
process {l.p. and rn.p.), each of whose lobes bears a long projection {p.) converging dorsally as shown
in Fig. 3 B.
The lateral lobes bear strong setae and are undoubtedly homologous with the lateral finger-like
processes of Nebalia, but the homologies of the median ventral lobe are uncertain. Thiele suggests
that it may be the sum of the two lateral ridges {l.r.) of Nebalia. If this were the case it might be
expected that some trace of the double nature would remain, but none could be found. It is more
likely that it is homologous with the ventral cardio-pyloric valve, the shifting of which posteriorly is
a slight change comparable with the other differences between the two species. A third alternative
is that the small pyloric pad of Nebalia (p.p. Fig. i) has been greatly enlarged, but this, like Thiele's
suggestion, is a major alteration involving the disappearance of the cardio-pyloric valve.
dr. d.r'
oesn
m.
Fig. 3. A. Anterior region of the fore-gut of Nebaliella looking forwards into a piece cut at region A of Fig. 2 to show the
relationships of the parts hidden by the median structures, a.m.p. anterior median projection; d.ch. dorsal channel ; d.r. dorsal
ridge; d.r.' small ridge on dorsal ridge ;^.s.i strong setae; 0.5.2 short stiff spines; m. mandible ; ow. oesophagus; o«.r. oesophageal
ridges. B. Posterior region of the fore-gut and the entrance to the intestine of Nebaliella looking backwards from region B
of Fig. 2 to show the spatial relations of the pads and the projections thereon in the pyloric part of the gut. d.gl. digestive
glands; /. intestine; l.p. lateral pad; m.p. median pad; p. long projections; v.p.ch. ventral pyloric chamber.
II. MID- AND HIND-GUT
Besides the dorsal glandular caeca already mentioned, Nebaliella resembles Nebalia in having three
digestive gland caeca {d.gl.) on either side. These unite and open into the ventro-lateral corners of
the ventral pyloric chamber {v.p.ch.) immediately posterior to the tripartite process. The openings
are smaller than those in Nebalia. Ventral glandular caeca were not found.
The lumen of the intestine is relatively wider than in Nebalia, and for a considerable distance is
roughly triangular, with one angle dorsal and two ventro-lateral, and with walls of highly vacuolated
cells. Passing posteriorly the cells become less vacuolated and a striated border appears on them.
Then the outline of the gut becomes oval and the striated border deeper. Finally the cross-section
THE GUT OF NEBALIACEA y
is almost circular, the cells are very dense and closely packed together, and some are elongated and
project as ridges into the lumen.
As in Nebalia the intestine and digestive glands are embedded together in a loose tissue of highly
vacuolated cells.
There is a very much reduced rectal gland and an anal chamber comparable to these structures
in Nebalia.
^■h.-^
a.m.
I
n
n
Fig. 4. Diagrammatic reconstruction of the right half of the fore-gut of Nebaliopsis typica in three sections (I-III). a.h. anterior
horn; a.m.p. anterior median projection; d.r. dorsal ridge; gl.r. glandular region; /. labrum; l.th. lateral thickenings; pa.
paragnaths ; pl.w. plated walls.
MUSCULATURE
The similarities between the musculature of the fore-gut of Nebaliella and that of Nebalia are very
striking, as the diagrams (Figs. 6 A, B) show. Differences are that the lateral dilator muscles {l.dil.)
of the oesophagus have five points of insertion as compared with two in Nebalia. The muscle corre-
sponding to the small median projection muscle is greatly enlarged. The anterior dorsal dilators
(a.d.dil.) are also enlarged but the posterior dorsal dilators (p.d.dil.) are reduced, though there is a
great thickening of the chitin at their point of insertion. The strong circular muscle (cont.) which is
so conspicuous in Nebalia is present in Nebaliella also, though slightly reduced. No muscles could
be found in the groove between the anterior horns of the stomach where Nebalia has a few thin
strands of fibres.
8 DISCOVERY REPORTS
In addition to these muscles which have their homologues in Nebolia, Nebaliella has a pair of very
strong muscles (t.p.m.) which stretch ventro-laterally from a thickening of the chitin of each lateral
lobe of the trilobed process immediately anterior to the point where the process splits from the
gut wall.
THE STRUCTURE OF THE GUT OF NEBALIOPSIS TYPICA (Sars)
FORE-GUT
The external features of Neboliopsis (Cannon, 193 1, pi. xxxii) indicate that it is a highly specialized
member of the Nebaliacea, and this is confirmed by the internal organization. Even in the gastric
mill group resemblances are few.
Fig. 4 shows the reconstruction of the right half of the fore-gut in three sections.
The molar processes of the mandibles are reduced and do not project into the mouth as in the
other forms. The mouth is a transverse slit between the labrum (/.) and the paragnaths {pa.). These
can be retracted by strong muscles, thus uncovering a flat plate of chitin with a median antero-
posteriorly directed slit leading into the stomach. The latter slit can be opened widely by the dilator
muscles [l.dil.), which slope upwards and outwards from the oesophageal wall. Very great variation
of both the size and the shape of the gape is thus possible by the combined action of these two slits.
There is no distinct separation into oesophagus, cardiac and pyloric parts, but the region surrounded
by the horizontal circular muscles, and to which the lateral dilators are attached, may be considered
oesophageal in comparison with Nebolia.
Setae are entirely absent from the gut. Except in the most anterior and dorsal regions the chitinous
lining of the lateral walls has the appearance of crazy-paving owing to the presence of grooves over
the junctions between the individual cells of the supporting tissue (Fig. 5 C, D). Posteriorly these
grooves are less distinct. The paved parts of the walls lie very close together (Fig. 5 C) and provide
a good gripping surface.
An anterior median projection (a.m.p.) is present as in Nebalia, but much reduced. There is a dorsal
ridge [d.r.) which is strongly chitinized and slightly grooved in the region immediately dorsal to the
mouth, but it arises anteriorly as a soft pad and becomes so again posteriorly. The lateral walls have
thickenings {l.th.) against which this ridge bites. The thickenings of the chitin are prolonged into short
anterior projections of the stomach, and may be homologous with the slight thickenings at the bases
of the spines (^-^.2) in Nebalia. There are no grinding tubes, but a strong grinding or biting action
probably occurs between these heavily chitinized regions.
Posteriorly where the dorsal ridge becomes a soft pad, the side walls open out slightly and they also
become soft. The chitinous lining of the fore-gut ends raggedly ; the walls become glandular and lose
the thick muscle sheath which encircles them throughout the stomach region.
MID- AND HIND-GUT
The glandular region mentioned above marks the beginning of the mid-gut. Here the anterior
digestive diverticula (d.gl.), which are comparatively small (Fig. 6 C), and extend only a short distance
forwards and which are probably homologous with the dorsal caeca of Nebalia, open by irregular
apertures. Some of these are small channels passing through the glandular region, but the largest
opens below a flap (gl.r. in Fig. 4) directly into an immense digestive sac {d.s. in Fig. 7 A), which
widens out suddenly and almost completely fills the body cavity back to the end of the fourth abdominal
segment, and which may be homologous with the digestive caeca of Nebalia, though it is difficult to
be certain of any homologies when the specialization is so great. It is important to note that this is
a plain sac without any convolutions and with only a few thin septa rising from its walls. The enlarge-
ment therefore does not provide a great deal of extra surface area for absorption, but it does give a
THE GUT OF NEBALIACEA 9
large volume for storage. The walls of the sac are formed of a thin layer of pavement epithelium made up
of huge highly vacuolated cells with a striated border and an average diameter of o-i mm. (Fig. 5 A, B).
A very thin basement membrane lies behind them.
The intestine is a very narrow tube lying dorsal to this sac (Fig. 7 A). In places the lumen is so
small that it is hardly distinguishable, but posterior to the end of the digestive sac it opens out into
a wider rectal region. A muscular sphincter separates it from a short proctodeum. No rectal gland
was found.
D
Fig. 5. A. Cells of the digestive sac in surface view. B. Same in section showing striated border. C. Section of the plated
side walls of the stomach showing the grooves in the chitin as it is laid down over each cell and the closeness of the opposite
walls of the gut : ch. chitin ; gr. intercellular groove in the chitin ; /. lumen of the gut ; n. nucleus. D. Surface view of the chitin.
MUSCULATURE
The musculature of the fore-gut of Nebaliopsis is shown in Fig. 6 C. The similarities to the other
Nebaliacea are striking. The oesophagus and stomach are sheathed in strong bands of circular muscles
[h.circ. and v.circ). These bands are many times thicker than the corresponding ones in Nebalia,
while the tissue between them and the chitin is comparatively much reduced. They cease abruptly at
the end of the fore-gut.
Acting antagonistically to these circular muscles are the dilator muscles. The dorsal dilators (d.dil.)
are probably homologous with the anterior dorsal dilators [a.d.dil.) of Nebalia, as the groove muscles
which lie close to the dorsal ridge between the anterior horns of the stomach run between them and
DISCOVERY REPORTS
are inserted on the dorsal wall more posteriorly. In Nebaliopsis the groove muscles consist of a very
thick bundle of fibres passing from the dorsal ridge as mentioned above to the anterior wall of the
oesophagus ventral to the anterior median projection, while only four pairs of slender strands were
found in Nebalia. The anterior lateral dilators {l.dil.^ and l.dil.^) differ only in that they slope dorsally
p.ddii
a.d.dil.
a.h.m.
y. arc.
arc.
3. oes.
pj.dil.
Fig. 6. Diagrams to show the musculature of the fore-gut of: A, Nebalia; B, Nebaliella; C, Nebaliopsis. a.d.dil. anterior
dorsal dilators; a.h.m. muscles of the anterior horns of the stomach; a.m. median anterior muscle; a.oes. anterior oesophageal
muscles; a.s.m. small anterior muscles; conl. strong contractor muscle; c.p.m. cardio-pyloric muscles; d.ca. dorsal caeca;
d.dil. dorsal dilator; d.gl. digestive glands; d.s. digestive sac; h.circ. horizontal circular muscles; /. labrum; l.dil.^, l.dil.^,
l.dil.^, l.dil.^, and l.dil.^, points of insertion of the lateral dilators; m. mouth; m.p.m. median projection muscle ;/)a. paragnaths;
p.d.dil. posterior dorsal dilators; p.l.dil. posterior lateral dilators; p.m. posterior muscles; p.oes. posterior oesophageal
muscles; t.p.ni. muscles of the trilobed process; v.ca. ventral caeca; v.circ. vertical circular muscles ; v.l.dil. ventro-lateral
dilators; v.m. small ventral muscles.
THE GUT OF NEBALIACEA
n
instead of ventrally. A small ventro-lateral dilator muscle {v.l.dil), three small muscles (a.s.m.)
extending anteriorly and a muscle (p.m.) pulling posteriorly from a triple insertion on the gut were
found. The anterior muscles (a.m. and a.h.m.) are small but correspond to similar muscles in Nebalia.
The digestive sac has dorsally and ventrally a pair of longitudinal muscle bands. Small segmental
muscles which support the thoracic limbs lie in the wall of the sac and cause slight ridges in it. No
musculature was found on the intestine and only a small sphincter at the anus {a.sp.).
Fig. 7. Diagrams of the right halves of A, Nehaliopsis, B, Nebalia, and C, Nebaliella, showing the position of the gut in the
body cavity of each. (The positions only of the limbs are indicated in B and C.) a. anus; a.ch. anal chamber; a.sp. anal
sphincter; d.gl. digestive glands; d.p. dorsal process projecting down the intestine; d.s. digestive sac;/.^. fore-gut; i. intestine;
m. mouth ; md. mandible ; r. rectum ; r.ca. rectal caecum ; sep. septum.
THE STRUCTURE OF THE GUT OF PARANEBALIA LONGIPES
(Wilemoes Suhm)
No specimens of Paranebalia were available for examination. Nevertheless, to complete the survey
of the Nebaliacea Thiele's description may be quoted. He found that the gut is, on the whole, like
that of Nebalia, but records these differences: (i) strong spines are present under the long setae on
the ventral side of the ' hypopharynx ' (the ventral lip) ; (2) the rows of setae on the dorsal ridge in the
pyloric region do not extend far back, and the funnel formed by the dorsal process bears no setae ;
(3) two ventral bristle plates take the place of the small lateral ridges {l.r.) of Nebalia. These plates
are composed of a transverse row of thick setae which meet each other across the lumen of the gut.
The midmost setae are longest.
Thiele's diagram does not show an anterior median projection, lateral plate setae (but he has depicted
the left side), setae on the walls of the oesophagus or cardiac region, or a ventral cardio-pyloric valve.
12 DISCOVERY REPORTS
but he does not mention these points as differences. Therefore either his description is inadequate
or his diagram incorrect. In the absence of further material the answer to this question cannot be given.
MODE OF FUNCTIONING OF THE GUT OF NEBALIA BIPES
The mode of functioning of the gastric mill of Nebalia bipes may be deduced from evidence furnished
by the structure of its parts, the distribution of particles within the gut, and also from observation of
living animals.
Specimens kept in shallow water in a jar, on the bottom of which was mud from their natural
habitat, were observed undisturbed. They occasionally swam about, but usually lay on the surface
of the mud (often in the shadow of large pieces of seaweed or stones), where it could be seen that
the thoracic limbs seldom ceased their regular rhythmic motion even when the animal as a whole was
stationar}^ They appeared to burrow only when disturbed.
When isolated in small dishes and placed under the microscope they swam rapidly but at times lay
quiescent and could then be studied. The thoracic limbs continued to beat unless the specimens
were kept long in these conditions when they frequently became completely inactive for considerable
periods though often reviving later. When all the movement of the limbs ceased in this manner the
rate of heart beat slowed down and large particles were seen floating in the blood stream. This effect
has not been studied in detail but it is probably caused by the unnatural conditions in the dishes, as
no such long pauses in the motion of the thoracic limbs were noted when watching the animals
in the jar.
The currents produced by the movements of the thoracic limbs bring particles to the filter apparatus
(Cannon, 1927). There is thus a continuous supply of food depending only on the concentration of
suspended matter in the water. If excessive amounts are collected the particles are gathered into balls
and shot out ventrally in the anterior region of the carapace. This mechanism probably helps to
prevent the filter apparatus from being choked with mud when the animal is burrowing and also
indicates that the movement of the thoracic limbs serves another purpose besides feeding. It is
possible that the continuous current of fresh water is necessary for respiration and must be maintained
whether it bears many or few particles. Thus normally there is a constant stream of filtered material
being passed to the mouth. Large particles have a preliminary grinding by the maxillary endites
(see Cannon, 1927, for details), and are also ground between the mandibles.
Rows of setae on the lips prevent pieces from falling off into the grooves on either side of the
mandibles and direct them into the oesophagus. Strong contraction of the circular muscles keeps the
passage from the oesophagus to the stomach closed most of the time, but periodically these muscles
are relaxed, and simultaneously the lateral dilators work actively causing a 'puff' of particles to pass
into the stomach and swiftly back into the pyloric region and the intestine. Setae on the walls of the
oesophagus point dorsally and prevent backflow. All the gut muscles move violently during this
operation.
The anterior median projection and the lateral plate setae {a.tn.p. and l.p.) help to direct the current
round the angle between the oesophagus and the stomach so that large amounts of material do not
pass dorsally and choke the grinding tubes. The setae are, however, not close enough to form a strict
filter, and some particles pass up into the grinding tubes and are ground between the setae (^.^.1)
and the vertical ridges {v.r.) and spines {g.s.o), which are rubbed across one another by a complex
circular and see-saw motion of the dorsal ridge, easily seen in living specimens and probably caused
by alternating contraction of the dorsal dilators combined with peristalsis of the circular muscles.
Only liquid was found in the dorsal channels (d.ch.), which are open posteriorly, and it is possible
that a secretion from the dorsal caeca may flow forwards in them and be poured upon the food as it
is being ground up in a manner analogous to Yonge's suggestion for Nephrops (Yonge, 1924).
THE GUT OF NEBALIACEA 13
The particles from the grinding tubes are passed back and on to the long setae of the pyloric region.
The narrowness of the lumen of the gut in the posterior part of the cardiac section and the presence
of the ventral cardio-pyloric valve {v.v.), whose tip moves violently describing an ellipse, causes
particles which have been driven directly back without secondary grinding in the grinding tubes to
pass up on to these setae also. The latter filter off the larger pieces and bear them back beyond the
openings of the digestive glands and far down the intestine in the tubular extension of the dorsal
process. The smaller pieces fall through into the ventral pyloric chamber and pass into the digestive
glands {d.gl). Muscle bands on the walls of these glands probably cause pumping in and out of fluid
bearing small particles as in Nephrops (Yonge, 1924). Certainly in ink- fed specimens grains were
found far into these caeca indicating that particles from the stomach are passed into them. Particles
appearing like finely ground food were frequently found in them also. No ink grains or other particles
could be seen in either the dorsal or ventral caeca {d.ca. and v.ca.). This suggests that absorption as
well as digestion probably takes place in the digestive glands, while the dorsal and ventral caeca
secrete a digestive fluid only.
The arrangements of the gastric mill of Nebalia are such that there is continuous action of the
secondary grinding apparatus which increases the number of particles small enough to pass into the
digestive glands, while at the same time the animal is able to deal casually with the large quantities
of potential food which are automatically available and whose amount depends only on the concen-
tration of particles in suspension in the water filtered and the proportion of inorganic to organic matter.
MODE OF FUNCTIONING OF THE GUT OF
NEBALIELLA EXT REM A
Nebaliella is a mud-living form. The eyes and rostrum have been shown by Cannon (1931) to be a
mechanism whereby mud is prevented from entering the space within the carapace and choking the
filter apparatus as the animal burrows. Variations in the completeness of closure of this apparatus
control the current entering the filter chamber. Particles found amongst the mouthparts and also
within the gut include large pieces of diatom skeletons, radiolaria, and many unidentifiable broken
pieces showing that the animal is an indiscriminate mud feeder and also that it can deal with relatively
coarse filtered food. That these particles are present far down the intestine indicates that there is no
very efficient grinding of the food. It is probable that as in Nebalia much material is passed through
rapidly and a little is more carefully treated.
That particles passed on to the mandibles receive only slight grinding before entering the oesophagus
is shown by the state of the food within the gut. The sheath of circular muscles probably functions
in the same way as in Nebalia and releases particles spasmodically.
The angle between the oesophagus and the stomach is more obtuse than in Nebalia, the anterior
median projection is much reduced and the lateral plate setae are absent, but clogging of the grinding
setae {g-s.-^) is prevented by an entirely different mechanism. In the anterior region the edges of the
horizontal shelves of the side walls almost touch the small ridges {d.r.') on the dorsal ridge so that
the channels containing the setae (^.^-i) are nearly closed and only relatively small particles can enter
them. These particles are ground between the setae and the shelf and when fine enough pass between
the former and are found as a ' felty ' layer on the dorsal side of them. This arrangement and the
general reduction in the numbers of setae are almost certainly correlated with the coarse texture of
the food against the passage of which fine setae would have no effect. Such setae would soon be
broken or worn away. The spines (^.^.2) may have some guiding effect on the current, but as they are
so short they are probably only a relic of their homologues in Nebalia.
There is no filter mechanism in the pyloric region. The openings of the digestive glands are small,
and only most minute particles were found within them. In the absence of more material the mechanism
14 DISCOVERY REPORTS
which prevents the openings of the digestive glands from being occluded by large particles is uncertain,
but the following is a possible interpretation of the structures found. The long projections on the
lateral and ventral pads form a triple barrier across the lumen. This barrier is augmented by the long
setae on the lateral pads. As a large mass of food is passed back it comes up against the barrier and
depresses the projections so that a bridge is formed which guides the particles across the pyloric
chamber into the intestine. This movement causes the lobes from which these projections rise to be
bent backwards and downwards to fill a large part of the ventral chamber, occlude the openings of
the digestive glands and at the same time press digestive secretion from the chamber out on to the
food as it enters the intestine. Elasticity for this movement is provided by the large blood sinuses
within the lateral pads and below the ventral one just anterior to the point where they become free.
When the food has passed, the projections spring back to the vertical position assisted by the powerful
muscles in the lateral lobes of the process. The pyloric chamber is thus opened once more and ready
to be refilled with secretion from the digestive glands.
The structure of the digestive glands is such that they are probably almost entirely secretory, while
a little digestion and absorption of the small amount of finely divided material which enters them
may also take place.
The dorsal caeca are entirely secretory as in Nebalio.
The structure of the intestinal wall suggests that besides absorption there is additional secretion
of digestive enzymes especially in the anterior region.
MODE OF FUNCTIONING OF THE GUT OF NEBALIOPSIS TYPICA
The mode of functioning of the various parts of the gut of Nebaliopsis cannot be described with
certainty as yet, for in specimen E hardly any particles were present and in specimen F2 the digestive
sac was full of an almost homogeneous mass resembling coagulated yolk, but two alternative mechanisms
are here suggested, the second being the more probable.
I. Fine particles filtered out of the water by the maxilla and first trunk limb (Cannon, 193 1) may
be sucked into the stomach by the action of the lateral dilators and the circular muscles. There can
be no preliminary grinding owing to the structure of the mouthparts, but once within the gut any
large pieces may be ground between the dorsal ridge {d.r. Fig. 4) and the lateral thickenings {l.th.)
and also between the side walls which approach each other very closely and are heavily chitinized
and grooved (Fig. 5 C, D). There are no setae to hinder direct passage of food into the digestive sac,
therefore it cannot remain long in the fore-gut. Digestive secretion is poured on to it as it passes
the openings of the anterior digestive diverticula {d.gl. Fig. 7 A). These openings are large and
unprotected and particles could easily enter them, but the structure of the glands does not suggest
that any absorption takes place within them.
There is no possibility of any food passing straight from the fore-gut to the intestine as it does
in Nebalia and Nebaliella. Everything must enter the sac where both digestion and absorption
probably take place.
It is difficult to visualize how the sac does not become clogged with indigestible matter, as there
is no apparent means of circulating the material in it. A possible explanation is that a deep pelagic
filter feeder will obtain very little particulate inorganic matter such as is so abundant in and near
the surface of mud so that digestion will be almost complete. Filterable particles are scarce in this
zone, and the blind diverticulum permits the retention of all material until it is thoroughly digested
thus preventing waste.
This suggested mechanism agrees with Cannon's belief that Nebaliopsis is 'entirely a filter feeder'.
His conclusions were reached from a study of the mouthparts alone, particularly important being the
facts that 'the whole mouth armature is extremely soft and unsuited for dealing with large food
THE GUT OF NEBALIACEA 15
particles', and that 'in addition there is a compUcated structure which, in my opinion, must be a
fiher'. The internal organization, however, and other considerations, make probable the following
alternative mechanism, in which the first of these facts plays an important part.
II. Nebaliopsis is probably an egg sucker. As far as information is at present available eggs of
various types have been found in small numbers in all the hauls in which Nebaliopsis has been collected.
The mechanism by which it sucks the yolk from these eggs is probably as follows.
As Cannon reports, the mandibular palps are long and armed at the tips with stout claw-like setae,
which grip the slippery surface of the egg. The eddy currents round the mouth caused by the move-
ments of the trunk limbs and mouthparts also help to keep the egg pressed against the mouth. The
molar processes of the mandibles are soft and useless for biting ; thus the egg is not punctured outside
the mouth, where there would be great risk of the contents being washed away by the water currents
in that region.
While being held close to the mouth one side of the soft egg is sucked into the oesophagus by the
pumping action of the muscles on its walls. It is then gripped tightly by the plated surfaces of the
lateral walls, while the biting action of the dorsal ridge against the dorso-lateral thickenings of the
chitin makes a hole in the egg membranes. The liquid yolk is then pumped through this puncture into
the digestive sac, digestive secretion being poured on to it as it passes the openings of the anterior
glandular caeca. The great thickness of the muscle sheath of the fore-gut may be associated with this
strong pumping action. The steadying action of the mandibular palps and the eddy current round
the mouth are most important during this process. The empty egg case would then be thrown away.
The capacity of the digestive sac is sufficient to accommodate the contents of several average-sized
fish eggs. As suitable eggs are likely to be found in groups near where they have been spawned, and
only at certain times of year, a meal is available only at infrequent intervals. Much food is thus taken
at one time and is stored in the immense digestive sac where it is assimilated slowly as required.
There is hardly any solid waste in this method of feeding, which agrees well with the observed
structure of the extremely narrow intestine and the lack of through current or strong muscles by which
solid waste could be evacuated from the blind digestive sac. It also agrees with the fact that no
' structure ' was found in the solidified mass in the sac, as would be expected if particulate matter were
collected indiscriminately by a filter mechanism or indeed if Nebaliopsis fed on anything but liquid
or semi-liquid food. There is nothing in the structure of the mouthparts or gut to suggest that it is
a blood sucker, and the appearance of the food undoubtedly suggests coagulated yolk.
In the depths from which Nebaliopsis has been collected there can be very little finely divided
material for a filter feeder — only the slow rain of dead plankton from the surface layers. An animal
of the size of Nebaliopsis will require a considerable amount of food. The second theory would supply
this better than the first. As has been shown, the structure of the gut and especially the presence of
the large digestive sac also indicate that an occasional large meal is taken. It is possible that the
animal depends chiefly on sucking eggs, but has a filter mechanism which provides a small additional
supply of food, alone insufficient, but valuable when prey is scarce.
Without intermediate forms it is impossible to tell how this complex and highly specialized mechanism
originated. It is undoubtedly, however, well adapted to the environment in which the species now lives.
CONCLUSIONS
The structure of the gut diff'ers considerably in the different members of the Nebaliacea, and many
of the changes may be correlated with the feeding habits.
The greatest similarities are found in the musculature. It is obvious that for the efficient working
of a complicated chitinized apparatus simple peristalsis of circular muscles is insufficient. Opposing
dilators are necessary. The oesophagus of Crustacea almost invariably has lateral dilator muscles
i6 DISCOVERY REPORTS
and others are associated with the teeth and other grinding parts. In the Nebaliacea the dorsal ridge
always forms part of the grinding organ of the gastric mill and at least one dorsal dilator is present
in all species.
Other muscles are developed in association with special parts or functions ; for example, the muscles
{t.p.ni. Fig. 6 B) which move the trilobed process of NebalieUa back to the vertical position after the
food has passed, and those (v.m. and c.p.m. Fig. 6 A) which cause the elliptical motion of the tip of
the cardio-pyloric valve in Nebalia.
The numerous small muscles which are not attached to particular structures in the gut probably
function in steadying the whole organ in relation to the other parts of the body.
Thus, though the plan of the musculature is simple and constant, the changes may be associated
with the structure of the chitinous parts, and these in turn may be correlated with the habits of the
species concerned.
Nebalia and NebalieUa both live where the bottom deposits are muddy, but observations of the
former, when living, show that it lies most of the time above the mud just beneath or amongst larger
debris of pieces of seaweed, shells and stones. The particles on which it feeds are thus the small ones
in suspension in this zone. NebalieUa, on the other hand, appears to be a true mud dweller. The
specializations of the eyes, rostrum, and antennae are adaptations to burrowing, and the food particles
found amongst the limbs and in the gut indicate that it feeds indiscriminately on the mud. Many of
the particles are too large to stay long in suspension. Therefore it must either allow some to pass
into the carapace chamber as it burrows or kicks up the mud and then filter rapidly before it settles
(as on occasions does Chirocephalus).
In this mud there is a much higher percentage by volume of silica and other inorganic matter than
in the suspension of finer particles taken by Nebalia. This means that the material which NebalieUa
swallows has a lower food value, and there must be more of it. The mechanism, which is already
present in Nebalia, for rapid dealing with much food is elaborated and that for efficient grinding of
a little is reduced. The food is largely retained in the through passage of the gut and not passed into
the diverticula. In this way the indigestible particles are passed on rapidly, and such nutriment as
can be easily extracted by the digestive enzymes is obtained. In morphological association with this,
the openings and lumina of the digestive glands, are reduced and the lumen of the intestine increased,
and in place of the filter allowing the passage of selected finer particles into the glands there is a
mechanism whereby they are almost all excluded and passed straight on down the intestine.
Thus the differences between the structure of the gut of Nebalia and that of NebalieUa may be
definitely associated with the habits of these animals and the food thus made available.
Nebaliopsis, which has so far been found only at great depths in the open ocean, is in very different
surroundings from the bottom-living forms. It is only to be expected that adaptations to these
conditions would cause specializations, such as are found both in the external and in the internal
structures. The form of the gut may be correlated with the difference in food.^
Filterable particles are much scarcer in this zone, and, as has already been shown, the mechanism
whereby much useless material is passed rapidly through the gut has disappeared. A special method
for dealing with an entirely different type of food has been developed. This food is almost certainly
eggs, and in adaptation to the periodic abundance and scarcity of these the large digestive sac has
been developed as a store chamber and the lumen of the intestine has been reduced to insure that
1 Since the above was written my attention has been drawn to a description by T. J. Evans (Q.J. M.S. 1922, lxvi N.S. p. 439)
of Calma glancoides, an Aeolidiomorph Nudibranch which feeds exclusively on ' the eggs and embryos of the smaller shore
fishes'. The amazing similarity between the adaptations of this mollusc to an egg diet and the specialized structure of
Nebaliopsis forms additional evidence that the latter also feeds on eggs. This is a remarkable case of parallel adaptive evolution
in two animals widely separated in phylogeny, habits and habitats and it is hoped to elaborate the comparison elsewhere.
THE GUT OF NEBALIACEA 17
nothing escapes thorough digestion. The fore-gut is adapted to the puncturing and sucking of the
eggs and the mandibles to holding them in position during these processes.
Interesting parallels to the development of a large storage chamber when an occasional meal is
taken are to be seen in the Decapoda and in the Anaspidacea. In the former group there is a swelling
of the anterior region of the cardiac portion of the stomach in all the predatory forms examined, while
in Porcellana, which has been shown by Nicol (1932) to be a filter feeder, there is no such swelling.
Similarly in the Anaspidacea, Koonunga cursoria, which has been shown by Cannon and Manton
(1929) to have 'given up fiher feeding completely', has a long tubular storage section of the fore-gut
which is absent in Anaspides and Paranaspides, which are filter-feeding forms.
In the above three examples the same result has been attained by entirely different means.
The gut is in more direct contact with the environment than any other internal organ and is thus
more subject to the same influences as act upon the external features. The type of food available not
only influences the method of capture and the mouth parts, but also the structures which have to
deal with it later on. An attempt has here been made to show how the digestive mechanism of the
Nebaliacea may be correlated with the habits and habitats of these animals as far as can be deduced
from present knowledge of this rare group in which so many evolutionary links are missing.
The greater part of this work was carried out in the Zoology Department of the University of
Manchester while holding the Grisedale Research Scholarship. I wish to thank Professor Graham
Cannon and Dr S. M. Manton for the loan of fixed material, Mr G. A. Steven for the living specimens
of Nebalia, specially collected along with characteristic elements of their habitat, and all three for
much helpful advice and criticism.
BIBLIOGRAPHY
Cannon, H. G., 1927. On the feeding mechanism 0/ Nebalia bipes. Trans. R. Soc. Edinburgh, lv, pp. 355-70.
Cannon, H. G., 1931. Nebaliacea. Discovery Reports, in, pp. 199-222.
Cannon, H. G. and Manton, S. M., 1929. On the feeding mechanism of the Syncarid Crustacea. Trans. R. Soc. Edinburgh,
LVi, pp. 175-89.
Claus, C, 1889. Organismus der Nebaliiden und Systematische Stellung der Leptostraken. Arb. Zool. Inst. Univ. Wien,
viii, pp. 1-149, pis. 1-15.
Jordan, H., 1909. Die Pylogenese der Filtervorrichtungen in Pylorttsmagen der Malacosiraca. Verb. d. Zool. Ges., Leipzig,
19. PP- 255-66.
Jordan, H., 1912. Der Magen der hoheren Krebse. Naturw. Wschr. xi.
Nicol, E. A. T., 1932. The feeding habits of the Galatheidea. J. Mar. Biol. Ass. U.K. 1932, pp. 87-106.
Ohlin, 1901. Arctic Crustacea collected during the Swedish Arctic E.xpeditions 1898 and 1899. Bihang Svenska Acad, xxvi, 4, 12.
Thiele, J., 1904. Die Leptostraken. Wiss. Ergebn. d. Tiefsee Expedition 'Valdivia', vii, pp. 1-26, pis. 1-4.
Thiele, J., 1905. Ueber die Leptostraken der Deutschen Siidpolar Expedition, 1901-1903. D. Siidpolar Exp. ix (Zool. i),
pp. 61-8, pi. 2.
Yonge, C. M., 1924. Mechanism of feeding, digestion and assimilation in Nephrops norvegicus. J. Exp. Biol, i, pp. 343-89.
[Discovery Reports. Vol. XXIII, pp. 19-36, March 1945]
ON A SPECIMEN OF THE SOUTHERN BOTTLE-
NOSED WHALE, HYPEROODON PLANIFRONS
By
F. C. ERASER, D.Sc.
CONTENTS
Introduction page 21
Lateral view of skull 21
Dorsal view of skull ......•• 24
Ventral view of skull ......•• 25
Mandible 26
Teeth 26
Vestigial teeth 27
Vertebrae -27
Chevron bones 3^
Ribs 32
Sternum ......•••• 33
Scapula ......••••• 34
Hyoids 34
Appendix 34
Acknowledgments 3^
References .....•■••• 3°
ON A SPECIMEN OF THE SOUTHERN BOTTLE-
NOSED WHALE, HYPEROODON PLANIFRONS
By F. C. Fraser, D.Sc.
Department of Zoology, British Museum (Natural History)
(Text-figs, i-ii)
INTRODUCTION
THE specimens of Hyperoodon planifrons, the Southern Bottlenosed Whale, of which there are
published accounts, are few enough in number to be detailed. The type of the species in the
British Museum collection is an imperfect, partly waterworn skull (Reg. no. 1814A) from Lewis
Island, Dampier Archipelago, North- Western Austraha, described and figured by Flower in the
Proceedmgs of the Zoological Society (1882). In the Anales del Museo de la Plata (1895), F. P. Moreno
gives a brief account of three specimens :
(i) Skeleton of an adult from the coast of the province of Buenos Aires.
(2) Skull of an adult, Chubut Territory, Patagonia.
(3) Skeleton of a young animal, Santa Cruz Bay, Patagonia.
Finally, the Records of the South Australian Museum, vol. iv, no. 3, 1931, contains an account by
H. M. Hale of a male which stranded near Port Victoria, Yorke Peninsula, South Australia.
The present paper is concerned with the description of a skeleton from South Georgia, presented
to the British Museum (N.H.) by the Discovery Committee, with an appended note about two
additional specimens, no part of which has been preserved, from South Georgia and the South
Orkneys respectively, in the Falkland Islands Dependencies.
The widely separated regions from which the Southern Bottlenose has been recorded indicate the
great area of distribution of this species. It may be presumed that its range includes the Southern
Ocean generally and extends into the warmer parts of adjacent seas in the southern hemisphere.
The Discovery skeleton (Reg. no. 1934.7.23 .3) belonged to an animal 6-5 m. long, a female, which
was presented to the Discovery Committee by Capt. Sorlle, Westfold Whaling Co., Stromness, South
Georgia.
The skull and axial skeleton are in very good condition and almost complete, only the slender
zygomatic arches in the skull, one or two of the terminal bones in the caudal series of vertebrae and
probably one chevron being lacking. The appendicular portions of the skeleton are missing except
the scapulae which are damaged.
The sutures of the skulls of the Discovery specimen are all well defined and the epiphyses throughout
the length of the vertebral column are not fused to the centra. In the South Australian specimen,
which was only 0-4 m. larger. Hale states that the sutures of the skull are more or less ankylosed, and
the figured vertebrae show no trace of separate epiphyses. These features suggest that, unlike the
northern H. rostratus, in which the physically mature female is appreciably smaller than the male,
in H. planifrons the two sexes must be about the same size when fully grown.
Recorded dimensions of skulls of //. planifrons, together with the dimensions of a skull of H. rostratus
for comparison, are given in Table i.
LATERAL VIEW OF SKULL (Fig. i)
In the description of the type specimen Flower (1882) drew attention to two features distinguishing
H. planifrons from H. rostratus, both of which are most obvious in the lateral view of the skull. The
first, the character which gives H. planifrons its specific name, is the relatively low development of
a
0^
w 0 O^O^O^r^J^r';0pC^
in Tl-
in ON
►-t M
t^ -1 00 NO
II
ON " iri K 0^ f^ob 00 " CO
t^ i/-i
ON ON
N N
ON N 'i- fO
ON m r^, ^-^hNroO
M »-
N t^
00 1-
M 00 N i-i
0 «
1^
>j
fe is
e
f^tn m oou-ju-iMOvMroC^
•^ in
rn in
0 "H
m 1 ^ " N
ro 0 ■-■ r^OvOrOt^i-iN«
m 0
ON NO
On NO
0 ON moo
E
roN -^ wivO^roK-lOO"
N N
m 0
0 "
N ' 0 en i-i
•^
•■^
" <o
■^
s
'•-0
OOn -^ ^J-f^O^wTw
1 A
Z) NO m
0 in Ht N l^
^ §
0^
1 1
?
f^ N in CO i^
ON ro -^ ro vO ^
N
t^ "
1-1 ■-■ 00 N M
^-f '^^
M •— <
00
«
§"2
s<
^ tS
^
1
0
g
3
6
"O lo o^iiOir^io
1 1
1 •"•
in in 0
t^ T)- -1- N 00
<
O" 00 O^OOuiON
1
'^
« NO «
m r^co ON i-i
6
m Tj- Tj- vo ' ' m 1 00 0
' ¥•*
« 0 N
N « « rn N
^J
H. # N
t~*
>-<
M
^
^
Vi
"■^
II 1 1 1 1 1 1 1 1 1
1 1
1 1
1 1
1 1 1 1 1
3
C
0^
II 1 1 1 1 1 1 1 1 1
1 1
1 1
i 1
1 1 1 1 1
•4->
Q. £
rt
}_i
03
e
E
0
E
1 iri ir^ 00 1 1 1 N m 1 0
CO 0
1 1
1 0
N 00 1 1 1
*-i
l^ t~- vO ro " 0
ON NO
1 1
1 2^
" en
10
00
g
' m m 1 1 1 ^ i/^ 1 «
N
N M ' ' '
ffi"
0
%
*>>
•2
s
&^ =
0
1^ p oo^f^r^^or^oo
OC> " io^66^N"t^"
rn r^
CO rn
0 -t
W ON
N NO
in in 0 N rn
f^ N N nO ih
II
ON r^ ^t/^T)-ro-^"-i"
N
m t^
r-, M
M M 00 N M
"i^
g
HH
CU
' a speci
6
mCO tJ- Oi^'-tr*^^OGO'-t
hH rr^,
0 CO
m 0
ON ONCO N •-'
■g
ov ^ t^ \0 N 0 ONOO c^oo -i-
0 CO
r^ -jt-
NO r--
0 Ti- r^ « f^
rt
6
w ro ro tosO U-) r^ -i-O "-i
" N
m ON
00 "
N ^ ON m M
^
i
\0
0^
II 1 1 II 1 ll-ll
1 1
1 1
1 1
Mill
>
0
-s
s"
s
,0 0 0 0 K-i r^OO 0 1 CO
0 rn
0 1
1 1
1 I 1 1 1
E
s
r^ 0 sO " 0 t^ t^CO r^
r^jNO
00
1 1
III
s
'ro r^ tovDiorOTh'n
« N
en '
3
CO
cl
03
t— 1
0
E
CO
<il
C4
0 7 vO ^ 0 ^0 tJ- 71-00
m ON
9^ 1
m On 0 Tj-
•s
-v^
OCO b^ vOr^r^NrOMON
On fn
^
1 1
nO 4t- in ^
ON N ■'J- lO VO ro Tj-vO M
N
tn
" 00 N "
"a,
1h
E
ON "-> u-iOOrot^t^"-lt^
r^ in
0 1
1 1
0 1 moo NO
0
u^N r^i r^O>-'OOOt^O-<*-
0 t-^
ON
ON t-»CO NO
-a
E
•H ro fO u-isO ^ ro -^^ ►""
■-• N
m '
1 1
- ' ON N "
g
0
2
"
>n
00
^0
0^
On ro t^ c^ r^ 9^^ ?^ 7*" 9^
0 00 ^ t^db ^ " ro i/^ "-1 N
r^- 0
ON ^-
ON
NO
1 1
>
p-i
ON 0 Ti-u->inc<^Ti-ioi-i
N
m
" 00 N "
"So
^
E
r«-iio 0 r^row^NmOOOi^j
NO m
0 1
1 1
m 1 N N "
C4
42
t-H ro t^ vO ON -^ N r^vO »o 0
(^ ON
0
1
N ONOO N
*
E
CON N rort--<i-NrO-<l-M
M
m '
w 1 NO w w
V
"^ B « § i .^
^^
^*-
«
2
0
tf)
1
ill? Sl^l°
"5-
0
-a
0
»*■<»
0
J2
Mi3
s
-ai
CO
iA%i lull
J5 >.
1'i
U
^
3
,
E^
^ s
g^
'0
V 0
M
00 Og BccS=
<- S
'0 h
c .E
-a m ^
2S-al|g= °gl 0 1
^il^S2<^|EEp^
-^ Ej= gj:^^^J=J= f,J= ^
«.S^2-SS^.SfSS-5-S^-SS
0 a
0
_c E
J 1
-a 0
0 tffl t«
OX
.1^
"Oh
;" 4) U S u
THE SOUTHERN BOTTLENOSED WHALE
23
the maxillary crests. These in H. rostratus originate anteriorly approximately equidistantly between
rostrum tip and antorbital notch, and ascend in a slope which varies according to age to a summit
which overtops the skull vertex in all but the most juvenile specimens. Posterior to the summit
there is a decline in level which is generally more abrupt than the anterior slope, the maxillar}' bone
attaining normal thickness again before it ascends posteriorly in contact with the vertical portion of
the frontal. The crests in H. planifrons originate anteriorly about two-thirds from the anterior end of
the distance between rostral tip and antorbital notch. The slope is gradual to a low summit above the
antorbital notch, and the decline posteriorly is equally gradual so that there is no horizontal thin
portion of maxillary before it rises vertically in contact with the frontal in the occipital crest. In the
Discovery specimen the maxillary crests are less massive than in the South Australian specimen.
It may be that the difference is associated with sex, but it may equally v/ell be due to difference in age.
Fig. I. Lateral view of skull and lower jaw. ( x \.)
The second feature referred to by Flower, and visible in the lateral view of the skull, is the much
larger size in H. planifrons of the crest formed by the vertex behind the nares. Not only is it much
larger than in H. rostratus but it differs somewhat in shape, overhanging the narial area considerably,
whereas in H. rostratus the anterior face of the crest viewed in profile is approximately vertical. Other
differences will be mentioned when describing the dorsal aspect.
In skulls of comparable size the distal portion of the rostrum is more slender in H. rostratus than
in H. planifrons. The differences which exist in the proximal portion are associated with the dis-
similarity of the maxillary crests. The distance of the maxillary tip from the tip of the premaxilla is
alike in both species.
The extent to which the lachrymal is seen in the lateral view appears to be equally variable in both
species. The type specimen of H. planifrons has the left lachrymal completely separating the malar
anteriorly from the orbital process of the frontal posteriorly and it has a wide contact with the maxilla.
In the Discovery specimen it hardly appears in lateral view. It does not nearly reach the maxilla, and
24
DISCOVERY REPORTS
the malar and frontal are in contact above it. The H. rostraius specimens examined showed a variety
of form in the lachrymal. In one it did not appear in lateral view, in another there was a ventral
portion separated from a more dorsal portion by a considerable extent where malar and frontal were
in contact, and in a third the lachrymal completely separated malar from frontal and was in contact
dorsally with the maxilla. This variation in the H. rostratus lachrymal is apparently connected neither
with the age nor the sex of the animal.
The temporal fossa in the Discovery specimen, like that of the type, is higher and shorter antero-
posteriorly than that of H. rostratus. Apart from these differences the general form of the skulls is
very similar and confirms the generic affinity of the two species with each other.
DORSAL VIEW OF SKULL (Fig. 2)
The differences between H. rostratus and H. planifrom are again clearly seen in the dorsal view. The
massive prominences over the nares in H. planifrons
extend forward so that the anterior boundary of the right
one is almost vertically above the premaxillary foramen.
The left, smaller in size, does not extend forward quite
so far, but both in this aspect shut out any view of the
narial apertures. The two bones are separated from each
other by a parallel-sided gap bounded by the nasals. Hale
states that in the South Australian specimen ' The inner
anterior edge of each nasal (at the bottom of the groove)
drawn up into a low thin flange'. This is the condition
in the type, but in the Discovery specimen the flanges
are wanting, so that there is no median ridge at the hinder
end of the groove. All the specimens show the internasal
gap diverging to the left from behind forwards and con-
tributing to the asymmetry which characterizes the whole
of this region of the skull. In H. rostratus the narial
prominences do not extend forward so as to shut out
completely the view of the nares ; the right boss, still larger
than the left, has a greater width to length proportion than
in H. planifrons. The internarial groove is wider, and has
divergent, not parallel, sides. The nasal septum is similar
in both species. It is strongly deflected to the left anterior
to a pronounced emargination, and overlays to some extent
the left premaxillary. Its extension forward in the gutter
of the vomer is similar in the Discovery and the type
specimen, in both of which it ends in the region of the
posterior edge of the maxillary foramina about 100 mm.
behind the level of the antorbital notches. The South
Australian specimen has this ossification extending forward
nearly to the antorbital notch level.
In the region of the antorbital tubercle the outline of the skull is consistently different in the two
species. In H. planifrons from the apex of the tubercle the external outline of the antorbital region
extends posteriorly at an obtuse angle with the external edge of orbital process of the frontal, whereas
in H. rostratus it is very nearly a right angle.
Fig. 2. Dorsal view of skull. ( x J.)
THE SOUTHERN BOTTLENOSED WHALE
25
Flower drew attention to the large size of the premaxillar}^' foramina in H. rostratus as compared
with H. plaiiifrom, and this is consistent in all the specimens so far described and figured.
The maxillar)' crests of H. rostratus rise vertically from the external margins of the maxillary
foramina, and the inner faces are nearly parallel to one another. In old males especially, the crests
approximate to such an extent as nearly to touch and thus form an arch over the prenarial portion
of the premaxillae. The medial margins of the maxillary crests of H. planifrons overhang gutter-like
extensions forward of the maxillary foramina. The inner faces diverge from each other at a very wide
angle, and this feature, together with the lesser height of the crests in H. planifrons, provides one of
the most conspicuous diagnostic differences between the two species.
The vomer, which is without mesorostral ossification, is visible between the overarching anterior
portions of the premaxillae. Its anterior tip is nearer the end of the snout in the Discovery specimen
than in the South Australian specimen, the distance being 257 mm. as compared with 380 mm. In
H. planifrons the greatest width of the premaxillae anterior to the foramina is about midway between
the foramina and the premaxilla tip. In H. rostratus the greatest width is at about two-thirds of the
distance from the tip.
VENTRAL VIEW OF SKULL (Fig. 3)
In ventral view such difl^erences as exist between the skull
of H. rostratus and H. planifrons are of detail rather than of
fundamental structure.
The vomer in both species appears as two lenticular areas
in the middle line of the rostrum. The anterior area separates
the premaxillaries posteriorly and the maxillae anteriorly.
There is then a short length where the maxillae are in contact
in the middle line before the vomer appears again, when it
is bounded partly by maxillae and partly by the palatine and
pterygoid bones.
The anterior portion of the vomer appears to be consistently
shorter in H. rostratus than in H. planifrons.
The palatine bones in both species are in two portions,
palatal and lateral, separated by the pterygoid coming into
contact with the maxilla. The palatal portion is bounded by
maxilla, vomer and pterygoid, the lateral part by pterygoid
and maxilla. In H. planifrons the palatal portion is a narrow
strip with a greatest width, in the Discovery specimen, of
less than a centimetre, and a length of about 10 cm. Each
palatal portion in H. rostratus is roughly triangular in outline
and of greater expanse (width about 4 cm. and length 11-5 cm.,
in a specimen of size comparable to H. planifrons). This
diff'erence appears to be constant. The space between the
palatal and lateral portion of the palatine, where the pterygoid
anteriorly comes in contact with the maxilla, is much greater
in H. planifrons than, in proportion to skull length, in any
of the H. rostratus skulls in the British Museum collection.
Incidentally it may be remarked that in Berardius arnuxii
the palatal and lateral portions come into contact, being
Fig. 3. Ventral view of skull.
( X l)
26 DISCOVERY REPORTS
separated from each other only by a suture. The lateral portion of the palatine is smaller in H. planifrofis
than in H. rostrattis, and in general the impression obtained is that in the former species the pterygoid
anteriorly has expanded at the expense of the bones adjacent to it.
The pterygoids are of typical ziphoid form in both species, ' large, solid, backwardly produced,
meeting in the middle line, not involuted but simply hollowed on the outer surface' (Flower, 1871).
The zygomatic process of the malar has its origin much nearer the posterior border of the bone in
H. planifroiis than in H. rostratm, in which species it originates only a little way behind the antorbitai
notch. Differences in the anterior margin of the antorbitai region involve the malar bone and were
referred to in the description of the dorsal view of the skull.
The lachrymal, a distinct bone, has the same essential form in both species. The extent to which
it appears on the lateral border of the skull has already been referred to. It is long and narrow,
extending obliquely backwards from the external margin of the skull to the infra-orbital foramen.
It is bounded anteriorly by the malar and maxilla and posteriorly by the orbital process of the frontal.
The external margin of the orbital process of the South Australian specimen is more pronouncedly
concave than that of the Discovery specimen.
No marked differences are discernible in the squamosals either between the South Australian and
the Discovery specimens or between either of these and H. rostrattis.
The tympanic-periotic bones are very similar in H. rostrattis and H. planifrons, and as in the
former species so in the latter they are secured to the skull anteriorly by a slender inward-curving
process from the squamosal embracing the periotic, and posteriorly by a rugose wedge-shaped
extension from the tympanic between the squamosal and basi-occipital.
In the posterior view of the skull all the available specimens of H. planifrons show the characters
to which Flower drew attention in his description of the type, namely, the narrowness and greater
height compared with H. rostratus and also the inferior size of the occipital condyles in the southern
species.
MANDIBLE (Fig. i)
The jaws of H. planifrons compare closely with those of H. rostratus in general shape and in the
extent of the symphysial region. The two rami of the mandible are not ankylosed at the symphysis
in the Discovery specimen, whereas in the South Australian specimen Hale describes fusion as pro-
ceeding, the two rami being linked by ossified bridges. In the former specimen the tooth alveolus at
the tip of each ramus is continuous posteriorly with the dental groove, gradually merging into it.
In the South Australian specimen (Hale's Fig. 4) the alveolus appears to be sharply defined from the
dental groove.
These differences between the jaws of the two specimens are such as might be expected from their
difference in age.
TEETH (Fig. 4)
The outlines of the teeth of the Discovery specimen and of the South Australian specimen show the
main differences between the two. Those of the former are conical and slender, and have a widely
open pulp cavity. The dimensions are as follows :
Right Left
(i) Length 50 mm. 50 mm.
(2) Greatest diameter 18 mm. 18 mm.
(3) Diameter at right angles to (2) 17 mm. 16 mm.
THE SOUTHERN BOTTLENOSED WHALE 27
The greatest diameter is just a little distance above the lower edge of the tooth, which has this
indication of incipient closing of the pulp cavity. The tip of each tooth, an unworn crown of about
8 mm. length, projects from a thin investing coat of cement.
The South Australian specimen has much more massive, fusiform teeth. Their length is comparable
to that of the Discovery specimen — 57 and 59 mm. — but the greatest diameter is double. Apart from
the difference in the pulp cavity (the root is entirely closed in the South Australian specimen) which
is due to age, it is considered that the dissimilarity is associated with sex, and that in this as in other
ziphoid whales the teeth of the male are large, massive and projecting above the gum, whilst those
of the female are more slender, and, since the crowns are unworn, presumably concealed by the gum.
Fig. 4. Teeth of H. planifrons.
Upper pair, 9, Discovery
specimen ; lower pair, S, South
Australian specimen. ( x \.)
Fig. 5. Anterior view of atlas.
(xi)
VESTIGIAL TEETH IN THE UPPER JAW
When the Discovery specimen was received the skin and dried flesh on the ventral surface of the
rostrum were still attached, and on each side of the upper jaw was a row of teeth commencing at
about 24 cm. from the jaw tip and extending along the jaw about 16 cm. The teeth were spaced roughly
equidistantly about 8 mm. from each other. All of the teeth were not in situ ; some had either been
absorbed or had dropped out, but evidence of their existence was indicated by the fibrous follicles
in which they had rested. It was estimated that each row consisted of twenty teeth, but the difficulty
of dissection made exact computation impossible. Sixteen teeth were recovered on each side ; most of
them projected 2-3 mm. from the dried gum, but whether this post-mortem conspicuousness existed
in the living animal is doubtful. Their shape is fusiform and they are slightly to moderately curved.
A basal portion consisting of cement envelopes the dentine of the crown to a greater or lesser extent,
in some the junction between cement and dentine being clearly defined. The root portion of some of
the teeth is drawn out into a needle-like extension. This is considered to be due to absorption in
process, and in the shorter teeth, in which the extension has disappeared, it is presumed that the
process has gone still further. The length of the teeth ranges from 4 to 14 mm. with diameter up
to 2 mm.
VERTEBRAE (Figs. 5, 6)
Vertebral formula. Cervical 7, dorsal 8, lumbar 11, caudal 17 + .
Cervical vertebrae. The Discovery specimen, like the South Australian and H. rostratus, has all
seven centra fused together. The posterior epiphysis of the seventh is still distinct. In correspondence
with the superior size of the occipital condyles in H. rostratus the anterior articular surface of the
28
DISCOVERY REPORTS
I
THE SOUTHERN BOTTLENOSED WHALE
29
X
60
30 DISCOVERY REPORTS
atlas is also larger than that of//, planifrons. Otherwise the cervical mass is much alike in both species.
Such differences as exist between the South Australian and the Discovery specimens may be regarded
as coming within the range of individual variation. The former has the lateral process of the atlas
fused with the inferior lateral process of the axis, whereas in the Discovery specimen the inferior
lateral process of the axis is distinct. Both specimens show a short rugose superior lateral process
on the axis, the South Australian specimen having ' an incomplete foramen on the right and complete
foramen on the left between it and the inferior lateral process ', whilst the Discovery specimen has
this arrangement of foramina transposed. The superior lateral processes of the third to sixth
vertebrae are separate and of diminishing size antero-posteriorly in the Discovery specimen.
The South Australian specimen has the third ankylosed on the left with that of the preceding
cervical.
The neural arch of the sixth is not completely fused with the arches anterior to it, and fusion is
less on the left than on the right side. The corresponding arch in the South Australian specimen
appears to be completely fused. There is a strong forward-projecting inferior lateral process on the
sbcth vertebra of the Discovery specimen. Hale (1931) does not mention its presence in the South
Australian specimen, and his figure shows that the inferior lateral process of the seventh is of con-
siderable size and prominence and similar to that of the specimen of H. rostratus used for comparison
with the Discovery H. plmiifrons. The inferior lateral process of the seventh in the Discovery
H. planifrons is small and inconspicuous. Between it and the superior process is the articular facet
for the head of the first rib. The neural arch is free except at the tip, whereas the South Australian
specimen has the ' greater part of right side of neural arch free including apex which does not meet
the opposite member of the arch '.
Thoracic vertebrae. The Discovery specimen has eight pairs of ribs and therefore eight thoracic
vertebrae. As the South Australian animal had nine pairs of ribs the possibility was considered of the
ninth pair in the Discovery specimen having been overlooked. However, this is discounted to some
extent by the fact that in the La Plata examples eight, not nine, is the number recorded. The reduction
to this number represents the extreme reached in any of the Mammalia.
The series of thoracic vertebrae in the Discovery Bottlenose commences with one having a slender
neural spine, wide neural arch, widely separated zygapophyses, and short metapophyses at the
proximal ends of transverse processes, which last are directed downwards and forwards and bear a
facet for the tuberculum of the rib. There is a short centrum bearing a postero-lateral facet for the
capitulum of the second rib. Proceeding tailwards the neural spines increase in length and width,
the neural arches diminish in size, and the zygapophyses are very much reduced. The metapophyses,
from being stout and short, are, in the eighth thoracic laminar, almost semicircular in outline and
projecting from the anterior edge of the neural arch. The centrum at the end of the series is about
double the length of that of the first thoracic.
The arrangement of the articular facets for the ribs is interesting, and it is unfortunate that the
centra of the vertebrae were damaged by the harpoon which killed the animal just at the point where
detailed description is most required. However, enough remains to make some sort of interpretation
possible. As far back as the fifth thoracic vertebra the articular facets are conspicuous on the postero-
lateral edges of the centra. In the sixth vertebra the surface of the centrum on the left side has been
obliterated, but the right side which is entire has only the very slightest indication of a facet, whilst
the seventh vertebra has a distinct antero-laterally placed facet. It would appear therefore that as
far back as the fifth vertebra the capitular articulation is with the rib of the succeeding vertebra, that
the sixth is transitional between this arrangement and one in which the capitulum of the rib articulates
with the centrum of the same vertebra with which the tuberculum is associated, and that in the seventh
THE SOUTHERN BOTTLENOSED WHALE 31
this process is almost complete, with capitulum and tubercle of the seventh rib having articulation
almost completely restricted to the seventh thoracic vertebra.
The change in position of the transverse process from the side of the neural arch (upper transverse
process of Flower, Osteology, 1870, p. 60) to the side of the centrum (lower transverse process of
Flower, op. cit.) takes place in the eighth vertebra. There is not in the Discovery specimen as in the
South Australian specimen a vertebra showing the transition from the one to the other kind of
transverse process. The H. rostratus specimen used for comparison with H. planifrons showed in the
eighth vertebra a condition intermediate between that of the other two specimens. In it the upper
transverse process is in the form of a small knob-like and quite vestigial process on the lower margin
of the metapophysis.
Going tailwards the ventral surface of the centrum shows increasing development of the median
ridge which is in the form of a well-defined keel on th. VIII.
Lumbar vertebrae. There are eleven vertebrae in the lumbar series of the Discovery H. planifrons.
The South Australian specimen has one less, but this discrepancy may be accounted for by the
greater number of thoracic vertebrae in the latter specimen.
The neural spines increase in length to about the middle of the series and then diminish gradually,
so that a line joining their extremities makes a very shallow arcr There is an increasing inclination
backwards of the spines going tailwards, a widening of the spine as a whole and of the distal end as
well in the more posteriorly situated elements. The metapophyses are laminar, have rounded margins,
and show increasing approximation to each other. The neural canal diminishes in size ; the centrum
increases so that at the end of the series it is about i J times the length of the first lumbar ; the diameter
also is increased. The transverse processes are directed obliquely forward, flattened, beginning to
diminish in length, and get wider at the tail end of the series. The first lumbar transverse process is
somewhat different from those that succeed it, being disproportionately broad and rather stouter.
The hypophysial ridge is of increasing definition to about the middle of the series, whence it
diminishes in prominence ; and in the last lumbar it is a low, flattened, inconspicuous keel.
No obvious differences distinguish the vertebrae in this region from those of H. rostratus.
Caudal vertebrae. The caudal series of vertebrae is incomplete in the Discovery specimen. Seventeen
remain and the missing elements are at the posterior end. The South Australian specimen has
20 caudals.
The neural spines diminish tailwards and disappear after the tenth caudal. In lateral view they
are broad distally with a slight narrowing towards the neural canal. There is a corresponding diminution
of metapophyses which anteriorly in the series are laminar with rounded border, and posteriorly are
rather stout short tubercles which finally disappear. The neural canal continues the diminution in
size observed in the lumbar series.
Anteriorly the centra have the massiveness which characterizes the more posteriorly placed lumbars
and, going tailwards, although length diminishes gradually, the decrease in transverse diameter is
not noticeable until near the end of the column where the diminution becomes more marked and the
vertebrae adopt a subcuboid shape unlike the cylindrical form of the more anterior elements.
The transverse processes disappear as distinct prominences after the seventh caudal. While still
distinguishable they maintain the obliquely forward direction noted in the lumbar vertebrae. The
perforation of the transverse process of the seventh, noted by Dale, is represented in the Discovery
specimen by a pronounced emargination of the outer edge of the process on each side near its posterior
end. This is visible, although much less obvious, on the transverse processes of two vertebrae im-
mediately anterior to the seventh caudal.
On the lower surface of the centrum anteriorly and posteriorly are the paired facets for the chevron
32
DISCOVERY REPORTS
bones. Two longitudinal ridges with concave margin join the anterior to the posterior facets. The
concavity is ill defined at the anterior end of the series, and is correlated with the lesser prominence
of the facets themselves ; but going tailwards with the greater development of the articular surfaces
and the shortening of the length of the mass of the centrum, the emargination becomes increasingly
pronounced until on the ninth (in both the South Australian and the Discovery specimens) a foramen
is enclosed.
CHEVRON BONES (Fig. 7)
The nine chevron bones figured are an incomplete series; at least one is considered to be wanting.
However, those remaining give an adequate idea of the form these bones assume in H. planifrons.
Fig. 7. Chevrons. ( x \.)
Only one side of the first chevron is present, a slender lamina of bone which has no evidence of
having been fused to the element of the other side. The second chevron, a single bone, has a broad,
short, spinous process with obliquely rounded ventral margin. The third has the spinous process
greatly elongated with rounded antero-ventral margin, and with hinder and ventral margins meeting
at roughly a right angle. From the third tailwards there is a progressive diminution in the spinous
process length and a reduction in size of the bone as a whole, in the last of the series the spinous
process being only about one-half as long as it is wide. The chevrons show no distinctive difference
from those of the South Australian specimen or of H. rostratus.
RIBS (Fig. 8)
The Discovery H. platiifrons has eight ribs on each side, in this number agreeing with the La Plata
Museum specimens. The South Australian specimen has nine pairs of ribs, the ninth pair being
small, asymmetrical and obviously vestigial. H. rostratus normally has nine pairs of ribs also, but at
least one specimen in the British Museum collection has only eight pairs.
In the Discovery specimen the first pair of ribs is short, broad, flattened and with sternal end
directed at a slight angle forward from the remainder of the shaft of the bone. The second rib is
moderately broad, more elongated than the first and without forward trend of the distal end. The
third to the sixth are similar to each other, long, slender and subequal in length. In the seventh,
shortening of the shaft has become pronounced, but otherwise the essential features of the four
preceding ribs are maintained. The eighth is still shorter, and in the absence of a capitular portion is
distinguished from all the ribs that precede it.
The first seven ribs have the capitulum defined to a greater or lesser degree. In the first the
capitulum and tubercle are almost confluent, in the following five the capitulum is situated at some
distance from the tubercle. In the seventh the tubercle and capitulum approximate again and the
eighth, as just stated, has no capitulum.
THE SOUTHERN BOTTLENOSED WHALE
33
STERNUM (Fig. 9)
The sternum consists of three elements, the largest of which is the manubrium. The manubrium
is roughly rectangular in outline. The anterior emargination is semicircular and not so pronounced
as in the South Australian specimen. There is a small posterior notch, and the bone extends tailwards
on the right side of this to a greater extent than on the left. Asymmetry is also displayed on the
lateral margins. The facets for the first pair of sternal ribs are equally prominent, but whereas the
Fig. 9. Sternum. ( x |.)
Fig. 8. Ribs. ( x i.)
right side bears a facet a little way posteriorly to the first there is no corresponding one on the left.
The external surface of the bone is convex and the internal concave.
Anteriorly, the second sternal element has a median notch, on the right side of which the anterior
margin is a little way behind that on the left side. This asymmetry is repeated on the posterior margin,
in which, however, the notch is wanting. The lateral margins are shallowly concave, and at the
antero- and postero-lateral corners are facets for the appropriate sternal ribs.
34
DISCOVERY REPORTS
The last sternal bone has again an uneven anterior border, the left side being in advance of the
right. It is without anterior notch. The posterior margin has a deep, angular notch extending nearly
to the middle of the bone ; in the South Australian specimen it is wide and shallow. There are three
facets on each side for sternal ribs, one at each antero- and postero-lateral corner and one midway
between these.
The ventral surface of the bone is raised into a low, ill-defined tubercle.
SCAPULA (Fig. lo)
Both the scapulae of the Discovery specimen are damaged posteriorly. Anteriorly the evenly convex
dorsal margin meets the straight anterior margin at almost a right angle, not being broadly rounded
as in the South Australian specimen. The acromion, as in the latter specimen, is bent upwards and
inwards, the superior and inferior margins being parallel to each other and the distal margin rounded.
It is shorter than in the South Australian specimen. The coracoid is without the distal expansion
noted in the South Australian' specimen, but is otherwise similar in position and shape.
Fig. 10. Scapula. ( x ^.)
Fig. II. Hyoids. ( x ^.)
HYOIDS (Fig. II)
The thyro-hyals are not fused to the basi-hyal. The basi-hyal has a short, straight anterior margin
and deeply concave posterior margin. The lateral portions of the bone which are convex are rugose,
and are completely occupied by the facets for connexion with the thyro-hyals.
The thyro-hyals are wing-like in shape, and stoutest at their proximal ends where there is a broad
area for attachment to the basi-hyal. The bones diminish in thickness from the anterior to the posterior
border, where the upper and lower surfaces meet in a ridge at a very acute angle. The distal tips of
the thyro-hyals are truncated and rugose.
The tympano-hyals are elongate, flattened and tapering at each end to a truncated rugose tip.
The thickness of the bone diminishes from the front to the hinder margin, which last has a fairly
acute edge.
APPENDIX
(i) A male specimen of H. planifrons was measured and examined by Dr L. Harrison Matthews,
at Leith Harbour, South Georgia, on 3 January 1927. It was intended that the skeleton should be
preserved, but before it could be despatched to England an avalanche, which obliterated part of the
whaling station, buried the specimen, and it was not recovered.
THE SOUTHERN BOTTLENOSED WHALE 35
The external measurements recorded by Dr Matthews are as follows :
m.
Total length, tip of snout to notch of flukes 4-63
Projection of lower jaw beyond tip of snout Nil
Tip of snout to blowhole 0-74
Tip of snout to angle of gape 0-85
Tip of snout to centre of eye 076
Tip of snout to tip of flipper 1-72
Notch of flukes to posterior emargination of dorsal fin i -27
Width of flukes at insertion 0-39
Notch of flukes to centre of anus i'33
Notch of flukes to umbilicus 2-46
Centre of anus to centre of reproductive aperture 0-38
Vertical height of dorsal fin 0-25
Length of base of dorsal fin 0-37
Axilla to tip of flipper 0-42
Anterior end of lower border to tip of flipper 0-51
Length of flipper along curve of lower border 0-55
Greatest width of flipper 0-17
Length of severed head from condyle to tip 0-697
Greatest width of skull 0-369
The following notes were also made :
Colour Black dorsally shading to grey ventrally
External genitalia Normal
External parasites None
Hair None
Ventral grooves Two grooves on the throat, one on each side situated under the ramus of
the mandible, 22 cm. in length
Blubber 5 cm. thick on the side below the dorsal fin
Palate Grey
Tongue Flesh-pink
Food Stomach contained a few crystalline lenses from the eyes of cephalopods
Internal parasites None seen
Mammary slits Each 4 cm. in length, situated 12 cm. anterior to the anus
(2) Mr A. G. Bennett, at one time naturalist to the Government of the Falkland Islands, has
provided another record of the occurrence of H. planifrons. He obtained photographs of a specimen
killed in the vicinity of the South Orkney Islands in January 191 5.
One of the photographs, in which the carcass is floating in the water alongside the factory ship,
shows the surface of the skin scored by numerous irregular marks. Similar streaks have been noted
in other ziphoids and are presumed to be the teeth marks of other individuals of the same species.
In addition to these elongated scratches one or two oval marks can be seen. They are reminiscent
of the scars described and figured by Mackintosh and Wheeler (1929) as occurring in various members
of the whalebone whales. Other features which can be observed in the photograph are the pronounced
' forehead ' which rises at almost a right angle from the well-defined beak ; and the right flipper which
is of typical ziphoid form, having a very shallowly convex lower border and slightly more convex
upper edge.
A second photograph gives a ventro-lateral view of the anterior portion of the body, lying on the
deck of the whaling vessel. The region of the mouth and throat, as far back as the two ventral grooves
is of a much lighter colour than adjacent portions of the body. The ' forehead ' appears to be quite
darkly pigmented. The rostrum is stout and well defined and the upper and lower lips meet in a line
which anteriorly is horizontal but farther back swings obliquely upwards.
36 DISCOVERY REPORTS
ACKNOWLEDGMENTS
I have to thank Dr L. H. Matthews and Mr A. G. Bennett for the information and assistance they
have given me. The figures illustrating the paper are the work of Col. M. St L. Simon, and it is with
pleasure that I acknowledge my indebtedness to him, and to my colleague W. H. T. Tams, Esq.,
who took the photographs from which the figures of the axial skeleton were executed by Col. Simon.
I have also to thank Mr E. J. Manly who has helped me with the compilation of the report.
REFERENCES
Flower, Sir Wm., 1871. On the recent ziphoid whales, with a description of the skeleton of Berardius arnouxi. Trans. Zool.
Soc. London, vol. viii, part in.
Flower, Sir Wm., 1882. On the cranium of a new species of Hypeioodon from the Australian Seas. Proc. Zool. Soc. London.
Flower, Sir Wm., 1885. An Introduction to the Osteology of Mammalia, 3rd ed.
Hale, H. M., 1931- Beaked whales — Hyperoodon planifrons a7td Mesoplodon layardii — from South Australia. Records of
the South Australian Museum, vol. iv, no. 3.
Mackintosh, N. A. and Wheeler, J. F. G., 1929. Southern Blue and Fin Whales. Discovery Reports, vol. i, pp. 257-540.
Moreno, F. P., 1895. Nota sobre los Restos de Hyperoodontes conservados en el Museo de la Plata. Anales Mus. de la Plata.
Secc. Zool. III.
[Discovery Reports. Vol. XXIII, pp. 37-102, Jw/zc, 1945]
REPORT ON ROCKS FROM WEST ANTARCTICA
AND THE SCOTIA ARC
By
G. W. TYRRELL, A.R.C.Sc, D.Sc, F.G.S., F.R.S.E.
(Lecturer in Geology, University of Glasgow)
CONTENTS
Foreword, by J. M. Wordie, M.A page 39
I. Petrography of the South Shetland Islands, West Antarctica . . 41
II. Petrography of Rocks from the Graham Land Peninsula and Adelaide
Island, West Antarctica 66
III. Petrography of Rocks from the Elephant and Clarence Group 76
IV. Petrography of Stones dredged from the Vicinity of the Shag Rocks . 89
V. Petrography of the South Sandwich Islands 92
REPORT ON ROCKS FROM WEST ANTARCTICA
AND THE SCOTIA ARC
By G. W. Tyrrell, a.r.c.sc, d.Sc, f.g.s., f.r.s.e.
Lecturer in Geology, University of Glasgow
(With Geological Notes by N. A. Mackintosh, D.Sc, and J. W. S. Mark, M.A., B.Sc.)
(Text-figs. 1-14)
FOREWORD
By J. M. WORDIE, M.A.
In the second volume oi Das Antlitz der Erde published in 1888, and again in more detail in the final
volume in 1909, E. Suess put forward the view ' that the Andes are to be seen again in Graham Land '.
By this dramatic phraseology he implied that the folded mountain border of the Pacific, as exemplified
in the Andes, swings eastward from Tierra del Fuego to South Georgia and then curves back from the
South Sandwich Islands through the South Orkneys to Graham Land and the South Shetlands. Suess
based his views on a memoir by H. Reiter in 1886,^ who there gave substance to an idea put forward
as far back as 1831 by Sir John Barrow."^ Suess characteristically gives the credit for these arguments
to Reiter, whose paper I have not seen, but it is not unlikely that it was Suess himself who suggested
this work; the first volume of the Antlitz had appeared in 1885, and there can be no doubt but that
the ideas of the second volume would already have formed themselves in the author's mind, and this
was a problem which required to be examined. Andersson, in his Geology of Graham Land, in fact
mentions that Reiter had been stimulated by Suess's first volume. In the interval between Suess's
first statement in 1888 and his more detailed advocacy in 1909, Dr Otto Nordenskjold led the Swedish
Antarctic Expedition to the east coast of Graham Land in 1901-3, and J. Gunnar Andersson who was
with him published his important Geology of Graham Land in the Bulletin of the Geological Institute
of Upsala, vol. vii, Upsala, 1906. Nordenskjold himself was also much alive to the problem and has
both described the rocks, Petrographische Untersuchungen aus den Westantarktischen Gebiet, Upsala,
1906, and also put forward an authoritative statement of the whole problem in Handbuch der
Regionalen Geologie: Antarktis, Heidelberg, 1913. Nordenskjold and Andersson carried out in the
field what Reiter had sensed in the study. Andersson, Nordenskjold, and Suess together may
, therefore be regarded as the main advocates of 'two groups of Antilles'. 'South Antilles' was the
name first given to the islands of the southern arc; but more recently the sea enclosed by these
islands has been named the Scotia Sea, and the name South Antillean Arc has now automatically
been replaced by the more appropriate title of Scotia Arc.
Andersson and Suess could base their arguments only on imperfect data, some of which are now
known to be incorrect. Since then many new rock specimens have been obtained and worked on
by qualified geologists. The activities particularly of the Discovery Committee have succeeded in
providing collections surpassing all previous material. Dr Tyrrell has already dealt with some of
the collections in earlier papers on South Georgia, the South Sandwich Islands and the South
Shetlands ; and in the present memoirs he is at last able to make authoritative statements on the remain-
ing portions of the arc either scantily known or completely unexplored at the time when Suess made
his great analysis of the plan of the Earth.
1 H. Reiter, Die Siidpolarfrage imd Hire Bedeutimg fiir die genetisctie Gliedenmg der Erdoherflache, Weimar, 1886.
" Sir John Barrow, Journal of tlie Royal Geographical Society, vol. i (1832), p. 62.
40 DISCOVERY REPORTS
Dr Tyrrell's main conclusions are as follows :
Two dredgings were made from 'Discovery IT in the neighbourhood of the Shag Rocks in
November 1930. Of the nineteen specimens obtained fifteen are described as tremolite-epidote-
greenstone or greenstone-schist. This is an important find, as it can be paralleled both with rocks
from Clarence Island and with specimens from Tierra del Fuego.
Fresh material has been obtained in the South Sandwich Islands both in situ at Saunders Island
and from dredgings elsewhere in the group. These rocks are all volcanic in origin and of Recent age.
The new material, along with earlier collections, shows that the South Sandwich rocks have more in
common with rocks from the Antilles of North America than with any specimens so far known from
the Andes. Dr Tyrrell considers that the South Sandwich Islands probably lie on a ridge parallel
to, but east of, the main Scotia Arc.
Elephant Island and Clarence Island and others east of the main South Shetland Islands not
only lie at some distance from the South Shetlands proper but also differ from them geologically.
A greenstone-greywacke-mudstone association is present, such as is formed in the geosynclinal stage
of a mountain-building cycle and is affected as would be likely by low-grade metamorphism.
Assemblages of this character are found not only in the Elephant and Clarence Group but also in the
South Orkneys. They are paralleled near Ushuaia in Tierra del Fuego, and a somewhat similar
assemblage occurs in South Georgia. Dr Tyrrell considers that these types may also be expected to
form the at present unknown rock basement of Graham Land.
There are extensive collections from the South Shetlands which modify earlier conclusions. The
occurrence of sediments of presumed Mesozoic age on certain of the islands has apparently been
over-emphasized, and one should now regard the South Shetlands as of preponderatingly volcanic
origin, made up either of lavas, mainly andesites, dacites and rhyolites, or of their associated tuffs,
breccias and agglomerates. Plutonic rocks may, however, be commoner than so far supposed. There
were two lava periods, and the intrusive rocks, such as the diorite on King George Island, are regarded
as the underground equivalents of the later period. The Recent volcanoes along Bransfield Strait are
still younger than either of the above lava periods, and it is even probable that Deception Island and
Bridgeman Island have been active in historic times. The chemical characters of the Deception Island
lavas indicate a soda-rich andesite, not readily paralleled in the Andes. Elsewhere the andesites and
basalts are of normal circum-Pacific, that is to say undoubted Andean, type.
Finally, a fifth section deals with some specimens from Graham Land. These are less numerous as
a collection, but they include a quartz-porphyry formation at Adelaide Island of the same nature as
the rocks of a belt 400 km. in length already known from Patagonia.
No new rocks are to hand either from the South Orkneys or from South Georgia. Both localities
are now well known. The importance of the new material lies in the nature of the rocks themselves,
and Dr Tyrrell, in these five papers, has provided petrographic arguments for what was up till now not
more than a matter of inference. The petrographic evidence is more or less complete. To settle the
actual line of the Arc, however, requires that the bottom contours should be better known. Soundings
over a wide area are much to be desired, and will decide whether there is a single arc or a series
of concentric curves. Meantime one can safely say that Suess's, Andersson's and Nordenskjold's
arguments no longer relate merely to a possibility, and that Suess's vision of the Pacific structure
advancing into the Atlantic must now be regarded as firmly established.
41
PART I. PETROGRAPHY OF THE SOUTH SHETLAND ISLANDS
INTRODUCTION
TH I s work is based on two collections of rocks, made during the third and fourth commissions of
the 'Discovery II' in 1934 and 1937 respectively. The specimens were accompanied by excellent
geological and geographical notes, those of 1934 by Dr N. A. Mackintosh, and those of 1937 by
J. W. S. Marr, M.A., B.Sc. Relevant points from these notes have been incorporated, with appropriate
acknowledgement, in the following descriptions.
Bibliography. A full bibliography of the earlier literature relating to the geology and petrography
of the South Shetland Islands (and adjacent lands) is given in my paper listed as (i) below. Only
papers which have been published since 1920 are given in the following list:
(i) G. W. Tyrrell. 'A Contribution to the Petrography of the South Shetland Islands, the Palmer
Archipelago, and the Danco Land Coast, Graham Land, Antarctica.' Travis. Roy. Soc. Edinb. Liii,
pt. I, 1921, pp. 57-79.
(2) H. H. Thomas. 'On the Innes Wilson Collection of Rocks and Minerals from the South
Shetland Islands and Trinity Island.' Ibid. pp. 81-9.
(3) O. Holtedahl. 'The Geology and Physiography of Some Antarctic and Sub-Antarctic Islands.'
Scientific Results of the Norwegian Antarctic Expeditions, 1927-28 and 1928-29, instituted and financed
by Consul Lars Christensen, No. 3, Norske Vidensk.-Akad., Oslo, 1929, 172 pp.
(4) T. W. F. Barth and P. Holmsen. ' Rocks from the Antarctandes and the Southern Antilles
(Being a Description of Rock Samples collected by O. Holtedahl, 1927 28, and a Discussion of their
Mode of Origin).' Ibid. no. 18, 1939, 64 pp.
General. The South Shetlands comprise a group of ten large and small islands extended in a
north-east to south-west direction parallel to, and at a distance of from 60 to 70 miles from, the
coast of the Graham Land peninsula, from which they are separated by Bransfield Strait. From
north-east to south-west the islands are Bridgeman Island, King George Island, Nelson Island,
Roberts Island, Greenwich Island, Livingston Island, Deception Island, Snow Island, Smith Island,
and Low Island. Of these, practically nothing is known of the two last-named. Deception Island,
a sea-flooded Recent crater, is the best known. Bridgeman Island, too, is a Recent volcano and may,
like Deception Island, have been in comparatively recent eruption. Mr Marr's notes make it clear
that Penguin Island, off the eastern horn of King George Bay in King George Island, is also a Recent
volcano comparable with Deception Island and Bridgeman Island.
The rock specimens collected during the recent Discovery II expeditions number in all 141, of
which 81 come from King George Island, 19 from Deception Island, 17 from Roberts Island, 16 from
Livingston Island, 4 from Nelson Island, and 4 from Snow Island.
The plan of the present paper is to describe the collections from each of these islands in turn,
incorporating as much of the geology as can be gleaned from the field notes made by Dr Mackintosh
and Mr Marr. The chemistry of the igneous suite will then be studied with the aid of previously
published and two new analyses, and finally a conspectus of the geology of the South Shetlands will
be attempted from the material now available.
43
PETROGRAPHY
KING GEORGE ISLAND
Admiralty Bay. The Ullmann Range, a ridge trending north and south, projects into Martel Inlet
(north-east arm of Admiralty Bay) and forms the eastern side of Visca Anchorage. Specimens were
collected from the western side of this ridge. In his notes, Dr Mackintosh has given an excellent
sketch of the Ullmann Range as seen from Visca Bay, and has called attention to a prominent dike
which climbs the scarp and culminates in a sharp pinnacle near the central point of the ridge. This
view is undoubtedly the subject of Mr Ferguson's fine photograph (Ferguson, op. cit. pi. iii, fig. 1),^
which clearly shows the dike and a series of lava scarps to the left (north) of it.
The dike consists of a highly porphyritic pyroxene-andesite with phenocrysts of plagioclase (basic
andesine, Ab^j), yellow augite, and chloritic pseudomorphs after orthorhombic pyroxene, in order
of abundance. There are also some large irregular masses of magnetite. The ground-mass is fine-
grained, but apparently holocrystalline, although somewhat altered. It contains a little quartz.
The lavas of which the Ullmann Range is composed are represented by several specimens mainly
collected from screes. Alongside the dike occurs a trachytic lava with a very dense fluxional ground-
mass, consisting of minute feldspar microlites, apparently orthoclase, in a cryptocrystalline base.
There are numerous small phenocrysts of soda-orthoclase and a plagioclase which is now mostly
albite, but the presence of epidote suggests that it may originally have been a more calcic variety.
The rock also carries numerous euhedral crystals of ilmenite rimmed with a leucoxenic alteration
product. Traces of ferromagnesian minerals are present, but are altered beyond recognition. This rock
is notable in containing a few crystals of pale blue pleochroic apatite.
A coarser textured specimen provides further data. The ground-mass is seen to consist of laths of
orthoclase mingled with oligoclase, and contains visible quartz. Still another specimen consists of
an angular breccia of fragments similar to the above. Many of the fragments are rich in quartz. The
shapes of some pseudomorphs outlined in iron ores suggest that the ferromagnesian mineral in these
rocks may have been hornblende.
These lavas may be provisionally classed as dacite or quartz-latite according to the amount of
quartz or orthoclase present. Similar types have been described from Admiralty Bay by the author
((i), p. 71). They also occur in the Fildes Strait area (p. 44).
Near the beach on the western side of the Ullmann Range was collected a lava which maybe described
as an altered quartz-andesite. It contains phenocrysts of plagioclase badly carbonated, and chloritized
pseudomorphs after pyroxenes. Quartz is comparatively abundant, but is partly of secondary origin.
Bluish apatite crystals are abundant, and the lava is therefore regarded as belonging to the same
series as those described above. From the screes to the south of this point a silicified and pyritized
volcanic tuff was collected.
Mr Marr collected three specimens from the western side of the Keller Range along the eastern
shore of Mackellar Inlet. He describes this coast as consisting of slopes of reddish brown tuff with
frequent outcrops of lava which are also prominent at sea level. While two of his specimens are so
highly carbonated and silicified that they can only be described as altered andesites, the third, which
is stated to have come from a fan-shaped columnar outcrop, is less altered, and can be described as
pyroxene-andesite. Feldspar phenocrysts are numerous and, although badly carbonated, can be
identified as plagioclase of composition about k\,An^,. The ferromagnesian constituent consists
of chloritized pseudomorphs after pyroxenes, usually found in crystal clots along with feldspar,
ilmenite, and large crystals of apatite. The ground-mass is dense, brown, and cryptocrystalline, the
only identifiable constituent being feldspar microlites showing straight extinction (.? oligoclase).
1 For full reference see p. 76.
44 DISCOVERY REPORTS
Two specimens were collected by Mr Marr from near Point Thomas, Admiralty Bay. One, from
the coast a little south of the Point, is a fresh hypersthene-andesite. This rock appears to be identical
with the rock called hypersthene-augite-bandaite of the volcanic vent of Three Brothers Hill, Potter's
Cove, Fildes Strait, described by the author ((i), p. 68) from Mr Ferguson's collection, and the
reader is referred to this full description for petrographical details. In fact, Mr Ferguson actually
collected material from the same area ((i), p. 69). The extreme freshness of this rock, as compared
with the extensive alteration suffered by the lavas from the interior of Admiralty Bay, suggests that
it belongs to the later of the two volcanic episodes on the mainland of King George Island.
On the other hand, the rock collected by Mr Marr from the coast of Ezcurra Inlet, one mile west
of Point Thomas, is an altered pyroxene-andesite which clearly belongs to the older series of lavas.
This occurrence suggests that the boundary between the older and newer series of lavas should be
drawn a little farther south than is shown on Mr Ferguson's map (D. Ferguson, op. cit. supra,
fig- 2, p. 38).
Fildes Strait. Fildes Strait separates King George Island from Nelson Island to the west.
Dr Mackintosh collected several specimens from a harbour (St. 1482) near the south end of the
strait, which may be identical with the ' Potter's Cove ' of Mr Ferguson, or it may be the ' Marian
Cove ' of the same author which is a little farther north. Dr Mackintosh describes the rocks as much
weathered, breaking down into screes through which solid rock appears here and there.
Three of the specimens from this locality are dark, very compact rocks of basaltic type. They
consist mainly of a very fine-grained ground-mass of intersertal type with numerous microlites of
a striated feldspar giving extinctions up to 20° (andesine), scattered patches of chlorite and obscure
brownish material probably representing pyroxenes, and particles of haematitized iron ore embedded
in a reddish cr>'ptocrystalline or glassy base. The few small phenocrysts consist of epidotized plagioclase
(originally labradorite), and, in one section, fresh, euhedral, colourless augites of small optic axial
angle (.? pigeonite). A chemical analysis (p. 59) shows that these rocks must be regarded as of
tholeiitic composition.
One specimen from this locality, however, is much more acid than the above, and must be classed
as soda-rhyolite or quartz-keratophyre. It is a whitish felsitic rock much reddened by haematitic
staining. In thin section it is seen to consist of a dense quartzo-feldspathic ground-mass with an
obscure hint of spherulitic structure, which carries numerous large phenocrysts of turbid albite and
haematitized biotite.
A single specimen was collected from another locality on Fildes Strait near the narrow northern
entrance (St. 1483). Dr Mackintosh states that the rock formation here appeared to be quite different
from that of St. 1482, an observation which is confirmed by examination of the specimen. One
adjacent islet consisted of a dome-shaped mass of rock, ' probably basalt ', with a pronounced columnar
structure, but the outcrop from which this specimen was collected was not columnar.
This rock turns out to be a feldspathic olivine-basah or olivine-andesite. Large phenocrysts of
fresh basic labradorite (Ab^Ang) are very abundant. Calcified and serpentinized olivines are numerous,
but a fresh pale augite is quite subordinate in amount. These are embedded in an intergranular
ground-mass consisting of plagioclase laths, augite and iron-ore granules, and a dark crypto-
crystalline base.
North Foreland District. The North Foreland is the tip of a long narrow peninsula springing from
the north-eastern corner of King George Island. A shorter peninsula ending in a steep blulT headland
called Brimstone Peak occurs a mile or two to the west, and the two peninsulas enclose a deep bay.
Still farther west comes the well-known Esther Harbour, which was apparently not entered on this
occasion. This district (St. 1949) was visited by Mr Marr.
SOUTH SHETLAND ISLANDS 45
Mr Marr writes that ' the cUffs forming the west side of the Foreland ... are composed of a massive
grey rock much traversed by cracks and joints, giving it a very shattered appearance '. This is borne
out by the three specimens collected here, which are all parts of a plutonic igneous rock of variable
grain size. This may be described as quartz-hornblende-pyroxene-diorite, and represents a ver)'
abundant type in West Antarctica ((i), p. 6i). Its three principal minerals are plagioclase (core
andcsine; outer shell oligoclase) ; pale green hornblende, sometimes with a pale brown tint; colourless
diopsidic pyroxene which is altering into a pale green amphibole. The accessory minerals are quartz,
filling the interstices between the main constituents; some large flakes of reddish biotite; abundant
ilmenite altering to leucoxene ; and a considerable amount of apatite. The amphibole and pyroxene
tend to form well-shaped crystals, and to enter into clots with biotite and ilmenite. One of the
specimens is a true plutonic type with allotriomorphic texture and comparatively coarse grain. Another
is a fine-grained aplitic type poorer in the mafic minerals, which may be styled quartz-microdiorite ;
and the third is a porphyritic type in which the feldspars, hornblendes and pyroxenes (including both
augite and hypersthene) occur as phenocrysts in a fine-grained granulose ground-mass. A few large
crystals of bluish apatite occur in this rock. This type may represent a chilled marginal phase of
the intrusion.
It is clear that the vicinity of North Foreland is occupied by a large plutonic intrusion of the same
type as occurs at Noel Hill, Marian Cove ((i), p. 6i), and at Le Poing on the west side of Admiralty
Bay ((i), p. 62). This mass may occupy the whole of the eastern side of King George Island, as
Mr Marr states that the cliflFs to the east and south of the Foreland, and probably as far as Cape
Melville, are high and sheer, and seem to consist of the same grey massive rock.
Brimstone Peak is said to be composed of perpendicular 'basalt' cliffs rising sheer out of the sea
to a height of 150 ft. The single specimen obtained from this locality shows, however, that the rock
is a fresh hypersthene-augite-andesite of the Recent type so common elsewhere in King George Island.
The hypersthene is mostly altered to chlorite or bastite, and often forms the core of an augite crystal.
A single crystal of magnetite-rimmed brown hornblende was present in the thin section.
Bolinder Beach (St. 1953) is situated a few miles west of Esther Harbour and Brimstone Peak.
It is described by Dr Ommanney as a bluff peak crowned by three buttresses of dark grey and light
brown rock veined by what, on closer examination, proved to be finely crystalline rose and amber
quartz. All the rock specimens collected here were lost in a boat accident except a few from a 100 ft.
cliff at sea-level on the northern face of the bluff.
This rock proves to be an enstatite-andesite of micro-porphyritic and intersertal texture, consistmg
of very numerous feldspar laths (andesine, AbgAn,), and less abundant pseudomorphs in chlorite
after enstatite (typical square prisms with truncated corners), in a dense, brown, cryptocrystallme to
glassy ground-mass. It probably belongs to the older series of lavas, as it is intersected by mineral
veins which may represent the same group of veins (quartz and pyrites) as that described by Ferguson
from the islands of Esther Harbour {op. cit. supra, p. 41). These veins run nearly east and west, and
might thus probably intersect the region of Bolinder Beach.
Pengum Is/and and Adjacent Mainland. Penguin Island is situated off the eastern horn of King
George Bay. That Penguin Island is a Recent volcano, one of the line of volcanoes fringing Bransfield
Strait, is Mr Marr's important and most interesting discovery. The following is a description of
Penguin Island quoted from Mr Marr's report:
The southern half of Penguin Island is a volcanic cone. The northern half consists of a long, very low plateau,
much of it only about 50 ft. high. The western face of the cone is steep and has a deep brick-red tint. On its south-
eastern and eastern sides the cone slopes down to a plateau roughly 100 ft. high, which is continuous in a wide
sweep with the lower plateau which forms the northern half of the island. On the southern side the cone ends
46 DISCOVERY REPORTS
abruptly in sheer and inaccessible cliffs from 50 to 100 ft. higii which continue round the coast to the eastern side
of the island. The rock is lava, at a distance dark in colour, and much broken with cracks and fissures. . . .The island
is remarkably free of snow and ice, and although snow may lie thinly on it after a heavy fall it does not remain for
long. [This fact strongly suggests that there is still much residual heat in the cone, and that it may only be dormant.]
Penguin Island is a volcanic cone in the shaping of which three, and perhaps four, periods of activity seem to
have been involved. What seems to have been the earliest and biggest eruption is represented now by the concave
section of a very large, but almost entirely cut away crater which occupies nearly the whole of the western face of
the cone, from the shingle beach up to the summit. The degree of concavity is not very high, yet it is unmistakable.
The sides of the interior of this now almost destroyed cone are composed of rather finely divided volcanic clinker
of a rich brick-red colour which gives this side of the island its characteristic tint. The clinker fragments have the
even consistency of a coarse gravel. Projecting out of this eroded crater, its base on a level with the beach, is a huge
plug [? dike] of lava from three to five feet in width and rising vertically like a wall for nearly a hundred feet. Similar
though less conspicuous plugs [dikes] occur elsewhere in this crater.
Main summit crater. A later eruption is perhaps represented by this crater, a third of a mile across and about
200-300 ft. deep, which occupies the summit of the cone. Evidently the rim of this crater has crumbled away
considerably, for it is highest to the north, but slopes downward towards the south (see sketch, Fig. 2). The bottom
is rather damp and shows signs of there having been water lying about. On the east side of the interior of the bowl
a gigantic plug of lava sticks up vertically for about 100 ft., the top, however, not projecting beyond the rim of the
crater. There is some quite deep snow, which is possibly permanent, inside the bowl on its north-east side.
Fig. 2. Penguin Island.
Another eruption, subsequent to that which produced the main summit crater, is represented by the small secondary
cone which rises concentrically from the bottom of the former. The secondary cone is about 100 ft. high and has
a crater less than 80 yards across at the rim, and about 20 ft. deep.
Ash beds. Much of the lower part of the cone, and a large part of the 100 ft. plateau to the south-east and east
of it, seem to be composed of horizontally stratified ash beds of a light colour.
The coastal cliffs throughout are composed of lava often broken by cracks and fissures. On the eastern side of
the island the crests of the cliffs are extremely rugged and often twisted into grotesque sliapes, evidently the result
of cooling in the surface of an ancient lava flow. At the south-west corner of the island a certain warmth was felt
on the lava and inside a fissure. The heat experienced was very slight, but we were of the opinion at the time that
it was unlikely to have been due to absorption from the sun.
Crater on east side. On the east side of the island, some 60-80 yards from the coast, another old crater occurs in
the 100 ft. plateau. Its rim is flush with the general level of the plateau, and it is rather a remarkable sight, strongly
resembling an old quarry. It is a perfect circle and about 150-200 yards across at the rim. The sides are steep,
descending for at least 50 ft. There is deep water at the bottom in which a few penguins were swimming; the water
was not icy cold. On its west side the crater cuts through horizontally stratified, light-coloured beds of volcanic
ash at least 30 ft. in thickness. On the eastern rim of the crater there is much glassy lava, obsidian, of various hues.
All specimens of the lavas collected from the volcanic cone of Penguin Island represent textural
variants of a typical olivine-basalt. The most fully crystallized type comes from the plug in the
summit crater. In thin section it is found to be highly porphyritic with numerous phenocrysts of
fresh olivine and pale brown augite, sometimes aggregated into clots, and very numerous micro-
SOUTH SHETLAND ISLANDS 47
phenocrysts of plagioclase (Ab55An4g) with both chemical and mechanical zoning, embedded in an
intergranular ground-mass consisting of feldspar microlites mingled with granules of augite and iron
ores. In other specimens the ground-mass contains some glassy matter usually blackened with
iron-ore dust, and is of intersertal or cryptocrystalline texture.
In one of the rocks olivine is serpentinized and much reduced in amount, but its place is taken
by a small quantity of pleochroic hypersthene, illustrating the affinities of these olivine-basalts with
the more common hypersthene-augite-andesite lava-type. This association suggests that the olivine-
basalts are possibly due to some accumulative process operating in the early stages of the crystallization
of a pyroxene-andesite magma from which olivine began to separate.
A closely comparable olivine-basalt has been described from Edinburgh Hill, a volcanic vent in
Livingston Island on the M'^Farlane Strait coast (Ferguson, op. cit. p. 44; (i), p. 66). Mr Ferguson's
fine photograph {op. cit. pi. i, fig. i) illustrates the magnificent columnar structure of this plug. An
olivine-basalt also occurs in the Desolation Islands, off the northern coast of Livingston Island (this
paper, p. ^i). Olivine-basalts of very similar characters have been described by H. H. Thomas
from Roberts Island ((2), p. 86). Basalts have also been described from the volcanoes of Deception
Island and Bridgeman Island.
The mainland coast opposite Penguin Island, according to Mr Marr, consists of cliffs of lava,
fronted by extensive raised shingle beaches. Only one specimen was collected from this locality.
This is a typical augite-andesite with a beautiful pilotaxitic texture. The few phenocrysts are small
and consist mainly of a colourless augite which is, however, occasionally zoned with cores and bands
of a yellowish variety. The remaining phenocrysts are of andesine feldspar (Ab5An4). This lava is
quite fresh and no doubt belongs to the younger lava series.
Many specimens of the coarse rounded shingle on the beaches of Penguin Island and the adjacent
mainland were collected. These consist of the older andesite lavas, together with many of the typical
plutonic rocks of the region — granite, adamellite, tonalite, quartz-monzonite, quartz-pyroxene-
diorite, etc., and two highly metamorphic types, quartz-chlorite-biotite-schist and hornblende-
granite-gneiss.
Martin s Head and The Lions Rump. These are conspicuous adjacent headlands on the western
side of King George Bay. Mr Marr's report states that the basal portion of both headlands consists
of a dark grey columnar 'basalt' about 100 ft. in thickness, and with the columns inclined at a steep
angle towards the south. At Martin's Head the ' basalt ' is overlain by a massive rock with a ' twisted
appearance' (? confused columns), and from 50 to 60 ft. in thickness. This in its turn is covered by
what appeared to be a tuff (Fig. 3). Behind the headlands are tuff slopes characterized by an
abundance of angular rock fragments of many different kinds (.'' agglomerate). About 200 ft. above
the Lion's Rump there is what appears to be an old volcanic crater, now almost completely filled with
dirty stagnant ice (Fig. 4). A little to the north of the headland is a conspicuous lava flow reaching
the sea. Near by, perched on the beach, are several gigantic erratics of conglomerate, one of which
must weigh more than 200 tons. The conglomerate is exceedingly coarse, containing rounded water-
worn stones from a few inches in diameter to some 2 ft. across.
The columnar lava of Martin's Head is a fresh hypersthene-augite-andesite of the type common
among the younger lava series. An andesite of similar type, but much richer in feldspar phenocrysts,
poorer in augite, and apparently devoid of hypersthene, was collected i mile east of the Lion's Rump.
From the same locality comes a green mudstone, consisting of finely divided quartz and vermicular
chlorite, much of the latter being aggregated into small rounded or ellipsoidal pellets. It is difficult
to diagnose this rock in the absence of data regarding its field occurrence, but it may be a muddy
sediment made up of decomposed wash from a surface composed of the older andesite lavas.
m^t^JtA
jmjii.
b<^c^c^U.^W*eir 'buv^t^U.tA
jMriOTVJMviv
Fig. 3. Martin's Head.
7ajA-''i<JBjV0
Cfl
O
Fig. 4.
SOUTH SHETLAND ISLANDS
49
The 'conspicuous lava flow' north of the Headland is a doleritic type of andesite characterized
by an exceedingly coarse intergranular texture. The rock is mainly composed of laths of plagioclase
(AbjAnj), pale brown augite, and serpentinous patches which may represent vanished olivine or
hypersthene or both. The ferromagnesian minerals form clots as is common in this lava series. A
sparingly developed earlier generation of feldspars, slightly larger than the laths, is highly zonal,
both chemically and mechanically, and gives rhomboidal cross-sections.
Four specimens of the boulders in the great conglomerate erratics consist of typical augite-andesites
differing among themselves only in the texture of their ground-masses. The numerous large pheno-
crysts consist of plagioclase (mainly about AbaAug), and pale brown augite. Many of the feldspars
show strong chemical and mechanical zoning. They are often full of inclusions except for a narrow
zone of oligoclase on the margins. Nevertheless, many of the feldspars are quite free from inclusions.
In fact the cloudy feldspars look rather like xenocrysts, especially when thev^occur in juxtaposition to
perfectly clear crystals. The facts that these rocks carry the bluish apatites, and occur in a hard
HARMONY cove., NELSON STI^-, SOUTH SHETLPiNDS V;.n» ^oU«.i^ +.1 iS,
67MI0N 1465
ISi-ftMO STl?>Mr
N A M.
F'g- 5-
coarse conglomerate of well-rounded boulders indicating a long period of erosion, suggest that,
notwithstanding their freshness, they belong to the older series of lavas.
Examination of a series of pebbles from the agglomerate in the vicinity of Martin's Head and
Lion's Rump shows that the majority consist of hornblende-augite-andesite lavas and their tuffs.
In addition, there is an altered doleritic andesite somewhat similar to that described above, a highly
epidotized andesite obviously belonging to the older lava series, and an altered tonalite in which the
feldspars have been thoroughly sericitized and epidotized, and the ferromagnesian minerals chloritized.
The hornblende-andesite is an unusual type which has not hitherto been described from West
Antarctica. In the best-preserved specimen brownish green pleochroic hornblende in well-shaped
crystals comes next to plagioclase in abundance as phenocrysts, and is greatly preponderant over
augite. The ground-mass is dense and cryptocrystalline.
"nelson island
Harmonv Cove. Very little is known about the geology of Nelson Island. Mr Ferguson {op. cit.
p. 43) visited Harmony Cove, a harbour at the western corner of the island where Nelson Strait
joins Bransfield Strait, and collected a quartz-diorite-porphyry which appeared to be intrusive into
an igneous breccia. Dr Mackintosh collected four specimens from Harmony Cove. His account is
almost entirely topographical, but he has provided an excellent sketch of the rock exposures (Fig. 5).
Study of these specimens confirms Mr Ferguson's results. One of them is a fine-grained norite
50
DISCOVERY REPORTS
consisting of labradorite (somewhat albitized and epidotized), fresh pale augite with which the
feldspar laths are sometimes in ophitic relation, numerous brown pleochroic pseudomorphs after
hypersthene, and much diffused chloritic matter. There is also a micro-granular variant of this type
with porphyritic feldspars and hypersthene (bastite), and highly epidotized. An outcrop near the
glacier (Fig. 5) consists of pyroxene-andesite of a type common among the older lava series. It
shows porphyritic feldspars (andesine), pale brown augite, and chlorite pseudomorphs after ortho-
rhombic pyroxenes, in a very fine-grained intergranular ground-mass. The fourth specimen, from
the shore, is an igneous breccia mainly composed of angular fragments of altered andesite, much
epidotized, and peppered with cubes of secondary pyrites. Mr Ferguson's specimen of igneous
breccia from the same locality, however, is rich in fragments of the more acid dacitic and rhyolitic
lavas.
COPPERMINE COVE , GnQLISM STRAiT, SOUTH SHETLAND:
*i.tfV^ ->-"*j-L.-i:rta^ii
STATION 14-85
TABL€ I
.CUM
ii.JXJ^a4>
X tnajiKd lEl tun«^j.t"iot^;ti^ CI ^jiuv^ ^<.Mo C«_ «^\ii.*. ahaUc.
■DVKE
^ ;^r1^f-?-^--rt-yi^?-rfT-^-.-y;^^Ty^^
Fig. 6.
1^^ "
ROBERTS ISLAND
Coppermine Cove. This anchorage is situated at the north-western end of Roberts Island, close
to the multitude of small islands and rocks which are scattered over the northern exit of English
Strait. Specimens were collected by Dr Mackintosh from a small peninsula ending in a flat-topped
columnar rock known as Fort William (Fig. 6). Opposite the anchorage (reports Dr Mackintosh)
are cliffs of reddish breccia, presumably volcanic, and Fort William appears to consist of columnar
basalt. In this respect it resembles Table Island, and many, if not all, of the islets and rocks in the
vicinity. Many rock specimens were collected between the anchorage and Fort William. A dike
about 5 ft. thick cuts the cliff opposite the anchorage.
The only previous description of rocks from Roberts Island is that by H. H. Thomas ((2), pp. 85-7).
He describes five specimens from Coppermine Cove, all porphyritic olivine-basalts and all showing
considerable variations in the relative abundance of the porphyritic constituents, and in the richness
of the ground-mass in ferromagnesian minerals. Most of Dr Mackintosh's specimens are also olivine-
basalts of varying composition and texture. Thus the columnar rock of Fort William is a feldspathic
olivine-basalt, or rather dolerite, with a ground-mass of excessively coarse intergranular texture
SOUTH SHETLAND ISLANDS 51
composed of lathy plagioclase (about AbiAiii), pale brown augite, and iron ores. Both feldspar and
augite occasionally attain micro-porphyritic dimensions. The abundant fresh olivine, however, forms
large phenocrysts.
A 'common type' along the shore is a basalt with numerous small feldspar phenocrysts, and less
numerous olivine and augite crystals, embedded in a ground-mass of intersertal texture. This recalls
the Dunsapie type of the Scottish Carboniferous, as was also remarked by Dr Thomas. Another type
which appears to be abundant in this locality is one with an intergranular ground-mass exceedingly
rich in augite. Dr Thomas described rocks of this type but, unlike our specimen, his material contained
much olivine. Some of these augite-basalts, as they might be called, carry numerous little prisms of
l(iw double refraction and straight extinction which are identified as enstatite, in the ground-mass
along with the monoclinic pyroxene. This is an enstatite-basalt. Dr Thomas described a similar rock
as hypersthene-basalt.
While most of the specimens collected here are basalts, one is an augite-andesite of the common
type belonging to the younger lava series. It is accompanied by an andesitic agglomerate. Beach
pebbles collected from Coppermine Cove consist of tonalite and granite-aplite.
LIVINGSTON ISLAND
Livingston Island is the second largest of the South Shetland group, but very little is known of
its geology. Mr Ferguson collected an olivine-basalt from a fine columnar exposure forming a small
island off the coast in M^Farlane Strait (Edinburgh Hill), and noted tuff's in the vicinity which, beside
basalt, contained fragments of quartz-diorite and black mudstone [op. cit. p. 43 and pi. i, fig. i).
Desolation Island. Dr Mackintosh collected a few specimens from Desolation island which lies
off the northern coast of Livingston Island. He gives no geological details except that the island is
mainly composed of a columnar igneous rock. It is noteworthy that on the Discovery Chart (Discovery
Reports, vol. vi, 1932, Chart 6) Desolation Island is represented in the shape of an irregular broken
ring, suggesting that it may be a breached crater flooded by the sea ; but this resemblance may, of
course, be quite accidental.
Two of the specimens were collected in situ from columnar outcrops. Both are very fresh and coarse-
grained hypersthene-basalts of an unusual type. The major part of both rocks consists of a coarse
intergranular admixture of laths of labradorite (Auen-An^o) with granules of pale green augite, prisms
of enstatite-hypersthene with faint pleochroism, and iron ores. The feldspar and augite occasionally
form somewhat larger micro-porphyritic crystals, but the rock is not conspicuously porphyritic.
Both kinds of pyroxene, moreover, tend to build small aggregations or clots, which stand out as a
glomero-porphyritic texture. Olivine occurs only sparingly as small pseudomorphs in brownish
serpentine. A small amount of dark brown glass fills up interstices in the ground-mass.
A basalt with orthorhombic pyroxene in the ground-mass was described by Thomas from Roberts
Island ((2), p. 86). OUvine did not occur in this rock, and the augite occasionally formed glomero-
porphyritic aggregates. A closely comparable rock from the same locality has been described in this
paper (p. 51). These rocks are no doubt closely related to the basic hypersthene-augite-andesites
above described, which are so common in the South Shetland Islands. In these rocks, however, the
hypersthene is porphyritic and does not occur in the ground-mass. Barth and Holmsen have given
an interesting discussion of the petrographical problem involved in the presence of hypersthene in
these rocks ((4), pp. 14^17)-
Numerous pebbles from the beaches and fragments from the screes of Desolation Island were
collected. These include tonalite and a sericitized and chloritized diorite, silicified andesitic breccia,
and a series of acidic volcanic rocks including a fluxional rhyolite or dacite with augite, a rhyolitic
52
DISCOVERY REPORTS
tuff made up of angular fragments of the fluxional rock, a biotite-rhyolite, and orthoclase-porphyry
or felsite with only sparse phenocrystic quartz. Finally, a fragment collected from the scree on the
cliffs of a rocky islet near the anchorage turns out to be a crushed sericitic quartzite of a distinctly
ancient aspect.
DECEPTION ISLAND
Deception Island is the best known of the South Shetland Islands. Dr Thomas ((2), p. 81) has
commented on the earlier literature of the island. Mr Ferguson added a few details and published
two excellent photographs {op. cit. p. 44; pi. iii, figs. 2, 3); but the fullest recent description is that
by Holtedahl ((3), pp. 29-47). Deception Island apparently represents a huge breached crater flooded
S£. WALL OF OecePTiON HflKBWR , SOUTH SHeTLAMOS. >«i- ^.S^'i OMctc^y . STATION 1484
ttju-46wu^
^//^<^'>>'
Souk
.(Sjj SeoLclv
dan*
I II
--^*"^4,
''^'':c''
N ■ A .M.
Fig. 7-
by the sea, of which the inner diameter is about 8 km. Holtedahl believes, however, that it is not a
single large crater, but a volcanic ring mountain built around a caldera subsidence bounded by
a circular fault or series of faults.
Dr Mackintosh collected material from the cliffs and slopes on the south-east side of the whaler's
anchorage near the entrance to Deception Harbour. These form a narrow ridge of land separating the
anchorage from Bransfield Strait (see Dr Mackintosh's sketches. Fig. 7). He reports that the whole
of the cliffs shown in the sketch, except beyond Neptune's Bellows,^ consist of an 'agglomerate of
ashes in a yellowish matrix'. It is possible that the yellow colour is mainly superficial, as freshly
broken surfaces generally seem darker. The slopes below the cliffs are mainly of a soft gravel obviously
formed from the disintegrated agglomerate, carrying a fair proportion of solid boulders of agglomerate,
and here and there boulders of a harder dark rock presumably derived from intrusions in the agglo-
merate (andesitic basalt).
1 Apparently the name given to the entrance channel of Deception Harbour.
SOUTH SHETLAND ISLANDS 53
A visit was also made to the bluff on the south-west side of Neptune's Bellows (Fig. 7). The lower
part consists of conspicuous red cliffs, but higher up there are outcrops of the yellowish agglomerate
characteristic of the other side of the channel. The main range of hills in this locality appeared to be
composed of ' cindery lava or scoria ' with reddish black tints. It appears to be the weathered surfaces
of this rock which impart the striking red colour to the lower cliffs. Three rock specimens were
collected from this locality, and a few from localities north of Whaler's Bay (Anchorage?).
The petrography of Deception Island has been dealt with by the writer ((i), pp. 67, 71), who
described olivine-basalt and basaltic tuffs, ^ and hyalo-dacite (ungaite). Dr H. H. Thomas ((2),
pp. 81-5) described ophitic olivine-dolerite, various types of andesite and their tuffs (mostly glassy),
and soda-trachyte (oligoclase-trachyte). He also noted the presence of tridymite and iron-olivine
(fayalite) in some of the more acid types, and of anorthite in the hyalo-andesites. Barth and Holmsen
((4), pp. 8-17) described andesine-basalt and a vesicular, glassy 'pillow-lava', both of which they
regarded as of bandaitic composition, a view which is borne out by their chemical analyses. Further-
more, they gave a full description of a rock which seems to be identical with my oligoclase-dacite
and Thomas's oligoclase-trachyte. Barth and Holmsen find the closest analogues of this rock in the
products of the Santorin volcano in the Aegean Sea, and as it contains 17 per cent of tridymite they
call it tridymite-santorinite.
From the study of Mr Ferguson's original specimens on which I based my first account of the
rocks of Deception Island, of Dr Mackintosh's new material, and of the above literature, it seems clear
that four main types of rock have been erupted from the Deception Island volcano, namely, olivine-
basalts or dolerites (of which there are no analyses), lavas of bandaitic composition, hyalo-andesites
of more acid type, and finally, the trachytic type which has been variously called oligoclase-dacite,
oligoclase-trachyte, and tridymite-santorinite. Eight analyses of Deception Island rocks have been
published (p. 58) from which it seems clear that they form a perfectly gradual series varying from
basic to acid, all of which (except the olivine-basalts) are highly sodic and relatively poor in potash ;
and are mineralogically characterized by the presence of calcic feldspars, orthorhombic and monoclinic
pyroxenes, and, in the more acid types, by fayalite, tridymite, and sodic feldspars.
The following account of the petrography of Deception Island is based on the study of the specimens
collected by Dr Mackintosh, and on the re-study of the material collected by Mr Ferguson ((i),
pp. 58 et seq.).
Olivine-basalt. Only two rocks, both from the Ferguson collection, belong to this type. One is
described in the following terms ((i), p. 67): 'A beautifully fresh rock showing more or less rounded
olivine phenocrysts in a ground-mass of good fluidal texture, which consists of elongated microlites
of labradorite with subordinate granules of augite and magnetite.' The texture can be described more
exactly as fluxional intergranular. A few of the augite crystals are of slightly larger dimensions and
more euhedral than the granules of the ground-mass, and can be regarded as micro-phenocrysts.
The rock has a close resemblance to the Dalmeny type of the Scottish Carboniferous basalts. Its
occurrence is as a pebble in a tuff or agglomerate.
The other olivine-basalt is flow-banded in the hand specimen, but its ground-mass is not so con-
spicuously fluxional as the above. The ground-mass is of coarse intergranular type and consists of
laths of andesine, with granules of pale augite and magnetite. Numerous phenocrysts and glomero-
porphyritic aggregates of fresh olivine and brown augite, together with smaller and much less numerous
feldspar crystals (labradorite) are embedded in the ground-mass. This rock has aflinities with the
Craiglockhart and Dunsapie types of the Scottish Carboniferous basalts.
Basaltic andesites of bandaitic type. These rocks differ from the basalts described above in not
1 These are now regarded as andesitic tuffs.
54 DISCOVERY REPORTS
being conspicuously porphyritic, and in being almost or quite devoid of olivine. All but one of the
six specimens available come from Dr Mackintosh's collection, and were obtained from both sides
of the entrance channel to Deception Harbour. The ground-mass is of the same type as that of the
basalts, that is, composed of andesine laths, and granules of augite and iron ores. A plagioclase of
somewhat more basic character forms numerous laths which run in wavy flow-lines through the
ground-mass. A few large phenocrysts of augite may occur, but olivine, if present at all, is always
in very small quantity, and is altered to brownish serpentine. The ground-mass varies in texture
from coarsely intergranular to fine-grained intersertal, with a brownish glassy base blackened with
iron-ore dust.
These rocks are adjudged to be the same as those described by Barth and Holmsen ((4), p. 9) as
andesine-basalt and pillow-lava of bandaitic type, of which they have provided chemical analyses
(p. 58). Dr Thomas, too, described what is apparently the same type, in the more basic varieties
of his 'hyaloandesites' ((2), p. 82). Both Barth and Holmsen, and Dr Thomas, mention hypersthene
as a constituent of this rock type, but the writer was unable to identify orthorhombic pyroxene with
certainty in the material at his disposal.
Andesite {hyalo-aiidesite). This is the most abundant rock type in both Mr Ferguson's and
Dr Mackintosh's collections. As the analyses show (p. 58), there is a continuous series of com-
positional types from the basic bandaites to the relatively acid oligoclase-andesites (santorinites),
varying chiefly in silica percentage and proportion of ferromagnesian to feldspathic minerals and
quartz. As many of the rocks are of glassy facies, these variations are masked, at least mineralogically,
by the glassy matrix; in thin section the rocks present a relatively unvarying appearance and,
except for one or two more crystalline types, may be grouped as hyalo-andesites. Dr Thomas ((2), p. 82)
described several rocks from Deception Island under this heading.
In hand specimens these rocks are black or dark grey in colour, usually slaggy, vesicular or even
pumiceous, and are obviously of glassy nature. Even the more crystalline varieties are black and of
dense texture. From these black slaggy types there are all transitions to dark, non-vesicular, glassy
rocks, resembling pitchstones, which are, however, more acid than the majority of the types grouped
under the name hyalo-andesite, and properly belong to the oligoclase-andesites or santorinites.
In thin section many of these slaggy rocks are found to be composed of a brownish glass, dusted
thickly with black specks of iron ores, and often highly vesicular. They always show swarms of
plagioclase microlites (oligoclase to andesine), usually in parallel fluxional streams, but occasionally
felted together with the production of pilotaxitic texture. Microlites of pyroxene can often be detected
in varying numbers by their bright polarization tints and oblique extinction. Some microlites,
however, which are indistinguishable from the pyroxenes in their appearance under ordinary light,
have a very high double refraction and straight extinction. ^ It is probable, therefore, that these are
olivines. Olivine does actually occur in very small amount in a few of the rocks as micro-phenocrysts,
and is almost invariably altered with the production of a reddish serpentine. There are also occasional
micro-phenocrysts of andesine and augite.
From these highly vitreous types there are all gradations to almost holocrystalline (micro-crystalline)
types consisting of a very dense intergranular admixture of plagioclase microlites with granules ot
augite and iron ore, which carries fluxional streams of plagioclase laths.
Dr Thomas detected well-formed ciystals of tridymite lining steam cavities and planes of flow in
these rocks ((2), p. 84). Barth and Holmsen ((4), p. 1 1 e^ seq.) found no less than 17 per cent of
tridymite lining steam cavities in one of the more acid types. The writer found abundant tridymite
in only one of the vesicular hyalo-andesites. It lines and fills steam cavities and fracture cracks m
Barth and Holmsen ((4), p. 9) have also noted small elongated crystals of olivine in the ground-mass of these rocks.
1
SOUTH SHETLAND ISLANDS 55
the rock. Associated with and apparently passing into the tridymite aggregates there are a number
of small spherulites giving a perfect extinction cross, of which the constituent fibres have straight
extinction and a refractive index much lower than that of canada balsam. While these may be tridymite,
it is possible that they represent cristobalite. A. G. MacGregor has described both tridymite and
cristobalite from the Recent lavas (pyroxene-bandaite) of Montserrat.' He writes: 'The cristobalite,
besides obviously replacing tridymite laths and twins, often occurs as innumerable rounded to
irregularly shaped spots up to o-i mm. across', but he does not mention any spherulitic structure.
Oligoclase-andesite (oligoclase-trachyte — Thomas; santorinite — Barth and Holmsen; oligoclase-
dacite (ungaite) — Tyrrell). This rock represents a somewhat more acid development of the magma
which gave rise to the hyalo-andesites above described. Its nomenclature presents a rather per-
plexing problem, and it has been given various names by different authors as shown above. As
indicated by the analyses (p. 58), the free silica works out at between 15 and 20 per cent. The writer
has shown that the average andesite contains round about 15 per cent of normative quartz;- and as the
principal feldspar in the rocks under discussion is oligoclase, it is thought that oligoclase-andesite is the
best name for the type. It is, however, of somewhat unusual composition, as shown by Barth and
Holmsen ((4), p. 13), in that the ratio of soda to potash is much higher than in normal andesites.
They have marked this distinction by conferring the name santorinite, since the lavas of Santorin are
found to be the closest analogues of this rock type. Perhaps the most acid types should be called
oligoclase-dacite to mark the presence of as much as 20 per cent of free silica.
In hand specimens these rocks vary from light grey compact 'stony' to black pitchstone-like
material, which carries scattered whitish crystals of feldspars and often shows marked parallel banding
due to flow.
In thin section they are seen to contain very sharply bounded micro-phenocrysts of plagioclase,
augite, enstatite, olivine (fayalite) and magnetite, embedded in a ground-mass which varies greatly
in its proportion of glass to crj'stals. The glass may form at least 50 per cent of the ground-mass ; at
the other extreme the rocks may be almost completely crystalline. The glass is usually yellowish brown
in colour, but may be colourless; it contains many minute needle-like crystallites. Numerous
microlites of oligoclase-albite (and perhaps a potash-soda feldspar) stream through the glass in fluidal
fashion, mingled with minute granules of pyroxenes and iron ores. The feldspar micro-phenocrysts
were identified in my earlier memoir as anorthite ((i), p. 71). Dr Thomas also found anorthite in his
material ((2), p. 82), but Barth and Holmsen ((4), p. 1 1) apparently noted only andesine of composition
AbesAnas . The ferromagnesian phenocrysts include augite (probably diopside) in well-shaped prisms
and octagonal basal sections, enstatite and fayalite. The micro-phenocrysts often cluster in groups.
Only one of the rocks w^as vesicular, and in it was found tridymite lining steam cavities exactly as
reported by Barth and Holmsen.
These rocks resemble some of the more basic pitchstones of the Tertiary igneous episode in the
west of Scotland, notably the types called leidleite and inninmorite,^ especially the latter, which is
reported to contain anorthite phenocrysts. Indeed, the text-figures of the microscopic appearance of
leidleite and inninmorite (e.g. figs. 47, 48) given in the Mull Memoir cited above might pass for some
of the hyalo-andesites and oligoclase-andesites of Deception Island.
Tiijf and agglomerate. Every account of Deception Island emphasizes the abundance of fragmental
volcanic rocks — tufi^ and agglomerate — in the constitution of the volcano. Five specimens from
1 The Royal Society Expedition to Montserrat, B.W.L: 'The Volcanic History and Petrology of Montserrat, with Obser-
vations on Mont Pele, in Martinique', Pliilus. Trans., B, ccxxix, 1938, pp. 58-61.
- G. W. Tyrrell, ' Some Tertiary Dykes of the Clyde Area', Geol. Mag. 1917, p. 31 1-
3 'Tertiary and Post-Tertiary Geology of Mull, Loch Aline and Oban', Mem. Geol. Surv. Scotland, 1924, pp. 281-4.
3-2
56 DISCOVERY REPORTS
Dr Mackintosh's collection have been sliced, and they are found to be singularly uniform in com-
position. They are made up of irregular, angular, and highly vesicular lapilli and scoria, the fragments
usually var^'ing in size between a hazel-nut and a walnut. The fragments consist of glassy forms of
both the basic and acid andesitic types, the black opaque slaggy form and the clear glassy form being
about equally abundant. The glassy fragments are frequently of a bright yellow colour, but some are
brown and a few others of a greenish tint. Many of these fragments have a narrow border of the
black opaque variety, suggesting that the separation of magnetite dust in the glass which gives rise
to the opacity may be due to a reheating or annealing process. There is little or no matrix of finer
material between the fragments, and they appear to be welded together along their contacts. This
material therefore might be better classed as agglutinate'^ than as agglomerate.
Tridymite (and cristobalite?) occurs abundantly in these fragmental rocks, not only lining the
vesicles of the glassy fragments, but also as an edging around the individual fragments. This suggests
that, in these rocks at any rate, the tridymite is of deuteric crystallization. It has been formed shortly
after the consolidation of the fragmental material, and is no doubt due to late emanations derived
from the parent magma.
SNOW ISLAND
This is a small island west of Livingston Island, and west-north-west of Deception Island. It is
geologically unknown, and no description and no record of any landing is known to me. Four specimens
of rocks from Snow Island, however, were found in the first set of material sent to me by the Discovery
Committee, with no record when and by whom collected. Three of the rocks appear to have been
collected in situ from actual exposures, but the fourth is a pebble from a raised beach at 50 ft. above
present sea-level on the eastern coast of the island.
Of the three specimens collected />/ situ on the eastern side of the island one is a quartz-pyroxene-
diorite or feldspathic quartz-gabbro of a type identical with other occurrences in the South Shetland
Islands and the Palmer Archipelago; the second is an oligoclase-andesite breccia with a tuffaceous
matrix containing a good deal of quartz. The third is a quartz-felsite or rhyolite with a scanty crypto-
crystalline matrix. The pebble from the raised beach is quartz-augite-microdiorite, identical with the
quartz-pyroxene-diorite above mentioned except that it contains patches of fine-grained ground-mass.
Even from this scanty material, therefore, the indications are clear that the constitution of Snow
Island is the same as that of the other islands of the South Shetlands group, and that rocks of the
older igneous series are here represented.
DREDGINGS FROM BRANSFIELD STRAIT
A few score of stones dredged from two stations in Bransfield Strait were included in the first
collection of rocks received from the Discovery Committee. These came from St. 175, about 25 miles
south-east of Deception Island, and St. 177, about 27 miles south-west of Deception Island, and were
dredged from depths of 200 and 1080 m. respectively. The stones were probably dropped from the
ice which formerly occupied Bransfield Strait, and which probably moved from the west and south-
west. Some of the material may have been carried by icebergs breaking away from glaciers on the
South Shetlands and the Graham Land coast. The specimens range in size from blocks 6 in. across
to \ in. pebbles. Most of them are angular and facetted, with corners and edges roughly rounded
oflF; only a few appeared to be well-rounded, apparently water- worn pebbles.
As was to be expected, the great majority of the seventy-nine stones sliced consist of the older
series of andesites, dacites, rhyolites, agglomerates and volcanic breccias, which appear to constitute
^ G. W. Tyrrell, Volcanoes {Ylome. University Library), 193 1, p. 66.
SOUTH SHETLAND ISLANDS 57
the main part of the South Shetlands, and perhaps some part of the Palmer x'Vrchipelago and the
Graham Land coast. There is also one hyalo-andesite with good tridymite which certainly comes
from Deception Island and two others which probably come from the same source. Rocks of plutonic
aspect are also well represented in this collection. They include the quartz-pyroxene-diorites and
their porphyries which are common in the South Shetlands and adjacent regions. Diorite, tonalite,
granodiorite, biotite-granite, and their porphyries, together with granophyric granites and true
granophyres, which more probably come from the Palmer Archipelago and adjacent parts of Graham
Land, are also fairly abundant. Rarer types are represented by a basic diorite with abundant brown
hornblende, biotite, and apatite ; and a serpentine derived from augite-peridotite.
The most interesting material, however, is provided bv specimens of sedimentary and metamorphic
character, which are unrepresented among the rocks in the Discovery collections obtained from actual
exposures. Little is known of these rock types in the South Shetlands and adjacent regions as they
have attracted little attention, perhaps owing to the relatively great abundance and conspicuous
characters of the igneous rocks.
Many of the sediments represented among the dredged stones have suffered a low-grade cataclastic
metamorphism by crushing and shearing. Among the unaltered sediments are mudstone, siltstone,
greywacke, arkose and sandstone. There are two mudstones, and both appear to represent exceedingly
fine-grained washes from the weathered surfaces of basic lavas. Microlites of plagioclase can be
recognized in a chloritic and ferruginous clay matrix, and in one of them there is a sparse sprinkling
of angular quartz grains of silt grade. Another mudstone of similar type has undergone a little
crumpling and shearing with the development of thin quartz-chlorite veins.
Seven pebbles appear to represent laminated sediments consisting of alternate beds of grey\vacke
and siltstone or slate in various stages of shearing and crushing. The least altered specimen shows
angular grains of quartz and subordinate feldspar in a siliceous ground-mass of silt grade in which
quartz is mingled with finely divided sericite, chlorite, epidote and iron ores. This material is pene-
trated by thin veins of secondary silica, now recrystallized to lines of granular quartz. The other
members of this series have undergone severe cataclasis, whereby ultimately quartz-chlorite-schist
has been developed from the greywacke bands and phyllite from the slaty bands. Three of the
specimens show signs of having first been broken up by crushing into an angular breccia in which,
by further shearing, the fragments have been drawn out with the production of a kind of mortar
structure, and with the development of much coarse chlorite and white mica. In one specimen, which
is relatively poor in quartz and rich in chlorite and epidote, it is probable that basic igneous rock
fragments made up the greater part of the original greywacke. The extreme term of alteration is
represented by a true schist consisting largely of quartz, biotite and sericite, in which mortar structure
is finely developed.
One specimen is an interesting arkose consisting of extremely angular grains of quartz, alkali-
feldspar and plagioclase, small chips of andesite and keratophyre (?), a few bits of garnet and epidote,
and many flakes of unaltered biotite, in a ferruginous clay matrix. This composition suggests the
rapid waste of a mixed terrain consisting of granitic rocks, andesitic lavas, and perhaps some meta-
morphic rocks.
Finally, there is a true sandstone consisting mainly of angular to subrounded grains of quartz, with
less abundant grains of alkali-feldspar and plagioclase, a few chips of slate and siltstone and, above
all, many large angular fragments of pale garnet.
Mudstones, greywackes, quartzites and igneous breccias have been described from the South
Shetlands, but especially from the Palmer Archipelago ((i), p. 74; (4), p. 28). Ferguson {op. cit., p. 37)
described siliceous and argillaceous sediments interbedded with the lavas and tuffs of the older
58
DISCOVERY REPORTS
igneous series in Admiralty Bay, King George Island. The present study of dredged stones from
Bransfield Strait has brought out the fact that somewhere in the surrounding region there must be
a basement series of greywackes, mudstones and slates, which has undergone severe cataclastic meta-
morphism. There is good evidence from contact-metamorphic effects that the plutonic masses of
the South Shetlands, the Palmer Archipelago and Graham Land, have broken through this sedimentary
basement ((i), pp. 75-7), and also through the older series of andesite lavas. Hence the metamorphosed
sedimentary basement must be at least of early Mesozoic age, and quite possibly Palaeozoic.
CHEMICAL CHARACTERS
For a discussion of the chemistry of the igneous series of the South Shetland Islands there are
available twelve previously published analyses and two others made for the present investigation
and here published for the first time. Seven of the twelve published analyses were given by E. Gourdon
Table i a. Analyses of
igneous rocks from Deception Island
I 2
3
A
4
5
6
7
8
SiOa
69-01 68-28
67-71
68-33 60-62
56-89
52-93
53-50
49-84
AI2O3
14-21 15-95
14-65
14-94 16-22
16-07
15-86
17-62
19-37
FcOg
2-23 2-00
1-59
1-94 1-76
i-8i
2-01
2-58
3-42
Feb
2-89 1-82
3-29
2-67 5-67
7-08
8-90
6-07
3-69
MgO
0-62 0-09
0-85
0-52 1-62
2-79
3-63
4-39
4-71
CaO
2-II 1-78
2-34
2-08 ' 4-18
5-89
7-60
9-22
12-35
Na^O
6-30 , 7-03
6-09
6-47 : 6-25
5-89
5-03
4-15
2-50
K2O
2-07 1-75
1-99
1-94 1-20
0-94
0-64
0-75
0-87
H,0+)
H,0-)
0-09 0-24
o-i6
0-16 0-56
(0-56
1 0-08
0-42 1
0-04)
0-00
1-79
TiO,
0-58 0-70
I -00
0-76 1-54
1-79
2-29
1-65
1-32
P2O5
0-12
0-07
0-16
0-12 , 0-24
0-21
0-35
0-36
o-ii
MnO
—
—
—
—
o-o8
O-II
—
S
—
—
—
—
0-06
0-06
—
100-23 9971
99-83
99-93
99-86
100-14
99-87
100-29
2-0
99-97
Q
2I-I 18-5
20-1
1
20-2 8-1
1-9
— I-O
2-7
F'
64-9 70-5
62-5
65-3 60-5
55-2
46-2
39-0
26-7
M'
14-0 II-O
17-4
14-5 31-4
42-9
54-8
59-0
70-6
link
89-2 84-1
82-6
84-9
71-7
66-4
56-4
43-1
26-3
k
177 j 14-3
17-7
i6-i
1 1-4
9-5
8-0
lO-I
20-0
I.
2.
3-
A.
4-
5-
6.
7-
8.
Trachyandesite, Deception Island. E. Gourdon, C.R. Acad. Sci., Pan's, clviii, 1914, p. 1906.
Tridymite-santorinite, Deception Island. Barth and Holmsen ({4), p. 14).
Trachyandesite, Deception Island. Gourdon, op. cit.
Average of nos. i, 2, and 3.
Andesite,* Deception Island. Gourdon, op. cit.
Bandaite,f pillow-lava. Deception Island. Barth and Holmsen ((4), p. 11).
Andesine-basalt, Deception Island. Barth and Holmsen {{4), p. 11).
Basalt (' Labradorite' — Gourdon), Deception Island. Gourdon, op. cit.
Doleritic basalt, block (in tuff or agglomerate.') Deception Island. Gourdon, op. cit.
* The alkalis in this analysis are given as recorded in Gourdon 's first paper of 1914, i.e. NajO, 6-25; K„0, 1-20. In Washing-
ton's Tables (U.S.G.S. Prof. Paper 99, 1917, p. 466) the alkalis are given as Na^O, 6-67; K2O, 0-78, and as the summation
remains the same it seems clear that 0-42 per cent has been transferred from KjO to Na,0. This may have been a correc-
tion of the original analysis when it was transmitted to Washington by Gourdon, but it has been thought best to leave the
original figures intact, especially as they are repeated in Gourdon's later work published in Deuxieme Expedition Antarctique
Franfaise (1908-1910), commande par le Dr Charcot: Mineralogie, Geologic, Paris, 1917, p. 7. The earlier figures for the alkalis
are also more accordant with the serial characters of the Deception Island suite than the later.
•f- Correct summation, 100-14, given in the table. Barth and Holmsen give 100-08.
SOUTH SHETLAND ISLANDS
59
in a short paper, ' Sur la constitution mineralogique des Shetland du Sud ' {sic),' with only exiguous
petrographical notes. Four new analyses are given in the 1939 paper of Barth and Holmsen ((4),
pp. II, 14, 25). The remaining analysis is a computation made from a Rosiwal estimate of mineral
proportions in a quartz-gabbro from King George Island by the writer ((i), p. 65). The two new
analyses made for this work are of a tholeiitic lava type from Fildes Strait (p. 44), and of the Recent
olivine-basah lava of the Penguin Island volcano (p. 45). Thus there are now available analyses
of eight rocks from Deception Island, five from King George Island, and one from Bridgeman
Island.
Table I b. Analyses of igneous rocks from King George Island and Bridgeman Island
SiO.,
AlA
FcaOj
FeO
MgO
CaO
Na^O
KaO
H,0~)
CO.,
Tido
P205"
MnO
57-30
17-97
2-17
379
2-57
6-72
3-25
0-96
4-26
0-56
0-20
Q
F'
M'
nak
k
99-75
17-3
34-7
48-0
35-8
17-5
54-9
15-6
5-4
7-0
27
9-1
2-9
1-7
0-7
II
loo-o
10-6
34-7
54-7
42-5
27-7
53-45
19-37
3-37
4-09
4-42
8-iS
3-55
1-35
1-69
0-66
0-04
12
13
100-17
53-02
15-57
4-40
6-58
3-93
8-15
2-38
1-68
(2-02
jo-so
tr.
i-i6
0-35
o-i6
99-90
48-26
17-42
3-36
5-6i
8-83
11-56
2-44
0-89
0-24 1
o-i6|
nil
1-07
0-22
0-14
14
I00'20
3-8
38-5
57-7
37-9
20-8
10-7
30-9
58-4
37-2
31-6
-5-2
25-0
80-2
28-6
i8-4
54-24
17-20
2-81
4-98
5-84
10-19
2-91
0-92
0-09
0-91
0-09
ioo-i8
5-3
29-3
65-4
33-7
17-5
9-
10.
II.
12.
13-
14.
Hypersthene-andesite, Admiralty Bay, King George Island. Gourdon, op. cit.
Quartz-gabbro, intrusion, Le Poing, Admiralty Bay. Tyrrell ((i), p. 65).
'Dolerite',1 dike, Admiralty Bay. Barth and Holmsen ((4), p 25).
Tholeiitic basalt, lava, Fildes Strait, King George Island. New analysis by F. Herdsman. , . ,
Olivine-basalt, lava of Recent volcano, Penguin Island, King George Bay, King George Island. New analysis by
F. Herdsman.
Basalt, Bridgeman Island. Gourdon, op. cit.
X The description of this rock by Barth and Holmsen makes it tolerably clear that it is a porphyritic hypersthene-augite-
andesite, practically identical with the rock of the dike in Admiralty Bay described in the present paper (p. 43). As this is
a very conspicuous feature in Admiralty Bay, it is very probable that the two specimens come from the same dike.
The fourteen available analyses are set out in Tables i a and i b in the above geographical order.
The von Wolff normative parameters as modified by the writer are also given.- In these O represents
the excess or defect of molecular silica, a positive number giving the amount of normative quartz,
and a negative figure representing the amount of olivine. F is the percentage amount of normative
alkali-feldspar (orthoclase and albite), and M' the combined percentage of anorthite, pyroxene,
1 C.R. Acad. Set., Paris, ci.viii, 1914, pp. 1905-7.
2 A full account of this method of calculation will be published as soon as possible.
6o
DISCOVERY REPORTS
iron ore and apatite. The symbol nak represents the percentage of alkah-feldspar in total feldspar,
and k the percentage of potash feldspar in total alkali-feldspar. Thus :
salicCNa^O-KaO)
nak-
100,
X I GO.
salic(Na20.K20.CaO)
, _ salic K2O
salic (Na^OTK^O)
The geographical arrangement of the analyses in Tables i a and i b shows at once that there is a
considerable difference between the Deception Island series at the southern end of the South Shetland
archipelago, and that of King George Island and Bridge-
man Island at its northern end. The Deception Island
series is characterized throughout (except no. 8) by
comparative richness in alkalis as against lime, as shown
by the high nak ratios. Moreover, in the alkalis, soda is
extraordinarily high in relation to potash, as is shown by
the low k ratios. The members of this series show regular
chemical variations throughout, again with the exception
of no. 8, which stands apart in several particulars. This
rock is described by Gourdon as ' doleritic basalt '. It is
stated to occur as ' blocks ' (.'' in agglomerate or tuff), and
is not found in situ} As its analysis agrees fairly closely
with those of the Recent basalts of King George Island
(no. 13) and Bridgeman Island (no. 14), it is possible that
the rock represents a fragment torn from a foundation
of Recent basalts through which the Deception Island
volcano, of quite different constitution, has burst. It will
be so regarded in the present investigation.
The serial relations of the Deception Island series are
shown in the variation diagram (Fig. 8). The silica per-
centages, and the values for F' and M' , were tried as
abscissae against which the other constituents were plotted.
It was found that F' gave the smoothest curves. In all
cases analysis no. 7 (Gourdon 's ' labradorite ') was some-
what discrepant from the others. The curves show the
same general trends as for other andesitic series. The
distinguishing feature of the diagram, however, is the
height of the NagO curve and its distance from the K,0
curve.
The Deception Island rocks may thus be regarded as an andesitic series of quite abnormal sodic
composition (Barth and Holmsen, (4), p. 13). On the other hand, the King George Island and
Bridgeman Island suite, together with the block of doleritic basalt (no. 8) from Deception Island,
constitutes a quite normal series of pyroxene-andesites ranging to olivine-basalt, with accompanying
plutonic types, and belongs to the great circum-Pacific petrographic region of which the characteristic
lava type is hypersthene-augite-andesite.
1 E. Gourdon, ' Sur la constitution mineralogique des Shetland du Sud (lie Deception)'. C.R. Acad. ScL, Paris, CLViii,
1914, pp. 583-6.
Fig. 8.
SOUTH SHETLAND ISLANDS
6i
The Deception Island series. It is difficult to match the rocks of the Deception Island series with
those of other andesitic fields. Very occasionally one finds soda-rich andesites as, for example, in the
Andean petrographic region, and in that of western North America ; but the more normal andesitic
types are overwhelmingly predominant in these regions. As a suite the Deception Island rocks are
almost unique. The only other series which approaches them in richness in soda is that of the Santorin
volcano in the Aegean Sea, as has already been pointed out by Barth and Holmsen. But even among
the Santorin analyses only two are closely comparable to the ' santorinite ' of Deception Island. In
Table 2, col. B, the closest Santorin analogue of the Deception Island santorinite (Table 2, col. A)
Table 2. Deception Island ' santorinite ' and comparable analyses
A
B
C
D
E
SiOa
AUOa
Fe.03
FeO
MgO
CaO
Na.O
K.,6
H.,0+|
HoO-j
TiOa
P2O5
MnO
S
CI
68-33
14-94
1-94
2-67
0-52
2-08
6-47
1-94
o-i6
0-76
0-12
64-99
14-32
1-30
4-01
1-12
3-94
6-20
1-99
1 0-05 1
I nil J
2-23
o-oi
0-07
65-9
15-8
1-6
3-4
I-o
3-5
5-1
2-1
0-4
I-O
O-I
O-I
69-00
14-48
1-25
I-OI
0-36
2-34
6-00
2-76
2-19
0-24
66-05
13-29
3-22
S-°7
1-36
0-50
6-67
0-87
(1-88
(0-96
0-49
0-09
?tr.
?tr.
?tr.
9993
100-23
lOO-O
99-63
100-45
0
F'
M'
)iak
k
20-2
65-3
14-5
84-9
i6-i
14-2
63-0
22-8
86-4
17-3
20-6
54-8
24-6
67-1
2I-I
21-8
67-7
10-5
89-4
237
20-8
62-8
16-4
93-5
7-8
A. Average santorinite, Deception Island (Table i a).
B. Hyalodacite, east lava flow, August 1925, Fouque Kaimeni, Santorin, Aegean Sea. Quoted from H. S. Washington,
'Santorin Eruption of 1925', Bull. Geol. Soc. Atner. xxxvn, 1926, p. 378.
C. ' Santorinite', average of eleven analyses of the Recent lavas of Santorin volcano, Aegean Sea.
D. Biotite-andesite, Inca-loma, Cotopaxi, Ecuador. A Young, Hochgeb. Republik Ecuador, 11, 1904, p. 256. Quoted from
Washington's Tables [op. cit. supra), p. 154.
E. Keratophyre, Trevennen, St Goran, Cornwall. Quoted from Cliem. Anal. Ign. Rocks, etc. Geol. Surv. Gt. Brit.
1931, p. 85.
is tabulated. It agrees closely with the Deception Island analysis except for silica, which is 3 per cent
lower. The von Wolff parameters also show concordance except for 0. Even Santorin is not a very
close analogue for the Deception Island volcano, as is shown by the average of eleven accordant
analyses of the lavas of that volcano (Table 2, col. C). The Deception Island rock is distinctly richer
in soda and silica, and poorer in potash than that of Santorin.
Among Andean andesites the biotite-andesite of Inca-loma, Cotopaxi (Table 2, col. D) provides
a close comparison with the santorinite of Deception Island. Further, some rocks of the keratophyre-
spilite association are chemically similar to those of the Deception Island series, as is shown by an
analysis of a Cornish keratophyre (Table 2, col. E) ; but the k ratio of this rock is notably smaller, and
the nak ratio higher, than those of the Deception Island rock (see also Table 4).
62
DISCOVERY REPORTS
The intermediate rocks of the Deception Island series are even more difficuh to match. The
oUgoclase-andesite (Table 3, col. 4) can be paralleled, and that not very closely, by an andesite from
the Sincholagua volcano in Ecuador (Table 3, col. F), and by a trachytic andesite from the Recent
lavas of the Modoc Quadrangle, California (Table 3, col. G). The bandaitic pillow-lava of Deception
Island (Table 3, col. 5) can be most closely compared with a hypersthene-augite-andesite from
Grenada, B.W.I. (Table 3, col. H); and less closely, at least in respect of the nak and k ratios, with
an andesitic ash from Cotopaxi, Ecuador (Table 3, col. I). It is to be noted that the Ecuadorian
Table 3 . Intermediate lavas of Deception Island and comparable analyses
4
F
G
5
H
I
J
SiO,
AlA
Fe,03
FeO
MgO
CaO
Na,0
K„0
H^O t !
H„'o )
TiO„
P2O5
MnO
S
CI
60-62
l6-22
1-76
5-67
1-62
4-i8
6-25
1-20
0-56
1-54
0-24
58-82
i6-35
5-50
2-36
4-37
4-06
5-31
2-02
1-05
0-36
0-25
59-98
16-71
2-52
5-04
2-22
4-84
5-12
1-63
0-19
1-30
0-43
o-ii
56-89
16-07
1-81
7-08
2-79
5-89
5-89
0-94
1 0-56!
I0-08J
1-79
0-21
0-08
0-06
56-51
14-07
4-04
4-65
3 95
8-44
5-32
0-79
1-51
0-19
0-23
tr.
56-89
19-72
4-06
3-65
I-9I
5-87
5-14
1-96
0-62
tr.
tr.
tr.
tr.
54-53
13-06
6-85
4-86
3-14
9-83
4-62
1-59
0-52
0-96
99-86
100-45
100-09
100-14
99-70
99-82
99-44
Q
F'
M'
nak
k
8-1
60-5
31-4
71-7
II-4
7-6
56-5
35-9
65-8
19-8
11-2
52-4
36-4
60-4
17-2
1-9
55-2
42-9
66-4
9-S
3-4
49-5
47-1
68-1
9-6
3-8
56-4
39-8
53-9
20-2
3-9
48-3
47-8
71-1
18-7
4-
F.
5-
H.
I.
J-
Oligoclase-andesite, Deception Island (Table i, col. 4).
Pyroxene-andesite, Ceballos-chupa, Sincholagua volcano, Ecuador. A. Young, op. cit. supra, p. 24S. Quoted from
Washington's Tables, op. cit. supra, p. 452.
Trachytic andesite (Platy Andesite Group), south of Medicine Lake, Modoc Quadrangle, California. H. A. Powers,
'The Lavas of the Modoc Lava-bed Quadrangle, California', Amer. Min. xvii, 1932, p. 292.
Bandaite (hypersthene-augite-andesite), pillow-lava. Deception Island (Table i, col. 5).
Augite-hypersthene-andesite, Grenada, B.W.I. J. B. Harrison, Rocks and Soils of Grenada, 1896, p. 10. Quoted from
Washington's Tables, op. cit. supra, p. 466.
Andesitic ash, Cotopaxi, Ecuador. J. W. Mallet, Proc. Roy. Soc. XLii, 1887, p. 2. Quoted from Washington's
Tables, op. cit. supra, p. 764.
Augite-hypersthene-andesite, Mt Kouragio, Aegina, Greece. H. S. Washington, 'A Petrographical Sketch of Aegina
and Methana, Part III', J. Geol. in, 1895, p. 150.
volcanoes have provided two of the comparable analyses in Table 3. It would appear that the andesites
of these volcanoes are more sodic than the usual run of Andean andesites. It is interesting to find,
also, that an augite-hypersthene-andesite from the Aegean region (Table 3, col. J) has some chemical
characters in common with the bandaite of Deception Island.
It will be noted that all the Deception Island rocks and the comparable types dealt with in Tables 2
and 3 have been characterized by a ratio F'jM' greater than unity. In the remaining rocks of the
Deception Island series, the andesitic basalts (Table 4, cols. 6, 7), however, this ratio is less than unity.
The andesine-basalt (Table 4, col. 6) is closely comparable with another Ecuadorian lava, a basalt
SOUTH SHETLAND ISLANDS
63
from the Ruminahui volcano (Table 4, col. K). Some spilites as, for example, those of Oregon
(Table 4, col. L), are also quite similar. The basalt (' Labradorite ' — Gourdon) of Deception Island
(Table 4, col. 7) differs from the andesine-basalt only in its positive O. Comparable analyses are those
of a hornblende-soda-andesite-basalt, an inclusion in dacite lava from the San Franciscan volcanic
field of Arizona (Table 4, col. M), and a hypersthene-augite-andesite from the Czerhat Mountains of
Hungary (Table 4, col. N). These rocks, however, are only isolated examples of the type, for in both
the Arizona and Hungarian fields the great majority of the andesites otherwise comparable to the
Deception Island rocks have a much higher k ratio.
Table 4. Andesitic basalts of Deception Island and comparable analyses
6
K
L
7
M
N
SiOa
Al.,03
Fe,03
FeO
MgO
CaO
Na^O
KjO
H2O+
H2O-
CO2
TiO,
P2O5
MnO
SO3
CI
52-93
15-86
2-01
8-90
3-63 '
7-60
5-°3
0-64
0-42 i
0-04 j
2-29
0-35
52-92
i6-66
476
4-89
7-96
5-71
5-12
0-89
o-8o
078
53-15
14-39
1-28
9-33
4-74
7-04
4-58
I-OI
( 2-02 1
I0-I9)
O-IO
1-50
0-19
0-14
53-50
17-62
2-58
6-07
4-39
9-22
4-15
0-75
0-00
1-65
0-36
53-97
1 6-00
4-56
3-63
6-36
7-47
4-38
1-23
(1-31
1 0-03
nil
1-46
o-io
nil
tr.
52-80
19-44
3-47
5-15
2-33
8-70
471
1-12
1-26
0-21
1-05
0-24
O-II
99-87
100-49
99-66
100-29
100-50
100-59
Q
F'
M'
nak
k
-i-o
46-2
54-8
56-4
8-0
-3-6
47-5
56-1
56-1
10-9
-1-5
44-9
56-6
60-3
12-9
2-0
39-0
59-0
43-1
lO-I
2-0
43-8
54-2
53-5
15-^
0-3
47-5
52-2
46-3
13-6
6.
K.
7-
M.
N.
Andesine-basalt, Deception Island (Table i, col. 6).
Basalt, Panang Hondon, Ruminahui volcano, Ecuador. A. Young, op. cit. supra, p. 243. Quoted from Washington's
Tables, op. cit. supra, p. 538.
Spilite, Poorman Mine, Oregon. J. Gilluly, ' Keratophyres of Eastern Oregon and the Spilite Problem', Amer. J. Set.
XXIX, 1935, p. 235.
Basalt ('Labradorite' — Gourdon), Deception Island (Table i, col. 7).
Hornblende-soda-andesite-basalt, inclusion in hornblende-soda-dacite, Bill Williams Mt, San Franciscan Volcanic
Field, Arizona. H. H. Robinson, ' The San Franciscan Volcanic Field, Arizona', U.S.G.S. Prof. Paper 76, 1913, p. 147.
Hypersthene-augite-andesite, Czerhat Mountains, Hungary. A. Vendl, 'Ober die Pyroxenandesite des Czerhat-
gebirges (Ungarn)', Min. u. Petr. Mitt, xlii, 1932, p. 516.
The Deception Island series has been treated at some length because, chemically at least, it appears
to be almost unique among andesitic series, especially in its richness in soda. As a series, only that
of the Aegean volcano Santorin approaches it in chemical character, although sporadic examples of
similar rocks occur in andesitic regions of the normal type, and especially among the volcanoes of
Ecuador.
It is not necessary to deal with the King George Island and Bridgeman Island series in such detail,
for it consists of perfectly normal andesites and basalts conforming closely in their minerals and
chemistry with the great circum-Pacific granodiorite-andesite region, and other similar regions (western
4-2
64 DISCOVERY REPORTS
North America, Hungary, New Zealand, etc.). The hypersthene-andesite of Admirahy Bay (Table i b,
col. 9) closely accords, except for lower potash, with an average hypersthene-andesite computed by
the author from 114 analyses derived from the circum-Pacific region, including the East and West
Indies, and certain European fields (Sardinia, Hungary, Aegean Sea).i The quartz-gabbro (Table 16,
col. 10) agrees well with an average of 11 analyses of rocks so called taken from Washington's Tables
{op. cit. siipra).^ The tholeiitic basalts of the series (Table i^, cols. 11, 12, 14) are accordant with the
average Non-porphyritic Central Magma-type of MuU,^ and with as yet unpublished average analyses
of tholeiitic types from the Tertiary igneous region of Scotland. They also accord with the sparsely
developed basalts which are found in the great andesitic regions.
The above-mentioned rocks are all over-saturated with silica (positive O) ; and in this respect the
under-saturated olivine-basalt (0=— 5-2) of the newly discovered Penguin Island volcano (King
George Island) stands quite apart from the rest. With M', 80-2, it is also the most basic lava type
from the South Shetland Islands so far analysed. Its closest analogue appears to be the olivine-basalt
or 'plateau-magma type' of the Tertiary igneous series in Scotland,^ although it is richer in alumina
and lime and poorer in the ferromagnesian oxides than that type, and is thus richer in plagioclase
feldspar and poorer in olivine. It is precisely in these chemical and mineral characters that the
comparatively rare basalts occurring in andesitic regions differ from the olivine-basalts which are the
most abundant and characteristic types of oceanic regions and of many mildly and richly alkaline
regions on the continents. Thus the olivine-basalt of Penguin Island preserves its relationship with
the associated andesites, notwithstanding its superficial similarity to the olivine-basalts of quite
different petrographical regions.
CONCLUSIONS ON THE GEOLOGY OF THE SOUTH
SHETLAND ISLANDS
A synopsis of the geology of the Danco Land Coast (Graham Land), the Palmer Archipelago, and
the South Shetland Islands was given in my memoir of 192 1 ((i), p. 75). The following are relevant
excerpts from that summary :
The oldest rocks in the region (excluding a possible basement of crystalline schists and gneisses) appear to be
a series of folded bluish slates and mudstones, with subordinate fine-grained sandstones and greywackes, and
abundant intercalations of coarse breccias made up principally of igneous fragments.. . .The igneous breccias. . .may
possibly be as much due to the rapid denudation of an earlier range of porphyry mountains under arid conditions,
as to explosive igneous action.. . .
Because of the abundance and size of the plutonic masses the sedimentary series is only visible in small fragmentary
exposures on the Danco Land coast. It appears, however, to occur in great force on the islands of the Palmer
Archipelago, in which the igneous breccias are also especially prominent. The sedimentary series constitutes a large
part of the South Shetland Islands, especially King George Island. Blue mudstones are intercalated with the older
andesites around Admiralty Bay, and are intersected and metamorphosed by the intrusion of Noel Hill, in Marian
Cove. . . .
The presumably Mesozoic mudstones are interbedded with an early series of andesite lavas in King George
Island, and possibly also in the other islands of the South Shetland group. The plutonic masses of Noel Hill and
Le Poing intersect and cause hornfelsing in both sediments and lavas.. . .
The next event in the geological history of the region seems to have been the extrusion of a great series of later
andesites, which, in King George Island, are regarded by Mr Ferguson as being banked up against the older series
and interbedded mudstones to the north-west. An eruptive focus of this period is probably to be seen in Three
Brothers Hill, Potter's Cove, a columnar plug of typical fresh bandaite lava. . . .
1 G. W. Tyrrell, 'The South Sandwich Islands. Report on Rock Specimens', Discovery Reports, iii, 1931, p. 195.
2 G. W. Tyrrell, The Principles of Petrology, 1926, p. 120.
3 'Tertiary and Post-Tertiary Geology of Mull', Mem. Geol. Surv., Scotland, 1924, p. 17.
* G. W. Tyrrell, 'The Geology of Arran ', Mem. Geol. Surv., Scotland, 1928, p. 121.
SOUTH SHETLAND ISLANDS 65
The latest volcanic episode seems to have been the extrusion of olivine-basalt lavas mainly from a series of volcanoes
in the north-west side of Bransfield Strait (Deception Island; Edinburgh Hill, Livingston Island; Bridgeman
Island). These volcanoes are largely built of basalt tuffs with subordinate basalt lavas and intrusions. Deception
Island, however, contains hyalodacites and oligoclase-trachytes, as well as basalts. Nordenskjold (Antarctis, 1913,
p. 11) suggests that these volcanoes may have some relation to the subsidences of the Bransfield Strait region.. . .He
regards the Bransfield Strait volcanoes also mainly as of early Quaternary age; but Deception Island, and probably
Bridgeman Island, continued erupting until recent times.. . .
The main addition we have been able to make to Nordenskjold's account of the region is the recognition of folded
sediments in the South Shetland Islands, similar to those of the Palmer Archipelago and the Danco Land coast,
but here interbedded with, and covered by, typical Andean lavas. It seems probable that a tectonic zone parallel
to those of the Palmer Archipelago and Graham Land runs through the South Shetland Islands. It is worthy of
note that the intensity of plutonic action diminishes towards the outer (north-western) part of the region. Plutonic
rocks build up the greater part of the mainland ranges; they are also abundant in the Palmer Archipelago, but folded
sediments are here also very conspicuous, while in the South Shetlands plutonic masses are small and isolated, and
very subordinate in bulk to the sediments and lavas. Conversely the volcanic rocks are very largely confined to the
South Shetlands, and are rare in the Palmer Archipelago and the Danco Land coast.
The new Discovery II collections described in this memoir make it clear that King George Island,
at any rate, and probably all the larger islands, are mainly composed of the older series of andesites,
dacites, rhyolites, etc., with their tuffs, volcanic breccias and agglomerates, which are interbedded in
places (Admiralty Bay; Marian Cove) with argillaceous and arenaceous sediments, all conjecturally
of late Mesozoic age. This series is intersected by a number of tonalite, diorite and gabbro intrusions.
Although Ferguson {op. cit. p. 37) has tabulated a thick section of the older andesites, tuffs, agglo-
merates and sediments in Admiralty Bay, it seems possible that the importance of the sedimentary
intercalations has been exaggerated in previous accounts. Ferguson himself collected only a very
few of these sediments, and other collections from many localities in King George Island have not
included any. If the sediments had been at all prominent in the field, it seems likely that they would
have bulked much more largely in the collections, notwithstanding their inconspicuousness in contrast
with the more spectacular igneous rocks.
On the other hand, the importance of the plutonic intrusions in the make-up of the South Shetland
Islands may have been minimized in previous accounts. The Discovery II collections have brought
to light the existence of a large mass of diorite on the eastern coast of King George Island ; and diorite
seems to form a part of the previously unknown Snow Island. Diorites are also known to occur in
Livingston Island, Greenwich Island, and Nelson Island. These rocks are certainly intrusive into
the older series of andesites and sediments, as shown by their contact-metamorphic effects. It may
be conjectured that these plutonic masses are the underground equivalents of the later and fresher
series of andesite lavas which appear to be unconformably banked up against, and superposed upon,
the older andesite series. That a long period of erosion succeeded the extrusion of the older series is
shown by the occurrence of large erratics of coarse conglomerate at Martin's Head (p. 49), which
contain well-rounded boulders of the older andesite, altered tonalite, and comparatively fresh augite-
andesite. Since the last-named contains the blue apatites characteristic of the older series of lavas, it
is a reasonable assumption that all the boulders and pebbles belong to the older series.
The latest volcanic episode is represented by a series of Quaternary or Recent volcanoes along
Bransfield Strait, the craters of which are still well preserved. It is probable that the Deception Island
and Bridgeman Island volcanoes have erupted within historical times (Ferguson, op. cit. pp. 36, 45).
A very notable addition to our knowledge has been provided by Mr Marr's discovery of the Penguin
Island volcano (p. 45). The lavas of Penguin Island and Bridgeman Island are olivine-basalts.
Olivine-basalt was also erupted at Deception Island ; but the main products from this volcano were
slaggy and glassy andesites of peculiar composition (p. 54).
66 DISCOVERY REPORTS
Another noteworthy addition to our knowledge made by recent Discovery II expeditions is the
existence of several basaltic volcanoes on the north-western side of the South Shetland Islands.
Desolation Island, off the northern coast of Livingston Island, consists of columnar basalts of Recent
aspect. On M^Farlane Strait, not very far to the east, is the beautiful columnar basalt plug surmounted
by agglomerate of Edinburgh Hill, discovered and figured by Ferguson {op. cit. pi. i, fig. i). Then
again at Fort William, Coppermine Cove, on Roberts Island, the islands at the northern end of
Fildes Strait, and on the mainland of King George Island along Fildes Strait, fresh columnar olivine-
basalts were collected which probably mark the sites of Quaternary or even Recent volcanoes. All
these volcanic centres on the north-western side of the South Shetlands have obviously suffered
considerable denudation, and are therefore somewhat older than those on the Bransfield Strait side.
There can be no doubt but that these occurrences will be augmented in number when the geological
survey of the South Shetland Islands is carried out in detail.
Finally, it is possible that the South Shetlands rest on a basement of crystalline schists and gneisses,
with sedimentary rocks in various stages of cataclastic metamorphism. Boulders and pebbles of these
rocks are numerous in shore and glacial accumulations, and among the dredged material from Bransfield
Strait (p. 57). Quite possibly some of this material has been derived from exposures on the South
Shetland Islands, although it is more probable that the bulk of it has come either from the Graham
Land peninsula to the south-east or from the Palmer Archipelago to the south.
PART II. PETROGRAPHY OF ROCKS FROM THE GRAHAM LAND
PENINSULA AND ADELAIDE ISLAND, WEST ANTARCTICA
INTRODUCTION
Among the material sent me for description by the Discovery Committee during recent years I found
small collections of rocks from Cape Roquemaurel, Wiencke Island, and the Marin Darbel Islands,
as well as a large collection of stones dredged a few miles off the west coast of Adelaide Island.
Dr N. A. Mackintosh kindly provided me with a copy of the short geological notes he had made on
Cape Roquemaurel and Port Lockroy in Wiencke Island. These notes have been incorporated with
suitable acknowledgement in the following descriptions. The collections, especially that from Adelaide
Island, have proved valuable in extending our knowledge of the geology of West Antarctica, and in
providing confirmatory evidence in favour of previously expressed views on the relationships of
West Antarctic rocks with those of the southern Andes in Patagonia and Tierra del Fuego.
PETROGRAPHY
STATION 1490 (20 JANUARY 1935), CAPE ROQUEMAUREL, TRINITY PENINSULA,
GRAHAM LAND
Cape Roquemaurel is situated on the northern coast of the Trinity Peninsula, the eastern termination
of Graham Land, in long. 58° 30' W., lat. 63° 30' S. In his notes on this locality Dr Mackintosh
states that: 'The headland consists of several high rocks projecting from the ice-sheet of Trinity
Peninsula. On the south-west side of the outermost rock is a good boat harbour with a very small
beach. The rocks of the headland are said to be about 600 ft. high, and consist of a pale granite-like
rock traversed by conspicuous dikes of fine-grained blackish rock. On the south-west side of the head-
land beneath the granite (?) a yellowish brown rock could be seen for several hundred yards just
showing itself above the water line. This seemed to be a different kind of rock, though its structure
GRAHAM LAND
67
65°
60"
Smith I. ^ ^Deception
(7 Low I
Palmer
Archipelagro
Anvers
65
s
'■ -I
Victor Hugo \.o I ,^tf ^^><„. , ...„>
70 -
55° W
Fig. 9. Graham Land.
68 DISCOVERY REPORTS
and cleavage [jointing?] did not look much different from the crystalline rocks above it.' [This
may have been a discoloration of the granite due to intensified weathering between tide-marks. —
G.W.T.]
Dr Mackintosh collected four specimens from this locality, two from the main rock formation
(granite), and two from dikes. He remarks that the granite showed some variation within short
distances, especially in the proportions of the darker minerals, and that his two specimens may have
a smaller proportion of the dark minerals than is typical of the rock as a whole.
The main rock is a true granite consisting of quartz, orthoclase, albite-oligoclase, and a very small
amount of biotite largely replaced by pale green chlorite. One of the specimens is verv coarse-grained,
the crystals ranging from ^ in. to h in. in greatest diameter. The feldspar is pinkish white and the
quartz milky blue in colour. The other specimen is finer in grain and shows a white vein of aplite
with a knife-edge contact against the granite.
In thin section the feldspars are seen to be thickly dusted with kaolinitic and sericitic alteration
products. The orthoclase seems to be almost pure, with only obscure traces of albite lamellation.
The albite-oligoclase occasionally shows an approach to the typical chequer-twinning, and is sub-
ordinate in amount to the orthoclase. The quartz and feldspars are sometimes intergrown in a coarse
and obscure graphic structure, especially in the finer-grained specimen. The only ferromagnesian
constituents are a very few flakes of chloritized biotite. The aplite vein consists of a very fine-grained
base of quartz and sericitized orthoclase with a saccharoidal texture, which carries small micro-
phenocrysts of quartz, orthoclase, and albite. It is quite devoid of coloured constituents.
Conspicuous dikes of a blackish rock traverse the granite. One of Dr Mackintosh's specimens is
' probably characteristic of all the black dikes in the headland '. In hand specimen it is a fine-grained
dark grey rock with a few large fresh phenocrysts of feldspar and a sprinkling of pyrites. In thin
section it consists mainly of a panidiomorphic plexus of andesine feldspar and a pale green hornblende
in about equal amounts. In addition, there are a few phenocrysts of labradorite (extinction 30 ),
a little quartz, and numerous fine-grained irregular aggregates of a reddish brown biotite which, in
many cases, are apparently growing at the expense of the hornblende. As these aggregates are
invariably associated with pyrites, they are probably of secondary origin, and connected with the
ingress of sulphide solutions into the rock. This rock is identical with some of the lamprophyres
described by Rosenbusch as spessartite}
The remaining dike specimen was taken from the inner portion of what is probably a composite
dike. This dike was of the same blackish tint as the others. It was about 8 ft. thick, and had a central
part of greenish colour and a foot in width. This, however, is only a surface coloration. When broken,
the fresh rock is of a greyish blue colour and is very dense, with a flow-banding delineated by the
alinement of small pink feldspar crystals. In thin section it proved rather hard to interpret owing to
its denseness and opacity. It appears to consist mainly of straight-extinguishing feldspar microlites
( .^ oligoclase) arranged in a wavy flow-banding, with somewhat larger feldspars (? orthoclase), and
quartz in smaller quantity. The feldspars are all highly sericitized. In this ground-mass material
there are embedded micro-phenocrysts of quartz, oligoclase, and a few pseudomorphs in pale green
fibrous hornblende of what may have been an earlier amphibole. As some epidote is always associated
with the oligoclase, the original crystals were probably of a more calcic composition. On the whole, the
rock has the mineral composition of a dacite. Perhaps an earlier generation of petrographers would
have called it quartz-porphyrite.
1 Osann-Rosenbusch, Elemente der Gesteinslehre, 4th ed. 1922, p. 333.
PORT LOCKROY 69
PORT LOCKROY, WIENCKE ISLAND
Port Lockroy is a small harbour on the west coast of Wiencke Island, opening out on to the
Neumayer Channel which separates the large Anvers Island from Wiencke Island. Rocks from
Wiencke Island and Doumer Island, as well as from the islands in the Neumayer Channel, and on the
south and west of Wiencke Island, have been collected by several expeditions. Thus Pelikan^
described quartz-diorite and gabbro, the former cut by diorite-porphyry and diabase dikes. Gourdon^
described quartz-mica-pyroxene-diorite, quartz-diorite, and micro-diorite, with numerous ' labradorite '
(hornblende-andesite) dikes penetrating the quartz-diorite massif. Ferguson wrote: 'Wiencke Island
is bounded on the side facing Neumayer Channel by almost vertical walls of sedimentary rocks in-
cluding bluish black mudstone ; it is, however, largely formed of gray diorite, which is the only rock
present in Doumer Island and the Cairn Islands.'^ From Ferguson's collection the writer described
tonalite, igneous breccias, and a siliceous mudstone.^
The most recent work on the petrography of this part of the Palmer Archipelago is that of T. Barth
and P. Holmsen.* They described eucrite and anorthosite (with chemical analyses) from an islet
near Victor Hugo Island (west of Wiencke Island). The Joubin Islands, also west of Wiencke Island,
consist mainly of igneous breccias, and an analysis is given of a prehnitized rock fragment from these
breccias. From Port Lockroy, Barth and Holmsen described quartz-diorite and adamellite, with
analyses. They remark that the whole region from Port Lockroy westward to the Joubin Islands and
Victor Hugo Island is penetrated by ' diabase ' dikes. The general picture of the geology of this region
is then that of an ancient basement consisting of sediments and igneous breccias, cut by plutonic
intrusions of tonalite and adamellite, the whole being penetrated by numerous dikes, especially
' diabase '.
Dr Mackintosh collected two rock specimens from an island in Port Lockroy harbour. Both consist
of tonalite identical with that described by me from Ferguson's collection, but the larger specimen
shows a sharp contact of tonalite with a dike of fine-grained grey micro-porphyritic rock which is a
porphyritic micro-tonalite. In thin section the tonalite shows biotite, hornblende, and magnetite as
mafic constituents, with very abundant euhedral plagioclase (andesine, Anjo), all of which are embedded
in a coarse ground-mass consisting of interlocking crystals of quartz with subordinate orthoclase.
Biotite and hornblende are present in roughly equal amounts. The hornblende is variegated in
shades of green, the larger crystals breaking up into aggregates of smaller, diflterently coloured
grains.
The dike rock shows numerous phenocrysts of andesine with heavy mechanical zoning, and
somewhat fewer phenocrysts of a fibrous, pale green hornblende, enclosed in a very fine-grained
equigranular ground-mass consisting of quartz, orthoclase, andesine, hornblende, biotite passing
into chlorite, and cubes of magnetite. It is a quartz-diorite porphyry or tonalite-porphyry ; or, if
it be desirable not to use the ambiguous term ' porphyry ', it may be designated as porphyritic micro-
tonalite.
1 A. Pelikan, ' Petrographische Untersuchungen der Gesteinsproben ', Resultats du Voyage de S.Y. 'Belgica', Exped.
Antarctique Beige; Geologie, Anvers, 1909.
" E. Gourdon, 'Geographic physique, Glaciologie, Petrographie', Exped. Antarctique Frattfaise, 1903-5, Paris, 1908.
3 D. Ferguson, 'Geological Observations in the South Shetlands, the Palmer Archipelago, and Graham Land, Antarctica',
Trans. Roy. Sac. Edin. Liii, 1921, p. 49.
« G. W. Tyrrell, 'A Contribution to the Petrography of the South Shetland Islands, the Palmer Archipelago, and the
Danco Land Coast, Graham Land, Antarctica', ibid. pp. 59, 73, 74.
^ ' Rocks from the Antarctandes and the Southern Antilles ', Scientific Results of the Nonvegian Antarctic Expeditions 1927-28
and 1928-29, instituted and financed by Consul Lars Christensen, No. i8, Norske Vidensk.-Akad., Oslo, 1939, pp. 17-33-
70 DISCOVERY REPORTS
THE MARIN DARBEL ISLANDS
This group of small islands and rocks lies a few miles south-west of Cape Bellue at about long.
66" 20' W., lat. 66 00' S. In a brief note accompanying the specimens, in which the above location
is given, they are wrongly allocated to the Biscoe Islands, which form a long chain of islands north-east
of Adelaide Island. The above-given latitude and longitude are those of the Marin Darbel Islands.
I have been able to find no previous reference to the geology of these islands.
The specimens collected are stated to come from a small uncharted island lying to the south-west
of Cape Bellue. This island, like all those in the vicinity, consists of an ice-worn mass of igneous rock.
Two large specimens of this rock (norite) were taken ; five others represent dikes penetrating it.
The main rock of the island is a coarse plutonic type of a mottled, greenish grey tint, consisting
of white feldspars and greenish black ferromagnesian minerals. In thin section the appearance of
coarse grain is seen to be illusory, for the rock consists of large areas of fresh labradorite (An55) in small
crystals, alternating with larger and more isolated crystals of hypersthene, augite, and magnetite. The
hypersthene is mainly fresh and distinctly pleochroic, but some crystals are in process of alteration
to a pale green fibrous bastite mineral, and a few to brown biotite, both modes of alteration being
accompanied by the disengagement of magnetite. There is also some primary iron ore. The hypersthene
is apparently slightly preponderant over the pale diopsidic augite, and the periods of crystallization
of the two minerals appear to overlap. Thiis the rock is a norite or more exactly a hyperite, since the
hypersthene is accompanied by a notable amount of monoclinic pyroxene. In another specimen the
hypersthene has gone over completely to bastite.
Three of the dike rocks are dark, greenish grey, aphanitic types in which numerous micro-
phenocrysts of serpentinized olivine and feldspar can be made out with the lens. In thin section they
turn out to be olivine-basalts with very numerous micro-phenocrvsts of bytownite (Augo) and almost
equally numerous olivines which are perfectly euhedral but completely altered to pale green serpentine.
The ground-mass is very minutely crystalline, and consists of microlites of plagioclase, augite, and
magnetite. Numerous spherical steam cavities are present which are usually filled with fibrous,
radiating, pale green delessite. A fourth specimen is much coarser and is highly carbonated. It appears
to represent a coarse basalt or dolerite.
That part of the Graham Land peninsula and the Palmer Archipelago which lies between lat.
64°-67° S. and long. 62 "-66° W. seems to be rich in gabbroic intrusions and basic dikes. Thus
Pelikan {op. cit. siipro) described gabbros and dolerite dikes from Anvers Island, Bob Island (off
south coast of Wiencke Island), and Cape Anna (Danco Land). Gourdon, likewise {op. cit. supra)
described basalt dikes from Wiencke Island and Doumer Island, diabase dikes from Booth Island
(Wandel I.), diabase and gabbro from Petermann Island and Cape Tuxen. From the Andvord Bay
region the writer {op. cit. supra) described basalt dikes and an intrusion of fresh olivine-gabbro (Bruce
Island). Barth and Holmsen {op. cit. supra) commented on the abundance of basic dikes in the region
between Victor Hugo Island and Port Lockroy (i.e. along the line of lat. 65° S.), and described
eucrite and anorthosite from Victor Hugo Island.
ADELAIDE ISLAND
Adelaide Island is a large island off the coast of Graham Land at about lat. 67" S., long. 69° W.
Geologically nothing is known of the main island, but the French Expedition of 1903-5 collected rock
material from three small islands, Jennv, Leonie, and Webb Islands, off its south-eastern coast.
Gourdon {op. cit. supra) described them as consisting of gabbro cut by numerous dikes of basalt,
diabase, and andesite, and has given no fewer than ten analyses of these rocks.
ADELAIDE ISLAND 7,
Among the first set of Discovery II material sent me I found a box of stones dredged from St. 599,
off the west coast of Adelaide Island at a depth of 203 m. The exact position of the Station is lat.
67 08' S., long. 69 o6|' W. Forty-six of these stones were examined and thin sections made. They
ranged in size from boulders 9 in. in greatest diameter to pebbles less than i in. across. As these
dredgings were taken only a few miles oif the western coast of Adelaide Island near the central point
of the western coastline, it is likely that many, if not all, were derived from this geologically unknown
land.
Ten of the stones belong to the granite family, including ordinary granite, granophyre, granodiorite,
and tonalite. Eight are quartz-diorites, three dioritic lamprophyres, and one quartz-gabbro. No fewer
than fourteen of the specimens are quartz-porphyries or allied rocks, all of which show signs of
crushing and brecciation, in extreme cases reducing them to ' porphvroids ' and even to types which
might be regarded as metamorphic quartzite. Five of the stones are lavas, including rhyolite, dacite
(or dellenite), and andesite. Finally, the collection includes five andesitic breccias similar to those
which have been described from other parts of the Antarctandes.
One of the two true granites consists of a coarse-grained allotriomorphic mixture of quartz, micro-
perthitic orthoclase, and somewhat less abundant albite-oligoclase which is much more heavily dusted
with clayey alteration products than the orthoclase. The sparse ferromagnesian constituents are
mainly chloritized biotite, and there are a few crystals of fibrous hornblende.
The second granite, like the first, is of a pale flesh-pink colour, but is of finer grain and obviously
richer in dark constituents. The feldspars consist of micro-perthitic orthoclase and oligoclase (Ab,o)
in roughly equal quantity. The oligoclase frequently forms well-shaped crystals which are enclosed
in the larger plates of orthoclase. Both feldspars tend to be poikilitically enveloped in a mosaic of
large grains of quartz, and both exhibit coarse intergrowths with quartz. The chief ferromagnesian
constituent is biotite which is mostly chloritized. With abundant magnetite, sphene, and apatite, the
chloritized biotite mainly occurs in small clots or segregations which appear to be of cognate origin.
Both these granites are, strictly speaking, adamellites, as plagioclase occurs to the extent of more
than one-third of the total feldspar.
One of the pebbles is a good granophyre consisting almostentirely of a fine micro-graphic intergrowth
between quartz and very turbid orthoclase. This encloses a few larger crystals of rounded and embayed
quartz. The original ferromagnesian minerals appear to have been biotite, now chloritized, and a few
flakes of muscovite; but a later mineralization has brought in some large aggregates consisting of
calcite, radial sheaves of muscovite, and irregular masses of pyrites.
Next comes a granitoid rock which bears a considerable resemblance to the second adamellite
described above, as it carries the same clots of chloritized biotite, but with epidote and pyrites instead
of sphene and magnetite. It differs, however, in its more richly ferromagnesian character, and
especially in the relation between the feldspars. In this rock oligoclase occurs in distinctly superior
amount to the orthoclase. It is therefore to be classed as granodiorite. Another stone is a porphyritic
micro-crystalline variety of this type, and may be called granodiorite-porphyry or porphyritic micro-
granodiorite.
Five stones belong to the tonalite group. Tonalite, in the author's opinion, is a granitoid rock inter-
mediate between granodiorite and quartz-diorite, distinguished by its abundant plagioclase relative to
orthoclase while retaining an amount of quartz sufficient to exclude it from the quartz-diorite group.
Its ferromagnesian constituents are mainly hornblende and biotite. They are more abundant than in
the granites and less abundant than in the quartz-diorites.
Each of the five stones assigned to this group conform more or less closely to the above definition.
Two of them contain biotite, mostly altered to chlorite and epidote, as their sole ferromagnesian
72 DISCOVERY REPORTS
mineral, with magnetite and apatite as accessories. In one of these rocks the biotite is interleaved with
narrow lenticles of a colourless mineral of high refraction and birefringence, straight extinction,
and good cross-fracture, which is doubtfully identified as sillimanite. The remaining three tonalites
have a considerable amount of green hornblende in addition to biotite, and sphene is a rather abundant
accessory. One of these rocks, however, has a well-marked granulose structure, and the irregular
grey-green plates of hornblende are spotted with rounded inclusions of quartz and feldspars. This is
the 'sieve structure' which is often taken as a sign of hybridism.
The diorite family is represented by eight rocks of which six are typical quartz-mica-diorites,
consisting of plagioclase (oligoclase to andesine), hornblende, and biotite, with a small residuum of
quartz and occasionally a little orthoclase. Magnetite and apatite are the most important accessory
minerals, and the apatite often occurs in some abundance as comparatively large crystals. Pyrites,
epidote, and chlorite occur as secondary minerals, the two last-named being the products of alteration
of feldspar and biotite respectively. The six quartz-diorites vary among themselves within narrow
limits in the proportions of dark to light minerals, and in the relative amounts of hornblende and
biotite.
The seventh quartz-diorite is distinguished from the above-described by containing a notable
amount of colourless augite, which occurs in small clots or segregations with hornblende, biotite,
magnetite, and apatite. It is therefore a quartz-mica-augite-diorite of a type approximating to
Stelzner's 'andendiorit' from the Argentinian Andes. The eighth rock assigned to the diorite group is
a micro-diorite of very fine grain and uniform, allotriomorphic granulose texture, consisting of andesine
and green hornblende in about equal quantity. A small amount of biotite is involved with the horn-
blende as well as a notable quantity of apatite and magnetite, and there is also a small residuum of
quartz. This rock may be regarded as a mesocratic quartz-micro-diorite which shows affinity to the
malchite of Osann.^
Only one of the stones in this collection falls in the gabbro family. It is a medium-grained rock
consisting of plagioclase, probably labradorite, but now intensely altered with the production of
aggregates of epidote and unidentifiable turbid matter; pale augite, and an almost equal amount of
faintly pleochroic hypersthene which is largely altered to chlorite. A little brown hornblende occurs
as an alteration product of the augite. Magnetite and apatite constitute the only accessory minerals,
together with a small residuum of quartz. This rock may therefore be described as quartz-hypersthene-
gabbro or quartz-hyperite. It is probably to be correlated with the quartz-gabbros of the Jenny
Island group off the south-eastern coast of Adelaide Island. -
The three lamprophyres in the collection all belong to the spessartite group, and consist essentially
of green hornblende and andesine with typical panidiomorphic texture. The hornblende is somewhat
in excess of the plagioclase. One of the rocks contains numerous phenocrysts and crystal aggregates
of hornblende in the lamprophyre ground-mass. Another contains patches of a pale bleached biotite
and of pale green chlorite, with a few micro-phenocrysts of feldspar. The third has much chlorite
and magnetite, and its hornblende is mostly of the brown variety. All these rocks carry a small
residuum of quartz. This group of lamprophyres appears to be abundant in the Graham Land
peninsula and the adjacent archipelagos.
We now come to the most interesting and important group of stones from Adelaide Island, namely,
the acid volcanic rocks, including rhyolite, dacite, and igneous breccias which contain a variety of acid
types. The breccias consist mainly of quartz-porphyry fragments which have suffered cataclastic
1 Osann-Rosenbusch, Elemente der Gesteinslehre, 4th ed., 1922, p. 321.
2 E. Gourdon, ' Sur la constitution mineralogique de I'lle Jenny (Antarctica)', C.R. Acad. Sci., Paris, 159, 1914,
369-71.
ADELAIDE ISLAND 73
deformation of the same kind as that described by Quensel from the ' porphyry formation ' of Patagonia
and Tierra del Fuego.^ Sixteen stones belong to this group.
Three specimens appear to belong to the rhyolite-dacite group. One is a dense whitish rock mottled
with pale green streaks which exhibit a rough parallelism. In thin section it becomes clear that this
is a coarse and even contorted flow-banding of alternating lighter and darker streaks, more obvious
when the slide is held up to the light than when it is viewed through the microscope. The rock consists
of a quartzo-feldspathic paste of variable but always fine grain, mingled with varying quantities of
sericite and a colourless to palest green, almost isotropic mineral of higher refractive index than quartz
or Canada balsam. This mineral occurs in reticulated areas with a flaky, fibrous, or vermiculate structure
under polarized light. These properties may serve to identify it tentatively as a variety of kaolinite.
Sericite and kaolinite are much more abundantly developed in the darker bands, although they are
not absent from the lighter streaks. The only other identifiable mineral is some secondary pyrites.
The rock is intersected by thin, thread-like, discontinuous veins of secondary quartz. The flow
structure may be primary and the rock therefore a rhyolite; but there is the possibility that it is a
pseudo-flow structure like that of the quartz-porphyries or porphyroids described later, and due to
cataclastic deformation. The facts that some of the larger quartz grains show undulose extinction,
and the considerable development of sericite, may perhaps be regarded as in favour of this view.
Another rock appears to be the same as that described by Quensel- from Patagonia as 'felsite-
porphyry'. This shows small phenocrysts of bipyramidal quartz, orthoclase, and oligoclase, in a
largely cryptocrystalline, quartzo-feldspathic ground-mass. There is, however, a large amount of
recrystallized quartz forming irregular areas which carry inclusions of ground-mass material, and
which impregnate feldspar phenocrysts in their vicinity. Both quartz and feldspar phenocrysts are
euhedral, and the latter enclose large, well-developed crystals of epidote and zoisite. The only coloured
minerals present are a few areas of leucoxene representing altered ilmenite, and some secondary
pyrites. Veins of secondary quartz traverse the rock and cut through some of the feldspar phenocrysts,
but appear to merge into the areas of recrystallized quartz in the ground-mass. This rock is a quartz-
felsite or quartz-porphyry which differs from those later described in its comparative lack of alteration
and in its much smaller proportion of phenocrysts to ground-mass. Its mineral composition roughly
corresponds to that of adamellite or granodiorite, and it might therefore, if a lava, be styled dellenite.
A third member of this group is obviously a fragmental rock of composition similar to the above
except that plagioclase feldspar is much more abundant. It contains numerous angular chips of
rhyolitic or dacitic composition in a uniform cryptocrystalline ground-mass of quartzo-feldspathic
composition. The rock has been heavily impregnated with secondary pyrites which has stimulated
local silicification of the ground-mass. It is best regarded as a dacitic tuff.
Next come three rocks interpreted as coarse tuffs or igneous breccias consisting mainly of fragments
and fine comminuted debris of the rhyolite and quartz-felsite (dellenite) just described. One of them
consists mainly of fragments similar in composition and structure to the above rhyolite, but in general
of coarser grain. There are nevertheless rapid variations in grain size across barely visible boundaries
between adjacent fragments. In fact it was only possible to identify the rock as a rhyolitic breccia
through the occurrence of a few angular fragments of a coarse feldspathic type apparently belonging
to the granite-porphyry described later. Some of the coarse-grained material may be due to secondary
silicification. The two remaining rocks of this group are clearly igneous breccias consisting mainly of
fragments of the dellenite above described.
1 P. Quensel, ' Die Quarz-porphyr- und Porphyroidformation in Siidpatagonien und Feuerland ', Bull. Geol. Inst. Upsala,
xn, 1913, pp. 9-40.
2 Op. cit. supra, p. 14, and fig. 10, p. 27.
74 DISCOVERY REPORTS
The ten remaining stones of the acid volcanic series consist of coarse quartz-feldspar-porphyries
and their tuffs or igneous breccias, in which a progressive series of cataclastic deformations have taken
place, resulting in the formation of typical ' porphyroids ' and, finally, a completely mylonized rock
which can only be distinguished with difficulty from a metamorphic quartzite. While the majority
of the porphyroids and igneous breccias consist of quartz-feldspar-porphyry fragments only, three
contain fragments of rhyolite, felsite, and oligoclase-andesite in subordinate amount.
The series begins with an almost normal, practically unstressed quartz-feldspar-porphyry or
granite-porphyry, containing very abundant phenocrysts of quartz, some a centimetre in length,
orthoclase not quite so large, and still smaller crystals of albite-oligoclase, in a fine-grained ground-
mass of aplitic type which consists of equidimensional crystals of quartz, orthoclase, and albite-
oligoclase. A few small crystals of altered biotite and a little iron ore represent the only ferromagnesian
constituents. The phenocrysts collectively make up considerably more than half the volume of the
rock. Only the large quartz crystals show the beginnings of stress. They are cracked and somewhat
rounded, with narrow zones of granulation along the fissures.
Next comes a series of rocks which may be described as igneous breccias consisting of shattered
fragments of the above quartz-feldspar-porphyry with, in some cases, a few pieces of rhyolite, quartz-
felsite, and oligoclase-andesite. These may, perhaps, be best interpreted as explosion breccias, but
they may possibly represent scree material at least in part. All these rocks have been subjected to
crushing and shearing stress of varying degrees of severity. The quartz phenocrysts have been
shattered and ground-mass material has been forced in between the fragments. Sometimes the
fragments have not been so far separated that the outline of the original phenocryst cannot be traced,
but in more severe cataclasis the fragments have been dispersed far and wide throughout the ground-
mass. Where the stress has not been great the feldspars have retained their crystal forms, but have been
more or less completely sericitized. With more severe shearing the feldspars have been broken down
and may show more or less rounded fragments enclosed in areas of comminuted and sericitized
material. In extreme cases the feldspars are represented merely by elongated areas of sericitized
material the margins of which fade out gradually into the ground-mass. The ground-mass itself has
been sheared and sericitized in the same way, but owing to its finer grain and its consequent greater
mobility under shearing stress, it has been forced to flow round the phenocrysts, producing what
Quensel {op. cit. supra) has called secondary flow structure. The rocks are then typical 'porphyroids',
with elongated strips of felted sericite flakes winding round the broken phenocrysts. Secondary
epidote and chlorite have been produced in some quantity, especially in the breccias that contain
andesite fragments.
What appears to represent the final stage of cataclastic deformation is reached in a quartzite-like
rock which, if seen in isolation away from the associated types, would certainly be regarded as a
metamorphic quartzite or quartz-schist. It consists of alternating strips of coarse and fine quartz
crystals. Some water-clear plagioclase feldspar is mingled with the quartz of the coarse layers, and a
very pale green, almost isotropic chlorite with the fine-grained quartz. The larger quartz crystals
interlock with their neighbours along crenulated margins. Chlorite and ilmenite decomposing to
leucoxene are somewhat concentrated in restricted areas presumably where fragments of andesite
occurred in the original breccia. Patches and veins of clear recrystallized calcite also occur. Not a
trace of sericitization is left. Presumably the sericite, together with particles of iron oxide, has been
reconstituted into chlorite. This rock is somewhat tentatively identified as the mylonized end-product
of extreme cataclastic deformation aflfecting a breccia composed of acid igneous rocks.
The connected series of rocks above described is thus regarded as a complex of acid lavas, or lavas
and intrusions (quartz-feldspar-porphyry, quartz-felsite, rhyolite, dellenite, dacite, and oligoclase-
ADELAIDE ISLAND 75
andesite), with their tuffs and explosion-breccias, which has been subjected to extensive crushing and
shearing. This complex appears to be identical with that described by Quensel {op. cit. supra) from
Patagonia and Tierra del Fuego.
The same or a similar complex of acid igneous rocks has also been noted in at least three localities
in the Graham Land peninsula and adjacent islands. Thus, O. Nordenskj6ld,i writing of the loose
blocks on the land surface and in the moraines, and of the boulders in the Late Mesozoic and Tertiary
conglomerates, found in the northern part of the peninsula, says that they include quartz-porphyries
of various types, some showing such a high degree of mechanical metamorphism that they have been
transformed into sericite-schists. He remarks the similarity of these rocks to the porphyry formations
of Patagonia which he had previously investigated. Again, in 1913, Nordenskjold'^ stated that at
Hope Bay, within the eastern ranges of Graham Land, there occurred acid porphyries and porphyry
tuffs apparently concordant with the folded and metamorphosed Jurassic sediments of that locality.
He further remarked that these rocks are probably the same as those that form part of the South
American cordilleras.
At Hope Bay, on the western side of Antarctic Sound at the northern tip of Graham Land, J. G.
Anderson^* described sediments with Jurassic plants overlain, in Mount Flora, by 200 m. of whitish
tuffs derived from acid volcanic rocks.
Finally, E. Gourdon* described an erratic from the north of Hovgaard Island as a 'rhyolite with
globular quartz', which he regarded as an 'ancient facies' of porphyry. This rock carries porphyritic
orthoclase and bipyramidal quartz, and the crystals are associated with sinuous flow lines. The quartz
is much corroded and surrounded by aureoles of ground-mass material. The rock, he says, has suffered
severe mechanical deformation. It obviously has a close resemblance to the porphyroids of Adelaide
Island described above.
The last remaining group of rocks from the Adelaide Island collection consists of oligoclase-andesite
lavas, and coarse tuffs or breccias consisting mainly of fragments of the same type. Eight stones
are assigned to this group. Two are normal lava types, two are slaggy and vitreous variants, and the
remaining four are coarse tuffs or breccias. The lavas exhibit numerous very small micro-phenocrysts
of fresh oligoclase, usually with well-marked parallel flow-orientation, embedded in a fine-grained
ground-mass consisting of microlites of oligoclase and orthoclase, with chlorite representing the
original ferromagnesian mineral (probably augite). This is peppered with numerous, irregularly shaped
particles of iron ore. t^i a- j
Slaggy variants of this lava contain much dark glass and are somewhat haematitized. The tuffs and
breccias consist of angular fragments of the above-described lava of varying textures, with an occasional
flake of mudstone or shale. Furthermore, volcanic mud has infiltrated into the breccias and acts as
a scanty cement.
These rocks recall the characteristics of some of the older group of andesite lavas which are so
conspicuous in the geological make-up of the South Shetland Islands (Tyrrell, op. at. supra and
preceding paper, pp. 43 et seq.).
CONCLUSIONS
The rocks from Graham Land and adjacent islands described in the foregoing pages strengthen the
already abundant evidence that the igneous rocks of the region, down to the latitude of Adelaide
Island at least, are identical with those of the Patagonian Andes. Of particular interest is the discovery
1 'Petrographische Untersuchungen aus dem westantarktischen Gebiete', Bull. Geol. Inst. Upsala, vi, 1900, p. 241.
2 'Antarctis', Handbuch der Regionalen Geologie, Bd. viii, Abt. 6, 1913, p. 9-
3 'On the Geology of Graham Land', Bull Geol. Inst. Upsala, vn, 1906, p. 24. or
* 'Geographie physique, Glaciologie, Petrographie ', Exped. Antarctique Franf.aise, 1903-5, Pans, 190b, p. 103-
76 DISCOVERY REPORTS
of a quartz-porphyry formation which has undergone intense cataclastic deformation in Adelaide
Island. This formation, which is of Mesozoic age (older than Upper Cretaceous) in Patagonia, and
extends in that country over a belt more than 400 km. in length, is thus shown to continue in Graham
Land to a further distance of about 1000 km.
The evidence of this rock collection thus strongly reinforces the conclusion the writer came to in
an earlier study, namely, that ' the Graham Land eruptives are identical down to the smallest chemical
and mineralogical details with Andean types as far as we know them. The chemical and petrological
similarities are so great that one can have no hesitation in subscribing to Nordenskjold's view that
the Graham Land ranges, and those of the contiguous islands, are the continuations in Antarctica of
the Patagonian chains In Nordenskjold's expressive phrase, Graham Land is a mirror-image of
the southern end of South America. '^
PART III.
PETROGRAPHY OF ROCKS FROM THE ELEPHANT
AND CLARENCE GROUP
6130
Minstrel ,
Bat/
Cornwallis I
SIS
'C Lookout-
^i
^^'^
■c'O'Brien I
Clarence I..
The Elephant and Clarence Group of islands, comprising Elephant Island, Cornwallis Island, and
Clarence Island, in its northern section, and Gibbs Island, Aspland Island, and O'Brien Island to the
south, is usually regarded as a part of the South
Shetlands archipelago (see map. Fig. 10). But
there is a good case for its separation as an
independent group, and for regarding it as on a
parity with the South Shetlands and the South
Orkneys. There is a wide sea gap between Gibbs
Island and King George Island (South Shetlands),
much wider than the distances between the in-
dividual islands of either group; moreover, the
Elephant and Clarence Group is geologically quite
different from the South Shetlands with their
thick coverings of andesite lavas, which are absent
from all the visited islands of the Elephant and
Clarence Group.
, . , , r 1 T-1 I ,. J Fig. 10. Elephant and Clarence Group.
Landmgs on the islands of the Elephant and & f
Clarence Group have been few, and consequently the geological data up to date are very scanty. In
the following pages the available information is assembled and supplemented by the investigation of
new material from Clarence Island and Gibbs Island, collected during expeditions ot the ' Discovery II '.
C SiO'N
;e5
SCALE OF MAU
SI 30
55 0
S"! 0 w
ELEPHANT ISLAND
During the Salvesen expedition of 191 3 the late Mr David Ferguson passed close to Elephant
Island, but was unable to land owing to stormy conditions. He made a few observations from the
ship, however, and has recorded them as follows i^ 'The rocks at the south-east corner of the island
[Cape Lookout?] are light grey to dark, and more or less banded. The grey rocks appear to be stratified
1 G. W. Tyrrell, 'A Contribution to the Petrography of the South Shetland Islands, the Palmer Archipelago, and the
Danco Land Coast, Graham Land, Antarctica', Trans. Roy. Soc. Edin. liii, pt. i, 192 1, p. 78.
2 D. Ferguson, 'Geological Observations in the South Shetlands, the Palmer Archipelago, and Graham Land, Antarctica',
Trans. Roy. Soc. Edin. Liii, pt. i, 1921, p. 35.
ELEPHANT AND CLARENCE GROUP 77
as the bedding is uniform, but some of the darker rocks may be bedded lavas. [Mr Ferguson was in
error here as shown by Prof. Tilley's observations on the Quest Expedition collection — see below.] . . .
Much of the island appears to be formed of stratified sediments. Along the extreme west coast, and
some eight to ten miles out to sea, is a series of sea-worn hummocks, roughly banded, with smooth
slopes, which resemble dark-coloured, table-topped lavas.' [Seal Islands?]
The first landing by a geologist on Elephant Island was made by J. M. Wordie in 1914 as a member
of the party marooned on the island during Sir E. Shackleton's Antarctic Expedition, 1914-17.
Although living under very difficult conditions Mr Wordie made rock collections at Cape Valentine,
the north-eastern point of the island, and at Cape Wild, 6 miles farther west, which were described
by the present writer in a section of Mr Wordie 's account of the geology.^
The rocks of the north-east coast consist of dark grey, indigo blue, bluish green and grey-green
phyllites of fine texture and glossy cleavage surfaces. Many of them are profusely veined and permeated
with secondary silica. The rocks consist of quartz, feldspar (plagioclase), chlorite of three varieties,
calcite, and opaque greyish (sericitic?) and black (carbonaceous) matter. The calcite is always, the
quartz' frequently, of secondary origin. These minerals are arranged in thin, elongated, parallel lenses
representing a small-scale flaser texture indicative of intense pressure metamorphism. These puzzling
rocks are difficult to interpret; some may represent ordinary argillaceous sediments, as Tilley believes
from a study of the similar rocks of Minstrel Bay on the west coast (see below), but others may have
been fine washes from an andesitic terrain, or even andesitic dusts.
These rocks are highly folded and tilted. At Cape Valentine Mr Wordie states that they dip south
by east at about 30°. South of Cape Valentine the rocks dip uniformly to the south and show no
folding. Between Cape Valentine and Cape Wild the dip is to the north and changes rapidly from
verticality to between 30 and 40°. At Cape Wild the dip of foliation is about 60° towards N. 15° W.
At the foot of Mt Houlder (south of Cape Wild) the most striking feature of the section is a reduplica-
tion of the beds by 'concertina' folding. There are thus indications of folding on both a small and
large scale; small-scale folding and foliation were probably contemporaneous, but the large-scale
folding was probably due to a later set of movements.
The Shackleton-Rowett Quest Expedition (192 1-2) landed parties at Lookout Harbour at the
extreme south of Elephant Island and at Minstrel Bay on the west coast. Rock collections made by
Mr G. V. Douglas- have been described by Prof. C. E. Tilley .»
Tilley describes the rocks from Minstrel Bay as dark grey to leaden grey phyllites, much contorted
and penetrated by numerous veins of secondary silica. The constituents are essentially quartz and
albitic feldspar, with scales and closely packed films of chlorite and white mica, abundant carbonaceous
matter and some granules of epidote. These rocks are regarded as normal sediments, and Tilley thinks
there is no reason to believe that volcanic material enters into their composition. These phyllites are
correlated with those of the Cape Wild area described by me (above). On G. V. Douglas's map
(Tilley p 56) signs indicate that the phyllites strike a little south of east and are vertical. Since
these phyllites have been found at Minstrel Bay, and in the area between Cape Wild and Cape
Valentine, it may be conjectured that the northern coast and perhaps the northern halt of the island
consists of these rocks.
1 J. M. Wordie, 'Shackleton Antarctic Expedition, 1914-1?: Geological Observations in the Weddell Sea Area', Trans.
""T'Grolfgtl Re^uUs of?hrShL7e:n:Rowett (Quest) ExpedU.on (Report of lecture)', Quart. Jourr.. Geol. Soc. .xxix,
''3Vp?t;ogra;hicaTNo;:fon Rocks from Elephant Island, South Shetlands', Quest Expedition Report, British Museum
(Natural History), London, 1930, pp. 55-62. ^
78 DISCOVERY REPORTS
On the other hand, the rocks of Lookout Harbour at the extreme south are of markedly different
mineral composition and metamorphic grade. According to Tilley they are divisible into three
petrographical groups:
(a) Garnet-hornblende-albite-schists,
(b) Amphibole-bearing marbles,
(c) Para-amphibolites.
The rocks of these three groups are linked by the general presence of hornblende, and, to a less degree,
albite. Their study, aided by chemical analyses, leads to the conclusion that 'they form a graded
series of related sediments ranging from limestones to impure types giving the amphibolites and
garnet-hornblende-schists rich in albite'. The original sediments were of abnormal composition,
inasmuch as abundant albite was present, probably derived from detrital plagioclase. The grade of
metamorphism is obviously much higher than that of the northern phyllites. No data are given of
the attitude or geological structure of the Cape Lookout series, which may occupy the southern half
of Elephant Island.
CORNWALLIS ISLAND
This is a small island lying in the strait between the much larger Elephant and Clarence Islands.
There is no record of a landing, and nothing is known of the geology except a brief note by Mr Ferguson
(op. cit. supra, p. 35). He says: 'It was not possible to land, but the steamer got very close in. It
[Cornwallis Island] rises sheer out of deep water in splintery crests, and is partly covered with snow.
The highest point of the island may be 1000 ft. or more above sea level. The slopes are very steep,
often quite vertical, and there is consequently much bare rock. ... It is formed of light-grey schistose
rocks, the foliation planes having a direction [of strike] about N. 70-80" E., with a nearly vertical
dip.' Cornwallis Island is not far to the east of Cape Valentine on Elephant Island, where Wordie
recorded the strike as east by north, i.e. about the same as that of the rocks on Cornwallis Island.
Wordie also says that 'the mountains along the coast [of Elephant Island], when of bare rock, have
precipitous slopes and serrated crests of the "frayed cardboard edge" type', which agrees well with
Ferguson's description of the topography of Cornwallis Island quoted above. It may therefore be
taken as probable that Cornwallis Island represents an eastern continuation of the same rocks as
those of the northern coast of Elephant Island.
CLARENCE ISLAND
So far as is known, no geologist had landed on Clarence Island until Prof. O. Holtedahl, in January
1928, managed with some difficulty to get ashore near the northern point (Cape Lloyd) during the
Norwegian Antarctic Expedition of 1927-8.^ But Ferguson, during the Salvesen Expedition of 1913,
passed close enough to Clarence Island to make a few observations {op. cit. supra, p. 36). He says:
' The north-east coast is a wall-like rampart, 500 ft. or more in height, of very regular and well-bedded
rocks, light grey, dark grey, and drab coloured. The west coast shows light grey, finely banded rocks
with a nearly vertical dip in places, and a broad band of brownish rock, evidently an intrusion, was
seen at one place cutting through them.' This description agrees well with Holtedahl's and with
photographic views of the north-western coast of Clarence Island published by Holtedahl {op. cit.
pis. xxiii, xxiv).
As regards the rocks, Holtedahl collected a number of characteristic specimens from the scree at
1 O. Holtedahl, 'On the Geology and Physiography of Some Antarctic and Sub-Antarctic Islands', Scientific Results of the
Norwegian Antarctic Expeditions of 1927-8 and 1928-9, instituted and financed by Consul Lars Christensen, No. 3, Norske
Vidensk.-Akad., Oslo, 1929, 172 pp. (Clarence Island, pp. 47-8).
ELEPHANT AND CLARENCE GROUP 79
the foot of a precipitous mountain wall rising behind the beach where he landed, and from wave-
rounded boulders. He gives the following brief particulars:
The rocks are rather highly metamorphic, grey or greenish in colour, with a more or less distinct schistosity,
rather fine-grained, most of them, however, showing a crystalline texture well already {sic) to the naked eye.
A grey rock is, according to Broch, a fine-grained albite-epidote-biotite-schist, with quartz and hornblende,
further muscovite, titanite, apatite. A chemical analysis shows an andesitic composition.'
A greenish chlorite-schist has a basaltic composition. A grey rock, with hardly any schistosity and less fine-grained,
is by Broch found to be mainly made up of albite, epidote, hornblende, biotite. It probably represents a highly
altered basic igneous rock.
These greenish or greyish rocks show a fairly distinct bedding that may be seen in pi. xxiii, fig. 3. The dip is there
rather var^'ing both as to inclination and direction. The main direction of the strike is probably south-west to north-
east, parallel to the north-western coast. Such a strike is at any rate typical of the extreme western part of the island.
The strike of the rocks in Clarence Island is thus not very different from that in Elephant and
Cornwallis Islands, and it is to be expected that the same or similar rock types will recur in Clarence
Island. From the above brief description of the rocks it would appear that they are comparable in
mineral composition and metamorphic grade with those described by Tilley from the southern point
of Elephant Island.
In the preface to his memoir Holtedahl says that his rock specimens had been assigned to O. A. Broch
for petrological investigation. Eventually, however, the work was taken over by T. F. W. Barth and
P. Holmsen.-
In regard to Clarence Island, Barth and Holmsen give very brief descriptions of a 'common
schistose greenstone' and a chlorite-schist, of which analyses are given. In their Table of Analyses
(p. 60, op. cit.) these rocks are designated respectively as: biotite-epidote-actinolitc-albite-schist, and
chlorite-actinolite-clinozoisite-albite-schist. These analyses are discussed later (see Table 6, p. 87).
DREDGED STONES FROM SOUTH OF CLARENCE ISLAND
Among the Discovery II material submitted to me was a box containing numerous stones dredged
on 23 February 1927 at St. 170 at a depth of 342 m. The exact position is long. 61° 25' 30" S., lat.
53° 46' W. On Chart no. 6^ a sounding of 342 m. is shown about 7 miles south-west of Cape Bowles,
the southernmost point of Clarence Island, but this sounding is shown on the chart at lat. 54° 15' W.,
the longitude being the same as that given above. The position of this sounding is about 30 miles
east-south-east of the eastern coast of Elephant Island.
The question of the provenance of the stones is rather difficult. It depends on the prevalent direction
of the marine currents near Clarence Island, both as affecting direct transport of the stones, and
as influencing the drift of icebergs which may have carried the stones or some of them from Elephant
Island, or even from more southern localities. It will be assumed that the majority of the stones came
from Clarence Island, some from Elephant Island, and possibly a very few from the south.
PETROGRAPHY
The stones range in size from about 3 in. in greatest diameter down to half an inch. They are all
covered with a thick growth of calcareous marine organisms. When this is chipped or dissolved off
it can be seen that most of the stones consist of fine-grained grey and green schistose rocks, often
profusely veined with quartz. Thirty-five of the stones were sectioned for petrographic examination.
Four were found to be igneous rocks, three sedimentary, and twenty-eight metamorphic.
1 This is presumably the analysis of a 'schistose rock' quoted on p. 109 of Holtedahl's memoir.
2 'Rocks from the Antarctandes and the Southern Antilles', Scient. Res. of the Norwegian Antarctic E.xpeditions, 1927-28
and 1928-29, No. 18, Norske Vidensk.-Akad., Oslo, 1939, 64 pp. (Clarence Island, pp. 59-60).
3 H. F. P. Herdman, 'Report on Soundings taken during the Discovery Investigations, 1926-32', Discovery Reports,
VI, 1932.
6-2
8o DISCOVERY REPORTS
IGNEOUS ROCKS
Porphyritic micro-diorite {quartz-diorite-porphyry). This is a fine-grained rock consisting of diversely
arranged laths of plagioclase (oligoclase-andesine), with subordinate chlorite representing an original
ferromagnesian mineral, probably hornblende, irregular grains of titano-magnetite, a little interstitial
quartz, and an abundance of thin needles of apatite. The porphyritic constituents are few and consist
solely of badly altered plagioclase (probably andesine). This rock resembles the quartz-diorite
porphyries which are abundant in the South Shetlands, the Palmer Archipelago and Graham Land.
Porphyritic honibleiide-micro-granite [Iioniblende-quartz-porpJiyry). This is an interesting and
unusual rock with very numerous euhedral phenocrysts of feldspar, quartz, hornblende, biotite, and
ilmenite, with apatite in well-formed crystals as an abundant accessory, embedded in a pale brown,
glassy to crypto-cr}stalline ground-mass. The feldspars are much sericitized and consist of orthoclase
and oligoclase (AbjAnJ in roughly equal proportions. Quartz occurs as large embayed cr^'stals up
to 0-5 cm. in greatest diameter, often with edges and corners rounded by corrosion. The hornblende
forms prisms and plates of green to pale yellowish brown pleochroism, and is often partially or
completely altered to chlorite of high d.r. The biotite is completely altered to a pale green chlorite
of anomalous 'ultra-blue' polarization colour, with the disengagement of magnetite. Ilmenite
altering to leucoxene occurs in large scattered crystals. The phenocrysts form more than half the rock.
Spherulitic quarts-porphyry. This rock contains a few small embayed phenocrysts of quartz, rather
more abundant euhedral phenocrysts of very turbid orthoclase and a few of albite, in a micro-
crystalline and spherulitic ground-mass. The spherulites are often perfect; they may be isolated in
the ground-mass, but more often they are grouped around the phenocrysts. The only ferromagnesian
minerals are a few small areas of chlorite with separated magnetite, and one or two large crystals of
titano-magnetite.
Rhyolite. This rock consists mainly of a crypto-cr^'stalline but obviously quartzose ground-mass,
with numerous parallel streaks of micro-granitic material. The latter consists of quartz and turbid
orthoclase intergrown with the production of a rough micrographic structure. A few small pheno-
crysts of oligoclase, orthoclase and quartz occur, but the only ferromagnesian constituents are
represented by ragged patches of titano-magnetite, and a few flakes of chloritized biotite, which are
associated with the streaks of micro-granite. This rock may be regarded as a rhyolite with flow structure.
It may represent a lava, or perhaps more probably, a small dike.
These acid volcanic or dike rocks may have come from the extreme northern tip of Graham Land,
where O. Nordenskjold has described a similar series, mostly tuflFs, at Flora Bay.^ Also, at Hoffnungs
Bay,- he found acid porphyries and porphyry tuffs, apparently concordant with folded and meta-
morphosed Jurassic sediments.
SEDIMENTARY ROCKS
Only three of the stones can be regarded as unmetamorphosed sediments. These are all greywackes,
one of sand grade, and the other two of silt grade.
The coarser greywacke is grey-green in colour and quartzite-like in aspect. In thin section it is
seen to consist mainly of ver^^ angular fragments of quartz and feldspars, with a little biotite (altered
to chlorite and magnetite), pale pink garnet, and some epidote, sericite, and chlorite developed as
secondary minerals. In addition to the mineral fragments there are numerous rock chips, including
carbonaceous shale, chert, fine-grained quartzite, sericite-schist, and fragments of the ground-mass
of trachytic and felsitic igneous rocks. Most of the quartz shows a marked undulose extinction
1 ' Untersuchungen aus dem westantarktischen Gebiete', Bull. Geol. Inst. Upsala, vi, 1900, p. 239.
^ 'Antarctis', Handbiicli dcr Rcgioiialen Geologic, Bd. viii, Abt. 6, 1913, p. 9.
ELEPHANT AND CLARENCE GROUP 8i
indicative of strain. The feldspars include orthoclase and albite (always turbid), and clear fresh
andesine (Ab^Anao). The rock is traversed by thin veins of secondary quartz, epidote and calcite.
The remaining tw^o rocks have the same composition as that above-described, but the grain-size
is coarse silty. They contain a greater abundance of biotite, chlorite and garnet, but rock chips are not
so much in evidence, probably because of the finer grain. A few crystals of apatite occur in these rocks,
and in one of them carbonaceous streaks delineate the bedding planes. The same slide shows a plane
of shearing along which coarse sericite and chlorite have been developed.
These rocks are probably due to the rapid waste of a terrain of miscellaneous rocks, including acid
and intermediate volcanic types, shales, cherts, quartzites, and schists. The abundance of quartz
with undulose extinction points to the presence of gneisses, or, more likely, of a quartz-porphyry
formation which has undergone extreme mechanical deformation, within the area of erosion.
A mylonized porphyry formation of this character covers great areas in Patagonia and Tierra del
Fuego, and has also been found in West Antarctica as far to the south as Adelaide Island (see this
Memoir, p. 74).
Greywackes and greywacke-siltstones of ancient aspect are common in Tierra del Fuego in forma-
tions of Late Palaeozoic and Early Mesozoic ages;^ and O. Nordenskjold {op. cit. supra, p. 238) has
described non-schistose slates and greywackes underlying fossiliferous sediments of Jurassic age in
Hope Bay at the northern end of Graham Land. He also asserts the abundance of porphyries and
porphyry tuffs in the same area. It is therefore possible that the above-described stones came from
this region ; but, from the identity in composition of the stones, and the fact that they were associated
together in the same dredging, it is considered to be at least as likely that they were derived from the
nearest land, i.e. Clarence Island.
METAMORPHIC ROCKS
Twenty-eight, or four-fifths, of the dredged stones belong to metamorphic types. The great majority
of these are due to the dynamic metamorphism of sedimentary rocks resembling the Scottish ' faikes ',
alternate laminae of carbonaceous shales and quartzose siltstone or sandstone. These rocks have been
intricately folded, sheared, crushed, and converted into carbonaceous sericite-phyllites alternating
with quartzose phyllite and quartz-sericite-schist. Some of the rocks contained a significant amount
of calcareous cement which has been recrystallized as calcite. This mineral is occasionally so abundant
that the rocks have to be recognized as calc-sericite-schists.
Thin flakes of sericite are profusely developed in both the siliceous and argillaceous laminae.
Calcite and chlorite are formed mostly in the coarser quartzose bands. The chlorite, developed from
ferromagnesian impurities in the original sediments, is usually a pale green variety with ' ultra-blue '
polarization colours. It is often vermicular and then almost isotropic. Epidote is sparingly developed
in the earlier stages of metamorphism, and generally in the slaty laminae.
Some of the rocks are minutely folded and puckered, even within the limits of a thin section (Fig. 11),
and the thicker laminae of phyllite acquire a strain-slip cleavage parallel to the axes of folds in the
coarser quartzose layers. Others are sheared and smashed into small fragments with the production
of crush-breccias. These crush-breccias are often rolled out and a kind of flaser structure is developed,
consisting of lenticular fragments of the brittle quartzose layers around which the phyllite laminae
have been forced to wind. The quartz grains grow during this process and uhimately form a coarse
mosaic. Similarly the size and amount of the sericite flakes increase with the degree of internal
movement. These rocks develop into well-crystallized quartz-sericite-schists at the climax of the
metamorphic reconstitution.
1 E. H. Kranck, 'Geological Investigations in the Cordillera of Tierra del Fuego', Ada Geographica, iv, no. 2, Helsinki,
1932, PP- 231-
82 DISCOVERY REPORTS
A few of the rocks, which must originally have been rich in calcareous and argillaceous matter,
contain abundant calcite and epidote. The latter mineral is no doubt produced by the well-known
reaction between calcareous and argillaceous matter during metamorphism. Quartz-calcite-epidote-
schists are thus formed. As the degree of metamorphism increases, epidote becomes the dominant
mineral with the dwindling or disappearance of calcite and sericite. The final product of this change
is a quartzose epidosite. All of these rocks are intersected by a profusion of secondary quartz veins.
A rock which may belong to the above series is a saccharoidal metamorphic quartzite which carries
scattered and irregularly bounded patches of coarse sericitic material. This may perhaps be interpreted
as representing one of the thicker beds of sandstone that may have contained clay galls.
Fig. II. Section of phyllite, showing folding.
Another specimen shows many points of resemblance to the above-described series, especially in
the abundance of argillaceous material and the presence of calcite, epidote and sericite. It differs,
however, in that some of the folia are rich in large, angular fragments of alkali-feldspars, including
orthoclase and albite, which are still comparatively fresh. This may perhaps be best interpreted as
a sheared rhyolitic tuff, intermingled with normal sedimentary material.
The remaining three stones of the metamorphic group are quartz-epidote-amphibole-schists which
have probably been derived from basic igneous rocks or their tuffs. One is a quartz-albite-tremolite-
epidote-schist ; the other two are calcite-quartz-glaucophane-epidote-schists.
The first is a fine-grained, apparently bedded rock with a schistosity coinciding with the bedding
planes. It consists mainly of a mixture of minute grains of epidote with microlites of albite, and
prisms of colourless to pale green tremolite which have a tendency to lie athwart the planes of
schistosity. This material carries large and small folia consisting of quartz and albite, both enclosing
innumerable needles and thin plates of tremolite. The albite often forms large, simply twinned,
blasto-porphyritic crystals developed in a mosaic of quartz and small albites. The largest and coarsest
of these folia has a distinct resemblance to an aplite vein. This rock is somewhat difficult to interpret,
ELEPHANT AND CLARENCE GROUP 83
but the conjecture may be hazarded that it is derived from a rock of the spihtic suite, perhaps a tuff.
It has a considerable resemblance to the slightly metamorphosed spilitic lavas of North Glen Sannox
(Arran).i
Of the glaucophane rocks, one is a quartz-albite-epidote-chlorite-glaucophane-greenstone devoid
of schistosity ; the other is schistose and carries abundant calcite in addition to the above-mentioned
minerals. In both rocks quartz, albite, and calcite, form a coarse, even-grained mosaic, within and
between the grains of which the coloured minerals are developed. In the greenstone the latter are
interspersed among the colourless minerals, and are non-schistose ; in the schist the coloured minerals
occur as streams winding through the colourless matrix, or they form folia alternating with broad
bands consisting of quartz, albite, and calcite.
The chlorite is of the deep green penninite variety with low birefringence and anomalous ' ultra-
blue ' interference colours ; it is associated with colourless to pale green muscovite. The epidote is of
the normal yellowish green variety and is associated with much leucoxenic material. Glaucophane
is abundant in both rocks. It has a striking pleochroism as follows:
X=pale yellowish green,
F= violet,
Z= azure blue.
In the schist it appears to be altering to a greenish blue soda-amphibole devoid of the violet pleochroism,
and with a rather high extinction angle (up to 20 ). This may be the 'abnormal glaucophane' rich in
a lime molecule, which is mentioned by Winchell.'^
These rocks are probably due to the recrystallization of igneous rocks of the spilitic series under
dynamothermal metamorphism. The abundance of quartz and calcite, with a little muscovite, may
indicate that the original rocks were tuffaceous and mingled with normal sedimentary material. Very
similar rocks are mentioned by Harker as forming the prosinite type of the Alps.'' Kranck* has
described a glaucophane-garnet-schist from Bahia Pliischow in Tierra del Fuego. Its mineral
composition is: garnet, glaucophane, quartz, sericite, biotite, chlorite, calcite, apatite, magnetite.
This rock is interbedded with garnetiferous quartz-schists and belongs to the Yahgan or Mt Buckland
formation. Kranck regards it as due to the metamorphism of a carbonate-rich sandstone [greywacke .?] .
THE GIBBS ISLAND GROUP
This is a group of three small islands, O'Brien Island, Aspland Island and Gibbs Island (with
Narrow Island joined to it), lying about 20 miles south-south-west of Cape Lookout on Elephant
Island. Practically nothing was known of the geology of these islands until 1937 when a landing was
made on Gibbs Island by a party from the ' Discovery II '. D. Ferguson, however {op. cit. supra, p. 35),
was caught in a terrific gale and had to shelter for some time under the lee of Gibbs Island. He says:
' The steamer was sufficiently near to show that the rocks were mainly stratified sediments. The rocks
on the west [south?] side of Gibbs Island are dark grey and banded, and dip about 40° W. A higher
horizon is represented by some uniformly and well bedded greyish-white rocks which dip about iS"" W.
They extend for about { mile, and look soft and friable in places. Aspland Island, 5 or 6 miles west
of Gibbs Island, is evidently formed of the same regularly bedded rocks, but they dip east.'
A landing on Gibbs Island and Narrow Island was made by J. W. S. Marr on 2 November 1937,
and the following facts concerning the geology of the island have been culled from his report (un-
published MS.).
1 G. W. Tyrrell, 'The Geology of Arran', Mem. Geol. Siirv., .Scotland, 1928, p. 26.
2 A. N. Winchell, Elements of Optical Mineralogy, Part II, 3rd ed., 1933, p. 259.
^ A. Harker, Metamorphism, 1932, p. 291.
* E. H. Kranck, op. cit. supra, pp. 52-4.
84 DISCOVERY REPORTS
Gibbs Island is high and steep, rising abruptly out of the sea which is deep close inshore. The
coast almost wholly consists of sheer and inaccessible cliffs reaching a maximum elevation of about
I GOO ft. These rock walls are remarkably ice-free, and only a thin mantle of highland ice crowns the
rising ground above them. Gibbs Island is joined to Narrow Island by a low shingle and boulder
spit, 50-80 yards long, which is probably awash at high tide. In its general features Narrow Island
is similar to Gibbs Island.
The south coast of Gibbs Island is largely composed of a fine-grained schistose rock penetrated
by occasional quartz veins. The planes of schistosity are conspicuous from the sea and dip south-west
at an angle of about 30°. Specimens of the rock were obtained from an outcrop near sea level on the
south coast near the landing place and from another outcrop about i ^o ft. higher. The steep screes
which descend to the sea are almost exclusively composed of slabs of the grey phyllite. Above the
screes, starting at 500 ft., is a vertical rock face reaching a height not far short of 1000 ft. As this
cliff has obviously provided the scree material it is undoubtedly composed of the same phyllite. About
100 ft. above the landing beach [in another direction?] is an outcrop of a massive, dark olive-green
rock [serpentine] which has given rise to boulders on the shore.
GIBBS I.
W.N.W.
E.S.E.
NARROW I.
SCHIST DUN \TE;- SERPENTINE
Fig. 12.
Narrow Island, on its south side, appears from the sea to be composed of a massive rock of reddish
brown hue, with no sign of the schistosity which characterizes the southern face of Gibbs Island.
A landing was made on the south coast near the connecting spit, and a specimen was obtained from
the cliff face a few feet above sea level. This rock is the dunite-serpentine described below.
From the data given above a tentative sketch section may be drawn showing the probable geological
structure of Gibbs Island (Fig. 12). The view is here taken that the serpentine has been intruded
parallel to the foliation planes of the schist.
PETROGRAPHY
The rocks of Gibbs and Narrow Islands comprise two sharply contrasted types, namely, schists
and serpentine.
Schists. Five of the specimens were sliced for microscopic examination. They can be described in
general terms as chlorite-sericite-albite-schists containing, in addition, quartz, calcite, and minerals
of the epidote group (clinozoisite, zoisite) in some abundance. Small garnets and a mineral of the
chloritoid group are found in one specimen, and the latter mineral also occurs in another rock. In
hand specimens the rocks show a fine, parallel schistosity yielding flat cleavage surfaces varying in
colour from light silvery grey to lead grey.
In thin section the rock containing garnet and chloritoid shows a thin foliation with somewhat
larger grains of quartz and feldspar taking part in a minute flaser structure. The garnets are small
and sparsely distributed; chloritoid is rather more abundant, and occurs as pleochroic grey-blue
prisms with good cross fracture.
Another rock consists of a mosaic of small grains of quartz through which wind thin folia of
ELEPHANT AND CLARENCE GROUP 85
interwoven flakes of sericite, and folia made up of large crystals of green pleochroic chlorite with
' ultra-blue ' polarization colours. In some of the intervening folia of quartz are remarkable ' trails '
consisting of small euhedral crystals of zoisite, strung out as a line of separate crystals, or occurring
in small clots. Both the slide and hand specimen of this rock show that it has been permeated by
vein quartz which has separated and isolated the individual folia.
A third type is rich in epidote. It shows alternating folia consisting (i) largely of quartz with
subordinate albite and calcite, but carrying films or thin folia of chlorite and epidote, and scattered
crystals of the same two minerals, and (2) mainly of chlorite flakes interwoven with epidote grains.
Sericite may form a notable constituent of these folia, but quartz only occurs as scattered fragments.
The most feldspathic type is a comparatively coarse schist consisting of more or less rounded
grains of albite, intermingled with smaller grains of quartz and patches of calcite, forming a mosaic
through which wind streams of flakes of chlorite and sericite, together with grains of epidote and
zoisite, and interwoven folia of these minerals. The albite is fresh and water-clear and is mostly
untwinned, but a few crystals show simple twinning or the more usual albite twinning. Many of the
albites contain curving lines of inclusions of the above minerals, suggesting their growth by accretion
during shearing as in the well-known case of ' snowball ' garnets. This rock closely resembles the albite
schists of the south-western Highlands of Scotland.^
As a whole the series of schists from Gibbs Island closely resembles those of Elephant Island and
Clarence Island, especially those of Minstrel Bay, but they are coarser, somewhat more highly
metamorphosed, and do not possess the abundant carbonaceous matter of those rocks.
Dtmite-serpentine. The least altered rock and the only one that contains unaltered olivine, is the
specimen which was collected from the south coast of Narrow Island. All of the serpentine rocks
collected show signs of intense shearing. They are, in fact, serpentine-schists of apple-green and
malachite-green colours and ornamental appearance. Some of the specimens show opaque patches,
streaks and veins of a black metallic mineral which turns out to be magnetite.
The Narrow Island rock must have consisted almost entirely of olivine crystals, but it is now made
up of olivine fragments in a mesh of serpentine. The only other mineral is magnetite, a little of which
may be primary but, for the main part, is undoubtedly of secondary origin. The olivine is a highly
magnesian chrysolite with 21=90" and positive sign, and therefore with a FeO content of about 13 per
cent. About half of it has been transformed to serpentine or allied substances. The alteration proceeds
as usual along the fissures and from the peripheries of the crystals. The first effect of alteration is to
produce a pale brownish yellow uncleaved mineral which is of very low birefringence or sensibly
isotropic (delessite?), shot through with colourless fibres of positive elongation which may be chrysotile.
These areas of delessite(?) and chrysotile roughly outline the original hexagonal forms of the olivine
crystals, and enmesh fragments of them. The next stage of alteration produces colourless antigorite
in irregular sheaves of platy crystals with negative elongation, which can be seen to be growing at the
expense of the areas of delessite(?) and chrysotile, with the liberation of iron oxides in the form of
ragged grains of magnetite.
In the remaining specimens of serpentine, all from the south coast of Gibbs Island, the alteration
is completed. Not a trace of olivine is left, nor of delessite (?) and chrysotile. The whole rock consists
of antigorite in closely woven felts of plates and prisms, with irregular ragged strings of magnetite
which have sometimes segregated into definite secondary veins about i mm. thick. The shearing to
which the rocks have been subjected has caused the reformation of the antigorite along the major
lines of movement, often with a superposed cross-lamellation. With a more severe crushing stress,
1 A. Harker, Metamorphism, 1932, p. 213. The rock figured on this page (fig. 95 A) strongly recalls the microscopic
appearance of the Gibbs Island rock.
86
DISCOVERY REPORTS
however, the crystals have been ground to powder, and wind in streaks around larger fragments which
have assumed a pseudo-spherulitic form.
The dunite-serpentine of Narrow Island has been analysed by F. Herdsman, A.R.S.M., with the
results shown in Table 5, col. i. For comparison an analysis of dunite-serpentine from Cornwall is
given. The resemblance between the two analyses is obviously very close. The calculated norms of
both rocks give about 50 per cent olivine and 40 per cent enstatite. While the Cornish rock is stated
to contain some enstatite and tremolite {op. cit. p. 64) not a trace of these minerals can be found in
the dunite-serpentine of Narrow Island. It may perhaps be surmised that in the alteration to serpentine
there has been some differential abstraction of magnesia and iron oxide relative to silica. This appears
to be the first record of dunite and serpentine in the West Antarctic region.
Table 5
I
A
SiO,
41-85
40-12
AUOj
1-37
0-98
FePa
2-62
6-52
I. Dunite-serpentine, Narrow Island, West Ant-
FeO
2-l6
I-2I
arctica. Anal. F. Herdsman.
MgO
39-44
35-78
CaO
tr.
0-12
A. Dunite-serpentine, Predannack, The Lizard,
Na,0
tr.
0-24
Cornwall. Anal. E. G. Radley. Quoted from
K,0
0-13
0-08
J. S. Flett and J. B. Hill, ' The Geology of the
H,0+
11-03
12-17
Lizard and Meneage, Mem. Geo!. Siirv.,
HoO-
0-45
1-69
England and Wales, Expl. of Sh. 359, 1912,
c6.
nil
0-15
p. 79.
Tid„
tr.
tr.
p.o;
0-22
o-io
MnO
tr.
0-52
(Ni, Co)0
0-24
0-15
CuO,
0-19
0-28
V2O3
—
tr.
BaO
—
nil
FeSa
—
o-oi
99-70
100-12
CHEMICAL COMPOSITION AND ORIGIN OF THE METAMORPHIC
ROCKS OF THE ELEPHANT AND CLARENCE GROUP
No new analyses have been made of the rocks described above, since none of them has been collected
in situ or located with exactitude except a few from Gibbs Island. Four analyses, however, have been
published, two each from Elephant and Clarence Islands, and these are collected in Table 6, together
with a few comparable analyses from Tierra del Fuego, South Georgia, etc.
Prof. Tilley regards the rocks of Lookout Harbour, Elephant Island, as a ' graded series of related
sediments ranging from limestones to impure types giving the amphibolites and garnet-hornblende-
schists rich in albite '. The amphibolites are closely associated, and even interbedded, with limestone
bands. Tilley surmises that the original sediments were somewhat abnormal inasmuch as abundant
albite was present. But there is one type of sediment, quite abundant and by no means abnormal,
which is often rich in soda and often rich in albite, namely, the impure sandstones known as greywacke.
The most typical greywackes are constituents of ancient fold-mountain ranges wherein they are
often associated with mudstones, slates, greenstones, ophiolites, and especially with igneous rocks
ELEPHANT AND CLARENCE GROUP
87
Table 6
I
A
B
2
c
° 1
3
4
E
SiOa
M"M
48-63
51-56
57-66
53-56
53-75
45-10
71-80
11-87
73-°4
ALO,
16-46
14-85
17-54
16-30
19-32
18-60
14-76
10-17
0-56
FejOg
1-92
1-91
1-80
3-46
1-06
2-04
4-5°
2-21
FeO
7-41
9-47
8-28
2-46
7-44
6-97
9-87
2-30
4-15
MgO
CaO
8-64
7-93
5-23
3-95
3-43
2-30
5-95
1-94
1-43
10-19
7-20
11-42
6-01
5-21
6-98
11-59
3-02
1-49
3-56
Na^O
274
2-98
2-18
4-39
3-86
4-06
2-55
3-27
K2O
0-06
0-30
0-33
2-68
1-96
1-32
0-47
1-02
1-37
H,0+
3-38
4-09
0-34
0-98
2-29
0-76
0-26
I-29I
0-48!
2-36
H2O-
o-io
0-21
0-22
o-io
0-06
0-07
O-IO
0-84
CO2
0-21
o-i8
nil
0-12
0-20
0-49
1-38
nil
TiO,
1-20
2-34
0-56
0-85
1-02
2-83
2-51
tr.
0-16
0-15
P,0=
0-14
0-15
o-oi
tr.
0-55
0-22
tr.
0-21
0-23
0-18
MnO
0-12
0-36
o-ii
0-12
0-18
0-26
0-45
(Ni, Co)0
BaO
s
0-02
0-21
—
o-o8
0-19
o-o6
0-07
—
tr.
0-23
nil
0-08
O-IO
SO3
—
—
—
—
—
—
0-20
""
CI
nil
—
—
0-02
tr.
F
nil
—
—
nil
tr.
—
—
—
0-17
C
~
99-99
100-43
99-82
99-91
99-88
100-35
99-94
99-89
99-80
I.
B.
D.
3-
4-
E.
Chlorite-actinolite-clinozoisite-albite-schist ('very schistose'), Clarence Island. Anal. E. Kluver Quoted fram Barth
and Holmsen, op. at. supra, p. 60. This rock is briefly described as containing chlorite actinolitic hornblende, clino-
TotiieZd aibite (An,,) Calcite and quartz occurred in fissures. It is stated that the latter minerals were removed
before the analysis was made (Barth and Holmsen, p. 59). ,,ttii r^ . a tv„„, F H Kranrk
Ophiolitic greenstone, north of Monte Olivia, Ushuaia, Tierra del Fuego^ Anal. L. Lokka. ^^^^^1 °f E" f^^'^'l;
Ob cit siJm p III. This rock is stated to be an 'effusive' associated with slates and phyllites of the Yahgan (or
Mt BuckIand)'^Formation (probably Lower Mesozoic). It is sheared and mylonized m places. The freshest material
2wfoligoclase (An,,) and augite altering to hornblende. Chlorite, epidote, actmohte, sphene altering to leucoxene
quartz, and aibite, occur in the highly sheared varieties. In its geological associations and petrography this rock is
T'rlmohL^rr— ne^slones dredged near the Shag Rocks, about 130 miles west of South Georgia. New analysis
So';^SSr:;t^iSlSSXr('not very schistose'), Clarence Island. Anal. E. KlU^^ Q^-^;;^- ;S
and Holmsen oP cit supra p. 60. This rock is stated to be a ' common schistose greenstone, the constituent minerals
Jrwhich are IreenbSit, '^actinolitic hornblende, ferriferous epidote, and aibite (An«). [From the analysis it is
toleTably certain that quartz should be added to this list.] The summation of this analysis is incorrectly given as 99-82
in R-irth and Holmsen but is correctly stated in Holtedahl, op. cit. supra, p. 109.
Sheared tuff from r;;ra^e, Virik Harbour, South Georgia. Anal. E. Kluver. Quoted from Barth and Holmsen,
IplTsu^ra pTo- These uffs contain fragments of keratophyres, trachyandesites and spilites (see G^W. Tyrrell
'Petrtraphv and Geology of South Georgia, 'Quest' Report {Brit. Mus. Nat. //»/.), 1930, PP- 35-7)- This aialys^^
- ^' ^""'','?• P\ T^;t Lnf the Yahgan fMt Buckland) Formation. It shows films of chlorite and mica winding
Grevwacke fKulm) Steinbach, Frankenwald, Germany. Quoted from R. Ligenteld, uie rvuimcu g
SuSmtz in. Fnt!;kenwalde,':4M. Math.-Phys. Kl. Sachs. Akad. W,ss. xlu, no. i, i933, P- 58-
88 DISCOVERY REPORTS
of the spilitic suite. These geosynclinal greywackes are rich in fragments of intermediate, basic and
ultrabasic igneous rocks and their minerals, especially spilites and their associates.^
Spilitic lavas are of submarine or at least subaqueous origin. The greywackes formed of their debris
may be regarded as due to disintegration by submarine eruptions aided to some extent by subaqueous
gliding (Bailey),^ which distribute an enormous amount of ' greenstone ' debris, mingled with sand
and mud, far and wide over the oceanic regions affected. In its descent through the water this material
would become sorted with regard to grain size and would form graded sediments ranging from
greywacke to mudstone. This view would explain the frequent passage of greywackes to siltstones
and mudstones on the one hand, and into tuffs on the other. Furthermore, limy material lying on the
sea floor, and also the radiolarian cherts and impure limestones which are often associated with spilitic
lavas, would be incorporated in these sediments. Moreover, spilitic lavas and their tuffs are very
frequently saturated with carbonate of lime, which would reappear as calcite in the greywackes
resulting from their disintegration.
Towards the deeper parts of the oceans these sediments would merge gradually into the blue
carbonaceous and ferruginous muds appropriate to this locus ; and towards the coasts they would pass
into the terrigenous sands and muds of the continental shelves.
The greenstone-greywacke-mudstone association is generally formed during the geosynclinal stage
of the orogenic cycle, and is therefore commonly affected by the low-grade metamorphism which
ensues when the later orogenic movements take place. Slates, phyllites, and quartz-sericite-schists
are thus formed from the mudstones and siltstones; fine-grained quartzites and qviartz-schists from
cherts and other siliceous rocks; schistose grits, quartz-chlorite-albite-schists, and greenstones such
as those found in the ' Green Beds ' of the Scottish Highlands, from the greywackes and greywacke-
tuffs; epidiorites, greenstones, chlorite-schists, hornblende-schists, amphibolites, etc., with epidote,
zoisite, garnet, and other accessory minerals, from the basic igneous rocks and their tuffs. Glauco-
phane-bearing schists may be formed from the soda-rich varieties of these rocks, or from greywackes
composed of their debris.
It is precisely an assemblage of this character which is encountered in the Elephant and Clarence
Group and the South Orkneys. South Georgia, too, is composed of greywackes and greywacke-tuffs
with slates and phyllites, and an occurrence of spilitic rocks is found at the eastern end of the
island. Such an assemblage may also form the basement of Graham Land and the adjacent archipelagos.
Above all, it is represented in Tierra del Fuego by the rocks of the Yahgan or Mt. Buckland formations, and
by some of the Central Schists of that region. Since radiolarian cherts are abundantly developed here,
it is probable that the whole assemblage belongs to the geosynclinal greenstone-greywacke-mudstone
association discussed above. It is difficult to read Kranck's descriptions of the petrography of these
rocks {op. cit. supra) and not to recognize that in West Antarctica we are dealing with exactly similar
groups of sedimentary and metamorphic rocks. The bearing of these considerations in favour of the
theory of the tectonic connexion between South America and West Antarctica put forward by
H. Arctowski, O. Nordenskjold, and E. Suess, is obvious.^
1 There are, of course, types of greywacke due to the waste of areas of miscellaneous rocks, including slates, basic igneous
rocks, etc. These may be styled continental greywackes, and are strictly equivalent to arkoses, which are derived from the
waste of a granitic or gneissic terrain.
^ G. W. Tyrrell, 'Greenstones and Greywackes', C.R. Reunion Internat. pour I'etude du Precambrien et des vieilles chaines,
Finland, 193 1, pp. 24-6. E. B. Bailey, 'New Light on Sedimentation and Tectonics', Geol. Mag. lxvii, 1930, pp. 77-92.
The writer does not accept Bailey's view that greywackes are merely 'muddy sandstones'.
^ For recent discussions of this problem see G. W. Tyrrell, 'Petrography and Geology of South Georgia', 'Quest' Ex-
pedition Report {Brit. Mus. Nat. Hist.), 1930, pp. 51-4; and H. F. P. Herdman, 'Report on Soundings taken during the
Discovery Investigations, 1926-32', Discovery Reports, vi, 1932, pp. 214-19.
89
PART IV.
PETROGRAPHY OF STONES DREDGED FROM THE
VICINITY OF THE SHAG ROCKS
INTRODUCTION
One of the most remarkable geological features of the West Antarctic region is the existence of an
eastwardly-directed loop of submarine ridges and islands which connects Staten Island in Tierra
del Fuego, through the Burdwood Bank, Shag Rocks, South Georgia, Gierke Rocks, South Sandwich
Islands, the South Orkneys, and the Elephant and Glarence Group, with the Graham Land peninsula
and its adjacent archipelagos. It represents an extension of Circum-Pacific orogenic structures for more
than I GOG miles into the heart of the alien geological region of the South Atlantic. This loop or arc
has been called the Southern Antilles on the basis of a supposed analogy with the Antilles con-
necting North and South America; but a better
term is the Scotia Arc, coined by J. M. Wordie,
since the loop surrounds the Scotia Sea. The
geological constitution of the Scotia Arc is con-
sistent with the view, put forward by E. Suess
and others, that it represents an orogenic tectonic
connexion between South America and Graham
Land.i
Something is known of the geology and petro-
graphy of all the connecting links of the Scotia
Arc with the exception of the Shag Rocks. It is
fortunate therefore, that two Discovery dredgings
have been made in the vicinity of the Shag Rocks
*'=. .Shag Rocks
Fig- 13-
(see map. Fig. 13), which have provided sufficient material to enable us to assess the geological character
of the Scotia Arc in this hitherto unknown region. These dredgings were made on 12 November 1930
by the 'Discovery 11' at Sts. 474 and 475. The exact positions and depths are as follows:
St. 474. One mile west of the Shag Rocks. Depth 199 m.
St. 475. Long. 53° 30^ S., lat. 42° 44* W. (about 25 miles west of the Shag Rocks). Depth 748 m.
PETROGRAPHY
Fourteen stones came from St. 474 and five from St. 475. Of these nineteen stones, fifteen are
practically identical and consist of tremolite-epidote-greenstone or greenstone-schist, one is a
feldspathic quartzite, and three are quartz-vein rocks. The four last-named stones all came from
St. 474, nearest to the Shag Rocks. The overwhelming preponderance of the greenstones in this
collection makes it tolerably certain that this rock constitutes the Shag Rocks themselves and the
submarine ridge on which they stand to at least 25 miles to the west.
The stones range in size from 4 in. to i in. in greatest diameter. Fifteen of them, as above stated,
are ' greenstones '—dense, compact rocks of grey-green colour, showing an ill-developed cleavage
along which they tend to split. Only two are definitely slaty or phyllitic in aspect. The quartzite is
a fine-grained rock of a pale buff tint, and obviously contains much feldspar. The quartz-vein rocks
are white and coarse-grained.
1 Recent summaries of the evidence have been given by O. Holtedahl, 'On the Geology and Physiography of Some
Antarctic and Sub-Antarctic Islands', Scientific Results of the Norwegian Antarctic Expeditions 1927-8 and 1928-9, instituted
and financed by Consul Lars Christensen, No. 3, Norske Vidensk.-Akad., Oslo, 1929, pp. 104-18. G. W. Tyrrell, 'Petrography
and Geology of South Georgia', 'Quest' Exped. Report {Brit. Mus. Nat. Hist.), 1930, PP- 5i-4- H. F. P. Herdman, Report
on Soundings taken during the Discovery Investigations, 1926-32', Discovery Reports, vi, 1932, pp. 214-19.
90 DISCOVERY REPORTS
Tremolite-epidote-gree7istone. The principal minerals, as disclosed by thin sections, are tremolite,
clinozoisite-epidote, chlorite, quartz, and albite. They are arranged in bands or elongated folia parallel
to an ill-defined slaty cleavage which, in two or three of the sections, develops into a phyllitic or
schistose structure. The cleavage planes are frilled and puckered by an imperfect strain-slip. The
bands consist of one or two of the above minerals to the almost complete exclusion of the others.
Folia consisting mainly of tremolite and clinozoisite or epidote are preponderant.
The tremolite occurs as colourless to pale green fibres, needles, prisms and plates, often arranged
in parallel position or with a slightly divergent, sheaf-like structure. It has a good cross-fracture and
longitudinal cleavage, although the typical prismatic amphibole cleavage is rarely seen. The extinction
is at io~20° to the cleavage direction [c). Its elongation is positive in sign, distinguishing it from the
colourless variety of pargasite (edenite). Both epidote and clinozoisite are present. Epidote is the
most frequent associate of the tremolite. It is of yellowish brown colour, and has usually undergone
considerable alteration converting it into a greyish cloudy material (leucoxene.?). This material forms
ragged areas or, in the more highly cleaved types, it is drawn out into streaks and lines. It is possible
that some of this material may represent altered sphene. Colourless clinozoisite occurs mainly as well-
shaped crystals associated with quartz and albite in lenticles which may be partly of secondary origin.
Chlorite of the pale green variety with ultra-blue polarization occurs in irregular pods or stout
lenticles. It is not abundant and, in a few places, appears to be growing at the expense of tremolite
and epidote. Quartz, always with undulose extinction, is abundant in some lenticles and bands, and
is associated with a little untwinned or simply twinned albite. Finally, in a few of the less altered
rocks, very slender microlites of plagioclase (oligoclase.'') can be detected. Discussion of the original
character of this somewhat unusual greenstone is deferred to the section dealing with its chemical
composition (p. 91).
Qiiartzite. This is a hard, yellowish, well-cemented sandstone or semi-quartzite. In thin section
it is found to consist mainly of quartz and feldspars (plus alteration products) in roughly equal
proportions. All the grains are angular and fit together like the stones in macadam. Only a few of the
quartz grains show undulose extinction. The feldspar is easily distinguished by its turbid appearance.
It includes soda-orthoclase and albite in about equal proportions. Many of the grains are com-
paratively fresh despite their turbidity, but others are completely altered to sericite and crystalline
kaolinite. These alteration products have insinuated themselves into fissures in the quartz grains
and between the grains, thus acting as a cement which has filled all open spaces. In addition to
quartz and feldspar there are a few small grains of epidote, sphene, and iron ores, and rather more
abundant fragments of what appears to be the ground-mass of dense acid igneous rocks like felsite
or rhyolite. In fact, the mineral composition of the rock suggests that it may have been derived from
the waste of rocks like the quartz-feldspar-porphyries which constitute the major part of a great
Porphyry Formation in Patagonia and Tierra del Fuego, and are also found in parts of West Antarctica
(see this Memoir, p. 75). The rock may thus be described as quartzitic arkose.
Quartz-vein rocks. These are all mainly composed of white quartz with films of a chloritic mineral.
In thin section one of them shows quartz, albite, chlorite, and a little calcite, all intensely sheared and
crushed. The quartz has marked undulose extinction and in the albite the twinning lamellae are bent
and twisted. The chlorite is greyish green, and shows the common ultra-blue polarization; it is
occasionally quite isotropic.
A second rock consists of intensely sheared and sliced quartz with some large crystals of greenish
brown epidote. While clearly later than the quartz, the epidote crystals have also been bent and sliced
by a movement in a diff'erent direction to that which first affected the quartz. The resulting fissures
have been healed by the infiltration of silica. No albite or chlorite occurs in this rock.
SHAG ROCKS 9i
A third quartz-vein rock consists of sheared quartz with films and foHa of almost colourless,
isotropic chlorite.
There is no evidence of the nature of the rocks penetrated by these veins. While there appears to
be secondary quartz in the greenstones, there are no sharply defined veins. However, from their
mineral composition and associations, it is likely that the quartz veins cut rocks of metamorphic type.
CHEMICAL COMPOSITION OF THE GREENSTONE
A composite sample from three of the least altered greenstones was analysed with the result shown
in Table 7, col. i. This analysis has a characteristically basaltic pattern with its comparatively high
lime and alumina which, in the rock itself, is accounted for by the abundance of epidote and tremolite,
and in comparable basahs, by richness in lime-plagioclase. The analysis is, for example, much like
that of the Porphyritic Central Basalt type of Mull (Table 7, col. A), and like the basalt of the South
Shetland Islands (Table 7, col. B). The latter, however, has a much higher k ratio than the Shag
Rocks greenstone.
Table 7
I
A
B
C
SiOg
51-56
48-51
48-26
47-37
AUO3
17-54
19-44
17-42
16-46
FeaOg
I -So
5-66
3-36
1-92
FeO
8-28
4-00
5-61
7-41
MgO
5-23
5-12
8-83
8-64
CaO
11-42
12-03
11-56
10-19
Na.,0
2-l8
2-53
2-44
2-74
K.,6
0-33
0-25
0-89
0-06
H.,0^-
0-34
0-48
0-24
3-38
H.O-
0-22
0-04
0-16
o-io
CO.,
nil
0-09
nil
0-21
Tid..
0-56
1-46
1-07
1-20
P2O5
tr.
0-16
0-22
0-14
MnO
0-36
0-23
0-14
0-15
(Ni, Co)0
—
0-04
—
—
S
—
—
—
0-02
CI
—
—
nil
F
—
—
—
nil
99-82
100-04
100-20
99-99
I. Tremolite-epidote-greenstone, stones dredged near the Shag Rocks, 130 miles west of South Georgia. Anal.
F. Herdsman. rr, ■ , n
A. Porphyritic basalt (Porphyritic Central Type), Mull. Anal. E. G. Radley. Quoted from 'The Tertiary and Post-
Tertiaiy Geology of Mull', Mem. Geol. Surv., Scotland, 1924, p. 24.
B. Olivine-basalt (Recent), Penguin Island, King George Island, South Shetlands. Anal. F. Herdsman. See this
Memoir, p. 59. • at ■
C. Chlorite-actinolite-clinozoisite-albite-schist, Clarence Island. Anal. E. Kluver. See this Memoir, p. 87.
The West Antarctic rock to which the Shag Rocks greenstone shows most resemblance is the schist
from Clarence Island (Table 7, col. C). There is obviously a close mineralogical similarity, and the
chemical analyses have the same pattern, although SiO^ is lower and (Fe, Mg)0 higher, in the
Clarence Island rock. The latter, however, is of more advanced metamorphic grade than the greenstone
of the Shag Rocks. From Tierra del Fuego, Kranck {op. cit. supra, pp. 43, 47, 54, no) has described
several ophiolitic greenstones, greenstone-schists, prasinites, etc., containing chlorite, epidote,
actinolite, sphene, leucoxene, and albite, but the only analysis given of these rocks (cited in Table 6,
col. A) does not accord very closely with that of the Shag Rocks greenstone.
The chemical affinities of this rock clearly accord with those of a common type of basalt, and it may
g2, DISCOVERY REPORTS
be regarded as due to low-grade metamorphism of basaltic rocks of this type. Its association with
quartz veins, and with a quartzite-like rock, and its chemical and mineralogical similarity to the
prasinitic schists of Tierra del Fuego and Clarence Island, make it congruous with the whole
assemblage of rock types found in the Scotia Arc, and adds confirmatory evidence for the theory of
tectonic connexion between South America and West Antarctica favoured by E. Suess and other
writers (see this Memoir, p. 89).
PARTY. PETROGRAPHY OF THE SOUTH SANDWICH ISLANDS
INTRODUCTION
The volcanic South Sandwich Islands are situated at the extreme eastern end of the Scotia Arc, and
form either a part of it, or a volcanic arc parallel to
and in echelon with it. They are fully described in
a recent publication to which reference will be fre-
quently made in the ensuing pages. ^ It is proposed
in this paper to summarize and collate the already
published petrographic data, and to supplement
them with descriptions of new material from five
localities, viz. material collected in situ by Mr G.
Rayner on Saunders Island, and dredgings from
four stations: (i) St. 363, 2-|- miles S. 80" E. of the
south-eastern point of Zavodovski Island, (2) St.
366, off the south coast of Cook Island, (3) St. 368,
in Douglas Strait between Cook Island and Thule
Island, (4) St. 370, 2 miles north-east of Bristol
Island. In addition, there are some stones collected
from a piece of floating ice near Bristol Island.
Many observations on the volcanology and on
the rocks of the South Sandwich Islands as seen
from a distance are published in the above memoir,
but the only petrographic data so far published are
to be found in the following three papers:
(i) O. Backstrom. ' Petrographische Beschreibung
einiger Basalte von Patagonien, Westantarktika, und
den Siid-Sandwich Inseln', Bull. Geol. Inst. Upsala,
XIII, pp. 115-82 (1915). South Sandwich Islands,
pp. 163-76.
(2) G. V. Douglas and W. Campbell Smith.
' Zavodovski Island, and Notes on Rock Fragments
dredged in the Weddell Sea ', ' Quest ' Exped. Report
{Brit. Mus. Nat. Hist.), pp. 63-7 (1930).
(3) G. W. Tyrrell. ' Report on Rock Specimens
from Thule Island, South Sandwich Islands',
South Sandzvich Islands Memoir, pp. 191-7 (193 1).
1 Stanley Kemp and A. L. Nelson, 'The South Sandwich Islands', Discovery Reports, in, pp. 133-98, pis. xi-xxxi (1931).
Hereinafter referred to as South Sandwicti Islands Memoir.
4IZAV0D0VSK
1.
t LESKOU I.
5OK0I I.
57'
VINOICATIO
^ 1 ,% CANOLtHAS 1,
^' SAUNDERS 1.
58-
^ MONIAOU 1.
5*
FRCEZEL
^....■■♦e«lSTOLI.
THULE I <^*
SEUINGSHAUSCN 1.
OOK 1.
29- W
. .^r '
7- 2j6°
. , , ; . T^ 1 . r
M"
Fig. 14. The South Sandwich Island.
SOUTH SANDWICH ISLANDS 93
PETROGRAPHY
General. The South Sandwich Islands, so far as present observations go, are composed exclusively
of Recent volcanic rocks, the products of present-day and recently extinct volcanoes. Five of the
islands, Zavodovski, Candlemas, Bellingshausen, Saunders, and Visokoi (map, Fig. 14), show definite
signs of volcanic activity and emit vapour and fumes; another three, Leskov, Vindication, and
Montagu, show no activity at present, although large areas of ice- and snow-free ground, indicating
residual warmth, exist on the islands. The remaining islands, Bristol, Cook, and Thule, are heavily
glaciated, and show no signs of volcanic activity or warm ground.^
The South Sandwich Islands Memoir (p. 150) states that the rocks consist mainly of 'reddish tuff
and black basaltic lava ', and this is supported by the petrological examination of the collected rocks.
The 'reddish tuif' may include reddened slags, and the lavas, while mainly basaltic, include basic
andesites and even more acid types such as dacite. A possible exception to this generalization is
Freezeland Peak, a small islet to the west of Bristol Island, which is referred to later (p. 99). It is
proposed to describe the petrography of each of the islands in turn, starting from the northern end
of the chain.
Zavodovski Island (South Sandwich Islands Memoir, pp. 156-60). This island is nearly circular in
outline and is 9 miles in circumference. It consists mainly of an active volcanic cone which rises
from a lowlying plateau of black basaltic lava most conspicuous on the eastern side of the island.
There are subsidiary craters on the slopes of the main cone, and to the south of West Bluff there are
fumaroles in reddish ground with some patches and streaks of sulphur. At one point horizontal strata
apparently consist of alternate beds of ash and tuff.
In 1908 the Norwegian, Capt. C. A. Larsen, landed on many of the South Sandwich Islands and
collected rock specimens of which, unfortunately, some were lost by accident. The collection was
presented to Goteborg Museum, and was later described by O. Backstrom (i). Larsen landed
at the north-west end of Zavodovski Island, which was found to consist of a porous lava carrying
zeolites in the vesicles. These specimens were lost by the upsetting of the boat. Only a few small
pebbles and lapilli were retained, which Backstrom identified as olivine-basalts and their tuflrs. The
fragments of which the latter were composed showed fresh phenocrysts [feldspars?] in a ground-mass
which had been altered by the action of solfataric gases. Their richness in phenocrysts and in shattered
basaltic ground-mass material showed that they represented a common type of 'Aschentuff' which
was probably rather glassy.
During the Quest Expedition of 192 1 G. V. Douglas saw the island at close range although he was
unable to land (2). He does not state from which direction the 'Quest' approached the island, but
from the fact that he mentions a cliflF 40 ft. high with a long gentle slope inland, it may be assumed
that he saw the low plateau on the eastern side. Douglas states that : ' The lava flows seen on the cliff
face appeared to consist of a compact columnar basalt at the base. Above, there was a line of red
cinder, and above this again what looked to be rough paehoehoe lava.'
Material obtained by dredging at 19 fm. corresponded with the above-described section. The
sample consisted of rounded black pellets of diameters between i and 5 mm. Twenty of these were
sectioned, and ten of them were found to consist of dense black glassy basalts free from olivine. Some
were crowded with minute laths of plagioclase; others showed a few small phenocrysts of plagioclase
and augite. Four of the pellets consisted of dense, dark brown, glassy olivine-basalts, some containing
many crystals of plagioclase and only a few of olivine and augite. Four others were paler basalts of
holocrystalline-porphyritic texture with small phenocrysts of plagioclase and sometimes augite in an
1 South Sandmch Islands Memoir, pp. 151-2.
94 DISCOVERY REPORTS
intergranular ground-mass consisting of minute microlites of feldspar and grains of augite and
magnetite. Tlie two remaining pellets consisted of basalt glass of a deep olive-buff colour. In one
of these microlites were absent, but in the other microlites of plagioclase and augite were abundant,
and a little olivine was probably present.
The Discovery II material submitted to me was dredged at St. 363 from depths between 278 and
329 fm. at a locality 2| miles S. 80° E. of the south-eastern point of Zavodovski Island. It consisted
of two bags, one containing grey scoria or lapilli, very rough and angular, the largest being about
I in. in greatest diameter; the other contained a few of the larger stones picked out from the scoria.
Five thin sections were prepared from this material.
The scoria and lapilli consist of a highly vesicular, opaque, pumiceous glassy basalt. The glass
varies in colour from black, even in thin section, to pale brown, and carries minute microlites of
plagioclase and pyroxene, the latter being noticeably more abundant in the pale brown glass. A few
large crystals are entangled in the glassy sponge; these include plagioclase (bytownite, Angj), pale
brownish green diopsidic augite, and olivine, all perfectly fresh. In one specimen the glass is much
haematitized, and carries much larger and more numerous feldspar microlites which can be identified
as labradorite (Anjo). These rocks are on the borderline between andesites and basalts. Their content
of olivine is small and sporadic; and as the glassy ground-mass probably contains much free
silica it may be presumed that if the magma had not been so rapidly quenched the olivine would
have been made over into pyroxene by reaction, and the rock would then have been revealed as
a basic andesite. This description agrees with that of the dredged material off Zavodovski given by
Douglas (p. 93).
One of the dredged stones, however, the largest, is undoubtedly a sedimentary rock. It is a very
dense, dark grey material which looks like cementstone. In thin section it shows a carbonate mineral
intermingled with argillaceous matter. The rock effervesces only when powdered and treated with hot
concentrated acid, and may therefore be identified as a dolomitic mudstone.
Leskov Island {South Smidwich Islands Memoir, pp. 161-2). This island, the smallest of the South
Sandwich Group, lies some distance to the west of the arc on which all the other islands are situated.
Its circumference measures only about i| miles. There is no record of any landing on this island, but
it was observed at close range by Capt. Larsen ((i), p. 166), Lt. Filchner,^ and by members of the
Discovery II party. The last-named state that the island is crescentic in outline and is doubtless
a fragment of a volcanic cone. Material dredged by Larsen at a depth of 75 fm. proved to consist of
basaltic rocks ((i), p. 167). At the south-eastern corner of the island a conspicuous conical rock
consists of columnar basalt; the cliffs round the southern and western sides are formed of rugged
flows of basaltic lava inclined towards the sea on the south side at an angle of 45°, but gradually
becoming vertical towards the west. The rock walls of Crater Bay are reddish and yellowish in colour
and apparently consist of tuff which shows no definite bedding and is much contorted {South Sandwich
Islands Memoir, p. 162).
Visokoi Island {South Sandzvich Islands Memoir, pp. 162-5). This island is one of those that show
definite volcanic activity. There is no known record of a landing and most of the information regarding
Visokoi was obtained during the visit of 'Discovery II'. The only geological information available is
that provided by a sketch of rock exposures on the north coast by Mr F. C. Fraser {South Sandwich
Islands Memoir, fig. 8, p. 164), which shows columnar basalt, dark grey rock intersected by dikes and
surmounted by light grey stratified rock [tuff?], reddish and grey rocks cut by dikes, and an exposure
of stratified rocks [tuffs?] in alternate layers of grey and red tints. The general impression was that
the rocks were basaltic lavas and tuffs similar to those seen on Zavodovski.
^ Zum Sechsten Erdteil, pp. 1 14-15, figs. 32-6 (Berlin, 1923).
SOUTH SANDWICH ISLANDS 95
Candlemas Group {South Sandwich Islands Memoir, pp. 165-72). This group consists of Candlemas
Island itself, and a smaller one to the west which is now called Vindication Island. A full account of
the geography and volcanic phenomena is given in the Memoir. A large collection of rock specimens
from the southernmost point of Candlemas Island, made by Capt. Larsen, has been described by
Backstrom in the following terms ((i), pp. 169-70, translated):
[The rocks] are mostly reddish and porphyritic with rounded feldspars which sometimes give an almost white
colour to the specimens. Under the microscope they are found to be extraordinarily rich in feldspar of composition
Ans5, which is zoned with glassy inclusions and shows both albite and pericline twinning. The main pyroxene is
hypcrsthene which is often invested by monoclinic pyroxene, but both pyroxenes may occur as independent cr\'stals.
The augite shows the usual polysynthetic twinning, which is also seen in the investments around the hypersthenes.
Strongly corroded olivine also occurs but is not common. It is mostly altered to a blackish brown dust, but all the
other constituents are fresh. In regard to the systematic position of the rocks, their richness in plagioclase suggests
that they represent a transition between the basalts and the andesites. It is difficult to assign some of the rocks to
either group, but others which are richer in olivine and pyroxenes should be relegated to the basalts.
Another type has an extremely fine-grained but holocrystalline texture. It is, however, little different to the above
in mineral composition. Its plagioclase is lath-shaped not equidimensional, its pyroxene is sharply euhedral, and
olivine is absent.
Fragmental rocks also occur as very fresh, reddish brown, sandy tuffs which consist of lapilli of hazel-nut size.
The latter consist of vesicular lavas with a glassy ground-mass full of crystallites, and carrying numerous crystals
of plagioclase, augite, and hypersthene.
It will be seen how closely comparable these lavas and tuffs are to those of Zavodovski Island and
Saunders Island (p. 96).
Members of the Discovery II party were not able to land on Candlemas Island, but they made
numerous observations at close range, noting rugged flows of black basaltic lava in the northern
plateau often showing columnar structure {South Sandwich Islands Memoir, pi. xvii, fig. 3). Mr F. C.
Fraser has also provided an excellent sketch of rock exposures on the east coast {ibid. fig. 12, p. 169)
showing what are obviously stratified tuffs and a coarse agglomerate.
It was found impossible to land on Vindication Island, but the geological structure of the island
was well seen in a sheer cliff face on its north-western side. The rocks here consist of irregular masses
of red and brown colours, presumably tuffs, cut by dikes of grey rock which run obliquely, vertically,
and sometimes horizontally, not infrequently intersecting one another. Two islets, Cook Rock and
Trousers Rock, both of which are tunnelled by wave erosion, show horizontal strata of red tuff and
hard grey rock.
Saunders Island {South Sandzvich Islands Memoir, pp. 172-4). Saunders Island, with a circumference
of 17 miles, is one of the largest of the group, and is, perhaps, the best known geologically. At its
centre is the glaciated but actively volcanic cone of Mt Michael (2640 ft.). The south-eastern part of
the island is composed of bare hills (700-800 ft.) apparently consisting of loose ash or volcanic mud,
and with several extinct craters. A very fine photograph of a half-section of a crater on the south
coast is given in pi. xx, figs. 2 and 3, of the Memoir. The northern part of the island is a low plateau.
All the rock exposures show that the basement of the island consists of columnar basalts similar to
those of Candlemas and Zavodovski.
Capt. Larsen landed with difficulty on the south-eastern coast ((i), p. 170), and Backstrom
describes the rocks collected here as, in the main, different from the type common in the South
Sandwich Islands in being very dense and non-porphyritic. Under the microscope these rocks show
a well-developed fluidal structure delineated by the alinement of the minute feldspar laths in the
direction of flow. The mineral composition is plagioclase (An^s-gs), almost colourless pyroxene in
rounded grains which belongs to the enstatite-augite series, and magnetite. This rock is free from
8-2
96 DISCOVERY REPORTS
olivine, and a little analcite was found in one of the thin sections. A chemical analysis of the principal
type, free from analcite, is published, which is set out with others from the South Sandwich Islands
in Table 8 (p. loi) of this memoir. Biickstrom calls the rock a basalt.
Owing to unfavourable conditions the Discovery II party was unable to land on Saunders Island,
but on 28 November 1937, Mr G. Rayner was able to get ashore for a few hours from the 'William
Scoresby'. He made some geological observations and collected a small number of rock specimens
which are described below. The observations that follow are condensed from his MS. report.
Mr Rayner landed near the penguin rookery on the south side of Cordelia Bay (see Chart in the
South Sandwich Islands Memoir, pi. xix). The beach material consisted of a loose black volcanic ash,
the size of coarse sand or grit. Behind a low cliff of compressed snow heavily loaded with the same
ash was a level area extending back to the hills. This platform consisted of a loose ash-like material
to a depth of some inches, with occasional small boulders up to 18 in. in diameter of a heavy dark
basaltic rock resting upon it.
From this point Mr Rayner walked along the shore eastward until he reached the first outcrop of
hard rock which forms the basement of the Nattriss peninsula. Here he ascended the hill to the south
near the point marked 800 on the Chart. On its northern slopes there were several outcrops of a soft
volcanic mudstone with a sub-horizontal stratification, standing up as buttresses and ridges between
steep-sided ravines. Mr Rayner thus gained a ridge which sloped eastward to Nattriss Point. The
higher parts of this ridge still consisted of the stratified mudstone, which was undergoing extremely
rapid atmospheric erosion. At one place he encountered a remarkable pillar 15-20 ft. high carved
out of the soft material (' The Beacon '). Elsewhere along the ridge a light, vesicular, reddish, scoriaceous
rock was found.
Descending eastward towards Nattriss Point Mr Rayner found that the rock became coarser in
texture, and took on the appearance of volcanic tuff, light buff in colour, in which many large fragments
of rock were embedded. This series of coarse tuffs rested on the roughly horizontal platform of dark,
vesicular, basaltic rock of which Nattriss Point is composed. This rock falls in sheer cliffs to the sea
and has a columnar appearance owing to wave erosion along vertical joints.
With, as the writer thinks, considerable probability, Mr Rayner concludes that ' a volcanic explosion
has occurred at no very distant date, possibly from the crater to be seen to the south-west of our
landing-place, and near the junction of the ice-covered main part of the island and the earthy region
explored. This explosion has thrown up the clastic material forming the hill now resting on a hori-
zontal table of rock of which Nattriss Point is the visible part. The finest material would be the last
to settle, and this has formed the upper strata of soft mudstones seen in the fast dwindling ridges and
buttresses along the hillside and in the pillar at the summit.'
Six thin sections were made from the specimens collected by Mr Rayner. The lava which forms
the basement of the Nattriss peninsula is a black, highly vesicular rock which, in thin section, shows
an abundant ground-mass of minute microlites of plagioclase with granules of augite and magnetite,
within which is set a generation of somewhat larger feldspar laths, and finally a few micro-phenocrysts
of feldspar and yellowish augite. Owing to their small size it is difficult to make out the composition
of the plagioclase microlites of the ground-mass, but they give extinctions up to about 15'' indicating
a composition Aug,,. The larger microlites and micro-phenocrysts are highly zonal, and their com-
position ranges about Augs, which is the composition ascertained by Backstrom. The pyroxene, too,
is zonal, as shown by an undulatory extinction. It is a pale yellow variety of moderate double refraction,
and is probably, as Backstrom surmises, a member of the enstatite-augite series. The larger feldspars
and pyroxenes, while occurring independently, are often aggregated into clots of which the feldspar
forms the greater part, and the microlites of the ground-mass are stream-lined around these clots.
I
SOUTH SANDWICH ISLANDS 97
Olivine does not occur in this type which, owing to its feldspathic composition, would be better termed
andesite than basalt.
Another specimen was taken from what appeared to be an inclusion within the above-described
lava It is not so dark in colour, but the thin section shows that it is the same lava with, however, a
somewhat finer grain and a few sporadic olivine crystals, most of which are altered to green serpentine.
This rock is probably a portion of the same lava, but consolidated slightly earlier than the main mass
of the flow, and thus retaining a few of the early crystallized olivines. It may have been carried as
a solidified lump of slag on the surface of the moving flow, and have been incorporated in it by
over-rolling. . . , ■ n
The coarse agglomeratic tuff which overlies the lava basement of the Nattnss peninsula is a well-
consolidated material of light buff colour containing numerous fragments of gravel size. In thin
section it proves to be a coarse lithic tuff consisting mainly of large angular fragments of the lavas
embedded in a matrix of smaller fragments and broken crystals. The lava fragments are vesicular
andesitic basalts of the same type as that described above, but they show every gradation of texture
from purely glassy to holocrystalline-micro-granular. The broken crystals include plagioclase, augite,
and fresh olivine. Conspicuous among the rock fragments are glasses of a bright green colour. An
isotropic or very feebly birefringent zeolite with cubic cleavage forms a scanty cement in some parts
of the sUde This may be the analcite recorded by Backstrom ((i), p. 171)- This rock must have
been formed by an explosion in or under a fully consolidated lava, and it may be suggested that it
was produced by renewed activity in a nearby volcano which had been temporarily sealed by a plug
of solidified lava. r , m .. •
The volcanic mudstone which overlies the lithic tuff and forms the higher parts of the Nattnss
peninsula, in contrast to the lithic tuff, is a vitric ash consisting almost entirely of small angular
fragments of clear brown glass. The only other constituents are a few small fragments of feldspar,
aueite and magnetite. This was undoubtedly formed by explosions within a still liquid lava. Hence
the sequence of events pictured by Mr Rayner (p. 96) must be slightly amended. The vitric ash does
not represent the finer, and the lithic tuff the coarser, material derived from one and the same
explosion- but the lithic tuff probably represents the disintegration by explosion ot a solidified plug,
and the vitric ash a subsequent explosion within the liquid lava that welled up into the crater
The coarse black sand at the landing-place in Cordelia Bay consists of angular fragments of brown
glass often blackened with separated magnetite, and crystals, in about equal proportions. The crystals
fnclude plagioclase, augite, and olivine, the last-named being rather more abundant than usual This
material may have been formed by explosion in an olivine-basalt magma within whic4i while still
liquid crystallization had advanced to a considerable extent. Examination of a small pebbe enclosed
in the sample bears out this diagnosis. It is an olivine-basalt with large phenocrysts of labradorite
(An ) abundant fresh yellowish olivine, and some magnetite, in a very dense ground-mass consisting
of augite granules and feldspar microlites, in which the augite is decidedly predominant.
MontcZ Island {South Sandwich Islands Memoir, pp. 174-6). Montagu is the largest island ot he
group with a circumference of about 24 miles, and is one of the least well known. It contains wha
fs probably the highest summit of the group, Mt Belinda (4500 ft.), almost certainly an ext.nc
volcano. Montagu is the most heavily glaciated island of the arc, and has fewest signs of residual
warmth in the shape of areas free from snow and ice. • r , • ^u^
The Discovery II party did not land on the island, but they had the opportunity of making the
following observations on the rock exposures as seen from a distance:
As on other islands the lowest strata seen in rock exposures are usually of black bas^alt, often columnar in structure,
and itTs ol blalt that the outlying rocks are formed. Above it red and yellowish tut^s with some hard grey rock are
98 DISCOVERY REPORTS
to be found. At several points the rocks are clearly stratified, showing three or more horizontal layers of dark grey
rock separated by narrow bands of red tuff. Sometimes yellow tuff with red inclusions was to be seen and frequently
the rocks were much contorted and intersected by dykes. At the north-eastern corner of the island are low cliffs
formed of a light grey rock, perhaps volcanic ash. {South Sandwich Islands Memoir, p. 175.)
Capt. Larsen landed at the south-eastern corner and mentions a crater here, as well as at the
north-eastern point of the island. Biickstrom ((i), p. 175) described the rocks collected as rather
uniform types of vesicular olivine-basalts in which phenocrysts of olivine, augite, and plagioclase
(Ansa) predominate over the ground-mass. The ground-mass consists of small granules of pyroxene,
laths of plagioclase, and some magnetite. The resemblance of these rocks to the olivine-basalt of
Saunders Island (p. 97) is obvious.
Bristol Island (South Sandwich Islands Memoir, pp. 176-8). Bristol Island is an irregular oval in
shape and has a circumference of 14 miles. The highest point is Mt Darnley (3600 ft.). Its profile
seen from the north has the shape of a horse-shoe, and is conjectured to represent part of the rim of
a crater. Bristol Island is heavily glaciated and the Discovery II party were satisfied that all volcanic
activity had ceased. Three rocky islets, Grindle Rock, Wilson Rock, and Freezeland Peak, stand in
line off the western coast of the island.
Capt. Larsen landed on the north-eastern side of the island^ and collected some rock specimens.
Biickstrom ((i), pp. 175-6) describes them as of reddish grey tints, and as showing numerous small
crystals of feldspar. In thin section numerous micro-phenocrysts of zonal plagioclase are disclosed,
of composition An75_85 . Pyroxene is confined mainly to the ground-mass and belongs to the enstatite-
augite series. Olivine is only sparingly present. A photomicrograph of this andesitic basalt type is
given by Backstrom ((i), fig. 20, p. 176). It conforms closely to the main type of lava erupted from the
South Sandwich Islands volcanoes.
Although no landing was made, the geological observations made by the Discovery II party (South
Sandwich Islands Memoir, p. 177) are important and must be quoted in full:
The rocks on Bristol are similar to those on the other islands. At Fryer Point black basaltic lava is to be seen and
the rock exposures on the bluff on the south side, at the western headland and in other parts, are of yellowish and
red tuff, or tuff conglomerate, sometimes stratified with a grey rock interposed between the layers, but frequently
much contorted and with many intrusive dykes.
From a geological point of view the three large outlying rocks appear to be more interesting than any other place
in the entire group of islands. . . . The great pillar on Freezeland is composed of a pale brown rock of a kind not seen
elsewhere. It showed distinct signs of bedding and in the upper part of the column some broad reddish bands.
We believe this may be a sedimentary rock. The eastern part of Freezeland, forming the lesser of the two summits,
is different ; it is formed of a brownish rock, with vertical fissures and striation, and may be metamorphic. Wilson
Rock, nearer the mainland, is a vast mass of black columnar basalt, while Grindle Rock repeats the reddish and
yellowish tuff's seen on the adjacent headland of the island. Thus, if our conjectures are correct, the whole succession
of rock formations in the Sandwich group is to be found in these three islets. Freezeland shows the only likely exposure
of the underlying sedimentary series that we know to exist, Wilson is of the overlying basalt, here seen in far greater
thickness than elsewhere, while Grindle is formed of the superposed tuffs which are characteristic of all the islands
of the group.
Among the material from the South Sandwich Islands submitted to the author there were specimens
from near Bristol Island. One of these was a bag of scoria and lapilli dredged from St. 370 at a point
two miles north-east of Bristol Island, and a bag of small stones, including lapilli, which were picked
off a piece of floating ice near the island.
A thin section of the dredged scoria from St. 370 shows that it is a sponge of opaque black glass
with minute microlites of feldspar and augite, and a few micro-phenocrysts of plagioclase (Augo)
' The position of the landing-place is mentioned in Biickstrom's memoir ((1), p. 175). Cf. South Sandwich Islands
Memoir, p. 178.
SOUTH SANDWICH ISLANDS 09
entangled in it. This seems to represent an extremely vitreous phase of the andesitic basalt lava
described by Backstrom, and carries the same lime-rich feldspar.
Most of the smaller fragments recovered from the piece of floating ice answer to the above de-
scription. A larger stone, however, is 2 in. in length and presents a microscopic appearance very
similar to that of the ' feldspathic basalt ' described and figured by Backstrom. It shows very numerous
micro-phenocrysts of plagioclase (An^j-g,,) with subordinate augite and olivine, in a dark glassy
ground-mass carrying microlites of feldspar and augite. All the phenocrysts are perfectly fresh and
euhedral. The augite is a yellowish, slightly-pleochroic variety belonging to the enstatite-augite series.
In this rock the olivine is much more abundant than in Backstrom's material, and it must be regarded
as an olivine-basalt.
Two other stones are interesting, as they are non-igneous. One is a fragment from a quartz-vein
rock, and the other is an epidote-biotite-gneiss. In thin section the latter shows a coarse mosaic of
quartz and oithoclase alternating with folia consisting of straggling crystals of bright yellow biotite
and epidote (with some clinozoisite). There is also a little ilmenite altering to sphene, and a few
fragments of deep green pleochroic hornblende. It is not possible to say whether this is an orthogneiss
or a paragneiss. The mineral composition favours the orthogneiss interpretation, but an arkose would
yield this type of gneiss on metamorphism.
The label attached to the material from floating ice does not state on which side of Bristol Island
it was recovered. As the metamorphic fragments were closely associated with scoria which indubitably
came from Bristol Island, it seems probable that they too were derived from that locality. It is possible
that the metamorphic pebbles came from Freezeland Peak which the Discovery II party believed to
consist of sedimentary and metamorphic rocks.
Southern Tlmle Group (South Sandwich Islands Memoir, pp. 178-89). This group consists of three
islands, Thule, Cook, and Bellingshausen, in order from west to east. Of these, Cook Island is the
largest, having a circumference of 9^ miles; Thule, the next largest, is more embayed than Cook
and has a coastline of 10 miles; Bellingshausen, the smallest, is only i^ miles wide.
Bellingshausen is still an active volcano, as shown by the steam and vapour rising from it, and by
the admirable sketches of Lt.-Cmdr. J. Irving (South Sandwich Islands Memoir, fig. 19, p. 184).
Cook and Thule, however, are buried beneath thick ice caps and there are no signs of present volcanic
activity. Nevertheless, soundings in Douglas Strait between Thule Island and Cook Island have
disclosed a steep-sided basin of elliptical shape and more than 400 fm. in depth. At the north and
south entrances to Douglas Strait the depths are less than 20 fm. This has been interpreted, correctly
in the writer's opinion, as the inundated crater of a volcano of which Thule Island and Cook Island
are the remnants. This view is reinforced by the parallelism of the eastern embayment of Thule
Island, and the western embayment of Cook Island, with the adjacent contours of the submerged
basin (South Sandwich Ishmds Memoir, fig. 16, p. 179), and by the photograph of the eastern side
of Thule Island (ibid. pi. xxx, fig. 4), which shows bedded lavas and ashes dipping westward and
outward from the Douglas Strait crater.
Of the geological constitution of the Southern Thule Group little is known. On Bellingshausen
the Discovery II party noted, as on other islands, black columnar basalt with overlying agglomerate,
tuff, and ashes.
Cook Island (South Sandwich Islands Memoir, pp. 185-6). Rock faces are exposed in the cliffs
bordering Douglas Strait. They are described as of yellow, red, or brown colours, sometimes showing
signs of bedding but always much crumpled and contorted, and seamed with dikes of grey rock.
Large, apparently intrusive, masses of brown rock showing a vertical striation were also seen.
Fortunately, however, some stones were dredged by 'Discovery 11' at St. 366, 4 cables south of
:oo DISCOVERY REPORTS
Cook Island at depths between 155 and 322 m., and a few small fragments of rock at St. 368 in
Douglas Strait, i mile north of the Twitcher Rock, dredged from a depth of 653 m. near the bottom
of the great submerged crater. These pebbles, which consist mainly of slaggy and vesicular lavas,
one or two being well rounded, range in size from about 2 in. in greatest diameter down to about | in.
Sixteen of these stones were sectioned for microscopical examination. All of them were found to
be textural variants of an olivine-basalt lava. Nearly holocrystalline varieties are grey and compact,
and the glassy types black, vesicular, and slaggy, in hand specimens. In thin section these rocks are
found to be highly porphyritic, carrying very numerous small phenocrysts of plagioclase, augite, and
olivine, in a ground-mass consisting, when holocrystalline, of minute crystals of plagioclase, augite,
and magnetite. In the more slaggy varieties the ground-mass becomes richer in dark glass and the
number of microlites diminishes. In fact a complete passage from holocrystalline to a purely glassy
ground-mass can be traced.
The plagioclase phenocrysts are generally most numerous, with augite and olivine following in
that order; but in a few rocks the olivine almost rivals the feldspar in abundance. The plagioclase is
both chemically and mechanically zoned and shows complex twinning ; its composition ranges between
An7o and Augs . The pyroxene is again the yellowish, slightly pleochroic variety of the enstatite-augite
series. The olivine is perfectly fresh and often euhedral, especially in the more glassy varieties of the
rock. It gives a dead straight isogyre and therefore contains about 13 per cent of the fayalite molecule.
This type is an olivine-basalt which compares closely with that from Bristol Island (p. 98), and
with the younger basalts of the South Shetland Islands (e.g. Penguin Island, p. 46). A chemical
analysis of one of the more holocrystalline types is recorded in Table 8 (p. 10 1).
Thule Island {South Sandwich Islands Memoir, pp. 187-9). The south-eastern plateau of Thule
Island appears to be composed of the usual black columnar basalt. Near Cape Flannery on the
west coast are beds apparently composed of yellowish tuff and ash, and farther north the rocks are
definitely stratified, three layers of ash separated by red tuff overlying black basalt.
A landing was made by the Discovery II party on Beach Point at the north-eastern corner of the
island. The ridge at Beach Point is composed of hard grey rock with outcrops of red tuff and a soft,
crumbling, black rock, perhaps volcanic ash, at its summit. The steep cliffs facing Douglas Strait
show contorted masses of red, yellow, and dark brown rocks with intrusive dikes.
Rock specimens collected here were described by the writer in an appendix to the South Sandwich
Islands Memoir (pp. 191-7). Of the fifteen specimens, eight were obtained from exposures and seven
were cobbles from the beach. Six rocks were obtained from an escarpment at 50 ft. above sea-level.
Four of these were acid lavas (dacite) with good flow structures, and two were pyroxene-andesites
containing both augite and hypersthene. As a black slaggy andesitic lava with red crusts was collected
at 100 ft. it is inferred that the upper part of the cliff probably consisted of andesite while the dacites
came from an underlying flow. At the top of the cliff, 150 ft. above sea-level, a true andesite-tuff was
collected, which may represent the final explosive discharge of this volcanic episode. The beach
cobbles and pebbles consisted mainly of dacites and andesites similar to those collected m situ. In
addition there was a specimen of olivine-andesite (or andesitic basalt) and one of andesitic pumice.
Thule Island is therefore notable as providing the only acid lavas so far known in the South
Sandwich Islands. The hypersthene-bearing andesites are also distinctive as they have hitherto only
been recorded from Candlemas Island (p. 95). Analyses of dacite and hypersthene-andesite from
Thule Island were published in the above Appendix, and are restated in Table 8 below.
I
SOUTH SANDWICH ISLANDS
CHEMICAL COMPOSITION OF LAVAS FROM THE SOUTH
SANDWICH ISLANDS
Four chemical analyses of the lavas are recorded in Table 8, in order of decreasing silica percentage,
along with comparable analyses of lavas from the South Shetland Islands, South America, and the
West Indies. In the lower part of the table the von Wolff parameters as modified by the author
(see p. 59) are given.
Table 8
I
A
B
2
C
3
D
4
E
F
SiO„
69-45
67-71
69-56
54-90
54-24
52-68
52-00
48-34
48-26
48-71
AIP3
14-20
14-65
15-65
17-62
17-20
16-38
19-22
13-45
17-42
18-40
Fe,03
2-83
1-59
1-24
2-70
2-81
3-II
2-73
1-12
3-36
3-70
FeO
3-24
3-29
0-91
6-80
4-98
7-98
5-61
11-34
5-61
5-25
MgO
0-25
0-85
0-82
3-93
5-84
7-47
5-54
6-62
8-83
10-30
CaO
3-05
2-34
2-52
9-05
10-19
8-o8
10-58
11-43
11-56
lO-II
Na,0
4-15
6-09
409
2-90
2-91
2-75
2-53
2-22
2-44
2-34
K26
I-5I
1-99
2-19
0-54
0-92
0-44
0-76
0-19
0-89
0-43
H,0+
0-40
o-i6 1
(0-301
1 0-20)
(o-20
294
0-24
0-25
H,0-
o-6o
- J
2-92
0-09
0-20
|o-i5
0-30
o-i6
CO,,
nil
—
—
nil
—
—
nil
0-12
nil
—
TiO,
0-15
i-oo
—
0-70
0-91
0-77
0-63
1-47
1-07
1-08
P2O5
0-14
o-i6
0-13
0-09
0-09
0-02
o-i I
tr.
0-22
0-06
MnO
0-07
—
—
0-23
—
o-i6
o-ii
0-32
0-14
o-o8
(Ni, Co)0
ml
—
nil
—
—
—
—
—
S
tr.
—
—
tr.
—
tr.
—
—
CI
—
—
—
—
—
0-05
—
—
100-04
99-83
100-03
99-96
100-18
100-09
100-17
99-86
100-20
100-71
0
34-6
20-1
34-7
lO-I
5-3
3-6
3-5
— i-o
-5-2
-3-6
F'
44-3
62-5
49-2
27-3
29-3
25-1
25-5
19-3
25-0
21-7
M'
2I-I
17-4
16-1
62-6
65-4
71-3
71-0
81-7
8o-2
81-9
nak
62-6
82-6
67-9
30-1
33-7
30-4
26-1
28-0
28-6
23-9
k
19-3
17-7
25-9
9-6
17-5
8-2
16-3
5-4
18-4
II-6
1. Dacite (Dacitoid) lava, Beach Point, Thule Island, South Sandwich Islands. Anal. F. Herdsman. Quoted from
G. W. Tyrrell, 'Report on Rock Specimens from Thule Island, South Sandwich Islands', Souih Samhvicli Islands
Memoir (1931), p. 192.
A. ' Trachyandesite ' (Gourdon); Santorinite (Earth and Holmsen), Deception Island, South Shetland Islands. Quoted
from E. Gourdon, C.R. Acad. Sci., Paris, clviii, p. igo6 (1914). Also see this Memoir, p. 58.
B. Dacite, Guaitara Slope, Loma de Ales, Colombia. Quoted from J. P. Iddings, Igneous Rocks, 11, p. 496 (1913).
2. Hypersthene-andesite lava, Beach Point, Thule Island, South Sandwich Islands. Anal. F. Herdsman. Quoted from
G. W. Tyrrell, op. cit. supra, p. 195.
C. 'Basalt' (Gourdon), Bridgeman Island, South Shetland Islands. Quoted from E. Gourdon, op. cit. supra, p. 1906.
Also see this Memoir, p. 59.
3. ' Olivine-free basalt' (Backstrom), Saunders Island, South Sandwich Islands. Quoted from O. Backstrom, Bull. Geol.
Inst. Upsala, xiri, p. 173 (1915).
D. Olivine-basalt lava. South Soufriere Hill, Montserrat, West Indies. Anal. F. Herdsman. Quoted from A. G. Mac-
Gregor, 'The Volcanic History and Petrology of Montserrat. . . ', Philos. Trans., Ser. B, ccxxix, p. 74 (1938).
4. Olivine-basalt lava. Cook Island, South Sandwich Islands. Anal. F. Herdsman. New analysis.
E. Olivine-basalt lava. Recent volcano. Penguin Island, King George Island, South Sandwich Islands. Anal. F. Herdsman.
New analysis (see this Memoir, p. 59).
F. Labradorite-basalt, Chateaubelair, St Vincent, West Indies. Quoted from A. Lacroix, 'Les caracteristiques litho-
logiques des petites Antilles', Livre Jubilaire, Soc. Geol. Beige, pp. 387-405 (1926).
There is a wide gap between the dacite (no. i) and the prevalent andesitic and basaltic lavas
(nos. 2, 3, 4) of the South Sandwich Islands— a gap which may be filled by future collections, although
it seems probable that all of the islands are built mainly of the basic lavas. Comparatively high lime
,02 DISCOVERY REPORTS
is characteristic of the whole series, inckiding the dacite in which the nak ratio is only 62-6. Another
general feature is the low F' jM' ratio (less than 0-5) in the prevalent basic lavas. This is in agreement
with the Recent basalt lavas of the South Shetlands, but is in strong contrast with the equivalent
lavas of the Andes, in which the F' jM' ratio fluctuates round about unity. It is a remarkable and
perhaps significant fact that West Indian or Antillean lavas agree best with those of the South Sandwich
Islands in this respect (cols. D and F, Table 8).
The dacite of Thule Island (Table 8, col. i), while of sodic type, does not compare very well with
the analogous santorinites of Deception Island (this Memoir, p. 58). Comparing the tiak ratios
(Table 8, cols, i. A) it is seen to be much less alkalic than the Deception Island rock, and that entails
a much larger amount of free silica (O). It compares rather closely, however, with an Andean dacite
from Colombia (Table 8, col. B). The hypersthene-andesite of Thule Island (Table 8, col. 2) compares
fairly closely with the Bridgeman Island basalt (Table 8, col. C), but no Andean lava of like silica
percentage could be found with even an approximately similar F'l'M' ratio. Backstrom's ' olivine-free
basalt' from Saunders Island (Table 8, col. 3) finds its closest analogue in an olivine-basalt from Mont-
serrat (Table 8, col. D). The olivine-basalt of Cook Island (Table 8, col. 4) is only slightly undersaturated
(^= — i-o), notwithstanding its comparatively large content of olivine. This illustrates its affinity
with the more basic types of andesite. It compares well with the olivine-basalt lava of the Penguin
Island volcano (Table 8, col. E), with the exception that it is slightly less undersaturated and somewhat
more potassic than that rock. Again, the closest analogue of this rock is a labradorite-basalt lava from
St Vincent in the West Indies (Table 8, col. F).
It would appear, therefore, that the predominant basic lavas of the South Sandwich Islands show
closer affinities with the comparable rocks of the Antilles than with those of the Andes. This may,
in turn, be regarded as evidence in favour of the view that the South Sandwich Islands do not lie
on the main line of the Scotia Arc, but form an easternmost ridge parallel to and in echelon with it.
On this view the main line of the Scotia Arc may curve southward from the eastern end of South
Georgia and join up with the South Orkneys. The most recent chart of the Scotia Sea^ shows South
Georgia trending to the south-east away from the line connecting it with the South Sandwich Arc,
and pointing towards a marked northerly projection of the 3000 m. depth-contour which, in turn,
leads towards the South Orkney Islands.
Of the basement on which the volcanoes of the South Sandwich Islands stand we possess only very
exiguous and doubtful scraps of information, namely, a comparatively large piece of dolomitic
mudstone dredged off Zavodovski Island (p. 94), and fragments of epidote-biotite-gneiss and vein
quartz taken from a piece of floating ice near Bristol Island (p. 99). Any future geological exploration
of the islands should therefore include search for exposures of this foundation, and examination of
coarse fragmental igneous deposits for non-volcanic material which may be presumed to have been
derived from the basement. The latter line of research is much more likely to be fruitful than the
former, except perhaps on Bristol Island.
Acknowledgements. The author's thanks are due to the Discovery Committee for defraying the cost
of the new rock analyses published in this work, to Prof. W. J. McCallien, D.Sc, for re-drawing
Fig. 8, p. 60, to J. M. Wordie, M.A., for his valuable introductory Foreword, and to Dr N. A.
Mackintosh for his editorial vigilance during the progress of this Memoir towards publication.
1 H. F. P. Herdman, 'Report on Soundings taken during the Discovery Investigations, 1926-32', Discovery Reports,
VI, pi. xlv(i932).
[Discovery Reports. Vol. XXIII, pp. lo^-ij^, Jntie, 1945.]
THE DEVELOPMENT AND LIFE-HISTORY OF
ADOLESCENT AND ADULT KRILL,
EUPHAUSIA SUPERB A
By
HELENE E. BARGMANN, Ph.D.
CONTENTS
Introduction
Material and methods
Acknowledgements .
Development
Larval krill
Adolescent krill
Adolescent males
Adolescent females
Adult krill
Mature males
Mature females
Pairing .
Spawning
Average growth rate
Factors influencing growth rate
Conclusions
Bibliography
Appendix . . . .
page 105
105
106
106
106
108
108
III
114
114
"5
117
118
120
128
130
131
132
THE DEVELOPMENT AND LIFE-HISTORY OF
ADOLESCENT AND ADULT KRILL,
EUPHAUSIA SUPERB A
By Helene E. Bargmann, ph.d.
(Text-figs. 1-3)
INTRODUCTION
^His paper is an extension of my short one, published in 1937- The stages in the development of
Tthe reproductive system described therein have been used here to work out the composition of the
euphausian population as a whole and its growth rate. This method was first employed by Ruud,
but as he was handicapped by lack of material he could not carry his work quite far enough.
I have been more fortunate in having access to the very extensive Discovery Collections; indeed,
there has been more material than I could cope with single-handed, and some selection became
necessary. My object has been to obtain as complete a series of observations as possible throughout
the whoL yelr. Unfortunately, weather and ice conditions in the Antarctic make it difficult to
fish nets in autumn and winter. The material for this time of the year is consequenUy very
scanty compared with that for the spring and summer months when there was such great
abundance that I could not examine it all. The voyages of the two ships R^R.S. D-o-ry 11 and
R R S 'William Scoresby', have covered between them the whole of the Antarctic zone, but the r
programme of work has kept them so continually on the move that regular observations in definie
LaUties are not available in consecutive months. I have therefore had to combine -tenal r m
different regions and different seasons in order to obtain records extending over all he months of he
year, and even so the material for the month of July is so scarce as to be neghgible. however 1 is
reasonable to conclude that, by using material for several seasons, a very fair general idea of the
average conditions in which Euphausia mperba grows and breeds is obtained.
MATERIAL AND METHODS
Material collected over a period of ten years was used. The me.ltod of examining specimens was the
same as that described in my previous paper. Each specimen was measured to .he nearest mrllune r
from the anterior margin of the eyes to the tip of the telson; the carapace was then opened under a
binocular microscope and the stage of development of the reproduct.ve system was determined, external
sexual characters also being noted ; 8029 specnnens were measured and dissected ,n this way.
The results of th,s intenL investigation are all set out in the appendix. The total catch from each
station has been divided into males and females, which are tabulated separately. All particulars of
ngth and internal and external development are given, together with the totals of the different stages.
Frfser's records of eggs and early larvae have been added .0 the lists of females to show as clearly as
possible the correlation between the occurrence of adults and eggs. t.t j-ff„.„,
' No statistical tests of validity have been applied to calculations of the average lengths of the different
stages, because the stages are in themselves always anatomically distinct. Nor have any fortnulae been
used in working out the curve of growth. There are too many factors involved for any of the existing
mathematical tfethods to be applied with any certainty. As Ottestad (1933) writes: "In the course of
1-2
^°^ DISCOVERY REPORTS
our studies of the problems of growth, it has gradually become manifest to us that, with our present
knowledge of the numerous factors determining growth, the problem of finding a law that will explain
the whole Cham of causes upon which growth depends is for the time being insoluble."
ACKNOWLEDGEMENTS
I am fortunate in having been able to discuss various problems arising during the course of this
work with Dr N. A. Mackintosh and with Dr F. C. Fraser, and I am very grateful for their criticism
and advice. My colleagues, Dr T. J. Hart and Miss D. M. E. Wilson, have helped me in many ways
the former by his work on the phytoplankton of the Antarctic zone, and the latter by her constant
interest and the practical way in which she has helped me to reduce the large body of evidence into
manageable shape.
DEVELOPMENT
LARVAL KRILL
During the first year of growth, Enphausia mperba passes from the egg through the successive larval
stages of nauphus, metanauplius, calyptopis and furcilia, umil it enters upon its second year of post-
larval or adolescent life.
Its larval history has been dealt with in detail by Fraser (1936) in his paper on the "Development
and distribution of the young stages of krill {Euphamia mperba) ". A summary of his work and a com-
parison with the observations of Taube and Lebour on euphausians of the northern hemisphere must
be given here, in order to present as complete a record of early growth as possible.
Fraser obtained, by analysis of plankton samples, records of eggs and their occurrence extending
from the first part of November to the latter part of March, a period of four and half months Just
before laying, the eggs of E. superba are so tightly packed that, on the outer surface of the ovary they
are approximately pentagonal or hexagonal in shape, while on the inner side they are roughly conical
I have measured sixty eggs from two gravid females, and I find that their average diameter is 0-55 mm
although their greatest diameter may be as much as o-68 mm. or even 072 mm. (Ruud, 1932), but that!
after laying, the eggs assume a spherical shape with a consequent adjustment in size, those' found in
the plankton and examined fresh measuring o-6o mm.
Fraser states that "eggs occurred in the plankton showing all stages of development, culminating
m the clearly distinguishable form of the ist nauplius". Only two free-swimming specimens of the
ist nauphus were obtained, one measuring 0-63 mm. in length and the other o-66 mm. These were
caught during the second half of December, together with three 2nd nauplii, measuring 0-65 o-68
and 070 mm. respectively. "The rarity of ist and 2nd nauplii and the smallness of numbers where
records exist may mdicate that these stages are passed through very rapidly in this species, as in other
euphausiids where the development is known".
Taube (1915) and Lebour (1926) found that in northern waters the euphausian egg can develop into
the metanauplius within a few days. Observations on Nyctiphanes norvegicus indicate that the free-
swimming nauplius is hatched from the egg three days after laying, and that by the fifth day the limbs
have taken on the metanauplius form, but that the mandible and lower lip characteristic of the fully
developed stage do not appear until about the fourteenth day.
Metanauplii occurred in the Discovery material in fair numbers from February onwards very big
catches being obtained at two stations in March. The average length of the larvae at this stage is
approximately 0-95 mm.
The measurements of these early developmental forms show that the larvae do not grow very
I07
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
rapidly until they begin to feed independently. Sars (1898) thought that this occurred in the meta-
nauplius, when the mouth opens to the exterior ; but Taube has shown that at this stage, in Nyctiphanes
norvegicus, there is as yet no connexion between the mouth and the mid-gut, and Macdonald (1927)
states that, in Meganyctiphanes norvegica, "although an open mouth is present in the metanauplius it
was not found to feed ". The internal yolk supply suffices until the gut is fully established. This occurs
after the calyptopis stage is reached, when the mouth and proctodaeum become connected with the
mid-gut. The cells of the mid-gut still contain a certain amount of yolk, but Sars (1898) writes that
the larva now begins to feed actively, " chiefly upon small Diatomeae, the remains of which could be
distinguished by microscopical examination of the contents of the intestine". In E. superba, "the
more typically oceanic species of diatoms are evidently digested rapidly: recognizable fragments are
rather rare even in the crop (Hart, 1934).. • .Two forms that appeared constantly in the stomachs of
adult specimens and remained clearly recognizable were Fragillaria antarctica^ and Thallassiosira
antarcttca".
Table i has been compiled from the data in Eraser's paper and gives the average lengths of the
different larval stages. It will be noticed that at the ist calyptopis stage, when the larva begins to feed,
its length is at once almost doubled, after which growth proceeds again more regularly throughout
the summer. "By the time the euphausiid reaches the 6th furcilia stage, the major developmental
changes have been effected and in appearance it is characteristically a euphausian."
Table i . Average lengths of larval stages
Average
Average
Average
Stage
length of
Stage
length of
Stage
length of
larvae in mm.
larvae in mm.
larvae in mm.
Egg
o-6o (diameter)
2nd calyptopis
2-71
3rd furcilia
7-32
I St nauplius
0-65
3rd calyptopis
3-98
4th furcilia
8-01
and nauplius
0-67
I St furcilia
4-50
5th furcilia
9-52
Metanauplius
°-95
2nd furcilia
5-II
6th furcilia
11-34
I St calyptopis
171
By plotting half-monthly average lengths of the larvae for the period of one year, Fraser found
that from November to March (the period of spawning) growth was slow, but that it increased steadily
from March to June, was retarded during the mid-winter months and began to increase again at the
end of August, by which time the first adolescents had made their appearance. Evidently, growth
from the egg of the adolescent occupies an average period of about nine months, although under
optimum conditions it can proceed more rapidly.
Eraser's work on larval krill shows clearly that spawning in E. superba is not restricted to one short
period, but is spread over most of the southern summer, with the result that new broods of larvae are
continually being hatched out, and the stock is constantly replenished. Taube (1915) and Ruud (1936)
found that in northern waters, Nyctiphanes and closely allied euphausians had a similarly extended
spawning season. Consequently, eggs, larval forms, adolescents in every stage of development, and
adult individuals can be, and frequently are, found to exist side by side, and the euphausian population
presents a very heterogeneous appearance.
The larvae of E. superba after one year of growth have attained by the following November an
average length of 13 mm. Their subsequent development from adolescence to maturity forms the
subject of this paper.
' Revised by Hendey (1937) and now called Fragillariupsis antarctica.
io8
DISCOVERY REPORTS
ADOLESCENT KRILL i
Of the 8029 specimens of E. superba which I have examined, 6006 were adolescent and of these 3073
were males and 2933 females. The youngest adolescents first make their appearance in any number
in August ; they show no trace of external sexual characters, but internally the reproductive system is
recognizable, and by dissection under a low-power binocular microscope the sex of each individual
can be determined. I have described the development of the reproductive system in the short paper
forming an introduction to this one, which has been published in vol. XiV of the Discovery Reports.
It will therefore be sufficient to summarize this development here, before discussing how growth
proceeds during adolescence.
Ruud (1932) has drawn attention to the fact that, in E. superba, investigation of the testis and ovary
is the only reliable method of determining maturity, and that the reproductive system of each in-
dividual must therefore be examined before the composition of any specific population can be estimated.
After examining the euphausian material obtained during the cruise of the S.S. ' Vikingen', he dis-
tinguished four stages of maturity in both males and females. These are listed in Table 2.
Table 2 (after Ruud). Stages of maturity
Males
Females
1
No spermatophores visible in the ejaculatory duct
1
Ovary small and immature. Eggs o-i 0-0-25 mm.
(depository)
in diameter
2
Visible spermatophores : not loosened by touching
2
Ovary large but immature. Eggs o-26-o-5o mm.
with a needle
in diameter
3
Visible spermatophores : loosened when lightly
3
Ovary large and mature. Eggs 0-5 1-070 mm. in
touched
diameter
4
Empty ejaculatory ducts. Mating has recently
4
Ovary small, mainly germinal layer. Eggs 0-54-
taken place
0-65 mm. loose in thorax. Spawning has taken
place
Ruud states that he does not know of " any practical method by which the degree of maturity of the
testicle can be ascertained". Consequently, he included within stage 1 all those male euphausiids
which were not fully adult (i.e. all those with no visible spermatophores in the ejaculatory ducts), and
he found that the specimens showed a very wide range in length: i6-6-44-4 mm.
As a criterion of development in the females, he used the diameter of the egg, and again found great
variation in length in the specimens included within stage 1.
It is clear that this first group of Ruud's, comprising as it does males and females of such difi^erent
size, covers the whole period of adolescence, during which time the reproductive system becomes
mature. A closer investigation of these adolescent forms has thrown more light on the development
of E. superba.
ADOLESCENT MALES
Although during adolescence the individual growth rate varies very considerably, five stages can be
distinguished in the development of the male sexual organs, both internal and external. These stages
Table 3. Growth stages in the male
Stage
2
3
4
5
Internal structures
Primitive condition. Small testis, simple uncoiled
vas deferens
Small posterior flexure appears on vas deferens
Lateral pocket appears on posterior fle.xure
Anterior flexure appears on vas deferens
Coiling on vas deferens near posterior flexure
Stage
B
C
D
E
External structures
Undifferentiated ist pleopod
Petasma appears as an undivided lobe
Petasma becomes divided into two lobes
Wing develops above petasma
Wing grows and curves over petasma
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
109
1>
„ C3
•A9p "ixg
1 oooooooo 1 1 0
Mil tooooo 1 1 1
•luiu ijiSuaq
1 r^u-iLnt^oooooK-il 1 '^
1 u-i u-t u-t yj~t ui\o 10 10 1 1 "•I
1 1 1 1 vS vO mvD mill
CiO -
CD
4-. C!
•A3p -ixg
loooooooo 1 1 1
MM ooooo 1 1 1
«
1'S "
■uim ijiSuaq;
1 N Ln o^ "^ :i :i'=2.°2. 1 1 1
-
1 1 1 lu-iLn-ct-rhTl-l 1 1
bD -
B
CO
to JJ
•A3p -1X3
E -
[iHpHtafeOlt^OtHl 1 !
1 1 1 OOfeOOO 1 1 1
M.S -
•turn m§u3q
•< '
1 1 lOOOOsOOOOl 1 1
■A3p -jxa
WfcOOfet^f^t^f^l 1 1
1 1 1 owooow 1 1 1
■uiui qiSuaq
OS 0 r- i^ 0 CO i2'=2.°S. 1 1 1
E '
■♦-»
<
1 1 |vDMOOC^J^;-C2| 1 1
1 1 |Tt-r'>c^'l--^T)-| 1 1
IT)
is "
•Asp ixg
CO
E
■4-*
a
8
CO
"o
-a
<1
WfeWMfeP^feWpH 1 |W
WWOt^feOC^I 1 1 1 1
•urni qjSuaq
00 0 w 10 N ro 0 "^00 1 1 00
^ ^ -"l- ^ "1 "^^ in ir> 1 1 ^
ly-l osvO M CO " m 1 1 1 1 1
«5 (U
•A3p 'ixg
QPlJnWWQt^Wta |WW
W|W00QQQ| 1 1 1
t»3
"a
••Kt
•UIUI ij:iSu3q
rorOfOc^-^-^-^f^'+l •^-*-
N|<X3^0'-r^r2l 1 1 1
B
CO
is "
hJ Oh
•Aap -5x3
Q WU QtaP^ P W Pi^ Q W 1
wwwwwo |ww 1 1 1
•UIUI qiSuaq
0 0 t~^N (jvr^osN r^sogs j
(^1 u-j o t^ -+^0 1 oo so 1 1 1
T^- 1/-J u-i r*^ ■n- n- 1 rf rr 1 1 1
•A9p -ixg
OOOPQmOOQQQQO
QQQQWO IWQ 1 1 1
s
•UIUI njSuaq
g
CO
•A3p -ixa
OUUUfeWWQQOU 1
E
CJ
CO
QQQU lUUUQQWQ
•UIUI mSugq
• U-, N rovo 0 '^ ,"t Ji D: li 2- 1
mcJO O O 1 O gvOO sp m OS "
•Aap •jxg
mUUOCQPQPQOmOOO
pauoomopqomoupa
J2
•UIUI m§U3q
o w (s u-)oo r^ " -h <3s r--so qs
CN) o (SI N c^ ro <^, c^ CM CO (^ csl
<u p
00. S
b 0
C^ nj
•A3p -ixg
m pq pa pa a m pa u « m m |
OOPUPQPSUUUOUm
•UIUI m§u3q
P u
•Aap •jxg
<; pq << < pa PQ pa pa < fp PQ 1
eQpapaPQ-<OP3P3paoapa-<
•UIUI qiSuaq
__~-)P> M o m^^ "-100 o 1
NNNNNMNNNNcnl
"3
•a
<
M ro iri M O « ■* "^^ sO O sD
(SiNNNNC^NNNNC^N
^—1
*->
u p
i; "
CO
•Aap •jxg
pq<j<;<pa<P3m<<| 1
<; pa<i<;ooo mOP3 pa |
•mui mBuaq
PjioOO'-'OOsOs'^'-' I i
U-, u-i o >- t^ b^^'^ ^ !i, ^ 1
(SJ <s N c^c^'^'^'^'^T'C^c^ I
■M P
to aj
« -3
p (U
& g-
•A3p •^xg
<<<<j<-<-<<d<;<:p5 1 1
<ii<C<<<<<'<<<< 1
•UIUI \\xSu3'-
[
„ ^„ O NCO OVOOOO 1 1
2 2 S 2 2'2^ S N S ?r 1
c
0
< CO O Z Q -^t^ § < S ►^►^
CO M
\0 sD
'XJ
3
3
-a
c
^c
V
OJ
An
M)
P
?■
u
<i)
<u
N
N
CD Cfi
o
sO
I
OS
^ CM-
c
c
o
CJ
m
o
n
-a
Tl
ca
;;
r
u
«)
OX)
0<)
c
C
CTl
cd
^
i-*
«)
u
N
N
no DISCOVERY REPORTS
have not been arbitrarily selected, but are part of the normal development of the reproductive system.
I have dissected 3073 adolescent male specimens and have found that growth proceeds by the addition,
in constant sequence, of definite anatomical structures. The appearance of each structure in turn
marks a stage, but internal and external development do not necessarily keep pace with one another.
These stages, five in number, tabulated below in Table 3, are collectively equivalent to the stage 1 of
Ruud. Stages A-E are the approximate external equivalent of stages 1-5.
The majority of adolescents of stage 1 show no differentiation of the copulatory organs : that is to
say, externally, development corresponds to stage A, the ist pleopod being unmodified. In some cases,
however, external growth proceeds more rapidly than internal, and the pleopod may carry a primitive,
undivided petasma, stage B. The total number of adolescent males of stage 1 which I have dissected
is 121 1, and of these 10 14 were at stage A externally and 197 at stage B.
Table 5
. Average length per month of each stagi
? in mm.
Month
Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
Stage 6
Stage 7
<S
?
(?
?
(?
?
S
?
cJ
?
S
?
S
?
Aug.
Sept.
Oct.
Nov.
Dec.
Jan.
Feb.
Mar.
Apr.
May
June
July
13
14
15
20
23
23
25
25
28
28
13
18
15
20
23
25
28
33
32
28
30
26
26
28
27
28
33
33
33
31
32
32
29
26
29
25
31
34
32
38
33
31
36
28*
32
31
31
34
40
42
43
44
40
38
37
30*
35
33
29
27
28
39
36
43
40
41
41
38
35
36
34
38
42
44
49
47
46
41
41
35*
39
42
40
31
39
44
46
42
40
41
37
41
47
47
53
49
53
40*
47*
46
49
45
43
45
42
46
43
44
48
43
41
46
48
54
52
51
47
46
50
5°
48
47
54
44
45
51
51
52
52
52
54*
57
54
52
50
47
* Neglect?
Average length of each stage per annum in mm. for the whole year
Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
Stage 6
Stage 7
Males
22
31
41
43
45
47
50
Females
24
32
36
40
44
49
50
Individual length varies very considerably. Table 4 gives the monthly minimum and maximum
lengths for all stages, and shows clearly that size alone is not necessarily a reliable criterion of develop-
ment. In November, for example, the length of the individuals of stages 1 and A varies widely over
a range of 10-30 mm., whereas in May the range is much narrower, lying between 26 and 31 mm.
Nor is size always associated with a more advanced stage of internal development : in February, for
example, a specimen measuring 39 mm. has a pleopod at stage B, but in May the largest specimen,
measuring 31 mm., is only at stage A, while the small specimen (26 mm.) is already at stage B.
The average monthly length of the specimens, on the other hand, shows a steady increase throughout
the year, starting at 13 mm. in August and reaching 28 mm. by the following May (see Table 5, which
also gives the average length of each stage for the whole year).
This great variation in the size of individuals within a stage is characteristic of all adolescents,
whether they are at stage 1 or at stage 5, or at any of the intermediate stages. By referring to Table 4,
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL iii
the whole range of size can be seen, while Table 6 gives the number of males and females in each stage
per month and their degree of external development.
Of stage 2, 795 specimens were examined. The stage of external development of the majority of
these (694) was stage B, though in 60 specimens the copulatory organs were as yet undeveloped
(stage A). Among the largest specimens, more advanced external development was found, 40 specimens
being at stage C and one even at stage D. This last specimen was not, however, the largest recorded;
it measured 43 mm., whereas the greatest length found was 46 mm., this specimen occurring in March
and being at stage C externally. The average length of stage 2 was 26 mm. at the beginning of the
southern spring and a steady increase was maintained throughout the year, rising to a maximum of
33 mm. in March, with a slight drop to 32 mm. in June.
No specimens at stage 3 were found showing the primitive condition, stage A, of external develop-
ment. The majority (268) were at stage C, iii were at stage B, 150 at stage D, 36 at stage E and 5 at
stage F (the first external stage usually characteristic of young adults), making a total of 570 in all.
External development is beginning to run ahead of internal development, with the result that super-
ficially some of the specimens appear to be approaching maturity. The maximum lengths in some
months are similarly deceptive, in December to February the largest specimens being over 50 mm. The
average monthly lengths, however, indicate steady growth to a size intermediate between young
adolescents and adults ; they increase from 32 mm. in August to 44 mm. in March, with a fall to 37 mm.
in June.
Among the 311 males at stage 4, much the same conditions obtain. The largest specimens were
well advanced in external development, being mostly at stages E and F, although the biggest one of all,
measuring 59 mm., was only at stage D. The majority of the specimens (150) were also at stage D, while
two specimens were at stage B, 47 at stage C, 85 at stage E and 27 at stage F. The average monthly
length of stage 4 increases from 35 mm. in August to 49 mm. in February, to fall again through 47 and
46 mm. to 41 mm. in May and June.
Stage 5 is the last adolescent stage. After passing through it, the male specimens of E. superba can
be regarded as being fully mature. In all, 186 specimens were examined, 19 being at stage D, 97 at
stage E, 68 at stage F and 2 at stage G, externally these last two being fully developed, although not
the largest specimens measured. The average length varies from 40 mm. in August through 37 mm.
in October to a maximum of 53 mm. in February and April.
ADOLESCENT FEMALES
No special sequence of structural additions marks the growth of the female reproductive system. It
has been pointed out already that Ruud used the diameter of the eggs as the criterion of development
in the female. There is, however, a period of growth before the eggs themselves can be measured under
an ordinary binocular dissecting microscope, during which time the ovary is clearly getting larger.
I have found that during this period, it is possible to use the size of the ovary as an indication of
maturity. Three stages are passed through before the eggs become of measurable size: these three
stages mark the period of adolescence. The thelycum, the thoracic pouch into which the spermato-
phores are inserted, also passes through three stages before the adult condition is reached. These
stages of growth are tabulated below. They are based on the examination of 2933 adolescent females.
The primitive, unlobed, saddle-shaped ovary can be distinguished in specimens as small as 10 mm.,
in which there is as yet no sign of the thelycum. It can also occur in specimens measuring 45 and
46 mm., in which the thelycum is at stage C. The average monthly length of the adolescents of stage 1
ranged from 13 mm. in August through 33 mm. in the following March to 30 mm. in June. The
thelycum was undeveloped in the majority of the specimens. Out of 1796 examined, 1150 were at
DISCOVERY REPORTS
Stage A, 403 at stage B, 220 at stage C, 21 at stage D and 2 at stage E, normally a stage characteristic
of the adults, these last measuring 38 and 39 mm. respectively. The analyses of the measurements of
adolescent females will be found set out in Tables 4-6.
Table 6. Analysis of total catch into stages
No. of individuals
in each stage per month
Total
No. of individuals in each stage per month
Total
Stage
A
Stage
B
Stage
C
Stage
D
Stage
E
Stage
F
Stage
G
Stage
A
Stage
B
Stage
C
Stage
D
Stage
E
Stage
Stage
G
Stage 1 :
Adolescent males
Adolescent females
Aug.
Sept.
4
7
I
—
—
5
7
14
6
12
z
—
—
—
—
14
18
Oct.
Nov.
234
182
z
—
234
182
254
211
2
z
z
z
z
—
254
213
Dec.
252
45
—
—
—
—
—
297
257
55
26
—
—
—
338
Jan.
199
—
—
—
—
—
—
199
191
40
9
—
—
—
—
240
Feb.
60
130
—
—
—
—
—
190
131
82
17
—
—
—
—
230
Mar.
Apr.
May
54
18
4
9
7
5
—
—
—
—
—
63
25
9
70
8
7
I II
83
15
74
94
20
I
2
—
—
275
188^
22
June
July
—
—
—
—
—
—
—
—
I
3
—
—
—
—
—
4
Total
1014
197
—
—
—
—
—
1211
1 150
403
220
21
2
—
—
1796
Stage 2:
Aug.
14
22
I
—
—
—
—
37
6
32
28
—
—
—
66
Sept.
—
17
—
—
—
—
—
17
2
7
8
—
—
17
Oct.
4
29
2
—
—
—
—
35
—
4
10
I
—
—
15
Nov.
17
23
—
—
—
—
—
40
I
15
2
— •
—
—
18
Dec.
—
79
3
—
—
—
—
82
2
2
3
I
—
—
8
Jan.
9
31
—
I
—
—
—
41
—
3
2
—
—
5
Feb.
—
297
19
—
—
—
—
3i6
135
142
25
2
—
—
—
304
Mar.
4
bg
14
—
—
—
—
87
3
51
30
22
4
—
—
no
Apr.
12
102
I
—
—
—
—
"5
7
46
13
3
2
—
—
71
May
—
21
—
—
—
—
—
21
—
28
6
—
—
—
—
34
June
—
4
—
—
—
—
—
4
—
14
II
3
—
—
—
28
July
—
—
—
—
—
—
—
X
2
—
—
—
—
3
Total
60
694
40
I
—
—
—
795
157
346
138
32
6
—
—
679
Stage 3 :
Aug.
—
5
19
—
—
—
—
24
—
9
37
58
6
—
—
no
Sept.
—
I
2
—
—
—
—
3
—
—
2
2
—
—
—
4
Oct.
—
2
21
—
—
—
—
23
—
I
26
25
2
—
—
54
Nov.
• —
—
3
—
—
—
—
3
—
—
5
3
—
—
—
8
Dec.
—
II
4
5
—
3
—
23
—
1
—
—
—
—
—
I
Jan.
—
II
4
4
8
—
—
27
—
—
2
—
—
—
—
2
Feb.
—
50
«S
44
25
—
—
204
I
99
22
2
—
—
—
124
Mar.
—
3
63
54
—
—
—
120
I
—
17
10
2
—
—
30
Apr.
—
26
53
41
3
2
—
125
—
17
16
13
4
—
—
50
May
—
2
4
I
—
—
—
7
—
—
3
3
—
—
—
6
June
—
—
9
I
—
—
—
10
—
—
27
26
8
—
—
61
July
—
—
I
—
—
—
—
I
—
3
I
4
—
—
—
8
Total
—
Ill
268
150
36
5
—
570
2
130
158
146
22
—
—
458
Stage 4:
Adult females
Aug.
—
—
13
23
I
—
—
37
—
—
—
13
18
3'
Sept.
—
—
I
I
2
—
—
4
—
—
—
8
25
— 1 —
33
Oct.
—
—
7
8
2
—
—
17
—
—
—
9
165
I 1 —
17s
Nov.
—
—
5
17
2
—
24
—
—
—
7
8
— ■
—
IS
Dec.
—
2
3
12
4
6
—
27
—
—
—
2
9
—
I
12
Jan.
Feb.
—
4
2
9
II
2
33
3
8
—
18
54
—
—
I
2
3
Mar.
—
—
I
30
9
I
—
41
—
—
—
S
12
I
—
18
Apr.
May
June
July
—
—
6
17
3
19
29
9
—
61
—
—
—
II
9
—
—
20
—
—
4
I
I
—
24
I
—
—
—
246
—
—
—
Total
—
2
47
150
85
27
—
3"
—
—
—
55
3
3
307
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
"3
Table 6 {cont.
)
r
No.
of individuals in each stage per month
Total
No.
of individuals in each stage per month
Total
Stage
A
Stage
B
Stage
C
Stage Stage
D E
Stage
F
Stage
Stage
A
Stage
B
Stage
C
Stage
D
Stage
E
Stage
F
Stage
G
Stage 5 :
Adolescent males
Adolescent females
Aug.
Sept.
Oct.
Nov.
Dec
—
—
—
8
2
4
4°
5
lO
7
I
to
3
2
3
I
48
17
14
13
4
—
—
—
I
5
I
20
48
39
3
12
24
2
18
50
5
I
25 ^
78
114
Jan.
Feb.
Mar.
—
E
4
I
2
s
17
8
6
I
I
14
12
19
—
—
3
1
I
4
10
2
24
7
41
10
I
Apr.
—
—
—
—
7
35
42
May
—
—
—
—
—
June
—
—
—
—
I
I
July
—
—
—
—
2
2
1
Total
—
—
~
19
97
68 2
186
—
—
—
6
117
SI
lOI
275 1
Stage 6 :
Adult males
Aug.
—
—
—
—
2
lO
12
Sept.
Oct.
Nov.
Dec.
Jan.
Feb.
Mar.
Apr.
—
—
—
6
2
38
48
2
2
3
6
S
ID
I
3a
3
S
7
9
I
39
80
5
7
3
21
14
II
!
—
6
3
6
2
5
I
12
IS
6
I
2
37
77
27
3
52
92
45
2
6
May
—
—
—
—
—
—
June
—
—
—
1
July
—
—
—
—
—
Total
—
—
—
6
4
124
58
192
—
—
—
6
16
35
143
200
Stage 7:
I
!
1
1
_
Aug.
—
—
—
—
—
Sept.
Oct.
—
—
5
67
5
67
—
—
—
—
—
—
—
—
Nov.
Dec.
—
—
z
—
34
169
34
169
—
—
—
—
—
I
3
5
4
6
Jan.
—
—
—
—
—
130
130
9
157
166
Feb.
—
—
—
I
—
19
123
143
2
127
129
Mar.
—
— ' —
5
3
183
186
Apr.
—
—
—
—
4
4
—
May
—
—
—
—
—
—
—
June
—
—
—
—
—
July
—
—
—
—
I
Total
—
—
—
I
—
19
538
558
—
—
—
—
3
13
475
491
Total adolescent ^'s 3073
Total adult ^'s 75°
Total adolescent 9's 2933
Total adult $'s 1273
Total ,^'s
3823
Total ?'s
4206
N.B. Adolescent total does not include Fraser's 124 specimens.
Stage
Table 7. Growth stages in the female
Internal structures
Primitive condition. Unlobed saddle-shaped
ovary. Oviducts clearly visible
Ovary becomes lobed. Oviducts become wider
Ovary extends down towards the legs, filling i of
thoracic cavity
Stage
A
B
C
External structures
Thelycum either not visible, or represented only
by a straight band across the sternum
Two small coxal outgrowths can be distinguished
at each end of sternal band
Thelycum half-developed: coxal part larger than
sternal part
114
DISCOVERY REPORTS
The smallest specimen at stage 2 occurred in December and measured 20 mm. ; the largest mea-
sured 48 mm. and occurred in February. The first was at stage A in external development, the second
at stage C. The six specimens recorded at stage E measured from 38 to 42 mm. Most of these adoles-
cents, however, were at stage B: 679 specimens were examined and it was found that 157 were at
stage A, 346 at stage B, 138 at stage C, 32 at stage D and 6 at stage E. The average length varied from
29 mm. in August through 25 mm. in November to 38 mm. in March, falling to 36 mm. in June.
Of stage 3, 458 specimens were examined. Only two of these were external at stage A; of the rest,
130 were at stage B, 158 at stage C, 146 at stage D and 22 at stage E. The average monthly length
varied from 35 mm. in August to 43 mm. in March, falling again in April to 40 mm. and rising once
more to 41 mm. in May and June. Great variation in the individual length of the specimens is met
with again and, as in stages 1 and 2, the largest measurements approach the average lengths of the
young adults very closely, showing that in the female adolescents, as in the male, size is no reliable
criterion of development. The smallest specimen was recorded in October: it measured 22 mm. and
was at stage C externally. The largest specimens occurred in February and June: they measured
49 mm. and were at stages C and E respectively.
The females now approach maturity. At the next stage (4), the eggs become measurable although
the ovary is only half-grown, and in some specimens spermatophores are found in the thelycum,
showing that externally these females are fully mature. I regard stage 3, therefore, as marking the
end of the period of adolescence in the female.
ADULT KRILL
MATURE MALES
There are two adult stages in the male, stages 6 and 7. In both, spermatophores are to be found in
the ejaculatory ducts, but in specimens at stage 6, the spermatophores are not perfectly formed, whereas
when stage 7 is reached, there are fully formed spermatophores in both the ejaculatory ducts and in
the spermatophore sacs. Two adult external stages in the development of the copulatory organs can
also be distinguished. These stages are tabulated below; they correspond to stages 2, 3 and 4 of Ruud.
Table 8. Growth stages in adult males
Stage
Internal structures
Stage
External structures
6
7
Coiling occurs on vas deferens in region of anterior
flexure. Imperfect spermatophores in the ejacu-
latory duct
Fully adult condition. Ripe spermatophores in
the ejaculatory ducts and in the spermatophore
sacs
F
G
Terminal process of inner lobe reaches to the tip
of the median lobe
Fully adult condition. Proximal process develops
a blade-like expansion at its tip
The average monthly length of stage 6 ranged from 44 mm. in August, through 41 mm. in Novem-
ber, to 54 mm. in February, falling again to 52 and 51 mm. in March and April. Between May and
August, no specimens at this stage were found. The smallest specimen, measuring 35 mm. occurred
in November, and the largest specimen measuring 64 mm. in February. The November specimen was
at stage G externally, and the February specimen was at stage F. In all, 192 specimens were examined,
6 being at stage D, 4 at stage E, 124 at stage F and 58 at stage G.
No fully adult specimens at stage 7 were found in May, June and August, and only one specimen
was found in July. The average monthly length varied from 54 mm. in September, through 44 and
45 mm. in October and November to 52 mm. in February, March and April. The smallest specimens,
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL iiS
each measuring 35 mm., occurred in October and December, the largest specimen measuring 60 mm.
in February ; all were at stage G externally. The majority of specimens (538) were fully developed
externally, but in 20 males growth of the copulatory organs was not complete, 19 being at stage F
and I at stage D.
MATURE FEMALES
I have classed all females in which the eggs are of measurable size as mature, i.e. likely to spawn
within the next 2 or 3 months.
The eggs become large enough to be measured when the ovary fills about half of the thoracic
cavity (stage 4). The thelycum, though not necessarily very heavily chitinized as yet, has assumed the
adult shape, and is ready to hold the spermatophores, when they are transferred by the male.
Table 9. Growth stages in adult females
Stage
Internal structures
Stage
External structures
4
Ovary fills -J thoracic cavity. Eggs o-o5-o-i2 mm.
in diameter
A
Thelycum of adult shape, but not heavily chitin-
ized
5
Ovary fills | thoracity cavity. Eggs o- 1 3-0-24 mm.
in diameter
E
Thelycum full grown and well chitinized
6
Ovary fills thoracic cavity. Eggs o-25-o-48 mm.
in diameter
F
Spermatophores in the thelycum : spermatophores
full
,7A
Ovary gravid or nearly so. Eggs o-49-o-70 mm.
in diameter
G
H.
Spermatophores in thelycum: spermatophores
empty, thelycum full
l7B
Ovary primitive again. Eggs spawned
Spermatophores torn away — thelycum empty
again
In stage 4 of the 307 specimens examined, 55 were externally at stage D, 246 were at stage E, 3 were
at stage F and 3 were at stage G. The majority had, therefore, fully mature though unfertilized
thelyca. In six of the specimens, those at stages F and G, copulation had taken place, and the
females were carrying full or empty spermatophores. The average monthly length varied from 39 mm.
in August through 31 mm. in November to 46 mm. in March, falling again in April to 42 mm. No
records were obtained in February, May, June or July. The smallest specimen measuring 27 mm.
occurred in November: it was externally at stage D. The largest specimen of 55 mm. occurred in
September and was at stage F.
By the time stage 5 is reached, the eggs are growing rapidly, and consequently the ovary is beginning
to fill up the thoracic cavity. Externally, 117 specimens were at stage E, but the majority carried
spermatophores, 51 being at stage F and loi at stage G. One large specimen measuring 43 mm. was
at stage D, whereas the smallest specimen recorded, measuring 30 mm., was at stage G, and the
largest specimen of all, measuring 58 mm., was at stage F. The average monthly length varied from
46 mm. in August, through 43 mm. in November and 42 mm. in January back to 46 mm. in February.
Only one specimen occurred in March, and no records at all were obtained in April, May, June or
July. Stages 4 and 5 correspond to Ruud's stage 1.
Stages 6, 7 A and 7 B may be regarded as covering the fully adult condition in the female and are
comparable with stages 2, 3 and 4 of Ruud. The three corresponding stages of external development,
F, G and H, show whether copulation has just taken place, or whether it occurred some time ago,
and lastly whether the female has spawned, although this is also readily seen from her altered shape.
When the spermatophores have been recently implanted in the thelycum, they still contain the sperm-
mass (stage F), but after a while this makes its way into the thelycum, leaving the spermatophores
empty (stage G). When the eggs are being laid, the spermatophores become loosened and break away
ii6 DISCOVERY REPORTS
from their attachment, so that females which have shed their eggs, generally have empty thelyca
(stage H).
Stages 7 A and 7B, comprising gravid females and those which have spawned, have been combined
together as stage 7 in all analyses of the catch, and in the tables of minimum and maximum length,
and in the average length estimations. This has been done for two reasons: in the first place, the
number of gravid females (i.e. females at stage 7 A) obtained was very small, and in the second place,
combining stages 7 A and 7B makes it easier to compare the adult females with the adult males.
The external stages G and H have been treated similarly. But in order to describe the life history
as fully as possible, I have, where necessary, analysed the catch of adult females into the two groups
of gravid and spawned.
Out of 200 specimens at stage 6, 6 were at stage D, i6 at stage E, 35 at stage F and 143 at stage G.
The smallest specimen, measuring 32 mm., was at stage E; the largest, measuring 60 mm., was at
stage G. No specimens at stage 6 occurred in May, June, July, August, September or October. The
monthly average length varied between 47 mm. in November through 50 mm. in January and February
to 47 mm. in April.
Stages 7A and 7B were represented by 491 specimens, of which 57 were gravid, i.e. at stage 7A,
and 434 had spawned, i.e. they were at stage 7B. The time of their occurrence ranged from December
to April. The smallest specimen, measuring 41 mm., was found at the end of the season; the largest,
measuring 64 mm., in February; both were at stage G. The average monthly length decreased from
57 mm. in December to 47 mm. in April.
Table 10. Occurrence of gravid and spawned females
Month
Stage 7 A (gravid females)
Stage 7B (spawned females)
Total
Stage E
Stage F
Stage G
Total
Stage F
Stage G
Stage H
Dec.
Jan.
Feb.
Mar.
Apr.
4
2
48
3
3
I
4
3
I
44
4
118
129
183
I
5
2
3
86
66
2
27
61
181
Total
57
3
6
48
434
8
157
269
Analysis of the total catch of these stages shows that the majority of gravid females (48) occurred
in February, and that all except four of the spawned females occurred in February, March and April.
Attention has already beeen drawn to the scarcity of the gravid stage. Of the 57 specimens obtained,
40 of the large catch in February were taken from deep nets (250-100 m., 750-500 m.); 3 of the
remaining 17 specimens occurred in the surface layer (0-5 m.) and 14 at depths varying between 137
and o m. In the catches examined, the marked failure to take gravid females at the surface in any
number seems to point to deep spawning. The euphausian egg contains a large quantity of yolk,
sufficient to feed the young larva for some little time after hatching, and the external food supply is
therefore not of immediate importance to the larva when it first leaves the egg. Presumably, this
quantity of yolk makes the egg heavy, for Eraser found that the bulk of the eggs and early larval stages
occur at depths below 250 m. Evidence from later commissions of the R.R.S. ' Discovery II ' confirms
this. Eraser suggested that the eggs become concentrated in basins on the submarine ridges, where
the homogeneous water column provides the uniformity of temperature and density necessary for
their development. A large catch of 95 gravid females in a vertical net, fished between 250 and 100 m.,
in February 1930 off South Georgia rather confirms this idea. The vertical 70 cm. net is small and
usually catches very few fully grown specimens of E. superba. The occurrence of 95 in one haul is
117
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
exceptional and suggests that the net passed through a dense swarm of them, which might well
account for a concentration of eggs. Some of the females in this particular catch subsequently spawned
in the ship's laboratory. The great increase in size of the ovary, when the females become fully gravid,
may affect their specific gravity, and this may account for their presence in the deeper layers and their
absence from the surface nets. In the most recent analyses made on board ship, so-called gravid
females were obtained at depths varying between looo m. and the surface, but the majority, however,
occurred in the 0-5 m. layer. These females may not have been fully gravid: measurements of the
eggs in the ovaries could not be systematically attempted in the ship's laboratory, and it is possible
that females have been classed as gravid which more properly belong to stage 6. On the other hand,
if these analyses are correct, the depths at which gravid females are found extend over a very wide
range. This problem really depends for its solution on a discussion of the factors influencing the
general distribution of E. superba, and is rather outside the scope of this paper.
Comparison of measurements of length with stages of development shows clearly that undue
importance must not be attached to size as an indication of maturity. The adolescent class contains
specimens, which judged by length alone would be regarded as fully aduh, but which on internal
evidence are far from mature. The smallest specimens in the adult class, on the other hand, might
well be regarded as being still adolescent.
It is interesting to note that although external and internal development do not necessarily keep
pace with one another, there is a very fair degree of coincidence of development in the majority of
cases. Thus most specimens of stage 1 are also at stage A, those at stage 2 are also at stage B, and so on
(Table 6). I would again emphasize the fact that the developmental stages are clearly defined: there
is no difficulty in determining to which stage a specimen belongs, so that evidence based on internal
and external anatomy, though laborious to acquire, is reliable.
PAIRING
The first records of fully adult males were obtained in September, when five specimens were found.
By October they appeared in larger numbers, and in this month the first females carrying spermato-
phores in the thelycum were noticed. These six females were not yet gravid, but were at stages 4 and 5,
the eggs being still in course of development. Pairing evidently takes place as soon as the females
have a fully developed thelycum, and it is generally possible to determine whether or not it has occurred
recently, because shortly after the spermatophores have been implanted, the sperm-mass passes into
the thelycum, leaving the spermatophores empty. The following table shows that of 556 females
carrying spermatophores, only 102 were full, showing that the migration of the sperm-mass into the
thelycum must be a rapid process.
Table 1 1 . Number of females with spermatophores
Month
No. of $'s
No. of $'s with spermatophores
Full
Empty
Torn away
Oct.
Nov.
Dec.
Jan.
Feb.
Mar.
Apr.
6
33
128
135
208
130
184
4
13
37
27
17
3
I
2
20
91
107
164
68
2
I
27
59
181
Total
824
102
454
268
556
ii8
DISCOVERY REPORTS
As a general rule, two spermatophores are found on each female, two being the number implanted
at each successful pairing, one from the right and one from the left ejaculatory duct of the male.
Sometimes, 4, 6 or 8 spermatophores occur, or 3, 5 or 7 may be found complete, with one broken
stalk attached as well. These are presumably cases of multiple pairing. I have seen as many as 6 empty
spermatophores on one specimen, but more often 2 or 4 will be full, as the thelycum is not large
enough to contain all the sperm-mass. Faulty implantation hardly ever occurs ; only one female was
found with spermatophores attached elsewhere than in the thelycum. This specimen carried 4 sper-
matophores, two fixed normally and empty, and two, full, on the base of the last right thoracic appen-
dage.
From December onwards to April, the majority of adult females have paired, as well as quite a
number of those nearly adult. The peak season comes in February, when the greatest number was
found to occur.
- SPAWNING
Spawning takes place when the eggs in the ovary have reached an average diameter of o-55-o-6o mm.
It would appear that the eggs take about four months to reach maturity.
Ruud distinguished three stages in the development of the egg, based on measurements of the egg
diameter: stage 1, diameter o-io-o-25 mm.; stage 2, diameter o-26-o-50 mm. and stage 3, diameter
0-5 1-0-70 mm. When stage 3 was reached, the eggs were spawned. I have measured the eggs from
1273 females and I obtained a slightly greater size range, with diameters lying between 0-05 and
0-70 mm. I first analysed the egg measurements from females caught in one season and in one area;
presumably they had all matured under the same conditions. It must not be supposed that all the
eggs in one ovary are of the same size. Great variation occurs, so I selected the smallest and the largest
for measurement and then worked out the mean diameter. As the southern summer advanced, the
monthly mean diameter in this one season was found to increase from o-o8 mm. in October, through
0-19 mm. in November and 0-32 mm. in December to 0-60 mm. in January, when the first females
which had spawned were also found.
Table 12. Average diameter in mm. of eggs per month
Month
Analysis of eggs measured
Av.
diam.
Class
I (stage 4)
Class
2 (stage 5)
Class 3 (stage 6)
Class
4 (stage 7 A)
of total
Av.
No.
Av.
No.
Av.
No.
Av.
No.
no. of
diam.
of?
7o
diam.
of?
%
diam.
of?
%
diam.
of?
/o
eggs
-
Aug.
0-09
0-07
31
86
0-20
5
14
—
—
—
—
—
—
Sept.
o-o8
0-08
33
97
0-14
I
3
—
—
—
—
—
—
Oct.
0-09
o-o8
17s
88
o-i6
25
12
—
— ■
—
—
—
—
Nov.
0-15
0-09
IS
16
o-i6
78
81
0-33
3
3
—
—
—
Dec.
0-23
o-io
12
7
0-19
114
63
0-33
52
29
0-50
4
I
Jan.
0-32
0-09
3
2
0-21
41
29
0-36
92
65
0-50
2
I
Feb.
0-46
—
—
—
0-20
10
5
0-40
45
20
°-57
48
22
Mar.
0-09
0-05
18
12
0-14
I
I
0-40
2
I
—
—
Apr.
fo-05
[0-48
0-05
20
9
—
—
—
°-45
6
3
0-S3
3
2
May*
June*
July*
No^$'s with measurable eggs.
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
119
Table 12 {co7it.)
No. of $'s examined
Month
With
eggs
1 Spawned
Grand total
No. of $
0/
,0
No. of I
0
0
No. of :
Aug.
36
■ 100
—
—
36
Sept.
34
100
—
34
Oct.
200
100
—
—
200
Nov.
96
100
—
—
96
Dec.
182
100
—
—
182
Jan.
138
97
4
142
Feb.
103
47
118
53
221
Mar.
21
14
129
86
150
Apr.
29
14
183
86
212
May
June
July
Total
839
434
1273
Using these figures as a guide, I then analysed the egg measurements from all the available material
(839 females)/ and I found that the eggs could be conveniently classified into four size groups, these
groups being used to differentiate the four internal stages in the adult females, stages 4-7. The first
group includes eggs measuring between 0-05 and 0-12 mm., with a mean diameter at o-o8 mm.
(class i); the second, eggs measuring between 0-13 and 0-24 mm., with a mean diameter at o-i8 mm.
(class 2); the third, eggs measuring between 0-25 and 0-48 mm., with a mean diameter at 0-36 mm.
(class 3) ; and the fourth, eggs measuring between 0-49 and 0-70 mm., with a mean diameter at o-6o mm.
(class 4), at which size the females are gravid and spawning occurs.
The analysis of these egg measurements, set out in Table 12, shows that, at first, development is
slow, but that, as the season advances, a steady rate of growth ensures that the eggs approximately
double their size each month, and take about four months to reach maturity.
The first measurable eggs occurred in females taken in August, and the majority of eggs in August,
September and October were in class i. In November, class 2 appeared in larger numbers and was
at a maximum in December. Class 3 appeared for the first time in any quantity in December and was
at a maximum in January. Class 4 was at a maximum in February, and in this month and in March
and April, the bulk of the females had spawned.
It will be noticed that in March the mean diameter of the eggs has dropped to the same figure as
in August and October, the majority of the females having spawned and those with measurable eggs
being mostly in class i. Again, in April, apart from the females which have spawned, there is a range of
small eggs. It is possible that these eggs never mature. I have found no females with measurable eggs
during the winter months of May, June and July, nor any full adults which have spawned. This mav
be due either to the sparse material from this time of the year, or else to the fact that the adults do not
survive the winter. One particularly striking fact emerges from this analysis of egg measurements:
class 4 containing gravid females is represented by very small numbers. The problem of the depth at
which spawning takes place has already been discussed, as having a possible bearing on the absence
of gravid females from the catch.
The evidence from these egg measurements points to a spawning season extending from January
to April, an increasing number of females having laid their eggs as the season advances. But if the
* These measurements are given in full in Table 20 in the appendix.
I20 DISCOVERY REPORTS
records of the occurrence of eggs are also taken into account, spawning begins a month or two earUer.
Eraser found eggs in the plankton from mid-November to March, and with this additional evidence,
it would therefore appear that E. sitperba lays its eggs over a period of 5I months.
AVERAGE GROWTH RATE
There are two ways of working out the growth rate of E. snperba : (i) by investigating the growth of
the developmental stages, and (2) by investigating the growth of the adolescent and adult population
as a whole.
(i) Grozvth of the developmental stages. By the first method, the total catch each month was sexed,
measured and divided into the seven growth stages, males and females of course being grouped
separately. These measurements are set out in detail in the appendix. Length frequency tables were
made for each stage and the average monthly lengths of the stages were calculated. The average length
of each stage for the whole year was also worked out (see Table 5, p. no).
These monthly and yearly averages for males and females are shown graphically in Fig. i . There is a
fair degree of correspondence between the lengths of the sexes at each stage, though on the whole the
males tend to be larger and grow more rapidly than the females. The growth of the females is slower
and steadier, but the same average length is reached in the final adult stage. The fall which occurs in
many of the curves in October and November, following a rise in September, may be due to the
stock being replenished, as the spring goes on, by a number of smaller specimens, whose growth has
been delayed by winter conditions. The general rise to a maximum in February and March, on the other
hand, occurs during the period of optimum conditions of food and temperature.
In both males and females, there is sometimes considerable overlap between the average lengths
of the stages. This is seen in Fig. i in which these average lengths are superimposed on one another.
Size is again shown to be an unreliable criterion of development, a point which has already been much
stressed.
In order to find out how long E. snperba takes to grow to maturity, I have worked out the frequency
of occurrence of the developmental stages expressed as a percentage of the catch each month. Table
13 A gives the actual figures^ and Table 13 B shows the months in which eggs, adolescents and adults
are at a maximum.
Eraser's work has shown that the spawning season is a long one. His results indicate that the
"greatest production of eggs is in November-December", when two large catches were recorded.
But he points out that once the eggs have been spawned, there are so many factors influencing their
dispersal, that a rather distorted impression of their abundance may easily be obtained, if it is based
on records of egg catches alone. The occurrence of gravid and spawned females gives a more reliable
picture. My results show that these are at a maximum in February, March and April, and it seems
reasonable to suppose that this will also be the period for the occurrence of the majority of the eggs.
Rough analyses made on board the R.R.S. 'Discovery II' during the ship's last two commissions
indicate that this maximum for gravid females and eggs is met with a month earlier, i.e. in January,
and extends through February to March. Annual variation may account for this, just as an early
southern summer is probably the reason for the large catches of eggs recorded by Eraser in November
1929. My results were obtained by examining material from many seasons and localities, and may
therefore be regarded as being characteristic of the average catch in average circumstances.
1 In the totals of stage 1, I have included the records of adolescents from Eraser's paper. He examined 124 young adol-
escents, and their measurements are given in Table 14. The majority of the lengths lie between 13 and 22 mm., and the
specimens can therefore be regarded as all belonging to stage 1. Males and females occur in the catch in approximately
equal proportions, so I have incorporated these lengths in the totals for stage i in both sexes.
) @ @
e3jt3jjt[];^-
:i
"-"="'I1'W
00
-3.
60
C
V
Q
i
.c
■d
(J
.j-
i
"^
y
^
_c
c
2
s
u
\.
C
"C
<u
■1
60
J3
60
'W
c
'2
s
'$
(A
<l>
o
o
V
rt
^
-I
*«
G
m
^d
£
Ji
^
oi
^
■■
•a
o
c
G
C3
-i
J3
"
, — ^
T-
Ln
U
•_^
'e
u
O-
^
U-t
o
w
s
s
<
-^
3
«
*
60
r
C
_c
.C
"^
u
^
[«
Q
u
s
o
a.
G
o
JS
J3
;
o
-
J3
v
1>
60
J,
60
60
y
rt
rt
V
S
u
u
<u
<ij
tn
a
>
>
o
^
< <
ex
G
2
o
5
^~'
■-^
iS
U
c5
<
m
;
h;
o
s
60
V
b
3-2
DISCOVERY REPORTS
■-5
5a
O
^
<u
O) jj
*
en
"is
1
o
o
m
>>
3
S?
1 1 8 S §- 1 S
1 t^ m 1 1 1 1
1 N r^ 1 1 1 1
o
1 1 „ „ N 1 «
m
1 '^°° 1 1 1 1
»H
u
C
3
1—1
55
1 « N vO II
■* o O 1 1 1 1
3
o
1 ^2 ;t" 1 1
■+00 "111!
N O 1 1 1 1
oo
OS
s?
M N CO CO 1 1 1
N lO -H III
so -J- O I 1 1 1
r*-; »n M 1 1 1 1
'a
+->
o
- =^N ^^ M 1
%■
N Tl-O 1 1 1 1
N ro 1 1 1 1
u
<
sO r'. O -t- 1
M vD
oo
-M
o
to lO u-l i-i N t-H -^
N " N O ^ ■-.
m
oo
m
OO M o O 1
OO r^ m N 1
vO so
oo
M
01
in
CO Ln -:j- N in ^ M
i
o^ o in en I
-1- N 1
1 "^
1 N
"(3
o
H
m t^ O " C^ -^ "-.
vO CO M -i- " "-
mo O oo M
r^ M m M
N M
N as
N
in
so
m
\0
O ^ M O " N lO
■o m Th 1 M
N m M 1
Tt- OS
13
o
N \0 "^ -t- N " m
M N Tj- ^ [ o
« m o N 1 -
^s
oo
00
d
E
r" N M M M o
mH O M
m 01
01
o
H
O Ti- N M M ro
N
in in N m M
01 ^O
OS
Os
oo
d
S^
C> m Ti- Th W M CO
-i- N O M M
sO N
O M
■M
o
>-' N m r^ -^ r^ CT^
O oo N rJ \o
N OO M N rt-
■+ t-H 1— 1
m M
in
m
<n
m
g
2
55
so " " :^ -J- « O
Cs m< 01 ro O
O N
" 1
"3
o
vD O CO Tt- ro in ■*■
m
in
m
r^oo CO moo
O M M f^
01
cn|
OS
OO
oo
o
O
^
ro t^ in r^j f^,0 r*-;
N r^; O M -+ 1 1
in Mm II
3
o
r^ in m r^ -^ o l^
vO m N M M oo \o
o
m
r^ in -^ m m 1 |
CO M in t^ 01
N M
so
m
in
1
-M
00 o m "^vo OO in
MM M n
•+ O m o M 1 1
r^ N ^ II
O
O t^ r^; -^ l^ C^ in
MM M rn
o
o r^ ^ m M
1 1
in
00
4-1
O
M M tJ- M t^vD 1
M N M N N 1
01 CO so 01 04 1 1
M 01 Ti- M II
O t^ -+- t^CO N I
M r^, N m -+ M
CO o O M m 1 1
N sO M ro I 1
o
01
u
an
1
^ M ro 'i- in o t^
13
c
^^ r.j ro ^ u-1 so r^
"c3
1
o
-a
C3
fe
3
?^
C
sr
<to
en
rt
<»
^
O
^
pq
a.
So
5r
tuO
3
<
a,
<
X ■ X X
X • X X
X • X X
X X
• X X
X I
X • X X
X • X X
X • X X
• X
a-
X X
X X
• X X
• X
X • X X
X X
X ■ X X ■ ■ • X
g^^H tN m ■+ m so t^
X
X X
• X
X
1 X
■ X
X •
X •
X •
X
X •
X ■
X
X ■
X
• • • ■ X • ■ ■
• X
X
•
• • 1 X • • ■ •
• X
X
• • X
X
• 1
•
•
• X
X
X X
■ X
1x
1 X
■ X
tM
X •
•
X •
X ■
X
X ■
X •
X
X •
X
• ■
• •
X
•
• X
X
•
• •
■ 1
X
• X
X
•
• • X
X
• 1
•
• X
X
X X
• X
X
1 X
•
■ X
So^^ cs^j en ■+ in so r^
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
Table 14. Adolescent lengths : from Fraser's report
123
Length
in mm.
Aug.
Sept.
Oct.
Nov.
Dec.
Jan.
Feb.
12
5
—
—
—
—
—
—
13
7
3
2
—
—
—
—
H
I
5
7
4
—
15
16
z
3
I
9
13
8
—
—
17
—
—
I
15
—
18
—
I
13
—
19
I
—
—
5
I
I
20
—
—
—
2
I
I
21
—
—
5
I
22
—
—
2
I
I
23
24
25
26
—
—
—
I
I
I
27
—
—
—
—
—
28
—
—
—
—
—
—
29
—
—
—
—
I
Total
14
12
33
54
4
5
2
NB. Measurements approximate to nearest whole number.
Table 15. Analysis of the adult female catch per month from later commissions (rough analysis)
Month
Stage 6
Stage 7 A (gravid)
Stage 7 B (spawned)
Total $
No. of ?
0/
/o
No. of ?
0/
,0
No. of $ %
No. of ?
Oct.
I
100
—
—
—
—
I
Nov.
—
—
2
100
—
2
Dec.
10
100
—
—
—
—
10
Jan.
5
I
609
98
2
I
616
Feb.
13
32
13
32
15
36
41
Mar.
12
8
50
35
83
57
145
Apr.
—
—
I
100
I
Eraser has established that growth from the egg to the ist adolescent stage occupies approximately
7-9 months. On this basis, if the eggs are most often met with in February, March and April, the
maximum occurrence of stage 1 should be in October, November and December. It is significant
that these are just the months when this maximum is encountered, with an extension into January
in the case of the males, and into March in the case of the slower growing females (Table 13 A and B).
The rest of the period of growth from stage 1 to maturity occupies another 12-16 months in the
males and 12-18 in the females. Ruud showed that the life cycle of E. superba occupied at least two
years. He drew attention to the two classes of kriU recognized by the whalers as "blue whale" and
"fin whale " kriU respectively, and diagnosed them as representing the two age groups, adolescent and
adult. But owing to lack of material, and not having devised a method of gauging the stage of develop-
ment of the adolescent males and females, the growth curve which he figures is too steep, and does
not give an accurate idea of the growth rate throughout the whole of the growth period.
Ruud suggested that after pairing and spawning the adults died off. I think this is very likely, for
although at the end of April I found females, which had spawned and were feeding actively (a thing
they are unable to do while gravid), no fully adult females or males were found between May and
October. This may of course be due to lack of material, which in all commissions of the R.R.S.
124 DISCOVERY REPORTS
'Discovery II ' has been scarce at this time of the year owing to the difficuhy of fishing nets in the
prevaihng weather and ice conditions, but adolescents have been obtained throughout the year, and
it is the half-grown specimens belonging to stages 3, 4 and 5, which are predominant during the winter
months. Young adults do not appear until September and full adults not before October. On the
evidence available, therefore, it is reasonable to conclude that adults do not exist in the catch after the
pairing and spawning season is over, because they have died out.
It would be interesting if the time of duration of each stage could be established with some cer-
tainty, but this is not so easy. Growth in the female is slower than in the male, and it seems likely that
the earlier stages anyhow take longer to pass through. There are indications in the table of maximum
frequency of occurrence of the stages (Table 13, p. 122) that each stage lasts 2 months in the male
and 2i months in the female, and on this basis it is possible to work out the months in which each
stage should be theoretically at a maximum. Supposing that: (i) the spawning period extends from
November to April, i.e. 5^ months, (2) the adolescents first appear in August, i.e. 9 months later,
(3) the maximum spawning period occurs in February, March and April, and (4) each stage lasts
2 months in males and 2^ months in females, then, in theory, the maximum occurrence of each stage
should be as in Table 16:
Table 16
Eggs
Stage 1
Stage 2
Stage 3
Stage 4
Stages
Stage 6
Stage 7
Male:
Nov. eggs
Dec. eggs
Jan. eggs
Feb. eggs
Mar. eggs
Apr. eggs
End of Aug.
End of Sept.
End of Oct.
End of Nov.
End of Dec.
End oi Jan.
Oct.
Nov.
Dec.
Jan.
Feb.
Mar.
Dec.
Jan.
Feb.
Mar.
Apr.
May
June
Feb.
Mar.
Apr.
May
June
July
Aug.
Apr.
May
June
July
Aug.
Sept.
June
July
Aug.
Sept.
Oct.
Nov.
Aug.
Sept.
Oct.
Nov.
Dec.
Jan.
Feb.
Female :
Nov. eggs
Dec. eggs
Jan. eggs
Feb. eggs
Mar. eggs
Apr. eggs
End of Aug.
End of Sept.
End of Oct.
End of Nov.
End of Dec.
End oi Jan.
End of Mar.
Beg. of Nov.
Beg. of Dec.
Beg. of Jan.
Beg. of Feb.
Beg. of Mar.
Beg. of Apr.
Beg. of May
Beg. oi June
Beg. of Aug.
End of Jan.
End of Feb.
End of Mar.
End of Apr.
End of May
End of June
End oi July
End of Aug.
Beg. of Apr.
Beg. of May
Beg. of June
Beg. of July
Beg. of Aug.
Beg. of Sept.
Beg. of Oct.
End of June
End of July
End of Aug.
End of Sept.
End of Oct.
End of Nov.
End of Dec.
Beg. of Sept.
Beg. of Oct.
Beg. of Nov.
Beg. of Dec.
Beg. oi Jan.
Beg. of Feb.
End of Nov.
End of Dec.
End of Jan.
End of Feb.
End of Mar.
End of Apr.
The months, in which the stages have actually been found to be at a maximum in practice, are
already known (Table 13), and have been italicized in, or (where necessary) added to, the above table.
There is on the whole a very fair degree of correspondence between expectation and reality, especially
when it is remembered that the material used came from very many seasons and localities. It is
interesting to note that, in practice, growth is slower during the winter months than it is in theory,
the actual maxima occurring later in the season than expected.
Of course the stages overlap one another in occurrence; this is the outcome of the protracted
spawning season. Some idea of the extent to which this takes place can be obtained by tracing the
growth of the generations arising from eggs spawned in different months. Suppose that the months
considered are those in which the eggs are known to be at a maximum, namely February, March and
April. We know that these eggs can, under optimum conditions, reach the first adolescent stage in
7 months, though normally they take an average of 8-9 months to do so. In each successive batch of
ill ^
Hi
<s=
UJ
UJ
_i
<
Ld
V
<
^
n
>^
<
>
1
s
■g.
cQ
<
UJ
-!3
Lu
C
C3
;7
a
<
ii 1
2
Q
fa
■T3
c
^
-a
■-n
^
c
s
•^
c3
o.
H
«
n
( )
c/5
<u
<-)
M
bJO
1)
JS
ti-i
■M
f-
O
0
n
K
LlJ
J3
1
JS
a
0
XI
a
UJ
<
r
<S
ifl
^
•T3
a
0
Z)
Si
J
JS
3
S
o
0,
UJ
0
a.
^
u;
3
0
>
<
2
1)
-o
c
a
0
'r?5
o
0
a.
D:i
2
0
Q_
d
u
<
6X3
-1
Qi
C
0
C
(1)
<
3
2
3
OJ
■u
j::
CD
(U
UJ
-Q
45
U_
h-
T"
1;
<r
0
-i
u
bC
111
a
u
0
"35
>
n
6
z
1-
Q
( J
,
o
feb
1—
fa
Q_
h
U)
J1
in
en
Cvl
in
in
in
in
S3yx^^Nl"^ll/^
S3yi3i^nniiAi
126 DISCOVERY REPORTS
eggs, therefore, the distribution in time of each developmental stage may be spread over a period of
three months. Thus, eggs spawned in February may become adolescent in the foUow^ing September,
October or November, that is, specimens at stage 1 met with in these months may have originated
from eggs laid in February. Similarly, eggs spawned in March may become adolescent in the following
October, November or December, or alternatively, specimens at stage 1 in these months may have
developed from eggs laid in March and so on.
The months, in which the other developmental stages may be theoretically expected to occur, can
also be worked out, by assuming that each stage lasts 2 months in the male and zh months in the
female. In this way, an explanation of the heterogeneous composition of the euphausian population
is obtained, and in Fig. 2 I have attempted to give some idea of this complexity at any given time by
a diagrammatic representation of the generations arising from the three batches of eggs, spawned in
Table 17
I
Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
Stage 6
Stage 7
Males:
time interval : 2 months
End of Sept.
End of Nov.
End of Jan.
End of Mar.
End of May
End of July
End of Sept.
14 mm.
End of Oct.
27 mm.
End of Dec.
42 mm.
End of Feb.
47 mm.
End of Apr.
End of June
End of Aug.
54 mm.
End of Oct.
15 mm.
End of Nov.
28 mm.
End of Jan.
43 mm.
End of Mar.
46 mm.
End of May
End of July
44 mm.
■ End of Sept.
44 mm.
End of Nov.
20 mm.
End of Dec.
33 mm.
End of Feb.
44 mm.
End of Apr.
41 mm.
End of June
End of Aug.
48 mm.
End of Oct.
45 mm.
End of Dec.
23 mm.
End of Jan.
33 mm.
End of Mar.
40 mm.
End of Mav
41 mm.
End of July
40 mm.
End of Sept.
43 mm.
End of Nov.
51 mm.
End of Jan.
23 mm.
33 mm.
38 mm.
—
41 mm.
41 mm.
51 mm.
Females :
time interval: 2k
months
1
End of Sept.
Beg. of Dec.
End of Feb.
Beg. of May
End of July
Beg. of Oct.
End of Dec.
18 mm.
End of Oct.
31 mm.
Beg. of Jan.
36 mm.
End of Mar.
Beg. of June
End of Aug.
Beg. of Nov.
57 mm.
End of Jan.
15 mm.
End of Nov.
34 mm.
Beg. of Feb.
43 mm.
End of Apr.
Beg. of July
46 mm.
End of Sept.
47 mm.
Beg. of Dec.
54 mm.
End of Feb.
20 mm.
End of Dec.
32 mm.
Beg. of Mar.
40 mm.
End of May
Beg. of Aug.
49 mm.
End of Oct.
46 mm.
Beg. of Jan.
52 mm.
End of Mar.
23 mm.
End of Jan.
38 mm.
Beg. of Apr.
41 mm.
End of June
39 mm.
Beg. of Sept.
45 mm.
End of Nov.
50 mm.
Beg. of Feb.
50 mm.
End of Apr.
25 mm.
33 niTi-
41 mm.
42 mm.
43 mm.
50 mm.
47 mm.
February, March and April. The average length per month of each stage has been plotted, and the
points marking the maximum average values have been joined up, as well as those marking the mini-
mum average values. The space between has been hatched in colour: blue for the generation arising
from February eggs, red for March and green for April. The months in which the stages have been
calculated to occur, and the values of the average lengths are set out in Table 17. It will be noticed
that in the winter months no values are given for stage 5 in the males and stage 4 in the females. In
the scanty material available from this time of the year, these stages did not occur, although theoreti-
cally they should be present. In the figure the actual period of time, in which each stage appears, is
shown by a solid black line, the theoretical period by a broken line.
The fate of early or late spawned eggs can be seen at a glance from the diagram. February eggs,
which have reached stage 1 at the beginning of the following season (i.e. September to November)
will have grown sufficiently to be at stage 4 or 5 before the winter sets in (i.e. June), and will be mature
by October or December of the succeeding spring ; they attain rather greater lengths than the later
generations. On the other hand, April eggs may not reach stage 1 until the following January, and will
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL 127
not therefore be approaching maturity, that is at stage 5, until the succeeding September, nor will
they be fully mature before the January (or April) after that. The overlap of the stages in the three
batches of eggs and the resulting mixed composition of the euphausian population is clearly shown.
The picture could be made even more complex, if the batches of eggs from every month in the whole
spawning season were represented, but I decided not to attempt this, because the diagram would lose
in clarity.
(2) Grozvth of the etiphmisian population as a whole. By the second method mentioned on p. 120,
the population was sexed, measured and divided into adolescents and adults, males and females again
being treated separately. The average monthly lengths of adolescents and adults were then calculated
(Table i8).
Table i8. Showing average length per month of larval, adolescent a?id adult Euphausia superba
Month
Nov.
Dec.
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Jan.
Feb.
Mar.
Apr.
Average length per month of
larvae and adolescents compiled
from Fraser's paper
Larvae
Av.
length
mm.
I
I
4
4
S
6
8
10
10
II
13
13
16
No. of
speci-
Adolescents
Av.
length
mm.
5
5
56
lOI
182
177
18
3
29
27
33
49
No. of
speci-
Average length per month of
adolescent and adult males
Adolescents
Total
13
14
15
18
20
23
24
Av.
length
mm.
No. of
speci-
mens
686
14
12
33
54
4
5
2
13
14
19
23
26
29
36
38
39
33
39
40
34
34
Adults
Av.
length
mm.
No. of
speci-
mens
Average length per month of
adolescent and adult females
Adolescents
Av.
length
124
19
19
356
316
437
304
778
330
368
40
39
4
146
41
3197
44
47
44
45
51
51
52
52
SI
12
44
147
39
176
133
164
19
15
13
18
18
21
23
25
31
35
34
31
39
35
32
27
No. of
speci-
mens
Adults
Av.
length
No. of
speci-
Month
749'
28
30
356
293
351
252
660
415
309
62
93
II
176
21
3057
40
42
41
41
45
48
51
50
47
36
34
200
96
182
142
221
150
212
1273
I adult in July neglected.
Larvae = nauplius to 6th furcilia.
Adolescent males = stages 1-5 ( + Fraser's adolescents).
Adolescent females = stages 1-3 ( + Fraser's adolescents).
Adult males = stages 6-7.
Adult females = stages 4-7.
Nov.
Dec.
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Jan.
Feb.
Mar.
Apr.
Total
128 DISCOVERY REPORTS
To obtain a complete growth curve, monthly averages of the larvae must be included. Using
Eraser's measurements, I recalculated these averages, which had been originally worked out on a half-
monthly basis, and I also made monthly frequency tables of his measurements of young adolescents.
These larval averages show the rate of growth during the first six or seven months, but when the
adolescents make their appearance in August, it is not sufficient to work out the average length of all
adolescents per month. Some selection is necessary, because early in the southern spring, that is, in
August, September and October, the overlap of generations brings about the co-existence in the catch
of young adolescents of stage 1 with late adolescents of stages 3, 4 and 5 of earlier generations, and the
inclusion of these larger adolescents in the calculations gives a wrong idea of the growth rate in these
particular months. Later in the season, the population becomes more sharply divided into adolescents
and adults, and the question of selection does not arise. I have, therefore, included in the calculations
for August, September and October, only the measurements of Eraser's adolescents and of my own
specimens at stage 1 . In calculating the average lengths of the adults, I have used all specimens which
could be expected to mature within the southern summer, that is stage 6 as well as stage 7 in the males,
and stages 4-7 in the females.
In Eraser's original graph of larval growth, there is a marked decrease during the winter months,
June, July and August. This tends to disappear when his results are combined with mine (Fig. 3),
and may have been due, in part, to scarcity of material. I think, too, that the apparent slowing-up of
growth during the second winter, in the transition period between adolescence and maturity, can also
be partly explained on these grounds, although the colder temperatures and less abundant food almost
certainly have some retarding effect upon the growth rate.
Before one year's growth is over, that is, as soon as the adolescents appear in August, it becomes
possible to distinguish between males and females, and the curve can therefore be divided into two
parts (Fig. 3). The rate of growth in the two sexes is very similar. Although the females are con-
sistently smaller than the males, the two curves follow approximately the same course. The period of
adolescence occupies, at a minimum, a whole year and is shorter in the males than in the females. In
the males, true adults, carrying fully formed spermatophores, appear for the first time in September.
In the females, true adults, fully gravid, appear three months later in December. The total period of
growth from the egg to the adult occupies a minimum of twenty-two months in the male, and twenty-
five months in the female.
FACTORS INFLUENCING GROWTH RATE
Obviously, the main factor which influences the growth rate of E. superba is the supply of food. Hart
(1934) writes that this "consists very largely, if not entirely of diatoms and other phytoplankton
organisms". He found that the most strongly silicified diatoms could be identified with certainty in
the stomach contents, but that those with thinner cell walls were too rapidly digested to be easily
recognizable.
In a later paper (1942), Hart discusses the factors which control the production of phytoplankton
in the Antarctic zone as a whole. He states that chief among them are the physical influences of
"light, the degree of stability of the surface layers and the (interrelated) effects of the pack-ice", and
that these three agents are "certainly the prime causes in determining the time of the onset of the
main increase" in the abundance of phytoplankton. This time, " falls later in the year as one proceeds
southwards ", as much as two months elapsing between its occurrence in the northern and the southern
regions of the Antarctic zone. However, Hart considers that none of these factors " adequately accounts
for the vastly greater richness of the neritic areas as compared with the oceanic regions". Recent
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
129
work strongly indicates that this is due to minute traces of organic compounds, iron and manganese,
derived from the land, which exert a "strongly favourable influence on phytoplankton production .
The importance of the pack-ice in this connexion is strongly emphasized. Hart regards it as "giving
rise to what might be termed pseudo-coastal conditions at vast distances from land, where neritic
species maintained by the ice flourish for short periods when the latter disperses ".
All these factors, since they influence the food supply, must have a bearing on the growth rate of
E superba but it is not a simple matter to produce evidence in support of this. Only an unbroken
series of observations extending over several seasons and made at short regular time intervals at the
same stations would provide reliable data. Unfortunately such a series is not available, the material
collected being too scattered and interrupted, so that there are many gaps in the chain of evidence
and attempts at correlation are always breaking down.
1
•so
50
40-
"30
20
..^^.•^- \-':>
\ U''^
V^o',..
Larvae -* A: — a — ^ —
Males — ■ — ■
Adolescents ■ ■
Adults IS 12 <-J
Females
Adolescents a a
Adults B a B
50
40
■30 E
2Q
Notf
First Year ot browth I
Fig. 3. Growth curve showing average length per month of larvae, adolescents and adults.
As the diatom maximum occurs earlier in the northern region of the Antarctic zone than in the
southern the average development of the northern E. mperba should be correspondingly more
advanced. But Fraser found no clear indication that larval development begins sooner in one area
than in another, though he obtained some suggestions that local variations in the abundance of food
may directly affect the average larval length. He did not feel justified however in concluding that
within a restricted area, food was the only factor involved, but decided rather that the effect was the
cumulative result of several factors acting locally. ^ ' ■ a .
Comparison of the size of larvae, adolescents and adults fron, the different Anrarcc regions do
not give a satisfactory result either. This may be due to the fact that length alone ts not a rehable
criterion of development, but even if the developmental stages are taken mto account the evidence s
n"t more definite From the material available, it cannot be shown that older adolescents or fuHy
mature aduks occur any earlier in the northern Antarctic region than in the southern. But on Ae
other hand, there is evidence to show that those E. ^perbo, which are hatched early anywhere wthm
"arc ic zone, are directly intfuenced by the abundance of the food supply. Hart points out h
he summer decrease in phytoplankton may be due in part to a ^^-^^-^^ ^^^\°^XXX
this is probably brought about to some extent in the oceanic areas, anyhow, by intensive^ graz ng
down by the herbivorous zooplankton". This occurs during and ™™*ately after the period o he
spring maximum, in December, January and February. These months "made with he fi.t ha^ o^
the spawning season in E. superba, and it would appear that those generations hatched early, which
I30 DISCOVERY REPORTS
are able to benefit fully from the spring maximum, reach greater average lengths than those developing
later, when grazing down has brought about a decrease in the food supply. This variation in size has
already been mentioned (p. no), and Fig. 3 shows that it is maintained throughout the life-cycle,
these larger adolescents giving rise to the very big adults, which occur at the beginning of the breeding
season.
The influence of the spring diatom maximum on the average lengths of the stages is also apparent,
though it is perhaps more consistently marked in the males than in the females. If plotted graphically,
the average lengths show a fairly rapid rise from October or November to a peak in February or
March (Figs, i and 3), after which they generally tend to decrease slightly or to remain almost sta-
tionary. The period of increase in length corresponds roughly with the time of the phytoplankton
maximum, and the succeeding period of slackened growth corresponds with the time of the post-
maximal decrease. A similar rise and fall at the same time of the year is seen in the maximum lengths
of the stages. The onset of the southern winter is doubtless also a factor which comes into play at this
time and influences the rate of growth, for Deacon (1933) has shown that the difference between the
summer and winter temperatures of the Antarctic surface water is as much as four degrees.
The other factors mentioned by Hart, light, surface conditions and pack-ice, except in so far as
their broad seasonal variations will certainly influence the growth rate, more properly affect the dis-
tribution of E. superba, and are outside the scope of this paper. This distribution was being worked
out by my colleague, J. W. S. Marr, but unfortunately its completion has been interrupted for the
time being by the war.
CONCLUSIONS
This investigation extends Fraser's work on the growth of E. superba from the egg to the beginning of
adolescence, and amplifies Ruud's sketch of a two year life-cycle.
In order to estimate accurately the composition of the euphausian population, a method was devised,
by intensive study of the reproductive system, for determining the degree of maturity of each in-
dividual. It was found possible to distinguish between males and females immediately the larval state
was left behind, and to divide the period of their growth to maturity into 7 stages. This method gives
a convenient way of checking deductions based on measurements of length alone, and has shown
clearly that individual length is not necessarily a reliable criterion of development, since there is
evidence to show that length may be the first thing to be influenced by variations in the factors
affecting the rate of growth. Division of the population into growth stages, combined with estimations
of the average length of these stages, however, gives a good idea of the life history.
The spawning season, which extends over 5I months, begins in November or December. Eggs
spawned then are probably adolescent by August, and mature about thirteen months later in Sep-
tember and October. The males grow more rapidly than the females, attaining slightly greater average
lengths on the whole, and requiring a probable minimum of 22 months to reach maturity, as against
25 months in the female. It seems likely that each state lasts 2 months in the male and 2\ months in
the female, though these times are not definitely established as yet, lack of material at certain periods
of the year making the evidence too scanty.
Pairing was first found to take place in October, before the females were fully adult. The sper-
matophores are therefore carried for some time before fertilization can occur, the evidence showing
that this is effected externally, while the eggs are being laid. Gravid females are present in surprisingly
small numbers. This fact, coupled with Fraser's records of eggs and early larval stages in the deeper
water layers, seems to indicate that the females go down deep to spawn. More evidence on this point
is needed, before this can be definitely established.
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL 131
The prolonged spawning season, which is characteristic also of euphausians from the northern
hemisphere, gives rise to a very heterogeneous population, the stock being continually replenished
by the addition of new generations. Adults were found between the months of August and April, but
when the breeding season is over, they appear to die off, being absent from the catch during the
autumn and winter months. It should be mentioned, however, that after spawning, females were
found in April at the surface feeding actively, a fact which does not suggest lack of vitality, and
therefore, since after this month the available material becomes very scanty, their apparent absence
may simply be due to lack of evidence, and not to a holocaust consequent on exhaustion after breeding.
The material available is not enough to show a correlation between the appearance of the spring
phytoplankton maximum in the different regions of the Antarctic zone and the precosity of develop-
ment of the euphausian population, but there is evidence to show that generations hatched early in
the season anywhere in the zone benefit directly from the abundance of food and the rising tem-
perature of the surface layers, while later generations develop more slowly, partly no doubt because
the food supply is becoming reduced by grazing-down, and partly because of the onset of the colder
weather.
BIBLIOGRAPHY
Bargmann H E 10^7. The reproductive system of Euphausii superb^. Discovery Reports, xiv, pp. 325-50, 5 pis., 26 figs.
Son; G. E: R., 1933- A general account of the hydrology of the South Atlantic ocean. Discovery Reports, vii, pp. 171-238,
Eraser, R^C.r"9T6. On the development and distribution of the young stages of Mil (Euphausia superba). Discovery Reports,
Hart, i:^]'.^J^2lon'the phytoplankton of the South-west Atlantic and the Bellingshausen sea, 1929-31- Discovery Reports,
viii, pp. 1-268, figs. 1-84. A A <; ^^
IQ42 Phytoplankton periodicity in Antarctic surface waters. Discovery Reports, xxi, pp. 261-350, hgs. i-9-
Hendey N I., 1937. The plankton diatotns of southern seas. Discovery Reports, xvi, pp. 151-364. Pls-/™.
Lebou? M. v., 1926. A general survey of larval Euphmmids, with a scheme for then tdenttficatwn. J. Mar. Biol. Assoc. N.S.
MacdonaTd' R^" lltr''Food% habits 0/ Meganyctiphanes norvegica. J. Mar. Biol Assoc. N.S^ xiv, pp. 753-84, 2 figs.
Ottes?ad P 1933 A mathematical method for the study of growth. Hvalradets Skrifter, Nr. 7, Oslo, pp. 30-54, Ags. 24-35-
RuuD T T 10^2 0«fAefo-o/oavo/w«/te-«Euphausiidae. Hvalradets Skrifter, Nr 2, Oslo pp. 5-105, 37 %».
— ^^6 £«?Wac... Report on the Danish Oceanographical Expeditions 1908-10 to the Mediterranean and adjacent
Sars, G. 0?ir98."b^/°S^o?a^a;fo« and early development of Euphausiidae. Arch. Math. Natur. Kristian.a, xx, Nr. 11,
Taube, l^'i^ls^^Beftrdge zur Entwicklungsgeschichte der Euphausiden. Zeit. wiss. Zool. cxiv, pp. 577-656, 7 pls-, 7 figS-
132
APPENDIX
Table 19. Measurements of all Specimens of Euphausia superba examined
MALES
FEMALES
Date August 28, 1928 Locality S. Georgia
St. No. WS264 Position |53°l3'-3pS
Net N7oB97-om. position ^ 34°5i'.ooW
Surface T. -1-65° C.
Date August 28, 1928 Locality S. Georgia
St. No. WS264 Position (53°"3'-3pS
Net N7oB97-om. i-osition -^ 34°5i'-ooW
Surface T. - lbs" C
Length
in mm.
Stages 1
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
II
12
13
14
22
23
24
25
26
27
28
30
32
33
34
35
36
11
39
41
42
43
45
46
47
51
• A
. a
. 3
1
1
1
2 .
2 .
I
1
4 ■ •
5 2 .
3 5 •
2 I
. 2 3
I
I
.
I .
2
2
I
I
I
I
4
3
I
I
5
4
2
2
5
2
I
I
I
4
2
2
I
I
I
10
II
12
13
14
21
22
24
25
26
27
28
29
30
31
32
33
34
11
39
40
41
42
43
45
55
2
2
4 .
4 .
I
. IC
1 . . . .
2 . . . .
I . . . .
21...
22...
21...
I
. 21..
2 .
I
2
2
4
4
I
I
2 . . .
4 • ■ •
2 . . .
3 • • •
12..
I . . .
I
. 3 . .
2
I . .
2
2
4
4
I
I
I
3
10
I
I
I
2
3
3
I
3
4
3
3
3
I
I
I
3
2
I
Total
Av. length
5 25 4 8 13 9 •
14 25 32 32 37 44 •
18 13 5 5 14 9 •
22 26 32 32 37 44 •
64
Total
Av. length
13 23 10 14 5 ■ •
12 26 33 37 46
13 8 21 13 10 .
12 25 28 35 43 . .
6s
Date August 16, 1938 Locality S. of Bouvet L
St. No. 2391 Position (55° 03'-3 S,
Net NiooHs-om. fosmon | 00° 21' E
N 100 B («°-22S m. Surface T. -132 C.
\700-0 m.
Date August 16, 1938 Locality S. of Bouvet I.
St. No. 2391 Position |"°°o'''?.%
Net NiooHs-om. I 00 21 E
NiooB ,f43c^225m. Surface T. -i-32''C.
(700-0 m.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
25
35
38
42
43
44
47
1 . I . .
. . . 2 . .
I
I
I
I
I . I . .
I
2
2
I
I
I
I
29
31
35
36
37
38
39
40
44
45
I
I . . . .
1 . . . .
2 . . . .
1 . . . .
2 . . . .
1 . . . .
2 . . . .
I . . . .
1 . . . .
I
I
I
I
I
2
I
2
I
2
I
I
13
Total
Av. length
. I I . 7 • •
. 25 35 .41 . •
.1116..
■ 25 35 38 42 . .
9
Total
Av. length
I 12 .
. 29 38 . . . .
.265...
. 30 37 41 • ■ •
Date August 17, 1938 Locality S. of Bouvet L
St. No. 2393 Pnsitinn /56" 42'-3 S,
Net N 100 H 5-0 m. Position | ^^, ^g,.^ j.
NiooBi28-om. Surface T. -l-8i°C.
Date August 17, 1938 Locality S. of Bouvet I.
St. No. 2193 Position |56°42'-3S,
Net NiooHs-om. Position ^ 00° 38'-3 E
NiooBi28-om. Surface T. -i-8l°C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
2S
30
31
32
33
34
35
36
39
I
1 2
2 2
I
I
I
4
4
I
I
I
I
I
I
26
27
28
30
31
32
33
34
35
38
39
40
I
I
I
51....
3
1 . . . .
2
II....
12....
2 . . . .
I . . . .
2 . . .
I
1 I
I
6
3
I
2
2
3
2
I
I
Total
Av. length
. 3 5 6 3 • •
• 30 33 34 44 ■ •
• 4553- •
• 30 33 34 44 ■ •
17
Total
Av. length
. 15 9 • • • •
. 31 36 . . . .
2 12 7 3 ■ • ■
27 31 34 39 • • •
24
'
Date August 18, 1938 Locality S.E. of Bouvet L
St. No. 2396 Pn.lfmn ,<sb°lf7S,
Net N 100 B 109-0 m. Position ^ ^^. ^^.^ g
Surface T. -i-65°C.
Date August 18, 1938 Locality S.E. of Bouvet \.
St. No. 2396 Position |S6° I7'-7,S,
Net N 100 B 109-0 m. l-osition ^ 03° o7'-9 E
Surface T. -I-65°C.
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
Length
in mm.
Stages
Total
in
sample
40
42
43
'.'.'.'. \ '. '.
1 . . .
. . . . I . .
I
I
I
1234567
A B C D E F G
40
I . . .
I . . .
1
Total
Av. length
. . . . 3 . .
.... 42 • •
. . . I z . .
■ ■ . 40 43 • •
3
Total
Av. length
. 1 . . .
. . . 40 • • •
. . . 40 . . .
I
1 HE DEVELOPMENT AND LIFE-HISTORY OF KRILL
133
MALES
FEMALES
Date August 1 9, 1938 Locality S.E. of Bouvet L
Date August 19, 1938
Locality
S.E. of Bouvet I.
St. No. 2399 Position (^4° 47 -i.S,
Net N 100 H 5-0 m. rosmon ^ ^^„ j, .^ g
St. No. 2399
Net N 100 H 5-0 m.
Position
(•54° 47'- 1 S,
\ o6°3l'-3E.
n; .„„ n /98-0 m. Surface T. -1-71° C.
NioobI^^-®""-
Surface T.
-1-71^ C.
\3oo-:50
m.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length
in mm.
m
sample
1234567
A B C D E F g!
123456
7
A B C D
E F G
29
30
2 I
I
. 3 . . ■
I . . .
3
I
11
I
I
I
:
29
3
33
I
I
I
30
3
35
3 ■ -
. 3 -
3
32
36
2 . .
2 .
2
33
37
3 I •
I 3 •
4
34
3
39
I 1
2
35
3
40
I
36
2
2
43
. . . . I . .
I
40
42
. 2
I
2 . .
2
I
2
Total
.2395-
.4573.-
19
Av. length
. 29 35 36 41 ■ ■
1
2 3 10 6
26
5 • •
Av. length
• 30 35 38 . .
28 29 33 36 39 • • 1
Date August 22, 1938 Locality E. of Bouvet 1.
Date August 22, 1938
Locality
E. of Bouvet I.
St. No. 2408 Position (54°52'-4SS,
Net NiooHs-om. rosition ^ 16° 22'-2 E
St. No. 2408
Net N 100 H 5-0 m.
Position
/54° 52'-45 S,
I l6°22'-2E
nI°o°b}'°^-°"^- Surface T. -f2i°C.
N 100 Bl „ „
N70B )-°8-°m
Surface T.
-1-21° C.
Stages
Total
Stages
Total
Length
in mm.
m
sample
in mm.
in
sample
1234567
A B C D E F G
1234567
A B C D
E F G
30
12....
■ 3
3
25
I . . . .
I . . .
I
I . . . .
I
27
.
32
1
I
1
30
33
2 .
2 .
2
31
34
I 2 .
2 I
3
32
1 4 .
I 4
35
36
_
• 3 ■
I 2
I
3
I
33
34
I 4 .
6 .
I 4
. . 6
6
37
I
.
I
35
J 6 .
10
40
. I 3
22..
4
36
6 .
I 4
I
6
41
2
2 .
2
37
I 3 •
4
. 4
. 13..
4
38
3 •
I
2 .
3
44
I
I
I
39
3 I
4
45
. .31.
. . 4 . .
4
40
3 I
2
2 .
4
47
I
I
I
41
I 4 •
4 • ■
5
42
I 2 .
1
2 .
3
Total
. I 4 10 14 2 .
■ 3 7 7 13 I •
31
Total
I 10 40 8
I 3 8 33
14 . .
59
Av. length
. -io 12 -14 42 46 .
- 30 33 37 42 47 -
Av. length
25 33 36 41 •
25 30 35 35
40 . .
Date August 23, 1938
Locality
E. of Bouvet I.
St. No. 241 1
Net N 100 B 35-0 m.
Position
(56°25'S
X 19° 54 7 E
Surface T.
-1-70° C.
Stages
Total
Length
in mm.
m
sample
I 2345
6 7
A B C D
E F G
26
. 1 . . .
I . . .
I
27
I . . .
I . . .
I
36
. 2 .
I I
2
Total
2 2..
2.11
4
Av. length
. 27 36 . .
27 . 36 36
Date August 24, 1938
Locality
E. of Bouvet I.
St. No. 2412
Net N 100 H 5-0 m.
N 100 B 107-0 m.
Position
fS5°4l-9'S,
\ 20^ 29-4 E
-i-8o=C.
Date August 24, 1938 Locality E. of Bouvet 1.
Surface T.
St. No. 2412 Position /5S°4I-9'S,
Net NiooH5-om. l-osition ^ 20= 294 E
Stages
Total
N 100 B 107-0 m. Surface T. — i-8o^C.
Length
in mm.
in
sample
12345
ft 7
A B C D
E F G
Length
in mm.
Stages
Total
in
sample
25
27
28
29
I
■ 3 ■ ■
I
3
1234567
A B C D E F G
■ 3 ■
. 2 .
27
I
I
I
I
I
28
i
I
30
• 4
4
30
12,.
2 I
3
31
• 3
. 2 I
3
31
I I
I I
2
32
1 4
2 2 I
5
32
21..
. 3 . .
3
33
• 5
• 5 ■
5
33
2 I
. . 3 ■
3
34
. 2
I I
34
. I I .
2 .
2
35
• 3
3
35
2
2
^2
. 2
37
. . I
I
38
I
I
38
I
I
40
I
39
I
I
I
42
I
Total
- 5 7 4 3 I •
.8732. .
20
Total
. 7 26 I .
. 13 12 9
34
Av. length
. 30 32 33 37 39 •
. 30 33 36 39 . .
Av. length
■ 28 33 42 .
• 29 32 37
134
DISCOVERY REPORTS
MALES
FEMALES
Date September 5, 1928
St. No. WS 277
Net N 70 B 124-0 m.
Locality
Position
S. Georgia
SSi° 52'-30 S,
I 38 09 30 W
Surface T. -054° C.
Date Septembers, 1928
St. No. WS 277
Net N 70 B 124-0 m.
S. Georgia
f53° 52'-3oS,
I 38° 09'-30 W
Surface T. — 0-54° C.
Locality
Position
Length
in mm.
Stages
234S67ABCDEFG
Total
in
sample
Length
in mm.
Stages
234567 ABCDEFG
Total
in
sample
23
24
25
26
27
29
30
31
32
33
34
3S
38
40
41
42
45
46
47
48
49
50
SS
Total
Av. length
IS 3 4 10 20
26 31 36 41 48
16 3 2 4 27
26 32 35 39 46
23
24
25
26
27
28
31
32
33
35
36
40
41
43
44
45
46
48
49
Total
Av. length
13 IS I 14 I
23 26 32 41 49
I 19 9 3 12
21 24 27 33 43
Date September 17, 1928
St. No. WS 282
Net N7oBi37-om.
Locality
Position
Surface T.
S. Georgia
fS4°22'-3oS,
I 34°43'-ooW
-1-35° C.
Date September 17, 1928
St. No. WS 282 .
Net N 70 B 137-0 m.
Locality
Position
Surface T.
S. Oeorgia
/54'' 22'-30 S,
I 34° 43'oo W
-1-35° C.
Length
in mm.
15
28
29
34
38
39
40
45
47
48
49
50
51
52
53
55
57
Total
Av. length
Stages
234567 ABCDEFG
7 2
14 29
7 18 5
41 49 54
7 2
14 29
Total
in
sample
Length
in mm.
I 2 21 6
34 39 48 54
14
15
16
31
33
35
36
37
38
39
42
45
46
51
52
55
Total
Av. length
Stages
34567 ABCDEFG
I 17
31 48
I 5 12
31 36 46
Total
in
sample
Date September 24, 1938
St. No. 2430
Net N 100 H 5-0 m.
N 100 B 117-0 m.
Locality S. of Bouvet L
Position (=*°'o+:5?,V
t, 00 29 'O r.
Surface T. —0-99° C.
Date September 24, 1938
St. No. 2430
Net N 100 H 5-0 m.
Locality
Position
S. of Bouvet T.
154° I4'l S,
I 00° 29'-o E
Surface T. —0-99° C.
Length
in mm.
Length
in mm.
Stages
1234567
ABCDEFG
Total
in
sample
25
26
32
38
42
46
Stages
234567
ABCDEFG
Total
in
sample
Total
Av. length
Total
Av. length
26 35 44
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
135
MALES
FEMALES
Date October 2, 1928 Locality S. Georgia
St. No. WS290 p„o>;„n /54°23'ioS,
Net N7oHo-5m. Position |='%5"-'44'.oo W
Surface T. -108° C.
Date October 2, 1928 Locality S. Georgia
St. No. WS 200 „ .. f';4° 2-!''io S
Net N7oHo-5m. P<>^'"°" { 35°4]'ooW
Surface T. - 1 08° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
13
15
16
17
22
28
29
30
31
32
33
35
38
41
.
I
I
2
I
3
2 ;
I
I
2
4
7
■
I
5
2
1
2
2
4
4
8
3
I
I
13
14
IS
l6
22
25
26
11
29
30
31
32
33
34
35
37
I
2
•1
3
2 .
2
5 ■
4
s .
2
1
2
3
I
I
2 .
4 . .
3 • .
I
. 2 I
I 2
• • s
• • s
s
I
1
I
2
4
3
I
I
3
3
5
5
7
2
I
2
3
I
I
Total
Av. length
9 10 10 4 3
15 30 30 34 38 . .
II 7 13 2 3 ■ •
17 26 32 35 38 . .
36
Total
Av. length
10 6 28 I . . .
15 28 30 37 . . .
0 4 22 8 I . .
5 26 29 33 37 • •
45
Date October 4, IQ28 Locality S. Georgia
St. No. WS295 Pn.,;Mnn f55° 23'-40 S,
Net N 100 B 97-0 m. Fosition | 34°4i'ooW
Surface T. -110° C.
Date October 4, 1928 Locality S. Georgia
St. No. WS295 P„»;tion fss'23'-4oS.
Net N 100 B 97-0 m. Position |= 34" ^r'oo W
Surface T. -110° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
13
14
15
16
17
25
26
27
28
29
30
32
11
35
36
37
(
^
)
I
4
5
4
3
.
I
4
5
4
3
2
4
I
3
I
7
2
2
2
3
I
13
14
15
16
17
26
27
28
29
33
37
I
I
I
2
2
.
3 ■
7 ■
15 .
7 •
2 .
I
I
2
I
.
.
3
7
15
7
2
2
4
2
2
I
I
I I
2
• 3
1 I
. 2
2 I
Total
Av. length
34 4 7 I . . .
15 27 29 37 . . .
!4 I 5 5 I • ■
IS 26 27 29 37 . .
46
.
.
.
Total
Av. length
17 18 2 8 I . .
«5 28 33 34 37 . .
19 15 6 5 I ■ ■
17 28 33 34 37 ■ ■
46
Date October 5. 1928 Locality S. Georgia
St. No. WSzgS Pr„;,;„„ (S5°27'-3oS,
Net N 100 B 94-0 m. 1 osition | 32°2r-4oW
Surface T. -1-76° C.
Date October 5, 192S Locality S. Georgia
St. No. WS298 Pn^i.inn (SS°27'-3oS,
Net N 100 B 94-0 m. Position -^ 32°2i'-4oW
Surface T. -1-76° C.
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
Length
in mm.
Stages
Total
in
sample
25
11
29
30
31
32
33
34
35
36
11
39
40
41
42
43
45
46
SO
51
52
S6
I
I
3
I
2
I
I
2
2
2
2
2
I
I
2
I
I
I
I
I
2
2
I
I
I
1234567
A B C D E F G
2
I
2 I
I
11
30
31
32
33
37
38
39
41
43
44
46
48
49
50
51
52
54
55
i
[
[ I
[
I
2
I
I
2 .
4
3 •
2
I
I
2 I
.
i
3
I
I
I I
2 .
4 •
3 ■
. . 2
1
I
. 2 I
4
8
I
I
2
I
2
I
I
2
I
2
4
3
2
I
1
3
Total
Av. length
5 II I 3 19 3
28 30 33 38 47 52
• 7 9 I 3 19 3
■ 29 30 33 38 47 52
42
Total
Av. length
. 2 14 9 10 .
. 26 31 38 47 . .
. . 5 10 iS . 2
. . 28 32 42 . 53
35
136
DISCOVERY REPORTS
MALES
FEMALES
Date October 6. igaS
St. No. WS 304
Net N 100 B llo-o r
Locality
Position
Surface T.
S. Georgia
f54° 54'-40 S,
(^ 30^* 2l''2o W
-l-58°C.
Date October 6, 1928
St. No. WS 304
Net N lOQ B 1 lo-o m
Locality S. Georgia
/54° 54'-4p S
t 30'^ 21 -20 W
Surface T. -1-58° C.
Position
Length
in mm.
25
26
31
35
43
45
47
48
Total
Av. length
Stages
34567 ABCDEFG
26
31 35 46
26 31
2 4
35 46
Total
in
sample
Length
in mm.
26
29
30
3t
34
40
44
Total
Av. length
Stages
234567 ABCDEFG
I 3 4
26 30 37
332
28 32 42
Total
in
sample
Date October 16-17, 193°
St. No. 453
Net N 100 B 164-0 m.
Locality
Position
Surface T.
Bouvet I. to
S. Georgia
/S4°05J' S,
I 03° 57t' E
-l-6o°C.
Date October 16-17, 1930
St. No. 453
Net N 100 B 164-0 ni.
Locality
Position
Surface T.
Bouvet I to
S. Georgia
/54° osi' S,
I 03= 57i' E
-l-6o°C.
Length
in mm.
Stages
34567
ABCDEFG
Total
in
sample
Length
in mm.
Stages
234567
ABCDEFG
Total
in
sample
34
35
Total
Av. length
Total
Av. length
Date October 17, 1930
St. No. 454
Net N 70 B 192-0 m.
Locality
Position
Surface T.
Bouvet L to
S Georgia
/53°42'oo S,
(. 04^ 42'-oo E
-1-38° C.
Date October 17, 1930
St. No. 454
Net N 70 B 192-0 m.
LocaUty
Position
Surface T.
Bouvet L to
S. Georgia
/53°42'-oo S,
(^ 04^ 42'-oo E
-1-38° C.
Length
in mm.
13
14
15
16
17
39
40
41
43
44
45
46
49
51
55
Total
Av. length
Stages
234567 ABCDEFG
18 2
4J 53
18 2
41 53
Total
in
sample
Length
in mm.
13
14
15
16
17
35
36
37
38
39
40
42
43
44
45
46
48
49
50
51
Total
Av. length
Stages
234567 ABCDEFG
2 26 I
39 41 51
3 26
36 42
Total
in
sample
15
4
7
Date October 18, 1930
St. >Io. 455
Net N 100 B ii6-om.
Locality
Position
Surface T.
Bouvet I. to
S. Georgia
f53° 55i' S,
\ 04° 47' E
-fS9°C.
Date October 18, 1930
St. No. 455
Net N 100 B 116-0 m.
Locality
Position
Surface T.
Bouvet I to
S. Georgia
|53° 55*' S,
I 04° 47' E
-1-59° C.
Length
in mm.
13
14
15
16
17
32
39
41
42
45
Stages
234567 ABCDEFG
Total
in
sample
Length
in mm.
13
14
15
i6
17
19
35
38
44
45
Stages
234567 ABCDEFG
Total
in
sample
13
4
3
Total
Av. length 1 1 5
I 3
39 43
Total
Av. length
5
40
5
40
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
137
MALES
FEMALES
Date October 19, 1930 Locality
St. No. 459
Net N too B 183-0 m. Position
Surface T.
Bouvet I. to
.S. Georgia
;S5°09i'S. 1
t 02' 00' E
-i-38°C.
Date October 19, 1930
St. No. 459
Net N 100 B1 ,0, „„
N-oB 1 '83-0 m
Locality
Position
Surface T.
Bouvet I. to
S. Georgia
t 02 00 E
-1-38° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
123456
7
A B C D E F G
13
14
15
16
\l
19
37
43
3
I
3
5
1
2
2
I . . .
2
3
I
3
5
3
I
3
5
I
2
2
I
2
12
ii
17
18
19
37
38
39
40
2
4
II ... .
. . . 4 ■ •
. . . 3 . .
2
2 . .
4 ■ ■
II . .
5 • ■
2
2
2
I
2
1
2 . .
2
4
II
5
2
I
1
4
3
2
I
2 . . .
2 . . .
I
. 2
Total
Av. length
17 . . I . . 2
16 . . 37 • -43
17 . I ■
16 .37 •
. 2
■ . 43
20
I
Total
Av. length
26 . . 10 .
IS • ■ 38 .
26 . . 5
15 . . 3
5 . ■
36
8 38 . .
Date October 20, 1930 LocaUty
St. No. 460
Net NlooBlS5-om. Position
Surface T.
Bouvet I. to
S. Georgia
/56° 46' S,
t oo°4irW
-i-29'C.
Date October 20. 1930
St. No. 460
Net N 100 B 155-0 m.
Locality
Position
Surface T
Bouvet I. to
S. Georgia
/56° 46' S,
I oo°4iil'W
-1-29° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
1234567
A B C D E F G
II
12
13
14
\i
11
I
2
I
7
3
3
I
I
2 . . .
7 . . .
3 • • •
3 ■ • •
I . . .
I
2
I
7
3
3
I
12
13
14
IS
i5
17
18
2
5
5
7
2
4
I
2
5
5
7
2
4
1
2
s
5
7
2
4
Total
Av. length
26 ... .
15 ... .
26
15
26
Total
Av. length
18 I
14 38
18 . . .
14 . . .
I
. . 38
19
Date October 22, 1930 Locality
St. No. 461 D
Net N 100 B 490-385 m. Position
Surface T.
Bouvet I. to
S. Georgia
f56"4l'ooS
X 02" 24 -co W
-1-72° C.
Date October 22, 1930 Locality
St. No. 461 D
Net N 100 B 490-385 m. Position
Surface T
Bouvet I. to
S. Georgia
|56^ 4l'-oo S,
\ 02° 24'- 00 W
-1-72'' C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
12345
5 7
A B C D E F G
33
. . . . I . .
' ■
I
36
37
40
49
I . . .
I . . .
I . . .
I . . .
I
I
I
I
1
I
I
Total
Av. length
. . . . I . .
. . . . 33 ■ ■
I
33 •
I
Total
. . . 4 ■
3 I ■
4
Av. length
. . . 41
38 49 •
Date October 22, 1930
St. No. 461 G
Net N 100 B 700-560
Locality
3'5)m. Position
Surface
Bouvet 1. to
S. Georgia
/56° 441' S
I 02"'2lJ'W
T. -I-74°C.
Date October 22, 1930 Locality
St. No. 461 G
Net N 100 B 700-560 (315) m. position
Surface T
Bouvet I. to
S. Georgia
/S6°44rs
I 02"2li'W
. -I-74°C.
Length
in mm.
Stages
Total
in
sample
12345
6 7
A B C D E F G
Length
in mm.
Stages
Total
in
sample
32
33
34
11
11
39
40
41
42
43
44
45
46
47
48
50
. . . 2 .
2
. . . 3 .
. . . 4 ■
. . . 4 .
. . . 8 .
. I . 7 I
. 12 .
. . . 13 ■
... 8 3
. 12 2
. . . 7 I
. I . 13 I
. . . 5 I
. . . 5 2
2
I I
I
2 .
2 .
3 • •
4 . .
t : :
8 . .
12
13 . .
10 I
't : :
I 14 . .
6 . .
6 I .
2
2
I
2
2
3
4
t
9
12
'3
II
't
7
2
2
I
1234567
A B C D
E F G
35
11
39
40
41
42
43
44
45
46
tl
49
SI
. . . . 2 . I
6 I
52
5 3
. . . . I 3 3
I 7
4 7
1 10
: : : : : 1 8
I 4
2 I
31
I 2
2 I
7
7
2 6
7
8
II
II
I 4
10
I 4
3
4
3
I
3
7
7
8
7
8
II
II
5
10
5
3
4
3
I
Total
Av. length
.... 3 38 52
. . . . 37 42 43
6 87
40 42
93
Total
Av. length
. 3 . 107 14 .
. 41 . 40 44
I
. . 38 4
I 118 3 .
♦ 41 46 .
123
5-2
138
DISCOVERY REPORTS
MALES
FEMALES
Date October 23, 1930 Locality Bouvet 1. to
St. No. 462 S. Georgia
Net N 100 B go-o m. d„..-»- ("56° oi'-oo S,
Position f 07° 28-00 W
Surface T. -1-55° C.
Date October 23, 1930 Locality Bouvet L to
St. No. 462 S. Georgia
Net N too B 90-0 m. p^^-^-^^ {^'^o7°'28°oo W
Surface T. -1-55° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
12
13
14
IS
16
\l
St
2
2
s
11
2
2
2
I
2
2
5
II
2
2
2
2
2
5
11
2
2
2
I
12
14
17
18
19
40
2
I
6
5
2
I
I . . .
2
6 '.'.'.'.'. '.
5
I
2
I
2
I
5
5
1
2
I
I
Total
Av. length
26 I
15 51
26 I
15 51
27
Total
Av. length
18 . . I . . .
16 . . 40 ,
18 ... I . .
16 . . . 40 • ■
19
Date October 23, 1930 Locality Bouvet L to
St. No. 463 S. Georgia
Net N too B. 32-0 m. p^^.,^^^ {''°J^''sTooW
Surface T. -i-8o°C.
Date October 25, 1930 Locality Bouvet L to
St. No. 463 S. Georgia
Ne, N,ooB,32^m. p^^.^^^^ {"".o^^'s^^oo' W
Surface T. -180° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Toul
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
II
12
13
14
IS
16
17
19
33
I
2
I
I
9
6
8
2
I . . .
I
2
I
I
9
6
8
2
I
I
2
I
I
9
6
8
2
I
11
13
14
IS
16
17
19
44
I
2
2
10
4
2
I
I . . .
I
2
2
10
4
2
I
. . . . I . .
I
2
2
10
4
2
I
I
Total
Av. length
22 . . I . . .
IS . . 44 • ■ ■
22 ... I . .
15 ■ • . 44 • •
23
1
Total
Av. length
30 . . I . . .
16 . . 33 . • •
30 ... I . .
i6 . . . 33 . .
31
Date October 26, 1930 Locality Bouvet I. to
St. No. 465 S. Georgia
Net N 100 B. 13-0 m. p^^-^-^^ {''',4»'o2°' W
Surface T. -i-68°C."
Date October 26, 1930 Locality Bouvet L to
St. No. 46s S. Georgia
Net N.ooBii3-om. p^^;,;^^ {"^^^''ojvw
Surface T. -i-58'C."
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234367
A B C D E F G
1234567
A B C D E F G
47
50
I
I
I
I
I
I
39
40
. 2 . . .
I . . .
. 2 .
1
2
I
Total
Av. length
2
49
2
49
2
Total
Av. length
. . . 3 . . .
. . . 39 • . .
. . . . 3 . ■
. . . . 39 . .
3
Date October 31, 1930 Locality Bouvet L to
St. No. 471 S. Georgia
Net N 100 B .6S-0 m, p^^i,;^^ |54°|7^-oo S^^
Surface T. -I-62°C."
Date October 26, 1930 Locality Bouvet 1. to
St. No. 464 ' S. Georgia
Net N too H 67(-o) m. Pnsitinn f 56° 03'oo S,
fosition -J ,3=,8..ooW i
Surface T —1-75-0. 1
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
38
I
I
I
48
I . . .
I
I
Total
Av. length
I
38
38
I
Total
Av. length
I . . .
. . . 48 . . .
I
.... 48 . .
1
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
139
Date
St. No.
Net
November 19, 1929
W.S. Alongside
Deception I.
N 100 B 0-5 m.
Locality
S. Shetland Is.
FEMALES
Date November 19, 1929
St. No. W.S. .Alongside
Deception I.
Net N 100 B 0-5 m
Locality
Bransfield Strait
Surface T. circa 0-15° C.
Length
in nun.
Stages
234567 ABCDEFG
Total
in
sample
Length
in mm.
1234567
ABCDEFG
Total
in
sample
42
44
47
34
38
48
Total
Av. length
S
44
5
44
Total
Av. length
4 I
37 48
38 . 41
Date November 13, 1930
St. No. 480
Net N 100 B i6i~o m.
Locality
S. Georgia
Position {"39°54'^W
Surface T.
-0-58' C.
Date November 13, 1930
St. No. 480
Net N 100 B 161-0 m.
N 70 V 1000-750 m.
Locality S. Georgia
Position {";*.i;?w
Surface T. -058° C.
Length
Stages
Total
Av. length
234567
ABCDEFG
Total
in
sample
Length
in mm.
13
46
Total
Av. length
Stages
234567 ABCDEFG
46
. 46
Total
in
sample
Date November 16, 1930
St. No. 484
Net N 100 B 73-0 m.
Locality
Position
Surface T.
S. Georgia
/S3° 52i' S,
I 37° os4' W
-056° C.
Date November 16, IQ
St. No. 484
Net N 100 B 73-0 m.
Locality
Position
Surface T.
S. Georgia
f53= 52I' S,
l 37° osi' W
— 0-56° C.
Length
in mm.
13
15
16
17
18
Total
Av. length
Stages
234567 ABCDEFG
Total
in
sample
Length
in nun.
Date November 18, 1930
St. No. 492
Net N 100 B 148-0 m.
Locality
Position
S. Georgia
/53° 12}' S.
I 37°04i'W
Surface T. — 0-35° C.
Length
in mm.
14
15
16
23
24
Total
Av. length
Stages
34567 ABCDEFG
23 5
19 24
27 I
20 24
Total
in
sample
13
14
15
16
17
18
Total
Av. length
Stages
1234567 ABCDEFG
Total
in
sample
36
Date November 18, 1930
St. No. 492
Net N 100 B 148-0 m.
S. Georgia
/53°i2rs,
I. 37°04i'W
Surface T. -035° C.
Locality
Position
Length
in mm.
14
15
16
17
Total
Av. length
Stages
234567 ABCDEFG
sample
140
DISCOVERY REPORTS
MALES
FEMALES 1
Date
November 19,
193
0 Locality S. Georgia
Date
November 19, 193
0 Locality S. Georgia
St. No.
Net
N 100 B 160-0
m.
Position {5^;i4;f.v
Surface T. -0-85° C.
St. No.
Net
494
N 100 B 160-0 m.
Position {5^;5°i'^S,^
Surface T. -0-85° C.
Stages
Total
Stages
Total
Length
in mm.
m
sample
Length-
in mm.
in
sample
12 3 4
5
6 7
A B C D E F G
I 2 3 4 S
6 7
A B C D E F G
33
I
I . . .
I
36
I
I
I
35
I
I
37
. . . . 2
I . I
2
38
I
I
38
2
2 .
2
39
I
I
I
39
2
2 .
2
40
I 2
• I 3
4
40
.... 7
....6.1
7
41
1
1 I
2
41
. . . . 3
. . . . 3 . .
3
42
3
■ 3
3
42
.... 6
. . . . 5 ■ I
6
43
2
2
2
43
.... 8
.... 5 I 2
8
44
I I
. 2
2
44
.... 7
. . . .3.4
7
45
6
b
6
45
.... 8
• • • • 5 - 3
8
46
4
• 4
4
46
. . . . 5
. . . . I 1 3
5
47
2
2
2
47
. . . . 3
I 2
3
48
4
4
4
48
I
I
I
51
I
I
I
49
. . . . 5
....14.
5
52
I
I
I
51
I
1
I
55
I
I
I
—
—
Total
.... 61
. . . -35 9 J7
61
Total
• 3
29
. I I 2 32
36
Av. length
■ ■ ■ ■ 43
. . . . 42 48 44
Av. length
4<
3 45
1
Date
November 29,
193
D Locality S. Georgia
Date
November 29, 193
3 Locality S. Georgia
St. No.
Net
523
N 100 B1 ,„_
N70B 1 '57
Position {553°«rS,^
St. No.
Net
523
N 100 B 157-0 m.
Position {";°-'^9rW
Surface T. -030° C.
Surface T. -030° C.
Stages
Total
Stages
Total
Length
in mm.
—
in
sample
Length
in mm.
in
sample
1234
f
> 7
A B C D E F (
1234s
6 7
A B C D E F G
18
19
2 . . .
4 . ■ •
2
4
2
4
15
17
2 . . . .
2
2
20
15 . . .
15
19
20
7 . . . .
20 . . . .
14 ... .
7
7
20
21
22
16 . . .
12 . . .
16
12
16
12
14
8 . . .
8
8
13 • • ■ .
12 ... .
13
24
10 . . .
10
10
12
24
13 ... .
:8 . . . .
]l
26
3 • • ■
3
3
18
27
9 . . .
9
9
26
10 ... .
10
10
28
5 . . .
5
5
I
27
28
6 . . . .
6
6
30
I . . .
I
4 - . . .
4
4
Total
95 • • ■
95
95
41
I
I
Av. length
23 . . .
23
45
48
I
I
50
I
^ •
I
Date
November 6, I
1)32
Locality Bransfield Strait
Total
121 ... 5
I
121 ... 2 4 ,
127
St. No.
Net
1009
N 100 B iio-o
m.
Pos.,ion {%irll'^
Surface T. -0-85^ C.
Av. length
23 . . . 45 48 .
23 . . . 45 46 .
Length
in mm.
Stages
Total
in
sample
Date
St. No.
Net
November 6, 1932
1009
N 100 B iio-o m.
Locality Bransfield Strait
Position {^^;|=5VS,^
Surface T. -0-85'' C.
12 3 4-
«
7
A B C D E F G
16
17
18
20
21
2 . . .
4 . • •
2
2
4
4
Stages
Total
2 . . .
I
2
I
2
I
Length
in mm.
in
sample
1234567
A B C D E F G
7 3 ■ •
10
23
2 2..
4
4
17
2 . . . .
2
2
24
I . . .
I
19
I . . .
I . . .
I
^1
. 9 . .
3 6 .
9
20
3 ■ • •
3 . . .
3
26
5 . ■
1 4 .
5
21
71..
7 I ■ .
8
^l
8 . .
3 5 •
8
22
5 • ■ .
41..
5
28
4 ■ •
4 •
4
23
44..
5 3-.
8
29
I
1
1
24
21..
12..
3
30
2 . I
3 ■
3
25
153.
153-
9
31
I
1
26
2 2
2 . 2
4
32
I
27
• 4 I I
.312
6
34
I
28
I 2
I 2
3
36
I 5
3 3
6
29
2
2
2
37
I
I
1
30
. I I 2
2 2
4
38
• 3
1 2
3
31
... 2
2 .
2
39
2 -
2 I 2
5
32
2
2 .
2
40
• 4 •
.42.
6
34
I
I
I
41
2 :
.4.1
5
37
. . . 3 2
. . . . 5 . .
5
42
2
. 2 . I
3
38
2
2
2
43
,
I
I
39
2
2
2
44
.
I
I
40
I
I
I
45
.
.
2 I
3
43
I
I
I
Total
19 35 3 21 1:
. :
32 23 7 18 7 s
92
Total
25 18 8 15 8
24 17 7 10 16 . .
74
Av. length
22 26 34 38 4
4'
\ •
22 27 35 39 42 4J
Av. length
21 25 27 31 39
21 25 27 28 36 . .
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
141
MALES
FEMALES
Date December?. 1926 Locality S.Georgia
St. No. Govt, jfetty, Gritvyken Surface T. circa 4-65 C.
Net N 100 H o-i m.
Date December 7, 1926 Locality S.Georgia
St. No. Govt. Jetty, Gritvyken Surface T. circa 465 C.
Net N 100 H O-I m.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
11
27
28
29
30
31
32
34
22
I
2
4 3
26
• 3
I
I
. 2
22
. I
II
3 4
3 5
.3
I
I
. 2
4
I
2
7
8
3
z
I
2
23
24
25
26
27
28
29
30
31
33
35
2
2
5
2
3
6
4
1
I
I
. 2
I
11
• 5
. 2
• 3
.6
. 22....
I
I
2
I
2
5
2
I
4
I
I
I
28
Total
Av. length :
9 20
7 29 '
9 20
29
8 29
Total 28
Av. length 28
. 24 4 . . . .
. 28 29 . . . .
Date December II, 1926 Locality S Georgia
St. No. King Edward's Cove Surface T. circa 4-65 C.
Net N 100 H o-i m.
Date December 11, 1926 Locality S.Georgia
St. No. King Edward's Cove Surface T. circa 4-65° C.
Net N 100 H 0-1 m.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
23
25
26
V,
29
30
31
32
33
34
35
36
37
38
39
I
2 .
I I
4 4 ■
1 2 .
J 3 •
2 I
I 3 •
• 7 •
■ 3 I
• 3 •
2 .
3 5 •
. 3 •
• 4 •
I 2 .
4
• 7 ■
• 4 I
■ 3 •
. 4 ■
• 3
I
I
I
I
2
2
8
3
4
3
4
7
5
3
4
3
I
I
I
24
25
25
27
28
29
30
31
32
34
37
2
4
6
9
4
5
8
3
3
11
I
II....
. 3 I • ■ • •
. 5 I . • • ■
.36....
22....
. 4 I . • • •
,26....
12....
. 21....
II....
I . . . .
2
4
6
9
4
5
8
3
3
2
I
I 2
1
Total
Av. length
46 I
28 34
. 24 23 . . . .
. 28 29 . . . .
47
I . . . .
Total
Av. length
13 30 7 2 . . .
27 31 36 35 • • •
6 45 1 ■ • • ■
27 31 33 . ■ • •
52
Date December 19, 1926 Locality S. Georgia
Ifet'"'- ^^LoH7om, P-«- {''l^zl^-fo^
Surface T. 1-45° C.
Date December 19, 1926 Locality S. Georgia
St. No. 125 Position (53 28 30 S,
Net N 100 H 70 m. l-osmon | 36" 2o'-3o W
Surface T. 145 C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
40
41
43
44
45
46
%
50
51
52
53
54
55
56
57
I
I
2
. . • 4
I
. . ■ 5
. • . 5
. I . 5
. . . 6
. . ■ 4
. ■ . 3
. . . 3
. 2
I . . I .
I
I I I
. . . . 5
. . . I 5
. . . . 6
. . . . 4
. . . . 3
. . . . 3
. 2
2
I
I
I
3
2
4
1
5
5
6
6
4
3
3
2
43
44
45
46
47
48
49
50
51
52
54
55
S6
57
58
60
61
I . . .
I
. . . . 3 • ■
I
I
2 .
. . . . 4 . .
. . . .41-
2 .
. 2 1
. . . . 6 . .
. . . .49-
I 2 .
I . I
1
I
I
1
I
I
I . I
. . . . 1 . 3
. . . . I . 4
2
I . 2
. . . . I . 5
2 II
3
2
I
I
I
3
I
I
2
4
5
2
3
6
13
3
2
1
I
Total
Av. length
. . I 3 I 4 40
■ ■ 40 43 52 43 52
. . 2 . 2 3 42
. .41 .45 46 51
49
Total
Av. length
. . . I 33 14 2
. . . 43 51 55 59
. . . . 7 5 38
. . . . 48 55 53
50
Locality S. Georgia
Date December 11, 1930 Position {'O'lor'w
St No s,zn *■ 34 29J w
Net N 450 H 122 (-0) m. Surface T. 0-35° C.
Length
in mm.
Stages
Total
in
sample
Date December II, 1930 Locality S.Georgia
&'et''°- ^n5oH.32(-o)m. P--" {'^^"I^VW
Surface T. 0-35° C.
1234567
A B C D E F G
14
15
i5
17
18
19
20
21
22
24
26
28
42
I . . . .
I
I
I
3
4
4
4
2
3
21
I
1
I
1
I
3
4
4
4
2
3
3
I
I
I
Length
in mm.
Stages
Total
in
sample
1 . . . .
3 ■ . ■ •
4 . . . .
4 . . . .
4 . . . .
2 . . . .
3 ■ • • ■
3 . ■ • •
1
1234567
A B C D E F G
16
18
19
20
21
23
24
5
2
4
2
I
4
I
5
2
4
2
I
4
I
5
2
4
2
I
4
Total
Av. lengtl
19
1 19
19
19
19
Total
Av. lengtl
28 .... I •
20 . . . . 42 .
27 I .... I
20 24 . . . .42
29
142
DISCOVERY REPORTS
MALES
FEMALES
Date December 13, 1930 Locality
Net"'" ^'°ooB,6S-om., P°-'-"
450-168 m. Surface T.
S. Sandwich Is.
f 55° 32J' S
I 33"I4'W
-O^S'C.
Date December 13, 1930 Locality
^'et^°- ^^!ooB,68-om., "-'•-
450-168 m. Surface T.
S. Sandwich Is.
/55'32rS,
I 33= I4'W
-0-95° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
1234567
A B C D
E F G
16
21
23
25
26
28
29
30
31
I
I
3 '■
I
2
I
12
I
1
I
2
I
I
I
I
3
I
I
1
2
1
20
25
26
27
28
30
32
I
I
2
4
I
I
I
I
I
2
31
I
I
I
1
2
4
I
I
1
Total
Av. length
II
27
7 4
26 28
II
Total
Av. length
12
26
48
21 28
12
Date December 14, 1930 Locality
gfet"""- 'N'tooBz64-om. P^"""
Surface T.
S. Sandwich Is.
fS7° 27' S,
\ 34°25'W
-0-90° c.
Date December 14, 1930 Locality
Net'"'- ^^ 100 B, 64-0 m. P--"
Surface T.
S. Sandwich Is.
/57° 27' S,
1 34°2S'W
— 0-90° C.
Length
in mm.
1
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
1234567
A B C D
E F G
25
27
30
I
I
I
I
I
I
J
21
22
24
25
2
2
2
I
2
r I
2
I
2
2
2
Total
Av. length
3
27
• 3
.27
3
Total
Av. length
7
23
61
23 22
7
Date December 17, 1930 Locality
St. No. 534 Position
Net N 100 B 172-0 m. l^osition
Surface T.
S. Orkney Is.
f6o° 08' S,
1 47°53'W
ois" C.
Date December 17, 1930 Locality
Net"""- ?l1ooB,7.-om. ?--"
Surface T.
S. Orkney Is.
/6o'> 08' S,
1 47°S3'W
0-15° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
1234567
A B C D
E F G
47
49
51
53
I
I
I
I
I
I
I
I
I
I
I
I
41
I
I
Total
Av. length
I
41
Total
Av. length
4
50
4
50
4
Date December 18, 1930 Locality
St. No. S.J5 Position
Net N 70 B 0 m.
Surface T.
S. Orkney Is
f6o° I3i' S,
\ 50° 51 J' W
o-ds" C.
Date December 18, 1930 Locality
Net''"- ?."7oBom. ?--"
Surface T.
S. Orkney Is.
(60° I3i' S,
I SO°5li'W
0-65" C.
Length
in nmi.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
1234567
A B C D
E F G
41
42
44
45
tl
49
50
51
52
S3
54
11
I
I
I
2
I
I
9
I 3
6
I
7
5
10
1
I
I
I
2
I
I
9
I
I
I
2
I
I
9
4
6
I
7
6
10
I
37
44
47
2
I
I
I
I
I
Total
Av. length
4
:::::: ^
I
7
6
zo
I
Total
Av. length
2 49
SI 51
SI
51
51
Date December i8, 1930 Locality
fj^t^"- 'N^':ooBt.2-om. P-"™
Surface T.
S. Orkney Is.
f6o° 43' S,
1 s2°29i'W
-0-30° C.
Length
in mm.
Stages
Total
in
sample
Date December 18, 1930 Locality
^'et"""- ?.^'ooB 122-0 m. P-'-
Surface T.
S. Orkney Is.
f6o° 43' S,
1 52°29rW
-0-30° C.
1234567
A B C D
E F G
35
42
50
51
54
55
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
36
I
Total
Av. length
6
48
6
48
6
Total
Av. length
I
. . . . 36 . .
36
1
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
143
MALES
FEMALES
Date
St. No.
Net
December 19, 1930 Locality
N''.ooB.37-om. P°^'"°"
S. Shetland Is.
/6i"o7i'S.
I 54°26-W
Date
St. .Xo.
Net
December 19, 1930
537
N 100 B 137-0 m.
Locality
Position
S. Shetland Is.
f6i°07*'S.
1 54°26'W
001° C.
N 70 V 500-250 m. Surface T.
o-oi' C.
N 70 V 500-250 m
Surface T.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
■ 2 3 4 5
6 7
A B C D
E F G
17
6
6 . . .
6
16
6 . . . .
6 . .
5
19
6
6
6
20
12 ... .
12 . . .
20
24
24
24
21
36 ... .
36 . . .
36
24
24
24
22
30 ... .
30 . . .
30
36
22
23
18
18
18
18
18
23
24
36 ... .
30 ... .
36 . . .
30 . . .
24
42 •
42
42
25
12 ... .
25
36 18
36 18 . .
54
29
I . . . .
I . . .
3t
12
6
6
6 6..
. 6 . .
. 6 . .
12
6
6
Total
163 ... .
163 .. .
163
44
49
I
I
I
Av. length
22 ... .
22 . . .
.
*
Total
198 18 . . . .2
180 36 . .
. 2
218
Av. length
23 25 . . . .47
23 27 . .
• • 47
Date
December ig, 1930 Locality
S. Shetland Is.
Date
December 19, 1930 Locality
S. Shetland Is.
St. No.
Net
nLoB 137-0 m. P°^'*i°"
f6i°2g'S,
\ S4°44rW
St. No.
Net
538
N 100 B 137-0 m.
Position
(•6l°29'S,
\ 54° 44i' W
Surface T.
-0-25° C.
Surface T.
-025° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
12 3 4 5
5 7
A B C D
E F G
17
I
I . . .
I
18
I . . . .
J
18
I
I
I
19
I . . . .
I
19
I
I
I
20
I . . . .
20
2
2
2
22
I . . . .
21
2
I
2
23
3 . . . .
3
22
3
2
3
48
I
I
24
27
46
I
I
1
1
I
I
Total
7 ... I
7 . . .
. . I
8
50
S3
S6
I
Av. length
21 ... 48
21 . . .
. . 48
I
I
Total
II I .... 4
9 3 • •
16
Av. length
21 27 . . . .51
21 23 . .
• • SI
Date
December 19, 193
0 Locality
S. Shetland Is.
St. No.
Net
flooM-"--
Position
Surface T.
;6i°48'S,
I 54° 514' W
-0-30° C.
Length
in mm.
Stages
Total
in
sample
Date
St. No.
Net
December 19, 1930 Locality
S. Shetland Is.
/6i°48'S,
I 54°Sii'W
12 3 4 5
5 7
A B C D E F G
N 100 B/ '37-0 m. Surface T.
12
I . . . .
,
I
-030° C.
17
z .
2
2
18
4
2
Stages
Total
19
2 .
2
Length
in mm.
in
sample
21
22
23
4 .
9 .
6 .
I
4
9
6
I
4
9
6
1234567
A B C D
EEC
12
I
I
28
I
I
16
2
2
30
I
2
18
3
3
31
I
2
19
I
I
32
: I
3
20
21
3
3S
2
21
2
2
36
I
22
2 2
4
37
I
23
62
8
38
5
24
I
I
39
2
27
I
40
I
28
I
I
41
2
35
I
1
42
4
n
I . . . .
I
I
43
I
38
I
I
44
I
4S
. I . . . .
I
45
I
*1
2
2
46
I
48
. . I . . I
2
47
I
49
3
3
48
I
SI
2
. . 2
2
50
I
52
3
• • 3
3
51
I
53
I
I
I
52
. 2
2
54
2
. 2
2
55
2
. 2
2
Total
20 7 2 I .16
13 15 • I
. I 16
46
Total
28 2 I I 8 25
2
29 2
I A
17 14
07
Av. length
20 23 41 48 . . 49
20 23 . 45
• 48 49
Av. length
20 27 28 38 38 41
55
20 29
38 3!
40 4S
144
DISCOVERY REPORTS
MALES
FEMALES
1
'Date December 19, 1930 Locality
N 70 V 500-250 m. Surface T.
S. Shetland Is.
/62° o6i' S,
1 S5°o8i'W
-0-48° C.
Date December 19, 1930
St. No. 540
Net N 100 I! 155-0 m.
N 70 V 500-250 m.
Locality
Position
Surface T.
S. Shetland Is.
(62° o6i' S,
l. 55°o8i'W
-048° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567ABCD
E F G
123456
7
A B C D
E F G
21
48
51
S3
I
I
2
I . . .
I
2
I
I
I
2
I
20
32
34
2
I
2 . . .
2
1
I
I
I
Total
Av. length
2 ... I I .
20 . . . 32 34 .
2 . . .
20 . . .
. 2
• • 33
4
Total
Av. length
I 4
21 SI
I . . .
21 . . .
• • 4
• . 51
5
Date December iQ-20, 1930 LocaHty
Net""" ^*",ooB,o8-om. P-"-
Surface T.
S. Shetland Is.
(bz° 22' S,
1 55°23'W
-0-85° C.
Date December 19-20, 1930 Locality
St. No. 541 „ . .
Net N 100 B 108-0 m. Posmon
Surface T.
S. Shetland Is.
f62° 22' S.
I 55°23'W
-085° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
123456
7
A B C D
E F G
23
I
I
I
18
19
21
23
25
27
32
44
Total
I
I
23
1
Av. length .23
I
I
Total
Av. length
6 ... I I .
22 . . . 32 48 .
6 . . .
22 . . .
. 2
. . 38
S
Date December 20, 1930 Locality
Surface T.
Bransfield Strait
/62° 16' S,
\ 57°2o'W
030° C.
Date December 20, 1930
St. No. 543
Net N 100 B 178-0 m.
Locality
Position
Surface T.
Bransfield Strait
/^62° J 6' S,
I 57" 20' W
0-30° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
■ 2 3 4 S 6 7
A B C D
E F G
123456
7
A B C D
E F G
20
I
I
I
IS
I
I
I
Total
Av. length
I
20
I
20
I
Total
Av. length
IS
I
15
I
Date December 20, 1930 Locality
St. No. 546 n ■.■
Net N 100 B 164-0 m. P°^'^'«"
Surface T.
Bransfield Strait
f62° 46i' S,
l S7°lli'W
-o-6o° C.
Date December 20, 1930
St. No. 546
Net N 100 B 164-0 m.
Locality
Position
Surface T.
Bransfield Strait
/62°46i'S,
I 57° III' W
-o-6o° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
123456
7
A B C D
E F G
48
53
I
I
I
I
I
46
I .
I
Total
Av. length
46 .
I
46
I
Total
Av. length
2
SI
2
51
1
Date December 20, 1930 Locality
Surface T.
Bransfield Strait
/62° 59i' S,
I 57°03'W
-1-02° C.
Date December 20. 1930
St. No. 547
Net N 100 B 37-0 m.
Locality
Position
Surface T.
Bransfield Strait
f62° 59i' S,
I 57-03'W
-102° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
123456
7
A B C D
E F G
16
I
I
I
20
I
1
1
Total
Av. length
16
I
16
I
Total
Av. length
I
20
I
20
1
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
145
FEMALES
Date December 21, I93<
St. No. 548
Net N 100 B 102-0 m.
Locality
Position
Surface T.
Bransfield Strait
/62°36!'S,
045° C.
Date
St. No.
Net
December 21, I93<
S48
N 100 B 102-0 m.
Length
in mm.
42
43
45
46
47
48
49
SO
51
52
S3
S4
Total
Av. length
Stages
67 ABCDEFG
4 4
4S 48
I 34
45 50
7 35
47 50
Total
in
sample
Date December 21-22, 1930
St. No. 549
Net N 100 B II 5-0 m.
Locality
Position
Surface T.
Bransfield Strait
f63° ooi' S,
0-41° C
Length
in mm.
23
31
32
33
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Total
Av. length
Stages
234567 ABCDEFG
7 6 9 IS 3 • 3
'-t 35 40 40 46 • 46
Total
in
sample
7 7 7 14 3 2 3 43
21 34 37 42 44 47 46
Date December 29, 1930
St. No. 558
Net N 100 B 146-0 m.
Bellingshausen Sea
f5s" 31' S,
t f,r 07 r W
Surface T. -092° C.
Localit>'
Position
Length
in mm.
Stages
Total
in
sample
12345
6 7
ABCDEFG
49
52
I
I
I
I
;
Total
Av. length
I
. . • 49
. 1
■ 52
I I
49 52
2
Date December 30. 1930
St. No. 55Q
Net N 100 Bi 13-0 m.
Locality Bellingshausen Sea
Position {^*;i.-»;f.w
Surface T. -o-8i° C.
Length
in mm.
40
47
Total
Av. length
Stages
234567
ABCDEFG
Total
in
sample
Bransfield Strait
f62' 36i' S,
i S8°58'W
Surface T. 0-43^ C.
Locality
Position
Length
in mm.
39
40
41
43
45
46
47
48
49
50
52
53
55
Total
Av. length
Stages
2 3 4 5 6 viABCDEFG
I 32 3
39 46 47
12 12 12
42 49 46
Total
in
sample
36
Date December 21-22, 1930
St. No. 549
Net NiooBli5-om.
Locality
Position
Surface T.
Bransfield Strait
f63° oof S,
{ 6i°i6i'W
041° C.
Length
in mm.
13
15
17
32
33
35
36
37
38
39
40
41
42
43
44
45
46
47
48
52
Total 6 4
Av. length 17 34
Stages
567
7 24
38 42
ABCDEFG
6 I 2 3 21 I 7
17 36 33 36 40 45 47
Total
in
sample
Date December 29, 1930
St. No. 5S8
Net N 100 B 146-0 m.
Bellingshausen Sea
/6s°3i'S,
(. 67= 07J' W
Surface T. -092' C.
Locality
Position
Length
in mm.
23
25
40
Total
Av. length
Stages
234567
3
40
ABCDEFG
3 ■
27 ■
3
40
Total
in
sample
Date December 30, 1930
St. No. 559
Net NiooBii3-om.
Locality Bellingshausen Sea
,, ... f66''2i}'S,
Position I 68''55rW
Surface T. -o-8i° C.
Length
in mm.
33
43
45
46
47
Total
Av. length
Stages
6 7
5
43
ABCDEFG
Total
in
sample
33 46 45
6-2
146
DISCOVERY REPORTS
MALES
Date December 30, 1930
St. No. 560
Net N 100 B 155-0 m.
Bellingshausen Sea
(66°47rS,
i 69°I9'W
Surface T. -069° C.
Locality
Position
Length
in mm.
29
49
Total
Av. length
Stages
1234567
A B C D E F G
FEMALES
Date
St. No.
Net
December 30, 1930
560
N 100 B 155-0 m.
Locality
Position
Surface T.
Bellingshausen Sea
f66° 47J' S,
I. 69° 19' W
-0-69° C.
Total
in
sample
Date December 31, 1930 Locality Bellingshausen Sea
St- No. 56. _ p„,itio„ i^^°Jji'3
Net
N 100 B 137-0 m.
\ 72° ogi' W
Surface T. — i-35°C.
Length
in mm.
30
37
42
45
47
50
Total
Av. length
Stages
234567
20 . 40 46 50
A B C D- E F G
3
47
Total
in
sample
Date December 31, 1930
St. No. 561
Net N 100 B 137-0 m.
Locality Bellingshausen Sea
„ .,. f66°47j'S,
Position I 72»o9j'W
Surface T. -135° C.
Length
Stages
Total
in
sample
in mm.
1234567
A B C D E F G
52
I
I
«
Total
Av. length
52
I
52
•
Date December 31, 1930
St. No. 562
Net N 100 B 113-0 m.
Locality Bellingshausen Sea
Position {"7° '54' S
Surface T.
75° 27' W
o-62''C.
Length
Stages
Total
in mm.
1234567
A B C D E F G
sample
47
50
I
I
I
I
Total
Av. length
1 1
. . • • 47 50 •
I . 1
. . . .47 -50
2
Date December 31, 1930
St. No. 562
Net N 100 B 1 13-0 m.
Bellingshausen Sea
I 75"_27'W
Surface T. — o-ba" C.
Locality
Position
Length
in mm.
48
Total
Av. length
Stages
234567
48
A B C D E F G
48
Total
in
sample
Length
in mm.
Total
Av. length
Stages
1234567
A B C D E F G
53
Total
in
sample
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
147
MALES
FEMALES
Date
January 22, l92g Locality
S. Georgia
Date January 22, 1020 Locality
S. Georgia
St. No.
Net
WS373EE p„^;,;„^
IN 100 B 70-0 m.
/54° 10' S.
\ 35'40'W
St. .N'o. WS 373 EE „ . .
Net N 100 B 70-0 m. Position
/S4° 10' S
1, 35'40 W
Surface T
circa 2-99- C.
Surface T
circa 2-99° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
36
I
I
33
,
I
I
37
t
I
34
I
38
r
I
36
2 .
. .
2
42
I . . .
I
45
.
I
I I
2
43
1
2 . . .
2
47
1
1
I
44
I
48
I
1
I
45
2 .
2
49
6 .
. I S
6
46
• . . 3
I • 3
4
50
3 I*
■ ■ 4
4
H
I . . 3
I . 3
4
51
10 .
. I 9
10
48
2
I . 2
3
52
6 .
• I 5
6
49
2
2
S3
2 2
. 2 2
4
50
. . . 5
• . s
5
54
7 .
. I 6
7
SI
• ■ ■ 3
■ . 3
3
55
2 .
. 2
2
52
. . ■ 3
I • 3
4
56
I
I
I
S3
. . . 6
. . 6
6
S7
1
I
I
S4
2
2
2
S8
1
I
I
55
. ■ ■ 3
I . 3
4
59
I
I
56
58
I
I
I
Total
Av. length
2 2 .. I 43 3
34 36 . . 45 52 52
I 3 •
34 35 .
■ 8 39
■ 52 52
51
Total
Av. length
• I 13 I • -33
• 43 45 47 • ■ SI
• 3
• 37
4
) 9 • 33
48 . 51
48
Date
January 21-22, 1930 Locality
S. Georgia
Date January 21-22, 1930 Locality
S. Georgia
St. No.
Net
N ^00 B loo-o m. P°^"'°"
fS3° 17' S,
^ 37°lo'W
Net''" r.ooB 100-0 m. P-"-
{";,-nf-w
Surface T
330" C.
Surface T
3-30° c.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length
in mm.
in
sample
1234567
A B C I
D E F G
1234567
A B C D E F G
35
2 1....
- 3 ■
3
37
36
2
38
37
I
39
38
I
40
4«
3
41
42
I
42
2
45
I
45
1
51
I
I
50
I
54
57
I
I
1
I
51
I
Total
. 8 4 ... 3
. 12 .
. . 3
15
Total
Av. length
8 .... 2 .
41 .... SI .
• 4 4
• 40 42
. 2
• ■ 51
10
Av. length
■ 39 39 . ■ .54
- 39 .
. . 54 1
Date
January 24-25, 1930 Locality
S. Georgia
Date January 24-25, 1930 Locality
S. Georgia
St. No.
Net
nIooB 150-0 m. P°^"'°"
/S3°39r S,
l 35°37i'W
Net'"'- ^'looBlso-om. ?-"-
("°39rs,,,
I 35 37rW
TSJ -7r. V / 1 00-50 'n. Surface T
' \sSO-250 m.
2-4S° C.
N 70 V 100-50 m. Surface T
2-45" C.
Stages
Total
Stages
Total
Length
in mm.
m
sample
Length
in mm.
m
sample
1234567
A B C I
) E F G
1234567
A B C D E F G
35
I
30
I
1
• .
41
1
I
50
45
.
51
I
47
57
I*
50
SI
52
2
I
I
62
Total
I ... I . 3
I . . .
. 2 2
5
54
SS
• ■ ■ 3
I
. . 3
I
Av. length
30 ... 51 • 56
30 . . .
. 57 54
S6
2
. 2
2
57
4
• . 4
4
Total
. I 3 ... 25
I I 2
. . 25
29
Av. length
. 35 44 • . .54
. 35 41 At
• . 54
Date
January 29, 1930 Locality
S. Georgia
St. No.
N^iooBro4-om. Position
;S3°39'S,
Date January 29, 1930 Locality
S. Georgia
Net
t 3S°24i'W
St. No. 318 Position
.J 53° 39' S
Surface T
3lo°C.
Net N 100 B 104-0 m. rosition
Surface T.
t. 3S^24A'\V
310° C.
Stages
Total
Length
in mm.
in
sample
Stages
Total
in
ABC!
) E F G
Length
1214567IABCE
E F G
sample
35
2 . . . .
. 2 .
2
37
I
I
I
37
4
4
39
I
I
1
39
z
43
. . I . . . .
I
1
40
2
2
Total
. 2 3 ■ • • •
.41-
5
Total
7
.43.
7
Av. length
. 38 38 ... .
. 37 43 .
Av. length
38
. 38 38 .
. . .
Note. Females marked with an asterisk have spawned.
148
DISCOVERY REPORTS
MALES
Date Januari' 8, igii
St. No. 575
Net N loo B 97-0 m.
Locality Bellingshausen Sea
Position |*7°53i'S.
1 91 23 W
Surface T. — i-47°C.
Length
in mm.
31
33
34
36
37
38
40
41
45
46
47
48
50
Total
Av. length
Stages
34567
226615
32 34 37 43 48 47
A B C D E F G
356215
33 36 41 46 48 47
Total
in
sample
FEMALES
Date Januarys, 1931
.St. No. 575
Net N 100 B 97-0 m.
Locality Bellingshausen Sea
Pos,tion {%pi^/ij,
Surface T. -1-47° C.
Length
in mm.
30
31
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Total
Av. length
Stages
234567 ABCDEFG
3
31 33
27 IS
39 45
2 3 2 I 7 32
33 31 34 43 43 42
Total
in
sample
Date January 8, 193 1
St. No. 576
Net N 100 B 132-0 m.
Locality Bellingshausen Sea
Position {%^°\%.^
Surface T. -1-15' C.
Date Januarys, 1931
St. No. 576
Net N 100 B 132-0 in.
Locality Bellingshausen Sea
Position {^'8<5°;S^.w
Surface T. — 1-15' C.
Length
in mm.
Stages
234567 ABCDEFG
Total I
Av. length 49
Total
in
sample
Length
in mm.
Total
Av. length
Stages
1234567 ABCDEFG
Total
in
sample
Date January 9, 1931
St. No. 578
Net N 100 B 128-0 m.
Locality Bellingshausen Sea
Position {^'8 54,fj.w
Surface T. -1-20° C.
Date January 9, 1931
St. No. 578
Net N 100 B 128-0 m.
Locality Bellingshausen Sea
Position C^^gS/ij-w
Surface T. -I-20°C.
Length
in mm.
21
28
50
Total
Av. length
Stages
34567
ABCDEFG
Total
in
sample
Length
in mm.
Total
Av. length
1234567
ABCDEFG
Total
in
sample
Date January lo, 1931
St. No 580
Net N 100 B 128-0 m.
Locality Bellingshausen Sea
Position (*7°4ji'S.
(, 75 56i' W
Surface T. -o-lo° C.
Length
in mm.
24
46
47
48
50
51
52
53
54
Total
Av. length
Stages
1234567 ABCDEFG
Total
in
sample
Date January 10, 1931
St. No. 580
Net N 100 B 128-0 m.
Locality Bellingshausen Sea
Position {^'„^-i,f.-w
Surface T. -o-lo° C.
Length
in mm.
45
48
Total
Av. length
Stages
1234567
ABCDEFG
Total
in
sample
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
149
MALES
FEMALES
Date January 10, 1931 Locality S. Georgia
St. No. Marine Station Jetty „ „ ,„
Net NH 0 m. Surface T. 2-40° C. (?)
Date January 10, 1931 Locality S. Georgia
St. No. Marine Station Jetty , r^ ,,■.
Net NHom Surface T. 240' C. (?)
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
«9
20
21
22
23
24
li
27
28
30
31
I
12
3°
26
27
25
26 2
12 I
31
I
I
12
30
26
27
25
26 2
12 I
31
I
I
12
30
26
27
25
28
13
4
I
I
I
18
20
21
22
23
24
25
26
29
30
I
17
25
24
41
30
13
3
I
I
I
17
25
24
41
30
13
12
I
I
I
17
25
24
41
30
13
3
I
I
Total
Av. length
56
23
152 4
23 28
156
Total
Av. length
63 6 163 6
23 27 23 27 1
169 .
Date January 12, 1931 Locahty Bellingshausen Sea
Surface T. -o-ig'^ C.
Date January 12, 1931 Locality BelUngshausen Sea
Surface T. -o-ig' C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
SI
I
I
I
3I
48
1
I
I
. . . I . . .
I
I
I
Total
Av. length
I
51
I
51
I
I
Total
Av. length
I . . . 2 . .
25 . . . 43 • •
I . . I . I .
25 . .38 .48 .
3
Date January 13, 1931 Locality Bellingshausen Sea
Surface T. -0-72° C.
Date January 13. 1931 Locality Bellingshai
St. No. 584 Position -f*'^^,^' ;
Net N too B 165-0 m. Position ^ ^^^ j
Surface T. -0-72° C.
isen Sea
5,
' W
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
46
I
I
1
23
48
I
I
I
I
I
I
Total
Av. length
I
46
I
46
I
Total
Av. length
I .... I •
23 .... 48 •
1 I
23 48
2
Date January 14, 193 1 LocaUty Bellingshausen Sea
St. No. 590 , Position l''5°^„°*' w',„
N" Nioob|9<^°'"- TT\- ^ 7,\.3oi'W
\310-om. Surface!. IS7 L.
N 70 B go-o m.
Date January 14, 193 1 Locality Bellingshausen Sea
St. No. 590 „ Position |*S,^o°*f.'w
Net N IOC B 90-0 m. Surface T. W^'a"* "^
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
49
50
52
54
55
I 1
2
2
I
I
I I
2
2
I
I
2
2
2
I
I
48
50
I
I
I
I
I
1
Total
Av. length
2 .
49 ■
2
49
2
Total
Av. length
....1.7
. . . .49-52
I 7
49 52
8
Date January 16, 193 1 Locality Bellingshausen Sea
N7oB/'70-om. Surface T. 1-51° C.
Length
in mm.
Stages
Total
in
sample
Date January 16, 1931 Locality Bellingshausen Sea
St. No. 596 ^ Position 1 'fi^='c?i'w
Net N 100 B 170-0 m. I (f SSi W
Surface T. 1-51 C.
1
1234567
A B C D E F G
45
tl
49
50
51
52
53
I
2
I
I . I
I
I
. . . I . . .
I
2
I
I I
I
I
I
I
I
2
1
2
I
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
43
45
46
I .
I
I
. . . . J . .
I
I
1
I
Total
Av. lengtl
...22.6
. . . 46 51 . 51
...1.36
. . .47 . 49 51
10
Total
Av. length
2 I
. . . . 46 43 •
1 I I
. . . . 45 43 46
3
ISO
DISCOVERY REPORTS
MALES
FEMALES
Date January 17. 1931
St. No. S99
Net N 100 B 142-0 m.
Locality Bellingshausen Sea
„ ... f67° 08' S,
Position I 'bg-obVW
Surface T. -0-71° C.
Date January 17, 193 1
St. No, 599
Net N 100 B 142-0 m.
Locality Bellingshausen Sea
Position {*';°»^;fi-w
Surface T. -0-71° C.
Length
in mm.
Total
Av. length
34567
A B C D E F G
Total
in
sample
Length
in mm.
47
48
49
Total
Av. length
Stages
I z 3 4 5 6 7
3
48
A B C D E F
Date January 19, 193 1
St. No. 602
Net N 100 B\ ,,„ „„
N70B ; "o-o"
Locality Bellingshausen Sea
Position {^^66»^2S' W
Surface T. — 0-02° C.
Date January 19, 193 1
St. No. 602
Net N 100 Bl ,,„ „„
N70B ; ''°-°"
Bellingshausen Sea
(66° 03i' S,
l 66' 25' W
Surface T. — 002° C.
Locality
Position
Length
in mm.
41
42
45
46
47
48
49
50
51
52
S3
54
Total
Av. length
Stages
234567 ABCDEFG
47 50 52 50
6 I 8 18
46 45 51 50
Total
in
sample
Length
in mm.
25
37
40
42
43
44
45
46
47
48
50
52
53
54
55
57
58
Total
Av. length
Stages
34567 ABCDEFG
2 3 7 18
25 • 39 44 44 53
2 7 19
39 • 43 51 49
Total
in
sample
Date January 20, 1931
St. No. 603
Net N too B 140-0 m.
Locality
Position
Bellingshausen Sea
/65° 04*' S,
I. 67 5l*'W
Surface T. i-o8° C.
Length
in mm.
47
50
53
55
Total
Av. length
Stages
1234567
ABCDEFG
Total
in
sample
Date January 25-26, 193 1
St. No. WS 537
Net N 100 B 67-0 m.
Locality Approaching
S. Sandwich Is.
n ■.■ (56° 10' S,
Position |5 ,j, ^^.-^
Surface T. 0-57'' C.
Length
in mm.
23
24
25
26
27
28
29
30
31
32
33
34
35
40
43
49
50
Total
Av. length
Stages
1234567 ABCDEFG
33 21 2
25 35 39
42 13 •
25 33 43
Total
in
sample
Date January 20, 1931
St. No. 603
Net N 100 B 140-0 m.
Bellingshausen Sea
|6s' 04*' S,
I. 67° 5 1 *' W
Surface T. 108° C.
Locality
Position
Length
in mm.
45
53
56
Total
Av. length
Stages
1234567
3
51
ABCDEFG
Total
in
sample
Date January 25-26, 193 1
St. No. WS 537
Net N 100 B 67-0 m.
Locality Approaching
S. Sandwich Is.
Position {">;;o°;f.v
Surface T. 0-57° C.
Length
in mm.
23
24
25
26
27
28
29
30
31
32
33
36
37
47
51
Total
Av. length
Stages
3 4 5 6 7 ABCD'EFG
33 26 I
25 29 37
Total
in
sample
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
151
MALES
FEMALES
Date February 22, 1928 Locality S. Georgia
St. No. WS 152 Position ^53° 12 00 S
Net NiooBiit^om. Position ^ 34' 52'-oo W
Surface T. circa 130 C.
Date February 22, 1928 Locality S. Georgia
St. No. WS152 Position (5^° 'o^''°? ^',.r
Net NiooBlio-om. Position ^ 34° 52'oo W
Surface T. circa i-30° C.
Length
in mm.
Stages 1
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
34
41
43
44
45
46
:^
49
50
51
52
S3
54
55
57
I
I
. 2
. 2
. ■ ■ 7
. . . 4
: : : t
. . . 14
. . . 9
. . . 6
. . . 5
I
. • • 3
I
I
. . I
2
I . 2
I • 7
■ . 4
• . 4
. . 6
■ • 14
. • 9
. . 6
• ■ 5
I I
• ■ 3
I
I
I
I
2
4
4
6
14
9
6
5
2
3
I
42
44
45
46
47
49
50
51
52
53
54
55
56
59
60
I
I .
. 2 2 .
. 2 2 .
2 .
2 .
I I I
31
12 1
I I
I I I
2 .
I
1
I
I
4
4
2
I I
3
4
I 3
I I
2
I 2
I I
I
I
I
I
4
4
2
2
3
4
4
2
2
3
2
I
Total
Av. length
. . . . 7 23 6
. . . . 43 50 S3
5 31
53 50
36
Total
Av. length
. . 4 . . .66
. . 45 • • -so
I
3 . 66
70
■ . 34
46 . 50
Date February 26, 1928 Locality S. Georgia
St. No. WS 156 Position •f53°4p'ooS
Net N 70 V 100-50 m. losition -^ 36' jz'-oo W
Surface!'. circa 2-31° C.
Date February 26, 1928 Locality S. Georgia
St. No. WS 156 Position /53°4p'ooS
Net N 70 V loc^so m. Position ^ ^g. ^^,.^^ ^
Surface T. circa 2-31° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
24
25
26
28
29
30
31
32
33
34
11
37
38
39
40
41
42
1 2
4 I
2 3
2 2
2 10
. 6
I 4
. 7
. 6
. 2
• 4
I 1
I
1
I . . . .
I . . . .
I
. I
I
. 3
• 5
5
■ 4
. II
. 6
. 5
. 5
. 6
. 2
I
2
2
2
r . . .
I
I
3
5
5
4
12
6
5
7
6
2
4
2
I
I
2
25
26
27
28
29
30
31
32
33
34
35
36
37
38
2
2
I
3
5
6 s '.'.'.'. '■
32
13
26
22
. 2
12
II....
2
2
I
12
• 5
II
16
• 5
.4
.8
13
II....
■ 3
2 . . . .
2
2
I
3
5
2
7
5
4
8
4
2
3
2
, 2 .
I
I I
Total
Av. length
29 20 I .
30 34 38 . . . .
8 39 3 • ■ ■ •
28 32 37 ■ • • •
50
Total
Av. length
16 so 2 . . . .
30 34 42 . . . .
. 58 9 I • • •
. 33 37 41 ■ • •
68
Date February 8, 1929 Locality S. Orkney Is.
St. No. WS 376 Position 157° 23 -op S,
Net N 70 V 750-500 m. Position ^ 42^52'ooW
Surface T. i-45 C.
Date February 8, 1929 Locality S. Orkn
St. No. WS376 Position (57^^.3
Net N 70 V 750-SOO m. „ ^ ,, '• "tV
Surface T. 145 C
ey Is.
■00 S,
52'-oo W
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
40
46
50
5«
. . . I . . .
I .
I
I . . ■
I
I
I
I
I
I
I
50
53
I
I
I
I
I
I
Total
Av. length
I I
S3 50
2
52
2
Total
Av. length
. . . I . I 2
. . .40 .46 51
. . . I . I 2
. . .40 . 46 51
4
-
Date February i, 1930 Locality S. Georgia
St. No. 325 Position (54° 53 S.
Net N 70 V 750-Soo m. position ^ 39° 57' W
Surface T. 3-32 C.
Date February 14, 1929 Locality ^. Orkr
St. No. WS381 Position (*' Ao
Net N7oV5tw>m. = ' -r *-fi/»P
Surface T. 165 C
ey Is.
'00 S,
l9'-00 W
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
41
I . . ' .
I
I
48
I . ■ • •
'
. I . .
Total
At. length
. . 1 . . . .
. . 41 . . . .
• I
• 41
I
Total
Av. length
I
.48
. . 48 . . ■ •
I
IS2
DISCOVERY REPORTS
MALES
FEMALES
Date February 8, 1930 Locality
S. Georgia
Date
February 8,
1930 Locality
S. Georgia
r55° 01' s,
\ 35°27i'W
St. No. 349 Position
Net N 100 B 60-0 m.
f55°oi'S,
l 35°27i'W
St. No.
Net
N'ooB6c«m. P°^'"°"
Surface T.
300° c.
N 70 V 50-C
m. Surlace 1 .
Stages
Total
Stages
Total
Length
in mm.
in
G sample
Length
in mm.
in
sample
1234567
A B C E
E F
1234
5 6 7
A
BCD
E F G
I
1
36
4 . . .
■ 4 •
4
I
I
3Z
4 . . .
I
. 2 2 I
S
40
21....
. 3 ■ •
3
38
I . . .
I
41
I
I
1
39
3 • • •
• • 3 ■
3
42
. 22....
• ■ 3 I
4
40
2 . . .
1
2 .
3
I
I
I
41
I . . .
44
45
49
I . . . .
. . I . . . .
I . . . .
I
I
1
I
tl
3 • ■ •
1 . . .
--
• 3 ■
I
3
I
Total
Av. length
19 . . .
39 • • ■
Total
Av. length
.85....
. 40 44 . . . .
• 743- • •
. 39 43 45 • ■ ■
14
39 • •
. 37 39 37
. • 40
Date February 9. 1930 Locality
S. Georgia
Date
February 9.
1930 Locality
S. Georgia
(54°2li' S,
I 35°42'W
St. No. 351 Pn^itinn
Net N 100 B 48-0 m. Position
(54°2li'S
\ 35° 42- W
St. No.
Net
351 Position
N 100 li 40-0 m.
Surface T.
3-78° C.
N 70 V 100-50 m. Surface F.
3-78- C.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length
in mm.
m
sample
1234567
A B C D E F G
1234567
A B C D
E F G
42
2 . . . .
2
•
2
38
I
I
I
43
2 I
I 2
3
39
2
I I
2
40
2 2
4
2 2
4
41
■ 3 I
■ 31.
t
2
4
42
■ 2 3
■ 2 3 .
47
I 2
.
2
3
43
■ I 4
113.
5
48
I 2
3
44
5
113.
5
1
45
5
I .
4 I
50
I
5
46
I
52
.
I
47
2
2
S3
.
2
48
I
I
I
2
49
. 3 •
3 . •
3
SS
56
I
J
55
57
2 .
. . I*
I
2
58
• . 5*
5
is
I
60
. . 2«
. . 2 1 2
59
3
60
I
Total
• 9 24
. 8 8
4 14 15
6 . 10
49
64
I
Av. length
• 41 43
• 49 58
42 42 44 •
47 ■ 58
Total
. 22 10 7 2 I
. 4 11 I
772
42
Av. length
. . 47 SI 56 60 so 1 ■ 43 46 4*
! S3 58 53
Date
St. No.
Net
February 9
352
N 100 B s8
1930 Locality
Position
-0 m.
Surface T.
S. Georgia
(54°I9'S.
I 35°24'W
2-95° C.
Date February g, 1930 Locality
gfet""" ?l'.ooB58-om. P-"""
S. Georgia
(54° 19' S
\ 35° 24 W
Surface T
2-95° C.
Stages
Total
Total
in
sample
Length
m mm.
I 2 3
4567
A B C D
E F G
sample
1234567
ABC
D E F G
u
J
37
. . I . . . .
. I
I .
I
38
3 . . .
■ 3 •
3Z
2 .
40
• 3 4 • •
• 7 ■
38
2 I
41
3 4 ■ •
. 6 I
39
42
43
12..
. . 3 ■ •
1 2
2 I
40
41
2 I
5 I
I 5
6
4S
3 • ■
■ • 3
42
3 •
• 3
46
2
I
43
2 2 I
5
47
52
I . .
I
I
I
45
47
I
'. '. !'
^
I
54
I
I
55
Total
. 10 20 1 1
. 21 8
I . I I
32
Total
17 It 2
2
9 17 3 I
2
32
Av. length
. 40 42 . 52 54 ■
. 41 44 4
5 ■ 52 54
Av. length
40 40 44
. . . SI
38 41 42 43
. • 51
THE DEVELOPMENT AND LIFE-HISTORY OF KRLLL
153
MALES
FEMALES
Date
February 9, 1930 Locality
S. Georgia
Date
February 9,
1930
Locality
S. Georgia
St. No.
Net
nIoo B 96-0 m. ^°'"'°"
(54° ISJ'S,
\ 34= 47r W
St. No.
Net
N 100 B 96
-0 m.
Position
\ 34 47* W
Surface T
2 35' C.
Surface T.
235' c.
Stages
Total
Stages
j Total
Length
in mm.
in
sample
Length
in mm.
]
in
sample
234567
A B C D E F G
234567
A B C D
E F G
35
I
I
I
34
I
I . . .
I
38
I
I
I
36
2 .
I 1
39
2
2 .
2
37
2
2
40
I
I
I
38
I
1
41
2 1
2 I
3
39
42
8 10
4 14
18
40
3
2
3
43
I 4
3
5
41
3 I
3
4
44
. 2
2
2
42
45
9
6 .
9
43
2
'
I
3
46
I
I
44
2 .
2 .
47
4
2 :
4
45
3
I 2
3
48
1
I
46
I
I I
49
3
2
4
47
1
50
2
I
2
49
51
2
5
■ •Si 7
SO
I"
52
. 3 7
2 I 6 1 10
SI
3*
I 2
3
53
. . 6
.33! 6
52
3*
}
■ • 3
3
54
55
• 2 7
2 14
-.25, 9
I 4 10 16
53
.
I
8*
. 10
10
56
• I 5
- • 4 I 1 7
54
.
6»
2 5
7
57
. . 6
. 3 2
6
SS
4*
• I 3
4
58
• 2 s
2 4
7
56
3*
. . 3
3
59
I I
I
2
57
S*
I 4
5
60
. 2
2
2
59
■
1*
. 2
Total
IS 38 I . II 61
13 30 I<
) 3 20 41
126
Total
18 7
• 7 35
6 II 6 6
. 6 32
67
Av. length
-
41 45 56 . SS S5
-
41 44 S
53 55 S4
Av. length
-
40 44
■ 48 S3
37 41 45 44
• 54 55
Date
February 10, 1930 Locality
S. Georgia
Date
February 10, 1930
Locality
S. Georgia
St. No.
Net
N 70 V 250-100 m. P°^'"°"
f54°ll'S.
1 33° 49' W
St. No.
Net
N 70 V 250-100 m
Position
/S4° 11' S.
1 33' 49' W
Surface T
2-02° C.
Surface T.
202° c.
Stages
Total
Stages
Total
Length
in mm.
•
Length
in mm.
in
sample
J 2 3 4 5 6 7
A B C I
D E F G
sample
1234567
A B C D
E F G
40
I . . . .
1
I
38
I
I
I
41
2
2 .
2
39
43
I
I
42
I
I I
44
2
2
2
49
45
3
3
3
SO
I
46
4
2
4
52
2
48
6 I .
a 3
I
7
53
2
50
SI
I I
. I
I
1
2
I
54
55
I*
6
6
6
54
1 I
S6
I
I
I
56
I
1
57
3
i
3
57
I
I
1
58
S
b
59
.... I
1
I
59
60
• I 3
1 9
4
10
Total
20 3 . I 3
6 II
i 3
4
27
Av. length
46 50 . 57 s6
44 46 4
3 SO
57
61
63
I*
1
. • 5
5
64
I
I
■ Total
3
J 42
I 2 2
. 3 40
48
Av. length
40
. 5
3 58
39 39 45
• 57 58
Date
Februarys, 193 1 Locality
Bransfield Strait
St. No.
1 62" oSi' S.
Date
February 8
1931
Locality
Bransfield Strait
Net
n'.oo B 128-0 m. P°^'"°"
1, 62' 57i' W
St. No.
609
/62° o8i' S,
1 62°57i'W
Surface T
203° C.
Net
N 100 B 12
i-o m.
Surface T.
2-03° C.
Stages
Total
Length
in mm.
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C I
3 E F G
123^
1 5 1
J 7
A B C D
E F G
47
I
I
I
48
I
I
I
49
52
3
• • 3
3
SI
I
53
3
. . 3
3
55
Total
8
. . 8
8
Total
i ■
2 I
3
Av. length
SI
. • SI
Av. length
. 5
! .
. 52 51
7-2
IS4
DISCOVERY REPORTS
MALES
FEMALES
Date
February 10, 193 1 Locality
Bransfield Strait
Date
February 10, 193 1 Locality
Bransfield Strait
St. No.
Net
n'.ooB 160-0 m, P°''"°"
(62° 42' S,
I 57°io'W
St. No.
Net
n'ioo B .60-0 m. P°="'°°
\ 57" 10'' W
N 50 V 1 00-0 m. Sm-face T
052" C.
Surface T.
0-52" C.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length _
in mm.
m
sample
1
I 2345671
\ B C D E F G
I 234567
A B C D E
F G
zo
: 1
1
I
16
I
I . . . .
I
21
2
2 .
2
19
I
I
I
22
I
I
I
21
I
I
I
23
6 . .
6
6
22
3
3
3
24
3 4 ■
7
7
23
5
5
5
25 I
■ 5 .
I IS
16
24
20
20
20
26
9 20 .
. 29
29
25
41
41
41
27 I
0 21
. 31
31
26
31
■ 3
31
28
3 13 .
. 16
16
27
. 34 •
3J
34
29
I 10 2
. 13
13
28
. 18 .
18
18
30
I .12 3
I 35
36
29
19 .
19
19
31
I 13 I
1 14
15
30
• 42
42
42
32
. 12 2
. 14
14
31
. 12
12
12
33
6 I
7
7
32
24
24
24
34
2 3
4 I
5
33
IS
'1
IS
35
.
7
6 2
8
34
6
6
36
.
2
I 2
3
35
20
19
37
4
2 3
36
6
6
6
38
5
6 2
8
37
3
3
3
39
.
3
2 I
I
4
40
2*
40
.
3 I •
I
3 I
5
41
I*
41
I 3 •
i ■
4
42
42
.
3 . ■
4
4
43
3;
3
3
43
3 • J
3 •
4
44
V*
7
7
44
I 2
3 •
3
45
14*
1 13
14
45
8 2 I
I 10
tl
46
7*
7
7
46
I I
2
47
4*
4
4
47
9 7 •
692.
17
48
6»
b
6
48
2 3 •
221.
5
49
i
5
49
6 3 .
2 7 ■ ■
9
SO
6
6
50
4 4 ■
.71.
8
51
51
2 2 I
• 321
■ 3 I •
131-
6
4
5
S3
I*
52
53
I 4 ■
Total
03 113 86
■ 58
185 116
I
1 57
360
54
12...
I 2 .
3
Av. length
24 29 33
■ 47
27 31
. 35
45 47
Total
« 146 77 39 4 2
3 212 16 27 47 II I
317
Av. length
25 29 41 48 48 49 . 29 29 38 4
5 48 50 51
Date
February 12, 1 93 1 Locahty
S. Orkney Is.
Date
February 12, 193 1 Locality
S. Orkney Is.
St No.
Net
N^ooB 182-0 m. P°^'"°"
|6o° 59}' S,
1 50°42i'W
St. No.
Net
N^ioo B .82-0 m. P°''"°"
f6o° 59}' S,
1 50°42i'W
Surface T
-021° C.
Surface T.
-0-2I° C.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length
in mm.
in
sample
I 234567
ABC
D E F G
I 234567
A B C D
E F G
45
I
I
I
22
23
26
35
42
I ...
I . . .
Total
I .
I
I
I ...
I
Av. length
45 ■
• • 45 ■
i<
> _
Total
3 I .... I
3 ■ I ■
. . I
5
Av. length
24 35 ■ • ■ -42
24 • 35 •
• ■ 42
Date
Februarv 18-19, 1931 Locality
S. Orkney Is.
St. No
Net
N^iooBii9-om., P°^"'°"
N 70 V 50-0 m. Surface T.
(59° 42}' S
I 43°57rW
1-29° C.
Date
St. No.
February 18-19, ^931 Locality
N''iooBil9-om., P°^'"°"
N 70 V 50-0 m. Surface 1
S. Orkney Is.
159° 42}' S
\ 43 57? W
r. 1-29° C.
Stages
Total
Net
Length
in mm.
in
sample
12 34567
A B C D
E F G
Length
in mm.
Stages
Total
- in
sample
26
27
28
2
2 ...
I ...
I
12..
2
I
3
I 234567
ABC
D E F G
27
2
2
2
30
4 3
7
30
21...
• 3 •
3
31
I I
2 ,
2
32
42...
4 2 .
6
32
■ 3
. 3 ■ •
3
33
15...
I 5
6
33
2 4
24..
34
9 . . .
9 .
9
34
2
35
4 ■ ■ ■
4 •
4
35
I II
II I
12
36
3 • • •
3 .
3
36
3
4 ■ •
4
37
4 . . .
4 •
4
37
5
5 . ■
5
38
3 . • •
3 ■
3
38
4
4
39
2 . . .
2 .
2
40
1
40
1 . . .
I
I
45
3
41
2 . . .
2 .
2
50
I
45
I
51
2
51
I
I
52
54
I .
. . .
I
54
4
4
Total
9 36 I . . 2 .
7 38 I
48
Total
15 37 I ... II
7 42 4 ■
. . II
64
Av. length
31 35 45 • -53 .
31 35 45
■ ■ • 5
Av. length
29 35 36 . . . 5C
29 34 38 .
■ ■ 5C
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
155
MALES
Date February 20, 1931
St. No. 6zi
Net N 100 B loo-o m.
Locality S. Orkney Is.
Position {=';|?53'''w
Surface T. 0-25' C.
Length
in nun.
39
42
Total
Av. length
Stages
1234567
A B C D E F G
Total
in
sample
FEMALES
Date February 20, 1 93 1 Locality S. Orkney Is.
St. No. 621 p„^,>i„n /58° 5oi' S,
Net N 100 B 100-0 m. Position | ^go 53. yj
Surface T. 0-25° C.
Length
in mm.
Date
St. No.
Net
February 20, 1931
622
N 100 B 155-0 m.,
N 70 V 50-0 m.,
N 50 V loo-o m.
S. Sandwich Is.
/59° 05*' S,
I 36°25'W
Surface T. -0-89° C.
Locality
Position
Length
in mm.
25
26
27
30
32
33
34
35
36
37
38
39
40
41
42
43
49
Stages
234567 ABCDEFG
Total
Av. length
29 24 6
32 37 41
3 51 5
26 35 41
Total
in
sample
60
Date
St. No.
Net
February 21-22, 1931
624
N 100 B 137-0 m.,
N 70 V 50-0 m.
Locality
Position
Surface T.
S. Sandwich Is.
/58° 34}' S,
I 3i°2ii'W
022° C.
Length
in mm.
23
24
25
26
27
28
29
30
31
32
35
36
37
41
Total
Av. length
Stages
234567 ABCDEFG
41 3 2
27 35 38
16 28 2
24 30 38
Total
in
sample
46
Date February 22, 193 1
St. No. 626
Net N 100 B 158-0 m.
S. Sandwich Is.
/57° 22' S,
1 25° 29i' W
Surface T. -009° C.
Locality
Position
Length
in mm.
23
24
25
29
Total
Av. length
Stages
34567
ABCDEFG
5 4
II 28
28
Total
Av. length
Stages
1234567
28
ABCDEFG
28
Total
in
sample
Total
in
sample
Date February 20, 1931
St. No. 622
Net N 100 B 155-0 m.
Locality
Position
Surface T.
S. Sandwich It
/59°o5S' S. ,
I 35° 25' W
-089° C.
Length
in mm.
23
25
27
28
29
30
31
32
33
34
35
36
38
30
40
41
50
S6
Total
Av. length 27 34
Stages
1234567 ABCDEFG
13 29
7 25 10
26 32 36
Total
in
sample
Date February 21-22, 1931
St. No. 624
Net N 100 B 137-0 m.
Locality
Position
Surface T.
S. Sandwich Is.
(58° 345' S,
'(. 3l°2lJ'W
022' C.
Length
in mm.
23
24
25
27
29
30
31
32
35
36
48
Total
Av. length
Stages
34567 ABCDEFG
10 15
27 30
48
9 15 I
27 29 36
48
Total
in
sample
Date February 22, 193 1
St. No. 626
Net N 100 B 158-0 m.
S. Sandwich 1
f57° 22' S.
Locality
Position {'■26° 291' W
Surface T. -009° C.
Length
in mm.
18
19
23
24
26
28
30
32
Total
Av. length
Stages
234567 ABCDEFG
15 4
21 30
14 5
21 29
Total
in
sample
156
DISCOVERY REPORTS
MALES
Date
St. No.
Net
February 23, 193 1
627
N 100 B 1 18-0 m.
Locality
Position
Surface T,
S. Sandwich Is.
f56°S3rS,
\ 23° 474' W
110° C.
Length
in mm.
23
24
25
26
27
28
29
30
31
32
33
34
35
35
37
38
40
45
51
Total
Av. length
Stages
3456
31 18 5
27 33 40
A B C D E F G
23 26 4 . I
27 33 37 • 45
Total
in
sample
Date February 24, 193 1
St. No. 628
Net N 100 B 126-0 m.
Locality S. Sandwich Is.
Position {"aa'l+'^W
Surface T. -015° C.
Length
in mm.
Stages
1234567
A B C D E F G
Total
in
sample
FEMALES
Date February 23, 1931
St. No. 627
Net N 100 B I iS-o m.
Locality
Position
Surface T.
S. Sandwich Is.
156° 538' S,
I 23'47rW
llo" C.
Length
in mm.
25
26
27
28
29
30
31
32
33
34
35
36
38
Total
Av. length
Stages
234567 ABCDEFG
2 44
23 29
13 32 I
27 30 38
Total
in
sample
46
Date February 24, 1931
St. No. 628
Net N 100 B 126-0 m.
Locality
S. Sandwich Is.
Position {'Qo^^^/i^
Surface T.
-015° C.
Length
in mm.
Stages
34567
ABCDEFG
Total
in
sample
26
30
Total
Av. length
Total
Av. length
26 30
Date February 25, 193 1
St. No. 629
Net N 100 B 152-0 m.
Locality
Position
Surface T.
S. Sandwich Is.
— S. Georgia
[55° 331' S,
X 30° 01' W
oii°C.
Length
in mm.
23
29
32
34
36
Stages
6 7
ABCDEFG
Total
in
sample
Date February 25, 193 1
St. No. 629
Net N 100 B 152-0 m.
Locality S. Sandwich Is.
— S. Georgia
Position {";o3o^i;?w
Surface T. o!i°C.
Length
in mm.
25
30
Stages
234567
ABCDEFG
Total
in
sample
Total
Av. length
5 4
25 35
3 6
21 33
Total
Av. length
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
157
MALES
FEMALES
Date March 14, 1926
Locality
S. Georgia
Date
March 14, 1926
Locality
S. Georgia
St. No. 23
Position
Cumberland Bay
St. No.
23
Position
Cumberland Bay
Net N 100 H 6o-(o) m.
Surface T.
267° C.
Net
N 100 H 6o-(o) m.
Surface T.
2-67' C.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length
in mm.
m
sample
123456
7
A B C D E
F G
123456
7
A B C D
E F G
26
I
35
3
I 2 .
3
37
I
.
. . I . .
I
36
3
. 2 I .
3
38
1
I
I
37
2
. I I
2
40
• • 5 .
. .23.
5
38
22....
. 2 2
4
4J
I
1
I
39
3
I 2 .
3
42
■ ■ 4 ■
. 2 2 .
4
40
3
. . 3 •
3
43
t- z . .
. . . 3 ■
3
41
4
. . 3 I
4
44
..63..
. . I 8 .
9
42
4
. I 3 •
4
45
. .32..
. . I 4 .
5
43
3 3 • • ■ .
• • 3 2
6
46
. .42..
. . . 6 .
6
44
I . . . .
1
I
47
2 . . .
2 .
2
48
l»
1
48
. . 2 2 .
. . . 3 1
4
49
:•
I
49
I
1
I
Total
27 6 ... .
2
. 6 21 5
I . 2
35
Total
I . 30 12 . .
I . 9 31 2
43
Av. length
39 42 . . . .
49
. 38 40 41
43 ■ 49
Av. length
26 . 43 45 ■ •
26 . 41 44 4?
.
Date March 14, 1926
Locality
S. Georgia
Date
March 14, 1926
Localitv-
S. Georgia
St. No. 24
Position
Cumberland Bay
St. No.
24
Position
Cumberland Bay
Net N 100 H 6o-(o) m.
Surface T.
circa 2-67° C.
Net
N 100 H 6o-(o) m.
Surface T.
circa 267° C.
Stages
Total
Stages
Total
Length
in mm.
m
sample
Length
in mm.
m
sample
I 2 3 4 5 f
7
A B C D E F G
123456
7
A B C D
E F G
40
I
I
I
35
2
. I I
2
42
I
I
I
36
2
. 2
2
45
. . I I .
I I
2
37
2
. 2 .
2
46
I
. . . I
1
38
5
5
■23.
I 4
5
5
Total
. .41.
. 122
5
40
41
14
12
. I 9 4
• 552
14
12
Av. length
• ■ 43 45 ■
• 40 44 46
42
■i
. 2 10 I
13
8
44
42....
• • 4 2
6
45
81....
■ 2 3 4
9
46
I
I
I
47
1*
. . 3
3
48
1*
I
I
49
I*
. . I
I
55
I*
1
I
Total
76 3 ... .
6
. 18 45 16
. . 6
8s
Av. length
41 44 • ■ ■ .
49
. 41 41 43
• . 49
Date March 19, 1926
Locality
S. Georgia
St. No. 38
Position
Cumberland Bay
Net N 100 H 50-(o) m
Surface T.
circa 2-85° C.
Stages
Total
in
sample
Date
March 19, 1926
Locality
S. Georgia
Length
in mm.
St. No.
Net
38
N 100 H so-(o) m.
Position
Surface T.
Cumberland Bay
circa 2-85' C.
1234s
6 7
A B C D
E F G
Stages
Total
36
11
39
I
I
I
3
Length
in mm.
in
sample
': 3 :
123456
7
A B C D
E F G
40
I 2
• • 3 •
3
35
I . . . .
I
I
41
. I 4
• I 3 I
5
37
2 . . . .
. 2 .
2
42
2
I I
2
38
I 3 . • • ■
. 2 2 .
4
43
■ ■ 3
. • 3 I
4
39
. 3 ■ • ■ •
■ ■ 3 •
3
44
• • 4
2 2
4
40
. 7 . . . .
. I 5 I
7
45
. . 8
■ • 4 5
9
41
12...
I 2
3
46
• ■ 3
. 2 I
3
42
■ 3 3 • • .
■ ■ 5 I
6
47
• • 4
. . I 4
5
43
. 2 . . . .
2 .
2
48
• . 4
. • I 3
4
44
. 6 . . . .
■ ■ 5 I
6
49
I
2 .
2
I
3
45
.36...
• I 5 2
9
50
I
I
I
46
. 2 5 . • •
.241
7
Total
. 7 36 6 .
. 7 21 20
I
49
Total
I 33 16 • . •
• 6 35 8
I . .
50
Av. length
• 39 44 47 •
• 39 43 46 4
9 • .
Av. length
38 38 44 . . ■
. 42 42 43
45 . .
IS8
DISCOVERY REPORTS
MALES
FEMALES
S. Sandwich Is.
Date
March 8, 1930
Locality
S. Sandwich Is.
St. No. 368 Position
Southern Thule
St. No.
368
Position
Southern Thule
Net N 100 B 146-0 m. Surface T.
011° C.
Net
N 100 B 146-0 m.
Surface T
on" C.
Stages
Total
Stages
Total
Length
in mm.
m
sample
Length
in mm.
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
38
I
. . I .
• •
I
36
. I .
I
40
2 . . . .
r
2
37
2 . .
2
42
I . . .
r
I
39
43
22...
3
4
40
44
. . 4 . . .
2 2
4
41
2
45
. 24...
4 I
6
42
46
III..
2 I
3
43
2 2
4
47
. . s • • •
4 I
5
44
2 2
I - 3
4
48
II..
. 2
2
+1
121..
4
4 • • ■
2 2
4
46
III..
I . I
I
3
I 3 ■ ■
I I
I
4
47
112..
4
I
1
48
I 3 • •
I
r I
4
52
55
56
2 . . .
. 2
. I
I
2
49
SO
51
I'
. . 2»
. . I*
I
2
I
57
. I
1
52
2
S8
.... 1
I
I
53
. 2*
54
• • 3*
. . 2*
. . 2*
I 2
2
. 2
3
2
2
Total
. 8 25 5 • 32
. 3 21 12
2 4
43
Av. length
. 43 47 49 • 55 57
■ 43 45 48 50 51 57
60
.
'
. . I
I
Total
3 12 10 7 . .16
6 8 9 6 2 3 14
48
Av. length
37 43 44 47 ■ . S3
41 41 46 46 47 51 54
Date March 9, 193 1 Locality
S. Orkney Is.
Date
March 9, 193 1
Locality
S. Orkney Is.
St. No. 638 Po<iition
Net N 100 B 155-0 m. I'osition
/6i° 00*' S,
1 49°48i'W
St. No.
Net
638
N 100 B 155-0 m.
Position
f6i' 00 J' S.
t 49^48i'W
Surface T.
-020° C.
Surface T
-0-20*' C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
25
I
32
I
I . .
37
34
40
43
SI
I
I
3^
46
48
1
*
I
. . I
53
I
49
51
52
55
*
I
2
I
2
I
Total
I . 2 . I 2 .
I . . 2
2
6
Av. length
25 .42 • 37 52 .
25 • ■ 42 37 • 52
3
7
3 . .
. . 7
Av. length
34 52
34 • .
■ ■ 52
Date March 9, 1 93 1 Locality
S. Orkney Is.
Date
March 9, 193 1
Locality
S. Orkney Is.
St. No. 639 Position
Net N 70 V 50-0 m. Position
(6i°57rS,
\ 5I°59'W
St. No.
Net
639
N 70 V so-o m.
Position
f6i ' 571' S,
l 51- 59'W
N so V loo-o m. Surface T.
-076° c.
Surface T
-0-76* C.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length
in mm.
in
sample
1234567
A B C D E F G
I 2 3 4 5 6
7
A B C D E F G
25
I
I . . .
I
25
I
I
I
26
I
1 . . .
I
26
1
I . .
29
I
I
33
-
I
I
I
Total
21
21..
3
Total
3
3 . .
3
Av. length
26 29
26 29 .
Av. length
28
—
28 , .
Date
March 10, 1 931
Locality
S. Shetland Is.
St. No.
Net
640
N 100 B 164-0 m.
Position
Surface T
/6i^ 26i' S,
1 53^47rW
-0-20'= C.
Stages
Total
Length
in mm.
in
sample
123456
7
A B C D E F G
Date March 10, 1931 Locality
S. Shetland Is.
^'et"""- n' too B. 64-0 m. P-"""
f6i° 26i' S,
1 53°47i'W
25
32
1
I
I
Surface T.
-020° C.
43
3*
- • ■ 3
3
45
46
I*
2*
I
2
2
2
2
I
2
Stages
Total
Length
48
49
2
2
in mm.
1234567
A B C D
E F G
sample
2*
23
I ...... .
I
I
51
r»
I
I
25
I
I
I
53
*
42
• I
I
I
54
Total
21
. 2 I .
3
Total
2 I
7
2 . .
. . . 17
19
Av. length
24 42 1 . 24 42 .
Av. length
29 48
29 . .
. . . 48
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
159
MALES
i
FEMALES
Date
March 10. 193 1 Locality
S. Shetland Is.
Date
March 10, 193 1 Locality S. Shetland Is.
St. No.
642 Position
1 S4°S8'W
St. No.
542 Position /6i°sii'S,
N 100 B 180-0 m I 54° 58' W
Net
N 100 B 180-0 m.
Net
Surface T.
0-70° C. 1
Surface T. 070° C.
Stages
Total
Stages 1
Total
Length
in mm.
in
sample
Length
in mm.
in
sample
1234567
A B C D E F G
1234567
A B C D E F C
27
I
49
2*
2
2
32
I
I
Total
2
2
2
Xi
. . . . I . .
Av. length
49
49
Total
2 I I . I . .
2 I I I
5
Av. length
28 32 42 .48 . ■
. 28 32 42 48
Date
March 10, 1931 Locality
S. Shetland Is.
Date
March 10, 1931 Locality S. Shetland Is.
St. No.
643 Position
I61 44i'S,
St. No.
643 Position /6i° 44 J S,
Net
N 100 Bl ,
N 70 B / 93-0 m. Surface T.
I 56= of W
Net
N 100 B1 „, „ „ I 56' 07' W
N 70 B ) 93-0 m. Surface T. o-5o° C.
o-6o° C.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length
in mm.
in
sample
1234567
A B C D E F G
1234567
ABCDEFG
30
■ 4
■ 4 ■ ■
.
4
23
I
I
33
, 2 .
. 2 .
2
30
3 • ■
3
34
• 5 I
5 I •
6
32
I
35
3 I
■ 3 1 ■
4
33
I I
36
3 ■
3
34
I 3 •
2 3
5
3 ■
3
35
8 . .
I I :: 3
8
38
■ 3 3
6 .
6
36
8 4 .
232s
I
I
H
4 I
I I I 2
5
40
. 232
5 2
7
38
3 2 I
2.13
1 2
3
3
39
3 4 ■
■ 4-3
7
42
I I
I
2
40
1411
223
7
43
.
I
41
44
2
2
42
4
45
1
I
43
I
46
I I
2
3
44
47
I
I
I
45
3
48
I
46
b'l
7
. . . 2 3
2
6
47
2
4
50
51
... 2 6
3
2
1
8
3
48
49
. 2
I 5*
■ 7!
5
7
7
52
. . . 2 I
1
3
50
"•.
14
14
S3
54
55
. . . . I
2
. . . . I
I
I .
I
I
I
2
I
I
3
2
51
52
S3
5*
8*
5
8
5
5
54
55
57
?:
5
1
5
I
Total
. 20 21 16 16 3 .
. 14 22 16 2
3 2 2
76
Av. length
■ 35 39 47 SI S3 •
. 33 38 45 5
I 54 52
58
j»
I
I
Total
36 22 2 II I 2 60
10 17 9 22 16
. 60
134
Av. length
35 38 39 45 43 48 SI
34 35 37 38 44
• SI
Date
S. Shetland Is.
Date
March 11, 1931 Locality S. Shetland Is.
St. No.
644 Position
f6i°2oi' S,
I s5°4o'W
St. No.
644 Position /6l° 20^ S.
N 100 B loo-o m. I 56° 40' W
Net
Net
Surface T.
i-Si°C.
Surface T. 1-51° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
1234567
ABCDEFG
4?
48
49
50
52
I
2 .
I
I
2 .
2
I
I
2
I
2
1
I
2
49
51
54
55
3*
:::::: 1:
2»
:•
2»
3
3
2
2
I
2
3
3
2
2
1
Total
....16.
I
3 3
7
Total
13
13
13
Av. length
. . . . 40 SO .
. . . 40
• 48 51
Av. length
50
so
Date
March II. 1931 Locality S. Shetland Is.
St. No.
Net
646 Position r6o° 22* S,
N 100 B 1 55-0 m. I S7° 43'W
Surface T 2-33"' C.
Date
St. No.
Net
March 11, 193 1 Locality
646 Position
N 100 B 155-0 m.
Surface T.
S. Shetland Is.
f6o° 22i' S,
1 S7°43'W
2-33° C.
Stages
Total
Length
in mm.
in
sample
Stages
Total
I 234567 ABCDEFG
Length
in mm.
m
sample
1234567
A B C D
E F G
43
1*
. . . . I*
I
I
I
48
52
I
I
I
I
I
I
51
2*
I»
2
I
2
I
Total
2
2
2
Toul
5
5
5
Av. length
SO
. • so
Av. length
. ' 47
47
i6o
DISCOVERY REPORTS
MALES
FEMALES
Date March 26, 1931
Locality
Approaching
Date
March 26, 193
Locality
Approaching
St. No. 6s8
S. Georgia
St. No.
658
Net N 100 B 120-0 m.
Position
t. 40° 27i' W
Net
N 100 B 120-0 m. Position
|53°38rS
t 40°27i'W
Surface T.
2-64° C.
Surface T
264° C.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length
in mm.
in
sample
1234567
A B C D
E F G
1234567
A B C D E F G
20
I . . . .
I . . .
I
20
1 . . .
. .
I
I
21
2
2
2
21
5 ■
5 • .
5
22
2
I I
2
22
6 .
4 2 .
6
7 .
6 I
7
23
17 ■
12 5 .
17
24 10 I
10 I
II
24
16 .
8 8 .
16
25 '11 5
13 3
16
25
25 8
8 25 .
33
26 1 7 5
8 4
12
26
15 I
6 10 .
5 7
12
27
S 8
13 .
'3
28 1
I 7
2 6
8
28
7 I
7 •
29
4
4
4
29
3 I
4 •
*
. 3
3
3
30
5 J
31
• 7
7
7
31
2 I
3
32
2
2
2
32
3 ■
3 •
3
34 '
I
I
I
33
2
2 .
35
I
J
I
34
2
.
2 I
36 1
I I
2
2
35
I
40 j
I
I
I
36
48
I
I
I
.
Total 48 42 I I .
Av. length 25 29 36 40 .
48 43 I •
25 29 40 .
92
i»
. . I
Total
Av. length
110 27 2 .
25 28 36 .
I
. 48
45 91 2 I . . I
24 27 33 38 . . 48
140
Date March 24. 1931
Locality
S. Georgia
Date
March 24, 193
Locality
S. Georgia
St. No. WS 572
Position
l53°iii'S,
\ 37°4I'W
St. No.
Net
WS 572
N 100 B 190-0
„ Position
m
(53° Vi'S.
1 37°4I W
Surface T.
204° c.
Surface T
204° C.
Stages
Total
Stages
Total
Length
in mm.
m
sample
Length
in mm.
sample
12345
J 7
f
L B C D
E F G
1234
; 6 7
A B C D E F G
I . . . .
.
I
21
I . . .
I
I
22
1 . . . .
I
24
2 . . .
.
I 1
2
23
I . . . .
1
25
3 . • ■
• •
I 2 .
3
24
2 . . . .
II..
2
30
I . . .
28
I . . .
I
I
31
I . . .
• ■
29
49
t
• • '
Total
Av. length
8 . . .
26 . . .
3 5 •
23 27 .
8
Total
Av. length
5 2...
23 29 . . .
I
49
4 3 ■ •
23 27 . .
I
• . 49
8
Date March 25, 1931
Locality
S. Georgia
St. No. WS 573
Net N 100 B 124-0 m.
Position
Surface T.
f52° S9r s,
1 37°48'W
213' C.
Stages
Total
Length
in mm.
m
sample
12345
f> 7
A B C D
E F G
22
I . . . .
I . . .
I
25
z . . . .
1 . . .
I
Total
Av. length
2
24
2
24
2
Date
St. No.
March 26, 193
WSS75
I Locality
S. Georgia
152° 35' S,
Net
N 100 B 78-0 m.
\ 38°09'W
Surface T
282° C.
Date March 26, 193 1
St. No. WS 575
Locality
S. Georgia
(52° 35' S,„,
Stages
Total
Net N 100 B 78-0 m.
Surface T.
I 38° 09' W
2-82° c.
Length
in mm.
in
sample
1234
5 6 7
-
\ B C D E F G
Stages
Total
'
I . . .
*
in mm.
12345
6 7
A B C D
E F G
sample
29
I . . .
I
I
29
31
2 . . .
2 .
2
30
32
32
33
. 3 . .
. 2 .
■ 3 •
. 2
3
2
11
It
. I . .
I
I
I
I
I
Total
. 5 • • ■
• 5 . ■
5
Total
67..
2 II
13
Av. length
. 32 • • •
. 32 . ■
Av. length
29 33 • .
25 32 .
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
i6i
MALES
FEMALES
Date April I, 1926 Locality S. Georgia
' St. No. 42 Position Cumberland Bay
Net N 7-T 120-204 m. Surface T. circa 2-08° C.
Date April i, 1926 Locality S. Georgia
St. No. 42 Position Cumberland Bay
Net N 7-T 120-204 m. Surface T. circa aoS" C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
35
38
39
40
41
42
43
44
n
47
I
. 2 .
. .41.
. .91.
. . 5 ■ •
..71-
• -73-
. .91.
. .42.
2 I
. . 4 . .
i I :
4 I
2 6 .
2 8 .
I
2
5
10
5
8
10
10
6
3
4
36
37
38
39
40
42
43
44 1
. 2
.41. ■ ■ •
II....
4 . . . .
. . 6 . . . .
. . 6 . . . .
I . . . .
I . . . .
. 2 . . . .
.32....
2 . . . .
. 22....
. 3 3 • • • •
.42....
I . . . .
2
5
2
4
6
6
I
I
21..
. 4 - • •
Total
Av. length
. 7 20 . . . .
. 37 41 . • • ■
. 12 15 . . • •
. 40 39 • • • ■
27
Total
Av. length
■ • 54 10 .
■ . 42 43
. 2 26 35 I • •
• 37 4J 43 46 ■ .
64
Date April 7, 1927 Locality Bransfield Strait
St. No. 207A Position f62° 54'oo S,
Net N7oHc^5m Position | sg^jo'-joW
c -c T" jFrom —078'
Surface T | ,0 -o-86° C.
Date April 7, 1927 Locality Bransfield Strait
St. No. 207A Position 1''^° 5/°,° S. „
Net N 70 H 0-5 m. i-osmon -^ jg» 5° -30 W
c.._f„.-. -r (From -078
Surface T. | ,0 -o-86» C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
32
33
35
35
11
39
40
48
49
50
51
52
• • 3 1
1 I
1
I
I
■
I
4
2
I
3
I
2
4
2
1
3
\
I
30
31
32
33
11
39
40
41
43
44
45
46
47
48
49
50
51
53
1
I
2
;:
;:
4*
I*
1
2
2
• 3
. 6
■ 4
I
I
2
2
Total
Av. length
. I 18 4 3 4 .
• 35 38 47 50 50 .
.557661
. 35 36 40 46 49 49
30
... I*
I*
. . . . I
I
Toul
Av. length
11
35 ■
. . 5 22
. ■ 46 47
. 3 8 . 6 . 21
. 32 36 .46 . 47
38
Date April 7, 1927 Locality Bransfield Strait
St. No. 207B Position C"^"'.*'"?^',,,
Net N7oHo-5m. fosmon -^ g. ^^..j ^
Surface T. {'^^^-'^1' C.
Date April 7, 1927 Locality Bransfield Strait
^'et''°- N^7oH<.5n,. P-''^™ i^'l^^^^V^
Surface T. {^^?-_-^,fc.
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
Length
in mm.
Stages
Total
in
sample
U
30
31
32
33
34
11
11
39
40
41
42
44
45
46
47
50
52
I
I
4
5
7
8
6
16
7
4
3
2
4
3
- 2
I
4
I
2
I
I
1234567
A B C D E F G
I
I
2 .
1 I
I I
z
i»
2*
I*
4*
I»
. z*
. 1*
. I*
I
. 2
I
• 4
I
. 2
I
I
26
11
30
31
32
33
34
35
36
37
38
39
40
41
43
45
52
. 4
■ 3
I 2
I
I
I
I
3
3
I
2
I
3
I . .
:
3
2
2
3
2
I
I
2
2
I
I
2
I
2 .
2 . .
I I
I
I
I
1
2
4
3
3
5
2
I
2
3
3
4
4
2
I
I
I
2
4 •
! 5 ■
^ 4 •
2 I
: 12 2
! 5 ■
2 .
Total
Av. length
4 13 18 6 I . .
30 32 39 4' 52 . .
7 15 9 4 5 2 .
30 34 38 +1 43 49 •
42
Total
Av. length
54 6 5 5 ■
33 37 40 37 •
. 13
. 45
. 23 39 5 3 '3
■ 34 35 37 38 . 45
83
8-2
l62
DISCOVERY REPORTS
MALES
FEMALES
Date
April 7, 1927
Locality
Bransfield Strait
Date
April 7. 1927
Locality
Bransfield Strait
St. No
Net
207 C
N 70 H 0-5 m
Position
_(62" S4''00 S,
I S9=So'-3oW
St. No.
Net
207 C
N 70 H 0-5 m.
Position
/62^ 54''00 .S.
I 59° So'-30 W
Surface T
(From —0-78^
\ to -o-86° C
Surface T.
/From —0-78°
\ to -0-85° C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
12345
6 7
A B C D
E F G
25
I . . .
,
26
I . . . .
I
I
28
I I
2 .
2
28
I . . . .
I
1
29
I
29
3 . . . .
. 2 I .
3
30
I
2
30
6 . . . .
•24.
6
31
1
3'
I . . . .
I
I
32
I
32
3 • • . ■
2 I
3
33
3 I
2 I
4
33
5 . . . .
• 23.
5
34
3
34
2 . . . .
2
2
36
I
I
2
35
5 . . . .
• 23.
5
37
2
2 .
3
37
• • 3 ■
3
38
3
2 .
4
38
.31..
1 I
2 .
4
39
3
3
39
3 I • ■
2
2
4
41
3 2
I 4
5
40
■ ■ 3 ■
2
I
3
42
2
2
41
I
I
I
44
1 2
3
43
. I*
.
. . I
I
45
I
1
44
2*
2
2
47
■ 4
■ . 4
.
4
45
1
50
I
..21.
3
46
• 3*
• . 3
3
SI
2
.
I •
2
47
■ 4*
4
4
53
. .
5
.33.
6
48
. 3*
• • 3
3
49
• 3*
• ■ 3
3
Total
I 16 14 14 i
i . .
3 16 10 7 12 5 .
53
5°
I 4*
• I 4
5
Av. length
28 32 39 46 5-
29 33 39 41 49 52 .
52
. I*
I
I
Total
30 6 2 5 .
I 22
. 11 20 5
7 I 22
66
Av, length
32 39 39 41 •
so 48
• 3J 33 39
40 so 48
Date
April 7, 1927
Locality
Bransfield Strait
Date
April 7, 1927
Locality
Bransfield Strait
St. No.
Net
207 D
N 70 H 0-5 m.
Position
(62' 54'-oo S,
I S9°So''3oW
St. No.
Net
207 D
N 70 H 0-5 m.
Position
J62" 54' 00 S,
i S9"5o'-3oW
Surface T.
1 From —0-78°
t to -o-86° C.
Surface T.
/From —0-78°
1 to -086° C.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length
in mm.
-
in
sample
1234
5 6 7
ABC D E F G
12345
6 7
A B C D
E F G
30
2 .
I
2
27
I . . . .
. I . .
I
33
I
I
2
30
2 . . . .
I I
2
34
I
31
3 . . . .
. 1 2 .
3
36
I
32
I . . . .
I
I
38
I
33
2 . . . .
2 .
2
47
I
I
I
34
2 . . . .
I
2
48
1
I
I
36
I . . . .
I
I
49
2
I I
2
42
.
*
I
I
51
2 .
2 .
2
45
• 4*
■ • 4
4
53
1
I
I
46
• *:
■ • 4
4
54
J . .
2 .
2
'*2
• 4*
• • 4
4
55
J .
2 .
2
48
• 9»
• • 9
9
56
.
I
I
49
. IO»
. 10
lO
50
. 8*
. . 8
8
Total
- 3 4 4
J 2
241.
! 9 ■
19
Av. length
• 31 35 48 5
> 51 •
32 33 38 .4
i S3 .
Total
Av. length
12 ... .
32 ... .
• 40
.48.
• ■ 40
• • 49
52
Date
April 7, 1927
Locality
Bransfield Strait
St. No
Net
207 E
N 70 H o-s m.
Position
(b2° S4''00 S,
(. s9°5o'-3oW
Date
April 7, 1927
Locality
Bransfield Strait
Surface T.
( From — 0-78°
St. No.
207 E
Position
(62" 54'-oo S,
(. to -0-86'' C.
Net
Surface T.
/From —0-78°
(, to -o-86° C.
Length
in mm.
Stages
Total
in
sample
6 7
A B C D
E F G
Length
in mm.
Stages
Total
in
sample
42
44
2
2
: : :
I
2
1234
! 6 7
A B C D 1
1 F G
.
40
. . I .
I
I
46
.
I
I
52
I
I
I
47
I'
I
Total
I I
I
I
2
Total
. . 2 . .
■ 4
. . . 2
2 . 2
6
Av. length
. . 40 52
. . . 40
52 .
Av. length
■ ■ 44 ■ •
• 45
• • • 44
44 • 46 1
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
163
MALES
FEMALES
Date
April 7, 1927
Locality
Bransfield Strait
Date
April 7, 1927 Locality
Bransfield Strait
St. No.
Net
207 F
^ 70 H 0-5 m.
Position
Surface T.
("62° 54'-oo S,
\ 59°5o'-3oW
/From -0-78°
\ to -086° C.
St. No.
Net
207 F
^i 70 H 0-5
Position
Surface T.
(62° S4'oo S,
I 59= 50'-30 W
( From —0-78'
I to -o-Sd" C.
Stages
Total
Suges
Total
Length
in mm.
in
sample
Length
in mm.
m
sample
I 2 3 4 £
t
7
A B C D E F G
1234
5 6 7
ABODE
F G
29
•
29
1 . . .
. . I . .
1
30
.
I
30
4 • ■ •
. . 4 . .
4
31
.
I
31
2 . . .
2 .
32
32
3 ■ ■ •
2 1
3
35
2
• 3
" i
2 . . .
}(•
I
34
2 . . .
1 I
39
1
35
4 • •
3 I
4
40
I
3'
2
42
I
38
48
2
2
39
SI
I
40
52
I 2
3
42
53
2 .
2
43
I*
54
I 3 ■
4
44
. . 2*
. 2
I
I
45
• • ^l
. 5
5
56
57
I
I
I
I
46
47
■ ■ 4*
. . 6*
■ 4
. 6
S
6
58
I
I
48
■ • '!
. 7
7
49
50
51
: : f.
. . I*
• 4
• 7
1
4
7
1
Total
Av. length
I 4 5 5 I
32 31 37 49 5
i 5
I
253.
31 33 38 .4
4 13 .
9 54 •
27
Total
23 • I
• • 37
. II II 2
• 37
62
Av. length
34 . 46 42 . . 47
. 34 33 44 39 ■ 47
Date
April 7, 1927
Locality
Bransfield Strait
Date
April 7, 1927 Locality
Bransfield Strait
(62° 54'-oo S,
^ 59 So'-3oW
St. No.
Net
207 G
N 70 H 0-5 m
Position
(■62° 54'oo S,
\ 59°So'-3oW
St. No.
Net
207 G
N 70 H 0-5
Position
m.
Surface T.
(From —0-78^
1 to -086° C.
Surface T
1, to -0-86 ' C.
Stages
Total
Stages
Total
Length
in mm.
in
sample
Length
in mm.
in
Sample
1234
5
6 7
A B C D
E F G
I 2 3
t 5 6 7
A B C D E F G
27
X
I . . .
•
40
42
: : ;
I
I
I
I
Total
I
I . . .
I
43
45
. . I
I
I
I
1
Av. length
. 27 . .
27 . . .
Total
. . 4
. . . 3
I
4
Av. length
. • 43
. . • 43 42 . .
Date
April 7, 1927 Ixjcality
Bransfield Strait
St. No.
207 H
/62°54-opS
Net
N 70 H 0-.
m.
Surface T.
1 59° 50 30 W
/From —0-78"
t. to -086° C.
Date
April 7, 1927
Locality
Bransfield Strait
Stages
Total
St. No.
207 H
Position
Surface T.
f62° 54'oo S,
Length
in
Net
N 70 H 0-5 m
I 59" 50 30 W
/From —0-78
\ to -086° C.
in mm.
I 2 3
4567
A B C D
E F G
sample
30
31
I
Stages
Total
I
Length
in mm.
in
sample
33
34
1
1234
5
6 7
A B C D
E F G
35
I
. I . .
I
43
I . . .
47
. . I .
I
I
44
3 . . .
• 3
48
I
45
50
1
I
3
46
47
^ : : 3
. 2
2 . 2
. . 3
52
3
I
48
49
. . ■ 4
. . . 2
• • 4
. . 2
2
55
57
1
. . . I
I
I
50
51
. • • 4
. . . 2
. . 4
. 2
4
2
Total
. I I 6
5
. I I
I 10
13
Total
7 • •
9 . . 19
.526
3 • 19
35
Av. length
. 35 47 52 5
2
• 35 • 47 1
8 52 .
Av. length
32 . . 4
5 . . 48
. 33 32 45 4
6 . 48
164
DISCOVERY REPORTS
MALES
FEMALES
Date April 7, 1927 Locality Bransfield Strait
Surface T. {•^r-'o^sl? C.
Date April 7, 1927 Locality Bransfield Strait
Net''" l?'7?Ho-sm. ^-tion \;'^l^^^^^
Surface T. {•'r--o.86'?C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
25
28
29
30
31
35
11
49
SO
52
S3
I . .
I
I
I
I
2
2
2
I
2
I
I
I
29
32
33
34
35
36
38
39
42
43
45
46
47
48
49
. I
. 2
X
2
I
I
1
3
I
-I
2
I
3
3
3
3
2
2
I I
2
■ '
2
. . . I . .
2 .
I
I
. . . I
2 . . . .
I . . . .
3*
3*
::;::: ^:
2*
. . 11..
I . . .
3
3
3
3
2
TotaJ
Av. length
26233. •
27 32 35 49 52 . .
46. .42.
29 33 . . 50 51 .
16
Total
Av. length
10 I 3 ... 14
34 39 42 . . . 47
. 4622 .14
■ 32 36 43 40 . 47
28
Date April 7, 1927 Locality Bransfield Strait
Net''"- ^°^7oHo-5m P°-- r^S9^5o°3oW
Surface T. {^^^^S^lT^.
Date April 7, 1927 Locality Bransfield Strait
Net"""- N'7oHo-5m. P°-ion {^^.^^^^^
Surface T. {^'"^^S^lfc,
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
18
24
I
I
I
I
I
I
31
32
I
I
I
I
I
Total
Av. length
2
21
2
21
2
Total
Av. length
2
• 32
2
32
2
Date April 7, 1927 Locality Bransfield Strait
^'et'^" N'7o''Ho-5m. ^osi^io. {"^i?'^^^
Surface T. {^Z"'--^l'c.
Date April 7, 1927 Locality Bransfield Strait
lie^- N'7o^Ho-sm. Pos.t.on {'^H^^^^
Surface T. {^r--o°86'='c.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
32
34
35
38
47
50
51
. 2
. z
. 2
2
I
2
28
30
31
33
35
40
43
44
46
47
48
2
2
2
. 2
2
2
4
2
I
I
I
2
4
5
I
2
...... 1*
I 1*
::.::: 4*
5*
I*
I . . .
4
5
'
Total
Av. length
.4122. .
. 33 35 43 SI ■ ■
. 5 . I I 2 .
.33 -38 47 51 .
9
Total
Av. length
II . 2 , . .12
31 .42 . . .46
. 562. .12
. 29 32 42 . . 46
25
Date April 7, 1927 Locality Bransfield Strait
^'et^°- ??^7?Ho-5m. P-^'io" {'^^'^'^SO^'o W
Surface T. {^-^"sI^C.
Date April 7, 1927 Locality Bransfield Strait
lfet''°- N^7?Ho-5m. P-"- {'ni-^'so^a'c' W
Surface T. {^'^?^S^lfc,
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
1234567
A B C D E F G
26
30
32
34
35
2 i ■ 2
2
I
2
I
so
53
I
2 .
I
I
2
I
2
I
I
I
I
Total
Av. length
. I ... 3 .
. I ... 3 .
.28 . .11
4
Total
Av. length
. 61....
7
^' .-.-....
•^
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
i6s
MALES
Date April 14, igzS
St. No. WS 196
Net N 100 B 103-0 m.
Locality
Position
Surface T.
S. Georgia
fS4°36'S,
\ 38'38'W
2-73' C.
FEMALES
Date April 14, 1928
St. No. WS 196
Net N 100 B 103-0 m.
Locality
Position
Surface T.
S. Georgia
/S4°36' S,
1 38=3S-W
2-73' C.
Length
in nun.
Stages
34567 ABCDEFG
Total
in
sample
Length
in mm.
Stages
234567 ABCDEFG
Total
in
sample
24
26
27
28
29
30
31
32
33
34
36
40
46
48
Total
Av. length
4 24 4 2
26 30 34 43
48
23
25
26
27
28
29
30
31
34
Total
Av. length
6 22
26 29
8 17 3
28 29 32
2S
48
Date April 28, 1929
St. No. WS 427
Net N 100 B 140-0 m.
Locality
Position
S. Georgia
/sa" 34' s,
( 40" 10' W
Surface T. 1-94° C.
Length
Stages
Total
in
sample
in mm.
1234567
ABCDEFG
48
50
54
I
I
z
. . . . I . .
I
I
I
I
Total
Av. length
I
... 48
. 2
. 52
I . 2
. . . . 48 . 52
3
Date April s. 1931
St. No. 663
Net TYF V 250-0 r
Locality
Position
S. Georgia
/53° 34i' S,
(_ 30°2srW
Surface T. 051° C.
Date April 5, 1931
St. No. 663
Net TYF V 250-0 m.
S. Georgia
.fS3° 34i' S,
l 30° 255' W
Surface T. 051' C.
Locality
Position
Length
in mm.
25
27
29
31
33
34
35
36
37
41
44
45
49
50
55
Stages
1234567 ABCDEFG
Total
in
sample
Length
in mm.
23
27
28
29
30
31
32
33
35
36
39
41
44
48
51
Stages
234567 ABCDEFG
Total
in
sample
Total
Av. length
8 10 I 3 2 I I 5 13
30 35 45 45 SO 55 55 30 31 48 43 49 53 55
26
Total
Av. length 29
19 2 I
31 42 41
Date April 17, 193 1
St. No. 665
Net TYF B 250-0 m.
Locality
Position
Surface T.
S. Georgia
/5i°4ii' S,
I 29° 581' W
250° C.
Date April 17, 193 1
St. No. 665
Net TYF B 250-0 m.
Locality
Position
Surface T.
S. Georgia
f5i°4iJ'S,
t 29°58i'W
Length
in mm.
26
27
28
29
30
3t
32
33
34
35
37
38
49
Total
Av. length
Stages
234567 ABCDEFG
3 29 I
25 31 38 .49
3 29 1
25 31 35
Total
in
sample
Length
in mm.
25
26
27
28
30
3t
32
33
34
35
36
37
38
40
55
Stages
1234567 ABCDEFG
Total 6 27
Av. length 26 33
5 26 2
127 32 39
Total
in
sample
1 66
DISCOVERY REPORTS
MALES
Date April 17-18, 1931
St. No. 666
Net TYF B 320-0 m.
Locality
Position
Surface T.
S. Georgia
149° SH' S,
1, 29° 52J' W
271= c.
Length
in mm.
Stages
Total
in
sample
1234567
A B C D E F G
28
30
41
I
1
I . . . .
I
I
I . . . .
I
I
I
Total
Av. length
. 21....
. 29 41 . . . .
. 21....
. 29 41 • ■ - ■
3
_
Date May 18, 1929
St. No. WS 434
Net N 100 B 91-0 m.
Locality S. Georgia
Position /53° 10' S,
Surface T,
/5.
\ 34°o8'W
3-26° C.(?)
Length
in mm.
Stages
1234567
A B C D E F G
Total
in
sample
FEMALES
Date May 18, 1929
St. No. WS 434
Net N 100 B 91-0 m.
Locality
Position
Surface T.
S. Georgia
(53° 10' S,
1 34°o8'W
326° C.(?)
Length
in mm.
567
A B C D E F G
Total
in
sample
30
31
33
34
35
35
39
43
46
28
29
31
34
37
39
40
Total
Av. length
4 3 3
32 37 41
424
32 36 40
Total
Av. length
9 I
33 40
5 4 I
31 37 40
Date May 6, 1934
St. No. 1359
Net N too H 5-0 m.
Locality
Position
Surface T.
South of
Cape Town
/63°4s'-2 S,
(, 36° 4i'-i E
-1-37° C.
Date May 6,
St. No. 1359
Net N 100
, 1934
H 5-0 m.
Locality
Position
Surface T.
South of
Cape Town
(-63° 45'-2 S,
( 36°4l'lE
-1-37° C.
Length
in mm.
Length
in mm.
26
27
28
29
30
31
32
33
36
37
39
40
41
Stages
34567 ABCDEFG
Total
in
sample
23
24
25
26
28
30
31
32
33
34
35
36
37
39
40
41
43
45
34567
ABCDEFG
Total
in
sample
Total
Av. length
9 17 4
28 32 40
4 24 2
29 32 41
Total
Av. length
22 25 5
28 31 41
7 38 5 2
27 35 37 44
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
167
MALES
FEMALES
Date June 2, 1936
St. No. 1780
Net N 100 B 114-0 m.
Locality
Position
S.E. of Bouvet I.
/56° lo'-i S,
I. 00' o8'-8 W
Date June 2, 1936
St. No. 1780
Net NlooBli4-or
Locality
Position
S.E. of Bouvet i.
(56° lo'-i S,
\ 00° o8'-8 W
Length
in mm.
Stages
Total
1234557 ABCDEFG| sample
Length
in mm.
Stages
I 234567 ABCDEFG
Total
in
sample
33
38
39
42
43
46
Total
Av. length
I 2 4
33 39 44
33 38 43
27
35
43
44
Total
Av. length
27 35 44
27 35 44
Date June 2, 1936
St. No. 1781
Net N 100 B 128-0 m.,
500-150 m.,
800-450 m.
Locality
Position
S.E. of Bouvet I
f57°4l'-8S.
( 00° l9'-8 W
Length
in mm.
30
31
34
35
36
38
40
41
42
43
45
49
Total
Av. length
Stages
34567
2 6 15 I
30 36 42 40
ABCDEFG sample
2 10 10 2
30 37 43 45
Date June 3, 1936
St. No. 1782
Net N 100 B 500-150 m.,
700-400 m.
Locality
Position
S.E. of Bouvet I.
758°44'-6S
\ 00° 01 -5 E
Length
in mm.
33
35
48
Total
Av. length
Stages
Total
33 35 44
Date Jime 2, 1936
St. No. 1781
Net N 100 B 128-0 m.,
500-150 m.,
800-450 m.
Locality
Position
S.E. of Bouvet I.
^57°4„-'-8S
(. 00° 19 -8 W
Length
in mm.
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
49
Stages
34567 ABCDEFG
Total
Av. ength
3 21 51
31 35 41
16 32 22 5
32 38 42 47
Total
in
sample
Date June 3, 1936
St. No. 1782
Net N too B 500-150 m.,
700-400 m.
Locality
Position
S.E of Bouvet I
/58° 44'-6 S,
\ 00^ ol'-5 E
4 5 6 7 ABCDEFG, sample
Length
in mm.
36
38
40
41
43
49
Total
I Av. length
Stages
34567
ABCDEFG
39 42 . • • • ] • . 36 40 46
Total
in
sample
Date June 6, 1936
St. No. 1786
Net TYF B 800-400 m.
Length
in mm
37
42
44
Locality
Position
Total
Av. length
S.E. of Bouvet I.
\ H°I4-7E
Stages I Total
-[ in
6 7 ABCDEFG! sample
Date
St. No.
Net
Length
in mm.
39
40
42
June 6, 1936
1787
TYF B 800-400 m.
Total
Av. length
Locality
Position
Surface T.
S.E. of Bouvet I
fsS" o5'-9 S,
I 12° 48'-6 E
-i-5o'-C.
Stages
1234567
ABCDEFG
Total
in
sample
1 68
DISCOVERY REPORTS
MALES
FEMALES
Date June 7, 1936
St. No. 1788
Net N 100 B 270-0 m.
Locality
Position
Surface T.
S.E. of Bouvet 1.
/S9° ir-7 s,
1. 17° 01 -9 E
-i-8o°C.
Date June 7, 1936 Locality
St. No. 1788 n . .
Net N 100 B 270-0 m. P'"""""
Surface T.
S.E. of Bouvet I.
fS9°ll'-7S
\ 17° 01 -9 E
-l-8o°C.
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
123456
7
A B C D
E F G
1234567
A B C D
E F G
40
. . . 1 . . .
. . . I . . .
I
38
2
Total
Av. length
. . . I . . .
. . . 40 . . .
. . . I . . .
. . . 40 . . .
I
Total
Av. length
. . . 2
38
. . 38 38
•
1
Date June lo, 1936 Locality
St. No. 1794
Net N 100 B 360-200 m. Position
Cape Town to
Ice Edge
/52° 37'-5 S,
I 18' 22'-7 E
Length
in mm.
Stages
Total
in
sample
1234567
A B C D
E F G
47
I . . . .
. . . . I . .
I
Total
Av. length
. . 47 ■ • • ■
I
.... 47 ■ ■
I
Date July 12, 1938
St. No. 2362
M., /N 100 H o-s m.,
Net 1 TYFB4S0-430
Locality
^ Position
Cape Town to
Ice Edge
f54° 59'-3 S,
\ 01 10 -2 E
Date July 12, 1938 Locality
St. No. 2362
Net r-i-°Y°F^r5i-«om. P°--
Cape Town to
Ice Edge
fS4' 59''3 S,
(, oi 10 -2 E
Length
in mm.
Stages
Total
in
sample
Length
in mm.
Stages
Total
in
sample
123456
7
A B C D
E F G
1234567IABCD
E F G
30
35
54
I . . .
I
I
I , . . .
. . I . . . .
I
1
I
I
26
29
30
31
35
40
11. . . . . 2 .
I
2
I
I
I
Total
Av. length
II..
. . 30 35 • •
I
54
2 .
I
■ • 54
3
I . . . .
. . I . . . .
. I . .
Total
Av. length
• 3 4 ■ ■ . ■
. 28 34 . . . .
151.
26 31 40 .
. . . ! 7
... 1
Date July 13, 1938 Locality
St. No. 2365
Net TYFB35C^20om. p^^^^.^^
Cape Town to
Ice Edge
(53" 23'4S,
I 04" 5o'-5 E
Date July 16, 1938
St. No. 2372
Net N 100 B 102-0 m.
Locality'
Position
Cape Town to
Ice Edge
fSi" io'-6 S,
(. 15° 47'-5 E
o,,„„
Total
in
sample
Length
in mm.
Stages
Total
in
sample
Length
in mm.
^ ^ . -A™lAD/-•^^
E F G
123456
7
A B C D
E F G
^0 1 ■ '
I
2
I
48
. . . . I . .
I
I
I
I
41
43
• 2 2
. . I I
Total
Av. length
. 2 .
■ 47 • •
. . . . 2 . .
.... 47 • ■
2
Total
Av. length
■ • 4 4 . . .
. . 41 .... J ... 41 .. .
4
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
169
Table 20. Egg Measurements from 839 Females
Date August 28, 1928 |
St. No. • WS264 j
Depth N 70 B 97-0 m.
Date September 5, 1928
St. No. WS 277
Depth N 70 B 124-0 m.
Length
in mm.
Egg classes
Total
Length
in mm.
Egg classes
1
)
I
2
3
4
I
i 3
4
Total
005-012
0-13-0-24
0-25-0-48
0-49-0-70
0-05-0-12
0-13-0-24
0-2S-0-48
0-49- 0-70
33
34
39
40
41
42
43
4S
55
005
o-os (2)
0-07
0-05
0-09
O-O7-0-09
0-07
0-09 (2)
0-09
0-09-0-I4
0-09-
014 (2)
0-18-
0-
023
18
2
3
31
33
35
36
40
41
43
44
45
■*5
48
49
oos
0-05 (2)
0-05
0-05
0-05-0-07
0-05-0-07
0'05-oo7
005-007
0-05-0-07
0-07-0- 09
0-05-0-09
009-0-14
0-09-0- 1 1
009-0-18
J
2
2
0-18-0-25
0-23
Total in sample
Average diameter of eggs (mm.)
15
007
Total in sample
Average diameter of eggs (mm.)
19
0-12
Date September 17, 1928
St. No. WS 282
Depth N 70 B 137-0 m.
Date August 18, 1938
St. No. 2396
Depth N too B 109-0 m.
33
35
36
11
39
42
tl
51
55
0-05
0-05
009
o-os
0-05
0-05
009 (2)
0-05
0-07
0-09
O-II
o- 09-0- 14
012 (3)
0-09
t
2
I
3
I
40 009
I
Total in sample
Average diameter of eggs (mm.)
I
0-09
Date August 19, 1938
Sj. No. 2399
Depth N 100 H 0-5 m.
NiooB-l'**-""'
1,300-150 m.
Total in sample
Average diameter of eggs (mm.)
17
33
35
36
38
40
42
0-05
005
O'og
009
0-07
009-0- 14
Date September 24, 1938
St. No. 2430
Depth N 100 H 0-5 m.
N 100 B 1 17-0 m.
Total in sample
7
Average diameter of eggs (mm.) • o-o8
46
0-09-0-14
0-09-0-14
I
I
Total in sample
Average diameter of eggs (mm.)
2
012
Date August 22, 1938
St. No. 2408
Depth N 100 H 5-0 m.
Date October 2, 1928
St. No. WS 290
Depth N 70 H 0-5 m.
39
40
41
42
0-05
oos
o-os (4)
0-07 (2)
1
4
2
37
0-05
.
-
1
I
Total in sample
Average diameter of eggs (mm.)
8
0-06
Total in sample
Average diameter of eggs (mm.)
I
0-05
Date August 24, 1938
St. No. 2412
Depth N 100 H 5-0 m.
N 100 B 107-0 m.
Date October 4, 1928
St. No. WS 29s
Depth N 100 B 97-0 m.
42
0-09
1
1
I
37
0-05
1
1
1
I
I
005
Total in sample
Average diameter of eggs (mm.)
I
0-09
Total in sample
Average diameter of eggs (mm.)
g-a
I70
DISCOVERY REPORTS
Length
33
34
35
36
38
39
40
4'
42
43
45
46
5°
51
52
56
34
35
35
36
38
39
40
42
43
44
45
46
48
49
50
51
35
38
44
45
Date
St. No.
Depth
October 5, 1928
WS 298
N 100 B 94-0 m.
Egg classes
005
007
0-09
0-07
007
O'og
o-og
009
0-25-0-48
0-14-
14
•14
-o 18
■18
■18
0-23
0-18-0-2:?
0-14-0-18
0-23
Total in sample
Average diameter of eggs (mm.)
Date
St. No.
Depth
October 6, 1928
WS 304
N 1 00 B II 0-0 m
30
0-05
34
0-05
40
0-05
44
007
Total in sample
Average diameter of eggs (mm.)
Date
St. No.
Depth
October 16-17, 1930
453
N 100 13 164-0 m.
0-05
0-05
Total in sample
Average diameter of eggs (mm.)
Date
St. No.
Depth
October 17, 1930
454
N 70 B 192-0 m.
0-05
0-09
o-os (3)
0-09
0-05
0-07
0-09 (2)
0-09 (2)
o-og (3)
009 (3)
O'll
o-ll
0-09
o-ll
o-og
0-09
0-09
o-ii
Total in sample
Average diameter of eggs (mm.)
Date
St. No.
Depth
October 18, 1930
455
N 100 B 1 16-0 m.
0-07
0-07
0-09
0-09
0-09
Total in sample
Average diameter of eggs (mm.)
Total
19
0-I3
4
0-06
27
0-09
5
0-08
Length
in mm.
38
39
40
JS3
34
35
36
37
38
39
40
46
48
50
35
37
40
49
Date
St. No.
Depth
October 19, 1930
459
N too B) ,0, „„
N70B / '83-0 m.
Egg classes
0-05
0-07
0-09
o-ll
0-07
0-09
0-09
0-07
0-25-0-48
Total in sample
Average diameter of eggs (mm.)
Date
St. No.
Depth
October 22, 1930
461 G
N 100 B 700-560 (315) m.
0-05
0-07
0-07 (2)
0-05
0-09
o-il
00s
0-07 (2)
0-09
0-05
0-07
0-09
o-il
0-07 (3
009 (3
o-ii {2.
005 (3
007 (2,
O-II (2'
o-os (2;
0-07 (4;
0-09 (6'
0-05 (2!
0-07 (5:
0-09 (5,
o-ii
005 (3
0-09 (4
O-II
0-05 (3
007 (3
0-09 (4;
O-II
0-05
007
0-09 (5:
00s (2
0-07 (3
0-09 (5
O-II (3
o-os (2
0-07
0-09
O-II
005
0'07
0-09 (2)
0-09
O-II
0-14 (2)
o- 1 1-0-23
14(2)
14(2)
Total in sample
Average diameter of eggs (mm.)
Date
St. No.
Depth
October 22, 1930
461 D
N 100 B 490-385 m.
Total
0-07
S
14
121
0-09
0-07
o-ii
O-II
O-II
Total in sample
Average diameter of eggs (mm.)
4
O'lO
Date
St. No.
Depth
October 23, 1930
462
N 100 Bl
, B / 9°"° ■"■
N70I
Total in sample
Average diameter of eggs (mm.)
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
171
Date October 25, 1930
St. No. 463
Depth N 100 B 132-0 m.
Date November 29, 1930
St. No. 523
Depth N 100 B 157-0 m.
Lengtli
in mm.
Egg classes
Total
Length
in mm.
Egg classes
Total
I
2
3
4
I
2
3
4
0-0S-0-I2
OI3-0-24
0-25-0-48
0-49-0-70
0-05-0-12
0-13-0-24
0-25-048
0-49-0-70
44
0-O9
I .
41
44
ti
so
0-14-0-32
0-18
0-18-0-27
023
0-18-0-23
0-14-0-45
I
2
I
I
I
Total in sample
Average diameter of eggs (mm.)
I
009
Total in sample
Average diameter of eggs (mm.)
6
0-23
Date October 25, 1930
St. No. 464
Depth N 100 H 67 (-0) m.
Date November 6, 1932
St. No. 1009
Depth NiooBiio-om.
48
O-II
I
Total in sample
Average diameter of eggs (mm.)
I
o-ii
27 0-07
0-14 (2)
0-14
0-14-0-18
0-14
018
■
I
2
2
2
2
Z
I
5
z
2
I
I
29
30
31
32
34
37
38
39
40
43
007 (2)
007
0-09
0-07
0-09
0-09 (2)
O-II
0-09 (2)
o-li
Date October 26, 1930
St. No. 465
Depth N 100 B 1 13-0 m.
39
40
0-05 (2)
O-07
2
I
Total in sample
Average diameter of eggs (mm.)
3
0-06
0-14
Total in sample
Average diameter of eggs (mm.)
23
oil
Date November rg, 1929
St. No. WS Alongside Deception I.
Depth N 100 B 0-5 m.
It
48
o-l8
014(3)
0-27 1
3
I
Date December 19, 1926
St. No. 125
Depth N 100 H 70 m.
Total in sample
Average diameter of eggs (mm.)
5
0-I7
43
44
45
46
47
48
49
50
SI
52
54
S6
11
60
61
0-
09
0-18
1
1
Eggs recorded at WS 477-WS 487 at depths between 1000-500 m.
2nd nauplii recorded at WS 480-WS 486 at depths between
750-200 m.
0-23
0-23
o-i8
0-23
0-18 (2)
0-23 (2)
0-14
0-18 (2)
0-23
0-18
0-23
0-23 (2)
0-18 (2)
0-23 (4)
0-18
0-23 (3)
032
0-27
0-27 (4)
032 (3)
0-
0-36-
49
-0-64
I
2
4
S
2
3
6
13
3
2
I
I
Date November 13, 1930
St. No. 480
Depth N 100 B 161-0 m.
N 70 V 1000-750 m.
46
041
I
Total in sample
Average diameter of eggs (mm.)
I
0-41
o-i8
0-41
0-4S
0-27
Date November 19, 1930
St. No. 494
Depth N too B 160-0 m.
0-27
36
37
38
39
40
41
42
43
44
45
46
47
48
49
SI
014
o-\t
014 (2)
0-14
01 4 (4)
o-i8 (3)
0-14 (2)
o-i8
014(6)
014 (3)
018(5)
014 (6)
018
0-14 (2)
01 8 (6)
014
018(4)
014(2)
018
0'14
01 4 (4)
018
014
I
2
2
2
7
3
6
8
7
8
5
3
1
5
I
Total in sample
Average diameter of eggs (mm.)
SO
0-24
Date December 11, 1930
St. No. 527
Depth N 450 H i22(-o) m.
42
0-23-0-55 1
I
Total in sample
Average diameter in eggs (mm.)
I
0-39
Date December 17, 1930
St. No. 534
Depth N 100 B 72-0 m.
41
0-23
1
I
Total in sample
Average diameter of eggs (nun.)
61
0-15
Total in sample
.Average diameter of eggs (mm.)
I
0-23
172
DISCOVERY REPORTS
Date December 18, 1930
Eggs and 2nd nauplius recorded at St. 540 at (500-100) m.
St No. 535
Depth N 70 B 0 m.
Date
St. No.
December 19-20, 1930
541
Egg classes
Depth
N 100 B 108-0 m.
Length
in mm.
2
3
4
Total
1
Length
in mm.
Egg classes
Total
0-05-
O.I2
0- 13-0-24
0-25 0-48
0-49-
0-70
I
2
3
4
37
023 0.23-0-45 1
2
0-05-0-12
0-
13-0-24
0-25-0.48
0-49-0.70
44
47
0-32
0-27 1
•
I
32
44
0-23
027
I
I
Total in sample
4
Average diameter of eggs (mm.) 0-29
Total in sample
2
Average diameter of eggs (mm,)
025
Date
December 20, 1930
St, No. 536
St. No.
W^ r
Depth N 100 B 122-0 m.
Depth
N 100 B 1(14-0 m.
36
0-23
I
46
0.27
I
Total in sample
I
Total in sample
,
Average diameter of eggs (mm.) 0-23
Average diameter of eggs (mm,)
0-27
Eggs and 2nd nauplius recorded at .St, 546 at (500-100) m.
Date
December 21, 1930
Date December 19, 1930
St. No.
548
St. No. 538
Depth
N 100 B 102-0 m.
Depth N 100 B 137-0 m.
39
40
41
0-14
0-14
018
0-23 (2)
0-14
48
023
1
.
4
2
Total in sample
I
Average diameter of eggs (rr
m.) 0-23
43
o-i8
0-14
0-18 (3)
5
0.,
45
0-14 (3)
0-
27
6
Eggs and ist r
auplius recorded from .St. 537 at (1000-
500) m.
0-18 (2)
Eggs recorded
at St. 538 at (750-500) m.
46
47
48
0-14
0-18 (2)
0-23
0-14
0-18
I
3
5
Date December 19, 1930
St. No. 539
0-23 (3)
Depth NiooBI ,,, „„
N 70 B 1 '37-° m.
49
0-i8
023
2
SO
0-18
0-23 (2)
■"
3
30
023
31
32
023 (2)
0-14-
0-36 (2)
2
3
52
53
55
0-1 8
0-23
o-
27
I
35
0-23
0-27-0-41
2
Total in sample
36
36
0-23-0-50
I
Average diameter of eggs (mm.)
0-19
37
o-ls-0-45
38 o-OQ-
.011
0-23-
0-27 (2)
5
0-41-0-45
Date
December 21-22, 1930
39
0-23-0-41
0-23-0-45
2
St. No,
Depth
549
N 100 B 115-0 m.
0-27-0-55
41
0 23 (2)
2
33
0-09 (2)
1
2
42
0- 1 8-03 2
0-23-0-55
0-27-0-41
0-27-0-45
4
^1
0-09
0-09
0.14
0-I4 (3)
014 (2)
I
2
3
2
43
0-32-0-45
I
39
0.09
014
3
44
0-23
0-18
45
0-32-0-45
I
40
014
3
46
0-23-0-32
I
0.18 (2)
47
018-0-27
I
41
0-18
2
48
0-27
I
0-23
SO
0-27-0-45
I
42
0-14
I
SI
0-23-0-55
I
43
0-09
I
52
0-18-0-32
0-36
2
44
0-09
0-I4
0-18
3
55
0-36-
0-
64 (2)
45
0-14(2)
0-18
3
46
0-23
I
Total in sample
35
'^l
o-iS
I
Average diameter of eggs (r
nm ) 0.32
48
52
0-14
023 (2)
2
Total in sample
31
Eggs and 1st
nauplius recorded at St. 539 at (looo-ioc
m.)
Average diameter of eggs (mm.)
OIS
Date December 19, 1930
St. No. S40
Date
December 29, 1930
Depth N 100 B 155-0 m.
St, No,
558
N 70 V 500-250 m.
Depth
N 100 B 146-0 m.
32
0-23
I
40
0-18
I
34
027
I
0-23 (2)
2
Total in sample
2
Total in sample
3
Average diameter of eggs (
-nm.) 0-25
Average diameter of eggs (mm.)
0-21
1
1
Eggs and 2nd nauplius
recorded at St. 558 at (200-50) m.
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
173
Date
St. No.
Depth
December 30, 1930
559
N 100 B 113-0 m.
Egg classes
Length
in nun.
33
43
45
46
47
37
42
45
47
SO
47
50
45
47
48
49
50
51
53
54
56
57
S8
59
013-024
o-l8
0-09-0-I8
018
0-09-0-I8
018
0-25-0-48
0-49-0-70
Total in sample
Average diameter of eggs (mm.) | 016
Total
Date
St. No.
Depth
December 30, 1930
560
N 100 B 155-0 m.
0-09
0-09
o-i8
023
028
Total in sample
Average diameter of eggs (mm.)
Date
St. No.
Depth
December 31, 1930
561
N 100 B 137-0 m.
Date
St. No.
Depth
December 31, 1930
56a
N 100 B 113-0 m.
032
Total in sample
Average diameter of eggs (mm.)
Date
St. No
Depth
January 23, 1929
WS 373 EE
N 100 B 70-0 m.
045
0-23-0-45
0-4I-0-45
027
' 0-32
0-4S
0-36-0-4S
0-32-0-45
0-4I-0-S0
0-27-0-45
036
0-32-0-50
0-27-0-32
0-4S-0-50
03 5-0-41
0-27-
045 (2)
0-27-0-32
0-45 (4)
01 8-045
0-45 (2)
0-4I-0-50
0-35-0-50
0-36-0-59
032-
0-45 (2)
0-4S (2)
0-27-0-45
036-
0-4S (2)
0-36-0-59
0-32-o-so
032
0-36-O-45
0-23-0-45
o-27-0'36
0-35-0'SO
0-32-0-45
Date
St. No.
Depth
January 21-22, 1930
N 100 B loo-o m.
Length
in mm.
Egg classes
50
51
5
0-2O
Total in sample 2
Average diameter of eggs (mm.) 0-28
50(2)
Total in sample
Average diameter of eggs (mm.)
46
0-40
31
33
35
36
37
38
39
40
41
42
44
45
46
47
48
45
48
38
48
0'2S-o-48
032
032
0'49-0'70
Total in sample
Average diameter of eggs (mm.)
Total
Date
St. No.
Depth
January 24-25, 1930
312
N 100 B 150-0 m.
Total in sample
Average diameter of eggs (mm.)
Date
St. No.
Depth
January 8, 193 1
N 100 B 132-0 m.
Total in sample
Average diameter of eggs (mm.)
Date
St. No.
Depth
January 8, 1931
575 „
N loo B 97-0 m.
018
018
0-23 (2)
0-23
o-i8 (2)
o-i8-0'23
o- 14-0-27
0-23
0-27
0-14-
0-23 (3)
0-23 (3)
023
0-14-0-27
0-09-0-23
0-14-
0-23 (2)
0-23 (2)
0-23 (2)
0-23
0-14-0-27
0-32
0-23-0-27
0-27 (2)
0-27
0-35-0-45
027
0-27
0-27-
0-45 (2)
0-36-0-45
0-23-0-32
0-32
Total in sample
Average diameter of eggs (mm.)
41
0-2S
Date
St. No.
Depth
January 10, 1931
580
N 100 B 128-0 m.
o-l8-0'36
0-18-0-45
Total in sample
Average diameter of eggs (mm.)
Date
St. No.
Depth
January 12, 1931
583
N 100 B 139-0 m.
0'I4
0-09-0-18
Total in sample
Average diameter of eggs (mm.)
174
DISCOVERY REPORTS
Date January 13, 1931
St. No. 584
Depth N 100 B 165-0 m.
Date January 25-26, 193 1
St. No. WS 537
Depth N 100 B 67-0 m.
Length
in mm.
Egg classes
Total
Length
in mm.
Egg classes
Total
I
2
3
4
I
2
3
4
005-0-I2
o- 13-0-24
0-25-0-48
0-49-0-70
0-05-0-12
0-13-0-24
0-25-048
0-49-0-70
48
.
0-23-048
I
47
SI
023
0-23-0-25
1
1
Total in sample
Average diameter of eggs (mm.)
I
0-34
Total in sample
Average diameter of eggs (mm.)
2
0-25
Eggs recorded at St. 585 at Cso~o) m.
Date January 14, 1931
St. No. 590
Depth N 100 B Qo-o m.
Date February 22, 1928
St. No. WS 152
Depth N 100 B iio-om.
48
50
,
0-18-0-45
0-18-0-45
I
2
0-32
42
44
45
46
47
49
50
51
52
53
54
55
56
59
60
0-18
0-18-0-23
0-18
0-18-0-23
0 18
023
0-23
0-27-0-55
0-32-0-45
0-23-0-55
0-27-0-45
0-23-0-36
0-23-0-45
0-23-0-45
0-36-0-55
0 32-
0-45 (2)
0-27-0-SS
0-18-0-45
0-45
0-36-0-45
0-32-0-45
0-41-0-55
0-32
0-45
0-36-0-55
0-23-0-45
0-23-0-55
0-36-0-55
0-23-0-55
I
I
4
4
2
2
3
4
4
2
2
3
2
I
I
Total in sample
Average diameter of eggs (mm.)
Date January i6, 1931
St. No. 5g5
Depth N 100 B 170-0 m.
0-45-0-64
0-4S-0-55
0-45-0-68
0-36-0-68
0-4S-0-55
43
45
46
•
0-23
009-0-18
0-23-0-45
I
I
I
Total in sample
Average diameter of eggs (mm.)
3
0-24
1
Total in sample
Average diameter of eggs (mm.)
36
0-38
Date January 17, 1931
St. No. 599
Depth N 100 B 142-0 m.
Eggs recorded at St. WS 147 at (250-100) m.
Date February 8, 1929
St. No. WS 376
Depth N 70 V 750-500 m.
49
•
032
0-18-0-45
0-23-0-45
1
I
I
50
53
•
0-45
0-55
I
I
Total in sample
Average diameter of eggs (mm.)
3
0-33
Total in sample
Average diameter of eggs (mm.)
2
0-50
Eggs recorded at St. WS 377 at (100-50) m.
Eggs recorded at St. WS 3S0 (250-100) m., WS 382 (400-250) m..
WS 385 (500-250) m., and WS 394 (200-100) m.
Date Januar\- 19. 1931
St. No. 602
Depth N 100 B1 ,,,
N70B ) '"°-°'"-
Date February 8, 1930
St. No. 349
Depth N 100 B 60-0 m
N 70 V 50-0 m.
42
43
44
45
46
47
48
50
52
53
54
55
57
58
0-09
0-09
0-23 (2)
0-23 (2)
0-23 (2)
0-23
0-23-
0-36 (2)
0-32-0-55
0-23-0-41
0-36
0-27-0-45
0-36-0-45
0-27-0-50
0-36-0-45
0-27
0-32
0-36
0-23-0-50
0-23-0-45
0 23-0-50
0-09-0-45
0-23-0-68
0-Z3-0-4S
2
I
2
5
I
I
I
4
2
4
2
I
I
37
40
0-18
018
I
I
39
Total in sample
Average diameter of eggs (mm.)
2
018
Eggs recorded at St. WS 505 at (250-100) m. on Feb. 4. 1930.
Date February 9, 1930
St. No. 351
Depth N 100 B 48-0 m.
N 70 V 100-50 m.
42
45
48
49
55
0-41
0-45
0-41
0-4S (3)
0-45 (2)
I
I
I
3
2
Total in sample
Average diameter of eggs (mm.)
28
0-31
Total in sample
Average diameter of eggs (mm.)
8
0-44
■
Date February 9, 1930
St. No. 354
Depth N 100 B 96-0 m.
Date January 20, 1931
St. No. 603
Depth N 100 B 140-0 m.
42
ti
53
54
59
0-32 (2)
0-4S
032
0-32
0-45
0-4S
0-50
0-50
2
I
I
2
2
I
45
53
56
\
0-18-0-32
0-18-0-50
0-18-0-64
•
I
I
1
Total in sample
Average diameter of eggs (mm.)
3
0-33
Total in sample
Average diameter of eggs (mm )
9
0-44
THE DEVELOPMENT AND LIFE-HISTORY OF KRILL
175
Date February lo, 1930
St. No. 356
Depth N 70 V 250-100 m.
Eggs recorded at St. 644 at (250-50) m.
Metanauplius recorded at Sts. 647, 648 at (1000-500) m. and
(250-0) m.
Length
in mm.
Egg classes
Total
Date April 7, 1927
St. No. 207 A
I
2
3
4
Depth N 70 H 0-5 m.
0-O5-O-I2
0-13-0-24
025-0-48
0-49-0-70
Length
in mm.
Egg classes
Total
42
49
50
52
53
S5
56
57
58
59
60
6t
63
64
0-41
0-64
0-55
0-64
0-55
0-59
o-SS (2)
0-59
0-64 (3)
o-SS
0-50
0-5S
0-64
0-55
0-59
0-63
0-64
0-S9
0-68
0-50
o-SS (3)
0-S9 (4)
0-68
0-55 (2)
0-59 (2)
0-68
0-64
I
I
I
2
2
6
I
3
6
3
10
4
1
I
2
3
4
0-4
5
O-0S-0-I2
o-l3-o*24
0-25-0-48
0-49-0-70
0-4
I
43
tl
47
48
0-4S
0-45
045
0-45 (2)
0-50
1
I
I
I
2
Total in sample
Average diameter of eggs (nmi.)
6
0-46
Date April 7, 1927
St. No. 207 B
Depth N 70 H o-s m.
35
11
40
o-os
oos (2)
0-05
0-05
I
2
X
I
Total in sample
Average diameter of eggs (mm.)
5
OOS
Total in samp
Average diamt
42
0-S9
ter of eggs (mm.)
Eggs recorded at a depth of (250-100) m.
Date February 8. 1931
St. No. 609
Depth N 100 B 128-0 m.
Date April 7, 1927
St. No. 207 C
Depth N 70 H o-s m.
49
51
55
0-27-0-45
0-18-0-45
0-23-0-45
I
I
I
40
4'
45
SO
0-05 (3)
o-os
005
0-45
3
I
1
I
3
0-34
Total in sample
Average diameter of eggs (mm.)
Total in sample
Average diameter of eggs (mm.)
6
0-I2
Eggs and metanauplius at St. 618 (Feb. 18-19, 193 1) at (750-100) m.
Metanauphus at St. 620 (Feb. 20, 1931) at (1000-500) m.
Date February 21-22, 193 1
St. No. 624
Depth N 100 B 137-0 m.
Date April 7, 1927
St. No. 207 E
Depth N 70 H 0-5 m.
48
0'l8-0-23
I
Total in sample
Average diameter of eggs (mm.)
I
0-21
46
•
050
0-59
I
I
Total in sample
Average diameter of eggs (mm.)
2
0-55
Date March 8, 1930
St. No. 368
Depth N 100 B 146-0 m.
47
48
0-05
0-07
o-os
0-07
o-os (3)
I
I
2
3
Date April 7, 1927
St. No. 207 F
Depth N 70 H o-s m.
Total in sample
Average diameter of eggs (mm.)
7
0-C5
42
0-05
.
1
1
I
Total in sample
Average diameter of eggs (mm.)
I
o-os
Eggs recorded at St. 637 (March 8, 1931) at (250-100) m.
Metanauplii recorded at Sts. 636, 637, 638, 639 (March 8-9. 1931)
at ( 1 000-0) m.
Date March 10, 1931
St. No. 643
S;o°B''}''3-om.
Date April 7, J 927
St. No. 207 H
Depth N 70 H 0-5 m.
40
41
42
43
44
45
46
47
48
0-05
o-os
o-os
o-os
0-05 (3)
0-05
005
0-05 (2)
0-05-0-23
3
I
2
3
c
0-2
'■45
3-0-45
43
44
45
46
0-05 1
oos (3)
0-05
0-05 (4)
I
3
I
4
Total in sample
Average diameter of eggs (mm.)
9
o-os
Total in sample
Average diameter of eggs (irmi.)
14
0-05
Metanauplii recorded at WS 197 (April 17, 1928) at (1000-750) m.
176
DISCOVERY REPORTS
Table 21. Totals of Males and Females taken in each month, including Eraser's Adolescents
Aug.
Sept.
Oct.
Nov.
Dec.
Jan.
Feb.
Mar.
Apr.
May
June
July
Males
Stage I
5
7
234
182
297
199
190
63
25
9
—
—
2
37
17
35
40
82
41
316
87
"5
21
4
—
3
24
3
23
3
23
27
204
120
125
7
10
I
4
37
4
17
24
27
18
54
41
61
3
24
I
5
48
17
14
13
4
14
12
19
42
—
I
2
6
12
39
80
5
7
3
21
14
II
—
—
—
7
—
5
67
34
169
130
143
5
4
—
—
I
Total
163
92
470
301
609
432
940
349
383
40
39
5
Stage A
18
7
238
199
252
208
60
58
30
4
—
—
B
33
18
31
24
137
42
478
81
135
28
4
—
C
30
3
30
7
10
8
102
78
60
4
13
2
D
29
3
8
19
17
18
66
85
58
4
20
—
E
43
6
12
8
5
12
64
26
39
—
2
2
F
10
49
49
7
14
14
41
7
5b
—
—
—
G
—
6
102
37
174
130
129
14
5
—
—
I
Total
163
92
470
301
609
432
94°
349
383
40
39
5
Females
Stage I
14
18
254
213
338
240
230
275
188
22
4
—
2
66
17
15
18
8
5
304
no
71
34
28
3
3
no
4
54
8
I
2
124
30
50
6
61
8
4
31
33
175
15
12
3
—
18
20
—
—
—
5
5
I
25
78
114
41
10
I
—
—
—
—
6
—
—
—
3
52
92
45
2
6
—
—
—
7
—
—
—
—
4
6
166
129
186
—
—
—
Total
226
73
523
335
529
389
879
565
521
62
93
II
Stage A
20
8
254
212
259
191
267
74
15
7
I
I
B
41
19
5
17
58
43
323
162
146
43
17
5
C
65
10
36
7
29
13
64
121
123
9
38
I
D
71
10
35
10
4
3
II
58
28
3
29
4
E
29
26
187
56
SI
4
6
20
25
—
8
—
F
—
—
4
13
37
27
17
3
I
—
—
—
G
—
—
2
20
91
108
191
127
183
—
—
—
Total
226
73
523
335
529
389
879
565
521
62
93
II
IDiscovery Report,. Vol. XXIII, pp. m-"^. ««'« '-^'^- 1"""y 'M'-l
THE ANTARCTIC CONVERGENCE AND THE
DISTRIBUTION OF SURFACE TEMPERATURES
IN ANTARCTIC WATERS
By
N. A. MACKINTOSH, D.Sc.
CONTENTS
Part I. The Antarctic Convergence page 179
Introduction j-q
Data j8o
The mean position of the convergence 181
The change of temperature at the surface 186
Part II. The distribution of surface temperature in Antarctic waters . . 194
Introduction .... ,„,
Treatment of the data j„.
Method of drawing the isotherms iqg
Notes on the distribution of temperature 201
References _„,
Appendix. Table 9 2or
Notes on the plates 2ii
Plates I-XIV following 212
THE ANTARCTIC CONVERGENCE AND THE
DISTRIBUTION OF SURFACE TEMPERATURES
IN ANTARCTIC WATERS
By N. A. Mackintosh, D.Sc.
(Plates I-XIV; Text-figs, i-ii)
PART I. THE ANTARCTIC CONVERGENCE
INTRODUCTION
r-p HE Antarctic convergence, which constitutes the northern hmit of the Antarctic surface water,
i was first observed by Meinardus (1923), and it is a boundary of far-reaching importance in the
Southern Ocean. It may be regarded as the fine at the surface along which the Antarctic surface water
sinks below the less dense sub-Antarctic water, and it is distinguished by a more or less sharp change
of temperature at the surface. The change of temperature is much less clearly defined in some longi-
tudes than in others, but it is probably correct to say that the convergence is continuous around the
Southern Ocean, even though there are regions where it can sometimes hardly be traced. It is essen-
tially a feature of the surface, but where it is ill-defined, or when surface temperature records are
insufficient, it can usually be assumed to lie in the latitude at which the coldest part of the Antarctic
surface layer sinks below 200 m. For further particulars reference should be made to Deacon (1933,
pp. 190-3, and 1937, pp. 20-4), and to Bohnecke (1938, p. 201, etc.).
The importance of the convergence does not lie only in the fact that it is the boundary between the
two principal water masses at the surface of the Southern Ocean. Its position is also related to the
distribution and movements of the deeper water masses, and there are reasons for believing that it
has a connexion with meteorological conditions in the Southern Ocean. Furthermore, it has a special
biological significance, for the Antarctic and sub-Antarctic zones, between which it forms the boundary,
have in certain respects, a distinct fauna and flora. That is to say, the convergence marks a limit
(though not always a very rigid one) to the range of certain species in the plankton (see Hart, 1934,
X937 1942; Mackintosh, 1934, i937; John, 1936), the fishes (Norman 1938) and the benthos
(Hastings 1943) The distribution of diatom ooze and diatomaceous mud as shown on Admiralty
charts, and by Neaverson (1934, plate xvii) indicates a relation also between the convergence and the
bottom deposits. , r 1 • *• *• ^
A number of oceanic species seem to be unaffected by the convergence, but further investigations
will probably increase the list of those whose distribution is influenced by it. It is probable that it
separates certain species which characterize different water masses or different systems of circulation,
but the simple effect of temperature also needs consideration, for the distribution of some species
whose upper or lower limit of tolerance approximates to temperatures at the convergence, may be
sharply bounded by the abrupt change of temperature. • j u .u w. ^.
Deacon (1937 p. 23) concludes that the latitude of the convergence is determmed by the latitude
reached by the Antarctic bottom water. He drew the line of the convergence on a circumpolar chart
fiQ.7 fi/4) and this was based on a number of positions fixed in the Falkland sector, and rather
more scattered positions in other sectors. Since then a very considerable number of additional positions
have been fixed. These show that Deacon's original line must lie very near the actual mean position,
i8o DISCOVERY REPORTS
but that the position in any given longitude is subject to rather more variation than the earlier data
suggested.
Before the mean monthly surface isotherms in Antarctic waters can be drawn it is necessary to
locate the mean position of the convergence as correctly as the available data permit. Furthermore,
records of both the mean position and the actual position at different times and places, and the change
of temperature at the surface, are likely to be required for various purposes in the future, and it
therefore seems worth while to publish particulars of every occasion on which the convergence has
been crossed by the Discovery Committee's ships, and to indicate the degree of accuracy with which
it could on each occasion be located, and the extent of the change of temperature.
DATA
The 'Discovery', 'Discovery II', and 'William Scoresby' have crossed or located the convergence
on 139 occasions. In nearly every instance at least some indication of its position was obtained, but
the degree of accuracy with which it was located varied considerably according to the observations
taken, to the extent to which the convergence was itself recognizable, and to the angle at which it was
crossed. The best indication of a crossing of the convergence is a sharp change of temperature shown
on the thermograph (see p. 195) an instrument which was in operation at all crossings by the ' Discovery '
and by the ' Discovery II ', except for a period in the fourth commission (1935-7) of the latter ship. The
'William Scoresby' did not carry a thermograph except on her last commission (1937-8). The other
important method of locating the convergence is by vertical stations taken on each side of it, whereby
the level of the minimum temperature of the Antarctic surface stratum can be determined. This is
a reliable method but is naturally of little value if the stations are very far apart. Sometimes it is
evident that the convergence lies between two stations at which only surface temperatures are read,
but there must be a clear difference, and of the correct range, if this method is to be trusted. Normally
the sea temperature is recorded every four hours as a routine by the ship's personnel. These records
are not altogether reliable as exact readings of the temperature, but in the absence of a thermograph
they may be of considerable value, especially when they show a sharp change of temperature at about
the expected position of the convergence, or when they show where the convergence lies when it is
known to have been crossed between two stations some distance apart.
The best determinations of the position of the convergence are obtained when it is crossed approxi-
mately at right angles, and the thermograph shows an abrupt change of temperature between two
vertical stations of which the more southerly shows a minimum temperature above 200 m. and the
more northerly a minimum below 200 m. Such complete indications are not ver)' frequent, though
they have commonly been obtained in the Scotia Sea. In many parts of the Southern Ocean it may
often be found that there is no clearly defined change of surface temperature, or the change may be
obscured by an oblique crossing of the convergence. Sometimes two successive stations may show
a minimum temperature at about 200 m., and it is then difficult to know where the true convergence
lies. A frequent source of difiiculty is the winding course of the convergence. At the surface the
junction of Antarctic and sub-Antarctic water seldom lies in a straight line, probably because it is
an unstable boundary. It forms twists and loops that may extend as much as 100 miles north or south,
and it possibly even forms isolated rings. The line is perhaps comparable to the edge of the pack-ice,
and may be even more tortuous. The latitude in which it is found may thus depend partly on whether
a ship happens to cross it where it bends to the north or to the south. Another consequence is that
a ship steaming in a straight line obliquely across the convergence may pass from Antarctic to sub-
Antarctic water and back again several times. It may also happen that before or after a crossing of
THE ANTARCTIC CONVERGENCE i8i
the convergence, the thermograph indicates that the ship's track passed near to a loop of the convergence
without actually cutting it. , i • i
It will now be seen that at many crossings of the convergence its position can only be approximately
determined Indeed the exact position in which it is assumed to be is sometimes little more than a
matter of opinion. However, even approximate positions are helpful in mapping out the mean position.
The latter must always be liable to modification (though in a diminishing degree) in the light of new
'^Table 9 (p 205) is a list in chronological order of all crossings of the convergence by the Discovery
Committee's ships. The position of the convergence is given as nearly as it can be estimated. This may
be taken as the middle of a sharp or comparatively gentle temperature gradient at the surface, as hal way
between two stations at which vertical observations were made, etc. Since some positions could be
determined with a good deal more accuracy than others, an indication of the extent to which they
can be relied upon is given in the column headed ' Degree of accuracy'. The quality of the evidence
on which they are based varies considerably, and it is difficult to assign a very exact meaning to the
terms entered in this column. As a rule 'V. good' implies that there was a sharp rise or fall in the
thermograph record, clearly identified as the convergence. The possible error in the estimated position
is not likely to be more than about 5 miles either way. ' Good' may indicate a thermograph gradient
which was well defined, but which extended over some hours of steaming ; it may mean that the crossing
was unmistakably indicated by the ship's 4-hourly routine temperature readings; or ^^ -ay be used
when there was no definite change at the surface, but when two vertical ^^at-ns, not too far apart
showed that the convergence lay between them. The maximum error here might be ^b-^ 2° -1 s
either way. 'Approx.' generally indicates that the convergence may have been crossed anywhere
between two pofnts abo.' 50-150 miles apart; or it may be used, for example, where the only evidence
was a vertical station with a minimum temperature at or about 200 m., for this does not necessarily
mean ^h the station lay exactly on the convergence. 'Probable' generally implies that there was
rv^ltlhaAhe converU hid been crossed at a position which could be d-mmed^^^^^^
precision but that it was not completely certain that it was in fact the convergence The erm is of en
ariicabk when there are several sharp changes in temperature which may indicate oops in th
clergenci 'Uncertain' is applied to any of the less satisfactory records, and may indicate doubt
as to the identification of the convergence or as to the accuracy of the estimated position.
In normal times this work might have been undertaken by Dr G E. R. Deacon^J-J^-S- D-^^^^
the war he has been in service with the Admiralty, but we have together reviewed the data on which
each position L Table 9 is based, and I am indebted to him for this assistance and for reading through
the manuscript of the paper.
THE MEAN POSITION OF THE CONVERGENCE
Many of the positions in Table 9 are from isolated "bse-ations but sometinaes Ae convergence^^^^
erossed more than once at positions not far apart and w.thm a few days or weeks The convergence
does no. appear to change its position very quickly, and such points can often be jomed up so as ,0
:°::a:Lr:.onofitsLrse^ta.ive„..™^^^^^^^^
Llralso F r, P .0.). Nos. 88 and 90 were a few weeks later and are jomed by a separate hne.
The tetf the convergence on these occasions may be compared wrth the mean hne ,n Plate I.
l82
DISCOVERY REPORTS
Figs. 2-4 include the positions listed in Table 9 except No. 20 which was vague and unreliable.
Different positions have been joined up, wherever it seemed reasonable to do so. This seems to provide
the best basis for estimating the mean position of the convergence, and it gives an impression of the
extent to which its position is liable to vary. The majority of positions are in the Atlantic sector. Data
are still rather scarce elsewhere, and the points plotted in Fig. 4 are joined by lines drawn simply in
what appear to be the most probable positions. In Plate I the convergence is drawn to represent as
nearly as possible the mean of the lines in Figs. 2-4. Wherever possible the average latitude has been
calculated within narrow limits of longitude, and those records which are marked ' Good ' or ' V. good ' in
Table 9, were counted twice in the averages so that some bias should be given to the more reliable data.
Fig. I. The Antarctic convergence in the Falkland Islands Sector, September to December 1934. Numbers refer to the serial
numbers in Table 9 (p. 205). Black dots denote well-defined positions, and rings approximate or uncertain positions. Pecked
lines show the ascertained position of the convergence (apart from unknown minor irregularities), and dotted lines show its
probable course where it is not checked. Continuous lines show the tracks of the 'Discovery 11' (see also Fig. o, p. 201).
The mean line of the convergence, shown in Plate I, like most of the results set forth in this paper,
is an estimate which is open to adjustment in the light of any additional data which may be obtained
in the future. It is obvious from Figs. 2-4 that it is based on far better material between 80° W and
30° E than elsewhere. In the Falkland region the estimated position must be very near the true mean
position, but in the Pacific sector, between no and 160? W it may be far from accurate.
Some comments are needed on the extent of variation from the mean. I have made a rough measure-
ment of the distance of each plotted point from the estimated mean position, and the resulting figures
given in Table i are perhaps worth noting. This table includes all the positions in Table 9 for which
a measurement could be made, and the possible extent of the displacement may not be so great as it
seems to suggest. Most of the larger deviations were at places where either the actual record of the
convergence was not very certain, or where the mean position and course of the convergence are based
on inadequate data. Furthermore, it is known that sub-Antarctic surface water is sometimes thrust
a considerable distance south of the normal position of the convergence, and the extensive loops in
the convergence (seen in Figs. 2, 3) show that there is an element of chance in the latitude at which
it is found at an isolated crossing. This last factor may indeed account for many instances of apparent
displacement from the mean position. Probably a displacement of 50 miles or so either way is not
uncommon, but it may be that the extreme displacement does not exceed about 100 miles.
THE ANTARCTIC CONVERGENCE
Table i. Deviation from the mean position of the convergence
183
Miles north or
south of mean
position
Well-defined positions All measurable positions
Number
0/
/o
Number
%
0- 24
25- 49
50- 74
75- 99
ioa-124
125-149
150-174
31
15
7
I
I
I
0
55
27
12
2
2
2
0
65
38
14
8
4
I
I
49
29
II
6
3
I
I
Total
56
100
131
100
The latitude of the convergence is not much affected by the time of year. Table 2 shows the average
deviation for each month. The figures for June, July and August are based on so few records that they
should probably be disregarded, and in the other months the average deviation is less than 20 miles
except in February. There is a suggestion of a small northerly displacement in the cold months
(September-November) and a southerly displacement in the warm months (January-March), but if
the latitude of the convergence really varied with the temperature we should not expect the average
for April to be so far north. The average monthly deviations are small compared with the actual range
of deviation, and the convergence may be found in very different positions even in the same month.
For example, Nos. 60 and 109 in Table 9 were both in November in different years and in about the
same longitude, but the former crossing was over 50 miles north, and the latter about the same distance
south of the mean position. Although there may be a minor seasonal oscillation, it seems probable
that variations in the latitude of the convergence depend mainly on local and temporary factors, such
as shifting loops and irregularities in its course, and perhaps to variations in winds and currents.
Table 2. Average monthly deviation from the mean position of the convergence
Month
Miles north or
south of the
mean position
Number of
instances
Month
Miles north or
south of the
mean position
Number of
instances
September
October
November
December
January
February
13 N
10 N
16 N
3N
19 S
28 S
9
II
16
17
9
17
March
April
May
June
July
August
11 S
15 N
12 S
(34 N)
(15 S)
(10 N)
16
16
9
4
2
5
Total
131
As already noted, the mean position of the convergence, drawn in Plate I, does not differ much from
that arrived at by Deacon (1937, fig. 4). It is farthest south about 80° W and 180° W, and farthest north
between 35° W and 70° E. It is possible that some of the minor bends in its course will be smoothed
out in the light of additional data. There is little doubt that the S-shaped turn between the Falkland
Islands and South Georgia is a normal feature, but it does not always exist, for there has been at least
one occasion when the 'Discovery 11' has crossed this region on a meridional course and found no
trace of the southern limb of the loop. There is good evidence for the sharp southward turn in 60° W.
i84 DISCOVERY REPORTS
New data may suggest a modification of the loop in 30° E, but some distortion of the convergence and
isotherms here is probably normal. Deacon (1937, pp. 34 and 92, and plate xliv) refers to a steep ridge
on the sea floor at this point, which may affect conditions at the surface. The slight northward bends
shown about 140, 97, 75, 67 and 35° W and 110° E need confirmation, and the whole of the con-
vergence between 1 10 and 160° W, and between 50 and 90° E is based on very little material.
Fig. 2. Positions of the convergence between 10° W and 80° W. Where two or more positions were ascertained within a short
period they are joined by a continuous line if the convergence was well defined, and a pecked line if the positions were approxi-
mate or uncertain. Single positions are indicated by a short line, continuous or pecked, drawn approximately parallel to the
mean line of the convergence.
Fig. 3. Positions of the convergence between 10° W arid 40° E. See legend to Fig. 2.
It is noteworthy that the convergence divides the Southern Ocean into two almost equal zones;
in fact it lies very nearly half way between the Antarctic coastline and the extremities of each of the
southern continents. From Cape Agulhas (South Africa) the distance to the convergence is rather
less than half the distance to the nearest part of the Antarctic continent, but from Western Australia,
Tasmania, New Zealand and Cape Horn the convergence is very close to the half-way point. It is
difficult to say whether this is of any special significance. The fact that the convergence lies in about
THE ANTARCTIC CONVERGENCE
i8s
Fig. 4. Positions of the convergence other than those between 80° W and 40° E.
i86 DISCOVERY REPORTS
the same relative position at each of these points rather suggests some balance of pressure between
water masses drifting through those parts of the Southern Ocean which are hmited to the north as
well as to the south by land masses. Sverdrup, Johnson and Fleming (1942, p. 607) say: 'The south-
ward displacement of the Antarctic Convergence to the south of Australia and New Zealand can
probably be ascribed to the relative narrowness of the passage between these regions and Antarctica,
and the similar displacement off South America can be attributed to the southerly location of Drake's
Passage separating South America from Graham Land.'
It is also worth while to compare the position of the convergence with the mean northern limit of
the pack-ice. The pack extends farthest north about September and October, and the line in Plate I
is a reproduction of the September-October line shown by Mackintosh and Herdman (1940, plate Ixix).
It is evident that the position of this line bears some relation to the position of the convergence. The
extreme limit of the ice will of course lie nearer to the convergence, but it is doubtful whether the pack
ever reaches the convergence itself.
The distance between the convergence and the mean northern limit of the ice shown in Plate I
varies from 120 miles in the Scotia Sea to about 550 miles in the Indian Ocean sector, with an average
of 370 miles. This may be taken as a belt which is practically free of pack-ice throughout the year.
There is some reason to suppose that it has a characteristic plankton fauna and flora. On the basis
of the distribution of macroplankton in the Falkland sector I distinguished a ' northern zone ' imme-
diately south of the convergence (1934, p. 150), and Hart (1942, p. 280 and Fig. 2) found that the
phytoplankton could be suitably divided into certain biogeographical areas, one of which he took as
the 'northern region' lying between the convergence and a line 330 miles to the south of it excepting
certain special areas. This region was in fact distinguished largely on the grounds that it is normally
free of pack-ice at all times. This is a matter to which I hope to return in a future paper.
THE CHANGE OF TEMPERATURE AT THE SURFACE
The convergence is generally distinguished by a sharp change of temperature at the surface, which
usually appears as a steep gradient on the thermograph, and it is important to consider what variations
occur in the position of the gradient on the temperature scale, and the range of the gradient. This is
best done by examining first the variations in the middle temperature of the gradient, and then the
variations in the range. The aim of the present section is to estimate (as nearly as the data permit)
what change of temperature we may expect to find in any longitude at any time of year, and hence
how to place the isotherms at the convergence when drawing charts of surface temperatures.
Deacon (1933, fig. 10, p. 190) gives an example of the gradient as shown on the thermograph. This
was No. 59 in Table 9, and it is an exceptionally well-defined crossing. Some further typical examples
are shown here in Fig. 5 A-E. The middle temperature is taken as the mean of the temperatures at
the beginning and end of the gradient representing the convergence, or, if the gradient is not very
clearly defined, whatever seems best to represent the central temperature of the convergence. It is
easy enough to decide the middle temperature and range in such examples as Nos. 33 and 106 (Fig. 5).
In such as No. 88 it is difficult, but these small and ill-defined gradients cannot be excluded from
calculations of the average. In this case the middle point was taken as 3-0° and the range as o-8°, on
the assumption that the rise of temperature immediately to the right of St. 1476 represented the
convergence.
Table 3 is a list of all the middle temperatures and ranges which I have been able to measure. It
includes some rather doubtful measurements, such as No. 88, but should serve for provisional estimates
of the changes of temperature to be found in different months and positions. Usually both the middle
HON I DAY
a .1 *
TUESsOAY
THE ANTARCTIC CONVERGENCE
* ■» 8 i "> o '• ' _" ~ . iq
H---dor-t--f t^THi-/ ; r''333Eirn m
IrZS'fr
::50
::,1S
10
• 451
1 . 5"^
0
-■'■■ v. ''■
-T^^.—n
452: \
\ A, :
„S- i.
i:;;,.,-0.., :^A...
:l_^..i.,i.I.i,.^i.a...l...x...,t...x„i
187
SATU I RSAV
n ? 4
4 8
[, 1 1 i [:::|i#H4^
* ; ♦■■-?■■■
iElESE.^ I i ; £
^-■•tESEs
=p$:U ' ; ' r '"^^T^t^-I—^
320
is
I
A.C.4:Z
4 8^4
T7-rTTTT7
i. t i-t
35 'r , ' '
.UHlcm ..ON.I.OAY . THUKlSSAr
^ «; 4 "
30
25
20
« a 5^ ■< 8 ' ■»
35 - , ' V ' ' /---r-T";-;
30
25
20
15
' r I I 7
AC^yV
AC 106
^?l:, \ 1357
f\
185&
\ \
4M36
i-5 ^ ^
.7^
00
r
\ \
5X1%,
V V \ ^ \
J
Fig. 5 A-E. E..a.p,.s .r .he™o,„ph records cro.!^, the A„.a.>^^^^^^^^^^^
tu? J ^iTry-t So" 33 Kbr.'C.f.n'i^Ir^f sr„Sler ,he periods dur.ng which ,he sh.p w.s .topped
or towing nets at slow speed. 2-2
i88
DISCOVERY REPORTS
temperature and the range are given, but sometimes the former could be distinguished while the
beginning and end of the gradient were too ill defined for measurement of the range. In some cases
where the change of temperature was small, and the convergence and its gradient could not be dis-
tinguished, the middle temperature is omitted, but, for calculating the average, a small figure is entered
as an estimate for the range. In one or two other instances an approximate figure could be given for
the range though the middle temperature was very obscure (e.g. No. 49).
Table 3. The change of temperature markmg the Afitarctic convergence at the surface.
Middle temperature and range of the gradient
u
Middle
temperature
u
n 0
Pi
u
.2 XI
Middle
temperature
Pi
t-t
:2^
Middle
temperature
1)
a 0
Pi
Middle
temperature
CQ 0
Pi
1
Serial
number
Middle
temperature
(3°
I
5-25
2-1
43
2-00
3-2
67
4-00
1-8
91
4-40
—
121
4-00
i-o
3
2-20
—
44
3-80
2-5
68
2-55
1-3
92
4-00
2-6
122
4-50
2-8
4
3-80
0-4
45
4-30
17
69
3 -20
3-3
94
4-50
2-0
123
4-20
1-4
6
4-50
—
47
4-10
2-2
70
4-50
07
95
3-40
o-S
124
4-15
2-5
7
4-90
2-0
48
5-25
2-5
71
4-00
i-i
96
3-40
II
125
4-40
1-5
12
2-00
—
49
—
2-0
72
3-50
1-5
99
5-50
—
126
3-50
2-2
13
3-20
—
50
3-10
1-0
73
4-10
3-4
100
5-80
—
127
4-05
1-3
15
5-89
—
51
4-45
1-9
74
4-45
2-3
102
5-3°
—
128
3-4°
1-4
17
3-8o
—
52
3-60
2-2
75
4-50
2-0
103
2-8o
—
129
2-85
17
22
2-6i
—
53
3-45
2-9
76
2-8o
2-0
104
2-95
—
130
2-50
1-0
24
—
0-3
54
4-00
4-0
77
2-50
1-0
105
2-00
1-2
131
170
1-0
27
4-08
—
55
2-00
1-4
78
3-20
2-4
1 06
3-15
2-1
132
2-40
1-0
29
3-40
2-8
56
375
3-5
79
—
2-4
107
3-50
3-0
133
2-85
0-9
30
3-50
3-0
57
4-25*
1-5
80
1-50
1-5
109
2-15
3-2
134
3-50
i-o
32
5-85
27
58
2-00
2-0
81
270
2-2
III
4-20
2-0
135
4-50
1-0
33
4-00
I-O
59
1-35
3-5
82
—
1-9
1 12
5-50
2-0
136
—
0-5
35
—
0-8
60
3-10
3-8
83
1-65
I "3
113
5-50
1-0
137
4-90
0-5
36
—
07
61
3-63
i-o
84
175
II
115
4-50
0-6
138
5-30
0-3
37
475
2-5
62
3-85
17
85
2-00
2-0
116
3-25
2-3
139
5-15
3-3
38
4-40
1-2
63
3-30
2-4
86
1-30
2-0
117
4-10
1-8
40
4-25
2-5
64
4-00
1-8
87
2-15
2-1
118
3-50
1-0
41
2-00
2-0
65
4-50
10
88
3-00
0-8
119
4-00
—
42
2-50
2-0
66
4-90
I-O
90
3-10
37
120
4-50
1-5
* Doubtful.
Analysis of this table indicates that the middle temperature varies mainly according to the time of
year, and to some extent according to the latitude of the convergence, while the range varies in different
longitudes and appears to depend on other factors.
Taking the middle temperature first, Table 4 gives the average and maximum and minimum values
for each month in arbitrary ranges of latitude. Divided in this way the quantity of data is rather small,
and it is evident that the middle temperature can vary considerably in any one month and latitude,
but it is obvious from the average figures that it generally falls to a minimum about September and
October, and rises to a maximum about February. This is naturally to be expected since the sea tem-
perature varies with the time of year while the latitude of the convergence is little affected. The figures
also leave little doubt that where the convergence is in a high latitude the middle temperature tends
to be lower than where it lies in a low latitude, though the difference is not very great. This implies
that surface isotherms near the convergence do not always run quite parallel to it.
THE ANTARCTIC CONVERGENCE
189
Table 4. The observed middle temperatures of the co?wergence gradient for month and latitude
North of 50° S
5o°-54° S
S4°-S8° S
South of 58° S
i i —
Mean
No.
2
Max.
Min.
Mean
No.
Max.
Min.
Mean
No.
Max.
Min.
Mean
No.
Max.
Min.
September
2-25
2-50
2-00
—
—
—
i-6s
3-15
2-82
3
375
2-00
175
1-32
2-38
2
2
2-00
i'15
1-50
1-30
October
November
2-63
3-63
3
5
4-00
4-10
1-70
2-8s
2-21
3-43
3
3-25
3-63
2-59
4
3-10
2-15
3
3-00
2-00
December
4-04
4
4-.;o
3-5°
3-57
6
4-20
3-20
2-7b
3
3-20
2-00
2-55
January
February
S-iS
4
.;-30
4-90
4-20
4-37
2
3
4-40
4-50
4-00
4-10
4-50
5-17
I
3
5-51
4-50
4'5°
4-47
4
5-89
4-00
March
April
4-64
4-6i
4
4
5-15
5-30
4-00
4-25
5-23
4-25
3
3
5-80
4-50
4-40
3-8o
4-30
379
3-60
372
4
2
475
4-08
3-5°
3'50
4-15
2
4-90
3-40
May
June
4-33
2-95
2
I
5-85
2-8i
3-45
3
4-45
2-8o
2
4-00
3-45
—
—
—
—
July
4-05
I
—
—
3-50
I
~~
_
—
August
2-99
4
3-40
2-5°
2-00
I
1
It is evident' that the average figures can be improved by some method of smoothmg, even though
the data are insufficient for an accurate estimate of the true averages for each month and latitude.
I have therefore adopted the rough and ready method of plotting the averages and drawmg curves
by eye to make as good a fit as possible. Figures showing the average for all latitudes in different
months, and for all months in different latitudes might be misleading, for there are insufficient winter
observations in the higher latitudes. In Fig. 6 A-D therefore the monthly averages are plotted separately
for each range of latitude and the curves are drawn both to represent the monthly trend of the points,
and to resemble each other, for it is assumed that the rate of change does not differ very much in
different latitudes. In Fig. 6 E the four curves are shown on the same scale and are marked with the
middle latitude of each range. Thus the second curve applies to all records between 50 and 54 S and
can be taken to represent 52° S. It will not be far wrong to say that those north of 50 represent 48 b
and those south of 58° represent 60° S.
Table 5. Provisional estimate of the average middle temperature of the convergence gradient
in different months and latitudes {figures smoothed)
46° S
48° S
50° S
52° S
54° S
56° S
58° S
60° S
62° S
Average
September
October
November
December
January
February
March
April
May
June
July
August
2-6
2-6
3-9
4-3
5-0
5-4
5-1
4-6
4-2
3-9
3-6
3-2
2-5
2-5
3-6
4-1
4-9
5-2
5-0
4-5
4-1
3-8
3-5
3-1
2-4
2-3
3-4
3-9
47
5-0
4-8
4-4
4-0
37
3-4
3-0
2-3
2-2
3-2
37
4-6
4-9
47
4-3
3-9
3-6
3-3
2-9
2-1
2-1
3-0
3-4
4-5
4-8
4-6
4-2
37
3-5
3-2
2-8
2-0
2-0
2-8
3-2
4-4
47
4-5
41
3-6
3-4
3-1
27
1-9
1-8
2-5
3-0
4-3
4-5
4-4
4-0
3-5
3-3
3-0
2-6
1-8
17
2-3
2-8
4-2
4-4
4-3
3-9
3-4
3-2
2-9
2-5
17
1-6
2-1
2-5
4-1
4-3
4-2
3-8
3-3
3-1
2-8
2-4
2-1
2-1
3-0
3-4
4-5
4-8
4-6
4-2
37
3-5
3-2
2-8
Average
4-0
3-9
3-8
37
3-5
3-4
3-2
3-1
3-0
3-5
Table r is derived from Fig. 6, and it includes figures for interpolated latitudes and a slight extra-
poltion to 46 Id 6." S. The'steps leading up to this table have been explained ■" <i«f --'_;'/-
Le idea of the extent to which it can be rehed on. This can be judged from F,g. 6 A-D and the
igo
DISCOVERY REPORTS
SEPT OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC
I I I I I I I I I I I I I I I I
50-
4 0
3 0-
50-
50 -54 S
50-
40-
3-0-
2 0-
I I I I I I I I I I I I I I I I
SEPT OCT NOV DEC JAN FEB MAR APR MAY JUNt JULY AUG SEPT OCT NOV DEC
Fig. 6 A-E. Variations of the middle temperature of the convergence gradient. Averages from Table 4 are shown as black
dots when based on three or more records, and as rings when based on one or two. The curves in Figs. A-D are reproduced
on one temperature scale in Fig. E.
THE ANTARCTIC CONVERGENCE 191
columns in Table 4 headed 'No.' (number of observations). A larger body of data would no doubt
result in corrections to most of the figures, but there are two purposes for which the table should be
useful. First it is helpful in drawing the charts of surface temperatures with which the latter part of
this paper is concerned, and secondly it should sometimes make it easier in the future to locate the
convergence. For instance, there is sometimes more than one sharp change of temperature in the
neighbourhood of the convergence and the figure given above under the appropriate month and latitude
may help to show which actually represents the convergence.
Fig. 6 shows a rapid increase of temperature from October to February, though the increase is
slightly retarded from November to December. Although the curves are drawn tentatively there is
evidence for this check between November and December. It can be seen in each set of plotted points
in Fig 6 and the averages for these two months are based on more data than most other months.
Before their work was interrupted by the war Mr A. J. Clowes and the late Mr J. A. Nicholson made
a calculation of the seasonal rise and fall of temperature in part of the Antarctic surface layer by
statistical treatment of the vertical temperature readings at a number of stations made by the
'Discovery IF on the Greenwich meridian, and the curve so obtained showed a similar reduction m
the rate of increase of temperature. This work was not published, but they suggested that the reduction
might be a latent heat effect resulting from the melting of ice.
Table 5 further indicates that the middle temperature varies through an average annual range of
about 2-7° C and by about i-o° C for 16° of latitude. The actual highest and lowest middle tempera-
tures were 5-89° (No. 15, Feb., 58° S) and 1-30° (No. 86, Oct., 62° S). Table 6 is a comparison of
actual middle temperatures with the expected temperatures given in Table 5.
Table 6. Deviation of observed temperatures from smoothed average middle temperatures
of the convergence
Degrees above or below
Number of
Percentage
average temperature
instances
O-00-O-49
62
59
0'5o-o-99
33
32
I-00-I-49
5
5
I-50-I-99
3
3
2-00-2-49
I
2-5<^2-99
0
0
104
100
When the convergence lies north of its normal position the middle temperature is usually (though
not always) a little higher than the average, and vice versa. This suggests that variations m the position
of the convergence are not dependent on variations of the sea temperature.
It is not certain that Table 5 is equally apphcable to all longitudes. There are insufticient data from
many parts of the Southern Ocean to test this, but there seems to be no reason why the middle
temperature should vary in different longitudes. „ j ,
When the convergence is crossed the amount of change of temperature at the surface, called here
the 'range', may be as much as 4-0° (No. 54, 162° E) or may be so small as to be mdistinguishable
from other minor fluctuations of temperature (e.g. No. 138 in 01° E). A typical well-defined con-
vergence is seen in Fig. 5 B (p. 187), and here the range is taken as from 1-5 to 3-5 , ^-e- 2-o C. Ihe
range varies mainly according to longitude; that is to say, it is generally distinctly greater in some
regions of the Southern Ocean, than in others. There seems to be no regular variation in the range
at different times of year; it does not seem to be affected when the middle temperature of the
192 DISCOVERY REPORTS
convergence is higher or lower than the relevant average figure in Table 5 ; nor does it appear to vary
according to the latitude of the convergence in its mean position ; but there is almost certainly a tendency
for the range to increase a little if the convergence is displaced to the north or south of its mean position.
This effect is not obvious, for it is obscured by the more important regional variation, and more data
would be needed before it could be measured.
For measurement of the regional variation the average range is plotted in Fig. 7 in convenient groups
of longitude, usually between meridians 10° apart. Thus between 20 and 30° W there were five crossings
with an average range of 2-4° C ; between 5° E and 5° W there were ten with an average range of 1° C.
Here again a curve is drawn to mark the trend of the plotted points. The data for this curve are rather
uneven. It is well supported in the Atlantic region, but in the Indian Ocean there is not very much
to rely on, and in the Pacific sector, between 100 and 160° W, the curve is little better than guesswork.
The reason for these variations in the range of the convergence gradient is not quite certain and will
need some further investigation.
For the purpose of placing the isotherms about the line of the convergence in Plates II-XII, I have
found it convenient to take the range of the gradient at every 10° of longitude from the curve in Fig. 7.
The resulting figures are shown in Table 7, which gives an overall average range of 17° C.
From Plate I and Tables 5 and 7 it is now easy to estimate the expected position of the convergence
and change of temperature at the surface, so far as the available data suggest. For example a ship
steammg south in 20° E in December would expect to meet the convergence any^vhere within about
50 miles north and 50 miles south of 48° 20' S (see Table i and Plate I). Table 5 gives the middle
temperature of the gradient as 4-1" and Table 7 gives the range as 1-4°. We should therefore expect
the temperature to fall from 4-8 to 3-4° C. The chances of the gradient being warmer or colder than
this can be judged from Table 6, but the chances of finding a larger or smaller range are more difficuh
to assess.
Table 7. Provisional estimate of the average range of the convergence gradient in different longitudes
180° w
1-0° C
60'
W
2-6° C
60=
E
0-9
170 w
i-o
50
w
1-8
70
E
0-9
160 w
17
40
w
19
80
E
0-9
150 w
2-4
30
w
2-3
90
E
i-o
140 w
2-4
20
w
2-3
100
E
i-o
130 w
21
10
w
17
no
E
i-i
120 W
17
0
10
120
E
1-6
no W
17
10
E
0-8
130
E
2-0
100 w
2-0
20
E
1-4
140
E
2-4
90 w
2-0
30
E
2-1
150
E
2-8
80 w
1-8
40
E
2-2
160
E
2-8
70 w
1-9
50
E
1-3
170
E
2-3
It is to be hoped that further data will be collected in the future, especially in the winter, and from
the Indian Ocean and Pacific sectors. It is desirable, for example, that Tables 5 and 7 should be
recalculated from fuller material, and much useful information might be obtained if the line of the
convergence at the surface could be followed and mapped out for considerable distances.
THE ANTARCTIC CONVERGENCE
193
"3 S33MD3a Nl 33NVM
a o o o
«i*. r) w —
1 I I I I I I
•Z'N3d¥D'S •
O
VINVWSVl .
3dV33 S
NIMn33T0 •
I
LI
U
o
5
z
3dOH0009J03-
z
U
O
u
P
z
<
-1
N^OH'O •
ZN 3dV:) S-
1 I
9
1 1 1
9
1 1
9
1
0
<
(«i
<u
—
to
«
>
i>
bi)
0
rt
^
1m
U
>
rt
D
0
^
■ nj
■ 1-
0
<
■S .
— uJ
a
w-a
CO 0
.5-3
• D
h
2'3
0 "J
o-c
■<o z
0
"^'^
_l
•" u
0 0
0
ID
-as
u ^
-fi "t::
H °
0
■^
U
^1
-a e
2S
.0
bC
fy
c "
_o -g
4J ^
-0-
ij
rtj t/i
_
it: m
•— 1-1
"O 1)
0
„ XI
(U
•s s
•w D
~
c C
1)
0
2 «
at) -
m
4J -0
0
0 S
"ID
c 0
1-
u u
\h
M U
~ iti
t, u,
0?
1> ^
"ttj u
D
_ D
0 0
0 ca
f-
u •5'
J3 tn
0 i3
4-*
-oz
0
u
60
0
C
-ru
rt
Lri
_
(U
60
0
"5
S
u
>
rt
u
0
j=
-tD
U.4
0
CO
0
.0
.2
*n
c^
0
>
r^
bib
•••4
fe
•3 533^930 Nl 39NVy
"94 DISCOVERY REPORTS
PART II. THE DISTRIBUTION OF SURFACE TEMPERATURE
IN ANTARCTIC WATERS
INTRODUCTION
In 1940 I published, in collaboration with Mr H. F. P. Herdman, a paper on the distribution of the
pack-ice, which consisted essentially of a series of charts showing the observed or estimated latitude
of the ice-edge in each month in all parts of the Southern Ocean. The present paper is the next step in
mapping the surface conditions in Antarctic waters, and its object is to estimate, as nearly as the data
allow, the mean positions in each month of the surface isotherms between the ice-edge and the
Antarctic convergence.
During the cruises of the ' Discovery ', ' Discovery II ' and ' William Scoresby ' between 1926 and
1939 a very large quantity of data on surface temperatures has been collected. These data of course
include far more than the surface temperatures published in the Station Lists {Discovery Reports
vols. I, III, IV, XXI, XXII, XXIV). It is the thermograph carried in these ships which furnishes the mosi
important body of data, for this instrument was in continuous use (except for one short period)
throughout the voyages of the 'Discovery' and ' Discovery II', and in the last commission (1937-8)
of the William Scoresby '. v vj / y
The data are unfortunately distributed very unevenly in time and space, for the movements of the
ships were generally dictated by other considerations than covering the surface as equally as possible.
u ', r ^"' ^'"""^^ ^^^'^ observations were obtained in each month, and it will be seen
that although the Atlantic sector is well covered at nearly all times of year there are large gaps in
other sectors, especially in the winter months.
Some data on surface temperatures south of the Antarctic convergence are of course available from
other sources. These are the observations made by other Antarctic expeditions and some published
observations made by whaling factories. I have not, however, obtained very much assistance from
this material. A large number of the observations were made within the pack-ice or near its fringe
and It IS not certain whether some of the others are sufficiently accurate (see pp. 197-8). The number of
reliable records from all other sources of temperature in the open regions of the Antarctic surface water
are very small compared with the Discovery Committee's material, and they cover few of the gaps
mentioned above. ^ ^
Since the surface temperature in any given locality depends primarily on the time of year, I have
set out to plot all appropriate records on separate charts for each month, irrespective of the year in
which they were made. In some years of course the temperature at a given place in any one month is
above or below the average, but these irregular annual variations seem to have only a local and tem-
porary significance, and their range is much less than that of the regular seasonal rise and fall of
temperature. The charts used were semicircular overlapping charts identical with those used for
plotting the ice-edge (Mackintosh and Herdman, 1940). The isotherms were then drawn to fit the
actual observations as nearly as possible. Some adjustments were made after a comparison of the
monthly position of each isotherm, and in areas where no observations were available in certain months
entative isotherms were filled in by interpolation as described below. Finally, the isotherms were
ransferred to the circumpolar charts reproduced here in Plates II-XII. The various steps in the
treatment of the data are described below in some detail because they raise certain points in technique
which seem worth recording. ^
DISTRIBUTION OF SURFACE TEMPERATURE iQS
TREATMENT OF THE DATA
As mentioned above, the continuous thermograph provides the most important body of data. In this
instrument the thermometer bulb Hes in a pocket in the ship's hull about 14 ft. below the surface,
and the temperature at that depth is recorded on a chart (see Fig. 5, p. 187) attached to a clockwork
drum in the laboratory on deck. Each chart covers a week. The instrument has given very satisfactory
results and, provided certain checks are made, the temperature can be read correctly to within less
than 0-2" C, and the time to within less than half an hour. Certain possible sources of error have to
be considered. The clock has generally been found to keep excellent time, but allowance had occasionally
to be made if the drum had not been accurately set when the new chart was attached. The timing could
generally be checked at points where the ship was stopped on station, for here the temperature is often
traced as a steady horizontal line which is usually distinguishable from the numerous small fluctua-
tions of temperature which appear while the ship is moving (for example, see Fig. 5 A, p. 187, St. 451).
The time when the ship is stopped can of course be checked from the logs. Correction of the tempera-
ture is more important as there is frequently a slight error which is possibly due to a small amount ot
play on the arm of the pen. This error seldom exceeds 0-5^ C and is usually constant until the chart is
replaced unless substantial changes of temperature occur. The thermograph record should, however,
always be checked against direct thermometer readings, preferably with the Nansen-Pettersson water-
bottle Since the ' Discovery ' and ' Discovery II ' normally worked full stations at frequent intervals,
there were few thermograph charts which could not be checked in this way at several points, but it is
verv evident that the method of reading the temperature in a bucketful of water drawn over the ship s
side is wholly unreliable as a check, however carefully the thermometer is read. Such readings seem
nearly always to be too high and the error may amount to i-o° C or more. This method has been
criticized elsewhere (see Lumby, 1928). , , r r. 1
A further question which arises is whether the thermograph, recordmg at a depth of 14 ft-, Proper y
shows the 'surface' temperature, and whether thermometer readings at o m. constitute a valid check.
In other words, what diflFerence may there be in Antarctic waters between the temperature at o m
and about s m.? At full stations the vertical series of temperatures, etc., included readings at o and
10 m and I have worked through the station lists and assembled the data on this point m Table 8.
This includes 987 vertical stations at which the temperature at o m. did not exceed 5-0 C (i.e. the
vast majority of stations south of the Antarctic convergence) and it will be seen that, except in the
summer months (December-February) about 90% showed less than o-z^ C difference between o and
10 m Even in the summer months when occasional superficial sun-warmed patches may be expected,
any significant difference is rare, and the difference between o and 5 m. must be still less. There have
beL one or two occasions when a vertical station has been worked in the vicinity of an iceberg, and
where a cold, less saline, surface stratum has depressed the temperature at o m. below that at 10 m
In any case where there is a significant diflference between o and 10 m. it is preferable to check the
thermograph chart from other stations. , , 1 . f*^„
Since ship's time changes with changing longitude the clocks are frequently changed at sea, often
several times in a week. The thermograph cannot conveniently be adjusted to ship s time and is there-
fore set at G.M.T. But the temperature readings must be linked with the ship s position the stations,
nd her observations, all of which are recorded by ship's time. It is therefore desirable to mark a
"ak of ship's time on the thermograph charts. I found it best to trace the corrected temperature on
to new charts, marked with the ship's time scale, periods on station, and other -notations.
The next step was to draw track charts, marked with ship's time, for all voyages south of 45 S
except such complex local movements as were undertaken in the oceanographic surveys around South
196
DISCOVERY REPORTS
Georgia and in the Bransfield Strait. The tracks were plotted on the usual semicircular charts. The
scale was 6 mm. to a degree of latitude, and it was thus possible to mark the ship's hourly positions
by dots at intervals of a millimetre or less (i mm. = 10 miles). The ship's positions recorded in the
logs, usually at 4-8 hr. intervals, were plotted first and the hourly positions interpolated. Such plotting
of the ships' tracks over many thousands of miles was naturally very laborious, but the resuhing charts
should in due course facilitate the plotting of various observations in addition to surface temperatures.
A small part of one of these charts is reproduced on the original scale in Fig. 8 A. These track charts
were then compared with the thermograph charts and the ship's track was marked in pencil at each
point at which the temperature line crossed the line of a whole or half degree C (-0-5, o, 0-5, i-o°, etc.).
The hues of the ships' tracks, together with the temperatures were then traced on to monthly tem-
Table 8. Frequency of differences between the temperature at o m. and at 10 m. {including all records
where the temperature at o m. does not exceed 5° C)
Range of
difference
rc)
September,
October,
November
December,
January,
February
March,
April,
May
June,
July,
August
No.
0/
/o
No.
/o
No.
/o
No.
%
0-00-0-09
0-10-0-19
0-20-0-29
0-30-0-39
0-40-0-49
0-50-0-59
0-60-0-69
0-70-0-79
0-80-0-89
0-90-0-99
I -00-1-09
1-10-1-19
1-20-1-29
1-30-1-39
I -40- 1 -49
239
22
3
I
I
89-8
8-3
i-i
0-4
0-4
368
46
26
12
6
8
3
I
2
I
I
2
77-3
9-7
5-5
2-6
1-2
1-7
0-6
0-2
0-4
0-2
0-2
0-4
196
II
5
2
I
91-2
5-1
2-3
0-9
0-5
27
2
I
90-0
6-7
3-3
266
100
476
100
215
100
30
100
perature charts, all data for any one month being included on one chart or set of overlapping charts,
irrespective of the year. The year was, however, marked against each track. A comparison of Figs. 5 and
8 will^show the method of plotting. For example, in Fig. 5 A (p. 187) the thermograph record crosses
the 3° level at noon on 14 October 1930. The position of this point is found in Fig. 8 A just north of
50° S, and hence 3° is marked in the corresponding position on the ship's track in Fig. 8 B.
A large number of temperature records at single positions were then added to the temperature
charts. These included the surface temperatures read at stations made by the Discovery Commhtee's
ships when no thermograph chart was available (mainly derived from the 'William Scoresby'), and
published records from other sources.
Part of a chart of October temperatures is reproduced as an example in Fig. 8 B. The mean isotherms
indicated here are in their final position, as shown in Plate II, after comparison with the corresponding
isotherms of other months. They represent the temperatures for both September and October, which
are very similar; but there is not room to show the observed temperatures for September on the same
chart, for in places the ship's tracks for the two months almost coincide. It should be mentioned that
DISTRIBUTION OF SURFACE TEMPERATURE i97
the mean isotherms cannot always be made to correspond with the obser^'ed temperatures so well as
they do in this chart. • . tv/t *
Of the published records from other sources, perhaps the most important are those m the Meteor
Reports (Bohnecke, 1936). Here a very large number of surface temperature readings are tabulated,
but they are from all parts of the Atlantic Ocean and only a small percentage he south of the Antarctic
convergence Bohnecke gives two principal tables. One shows the temperatures (also S°/oo, etc.)
recorded by the 'Meteor', with date and position (pp. 41-64). Only a small number of these are m
Fig 8 A. Part of a track chart for October 1930. showing hourly positions of the ^Discovery II ' during a voyage from Cape
Town to South Georgia. See Fig. 5 A for the thermograph record of part of this track.
12 n X- Ship's position at 1200 hours on 13th October.
20^00 (451): Stopped on Station 451 from c. 2000 to 0000 hours.
B. P^Jof^rCperature chart for October, including ^Je t.J in A w^^^^^^^^
rrtte-^s^trorie^^^^^^^^^ -P-- -' -'«- --''-'
from Plate^I, and the 2° isotherm lies on the mean pos.t.on of the convergence^^^
A.C. (No. 33): Observed position of the Antarctic convergence (No. 33 m lable 9, p. 20OJ.
('25): Meteor observations, 1925.
('26): 'Wi.lliam Scoresby' station, 1926.
Antarctic water, but they arc of some assistarrce although they fill no important gaps in *e Discovery
material For tkc other tabic (pp. 65-186) the Atlantic Ocean ts div.dcd into areas enclosed by ro
Tnati udc and longitude, and these are subdivided into 1° areas. Data from many sources are mcluded
aid o far as the data permit, the average temperature is given for each f area for each momh. Some
of A e a e n Antarctk waters, bu, it seems that such averages, compiled from m.seellancous sources,
1 u d be t Led with caution. B6hnecke drew charts of the Atlantic Ocean showmg the surface
'9^ DISCOVERY REPORTS
isotherms for each month (bound separately in an ' Atlas '), and although these no doubt give a correct
picture of the general distribution of Atlantic surface temperature, some features of the isotherms
south of about 50 S (e.g. o and ^ 1° in March) are certainly not compatible with the Discovery data
In his later paper (1938) Bohnecke gives a table of surface temperatures in 2° squares
The Norwegian Meteorological Institute published a series of records (1935) collected by whaling
factories for the International Polar Year, 1932^3. These include readings of sea temperature which
are no doubt accurate, but the great majority were taken near the ice-edge and for our present purpose
do httle more than confirm that temperatures here are about or below -- iC. Some temperatures
recorded in the first part of the homeward voyages, however, are helpful, though these again do not
fill any important gaps. A limited number of observations between the ice-edge and the convergence
are to be found also in results of the ' Norvegia ' expeditions (see Mosby, 1933 1934)
The Australasian expeditions under Sir Douglas Mawson provide material in the Australian Sector
in some months in which the Discovery Committee's ships collected no data in that region Some of
the temperatures recorded in the earlier expedition (1911^:4) seem very high, and here again it is
TaZVI^T 1"!''^'''^'^ -^ f""y ^-1-ble; but those recorded in the Station List of the
B.A.N.Z.A.R. Expedition of 1929-31 are undoubtedly reliable (see Mawson, 1940; and Johnson, 1937).
METHOD OF DRAWING THE ISOTHERMS
The distribution of surface temperature is of course subject not only to annual variations, but also to
innumerable local and temporary variations and complexities. The object here is to draw isotherms
to show as nearly as possible the average distribution of temperature for each month. Thus Plates
II-XII show the temperatures which .^ould probably be found in certain positions at different
times of year. For example, in January the 2° isotherm crosses longitude 30° W in c^° 30' S This
means that according to past experience a ship arriving at that position about the middle of January
IS more hkely to find that sea temperature than any other. If a higher or lower temperature is found
It IS probable but not certain, that in that region the surface water is warmer or colder than usual, and
the degree of probability will depend mainly on the number of different years from which data near
to that position are available. Data from the preceding and succeeding months will be relevant as
well as from the month in question.
Since in some regions data are lacking, and in some other regions insufficient to indicate the extent
. of annual variations. Plates II-XII can only be a first approximation to the average distribution of
temperature The data are, however, very extensive, and the charts should give a more correct view
of the distribution ot temperature in these waters than any that have been published before
It IS obvious that the basic pattern of the isotherms must be a system of concentric rings, generally
representing a falling gradient from north to south. The rings, however, are distorted in various ways
by water movements and the configuration of the land and the sea floor. The gradient is not usually
a steady one, but apart from small fluctuations it is not often inverted on a substantial scale. Inspection
of the temperature charts ^ indicates that the highest temperatures generally prevail in February and
suggests that the lowest occur about September and October (see also Fig. 6, p. 190). In drawing the
isotherm charts I have assumed that any isotherm should be placed progressively farther south in
each month from October to February, and similarly farther north from February to September
A break in the average seasonal rise and fall of temperature is not perhaps impossible, but the assump-
tion that there is no such break seems justified in the absence of evidence to the contrary. There are
are rlftVero'l^Ts^thr "chti " ""' '°^ '""^ '^"^^'"^' ^'^"^^ °" ^^^'^'^ '^^ ^^^^ temperatures were plotted. Plates II-XII
DISTRIBUTION OF SURFACE TEMPERATURE i99
comparatively few records for July and August, but these months are evidently not much warmer than
September.
Before any isotherms were drawn the Antarctic convergence was drawn in its mean position on
each of the temperature charts, and the mean position of the pack-ice edge for the month in question
was added. The latter was traced from the original charts reproduced in the report on the pack-ice
(Mackintosh and Herdman, 1940, plates Ixxi-xcv), and no alterations were made except for one small
adjustment. Inspection of the sea temperatures in the Falkland sector in November and re-examination
of the ice records suggested that the mean position of the ice-edge here should be drawn a little farther
south. Accordingly the line of the ice-edge in November between 10 and 70° W has been moved
about 20-40 miles farther south.
The lines of the ice-edge and the convergence form very useful starting points in the drawing of the
isotherms. The line of the ice-edge can be taken as coinciding approximately with the isotherm of
- 1-5° C. (It does not of course always coincide for the sea temperature at the ice-edge is sometimes
above -1° in summer or below -1-5° in winter, but some allowance can be made for this.) From
Tables 5 and 7 (pp. 189 and 192) we can calculate which isotherms should lie on the line of the con-
vergence. If these steps are accepted then the intermediate isotherms must lie at intervals withm a
limited belt.
The isotherms are drawn at intervals of 1° C, and in the first place were sketched in pencil on the
temperature charts. Those falling on the convergence were entered first, and they were derived from
Tables 5 and 7. For example, in April in 60° W (see Plate VIII) the middle temperature by Table 5
is 4-0^^ and the range by Table 7 is 2-6°. This gives a gradient from 27 to 5-3° which includes the
isotherms of 3, 4 and 5°. In 50" W, however, the gradient is from 3-2 to 5-0° which includes only the
4° and 5° isotherms. The f isotherm must therefore diverge away from the line of the convergence
between 60 and 50'' W. In this way the isotherms were drawn along the whole length of the con-
vergence for each month. 1 At some points (e.g. 10° E) the average range of the gradient is less than
1° C and the line of the convergence cannot accommodate more than one isotherm. Elsewhere the
convergence can be distinguished by the concentration of two or more isotherms.
The isotherms between the ice-edge and the convergence, and one or two also in sub-Antarctic
water north of the convergence, were then sketched in on the temperature charts. With data from several
different years included in a chart for one month, it was at once obvious that any attempt to join all
points of a given observed temperature would lead to absurdities. However, such points of equal
temperature were connected as often as possible provided this did not involve any improbable deviation
from the general trend of the adjacent isotherms and of the lines of the ice-edge and convergence.
If in any locality observations were available from only one year it was assumed that they represented
average conditions, unless observations from that and other years in adjacent regions indicated any
departure from the average. Where observations from different years suggested different latitudes for
the isotherm the line was drawn as nearly as possible in the mean position. At many points of course
it was difficult to decide where to place an isotherm, and even where observations were plentiful the
isotherms could be drawn with much more confidence at some points than at others. However, they
were drawn so as to make as good a fit as possible with the observed temperatures. Fig. 8 B (p. 197)
is an example of a fairly good fit. In this area the October temperatures for 1930 and 1938 corresponded
almost exactly In 1925 and 1926 in October and in 1936 in September the water was a little colder,
at least in the northern part of the area. September and October temperatures in 1938 were very
''whereas the majority of isotherms were in this way drawn by eye on the charts of observed
1 But see below with regard to special adjustments at one or two points.
200 DISCOVERY REPORTS
temperatures, the isotherms on the Hne of the convergence were derived from Tables 5 and 7. Ahhough
both are dependent for the most part on the same body of data this might have resuhed in some
inconsistencies, for Table 5 is no more than approximately correct, and as noted on p. 191 it is not
certain that it is equally applicable to all longitudes. Furthermore, the position of the isotherms not on
the convergence are in places influenced by extraneous data. However, the relative positions of
isotherms on the convergence and of those to the south of it were generally found to be quite con-
sistent, and only at one or two points was a compromise necessary. Thus in 150-170° E the 2° isotherm
in December and the 3° isotherm in January are placed a little south of the convergence, although they
should just be on it according to Tables 5 and 7, and a similar adjustment has been made in 60° W
in January.
Tentative isotherms were thus drawn for all months and for all regions except for the major areas
in which observations were lacking, and the next step was to compare single isotherms month by
month. For this purpose the isotherms of one temperature, for example all the 0° isotherms, for each
of the months October to February (i.e. coldest to warmest months) were traced on to one chart, as in
Plate XIII. It is to be supposed that during the whole of this period the sea temperature is on the
average rising, and that therefore the 0° isotherm (or any other) will lie farther south in each successive
month. This was found to be so, the 0° isotherms forming a succession of concentric rings which did
not touch or intersect each other except at one or two points. Such instances were assumed to be
due to abnormal conditions or annual variations, and the positions of the isotherms were adjusted
accordingly. For example, it was found at one point that the November isotherm, depending mainly
on temperatures in 1926, ran for a short distance a little north of the October isotherm which depended
on temperatures in 1936. Further inspection of the temperature charts indicated that in an adjacent
region the spring temperatures in 1926 were rather lower than in several other years. Therefore where
the overlap occurred the November isotherm was adjusted so as to lie a little south of the October
isotherm. The fact that such adjustments were few, and involved only small changes in the position
of the isotherms, suggested that they had been drawn not far from the correct positions. Some other
adjustments were made so that isotherms of adjacent months should be approximately parallel and
show similar features ; but this was only done where it involved small alterations which were not too
inconsistent with the observed temperatures. Sometimes the form of an isotherm changes considerably
from one month to the next (cf. the 0° isotherms for December and January between 30° W and 30° E,
Plate XIII).
One isotherm having been adjusted in this way, it was traced back in its new form on to the tem-
perature charts and the process was repeated for the other isotherms for September-February, and
then for all isotherms for February-September (the period of falling temperatures). It amounts of
course to a means of supplementing the data for one month with the data for other months on the
assumption that there is an unbroken rise of temperature from October to February and an unbroken
fall from February to September (see p. 198).
The revised isotherms were checked against the temperature charts in case of any unjustified
departures from the observed temperatures, and were then transferred from the original semicircular
charts to the circumpolar charts of a rather smaller scale which are reproduced in Plates II-XII. It
remained to fill in the isotherms in the regions in which no observations had been made (mainly in
the winter months). These were sketched on the circumpolar charts in what appeared to be the probable
positions by analogy with the isotherms already drawn. Each isotherm was then traced as before on
to a separate chart, adjusted with the same isotherms for other months, and traced back again. These
isotherms are shown as pecked lines in Plates II-XII. Often it was a matter of simple interpolation.
For example, between 80 and 110° W the 0° isotherms for November and January are based on
DISTRIBUTION OF SURFACE TEMPERATURE 201
observed temperatures, but there were no data here for December. The December isotherm, however,
must he between those for November and January, and since the observed temperatures nearly
everywhere indicate that it Hes nearer to the November than to the January isotherm there was no
difficuhy in deciding where to draw it here (see Plate IV). A good interpolation of this kind may some-
times give a more reliable position for the isotherm than a limited set of direct observations. Farther
west the November and December isotherms are both interpolated between those for October and
January, and in the winter months in the Pacific sector the 0° isotherms for five or six consecutive
months have to be interpolated (see Plate XIV). Here of course there is less certainty, especially as
the September isotherm rests on observations only from one year (1932). Even so there is not very
much choice in the positions in which they might be drawn.
NOTES ON THE DISTRIBUTION OF TEMPERATURE
It will be realized that Plates II-XII are semidiagrammatic charts which represent only the major
features of temperature distribution. Surface temperatures are naturally subject to some appreciable
widespread variations and to more numerous local variations ; and if the isotherms could be drawn in
full detail for any area at a given time they would usually
reveal elaborate complexities, and bear little resemblance
to the smooth lines in Plates II-XII. Fig. 9 shows the
apparent actual distribution of temperature between
South Georgia and the Falkland Islands in September
and October 1934, and may be compared with Fig. i
(p. 182) and with the corresponding region in Plate II.
The isotherms are correctly placed where they cross the
ship's tracks, and between these tracks they are drawn
in what seem to be the most probable positions ; but
there is not much doubt that, if additional lines of
observations could have been made at the same time, a
still more complex picture would be presented.
co-
Fig. 9. Observed temperatures in the Scotia Sea,
1st September to 14th October, 1934, showing the
apparent positions of the isotherms. Half-degree iso-
Most charts of temperature distribution as observed therms are shown as pecked lines, and the ship's tracks
at a particular time would be likely to show some features,
such as tongues of warm or cold water and other bends in the isotherms, which are temporary, and
which become smoothed out when the isotherms are drawn in their mean positions. Plates II-XII
are intended to show only those features which seem to be constant, or which at least are found in
most of the years for which observations are available. Even such constant features have been drawn
with caution, and it may be that some of them are not sufficiently emphasized.
The following points may be noted. South of the Falkland Islands in about 55^ W there is some
evidence of a northward thrust of cold water. This is noticeable in the chart for September and October
(Plate II) where the isotherms are crowded up to the convergence. For this chart there are data only
for one year (see Fig. 9), but the bend in the convergence here and the similar shape of the isotherms
in other months suggest that this is a constant feature. It is perhaps more noticeable in spring than m
summer. A more conspicuous tongue of cold water projects northwards around the South Sandwich
Islands (near 30° W). This is often very pronounced, and it is probably connected with a comparatively
narrow thrust of cold water off the north-east side of South Georgia, which was found during intensive
observations around the island. Warmer water is found to the east of the South Sandwich Islands,
and it is possible that the southward bend of the isotherms about 10-20° W is not sufficiently emphasized
in the charts. It is best seen in the charts for March and April, but it is not certain that it is really
202 DISCOVERY REPORTS
more conspicuous in these months than at other times of year. It seems probable that further examina-
tion of such features as these will show that they are correlated with the effects of bottom topography.
Between about io° W and 30° E the positions of the isotherms are very much influenced by the
eastward flow of cold water from the Weddell Sea, which also has an important effect here on the
distribution of pack-ice in the early summer (see Mackintosh and Herdman, 1940, p. 293). In winter
and spring the isotherms tend to be pressed up towards the convergence which here lies about its
lowest latitude, but in December, while pack-ice and cold water persist in a relatively low latitude
between o and 30° E, east of 30° both ice-edge and isotherms bend far to the south, and the ice belt
begins to break up internally, leaving an outer zone which contracts towards the South Sandwich
Islands (compare the ice-edge in Plates IV and V, December and January). As the outer zone of ice
60 ■
S-6 -
40-
3-0-
2-D-
1-0-
00-
-l-O-
1 1 1
52 53 54
1
55
1
55
1
57
1
53
1
59
1
60
1
61
1
62
I
63
1
64
I
65
t
6S
\
67
1
6B
DcGRCES SOUTH
Fig. 10. Comparison of average monthly temperatures, based on monthly isotherms.
melts away it tends to leave a long tongue of cold water which is seen as a conspicuous S-shaped turn
in the isotherms in Plates V-VII. The shape of the isotherms here is subject to a good deal of variation.
The data indicate that a line of observations running south from the convergence will usually, but not
always, reveal a slight rise in temperature about 60-65° S, and if there is no actual rise in temperature,
there will at least be a long expanse of ocean in which the temperature at the surface will change very
little. The existence of a belt of relatively warm water here, which melts the ice in a high latitude while
a zone of pack-ice still persists to the north of it, is noted by Deacon (1937, pp. 18, 28, etc., and Fig. 8),
who describes it as an ill-defined divergence region with an upwelling of warm water between the
Weddell Sea current moving eastwards in a lower latitude, and the current moving westward in a
higher latitude near the continental coast. Deacon also refers (p. 30) to the outer and inner belts of
pack-ice, and adds : ' It is just possible that even in winter there may be open water between the two
ice-streams in the eastern part of the Atlantic Ocean.'
203
DISTRIBUTION OF SURFACE TEMPERATURE
Meridional lines of observations in this region in April are only available for one year (1938). No
warm belt was then distinguishable, and the isotherms in Plate VIII are drawn accordingly, t is no
perhaps certain that the S-shaped bend in the isotherms is normally smoothed out in Apri bu it
may be that the advance of winter conditions has by then obliterated any signs of a warm belt at the
The loop in the convergence in 30° E is referred to on p. 184. It is possible that more of the adjacent
isotherms are involved in this disturbance than the charts suggest, but more data would be needed
to decide this point. . , ^^^^^„t
The course of the isotherms in the Indian Ocean and Australian regions call for no special comment.
The charts are self-explanatory, and it will be seen that more data are needed in some months.
The Ross Sea is another locality in which open water is found in a high latitude while pack-ice still
lies to the north. The isotherm for - 1° should perhaps reappear here in January, February and March,
but there is little material to show in what shape it should be drawn, and it is therefore omitted.
4 0-
3-0^
u 2-0-
o
kJ
a:
D
5G S
I 0-
0.
5
u
1-0-0-
-I 0-
, I I I I I I I I I I I I I ' '
SEPT OCT NOV DEC JAM FEB MAR APR MAY JU^.E JULY AUG SEPT OCT NOV DEC
Fig. II. Seasonal rise and fall of temperature, derived from Fig. 10.
In the Pacific sector the northward extension of cold water between 120 and 160° W, and the com-
pression of S^ isotherms in the winter and spring months in the Bellingshausen Sea and Drake Strait
^"XTs^e^^e^^^^^^^ of the temperature m different months and
diff nf iXdes. For Fig. :o the latitude of each isotherm in each month ^^^--f-]^
II-XII at every 30^^ of longitude, and the average was plotted. For example, in Plate VI (Februa^^
tL avLag lal^ of the 0° isotherm was found to be 65° 40' S. Thus m Fig xo the curve f r
the average latitu isotherms are correct the diagram should
'^eTrect ii; esl ^Z^^. i. ave.ge .e.pe..u.e between .he differen. .on.hs and
o ,he Ze of the temperature gradient south of the convergence. The slope of the gradient on the
cllrgence and to the north of it, and differences in the shapes of the monthly curves are perhaps
notve^ rel able The figure shows that surface temperatures are much the same m January and March,
: il Nov mber and j!ly, that south of the convergence the water is warmer „t December than m
T ,! b,„ if wou d OTobably not be justifiable to infer much more than this,
'"pt ;'is de W dtom' FTg .0, the teripera.ure for each month being plotted at arbitrary m.ervals
of Lfitude Ttflattening of the curves at the foot of the diagram indicates that the water ,s (on the
204 DISCOVERY REPORTS
average) covered by pack-ice in the months and latitudes in question. The figure suggests that when
the sea becomes free of ice in the early summer, the rise in temperature (December-January) is more
rapid in high than in low latitudes. Here again, however, the curves can only be taken as a rough
indication of the rise and fall of temperature.
REFERENCES
BoHNECKE, G., 1936. Temperatur, Salzgehalt wid Dichte an der Oberfldche des Alla7itiscben Ozeans. Erste Liefcrimg. Wissen-
schaftliche Ergebnisse der Deutschen Atlantischen Expedition auf dem Forschungs- und Vermessungsschiff 'Meteor',
1925-7, V, pp. 1-186, and Atlas.
BoHNECKE, G., 1938. Tetnperatur, Sahgehall und Diclite an der Oherfldche des AtlanHschen Ozeans. Ziveite Lieferung.
Wissenschaftliche Ergebnisse der Deutschen Atlantischen Expedition auf dem Forschungs- und Vermessungsschiff
'Meteor', 1925-7, v, pp. 187-250.
Deacon, G. E. R., 1933. A general account of the hydrology of the South Atlantic Ocean. Discovery Reports, vii, pp. 171-238,
pis. viii-x.
Deacon, G. E. R., 1937. The hydrology of the Southern Ocean. Discovery Reports, xv, pp. 1-124, pis. i-xliv.
Hart, T. J., 1934. On the phytoplankton of the South-West Atlantic and the Bellingshausen Sea, 1929-31. Discovery Reports,
VIII, pp. 1-268.
Hart, T. J., 1937. Rhizosolenia curvata Zacharias, an indicator species in the Southern Ocean. Discover}' Reports, XVI,
pp. 413-46, pi. xiv. . „ r r
Hart, T. J., 1942. Phytoplankton periodicity in Antarctic surface waters. Discovery Reports, xxi, pp. 261-356.
Hastings, .'\. B., 1943. Polyzoa {Bryozoa). Discovery Reports, xxii, pp. 301-510, pis. v-xiii.
John, D. D., 1936. The southern species of the genus Euphausia. Discovery Reports, xiv, pp. 193-324.
Johnson, T. H., 1937. Biological organisation and Station List. B.A.N.Z. Antarctic Research Expedition, 1929-31. Reports,
Ser. B, I, pt. 1, pp. 1-48.
LuMBY, J. R., 1928. Modification of the surface sampler with a vino to the improvement of temperature observation. Journal du
Conseii, iii, 3, pp. 340-50.
Mackintosh, N. A., 1934. Distribution of the macroplankton in the Atlantic sector of the Antarctic. Discovery Reports, ix,
pp. 65-160.
Mackintosh, N. A., 1937. The seasonal circulation of the Antarctic tnacroplankton. Discovery Reports, xvi, pp. 365-412.
Mackintosh, N. A. and Herdman, H. F. P., 1940. Distribution of the pack-ice in the Southern Ocean. Discovery Reports,
XIX, pp. 285-96, pis. Ixix-xcv.
Mawson, D., 1940. Ilydrological Observations of the Australasian Antarctic Expedition, 1911-14. Sci. Reps. Ser. A, I, pt. 4,
pp. 103-25.
Meinardus, W., 1923. Meteorologische Ergebnisse der Deutsche Siidpolar-Expedition, 1901-1903. Deutsche Siidpolar-
Expedition. HE Meteorologie, Bd. I, Heft 1.
Meteorological observations made on 9 Norivegian ■whaling floating factories during the International Polar Year 1932-1933.
Publications from the International Polar Year 1932-33, No. 1. Det Norske Meteorologiske Institut. Oslo, 1935, pp.
1-53-
MosBY, H., 1933. The sea-surface and the air. Scientific Results of the Norwegian Antarctic Expeditions, 1927-1928 et sqq..
No. 10. Det Norske Videnskaps-Akademi i Oslo, pp. 1-140.
MosBY, H., 1934. The waters of the Atlantic Antarctic Ocean. Results of the Norwegian Antarctic Expeditions, 1927-1928
et sqq.. No. 11. Det Norske Videnskaps-Akademi i Oslo, pp. 1-131.
Neaverson, E., 1934. The sea-floor deposits: 1. General characters and distribution. Discovery Reports, ix, pp. 295-350,
pis. xvii-xxii.
Norman, J. R., 1938. Coast Fishes. Part III. The Antarctic zofie. Discovery Reports, xviii, pp. 1-104, P'- i-
SvERDRUP, H. v., Johnson, M. W. and Fleming, R. U., 1942. The Oceans. Their physics, chemistiy and general biology,
pp. 1-1087. New York.
•
205
A.PPENDIX
Table 9.
r/(e Antarctic convergence. List
of all occasions oft which the convergence has been crossed
or located by ships of the
Discovery Committee
Abbreviations :
adj. adjacent
ship
s T.s Routine 4-hourIy temperatures (Fahrenheit) entered in ship's log
approx. approximate
St.
Station
betw. between
surf.
surface
D.R. dead reckon!
ng
temp. temperature
hr. hours
W.S
'Will
lam Scoresby'
min. minimum
Note. Where the hour is noted in the ' Remarks
' column
it refers to
ship's time on the date in the 'Date' column, and is given mainly
to facihtate reference back to the original data
.
Degree
No.
Ship
Voyage
Date
Estimated position
of
Remarks
accuracy
I
' Discovery '
Tristan da Cunha to
14-
ii. 26
47° IS' S,
25° 20' W
Probable
By min. temp. betw. Sts. 10 and 11. By
S. Georgia
14.
ii. 26
48° 05' S,
27° 05' w
»)
thermograph crossing loops of the con-
IS-
ii. 26
49° 10' S,
28°35'W
»j
vergence at 0800 and 2000 hr. on 14th,
IS-
ii. 26
49° 20' S,
28°55'W
»»
1500 and 1800 hr. on 15th, and 0100,
16.
ii. 26
49° S3' S,
29° 40' W
»»
0400 and 1700 hr. on i6th, the last
16.
ii. 26
50° 10' S,
30° 00' W '
crossing being the most definite.
16.
ii. 26
5°°3S'S,
30^ 45' W Good
2
»>
S. Georgia to
Falkland Is.
20.
iv. 26
52° 10' S,
43° 00' W
Uncertain
No adj. Sts. Thermograph not clear;
probably 0200 hr.
3
'W. Scoresby'
Cape Town to
S. Georgia
28.
X. 26
47° 00' S,
06° 45' w
Probable
Betw. Sts. WS II and 12. Ship's T.s
indicate about 1800 hr.
4
' Discovery '
>>
6.
xi. 26
47° 10' S,
14° 10' E
Uncertain i No adj. Sts. Thermograph suggests 1800 1
\ ' hr., but doubtful.
5
n
S. Georgia to
S. Orkneys
8.
ii. 27
S4° 35' S.
42° 50' w
Approx.
Thermograph rises to over 4° C at 1600
hr., indicating proximity of a loop of
the convergence.
6
'W. Scoresby*
S. Georgia to
Falkland Is.
20.
ii. 27
S3°i9'S,
45^ 40' w
Good
By min. temp. betw. Sts. WS 67 and 68.
By ship's T.s about 1500 hr.
7
* Discovery *
S. Shetlands to
C. Horn
18.
iv. 27
58° 28' S,
67°SS'W
Approx.
By min. temp, about St. 216. By thermo-
graph about St. 217.
8
*W. Scoresby'
Falkland Is. to
S. Georgia
13-
v. 27
52° 30' s,
48°i5'W
Approx.
No. adj. Sts. By ship's T.s about 1800 hr.
Position by D.R.
9
»»
S. Georgia to
31-
v. 27
49° 12' S,
35°5o'W
Probable No temps, at Sts. Ship's T.s show marked
Gough I.
I.
vi. 27
48° 40' S,
34° 00' W
Uncertain rise about 1800 hr. on 31. v. 27, and
I.
vi. 27
47° 47' S,
3i°34'W
))
indications of loops at about 0200 and
1500 hr. on I. vi. 27.
10
)»
Falkland Is. to
S. Georgia
II.
ii. 28
S3° 17' S,
47° 02' W
Approx.
By min. temp, about St. WS 140. St.
showed surf, cooled by nearby ice island
and brash ice.
II
)»
S. Orkneys to
26.
iv. 28
S6° 27' S,
54° 22' W
Good
By min. temp, about St. WS 204. Ship's
Falkland Is.
27.
iv. 28
56° 24' s,
S4° 30' W
Uncertain
T.s show rise at St. WS 204 and indicate
27.
iv. 28
56° 12' S,
55" 00' w
)»
loops about 0800 and 1400 hr. on
27. iv. 28.
12
))
Falkland Is. to
22.
viii. 28
52° so' S,
47° 40' W
Good
By min. temp. betw. Sts. WS 253 and 254.
S. Georgia
By ship's T.s about 0600 hr.
13
)»
S. Georgia to
Falkland Is.
4-
xii. 28
52°4S'S,
48° 25' W
»>
By min. temp. betw. Sts. WS 316 and
317. By ship's T.s about 0200 hr.
14
J)
Falkland Is. to
S. Georgia
14.
xii. 28
S2° 30' S,
49° 00' W
**
No. adj. Sts. Ship's T.s show clear con-
vergence at 1600 hr.
15
t»
S. Shetlands to
C. Horn
23-
ii. 29
58°2i'S,
65°37'W
»»
By min. temp. betw. Sts. WS 404 and 405.
No clear indication by ship's T.s.
16
1)
Falkland Is. to
S. Georgia
iS/i
5. iii. 29
.?
—
No temps, at Sts. Ship's T.s indeter-
minate.
17
)>
S. Georgia to
Falkland Is.
30-
iv. 29
52° 40' S,
48° 09' W
Probable
By min. temp, near St. WS 429, but ship's
T.s indicate on St. WS 430.
18
j»
Falkland Is. to
S. Georgia
7-
V. 29
52°4S'S,
46° 20' W
Approx.
No. adj. Sts. Ship's T.s show conver-
gence, but D.R. positions doubtful.
19
))
S. Georgia to
19.
v. 29
50° 55' s,
29° 53' w
Uncertain
By min. temp. betw. Sts. WS 435 and 437.
Cape Town
19.
V. 29
So° 13' S,
27' 40' W
jj
Ship's T.s suggest running parallel to
convergence between positions shown
(1200-0000 hr.).
206
DISCOVERY REPORTS
Table 9
[cont.)
■
•
Degree
No.
Ship
Voyage
Date
Estimated position
of
Remarks
accuracy
20
'W. Scoresby'
Cape Town to
19/27. X. 29
47° 12' S,
07° 50' w
Uncertain
No informative Sts. Ship's T.s showing
S. Georgia
52° 15' s,
28° 20' w
>»
fluctuating temps, betw. positions
shown. 1800 hr. on 19. x. 29 to 0200 hr.
on 27. X. 29.
21
»»
S. Georgia to
Falkland Is.
2.
xi. 29
52^34'S,
48° 48' w
»)
No temps, at Sts. Ship's T.s suggest at
or near St. WS 466.
22
»
Falkland Is. to
S. Shetlands
II.
xi. 29
57° 16' S,
57° 14' W
Approx.
By min. temp. betw. Sts. WS 469 and 470.
Ship's T.s fluctuating.
23
>>
S. Shetlands to
Falkland Is.
9-
i.30
57° 50' S,
59° 40' W
)»
No adj. Sts. Ship's T.s clearly indicate
0600 hr., but latitude doubtful.
24
' Discovery II '
Montevideo to
i5(?)-i-30
?
—
No adj. Sts. No indication by thermo-
S. Georgia
graph.
25
'W. Scoresby'
Falkland Is. to
S. Georgia
16.
i. 30
52°45'S,
44°35'W
Probable
No adj. Sts. Ship's T.s suggest 0200 hr.
26
i»
S. Shetlands to
Falkland Is.
20.
ii. 30
56°47'S,
58°i5'W
Approx.
No adj. Sts. Ship's T.s indicate 0200 hr.,
but positions not certain.
27
>>
NW'ward in 55° W
3-
iv. 30
55°45'S,
54° 50' W
Good
By min. temp. betw. Sts. WS 529 and
530. Ship's T.s indicate 0200 hr.
28
>»
E'ward in Scotia Sea
6.
iv. 30
55°28'S,
54° 30' W
Approx.
Noadj.Sts. Byship'sT.sat iSoohr. Sub-
sequently probably running parallel to
convergence.
29
'Discovery II'
S. Shetlands to
C. Horn
14.
iv. 30
59° 15' S,
64° 10' W
V. good
By min. temp, near St. 384. By thermo-
graph at 2330 hr.
3°
>i
C. Horn to
18.
iv. 30
55°47'S,
52° 50' W
Good
Surf, temps, only at Sts. By thermograph
S. Georgia
18.
iv. 30
55°4o'S,
5i°oo'W
J»
clearly running parallel to convergence
19.
iv. 30
55° 10' S,
47° 20' W
)»
and crossing loops at 0830 and 1900 hr.
on i8th, and 0530 on 19th. Some addi-
tional fluctuations of temp. E of 47° W.
31
'W. Scoresby'
S. Georgia to Rio.
?.
V. 30
?
—
No adj. Sts. Ship's T.s indeterminate.
32
'Discovery II'
S. Georgia to
Cape Town
13-
v. 30
48° 21' S,
2i°55'W
V. good
Surf, temps, only at Sts. Sharp rise in
temp, at 1800 hr. Temp, rather high
for May.
33
>»
Cape Town to
Bouvet I.
14.
X. 30
48° 20' S,
10° 00' E
Good
By min. temp. betw. Sts. 451 and 452.
Thermograph shows slight but distinct
fall at 0900 hr.
34
'W. Scoresby'
Montevideo to
S. Georgia
9-
j-3i
50° 26' s,
4i°24'W
Approx.
No adj. Sts. Ship's T.s show changes at
0600 and 1800 hr.
35
'Discovery II'
W'ward from
S. Georgia
4-
iii. 31
54° 05' S,
46 ' 00' W
V. uncer-
tain
Surf, temps, only at Sts. Thermograph
indeterminate but probably crossed
36
convergence at 0430 hr.
yt
S'ward in Scotia Sea
6.
iii. 31
55° 40' S,
5i°3o'W
>j
By min. temp, near St. 633. Thermograph
indeterminate. Crossing possibly at
0500 hr.
37
»»
S. Shetlands to
12.
iii. 31
58" 00' S,
60° 10' W
V. good
By min. temp. betw. Sts. 648 and 649. By
38
Staten I.
thermograph at 1930 hr.
»>
Falkland Is. to
24.
iii. 31
53° 17' S,
47°35'W
Probable
By min. temp. betw. Sts. 656 and 657. By
S. Georgia
thermograph at 1200 hr.
39
'W. Scoresby'
S. Georgia to
Falkland Is.
13-
iv. 31
52° 30' s,
47°45'W
Uncertain
No. adj. Sts. By ship's T.s about 0200 hr.
40
'Discovery 11'
N'ward in 30° W
18.
iv. 31
49° 50' S,
29°55'W
V. good
By min. temp. N of St. 666. By thermo-
41
1}
S'ward in 75° W
20.
xi. 31
59°47'S,
75°o5'W
Approx.
graph at 0900 hr.
By min. temp. betw. Sts. 730 and 731.
21.
xi. 31
62° 05' S,
75°oi'W
)»
Thermograph indicates confused loops
21.
xi. 31
62° 53' s,
75° 02' W
)»
with sharper fluctuations farther S.
Times adopted: 1400 hr. on 20th, 1330
and 1930 hr. on 21st.
42
>)
N'ward in 56° W
28.
xi. 31
56° 50' s,
55° 50' W
V. good
7 J
By min. temp. betw. Sts. 745 and 746. By
43
»»
S'ward in 49° W
3-
xii. 31
55° 50' S,
49° 02' W
»»
thermograph at 1700 hr.
By min. temp. betw. Sts. 754 and 755.
4-
xii. 31
56° 10' S,
48° 58' w
>»
Sharp convergence at 1845 hr. on 3rd,
4-
xii. 31
56° 30' s,
48° 56' w
»>
and well-marked loop or patch of sub-
Antarctic water at 0100 and 0300 hr.
on 4th.
THE ANTARCTIC CONVERGENCE
Table 9 (cont.)
207
Ship
44
45
46
47
'Discovery II'
48
49
5°
SI
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Voyage
North of S. Georgia
»> )»
Date
S. Georgia to
Falkland Is.
Falkland Is. to
S. Georgia
S. Georgia to
Cape Town
SE'ward from Cape
Town
NE'ward to
Fremantle
SE'ward from
Fremantle
NE'ward to
Melbourne
S'ward from
Melbourne
N'ward to
N. Zealand
SE'ward from
N. Zealand
NE'ward betw.
150° and 140° W
SE'ward in ii3°W
NE'ward to
Magellan Str.
S'ward in 80° W
S. Shetlands to
Falkland Is.
S'ward in 45° W
North of S. Georgia
S. Shetlands to
Falkland Is.
Falkland Is. to
S. Georgia
N'ward in 13° E
16. xii. 31
17. xii. 31
10. ii. 32
20. ii. 32
Estimated position
Degree
of
accuracy
Remarks
25. n. 32
25. ii. 32
25. ii. 32
25. ii. 32
26. ii. 32
14. iv. 32
50^48'S, 37^22'W
49"4o'S, 37'15'W
52^40'S, 43'30'W
52'=So'S, 43^15'W
49°45'S, 24°03'W
49° 25' S, 23° 00' W
48°S5'S, 2i°35'W
48°35'S, 2o°35'W
48° GO'S, i8''5o'W
5o°25'S, 3i°3o'E
Approx.
Good
I. V. 32
23- V. 32
31- V. 32
20. vi. 32
27. vi. 32
7. ix. 32
13. ix. 32
14. Lx. 32
14. ix. 32
26. ix. 32
30. ix. 32
28. x. 32
9. xi. 32
18. xi. 32
4. xii. 32
4. xii. 32
7- ii- 33
21. ii. 33
20. iii. 33
Probable
Good
By min. temp, near St. 776. By thermo-
graph near St. 775.
By min. temp, near St. 776. By thermo-
graph probably about 1 800 hr. , but some
indications of loops in the convergence.
No. adj. Sts. By ship's T.s about 1 100 hr.
By min. temp. betw. Sts. 829 and 830. By
thermograph at 0600 hr. betw. Sts. 830
and 831, showing surface stratum of
sub-Antarctic water drifted east of nor-
mal position. Surface position adopted.
Cf. No. 46.
Surf, temps, only at Sts. By thermograph
crossing loops near St. 835 at 0300,
0730, 1500, and 1900 hr. on 25th, and
0200 hr. on 26th.
52° 00' S, 95° 50' E Probable
52° 07' S, 121° 35' E
54° 25' S, 136° 00' E
53° 45' S, 151° 33' E
54° 15' S, 151° 55' E
55° 00' S, 152° 00' E
54° 30' S, 162° 35' E
57° 50' S, 165° 35' w
55° 30' S, 144'^ 55' W
53° 45' S, 141° 45' w
53° 05' S, 140° 50' W
56° 55, S, 113° 25' W
61° ID'S,
60° 20' s,
60° 10' s,
62° 45' s,
55° 17' S,
5i°2o'S,
49° 35' S,
50° 28' S,
93° 20' W
9i°i5'W
90° 45' W
80° 00' w
56° 00' w
44° 41' W
36°35'W
36° 57' W
58°2o'S, 6o°35'W
52°i3'S, 47°25'W
48°3o'S, i3°3o'E
By min. temp. betw. Sts. 849 and 850. By
thermograph at 2130 hr. Indication of
■ long northward bend in convergence
approaching ship's track betw. Sts. 848
and 849.
By min. temp. S of St. 866. By thermo-
graph at 1500 hr., but loop comes near
ship's track at 1800 hr. on 2nd.
Approx. By min. temp, near St. 883. By thermo-
graph about 1500 hr.
Good By min. temp. betw. Sts. 891 and 892. By
thermograph at 1000 hr.
By min. temp, near St. 903. By thermo-
graph at 0200 hr. with loop at 0630 and
1200 hr.
V. good By min. temp, near St. 920. By thermo-
graph at 0130 hr.
Probable By min. temp, near St. 949. By thermo-
graph probably at 1000 hr.
Good By min. temp. betw. Sts. 961 and 962. By
thermograph at about 0700 hr. Well-
defined loop of Antarctic water 0530
to HOC hr. on 14th betw. Sts 962 and
963-
Probable By min. temp. betw. Sts. 970 and 972. By
thermograph at oioo hr.
By min. temp. betw. Sts. 975 and 976. By
thermograph at 0330 hr., with loop of
Antarctic water at 1 130 and 1300 hr.
V. good By min. temp. betw. Sts. 990 and 991. By
thermograph at 0500 hr.
By min. temp. betw. Sts. 1017 and 1018.
By thermograph at 0500 hr.
Uncertain By min. temp, near St. 1027. By thermo-
graph indeterminate.
V. good By min. temp. betw. Sts. 1054 and 1055.
By thermograph at 0400 hr.
By min. temp. betw. Sts. 1055 and 1056.
By thermograph at 2300 hr. just S of
1056.
Approx. By min. temp. S of St. 1 1 17. By thermo-
graph probably at 1600 hr.
Uncertain By min. temp. E of St. 1 123. By thermo-
graph at 0400 hr.
By min. temp. S of St. 1162. Thermo-
graph suggests 0900 hr.
208
DISCOVERY REPORTS
Table 9 (cont.)
No. Ship
67
Voyage
Date
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
'Discovery II' Tristan to
S. Georgia
N'ward in 78° W
S'ward from
Falkland Is.
N'ward to Auckland
N. Zealand to
Ross Sea
N'ward in 79" W
Falkland Is. to
Elephant I.
NW of S. Georgia
Enderby Land to
Durban
Cape Town to
S. Georgia
SW'ward to
S. Georgia
S. Georgia to
S. Shetlands
N'ward in 78' W
Falkland Is. to
Elephant I.
N'ward in 46° W
NW of S. Georgia
S. Georgia to
Falkland Is.
S'ward in 79° W
E'ward betw. 105°
and 96° W
23-
24.
24.
25-
25-
i6.
•^'- 33
-^i- 33
xi-33
xi- 33
xi-33
-'^''- 33
Estimated position
47° 56' S, 23° 10' W
48° 30' S, 25° 15' W
49° 04' S, 26° id' W
49°48'S, 28"o3'W 1
49° 55' S, 28°34'W!
61° 40' S, 78" 00' vv
Degree
of
accuracy
Probable
29.
29.
^11- 33
■^>'- 33
29. xii. 33
29. xii. 33
29. xii. 33
29. xii. 23
29. xii. 33
3°- xii. 33
30- ^ii- 33
22. i. 34
19. n. 34
12. iii. 34
29. iii. 34
8. iv. 34
8. iv. 34
15- V. 34
II. viii. 34
21. viii. 34
22. viii. 34
23. viii. 34
23. viii. 34
4. ix. 34
5. ix. 34
13. ix. 34
26. ix. 34
55" 24' s,
55° 30' S,
55°48'S,
56° 25' S,
57° 20' S,
57'^ 40' S,
57° 58' S,
58° 20' S,
58° 45' S,
57°58'S,
60° 10' W
6o°i5' W
60" 28' W
60° 4s' W
6i°3o'W
6i°48'W
62° 05' W
62° 25' W
62° 50' W
170° 45' W
60° 47' S, 174" 50' w
63°3o'S, 79°o8'W
56° 35' S, 55° 40' W
5i°2o'S, 44" 09' W
5i°2o'S, 42°4o'W
50°20'S, 44°54'E
48'55'S, 04^15'W
47 4°'
48° 57'
49° 45'
49° 55'
55° 40'
56° 03'
61° 25'
56° 20'
I.
x-34
56° 06'
2.
x-34
55° 12'
3-
^•34
51° 55'
5-
x-34
51° 20'
12.
•^•34
52° 48'
29.
^•34
62" 10'
9-
xi. 34
61° 02'
10.
XI. 34
60° 15'
10.
XI. 34
60° 03'
II.
XI. 34
59" 35'
11.
XI. 34
59° 33'
II.
XI. 34
59" 40'
S, 23°56'W
S, 25°4o'W
S, 26°5o'W
S, 27'05'W
S, 44°2o'W
S, 45"2o'W
S, 78° 30 W
S, 55^40'W
S, 45°47'W
S, 45°45'W
S, 45°2i'W
S, 42°25'W
S, 47°57'W
S, 79'2S'W
S, 105" 05' w
S, 100° 45' w
S, 99°3o'W
S, 97°i5'W
S, 96°28'W
S, 96° 10' W
V. good
Uncertain
J)
Probable
V. good
Good
V. uncer-
tain
Good
Uncertain
Probable
Gt)od
))
Approx.
V. good
Good
Remarks
By min. temp. betw. Sts. 1190 and 1198.
Conditions confused at intermediate
stations. Loops probably crossed at
1500 hr. on 23rd, 0830 and 1330 hr. on
24th, and 0300 and 0530 hr. on 25th.
Course of convergence very confused.
By min. temp, near St. 1224. By thermo-
graph at 0400 hr.
By min. temp. betw. Sts. 1233 and 1234.
By thermograph a series of loops or
patches at 0030, 0200, 0330, 0730, 1400,
1600, 1800, 0100 and 0400 hr.
By min. temp. S of St. 1276. Thermograph
indeterminate, but 0800 hr. adopted
for position.
No. adj. Sts. Thermograph suggests 1200
hr.
By min. temp, near St. 13 15. By thermo-
graph at 0600 hr., but ship near con-
vergence until 0400 on 13th.
By min. temp. betw. Sts. 1325 and 1326.
By thermograph at 0400 hr.
By min. temp. betw. Sts. 1337 and 1338.
By thermograph at 0500 hr.
By min. temp. betw. Sts. 1338 and 1339.
By thermograph at 1700 hr.
By min. temp. betw. Sts. 1366 and 1367.
By thermograph at 0200 hr.
Surf, temps, only at Sts. Thermograph
indeterminate, but crossing perhaps at
0730 hr.
By min. temp, near St. 1391. By thermo-
graph sub-Antarctic water SW of 1391
but not SE of it. Crossings at 1400 hr.
on 2ist, about 2000 hr. on 22nd, and
0400 and 0530 hr. on 23rd.
Surf, temps, only at Sts. By thermograph
ship's track cuts loop of convergence
SW of S. Georgia at 1900 and 0300 hr.
By min. temp. betw. Sts. 1416 and 1417.
By thermograph about 1800 hr.
By min. temp. betw. Sts. 1424 and 1425.
By thermograph at 0700 hr.
Surf. temp, only at St. By thermograph
ship's track cuts loop of convergence
SW of S. Georgia at 1700 and 0500 hr.
By min. temp. betw. Sts. 1434 and 1435.
By thermograph at 2100 hr.
By min. temp. S of St. 1435. By thermo-
graph at 0200 hr.
By min. temp. betw. Sts. 1439 and 1440.
By thermograph at 0300 hr.
By min. temp. betw. Sts. 1446 and 1447.
By thermograph at 0500 hr.
By min. temp. St. 1467 very close to con-
vergence. By thermograph crossing
loops at 2100 hr. on 9th, 1230 and
1700 hr. on loth, and 0430, 0830 and
1330 hr. on nth.
THE ANTARCTIC
CONVERGENCE
209
Table 9
(cont.)
Degree
No.
Ship
Voyage
Date
Estimated position
of
Remarks
accuracy
88
'Discovery II'
N'ward in 80= W
17-
xi-34
60' 20' S,
79° 54' W
Uncertain
By min. temp, close to St. 1476. By
thermograph doubtful.
89
'W. Scoresby'
SE'ward from
Cape Town
xi./
xii. 34
?
No adj. Sts. Convergence doubtful by
ship's T.s. 30. xi. or i. xii. 34.
90
' Discovery II '
Falkland Is. to
4-
xii. 34
56" 05' S,
59° 00' W
V. good
No adj. Sts. By thermograph at 0030,
S. Shetlands
4-
xii. 34
56^ 55' S,
59° 20' w
})
0700 and 1430 hr. on 4th. Rise of temp.
4-
xii. 34
57' 54' S,
59° 45' W
)»
to 2-4° centred at 0800 hr. on 5th
5-
xii. 34
59° 47' S,
60" 20' W
Good
indicates proximity of a loop of the
convergence.
91
>>
N'ward in 44^ W
25-
i-35
52° 06' S,
44° 05' W
>»
By min. temp, at St. 1495. By thermo-
graph at St. 1495.
92
)»
NW'ward to Cape
Town
7-
"•• 35
49° 50' S,
3i°05'E
>»
By min. temp. betw. Sts. 1552 and 1553.
By thermograph at 0400 hr.
93 'W. Scoresby'
1
NE'ward to Cape
Town
24.
"i- 35
50" 57' s,
05°i5'E
Uncertain
No adj. Sts. By ship's T.s probably
about 1000 hr.
94
' Discovery II '
E'ward betw. 20°
2.
iv.3S
48° 50' s,
20= 45' E
Good
By min. temp. St. 1561 in Antarctic water.
and 35 E
5-
iv.3S
49° 35' S,
29° 07' E
j»
By thermograph crossings at 11 00 hr.
5-
iv-35
49° 30' S,
3i°oo'E
)»
on 2nd, 0500 and 1600 hr. on 5th, and
6.
iv.35
48° 40' S,
34° 10' E
Approx.
about 0900 hr. on 6th. Pronounced
bend in convergence about 30' E. Cf.
No. 49, 14. iv. 32.
95
>>
S'ward in 56' E
20.
xi. 35
48° 08' S,
56" 30' E
Uncertain
By min. temp, near St. 1618. By thermo-
graph probably 0800 hr. on 20th.
96
jj
NE'ward to
Fremantle
7-
xii- 35
52° 30' s,
117' 05' E
Probable
No adj. Sts. By thermograph at 1700 hr.
97
*W. Scoresby*
S'ward from Cape
Town
9-
xii- 35
49° 20' S,
18° 48' E
))
No. adj. Sts. By ship's T.s about 1600 hr.
98
* Discovery II '
N. Zealand to Ross
Sea
Balleny Is. to
6.
i.36
60" 22' s,
178° 15' E
)»
No adj. Sts. By ship's T.s about 2100 hr.
99
)>
8.
ii. 36
57° 20' S,
163° 00' E
Approx.
By min. temp. betw. Sts. 1679 and 1680.
Melbourne
By ship's T.s about 0600 hr.
100
)»
S'ward from
Melbourne
9-
iii. 36
53° 20' S,
146° 44' E
Good
By min. temp. betw. Sts. 1690 and 1691.
By ship's T.s near 1200 hr.
lOI
'W. Scoresby'
N'ward to Cape
29/30. iii. 36
?
—
No adj. Sts. Indeterminate by ship's T.s.
Town
102
'Discovery II*
N'ward to Fremantle
2.
iv. 36
48° 05' S,
109° 47' E
Good
By min. temp. N of St. 1730. By ship's
T.s near 1730 hr.
103
)>
S'ward in 0°
29.
V. 36
49° 37' S,
00' 06' E
Probable
By min. temp, near St. 1777. By ship's
T.s about 1400 hr.
104
)j
N'ward to Cape
Town
12.
vi. 36
48° 40' S,
18" 50' E
Approx.
By min. temp. S of St. 1798. By ship's
T.s about 1200 hr.
105
))
S'ward in 0"
26.
ix. 36
49° 54' S.
00^ 06' E
Uncertain
Bv min. temp. betw. Sts. 1810 and 1812.
By thermograph about 1200 hr.
106
>»
NW of S. Georgia
5-
xi. 36
5i°io'S,
41° 40' W
Good
By min. temp, near St. 1857. By thermo-
graph at 0400 hr.
107
»)
It yy »»
6.
xi. 36
52° 04' S,
43' 00' W
Approx.
By min. temp. N of St. i860. By thermo-
graph about 1600 hr.
108
)»
S'ward in 43° W
7-
xi. 36
54° 10' S,
42° 50' w
j>
Surf. temp, only at St. By thermograph
rise of temp, to 3-0° centred about
1200 hr. indicates proximity of loop of
convergence.
109
))
Elephant I. to
Falkland Is.
IS-
xi. 36
57° 08' S,
55°i5'W
V. good
By min. temp. N of St. 1877. By thermo-
graph at 0300 hr.
no
'W. Scoresby*
S'ward from Cape
Town
I.
xii. 36
47° 58' S,
i3°03'E
Good
No adj. Sts. By ship's T.s about 1800 hr.
III
'Discovery IF
E'ward to S. Georgia
3-
xii. 36
53°35'S,
48^ 30' W
Uncertain
By min. temp. E of St. 1916. By thermo-
graph about 0300 hr.
112
i>
S. Orkneys to
Falkland Is.
17-
ii-37
56°45'S,
5i°4o'W
V. good
By min. temp. S of St. 1969. By thermo-
graph at 1600 hr.
"3
)j
Falkland Is. to
S. Georgia
2.
iii- 37
53°25'S,
45° 3°' W
Uncertain
By min. temp, near St. 1975. By thermo-
graph about 0230 hr.
114
'W. Scoresby'
N'ward to Cape
Town
19.
iii- 37
48° 20' S,
ii°25'E
n
No adj. Sts. By ship's T.s perhaps at
2200 hr.
"5
'Discovery II'
N'ward in 0°
28.
iii. 37
49° 5°' S,
00° 20' E
Approx.
By min. temp. betw. Sts. 2022 and 2023.
By thermograph probably at 2300 hr.
2IO
DISCOVERY REPORTS
Table 9 (cont.)
No.
Ship
Voyage
Date
Estimated position
Degree
of
accuracy
Remarks
116
'W. Scoresby'
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
'Discovery II'
'W. Scoresby'
' Discovery II '
Whale marking betw.
Montevideo and
S. Georgia
SE'ward from
Cape Town
NE'ward to
Fremantle
S'ward from
Fremantle
N'ward to
N. Zealand
SE'ward from
N. Zealand
S. Shetlands to
Falkland Is.
NE'ward to
Falkland Is.
Falkland Is. to
S. Orkneys
N'ward to Cape
Town
S'ward in o
N'ward in 20' E
S'ward in o^
N'ward in 20^ E
S'ward in o"
N'ward in 20^ E
S'ward in 0°
N'ward in 20' E
S'ward in o"
N'ward in 20^ E
S'ward in o
N'ward in 20^ E
S'ward in o
N'ward in 20^ E
27- X. 37 47° 32' S, 33° 12' W
28. X. 37 ] 49° 00' S, 30° 32' W
30. X. 37 j 50° 04' S, 32° 10' W
5. xi. 37 j 48° 20' S, 33° 12' W
19. .XI. 37
10. xu. 37
5- i- 38
5- i- 38
6. i. 38
24. i. 38
14. ii. 38
24. ii. 38
47" 05' S, 2S° 00' E
52° 12' S, 98° 25' E
49° 47' S, 1 15° 47' E
50° go's, 115° 47' E
51° 12' S, 115° 50' E
59° 30' S, 165° 15' E
59' 52' S, 168^ 45' W
56°47'S, 58"oo'W
Probable
)?
Good
Probable
Uncertain
Probable
13. iii. 38 57° 57' S, 62" 15' W
22. iii. 38
29. iv. 38
29. iv. 38
30. iv. 38
10. vii. 38
23. vii. 38
14. viii. 38
27. viii. 38
23. ix. 38
3- X. 38
26. X. 38
5. xi. 38
2. xii. 38
14. xii. 38
16. i. 39
3- "• 39
23- ii- 39
15. iii. 39
56°3o'S, 52°i5'W
49° 45' S,
47° 50' S,
46° 28' S,
50° 12' S,
48° 20' S,
49° 28' S,
48° 30' S,
50° 00' S,
48° 55' s,
50° 06' S,
48" 10' S,
49° 45' S,
47° 55' S,
49° 45' S,
48" 20' S,
49° 33' S,
48° 20' S,
19° 22' E
19° 20' E
19° 18' E
00° 32' E
21° 00' E
00° 30' E
20' 30' E
00^"' 40' E
20° 00' E
02° 05' E
20" 15' E
01° 03' E
19° 55" E
02° 48' E
19° 45' E
oi°o6'E
i9°38'E
Uncertain
Probable
V. good
Probable
Good
Approx.
)»
Good
»j
Uncertain
Good
Uncertain
V. uncer-
tain
Uncertain
Good
Probable
V. uncer-
tain
Probable
Approx.
Good
Complex movements by ship and v. sharp
fluctuations of temp, by thermograph.
Loop of convergence approached at
0015 hr. on 27th. Crossings at 0600 hr.
on 28th, 1300 hr. on 30th and 0500 hr.
on 5th. Loops possibly shifted betw.
27. X. and 5. xi. 37.
By min. temp. N of St. 2088. By thermo-
graph at 0100 hr., S of St. Latter
position adopted.
Obscure by min. temp. By thermograph
probably at 11 00 hr.
By min. temp, near St. 2158. By thermo-
graph at 1500 and 1700 hr. on 5th and
0800 hr. on 6th. Track close to con-
vergence from 1700 to 0800 hr.
By min. temp, near St. 2206. Thermo-
graph shows rise in temp, before and
after St. 2206.
By min. temp. betw. Sts. 2220 and 2221.
By thermograph at 1330 hr.
No adj. Sts. By thermograph at 1000 hr.
By min. temp. betw. Sts. 2288 and 2289.
By thermograph at 0200 hr. Probably
oblique crossing.
By min. temp. betw. Sts. 2292 and 2293.
By thermograph at 0900 hr.
By min. temp. betw. Sts. 2347 and 2348.
By thermograph at 0300 and 1900 hr.
on 29th and 0800 hr. on 30th.
By min. temp. betw. Sts. 2358 and 2359.
By thermograph at 1330 hr.
By min. temp. betw. Sts. 2377 and 2378.
By thermograph at 0630 hr.
By min. temp. S of St. 2387. By thermo-
graph probably at 1500 hr.
By min. temp. betw. Sts. 2416 and 2417.
By thermograph at 1400 hr.
By min. temp. betw. Sts. 2426 and 2428.
By thermograph at 0100 hr. (doubtful).
By min. temp. betw. Sts. 2449 and 2450.
By thermograph at 1300 hr. (doubtful).
By min. temp. N of St. 2461. By thermo-
graph S of St. at 0300 hr.
By min. temp. betw. Sts. 2481 and 2482.
By thermograph at 0600 hr.
By min. temp. betw. Sts. 2495 and 2496.
By thermograph at 1600 hr.
By min. temp. betw. Sts. 2521 and 2522.
By thermograph at 0130 hr.
By min. temp. S of St. 2533. By thermo-
graph possibly at 0030 hr.
By min. temp. betw. Sts. 2573 and 2574.
By thermograph at 1030 hr.-
By min. temp, near St. 2585. By thermo-
graph indeterminate.
By min. temp. betw. Sts. 2622 and 2623.
By thermograph at 1400 hr.
211
NOTES ON THE PLATES
Place-names are shown in Plate I.
Plates II-XII show the mean surface isotherms for each month in Antarctic waters and for a short distance north of the
Antarctic convergence. The pack-ice edge, shown by a hatched line, is in the mean position for each month, and is reproduced
from Mackintosh and Herdman (1940) without alteration except at one point in Plate III between 10° and 100° W (see text,
p. 199). The positions of thermograph records, and single observations included in the Discovery Committee's data, are shown
in yellow; pecked lines in yellow represent ship's routine temperature readings.
DISCOVERY REPORTS, VOL. XXIII
PLATE I
GO.
,,,,.,'.,,/>•:./> . J ,'!'■• ,'r, I fr^'i-l ri I irfc
•■ ' i ','.■ I ■
60
30
120
150°
V/ 180°E
Estimated mean position of the Antarctic convergence. The mean northern limit of the pack-ice is the average
Lstimated mean po^s^i^^^ ice-edge for September reproduced from Mackintosh and Herdman (1940)-
DISCOVERY REPORTS, VOL. XXIII
PLATE II
Surface temperatures, SEPTEMBER and OCTOBER.
DISCOVERY REPORTS, VOL. XXIII
PLATE III
Surface temperatures, NOVEMBER.
DISCOVERY REPORTS, VOL. XXIII
PLATE IV
oO
120
120
150°
W I80°E
150°
Surface temperatures, DECEMBER.
DISCOVERY REPORTS, VOL. XXIII
PLATE V
120
150°
W 180-E
150°
Surface temperatures, JANUARY.
DISCOVERY REPORTS, VOL. XXIII
PLATE VI
30°
60
120
ISO-
W 180"E
ISO-
Surface temperatures, FEBRUARY.
DISCOVERY REPORTS, VOL. XXIII
PLATE VII
Surface temperatures, MARCH.
DISCOVERY REPORTS, VOL. XXIII
PLATE VIII
Surface temperatures, APRIL.
DISCOVERY REPORTS, VOL. XXIII
PLATE IX
Surface temperatures, MAY.
DISCOVERY REPORTS, VOL. XXIII
PLATE X
Gtf
150°
W 180"E
130°
Surface temperatures, JUNE.
DISCOVERY REPORTS, VOL. XXIII
PLATE XI
GO;
90
120
iso°
W 180°E
150°
60
30
120°
Surface temperatures, JULY.
DISCOVERY REPORTS, VOL. XXIII
PLATE XII
Surface temperatures, AUGUST.
DISCOVERY REPORTS, VOL. XXIII
PLATE XIII
0° ISOTHERM, summer months.
DISCOVERY REPORTS, VOL. XXIII
PLATE XIV
30°
go:
90
120
...,..■.,..>"■/-"■-■■. /.,.mmm/u riwriTi I I'rrrrI' r
.■Ml.f.. I m\m I'l I < ' IMM'l'V \'y'\'\'\' ^■\' \'">
150°
W 180-E
ISO"
0° ISOTHERM, winter months.
[Discovery Reports. Vol. XXIII, pp. 213-222, Plate XV, August 1946.]
NEBALIOPSIS TYPICA
By
H. GRAHAM CANNON, Sc.D., F.R.S.
NEBALIOPSIS TYPICA
By H. Graham Cannon, Sc.D., f.r.s.
Beyer Professor of Zoology in the Victoria University of Manchester
(Plate XV)
THE FORM OF THE BODY
IN a recent report which has just reached this country, Linder (1943) has called in doubt the shape
and configuration of Nebaliopsis typica which I described in an earlier Discovery Report (193 1) and
illustrated by photographs of a specimen which I stated (loc. cit. p. 201) was complete and undamaged.
The only specimen which, previous to my report, had reached the surface intact, I stated on the
authority of Dr Odhner of the Riksmuseum, Stockholm, was most probably lost. Dr Linder has now
found the missing specimen, and it is this that he maintains represents the normal appearance of this
rare deep-sea crustacean. My specimen was labelled F 2, and was illustrated by three untouched photo-
graphs on plate xxxii, and text-figs, i and 2, based on these photographs. Linder's chief criticism
is that the cephalothorax is distended, more especially in the posterior part, while the carapace has
remained unaffected (loc. cit. p. 5). As a result, part of the thorax is not covered by the lateral carapace,
and the length of the carapace relative to the rest of the body is abnormally small. The difference is very
marked by measuring Linder's figure and my own. A comparison shows that the length of the cephalo-
thorax in his specimen is about 54 % that of the carapace, while in my specimen it is 96 %. Obviously
such a large difference calls for further investigation.
I pointed out (loc. cit. p. 202) that measurements of all the specimens I had, together with that of the
specimen we may now call Linder's, established the existence of a considerable variation in the length
of the carapace and left the matter at that. I took the view that there was no doubt about it as, while
Linder's specimen and mine were both presumably as perfect as could be, his had a large carapace and
mine had a small one. I was relying for details of Linder's specimen on a sketch which had been
published by Ohlin (1904, fig. i). From this, however unsatisfactory it might have been in other
respects, there was no doubt about the length of the carapace— it was relatively long. The question to
be settled is whether the relative shortness of the carapace of my specimen F2 is abnormal.
Linder's explanation (loc. cit. p. 6) is that at the great depths at which Nebaliopsis lives, there must
be inside the body an enormous pressure. This is a point which no one will dispute. However, he then
says that as the specimen is brought to the surface in the collecting gear this pressure inside, acting in
all directions, enlarges the body so that by the time atmospheric pressure is reached at the surface the
body is completely distended. This distension will occur to different degrees in different parts, and
Linder assumes that the carapace is not affected by the pressure. Hence it retains its normal size while
the body becomes bloated, and thus the carapace appears abnormally short.
Now this argument represents a widespread fallacy, a fallacy which has arisen first from the known
fact that specimens of deep-sea fish are occasionally completely distended when they reach the surface,
and secondly from the persistent and erroneous belief that sea water is much more dense at great
depths than at the surface. Actually Linder's argument is valid, and then to a very uncertain degree,
only if the body of Nebaliopsis contained gas.
As long as the body contains no bubble of gas, no such distension as Linder describes can take place.
Now there is no reason to suppose that the body of Nebaliopsis contains gas any more than one may
expect to encounter gas on opening up a lobster. Its body can be looked upon very largely as a mass of
aqueous liquid and, moreover, a liquid closely similar in its physical properties to sea water. Now sea
2j^ DISCOVERY REPORTS
water, like other liquids, is only slightly compressible, or, more emphatically, is practically incom-
pressible. The coefficient of compressibility of sea water is roughly 44 x io-«, which means that at the
depth of a mile— the depth at which Nebaliopsis is known to occur— the density of sea water would only
be 1/130 the greater. From this figure a rough computation shows that a specimen such as Nebaliopsis,
if lowered to a depth of a mile, would diminish in length by approximately 1/400. Clearly, then, a
specimen raised up from i mile would expand by 1/400 of its length, that is, by an imperceptible amount
that would be extremely difficult to measure. Obviously such an expansion would lead to nothing like
the distension of my specimen F2.
I have stated that the presence of gas in a specimen, such, for instance, as occurs in the air bladder of
a bony fish, would make a considerable difference to the above argument, for a gas, unlike a liquid, is
extremely compressible. Now the pressure at a depth of 2000 m. is approximately 200 atm. Hence, if
there were a bubble of gas in such a specimen at such a depth it would enlarge to 200 times its original
volume by the time it was drawn up to the surface, or roughly a bubble the size of a pin's head at the
depth of i^ miles would expand to the size of a pea at the surface. This is a considerable expansion and
is of the order that might produce the enlargement which Linder maintains has occurred in the
Discovery specimen F2. Is there any reason, therefore, to suppose that gas is produced in the body of
Nebaliopsis as it is brought to the surface? It might be argued that as it passes upwards the resulting
reduced pressure would cause the dissociation of the oxyhaemocyanin that is presumably present in its
blood. But the solubility of a gas is proportional to its pressure, and hence oxygen at 2000 m. is 200
times as soluble at the surface. Hence any oxygen which might be set free from the dissociation of the
oxyhaemocyanin would immediately go into simple solution. For the sake of argument, however, let us
suppose that the dissociation took place so quickly that there was not sufficient time for the oxygen to
become dissolved before the animal reached the surface. What would be the result? Most probably the
animal would burst under the strain of the sudden expansion inside it and the gas would escape. But
if it were able to stand the strain without bursting one thing that is certain is that the oxygen would
remain in it as a bubble. Anyone who has handled preserved specimens of Crustacea in spirit knows the
difficulty of getting rid of a bubble of air in a specimen once it has got in. In my specimen F 2 there was
no bubble of air, as I think my photographs (loc. cit. plate xxxii) established without a doubt.
As far as I have argued, therefore, I have shown that my specimen of Nebaliopsis could not have
expanded as suggested by Linder merely by being relieved of the enormous pressure under which it
lived, simply because liquids are almost incompressible, which is the same thing as saying unexpansible.
Neither could it have become extended with gas, because there was in fact no gas in the specimen when
it reached the surface. My chief argument, however, in refuting Dr Linder's suggestion is the obvious
undistorted condition of specimen F2. Naturally, when I first received the specimen I was struck with
its bloated appearance, and this caused me to wonder whether it could be abnormal. Directly I examined
it ventrally and dorsally, however, I had no hesitation in deciding that it must be undistorted, for it
showed the complete ventral chain, of nerve ganglia, and, more striking, the complete tubular heart.
I am afraid I assumed that this perfection of the inner organs was so obviously a testimony to the
condition of the animal that I did not comment on it in my paper ; I merely relied on the photographs.
Now, surely it is practically impossible for the hind part of the cephalothorax to be enlarged without
producing an obvious distortion of either the nerve chain or the heart. Certainly if the supposed
expansion were caused by an expanding gas bubble (and I have shown that this, however remote, is the
only possibility) the expansion would not be bilaterally symmetrical, for the bubble would of necessity
lodge to one side of the gut.
The final demonstration, however, that the Discovery specimen is normal comes from the words of
Dr Linder himself. He states (loc. cit. p. 7) that 'a study of sections provides a certain proof. . .
NEBALIOPSIS TYPICA 2i7
specimens that are only slightly swollen show the muscles of the thorax torn away from the integument '.
Now, specimen F2 was sectioned down to the sagittal plain as I stated in my report (loc. cit. p. 204).
The sections are quite normal and are typical of deep-sea material. There is no such distortion of the
musculature as Linder predicts, or, in fact, of any other organ. The proof of this is that Miss Rowett
(1943) has used the series of sections to work out with conspicuous success the anatomy of the gut.
Miss Rowett 's work was published a few months after the appearance of Dr Linder's paper, and
throws a completely new light on Nebaliopsis. Moreover, it supplies an obvious explanation of the
differences between our two specimens. She showed that the large opaque mass in my specimen F2,
which I tentatively suggested was the ovary (loc. cit. p. 203), was in fact an enormous sac-like diver-
ticulum of the mid-gut. It was filled with a homogeneous coagulum ; that is, with a mass of food material
in which there was no structure (Rowett, 1943, p. 15). From this, with admirable argument, she puts
forward the view that Nebaliopsis is an egg-sucker. She then shows that, quite apart from the apparatus
for fiker-feeding, this remarkable form is fully adapted in a variety of ways to this peculiar diet. In
a paper that is now in the press she has gone further and has pointed out the extraordinary correspon-
dence that occurs between the adaptation of Nebaliopsis and of the nudibranch Calma glaucoides which
is known to feed exclusively on a diet of eggs.
The adaptation which is of importance in the present discussion is the enormous mid-gut sac.
Miss Rowett points out (loc. cit. p. 8) that this sac is 'without any convolution and with only a few
septa arising from its walls'. Therefore while it does not provide much extra surface for its digestion
it is admirably suited for a storage organ. Now eggs will certainly not always be present at the great
depths at which Nebaliopsis lives. During the breeding periods of neighbouring animals there will be
abundance, but in between whiles scarcity. However, it must be remembered that the neighbours of
Nebaliopsis are few and far between, so that even when eggs are present they will be patchy in their
distribution. A large storage organ is clearly an adaptation to this. It enables the animal to take a con-
siderable quantity of food on the infrequent occasions when it happens to encounter a patch. My
specimen F2 is clearly a specimen that had just had a meal and it became distended in the same natural
way in which, for example, a blood-sucking tick becomes bloated after a meal. Another parallel is to be
found in a deep-sea fish such as Chaismodiis niger (Murray and Hjort, 1912, p. 721, fig. 515) which has
an abdomen so distensible that it can accommodate a larger specimen of the same species.
The musculature of the body appears to be arranged so as to allow the body to expand in the hinder
trunk region. The abdominal region is a packed mass of muscles. The cephalic region contains all the
musculature of the antennae and mouth parts and the muscles extend ventrally in association with the
trunk limbs. They become less marked posteriorly in relation to the simple condition of the eighth
trunk limbs. The dorsal and ventral longitudinal muscles are practically non-existent in the thoracic
region. There is thus a large region of the body, the posterior and dorsal thoracic region, which is
almost devoid of muscles. The integument over this region is very thin and flexible and it is here that,
as my photographs show, the expansion takes place. u • u ^
To summarize, the enlarged appearance of the Discovery specimen F2 is due to the fact that it had
just taken a meal. It is not in any way an unnatural distension. A comparable specimen in which the
mid-gut digestive sac is empty is shown in Plate XV. This beautiful example occurs in a second small
Discovery collection. I consider this specimen much more perfect than Linder's, for it shows the
rostrum, eyes, and antennae in a normal position for one of the Nebaliacea. . , ^ , .
The photograph which Linder published (loc. cit. Taf. I, fig. i) agrees fairly closely with the sketch
published by Ohlin (loc. cit. fig. i) ; enough, in fact, to make it fairly certain that the sketch was made
after fixation and not while the animal was swimming round. How the animal was preserved we do not
know There is nothing to indicate the use of any special fixative, and so most likely it was placed in
2,8 DISCOVERY REPORTS
spirit or formalin in sea water. Also it was almost certainly moribund, when it was pickled, for as
Linder (loc. cit. p. 7) points out after quoting Ohlin, it could only swim on its side, a quite unnatural
position. Moreover, having obtained for the first time a living specimen of a deep-sea crustacean it is
only natural that the scientist would keep it alive as long as possible. Whatever happened, it was during
the fixation that the first damage to the specimen occurred. Now spirit and formalin are both slow
fixatives and take a considerable time to kill an organism. I have even seen a specimen of the fairy
shrimp Chirocephalus placed in the relatively fast fixative Bouin, and after one minute it was still
wriggling. During this time the outermost musculature becomes fixed, while the inner muscles still
contract. Naturally the result is a distorted specimen. It looks to me as though Linder's specimen,
during fixation, became distorted in this way, for in the head region the rostrum, together with the
antennules and antennae, appear to have been pulled inwards so as to become completely covered by
the carapace. This does not occur in any other nebaliacean. Even in badly damaged specimens of
Nebaliopsis, in which the body has become wrenched away from the carapace, the head region appears
to remain intact with the antennae and rostrum projecting beyond the edge of the carapace. This is
very clearly shown in one of Linder's own specimens (loc. cit. Taf. i , fig. 3), which shows just the same
arrangement of these parts as in the photograph I am publishing in this paper as well as in that of my
original F2 specimen.
In addition to this fixation damage I think there must have been slow shrinkage of the whole body
relative to the carapace after fixation, for in Ohlin 's sketch the hinder margin of the carapace reached
only to the front end of the seventh abdominal segment, while in Linder's photograph taken years later
it stretches to the hind end.
Some deep-sea Crustacea are able to stand a journey up to the top and appear quite normal when
caught ; thus Gigantocypris is often collected in numbers from great depths and will swim about actively
in bowls of sea water. My own studies of this form (Cannon, 1943) have, I think, demonstrated that
these forms are quite unaffected by the enormous reduction of pressure which they undergo in their
passage upwards of maybe more than a mile. Nebaliopsis, on the other hand, is now known from a
considerable number of specimens, and yet only three are anywhere near perfect. The greater number
are completely disrupted. Now Nebaliopsis is a much more delicate form structurally than the robust
Gigaiitocypris, but I feel certain that this does not explain the difference between them. Gigantocypris
has a firm and substantial outer shell, while that of Nebaliopsis is extremely delicate. But, on the other
hand, both have bodies constituted of cells containing living protoplasm and, quite irrespective of their
exoskeletons, that protoplasm is in equilibrium under enormous pressure. When they are brought to
the surface that pressure is relatively quickly diminished. Now, there can surely be no doubt that this
change will produce an immediate disturbance of the equilibrium of the protoplasm. In the majority
of cases it is probable that the protoplasm would be unable to readjust itself and so would be precipi-
tated, with inevitable death resulting. This is what I consider happens in Nebaliopsis. On the other
hand, if the protoplasm can so adjust itself as to remain in equilibrium, then the specimen will appear
living and normal at the surface as in the case of Gigantocypris. If my argument is correct, it means that
Nebaliopsis collected in deep-sea trawling is normally fixed (but not preserved) soon after leaving the
depth at which it occurs. It is fixed, that is, its protoplasm is precipitated, by suddenly reduced
pressure. Now this will act on all parts of the body quickly at the same time. Thus, there will be no
question of penetration of fixative. The protoplasm of all cells of the body, whether they are deep or
superficial makes no difference, will suddenly precipitate. There will be no distorting death struggles-
one part of the body still alive pulling against another part already fixed. There will be a sort of shock
suddenly immobilizing all parts of the body at the same instant, and the result will be a fixed prepara-
tion as nearly like the living form as possible. If, now, such a specimen can finish the journey to the top
NEBALIOPSIS TYPICA 219
and, more important, on to the boat, without being crushed, and further, if it is subsequently handled
and preserved by an expert, such specimens as my F 2 and the one I illustrate in this paper will be
obtained.
THE FILTRATORY FEEDING MECHANISM
In Nebaliopsis I described in my earlier report (1931, p. 210) a fihratory mechanism unique amongst
the Crustacea, for it was based on the joint action of the maxilla and the first trunk limb. I gave
evidence that this type of filter-feeding must have evolved from that of Nebalia, which I had previously
shown (1927) resulted from the combined action of all the trunk limbs, the maxilla being minute and
taking no part in the filtering process. I suggested (1931, p. 216) that along the lines leading to
Nebaliopsis the ancestral form ' developed the maxillary-first trunk limb filter mechanism, at first, to aid
its more posterior trunk limb filter. . . . Then, when this became sufficiently advanced, it opened up the
carapace — maybe to allow a greater inflow of water on to the maxillary region. . . . The maxillary filter
now became the chief feeding mechanism. . . . As it developed so the carapace widened out and water
came to be sucked in from all directions. The trunk limb filter was then almost abandoned, the trunk
limbs swinging forwards to act as a subsidiary mechanism supplying water to the . . . maxillary-first
trunk limb filter.'
Linder (loc. cit. p. 30) accepts my description of the new filter apparatus, but considers that in
addition the trunk limbs together still form an efficient filtering mechanism. He even goes further
(loc. cit. p. 31) and suggests that the filtering power of Nebaliopsis is more effective than that of Nebalia,
a point which I most strenuously deny. The filter process of Nebalia is, as far as I can judge, the most
efiicient of all those crustacean feeding mechanisms that I have studied, for water carrying suspended
food particles is sucked into a filter chamber just as if it were being sucked into a cylinder by a movable
piston, and after being filtered is passed to the exterior through a valve as efficient as a rubber gas valve.
The filter chamber of Nebalia into which the water is sucked is the median ventral space between the
trunk limbs. The entrance lies anteriorly in the mouth region. Posteriorly the chamber is closed by the
eighth trunk limbs uniting medially to form a wall. Laterally the trunk limbs form its walls, the spaces
between the limbs being spanned by continuous sheets of filter setae. Dorsally it is roofed by the mid-
ventral body wall, while ventrally there is a complete and thick floor formed by the endopodites of the
trunk limbs which recurve sharply backwards and slightly inwards so that their tops touch in the
middle line. Thus the filter chamber is a simple laterally compressed space with but one entrance, and
that is relatively small. It is a slit extending from the lower edge of the labrum to the 'elbow' of the
first trunk limb. Now the first trunk limb is itself small — it is only about two-thirds the length of the
middle trunk limbs which are the main limbs acting as pistons sucking water into the filter chamber.
Thus we have a relatively large filter chamber with a small opening into it. Clearly suction will be very
powerful, so that once water has been sucked in it will not escape out again through the same opening —
it must remain to be filtered. Obviously then any enlargement of this entrance will lead to a diminution
in the force with which the water is sucked in. In Nebaliopsis the whole floor of the filter chamber has
opened up by the disappearance of the long recurved posteriorly projecting endopodites of Nebalia.
Hence, even if the trunk limbs were acting as in the latter form, the suction into the filter chamber
would be relatively weak. However, it is wrong to call this median space between the trunk limbs of
Nebaliopsis the filter chamber. Apart from the fact that I showed that another filter chamber had
developed between the maxilla and the first trunk limb, this space, now that it is completely open
ventrally, is directly comparable with the mid-ventral space of a branchiopod. Thus, again supposing
that the trunk limbs were still filters, as they are in Nebalia, the efficiency of their combined efforts in
220 DISCOVERY REPORTS
Nebaliopsis would be comparable with that of the feeding mechanism of the fairy shrimp Chirocephalus.
Now this type of filtratory mechanism is not nearly so powerful as that of Nebolia. A floating particle,
once it has been sucked in between the limbs of Nebalia, cannot, as far as I can see, escape unless it is
forcibly ejected. On the other hand, a particle, on being sucked into the mid-ventral space of Chiro-
cephalus, as often as not is blown out again, and it would be the same with Nebaliopsis.
However, this is all on the supposition that the trunk limbs of Nebaliopsis are still efficient filters.
But I pointed out (1931, p. 21 1) that this cannot be so because, among other things, their lateral parts
do not form a valvular system as they do, in fact, in all other Nebaliacea. In Nebalia and Paranebalia
it is the epipodites together with the exopodites, and in Nebaliella, where the epipodites are absent, the
exopodites alone, which project backwards and completely and accurately span the gaps between
successive limbs, thus preventing any lateral entry of water into the inter-limb spaces during the
suction phase. In Nebaliopsis, on the contrary, there are wide lateral gaps both proximal and distal to
the epipodite through which water could pass unhindered. Linder (loc. cit. p. 30) denies this and says
that the gaps are simply due to the distension of the limbs. He implies, although he does not say so in
as many words, that the main axis of the limb has elongated under pressure, while the epipodite has
remained the same size, and hence a gap has occurred between the epipodite and the body wall. The
fallacy of this is obvious. There is no reason why the epipodite should not enlarge along with the rest of
the limb. In fact from the anatomy of the limb it is fairly certain that if this hypothetical swelling did
take place it would be the epipodite which would enlarge most and not the main axis of the limb, for
the latter is skeletally relatively rigid, while the epipodite is extremely delicate.
One further argument : if the lateral gaps which are so clearly seen in the upper photograph of my
plate xxxii (193 1) are due to artefacts, how is it that the lateral space between the first and second trunk
limbs is so accurately covered by the epipodite of the first trunk limb (loc. cit. p. 206, fig. 3)? Is this
also due to distension? Surely not — it is simply a manifestation of suction between the first and second
trunk limbs which is absent, or practically so, between the other limbs.
Having explained away the lateral gaps between the trunk limbs, Linder still has to deal with the
distal gaps that occur between the tips of the limbs. These gaps result from the absence of the pos-
teriorly curved endopodites. The typical trunk limbs ' are unsegmented, the exopodites being repre-
sented by a slight protuberance. . . .The endopodite must be considered as the tip of the limb distal to
this exopodite lobe' (Cannon, 193 1, p. 207). They are, as Linder correctly states (loc. cit. p. 10), turgor
limbs, that is, limbs that depend for their rigidity on an internal blood pressure. Moreover, ' the distal
two-thirds of each limb is devoid of musculature ' (Cannon, loc. cit. p. 210), so that the limb cannot be
bent by internal muscles — it can only be moved as a whole by the muscles at its base after being made
rigid by being pumped full of blood.
Now Linder explains the closure of the gaps between the tips of the limbs during the 'Abduktions-
phase ' (suction phase) as due to a bending over. of one limb on to the limb in front, as a result of suction
between the limbs (loc. cit. p. 30). This would imply that directly any two successive limbs commence
their suction stroke, the suction between them becomes so great that it causes the hind limb to bend in
the middle so that its tip comes down on the limb in front and closes the distal gap. But what causes
this original powerful suction? Unless there are efficient valves, the suction between the limbs will be
negligible. The gap is there to start with and hence the small suction caused by the inter-limb space
enlarging will simply draw in water through the gap which will at once neutralise the suction. Linder
says that similar conditions obtain in the Anostraca (loc. cit. p. 30) and he sees no reason why the same
thing should not happen in Nebaliopsis. But things are quite different in the Anostraca. There the limb
is jointed and the distal endopodite ffap is provided with a complete set of muscles by which it can be
pulled down and held, if necessary, on to the next limb.
NEBALIOPSIS TYPICA 221
Linder goes further and, as I have mentioned above, states (loc. cit. p. 31) that the filtering
mechanism of Nebaliopsls is an even more efficient system than that of Nebalia. He points out that it
differs in two respects — first that the eighth trunk Hmb is not fihratory, and secondly that the
maxilla, together with the first and second trunk limbs, form a separate filter system. It can only be
deduced from this that Linder considers that the Nebaliopsis mechanism evolved directly from that of
Nebalia, a fact with which all will agree, and further, that it only shows the two differences which he
enumerates. There is, however, another fundamental difference which he has overlooked. In Nebalia
the trunk limbs curve backwards so that one limb overlaps the limb behind. In Nebaliopsis, according
to Linder's hypothesis, the limbs bend forwards so that one limb overlaps the limb in front. This
being so it is clear that Linder's Nebaliopsis mechanism could not have evolved from that of Nebalia,
for such an evolutionary process would involve an intermediate stage when the limbs curved neither
backwards nor forwards and so would be unable to carry out any filtration. If, therefore, the trunk
limbs of Nebaliopsis do filter according to Linder's mechanism, then this would be an entirely new
development and this is not what Linder has been arguing.
There would be the same difficulty here as occurs when it is attempted to derive the malacostracan
maxillary filter from the anostracan trunk limb filter (Cannon, 1928, p. 820). Both these are based on
typical phyllopodia, but in the former the limb is concave anteriorly and in the latter posteriorly. But
what is more important, the functional activities of the limbs are dependent absolutely on these
arrangements of the limbs. The same applies to Nebaliopsis and Nebalia. The activities of their trunk
limbs, whether actually filtratory or only supposedly so, depend on their arrangement and since they
are arranged as mirror images, one pointing forwards and the other backwards, it is not possible to
derive one mechanism from another, without postulating an intermediate stage during which the limbs
could not function.
BIBLIOGRAPHY
Cannon, H. G., 1927. On the feeding mechanism 0/ Nebalia bipes. Trans. R. Soc. Edinburgh, lv, pp. 355-70.
Cannon, H. G., 1928. On the feeding mechanism of the fairy shrimp, Chirocephalus diaphanus Privost. Trans. R. Soc.
Edinburgh, LV, pp. 805-22.
Cannon, H. G., 1931. Nebaliacea. Discovery Reports, Cambridge, hi, pp. 199-222, pi. xxxii.
Cannon, H. G., 1943. On the anatomy of Gigantocypris muUeri. Discovery Reports, Cambridge, xix, pp. 185-244,
pis. xxxix-xlii.
Linder, Folke, 1943. t)ber Nebaliopsis typica G. O. Sars, nebst einigen allgemeinen Bemerkungen iiber die Leptostraken.
Dana-Report, No. 25.
Murray and Hjort, 191 2. The Depths of the Ocean. London.
Ohlin, a., 1904. tjber eine neue hathypelagisch lebende Phyllocaride. Zool. Anz., Leipzig, xxvii, pp. 59-61.
RowETT, Helen G. Q., 1943. The gut of Nebaliacea. Discovery Reports, Cambridge, xxiii, pp. 1-18.
PLATE XV
Nebaliopsis typica x 9. St. 1636. Date 30. xi-i. xii. 35. Net N 100 B. Depth 380-150 m.
DISCOVERY REPORTS, VOL. XXIII
PLATE XV
NEBALIOPSIS TYPICA
[Discovery Reports. Vol. XXIII, pp. 223-408, Plate XVI, December 1946]
REPORT ON TRAWLING SURVEYS ON THE
PATAGONIAN CONTINENTAL SHELF
Compiled mainly from manuscripts left by the late E. R. Gunther, M.A.
By
T. JOHN HART, D.Sc.
CONTENTS
Foreword, by N. A. Mackintosh page 226
Introduction 227
Field methods and preliminary observations 228
The first survey 230
The second survey 234
The third survey 236
Topography of the shelf 238
Hydrology 242
Plankton 246
Methods of presentation 250
General account of the fish fauna 251
Distribution and general notes on the species . ..... 259
Petromyzonidae ........... 259
Myxinidae 259
Lamnidae 259
Scyliorhinidae ........... 260
Squalidae 260
Squatinidae ............ 260
Torpedinidac 260
Rajidae ............. 260
Chimaeridae ............ 272
Summary of observations on Elasmobranchii ...... 274
Clupeidae ............ 275
Galaxiidae ............ 278
Aplochitonidae ........... 279
Syngnathidae 279
Macruridae 279
Merlucciidae . 280
Merluccius hubbsi . 280
Introduction: economic importance of allied species . . . 280
The sizes of Patagonian and European hake compared . . . 284
The distribution and relative abundance of Merluccius hubbsi within
the area surveyed, and the effect of latitude on numbers, size and
sex ratio . . : 289
The relation between length and weight of Merluccius hubbsi, and its
value as an indicator of the spawning season, and for other purposes 291
Migrations ^00
Conclusions on migration 106
The food and feeding of Merluccius hubbsi 308
Parasites ,jq
Macruronus magellanicus ' . .312
Comparison and contrast of the main features in the bionomics of
Merluccius hubbsi and Macruronus magellanicus . . . . ^20
Gadidae ^20
Micromezistius australis ^20
Salilota australis ~~.
Physiculus marginatus ^24
Muraenolepidae
Carangidae
Bovichthyidae ^
CONTENTS
Nototheniidae P^S^ 32?
Notothenia ramsayi 33
Harpagiferidae . ....••■•■• 34
Chaenichthyidae 34'
Summary of observations on Patagonian Nototheniiformes . . -342
Gempylidae ....•••••••" 347
Thyrsiies atun 347
Scombridae 35
Zoarcidae 35
Summary of observations on Zoarcidae 353
Lycodapodidae . . • • ■ ■ • ■ • ■ '354
Ophidiidae 355
Genypterus blacodes 355
Brotulidae 35^
Centrolophidae 35
Stromateidae 359
Stromateus maculatus 359
Atherinidae 374
Scorpaenidae 374
Congiopodidae ...•••••••" 375
Psychrolutidae 37
Agonidae 37
Liparidae •' ' '
Bothidae 377
Thysanopsetta naresi . ■ ■• 377
Other species 3
Features of general biological interest 3 2
Prospects of commercial development 3 7
The weight of catches 3 7
Conclusions 39
References 39
Appendix I. Particulars of trawling stations 39^
Appendix II a. Hake data: first survey 4°°
Appendix IIb. Hake data: second survey 4oi
Appendix lie. Hake data: third survey 402
Appendix III. Approximate positions of localities mentioned in the text but
not charted in text-figures '^°
pjgjg ' following page 408
FOREWORD
By N. a. Mackintosh
The position, extent, and physical features of the Patagonian Continental Shelf are such as to prompt
comparison with the major fishing grounds of the northern hemisphere ; but prior to the trawling
surveys of the R.R.S. 'William Scoresby ' the nature and magnitude of the population of demersal fishes
on the shelf had not been explored. Three surveys were carried out, for the most part at different times
of the year, in 1927, 1928 and 1931-2; and the principal results are set forth in Dr Hart's report.
The report is based to a considerable extent on the unfinished work of the late Mr E. R. Gunther,
and the circumstances in which it was prepared require a word of explanation. Several members of
the Committee's staff took part in the work at sea, but the largest part was played by Mr Gunther,
and he was in charge during the third and most extensive survey. It is very largely owing to his
untiring energy and enthusiasm, and his broad conception of the problems involved, that a most
comprehensive and thorough investigation was carried out. The surveys have resulted in a very great
mass of data, and the preparation of a general report on the results was undertaken by Mr Gunther.
Some delay in the completion of this report was inevitable. A taxonomic account of the fish fauna,
and a description of the marine deposits of the shelf, were needed first, and these have already
been published in the Discovery Reports, together with systematic papers on certain invertebrate
groups. In the years before the war, however, Mr Gunther had made good progress with the
general analysis of the material, though his work was unavoidably interrupted from time to time by
other responsibilities. Since he held a commission in the Territorials he was called upon for military
service just before the outbreak of war, and further progress was therefore suspended. His death on
active service in 1940 was a severe loss to the Discovery Investigations.
Owing to the dispersal of the Committee's staff in wartime no more could be done until 1943, when
Dr Hart was able to take over the work. The task of picking up the threads and collating the data
naturally involved considerable difficulties; and for many aspects of the subject it was necessary to
start again from the original data. The substance of the report can be regarded as the combined work
of Mr Gunther and Dr Hart, but the latter, as he explains below, has written the entire text in its
final form.
It will be realized that this report deals mainly with the general biology and ecology of the demersal
fish, and with the prospects of commercial trawling. The surveys were planned for this purpose and
did not include an investigation of the pelagic fish such as the Falkland herring. Various references
to these fish are included in the report, but there is still little information on the prospects of com-
mercial fishing by other means than trawling. The principal conclusion of the report is that hake, and
some other edible species, are obtainable in moderate numbers by trawling. Although the shelf has
been found to be less rich in trawlable fish than might have been expected, it is possible that enough
could be taken to support an industry if markets could be found, and problems of preservation and
delivery could be overcome. The report may be regarded as a contribution to our knowledge of the
fish faunas of the world, and it is hoped that it will be of assistance in any consideration of the future
economic development of the Falkland Islands.
REPORT ON TRAWLING SURVEYS ON THE
PATAGONIAN CONTINENTAL SHELF
Compiled mainly from manuscripts left by the late E. R. Gunther, M.A.
By T. John Hart, D.Sc.
(Text-figs. 1-53, Plate XVI)
INTRODUCTION
THE Patagonian Continental Shelf extends from the River Plate in the north to Staten Island in
the south, and from the South American coast to an average distance of some 250 miles offshore to
the eastward.' Outlying areas with depths of less than 200 m. extend round the Falkland Islands and
on the Burdwood Bank to the south. Beyond the 200 m. contour, which may be taken as the edge of
the shelf, the descent to oceanic depths is more or less abrupt. Faunistic writers referring to the
Patagonian region commonly include the Magellan Channels and the coast of southern Chile m their
geographical unit, but there is no need to qualify our definition of the Patagonian Continental Shelf
so as to exclude the west coast, for there the descent to oceanic depths is so immediate that a shelf
(in the accepted meaning of the term) can scarcely be said to exist. The investigations to be described
here covered the whole of the shelf south of lat. 42° S., an area of some 150,000 square miles which
is larger than the entire North Sea. Except for descriptions of small collections made in coastal
waters the marine fauna of the region was almost unknown when the Discovery Investigations began,
although it constitutes the largest expanse of shoal water (accessible to trawling) in the 'cold tem-
perate' or sub-Antarctic Zone of the southern hemisphere.
The need for a fisheries survey of the shelf, to gain information on the prospects of developing a
commercial fishery from the Falkland Islands, was recognized from the outset of the Discovery
Committee's work (Kemp, in Kemp, Hardy and Mackintosh, 1929, P- 148)- The greater urgency of
problems relating to whaling and sealing in the more southerly (Antarctic) waters of the Falkland
Islands Dependencies limited the scope of the trawling surveys however, and combined, with Mr E. R
Gunther's untimely death in 1940, to prevent publication of results until now. Vast collections of
benthos with lesser but probably representative collections of plankton and hydrological data, were
obtained Gunther had hoped to use the information gained from these, as they were worked up by
various specialists, in presenting the ecological study of the fish fauna in much greater detail than can
now be attempted. Continued work on many of the groups may not be possible for years, but by great
good fortune the taxonomic revision of the fishes had been completed by the late Mr J. R. Norman in
IQ.7 As the need for more knowledge of the bionomics of the fish fauna became urgent, Dr Mackintosh
asked me to prepare this report, working from Gunther's manuscripts. I found this unusually difficult
because for the first time in my experience it involved work upon data which I had not helped to
collect Moreover, Gunther had planned the production of five separate papers, and the manuscripts
were in widely different stages of incompleteness. As a single report was called for I have entirely
re-written the text myself, retaining Gunther's leading ideas and indicating our indebtedness to him
so far as I am able. I found it necessary to recalculate all numerical data, using the origmal log books,
except where the reasons for alterations made by Gunther himself could be traced. Any mistakes in
this part of the work are my own responsibility. Both Mr Gunther and I have gained much from
discussion of hydrological results with Dr G. E. R. Deacon, F.R.S., and the brief notes on the
hydrology of the region presented here owe much to him.
22g DISCOVERY REPORTS
My own experience of trawling in the southern hemisphere has been hmited to a few experimental
hauls off the Falkland Islands and off the south-west coast of Africa, but I have seen enough to recognize
the tremendous amount of hard work in the field that these trawling surveys must have entailed. This
is especially true of the last and most comprehensive of the surveys, when Mr Gunther was in charge
of the scientific work and had as his assistant Mr (now Comdr) G. W. Rayner. I think this survey
was one of the most arduous pieces of field-work ever completed by the Discovery Investigations ;
and, as my two colleagues were always anxious to point out, its success was largely due to the able
and willing co-operation of the Captain, net-man and ship's company.
The work of writing up the report has been carried out in the Laboratory of the Marine Biological
Association at Plymouth, by courtesy of the Director, the late Dr S. W. Kemp, F.R.S., who, while
Director of Research to the Discovery Investigations, planned much of the work here described.
Dr Kemp's personal kindness and encouragement have helped me throughout my working life, and I
am sure all members of the Discovery scientific staff would wish to say the same. I have gained
much from the advice and encouragement of the staff of the Laboratory and more especially from
hints on the handling of numerical data by Mr E. Ford and Mr G. M. Spooner. Had these two
gentlemen not been away on war service during most of the period my task would have been lighter.
The rapidity with which the librarian, Miss M. Sexton, procured obscure references under all the
difficulties of wartime conditions was a great help. I have gained much from an all-too-brief interview
with Mr C. F. Hickling of the Ministry of Agriculture and Fisheries (now Fisheries Advisor to the
Colonial Office), who also helped with the loan of some of his important papers on Hake. This brings
me to a point that needs emphasis if the work described here is to be justly appreciated — it was planned
and carried out before the results of the last decade of fishery research, prior to the war, were known.
If, for example, the final results of Hickling's prolonged work on European Hake had been available
there is no doubt that our programme could have been modified with advantage, but our data were
collected before Hickling's work was complete.
FIELD METHODS AND PRELIMINARY OBSERVATIONS
The three trawling surveys were carried out by the R.R.S. 'William Scoresby' in autumn (March-
April) 1927, winter (June-July) 1928, and throughout the whole of the warmer half of the year 1931-2
(October-April). A few additional observations were made within the area by the R.R.S. ' Discovery'
and by the R.R.S. ' Discovery II '. A description of the ship and of the gear will be found in Kemp,
Hardy and Mackintosh (1929). In the more detailed parts of this report abbreviated descriptions of
the gear have been used, as standardized throughout the station lists in Discovery Reports. Meanings
of the abbreviations relevant to the present work are repeated here for the convenience of the reader.
For this study of the bionomics of the fish fauna, the gear may be grouped under two main headings :
' Trawl -|- accessory nets' and 'Other gear'.
' Trawl + accessory nets ' comprises :
OTC Commercial otter trawl, 80 ft. headline, 3 in. cod-end mesh.
N7-T Net 7 mm. meshl Fine-meshed nets attached to the back of
N4-T Net 4 mm. mesh ,- the trawl as described in the work men-
NCS-T Coarse silk net J tioned above.
' Other gear' includes a motley collection of apparatus that helped to extend our knowledge of the
distribution of the fish fauna in lesser ways. (The finer plankton nets, which scarcely ever catch fish
except in their larval stages, and hydrological apparatus are not considered here.)
INTRODUCTION
229
OTL Large otter trawl, 40 ft. headline, 1 1 in. cod-end mesh.
BTS Small beam trawl.
NRL Large rectangular net.
BNR Russell's bottom net.
DC Conical dredge.
N450H Large plankton net 4I m. diameter at mouth.
TYF Young fish trawl, stramin, 2 m. diameter ring^ Fished either horizontally
N looH (or B) Plankton net i m. diameter at mouth / (H) or obliquely (B).
NH Hand net.
LH Hand line.
65"
•i
O
O
c^.«&
■ WSI09
'WS79 ""^3° •WS76_./VS75 • WS73 .ws72
WS92
wsea
• WS9l'
VVS7I
5d
•WSS4
WS87 •
^i<^/
•■•••■■ 200M-"
■1^ h"'-^'f f-
65°
55
60°
Fig. I a. First trawling survey, station positions in March and April 1927.
This report is primarily based upon the hauls of the trawl + accessory nets, and full details of date,
time, position etc. of all such hauls (apart from total failures) are tabulated in Appendix I. Similar
details regarding the working of all the other gear will be found in our Station Lists when completed.
For the present purpose reference to other gear has only been made where it directly affects our
knowledge of fish distribution.
230
DISCOVERY REPORTS
THE FIRST SURVEY
The first survey was aimed at investigating the area immediately surrounding the Falkland Islands,
with some observations on the shelf between them and the mainland, and on the Burdwood Bank.
Dr N. A. Mackintosh, now director of research to the Discovery Investigations, was in charge of the
scientific work. With him were Mr E. R. Gunther specializing on the fishes, and Mr D. Dilwyn John
+5
•WSEIB
WS237
O
■ wsz3a
W5
WS2I4.
ZI5
ai6
• VVS2Z5
'1Z33
210 'WS
(230
'1231
■.VV5229
'. •WS228
•WS243 .WS239
5rf
■•WS246^
• WS227
WS250
' WS245
..I
• WS248
55
65°
Fig. \h. Second trawling survey; station positions in June and July 1928.
working on the invertebrates and the food and parasites of the fish. The normal routine at the trawling
stations was: first, a sounding; then a haul with the conical dredge; then, unless the bottom had
proved impossibly rough, an hour's haul with the ' trawl + accessory nets'. The trawling was followed
by a second sounding, and the collection of water samples from surface and bottom for the deter-
mination of temperature and salinity. A line of full hydrological and plankton stations between the
Jason Islands and Port Desire was also worked, so that with incidental stations elsewhere and minor
digressions to land the fur seal guard on Elephant Jason, and to mend nets, much work was accom-
plished in the period of just over two months devoted to the survey.
INTRODUCTION
231
The species of fish were counted, and the important ones measured and sexed, and data on the
stages of maturity at different lengths collected. Occasionally observations on scales, gonads, and
pathological growths, and sexing of the less important species were possible. Provisional identifi-
cations proved adequate for the subsequent accurate determination of almost all the specimens taken,
in spite of the facts that it was only possible to preserve a small proportion of the catch, and that some
Fig. ic. Third trawling survey; station positions October to April 193 1-2.
Nototheniidae and Rajidae presented special difficulties which were only cleared up in the course of
the preparation of Norman's report years afterwards.
The invertebrates were more difficult owing to the great bulk of many of the catches. Quantities
of previously preserved species were estimated and noted when possible, and all or a noted pro-
portion of the others preserved; but the bulk of the catch and masses of broken sponges, coralline
polyzoa, large Scyphomedusae, etc., frequently made it impossible to sort the catch adequately m the
limited time available.
232
DISCOVERY REPORTS
Some forty species of fish were taken, and it was reckoned that the most important were :
Squalus lebruni
Psafnmobatis spp.
Rajidae (several spp.)
Clupea fuegensis
Macrurotius magellanicus
Merluccius hiibhsi
Salilota australis
Cottoperca gobio
Notothenia ramsayi
N. gunt fieri
Champsocephalus esox
Stromateus maculatus^
The existence of overlapping characters in the ' tesselata group ' of Notothenia was noted in the
field, also the profuse variation among the Rajidae of the locality, which necessitated a large collection
Fig. id. Plankton and hydrological observations; station positions 193 1-2.
for further revision. Five teleosts new to science were collected, and specimens of Cottoperca gobio
and Dissostichus eleginoides much larger than any previously recorded were obtained.
The most important result of the survey was to demonstrate that hake {Merluccius hiibbsi) were to
be found in moderate quantity on the shelf to the north and west of the Falkland Islands. This fish
was for long confused with Merluccius gayi, the west coast species, and was the object of several small
^ The names in this list are those determined after Norman's revision.
INTRODUCTION 233
trawling ventures from the great ports of Montevideo and Buenos Aires. These all worked very much
farther to the northward and close in to the land ; it is safe to say that no considerable haul of the
species had been taken south of lat. 42° S before. On this autumnal survey the hake were distributed
over the shelf to the north of a line from Cape San Sebastian to the northernmost of the Falkland
Islands (Figs, i a, 2). The largest catches were obtained north-west of the Falkland Islands, but there
was little evidence of any special concentration there. A moderately rich haul was obtained so far
away as St. WS90 near Magellan Straits. Females were commoner than males, males were com-
monest in the shallower water along with the smaller females, and the catches with the largest females
were almost devoid of males.
Notothenia spp. were the most widely distributed and most numerous fishes, but were obviously
less important than hake owing to their small size. A very heavy catch of Notothenia was obtained at
one station north-west of the Falkland Islands, and another on the Burdwood Bank.
Macruronus magellanicus was most frequent at hauls made in the centre of the plain of the shelf.
Its excellent edible qualities and freedom from superfluity of small bones were gratefully recorded.
Stromateus macidatus was commonest near the mainland, but was also found at two ofltshore stations
towards the end of the survey. The flesh of this species was described as resemblmg that of the
herring, but not quite so good. It was remarked of both these species that though not very abundant
they occurred along with hake and might serve to supplement catches of the latter.
John's observations on the food of fishes showed that squids, Clupea spp., Thysanopsetta naresi,
hyperid amphipods and small Euphausiidae formed the bulk of the hake food. Each predommated m
different hauls to the exclusion of some or all of the others. The fact that hake snatch up food while
in the trawl was noted. This habit is also common among European hake, a point that became well
known with the publication of Hickling's work later. The stomach contents of small numbers of other
less important species of fish were recorded, and John also collected large numbers of fish parasites.
Invertebrates formed the bulk of the nine catches made within 20 miles of the Falkland Islands in
depths of 80-130 m., where fish were scarce. Sponges, actinians, Alcyonaria, coralline Polyzoa, spider
crabs echinoderms and ascidians were extraordinarily abundant. These invertebrate hauls were
heaviest to the south-east, south and west of the Falkland Islands. Smaller invertebrates reached their
greatest abundance where coralline Polyzoa or Alcyonaria predominated. This ground was too rough for
profitable trawling, owing chiefly to the corallines, large catches of 'rubbish and few fish^ This dis-
appointing negative result must be regarded as the second important point proved during the first survey.
Invertebrates were somewhat less numerous between the Falkland Islands and the mainland, but
some very heavy catches were obtained. In general the bulk of invertebrates decreased to the west-
ward and was least near the Argentine coast. Of nektonic forms squids were taken at nineteen out
of twenty-nine stations, sometimes in considerable numbers. They form an important part of the
diet of the larger fishes, birds and seals, of the area. Large Scyphomedusae were taken at seventeen
stations, and sometimes completely smothered the rest of the catch. .^, ^ ^.
A broad tentative correlation between the nature of the bottom and possible fishing prospects was
ventured upon at the close of the survey, and may be summarized as follows. On the shelf and to
the north of the Falkland Islands a dark greenish brown sand predominated on fairly clean ground.
This was the best ground for hake also and is therefore by far the most P^^^ WSsV a"mt
east south and west of the Falkland Islands, on the Burdwood Bank and at St. WS 88, a similar
sand mixed with a high proportion of shell fragments occurred. This was associated with a rich
nvertebrate fauna and few fish, except perhaps Notothenia spp. The Burdwood Bank presents very
poor prospects owing to foul ground and prevalent gales with a steep breaking sea. Subsequent work
substantiates these pronouncements in greater detail.
234 DISCOVERY REPORTS
THE SECOND SURVEY
The second trawling survey was carried out in the winter (June and July, 1928). Mr D. Dilwyn John
was in charge of the scientific work and was assisted by Mr J. W. S. Marr. This survey was planned
as a continuation of the first, with additional observations designed to discover the conditions to be
found along the edge of the shelf in depths below 200 m. Routine methods at the trawling stations
were the same as those followed during the first survey, with the addition of extra water sampling
before as well as after the trawling. At this time the ship was (unavoidably) without an experienced
trawler hand, and Mr John remarked that this led to much difficulty until the necessary experience
was acquired. Much foul ground was encountered so that much time was spent in mending nets.
The weather was often atrocious, and some minor breakdowns hindered operations still further, so
that the ship. did well to complete a rather larger programme than before (Fig. i b) in about the same
time. Material was dealt with as in 1927, and a large representative collection of bottom samples and
of the fauna was made.
On this survey the numbers of hake taken were small (1071 in all) and good hauls few, but it is
probable that they were present nearby in greater abundance, as the following considerations show.
The best catches were obtained on the edge of the shelf in depths of 200-300 m., on a line running
north to a point some 300 miles north of the Falkland Islands. Later the same area was disappointing.
The bottom was of clean, fine, dark green sand. On the coarse brown sand, pebbles, and shells, of the
shallower waters of the shelf, very few hake were taken in the north, and fewer or none to the south.
In the trough of relatively deep water to the west of the Falkland Islands, two very moderate catches
were obtained. No hake were taken on the shelly bottom with heavy invertebrate fauna to the south
and south-west of the Falkland Islands.
Comparing these results with those of the previous autumnal survey, when most hake had been
captured on the shelf to the north and west of the Falkland Islands, it was seen that the different
distribution observed in winter would agree with an offshore migration. Such a seasonal movement,
connected with the sexual rhythm, was already known to occur in the closely allied European species,
and by analogy John concluded that the Patagonian hake were summer spawners also. The data on
condition of the gonads, though not entirely satisfactory, were compatible with such a view. The
comparatively good catches in shallower water of March and April 1927 would thus be accounted for
by closer proximity to the spawning season, and the poor hauls of June and July 1928 mainly by
fishing ' out of season '.
It was known that European hake were caught most readily when concentrated for spawning in
relatively shallow water, and that deep-sea trawlers from British ports followed the ' seasons ' south-
wards, sometimes as far as the Moroccan coast. Also some of the British boats were already working
' over the edge ', in far deeper water than any in which large-scale trawling had previously been carried
out, in order to keep the market going during the off-season. Hence John's decision to work extra
stations along the edge of the shelf. The fact that these were only moderately successful was almost
certainly due to the steepness of the slope (far greater than that off the west coast of the British Isles)
which left a very small area of moderately deep water in which fishing was possible. John concluded
that a commercial fishery would have to follow the spawning fish throughout the year, and that that
would mean going farther afield to the northward than the first two surveys had proceeded. Sub-
sequent work fully substantiates this view.
The most important result of the second survey lay in this recognition of the fact that Patagonian
hake would be found to move with the seasons in much the same way as their better known European
relatives (allowing for the reversal of the seasons in the southern hemisphere).
INTRODUCTION 235
Apart from the direct distributional study, the sex, lengths and maturity stages of all hake captured
were recorded. The great difficulty of recognizing clear-cut stages in the development of the ovaries
of these fish first became apparent at this point. The difficulty of standardizing such observations on
this subject as are possible by direct inspection remained a handicap throughout. Probably it can
only be tackled by large collections for subsequent microscopic examination such as Hickling (1930ft,
1935ft) developed during his prolonged work on the European species. Hickling found it possible to
distinguish major stages by naked-eye appearance, but our observers agreed that this was never
satisfactory with females of the Patagonian species. These are only slightly smaller than the European
fish, but they often mature when considerably smaller, with a consequent increase in the amount of
overlapping of ovarian developmental stages among fish of the same length class. Hickling was
working upon a single species, but a limited staff investigating virgin ground could not make a large
enough systematic collection of ovaries without neglecting other essential work.
On the winter survey female hake were markedly more numerous than male. As in the autumn
the males were associated with the smaller females, and very few males were taken where large females
preponderated. The majority of the smaller females were immature. Very few of the large females
were ripening, and the majority seemed to be 'spent'. Among niales the proportion of immature to
mature fish was roughly 3:2. Very few were 'ripe and running'.
The hake were found to be feeding very largely upon Euphausiidae, Clupeafuegensis, Notothema spp.
and squids. Apart from hake, no fish were taken in such quantity as to be considered of possible
commercial importance. Notothenia ramsayi were fewer than on the autumnal survey, and the best
catches were taken on the same grounds as the hake. All were measured and sexed. There was a
marked preponderance of females. Cottoperca gobio seemed to favour the rich invertebrate area to the
south of the Falkland Islands, but was found elsewhere as well.
Some seven or eight species of Rajidae were observed. Raja brachyurops being the commonest.
Very full notes on the large range of variation to be found among the members of this group
were made These were of great value later, when our data were brought into line with Norman's
revised taxonomy of the Patagonian species. Salilota australis was taken quite frequently (fifteen
stations), mostly in the more northerly part of the area, but in small numbers. Specimens of three
species of rays and two species of Zoarcidae, all new to science, were obtamed during the second
survey. ,,-11 c
Three well-defined associations of invertebrates were observed, correlated with three types ot
bottom deposit. Along the edge of the shelf north of the Falkland Islands to 45° S a fine green sand
giving a clean bottom for trawling was prevalent in depths of 150-300 m. This ground was very rich
in the smaller invertebrates-small ophiuroids, echinoids, other echinoderms, Serohs, amphipods,
cumaceans and ostracods. Small quantities of Cephalodiscus occurred at seven of these stations. The
Falkland trough yielded similar results but without the vast numbers of small ophiuroids, in slightly
greater depths of water. Both these grounds yielded bigger quantities of hake and of rays than the
other areas worked during the winter. . . . , , . ,
On the continental shelf to the north, west and south-west of the Falkland Islands there was darker,
coarser sand, often brown, with pebbles and shells. Some patches were heavy with large invertebrates
such as sponges and molgulids, and others comparatively clean. The invertebrates were heaviest close
in to the Falkland Islands, and especially to the west of the Jason Islands. . ^ ^
The coastal waters to the south-west, south and south-east of the Falkland Islands again showed a
strikingly rich invertebrate fauna of sponges, hydroids, coralline hydroids, Alcyonana and Polyzoa.
This sheltered a rich fauna of smaller invertebrates: holothurians (especially Synapta), polychaetes
and small Crustacea. Here the bottom was of light green sand with pebbles, shells (many of pectens),
236 DISCOVERY REPORTS
large stones and boulders. It was found that this area extended over the 200 m. line for some distance
beyond the immediate coastal shelf investigated during the previous survey.
Mr John concluded his preliminary report (which has not been published) with the statement that
Mr Marr emphatically agreed with his opinion that ' any decision on a commercial fishery in and near
the waters of the Falkland Islands must depend on a greater knowledge of the seasonal variations of
the hake of the locality, and of waters farther north, and a full knowledge of what is known concerning
hake in other waters '.
The preliminary results of the first two surveys have been treated at some length to show how
great was their value when the third survey was planned. Several of the pioneers (notably Mr John)
were engaged upon other studies after finishing their share of the field work, and the value of their
efforts should be recognized. The third, most extensive and important survey, was carried out by
Gunther himself with the assistance of Mr G. W. Rayner. The results form a large part of the main
substance of this report. In this section, therefore, I have referred to the field observations only where
they modified the general picture previously gained of the conditions. The longer period available,
and increased experience, permitted improvements in working methods which must first be made
clear.
THE THIRD SURVEY
At Dr Kemp's suggestion the plan of the third survey was designed to include five lines of stations
spaced at regular intervals over the shelf between 44 and 54° S. Rough ground was to be avoided
when possible, but eflForts made to keep the stations uniformly spaced. Each line was planned on a
course of 111° — roughly normal to the coast. The isotherms here run almost parallel to the coast,
and the value of observations upon temperature and salinity is much increased if they can be made as
nearly as possible at right angles to the isotherms, as our hydrologists have frequently pointed out.
After the two northernmost lines of stations had been completed, permission was obtained still further
to extend the scope of the survey. Gunther did this by interpolating four lines of more closely spaced
trawling stations, and by making many additional observations, including a north-south line that pro-
vided valuable evidence of the effect of latitude. This tends to be masked (in the absence of such
evidence) by seasonal movements, the effects of increased depth, and so forth.
As finally carried out the programme included vertical hauls of the Gran international net (N50V)
with water samples from surface and bottom at all stations. The vertical nets provide evidence on
the distribution of fish eggs and phytoplankton, the water samples gave temperatures and salinity
determinations. The trawling stations were of three types :
(i) On the lines A, B, C, D and E (those of the original plan) the stations were 60 miles apart and
the procedure the same as on previous surveys, with the addition of the N50V.
(2) On the intermediate lines W, X, Y and Z, the stations were 30 miles apart, and in addition to
the normal routine the trawl was shot for a further period of four hours whenever fish seemed plentiful.
(3) On the additional lines the trawl was shot for one hour at stations 60 miles apart, the conical
dredge was not used and no repeat hauls were made (see Figs, i c, d).
Russell's bottom-net, with plankton nets towed obliquely on the same wire, was fished on six rather
irregular lines during passages to and from the trawling stations, and on two more lines of additional
observations where a small beam trawl was also used.
The observations made included some important additions to the standard practice of the first two
surveys. Among the essential routine observations the sorted fish were weighed in addition to being
sexed and measured, and the larger species of invertebrates were also weighed in addition to being
counted and listed. At selected stations whenever possible the important fishes, after being sexed and
INTRODUCTION 237
measured, were sorted into length groups for weighing. Notes on maturity stages of ovaries, stomach
contents and parasites of fishes were made as before. In addition, an attempt was made to assemble
collections of scales and otoliths from prescribed length groups of Merluccius and Macruronus.
Unfortunately, it has not yet been possible to work them out. It is known that they were too small to
provide conclusive evidence as to growth rate, but they will give a clue which may help to clarify
conclusions drawn from other lines of inquiry that must still be regarded as tentative.
A series of carapace-breadth and weight records of the centolla crab, Lithodes antarctiais, were
kept. These also await examination. Some biometric data on squids (body lengths and weights) were
recorded, and a lot of data on the numbers, size, sex and incidence of bopyrid parasitism of Miinida spp.
These have been utilized by Rayner (1935) in an important study of the growth of these Crustacea.
In their pelagic stage they are predominant among macroplankton animals of the more coastal waters
of the area (in due season), a fact observed by the earliest navigators of these waters (see also Matthews,
1932). Munida spp. are important as food for fishes, birds and whales.
The much greater scope of the third survey, especially in more northerly parts of the area, led
naturally to more and different kinds of fishes being met with. To the north such forms as Callo-
rhyncJnis caUorhynchus, Seriolella porosa and two soles were unfamiliar, and there also the centolla crab
was to some extent supplanted by the large red oxyrhynch Libidoclea sp., though at some stations
both were found. To the south our fauna list expanded as a result of more extensive observations in
the deeper waters of the Falkland trough, and in shoal water close to the mainland. Two macrurids,
a ray, Parana signata and Sebastodes oculatus were the principal additions here. Only two species
new to science were recorded from the trawling stations, although the third survey was more ex-
tensive than the first two combined, and carried out under better conditions with all the added advan-
tage of previous experience. By contrast, ten new species were found during the first two surveys.
This is good evidence that the gear and methods used were adequate to provide a general picture of
that portion of the fish fauna that can be sampled by trawling. The two new species discovered during
the third trawling survey were Raja midtispinis and Notothenia macrocephala (Norman, 1937, pp. 20,
68).
A most important addition to ideas gained from previous surveys was the discovery of rough
ground at the edge of the shelf in several places, whereas Mr John had been fortunate enough to find
fine, clean sand there. This rough ground produced striking examples of specialized distribution of
fishes found elsewhere in shallow rocky waters but not on the shelf between.
The weight records constituted a big advance, and permit of a much better general idea of fishery
prospects, the relative importance of potentially useful species, and probable breeding seasons. The
observations made by Gunther and Rayner with spring balances seem amazingly accurate, from the
consistency of ponderal indices {K) calculated from them. Hickling (1930Z), pp. 7-8) has also shown
that very good results can be obtained with weighings made on small ships in rough seas. With less
detailed studies in view, Gunther and Rayner were weighing several fish at a time, in length groups ;
not individual fishes and organs as weighed by Hickling. Since the individual lengths were almost
always accurately known, it is possible that for the calculation of broad mean values, such weighings
are better than individual ones.
A second innovation during the third survey, that of repeat or control hauls, gave valuable evidence
on the shoaling of fishes. They also showed that the nature of the bottom changes so gradually on the
shelf that little would be gained (from the viewpoint of such studies as these) by a closer spacing of
the stations.
During this survey Gunther recorded his opinion that ' the extreme scarcity of fish in the immediate
vicinity of the Falkland Islands is unquestionably due to seals. Seals were frequently met far from
23g DISCOVERY REPORTS
the coast and seem likely to affect the fishing over a much wider area. This raises the question whether
fish or sea-lions are the greater asset. While trawling on rough ground near the Falkland Islands would
be difficult, so that here the sea-lion would be the more remunerative, it is possible that if their numbers
were reduced fish farther afield might be turned to still greater account.' It is now known that sea-
lions feed largely upon cephalopods (Hamilton, 1934), but there is little doubt that they eat con-
siderable quantities of fish also. The cephalopods are themselves among the most important fish foods,
and the way in which sea-lions frequented areas in the open sea where fish were shoaling was remarked
by all who took part in the trawling surveys. The Falkland Islands are also the breeding grounds of
vast numbers of oceanic sea birds, many of which eat fish whenever they can get them. These, too,
probably play their part in keeping down the local fish population.
Gunther's notes also include some pertinent statements about the invertebrate fauna of the region :
'Lithodes and squids are of direct economic importance to the South American market; macro-
planktonic forms like Parathemisto and Munida spp. are among the important constituents of fish-
food ; the rich benthos appears to reflect the character of the sea floor, and is thus bound up with the
distribution of fish. Very full notes upon the quantitative distribution of invertebrates were therefore
made, but it will be impossible to make full use of them until systematists have revised the taxonomy
of the several groups.' 'The plankton of the area appeared to be poorer than that of corresponding
latitudes off the west coast of South America, and on the sub-Antarctic whaling grounds.'
Unfortunately, one has to add that, owing to the war, most of the collections, both of plankton and of
benthos, still remain to be worked up. The magnitude of this task may best be judged from a quotation
from the summary of Gunther's unpublished report on the work of the third survey:
'. . .the OTC was shot at 80 out of 131 Sts. and repeat hauls made at 24 Sts. BTS was shot at
II Sts., BNR at 46, oblique plankton nets at 50, and N50V at all Sts. Roughly 150,000 square miles
between the Patagonian Coast and the Falkland Islands, and from 44° to 54° S were examined. The
trawling stations were arranged in 11 lines and at a rough average were 50 mi. apart.' This report of
Gunther's includes memoranda upon visits to the fish markets of Montevideo and Buenos Aires, and
upon the history of the trawling industry off the River Plate, based upon the account of Devincenzi
(1926). Both have been of great help to me in writing this work.
Throughout the period of these investigations and for some years afterwards, our collections were
supplemented by specimens of littoral fishes obtained at the Falkland Islands by the government
naturalists Mr A. G. Bennett and Dr J. E. Hamilton. Both these officers worked in close collaboration
with the Discovery Investigations; indeed, Dr Hamilton was seconded for service with us for many
years, his chief work being on the bionomics of the sea-lion, which happily gave opportunity for many
incidental observations on shore fishes. The value of their work, especially Mr Bennett's, can best be
appreciated from the frequent references to it in the main part of this paper and in Norman's (1937)
systematic report on the coast fishes. Norman's work gives us the taxonomic foundation without
which this could not have been written. In addition to the Discovery material he was aided by various
small collections made on the mainland coast by workers to whom he has already made acknowledge-
ment. Norman also had access to all the material from the region already preserved in the national
collection.
TOPOGRAPHY OF THE SHELF
The locality and extent of the region surveyed have already been described. The names of the features
most useful for general descriptive purposes are shown in the general chart of the area (Fig. 2). This
chart also shows the arbitrary division into northern, southern and intermediate regions best suited to
the handling of the data available. Most of the names are those of salient features of the coast and
INTRODUCTION
239
I I
63°
GO"
45 ■
SO
Rierto Mddrun \ \^ T
Golfo NuevD
Northern Region
RiaCoig
-^l Grande Bai:^ t=^
^^^^^-APuertx) G alleg-os
S^ — w)) C Virgins
^ Sir.
Sebastian.
9t'
FALKLAND Is/-.
Southern
Kegi.DtS
■^
..• C.Horn
°Diego Ramirez
1;
^-^ Statenl.
N I I '
W"'
J>
BURBY^OOD BANK
I ...-
■•• 200m""
II I f* '-n-'-l i-r-r-r-
65°
60°
-I " ' ' ' t^' ■ ' ' I
Fig. 2. General chart of the area surveyed.
45
50
55
240 DISCOVERY REPORTS
require no comment. I have preferred the name ' Falkland trough ' for the broad tongue of relatively
deep water running north between West Falkland and the mainland, to the ' Falkland channel ' of
some earlier writers. In describing faunal distribution in this part of the world it is particularly
important to avoid such misunderstandings as could arise by indiscriminate comparison of this
natural feature with the numerous and straiter 'Magellan Channels' to the westward. The broad
term 'plain of the shelf has been used to describe the central portion of the area where the gradient
of the sea floor is exceptionally slight.
The topography of the shelf has been admirably described by L. Harrison Matthews (1934,
pp. 177-9) ii^ ^^is valuable account of the bottom deposits sampled by the conical dredge during the
trawling surveys. The sections shown by him (loc. cit., Fig. i) give ample illustration of the features
that exert most influence in a study of the distribution of the fish. These are:
(i) The extremely slight gradient from 80 m. right out to the shelf edge at the 200 m. contour (the
distance exceeds 200 miles in places). This is most marked in the north; farther south the sea floor
slopes a little more steeply, but it is generally true to say that most of the shelf lies below water of
more uniform depth than is to be found over comparable areas elsewhere. The great importance of
this fact in studying fish distribution lies in the difficulties that result in interpreting depth relations,
especially of migratory species. Elsewhere hake movements lead to clear correlation between size of
fish and depth of water at appropriate seasons. Here the depth gradient is so slight that evidence of
similar movements can only be detected when distance offshore is substituted for depth. The diffi-
culty was augmented by areas of very slightly shallower water offshore, notably off the Golfo San
Jorge and south-east of Puerto Deseado. These very slight elevations can hardly be termed 'banks'.
(2) The slope from the shelf edge to oceanic depths is very steep. In the north it is almost pre-
cipitous, so that trawling in water below 200 m. was limited to within a very few miles of the shelf
edge. Farther south trawling between 200 and 450 m. was possible over a wider area, but was difficult
owing to rough ground.
(3) The Falkland trough and the area of deeper water separating the Falkland Islands from the
Burdwood Bank are also well illustrated by Matthews's sections.
It is possible (for our present purpose) to make a slight improvement on Matthews's general chart
of the bottom topography (1934, pi. iii), by including several more recent soundings and plotting the
80 m. contour. This is shown in Fig. 3, which should be compared with Fig. 2 and with the separate
distribution charts when depth relations are under consideration. ,
The main body of Matthews's work dealing with the grading and distribution of the deposits
themselves is very instructive. The distribution of the coarser grades (mostly in the south of the area
and along the landward and seaward margins of the shelf) has obvious practical significance : coarse
grades, especially large fragments (loc. cit., pi. iv), usually coincide with foul ground for the trawler.
Matthews's main conclusion was that deposits became finer as one proceeded northwards (fine sand
and silt clearly predominate to the north), and that this is due to elutriation by the prevailing north-
ward flowing Falkland current, which may be said to act as a natural levigator. This is one reason for
the better trawling conditions found to the northward.
To extract the full benefit from Matthews's work one would need to study first the relations
between the distribution of the deposits and the sessile benthic fauna which is more directly affected
by the nature of the bottom than are most of the fishes. Unfortunately, as already stated, this colossal
undertaking can only be begun when taxonomic revision of the main invertebrate groups is achieved.
However, one very striking correlation between fish distribution and bottom deposits can already be
demonstrated. The abundant occurrence of the flatfish Thysanopsetta naresi was almost entirely
restricted to a central area of brown sand delineated by Matthews (loc. cit., pi. xii, area 'C').
INTRODUCTION
241
Fig. 3. Topography of the sea floor. (Depths in metres.)
3-2
242 DISCOVERY REPORTS
Thysanopsetta is one of the important forage species for larger fish, so that the restriction to a particular
type of bottom deposit, later described in detail, affords a good illustration of the potential value of
Matthews 's work.
HYDROLOGY
The water movements over the area covered by the trawling surveys are comparatively simple. Over
the whole of the plain of the shelf and eastwards beyond the shelf edge, relatively cold sub-Antarctic
water flows northwards in what is known as the Falkland current. To the east and north of the area
the warmer subtropical water of the Brazil current flows southwards, and in the region of the con-
vergence between these two (corresponding to the subtropical convergence in the open ocean still
farther to the east) hydrological conditions are more complicated. Here streams of sub-Antarctic
and subtropical surface waters may alternate, giving rise to large differences in salinity and tem-
perature within a few miles. Klaehn (191 1) was able to trace the southward movement of the Brazil
current as far as 49° S to the north-east of the Falkland Islands, but Deacon (1937, pp. 5S-9) has
shown that south of about 431° S, the subtropical water is becoming more and more mixed with sub-
Antarctic water. The influence of the Brazil current is strongest in summer, when relatively unmixed
subtropical water may extend some 4° farther south than in winter. The complicated conditions
around the southern extremity of the Brazil current rarely impinge on the trawling area, but this is
the probable route by which fishes that normally live in warmer seas occasionally reach the north-east
coasts of the Falkland Islands. The main facts concerning the southward limits of the Brazil current
have been ably summarized by Deacon (1937) and an attempt has been made to depict them in Fig. 4,
which should be regarded as a pictorial representation of the current system, and not an exact hydro-
logical study.
The Falkland current, which bathes most of the shelf, is composed of sub-Antarctic surface water.
Deacon (loc. cit., p. 51) has described how the main west-wind drift of the south Pacific is com-
pressed while passing through Drake passage. This augments its speed, and the sub-Antarctic portion
of It IS swollen by a relatively small amount of warmer, poorly saline coastal water flowing southwards
down the south-west coast of Chile. The resultant of these forces is the so-called Cape Horn current,
which IS really a local intensification of the west wind drift. It sometimes reaches a speed of as much
as 40 miles per day.
To the east of Staten Island the Cape Horn current divides in the form of the greek letter y the
esser branch swinging north round the Falkland Islands, but mainly between them and the main-
land, to form the Falkland current; and the main branch proceeding north-east and then east until it
merges into the main easterly (i.e. 'west wind') drift of the open ocean south of the Atlantic
The Falkland current itself flows most rapidly on its right flank, well oflFshore and beyond the shelf
edge, outside our immediate area. It is here that the coldest water is found, but the lower tem-
peratures are not caused solely by the greater speed of flow than that obtaining over the shelf. Both
Krummel (191 1) and Klaehn (191 1) postulate upwelling as an additional source of the cold water.
Where the Brazil and Falkland currents are flowing in opposite directions alongside each other it is
natural to suppose, as did Klaehn, that the dynamic disturbances so set up favour the creation of
whirls with consequent upwelling. Deacon (1937) considers that Klaehn 's demonstration (191 1, pi. 35
hg. 4) of isolated patches of relatively cold water towards the northern end of the Falkland current is
proof of upwelling, but our own data did not then provide any fresh evidence of its mechanism.^
^^\^^^^^:^^X^£^^^ -' ^- -"-d by the Dis^very Cc.iUee's
INTRODUCTION
243
Fig. 4. Current system of the area surveyed. Black stipple indicates cold currents, and red
indicates relatively warm currents.
244 DISCOVERY REPORTS
The Steep slope of the shelf edge is almost certainly one of the factors involved. These complications
in the coldest part of the Falkland current occur mainly to the eastward of the trawling area, and
therefore need not concern us further here.
Along the mainland coast the speed of the Falkland current is greatly reduced and close inshore
southerly movements of surface water may predominate. Consequently the water close to the main-
land is warmer than that over the plain of the shelf. A definite counter-current close to the beach is
set up in the summertime, which may flow (more intermittently) at other seasons also. These inshore
conditions are clearly indicated by the direction of the isotherms in Klaehn's charts (191 1, pi. 35).
The warmer inshore water does not result from any southward translocation of subtropical water,
like the Brazil current offshore. The latter begins to swing away from the land well to the north of
our area, usually in about 30° S. The warmer inshore water is formed by 'warming-up' of sub-
Antarctic water, owing to the slower rate of advance on the left flank of the Falkland current. It may
be described as ' old shelf-water '.
In the area of the trawling surveys there are only these two main hydrological features to be
visualized, the northward flow of the Falkland current over the plain of the shelf, colder and faster^
offshore along the shelf edge ; and the old warmer water close inshore with a more or less definite
southerly trend.
It is helpful in considering the distribution of fishes to gain some idea of the annual cycle of tem-
perature of the water. This enables some direct comparison to be made with conditions on better-
known fishing grounds elsewhere. Here we are handicapped by the fact that our three surveys were
made at different seasons in different years. Klaehn's very thorough averaging of results from many
years' observations, at a time when the region was much more important for traffic,^ provides the
general picture we need, though it deals with surface waters only. From a careful check of Klaehn's
charts against our own results it appears that the three years in which our own surveys took place
must have conformed closely to the 'average' conditions depicted by him. Comparing observed tem-
peratures obtained with the best modern apparatus on our surveys with Klaehn's monthly means,
we find a resultant mean error of -0-26° C. in autumn, -0-40° C. in winter and -0-07° C. in
summer. The range of error introduced by 'assuming' Klaehn's mean values instead of our own
would be -i-oi to +0-49° C, -1-35 to +o-8i° C, and -3-21 to +1-95° C. respectively. Some
error must be introduced by the time interval between our individual observations and the middle
of the month, the rest may be confidently attributed to diurnal variations, as indicated by the dis-
crepancy being greatest in summer. Klaehn worked on ten years' records from sailing ships (doubtless
mainly nitrate clippers) and four years' records of the German Admiralty. We know that two years
(1896-7 and 1906-7) out of the total studied by him were exceptionally cold, with icebergs drifting
far north in the Falkland current (Krummel, 191 1, p. 606, fig. 172). Such conditions are rare, and
this fact alone is probably sufficient to account for the slight tendency of Klaehn's values to be lower
than ours, especially in winter. More important still, a comparison between the observed differences
m surface temperature between successive pairs of our own observations, and the differences that
would be expected from Klaehn's results revealed a close correlation {r= +0-85, with P much less
than o-oi). It seems clear, therefore, that Klaehn's results give a very adequate picture of the ' average '
conditions, which will give a more satisfactory idea of the annual cycle of surface temperature than
any scheme of plotting our less numerous and more scattered observations with interpolations for
the gaps.
1 Some 13 sea miles per day according to Klaehn (191 1).
sequently""^"^ °^^" ^^"' °^ '^^ southern ocean, this region was far better known during 'the last days of sail' than sub-
INTRODUCTION
245
Graphs for the annual cycle have been plotted from Klaehn's charts in Fig. 5. The three positions
have been chosen arbitrarily on account of their approximation to the centres of the most numerous
of our observations on the fishes of the shelf, in each of the three main latitudinal regions into which
our stations may be grouped. The curves show that the temperatures in our 'Northern region are
very slightly colder than those found (say) on the hake grounds south of Ireland. The range is similar,
but the winter temperatures nearly i^ C. lower. The temperature cycles plotted for our ' mtermediate
and 'southern' regions show a decrease with increasing latitude, as is to be expected; while the three
curves together show the diminution in annual range as one proceeds southwards. This is also well
illustrated by Klaehn (1911, pi. 34, ' Jahresamplitude der Wassertemperatur')-
15-1
14-
13-
u
o
uj 12-
D
^ 1|.
u, 9-
u
■ c Northern Region
--X-- Intermediate Region
— Jk— Southern Region
<
in
8-
6-
JULY 1 AUG I SEPT | OCT | NOV | DEC | JAN | FEB | MAR \ APR ] MAY | JUNE |
Fig. 5. Representative annual surface-temperature cycles for the plain of the shelf, after Klaehn (191 1).
For consideration of the bottom temperatures, which are more significant in relation to our trawled
samples of fishes, we have only our own scattered observations to go upon, ^he actual %^^^^^^^^
eiven in Appendix I. It is obviously impossible to show typical seasonal curves, like those derived
from Klaehn's monthly charts of surface temperatures, owing to the small proportion of these obser-
vatLs that fall within any one depth range over a reasonably restricted area. A single, partly hypo-
Ih Ucal, curve for the intermediate region is shown m Tig. 6. With this as a rough guide it is possibl
to perc ive some important features by simple inspection of the figures quoted in the Appendix. The
seasonal cycle appears to be centred later in the year m the bottom water, with maximum tempera-
Tre^i: Mich. I autumn and the first half of the winter there is much less dif^^^^^^^^^
between surface and bottom water than at other seasons; it is rarely more than 2 C. whereas in
summer it usually amounts to some 4° C. This is due to intense mixing over the shelf m winter,
aradrdescribed'by Deacon (.937, P- SD- ^ spring and early summer even in the -rthern -^^^^^^
some temperatures below 5° C. were recorded (lower than the winter values), but most of these were
in fairly deep water well ofTshore.
246
DISCOVERY REPORTS
These bottom temperatures are particularly important in their relation to the most promising fish
of the region, the Patagonian hake. It is quite clear, from the general run of the figures, that these are
plentiful at some seasons in waters where the bottom temperatures are around 6° C. Their lower
hmit seems to be some 3° below the figure 87° C. which Hickling (1927, p. 67; 1928, pp. 81-2, 88)
found to be the normal cold limit of penetration of European hake. The ' stumpies ' of the Norwegian
deep are commonly found at lower temperatures, it is true; but Hickling 's later work provides ample
evidence that they are a local race, unrepresentative of the main stock, quite apart from the tem-
perature of the waters in which they are found (Hickling, 1930ft, pp. 50-1).
The bottom temperatures north and south of that part of the shelf represented by our 'partly
hypothetical cycle' seem to vary with the seasons in similar fashion. To the north they probably
'average' 1° C. warmer, and to the south perhaps 1° C. colder (certainly so in summer). The main
JULY I AUG I SEPT OCT NOV I DEC I JAN I FEB I MAR I APR I MAY I JUNE
Fig. 6. Bottom temperature on the shelf; partly hypothetical annual cycle for the intermediate region,
corrected to an assumed mean depth of 100 m., with surface cycle for comparison.
feature is the lag of at least a month after the surface maximum, and the relatively high autumn and
winter values due to mixing. For this reason the seasonal graph does not fall anything like so steeply
as does that showing the surface temperatures, from the maximum to midwinter.
Observations of the nutrient sah content of the waters of the Falkland current are not yet available
for this area.
PLANKTON
During the trawling surveys some phytoplankton samples were collected from the waters of the shelf
with the Gran net. They have not been fully analysed because of pressure of other work. Most were
obtained during the third survey. Mr Rayner made some preliminary observations on them (un-
published) and has recorded his impression that they indicate a poorer phytoplankton than that found
off the west coast of South America. Some of our early phytoplankton work in the ' Discovery H '
extended northwards, to the extreme south of the trawling survey area, and there we found some
INTRODUCTION ^47
evidence of a secondary autumnal increase in March 1930, with Rhizosolenia spp. dominant (Hart,
'^OurTatlr wo/k from 1933 onwards was based on results obtained by Harvey's (1934) method.
I have not previously dealt with our results from the sub-Antarctic Zone, because they were relatively
few and widely scattered, not lending themselves to the methods of presentation adopted in dealing
with the Antarctic material, which was the main object of study (Hart, 1942). Although most of these
observations are from oceanic waters, I think it profitable to discuss them briefly here since they
seem to me to give a good indication of the type of seasonal cycle (of standing crop) that may be
expected on the shelf. In November 1936 several observations were made on the shelf and in the
oceanic waters just beyond the shelf edge. These showed very little phytoplankton on the shelf but
some rich hauls offshore. . ..^u^f^.oc
The widely scattered observations available throughout the sub-Antarctic Zone, north of 54 b
are shown as a scatter plot in Fig. 7. No grouping of these observations by time intervals would
permit calculation of means that could be plotted to show the seasonal variation effectively^ Owing o
the'mall numbers of observations and their wide dispersion, the differences between the monthly
rneans for example, would not be statistically significant. Wide dispersion is a common source of
rifficuity in all 'samples' of quantitative estimations of phytoplankton, whatever method is adopted^
I Teerns certam that it is due to the extremely patchy distribution of these orgamsms m nature, though
1 meThods have limitations that may exaggerate this feature. Plotting the logarithms of the individual
observations instead of the actual numbers of units per m.^ overcomes another great difficulty due to
t^^^^^ of finding a scale on which all the observations can be shown. The observations
Iwn we not all obtained in the same year and, as already stated, they were scattered throughout
Llblntarctic Zone, but apart from the logarithmic plotting they have not been otherwise mam-
pulated fnty way. Moreover, nearly all the species involved have a completely circumpolar distn-
Kiitinn as in the Antarctic Zone to the southward.
tZ2Z shows clearly that the only months in which hauls exceeding looo umts per m.» occurred
weTe Nov:ltr Dec "bl and March' Further, .he proportion of very small hauls, be ow too unrts
were NovemDer ^ ^^^^^^^^ ^._^.|^_^ ^^ ^^^^ postulated for he
rnhtroctreTa^r: IfThe Ant^arctie Zone (Hart, .,4. PP- 307-8). but centred slightly earher
" wLr 1: Ts Manation of the low values recorded over the shelf in Novenrber .,36? The
'Tild not fake place un^ December in our southern locality. But cond.tions in the two areas are not
"' I " n the Antar ti 'the time of the main increase in oceanic sub-Antarct,c waters ts remark-
s'^(^hirrths' of the year, as compared 7;;-^-:^ rTm r^!:::
L:^^SytSe:rr:eitres:^^^^^^^^^
248
DISCOVERY REPORTS
5-R
CO
CO
CO
n
• •
V
^
rvj
to
-•-•— A-
lb
CO
CD
CD
• •'•
• Sxg'
CO
3^
CO
is t
; s CO
CO
CM
CO
O
O
O
O
ID
O
O
o
• 9.
• CD
O
o
o
o
[3ll^Hliyv90l]rrj y3d SUNG A3AaVH
c
o
c
o
3
I
3
rr-
c
li 1
o
m
^
ca
If 1
1-4
^
>
Z
nt
C
o
CO
bo
INTRODUCTION ^49
shelf waters sampled in November 1936. The fact that the temperatures diminish as one proceeds
offshore supports this view. There should be an earlier tendency towards establishment of a thermo-
cline, favouring earlier onset of the main diatom increase, under the warmer conditions prevailing
'"'itTs^thus probable that over the area of the trawling surveys the times of maximal production of
phytoplankton are only slightly later than in corresponding latitudes in European waters, and not so
much later as in the oceanic sub-Antarctic waters beyond the edge of the shelf. , , , .
The effect upon fishes of a slightly later timing of this basic element in the 'plankton-calendar of
the locality is likely to be a corresponding shift in their own dispositions, in so far as these are affected
by the plankton, for it is generally true that zooplankton maxima follow the maximal ' standing crop
of phytoplankton. Some of the evidence on this point has recently been well summarized by Bogorov
^' One of the most striking features of the zooplankton of the region is the swarming of lobster-kriU
(the Grimothea post-larval pelagic stages of Munida gregaria\ which occurs most frequently during
the summer months. The aduhs of this anomuran also swarm at the surface occasionally. The swarms
of post-larvae are often thick enough to impart a reddish appearance to the surface of the sea, observed
by many of the earliest navigators of these waters (Matthews, 1932, P?" 479-8;, several references)
The later phases of this species are limited to coastal areas round the Falkland Islands and along the
mainland coast, within our area (Rayner, X935, %• 18). A closely allied iorr.M. suhrugosa, is even
more widespre;d and abundant on the shelf (Rayner, i935, P- ^38, fig- ^D- The later stages of this
species are more strictly bottom dwellers. Munida spp. are of great importance in the ecology of the
region, as a staple food of whales (Matthews, 1932, PP- 481-3)- seals (Hamilton, 1934. p. 295), bird
and fihes M. gregarra is also found fulfilling a similar role in New Zealand waters (numerous
eferences quotef by Matthews). Other allied species of Anomura are equally -P-^-^/^^^^^^ Jf^^
Pacific off Mexico and southern California (information supplied by Capt. Fagerh and by Dr Waldo
L Schmidt of the Smithsonian Institute, quoted by Matthews, 1932, p. ^']^).
Rayner's detailed work on the growth of the Falkland species of M...W. shows that they are c--
paratLly long-lived (5 years or more of post-larval life in M. subrugosa) and are sexually matu e
rom the'end of the first year of post-larval life. The Antarctic krUl, EupH..^a ^^P^rt^;^^^^^^
a less specialized group, and pelagic throughout its life history, lives for a much shorter time^ The
di St X differences in pelagic life of the two regions may be affected by this difference in the If
histo" of their respective key-industry animals, as well as by the more obviously important physical
^' Euphausians are quite important constituents of the macroplankton over the shelf, but are obviously
les! important than'm the Antarctic. None have been observed to form ^^nse -rms^ ^^^^^^^^^^^
the surface of the sea, as E. superha commonly does in the Antarctic, and as Meganychphanes
tZtallT^ inermis) more rarely do in north European waters. Two or more
Southern p ies of Thysanoessa are to be found over the shelf, and of these T. gregana is the mo
rindant From Tohn^s (1936) work we know that of the genus Euphausia, E. vallenUm is the most
wilpread Z.^My the Ust numerous species around the Falkland Islands and in our southern
lion £ ZL E. Lus, E. longrrostris and E. triacantha also occur. Judging by its general dis
Sion tough^ut the sub- Antarctic Zone, E. luce, may be the most important species on the shelf
farther north but quantitative data are not yet available. , . if ^ fi^I. wprP
The ub nu ,ous hyperid amphipod Para,hen.hio gaudichaudii is abundant on .he shelf, and fi h were
faqundy found .o'have been'feeding upon h. Numerous -'anoid copepods oecur and „^a -
dance of large Scyphomedusae is frequendy referred to m the log books. The general
250 DISCOVERY REPORTS
zooplankton is indeed not unlike that of north European waters, but important differences may be
expected to appear, especially in the times of maximum abundance of the different groups, when the
material is fully worked up. As already explained, however, this must wait upon the completion of
taxonomic revision of the groups. Moreover, the environmental features of an area of this size, both
physical and biological, present almost unlimited scope for further investigation. In general we may
say that most of the more important macroplanktonic forms have a wide, often a circumpolar distri-
bution within the sub-Antarctic Zone. They are noticeably less restricted than the benthos in their
latitudinal range also, including a higher proportion of species that may extend into the distinct water-
masses both to the south and to the north of the sub-Antarctic Zone. Among microplankton forms
this wide tolerance is even more marked ; many of them are cosmopoHtan.
METHODS OF PRESENTATION
It has not been considered advisable to attempt to publish the biometric data on fishes in full, for the
raw data alone in manuscript form weigh over i cwt. These papers will be stored by the Discovery
Committee, and it is hoped that much information may still be gleaned from them.
Full details of the station positions are appended, with some of the abstracted data on the most
important.fishes. The numbers of fishes recorded are shown with station numbers only, in the general
distributional accounts of each species. Wherever it has been found profitable to consider length and
weight measurements these are given in the form of summarized tables and diagrams with legends
which It is hoped contain sufficient explanation. The data relating to different species varies so much
that It IS obviously impossible to adopt uniform treatment throughout.
Such simple statistics as have been ventured upon relate mainly to length-frequency distributions
and mean lengths. Where mean lengths have been used to compare or contrast catches from difl^erent
stations or groups of stations I have included sufficient information as to dispersion to enable the
significance of the differences to be determined. Simpson and Rowe's book (1939) has been a great
help m dealing with the numerical data.
The systematic arrangement and taxonomy of the fishes follow Norman's report (1937) throughout
In discussing bionomics of the fishes I have stressed differences and resemblances between Patagonian
species and those most nearly allied to them that have already been studied on better-known fishing
grounds elsewhere. I believe that it is easier for readers without experience of this particular area to
visualize the conditions if they are presented in this way. Since hake are the most important species
on the shelf, I have tested out the theories developed by Hickling during his prolonged work on the
iiuropean species, m so far as our scantier data on the Patagonian species permit. With the possible
exceptions of cod, haddock, herring and plaice, Hickling's work on hake probably constitutes the most
complete picture we have of the life history of any marine fish. Consequently all members of the
Discovery staff who have been concerned in this work have studied Hickling's work intensively
Iwo conventions have been used to facilitate general descriptions offish distribution, and to ensure
that seasonal comparisons should be kept roughly comparable: .
First, the whole area has been arbitrarily subdivided into ' northern ', ' intermediate ' and ' southern '
regions, according to latitude. The interval chosen was 4° of latitude, so that the northern region
includes all trawling survey stations north of 46° S, the intermediate region all those between 46 and
50 b, and the southern region all those south of 50° S. This subdivision has already been indicated
mtig. 2. It IS an arbitrary division introduced solely for the purpose of reducing the data to manage-
abe proportions, but it approaches a natural division inasmuch as the physical conditions that change
with latitude afl^ect the flora and fauna. Provided that large-scale migrations in a north and south
GENERAL ACCOUNT OF THE FISH FAUNA 251
direction are not involved, observations at different seasons within any one of these regions are
obviously more fairly comparable than those over the whole area. In the specific distribution lists,
the regions to which individual stations belong are indicated by printing the numbers of northern
stations in heavy type, intermediate stations in ordinary type, and southern stations in italics.
Secondly, the reversal of the seasons in the southern hemisphere is indicated by beginning all time
scales on i July, comparable to i January in the northern hemisphere. It should also be remembered
that when the looser seasonal expressions 'spring', 'summer', etc., are used, a similar reversal is
'Twing to the shght gradient on the shelf, distance from the mainland coast is more significant in
relation to seasonal changes in distribution than changes in depth of water inhabited by migratory
species This figure has been calculated (in sea miles) for all the stations and is tabulated m
Appendix I Conditions round the Falkland Islands themselves are peculiar, and they do not appear to
be regarded as a coast at all by the main migratory species, which are almost absent f»-om their
immediate vicinity. I have therefore calculated the distance from the nearest point of the Falkland
Islands also, for those stations that fall within 100 sea miles of them. Beyond that distance migratory
fishes seem to proceed towards the mainland (in due season) as though the Falkland Islands did not
exist even though the mainland may be twice as far distant. The exact distances from the Falkland
Islands of all other stations has not, therefore, been calculated; they are tabulated merely as more
than 100 miles'. • ,• ■ r>-
Abbreviated descriptions of gear are the same as those used throughout the station lists m Discovery
Reports. The meanings of those used in this work have already been given m the section on field
methods.
GENERAL ACCOUNT OF THE FISH FAUNA
A list of all the species of fishes recorded on the Patagonian Continental Shelf and immediately
adjacent coasts is given in Table i. This also shows which species were obtained during the surveys
in ' Trawl + accessory nets ', in ' Other gear ' and by shore parties. Most of the specimens recorded in
the last category were obtained by Mr A. G. Bennett and Dr J. E. Hamihon. I have not included
some recent records of Pozzi and Bordale (1935) which are mentioned by Norman (1937, PP- HS-^)-
Some of these require confirmation. The list follows that of Norman (loc. cit., pp. 143-5) but omits
species recorded only from the west coast (columns A and B of Norman's list). The historical aspect
of the growth of our knowledge of the fish fauna is admirably dealt with by Norman (1937, PP- i37-4^).
and from his account the debt we owe to earlier expeditions-British, French, Swedish and American
— can be assessed. . . r
From Table i it can be seen that ninety-five species are recorded from the region, and specimens ot
seventy-eight of these were obtained by ships or shore parties in the course of the Discovery investi-
gations. Also an undoubted basking-shark was observed. The records of two of the species that we
did not capture seem somewhat doubtful: Bunocottus apus Kner is based on a single specimen said to
have come from the Burdwood Bank, no other Cottidae are known from the region and Kner s
description does not agree with his figure (cf. Norman, 1937. P- HS)- Alphestes afer (B och), a small
serranid common in the West Indies, has a normal range extending from Cuba to Brazil (Jordan and
Eigenmann, 1890, p. 35°). It is clearly a tropical species, so that its occurrence south of 42 ^ would
be most extraordinary, and I am unable to trace the authority upon which Norman placed it in the
Patagonian list. Among the other species not taken on the surveys, Notothenia tngramma Regan and
Crossostomm fasdat^^s (Lonnberg) are known only from their unique holotypes, and it is possible that
252
DISCOVERY REPORTS
Table i . Fish fauna of the Patagonian Continental Shelf
Taken in
' Trawl +
accessory
nets'
Taken in
Taken by
Common names
Family
Species
'Other
shore
adopted in
gear'
parties
this report
Petromyzonidae
Geotria australis Gray
—
—
X
—
Myxine australis Jenyns
X
X
X
Hagfish
*M. affinis Guntherf
—
—
—
—
Lamnidae
*Cetorhinus maximtis (Gunner)|
—
—
J.E.H. obs
Basking shark
Scyliorhinidae
* Scyliorhinus (Halaelurus) bivius (Smith)f
—
—
—
Carcharinidae
*Mustelus cams (Mitchill)
—
—
—
■ Dogfish
*Centroscy Ilium granulatum (Giinther)
—
—
—
Squalidae
Squaliis lebruni (Vaillant)
X
—
—
^
Torpedinidae
Discopyge tschudii Heckel
X
—
—
V
Rajidae
Raja flavirostris Philippi
X
—
—
R. doello-juradoi Pozzi
X
—
—
R. macloviana Norman
X
—
—
R. magellanica Steindachner
X
X
—
R. multispinis Norman
X
—
—
Rays
R. scaphiops Norman
X
R. brachyurops Fowler
X
X
—
R. griseocauda Norman
X
—
—
Psammobatis extenta (Garman)
X
P. scobina (Philippi)
X
X
—
Chimaeridae
Callorhynchus callorhynchus (Linnaeus)
X
—
—
—
Clupeidae
Clupea fuegensis Jenyns
X
X
X
'Herring'
C. arcuata Jenyns
X
—
—
' Sprat '
Galaxiidae
Galaxias attenuatus (Jenyns)
X
*G. maculatus (Jenyns)
—
—
*G. smitliii Regan
—
—
Aplochitonidae
Aplochiton zebra Jenyns
—
—
X
*A. taeniatus Jenyns
—
—
—
Syngnathidae
Leptonotus blainvilleanus (Eydoux and Ger\'ais)
X
—
—
Pipefish
*Entelurus aequoreus (Linnaeus)
—
—
—
—
Macruridae
Coryphaenoides holotrachys (Giinther)
X
—
—
Coelorhynchus fasciatus (Giinther)
X
—
—
Merlucciidae
Merlucciiis hubbsi Marini
X
X
Hake
Macruronus magellanicus Lonnberg
X
'Long-tailed hake'
Gadidae
Micromezistius australis Norman
X
Salilota australis (Giinther)
X
Physiculus marginatus (Giinther)
X
—
Muraenolepidae
Muraenolepis microps Lonnberg
X
—
—
M. orangiensis Vaillant
X
Lamprididae
Lampris regius (Bonnaterre)
X
Serranidae
*Alphestes afer (Bloch)
—
Carangidae
Parana signata (Jenyns)
X
Cheilodactylidae
Cheilodactylus bergi Norman
—
X
Bovichthyidae
Cottoperca gobio (Giinther)
X
X
Bovichtus argentinus MacDonagh§
X
Nototheniidae
Notothenia macrophthalma Norman
X
*A^. trigramma Regan
—
—
—
—
N. canina Smitt
X
X
A^. jordani Thompson
X
X
A^. tessellata Richardson
X
X
X
A^. brevicauda Lonnberg
X
X
A^. guntheri Normann
X
X
,
A'^. ramsayi Regan
X
X
N. wiltoni Regan
—
X
X
,
A'^. squamiceps Peters
—
X
—
—
N. sima Richardson
X
X
N. cornucola Richardson
"
X
X
A'^. elegans Gunther
X
X
—
GENERAL ACCOUNT OF THE FISH FAUNA
253
Table i {continued)
Taken in
' Travel +
accessory
nets'
Taken in
Taken by
Common names
Family
Species
' Other
shore
adopted in
gear'
parties
this report
Nototheniidae
Notoihenia macrocephala Giinther
—
X
X
—
*N. microlepidota Button
—
—
—
—
Dissostichus eleginoides Smitt
X
—
—
—
Eleginops maclovinus (Cuvier and Valenciennes)
—
X
X
'Mullet'
Harpagiferidae
Harpagifer bispinis (Schneider)
X
X
X
—
Chaenichthyidae
Champsocephalus esox (Giinther)
X
X
X
—
Gempylidae
Thyrsites atun (Euphrasen)
X
X
X
—
Scombridae
Gasterochisma melampus Richardson
—
—
X
—
Zoarcidae
Iluocetes fimbriatus Jenyns
X
X
—
—
/. elongatus (Smitt)||
X
X
—
—
Austrolycus depressiceps Regan
—
—
X
—
*A. laticinctus (Berg)t
—
—
—
—
Phucocoetes latitans Jenyns
X
X
X
—
*Crossostomus fasciatus (Lonnberg)
—
—
—
—
Pogonolycus elegans Norman
X
X
—
• —
Platea insignis Steindachner
—
X
—
—
*Maynea patagonica Cunningham
—
—
—
M. brevis Norman||
X
—
—
—
Melanostigma microphthalmus Norman
X
—
—
—
Ophidiidae
Genypterus blacodes (Schneider)
X
X
—
—
Brotulidae
Cataetyx messieri (Giinther)
X
—
—
—
Centrolophidae
Seriolella porosa Guichenot
X
■ —
—
—
Palinurichthys caeruleus (Guichenot)
X
—
—
—
P. griseolineatus Norman
X
—
—
—
Stromateidae
Stromateus maculatus (Cuvier and Valenciennes)
X
• —
—
' Spotted pomfret '
Atherinidae
Austromenidia smitti (Lahille)
—
X
X
} 'Smelt'
A. nigricans (Richardson)
—
—
X
Scorpaenidae
Sebastodes oculatus Cuvier and Valenciennes
X
—
—
—
Congiopodidae
Congiopodus peruvianus (Cuvier and Valenciennes)
X
—
—
—
Cottidae [?!]
*Bunocottus apus Kner [?!]
—
—
—
—
Psychrolutidae
Neophrynichthys marmoratus Gill
X
X
—
■ —
Agonidae
Agonopsis chiloensis (Jenyns)
X
X
■ —
—
Liparidae
Careproctus falklandicus (Lonnberg)
X
—
—
—
*Liparis antarctica Putnam
—
—
—
—
Bothiidae
Thysanopsetta naresi Giinther
X
X
—
'Scald fish'
Paralichthys isosceles Jordan
X
X
—
Xystreurys rasile (Jordan)
X
—
—
Mancopsetta maculata (Giinther)
X
• —
—
'
Achiropsetta tricholepis Norman
X
—
X
* Species not taken during the surveys.
f Species not taken in this area, but specimens obtained by the expedition elsewhere. \ Observed, but not taken.
§ Taken by Mr MacDonagh at Puerto Madryn. || Species taken only in 'Accessory nets' when trawled.
the latter may prove to be a young example of Austrolycus depressiceps Regan (Norman, 1937, p. 106).
Specimens of three species, recorded from the shelf but not taken there by us, were obtained on other
occasions among the channels to the westward.
Of our seventy-nine 'shelf species, sixty-one were obtained in the ' Trawl + accessory nets', thirty-
five in ' Other gear' (including ten that were not trawled), and twenty-three by shore parties (including
seven not taken by other means), the total being made up by the basking shark seen by Dr Hamihon.
Thus we see that of the ninety odd species recorded from the region, two-thirds were obtained by
trawling. The remainder are mainly littoral fishes; for example, eight of the notothenias and two
zoarcids have never been taken in water deeper than 46 m. (25 fm.). The difficulties of obtaining
littoral fishes in this area are great : the tidal range is small and the surf often heavy, while the great
254
DISCOVERY REPORTS
Table 2. Fish fauna of the Patagonian Continental Shelf (trawled fish only). Total numbers
taken and frequency of occurrence
I
Numerical
Total nos. of
Frequency of
Frequency of
class
Species
fish taken
occurrence in
178 hauls
occurrence as
percentage
> 1000
Notothenia ramsayi
9599
130
73-0
Merlitccius hiihhsi
5748
109
6l-2
Macruronus magellanicus
5336
63
35-4
Thysanopsetta naresi
1916
33
i8-5
Stromateus maculatus
1044
51
287
100-999
Clupea fuegensis
694
28
157
Micromezistius australis
557
21
II-8
Salilota australis
485
65
36-5
Cottoperca gobio
414
50
28-1
Raja brachyurops
274
48
27-0
Notothenia guntheri
267
11
6-2
Psammobatis scobina
153
45
25-3
Coetorhynchus fasciatus
140
12
6-7
Champsocephalus esox
124
15
8-4
Notothenia tessellata
102
9
5-1
50-99
Coryphaenoides holotrachys
95
8
4-5
Genypterus blacodes
74
33
i8-s
Raja doello-juradoi
68
12
6-7
R. magellanica
54
21
II-8
Congiopodus peruvianus
5°
16
9-0
25-49
Agonopsis chiloensis
47
IS
8-4
Thyrsites atiin
45
6
3-4
Raja flavirostris
41
24
13-5
Clupea arcuata
39
I
0-6
Iluocetes jimbriatiis
38
22
12-4
Parana signata
35
2
II
Notothenia canina
30
5
2-8
Callorhynchus callorhynchus
28
6
3-4
Notothenia jordani
25
3
1-7
Neophrynichthys marmoratiis
25
7
3-9
1-24
Myxine australis
24
7
3-9
Raja macloviatia
23
12
6-7
Physicidus marginatus
22
4
2-2
Dissostichus eleginoides
9
5
2-8
Squalus lebruni
8
6
3-4
Seriolella porosa
8
I
0-6
Raja griseocauda
8
6
3-4
Paralichthys isosceles
8
5
2-8
Raja scaphiops
7
5
2-8
R. albomaculata
7
6
3 "4
Notothenia elegans
7
S
2-8
Iluocetes elongatus
6
0-6
Careproctus falklandica
6
06
Disco pyge tschudii
4
0-6
Leptonotus blainvilleanus
4
0-6
Phucocoetes latitans
4
0-6
Maynea brevis
4
4
2-2
Xystreurys rasile
4
3
1-7
Harpagifer bispinis
3
I
0-6
Palinurichthys griseolineatus
3
3
1-7
Sebastodes oculatus
3
3
17
Pogonolycus elegans
2
2
i-i
Melanostigma microphthalma
2
' 2
i-i
Cataetyx messieri
2
2
i-i
Palinurichthys caeruleus
2
2
I-I
Raja multispinis
0-6
Psammobatis extenta
06
Muraenolepis microps
0-6
M. orangiensis
0-6
Notothenia macrophthalma
0-6
Mancopsetta maculata
0-6
Achiropsetta tricholepis
0-6
255
GENERAL ACCOUNT OF THE FISH FAUNA
beds of kelp {Macrocystis, Durvillea, etc.) may extend out to 30 fm. and render any form of fishing
difficult, though they certainly harbour many species of fish.
The trawl seems an adequate sampling instrument within the limits set by the mesh used. Hicklmg
(i933' PP- I ^-^9) ^^^ g^'^^^ adequate demonstration of this, and our results with a closely allied species
of hake (pp. 284-9) conform with his as closely as could reasonably be expected. Faulty hauls cannot
prove absence or relative abundance, but have helped to prove presence of certain species on some
occasions. The trawl cannot provide adequate data for pelagic species such as herring, or very small
species such as some Zoarcidae. Immature specimens of some of the larger, more important species
also escape through the meshes. This loss is most serious with the more slender forms, such as very
young hake (especially males), immature Macrurontis, Mkromezistim and Genypterus. The fine-
meshed nets attached to the back of the trawl caught enough of these to enable us to outline their
probable distribution, but do not provide comparable quantitative data. Provided that these limita-
tions are borne in mind. Table 2, which gives the total numbers of the species in ' Trawl + accessory
nets', and their frequency of occurrence, helps to extend the outline of our general picture of the fish
fauna begun by the first table.
The fish fauna of the Patagonian shelf is not rich in species, as the full list in Table i shows. In a
preliminary account of the trawling surveys,^ Gunther pointed out that it is less than one-third of the
strength of the British list, and nearly twice as many species occur in the Gulf of Maine (Bigelow and
Welsh 1925) Apart from the numbers of species, there are big qualitative differences from the types
of fish' faunas known from other parts of the world. The marked predominance of the percoid group
Nototheniiformes, with four families, seven genera and twenty-one species from the area of the surveys,
and relatively large number of Zoarcidae with eleven species representing eight genera, are a most
peculiar feature. Among elasmobranchs Rajidae show many species and remarkable diversity for an
area where nearly all the known changes in environmental conditions are gradual. Dogfish are not
common on the shelf, and the numbers of species of true codfishes (Gadidae) is small; in these
features we see a great contrast to the fish faunas of northern Europe and of the New England states.
Table 3 showing the relative strengths of some of the important groups in British seas, on the Pata-
gonian shelf and in the Gulf of Maine, summarizes these points. The comparison between British
and Patagonian fish faunas was first made by Gunther in the preliminary account mentioned above,
and I have abstracted figures from Bigelow and Welsh (1925) for the Gulf of Maine to make the
comparison wider.
Table 3. Relative proportions of certain taxonomic groups in British
seas, on the Patagonian shelf and in the Gulf of Maine
No. of spp.
Rajidae
Percomorphi
Gadidae
Heterosomata
British seas
350
5% ..
31%: No Nototheniiformes
Zoarcidae i sp.
8%
Patagonian shelf
95
10%
48%: Nototheniiformes 21 spp.
Zoarcidae 11 spp.
3%
5%
Gulf of Maine
173
3*°'
27%: No Nototheniiformes
Zoarcidae 3 spp.
7%
-710/
72 /o
The Nototheniiformes are, of course, an essentially southern group, but it has further to be noted
that most of the Patagonian species are distinct from the Antarctic ones; and the number of species
comln to other sub^Antarcdc localities, such as the Antipodes, is small. The relations between the
1 'A Fishery Survey of the Patagonian Continental Shelf read before Section D of the British Association, July 1938-
256
DISCOVERY REPORTS
Patagonian fish fauna and that of Kerguelen and Heard Island were dealt with by Regan (1914, p. 36),
who treated the latter as a peripheral district of the Antarctic Zone. Norman (1937, p. 148), reviewing
the question with more recent evidence, reached the conclusion that the dissimilarity between the
fish faunas of the Kerguelen and Patagonian regions was not so great as Regan had supposed, and
pointed to several pairs of species of the closest phylogenetic relationship from the respective regions.
However, it is easy to demonstrate similar close relationship between Patagonian species and others
from very widely remote regions, as we shall presently show. Norman himself, in a later work (1938,
pp. 100 et seq.), summed up the present position in regard to this question with the statement: 'it is
clear that, although the coastal fish fauna of the Kerguelen district shows certain features of resem-
blance to that of the Patagonian region and the Antipodes, its affinities are mainly with that of
Antarctica.' This is precisely what one would expect from our latest knowledge of the hydrology of
the regions concerned (Deacon, 1937).
In strong contrast to the marked differences between the Patagonian fish fauna and that of other
regions, some of which have been shown in Table 3, several important and familiar species from
better-known grounds can be 'paired-off' with Patagonian species closely allied to them. This feature
was first made clear in Gunther's unpublished work, and was subsequently expanded by Norman
(i937» P- 146), and by Gunther himself in his address to Section D of the British Association during
the following year. These workers were concerned to show the parallels between allied British and
Patagonian species, as an aid to general description of the fauna. I have attempted to widen the basis
of this comparison with further parallels from the Gulf of Maine and from South Africa. Naturally
there are fewer closely allied species from such widely diverse regions. South African waters are
subtropical, though with low temperatures in the Benguela current, and the Gulf of Maine is a very
specialized ' cold-temperate ' area with exceptionally high summer temperatures due to the influence
of the Gulf Stream. Nevertheless, I find these broad comparisons helpful in gaining an idea of the
character of the Patagonian fish fauna, especially in conjunction with the roughly quantitative work
to be described later. I therefore include them here in the hope that those who read this report may
similarly be aided.
Table 4. Closely allied species from the Patagonian shelf, British seas,
the Gulf of Maifie and South African seas
Patagonian shelf
British seas
Gulf of Maine
South African seas
Myxine australis
Squabis lebruni
Raja flavirostris
Clupea fuegensis
C. arcuata
Merluccius hubbsi
Micromezistius australis
Salilota australis
Sebastodes oculatus
Stromateus maculatus
M. glutinosa
S. acanthias
R. batis
C. harengus
C. sprattus
M. merluccius
M. poutassou
Urophycis blennioides
Sebastes marinus
M. glutinosa
S. acanthias
R. stabuliforis
C. harengus
M. bilinearis
Microgadus tomcod
U. tenuis et spp.
Sebastes marinus
Poronotus triacanthus
M. capensis
S. acanthias
R. batis
M. capensis
Sebastichthys capensis
Stromateus fiatola
It IS clear from this table that quite a number of Patagonian species have close relatives elsewhere,
but when the quantitative aspect is taken into account we find that with the exceptions of the hake,'
Micromezistius and Stromateus,^ the Patagonian species listed are relatively far less numerous than are
their nearest counterparts on the better-known grounds we have considered. Thus the differences
1 Also the Falkland herring, but this could not be adequately sampled by the trawl.
GENERAL ACCOUNT OF THE FISH FAUNA 257
between the Patagonian fish fauna and that of these better-known fishing grounds outweigh the
resemblances, and .f we are to succeed in descnbing the general facies of fePatagoman fauna by
analogy with that of other regions it will be necessary to cast our mmds still farther afield. The differ-
ences have already been summarized (in part) in Table 3. Four notable ones are: predominance of
Nototheniiformes the relative unimportance of Gadidae and of flatfish, and an absence of Salmomdae
of useful size (which was not previously mentioned). In all these respects the Patagonian fauna differs
markedly from that of better-known fishing grounds in the northern hemisphere. It is also most
Unfortunately true that relative scarcity and small size of flatfishes and Salmomdae are features
common to all the other fishing grounds of the southern hemisphere. Consider, then, an area in the
northern hemisphere which we have not yet taken into account-the North Pacific. If we allow
ourselves to imagine this fauna without its two best fishes, salmon and halibut, what would be its
emating characteristics? Predominance of Scorpaemdae and allied families, especially Hexagram-
mle (with Ophiodon elongatus, the cultus cod) and the Cottidae or sculpins. Herring would be
Cort nt, but though Gadidae would be fewer than in other parts of the northern hemisphere they
n s'ili rank high The relative importance of Merluccius productus, a true hake hitherto despised
on that favoured coast, would be much enhanced. • , . .
Now Jordan (,,05, vol. „. pp. 50. e. seq.) has pointed to the c ose analogy wh.eh ex.sts between
the exdusnely southern Notothemiformes and certain families of the great group of ma.l-cheeked
fishes whch L calls Pare.oplitae (including the Hexagrautmidae, etc.). Of the Notothenndae he
retarklAeir general resemblances to small Hexagrammidae •, and a little later he speaks of them as
■ he n ipod s of the Cottidae and Hexagramm.dae ; although lacking the bony stay of the latter, they
show severa analogical resemblances and have very similar habits', and aga.n ' . . ^Harpagrtendae
resemb e u pins even more closely '. To th,s may be added the close superfiaal resemblances between
o hewia nt Antarcttc) Nototheniiform fishes (Ba.hydraconidae and Chaemch.hydae) and ye other
?r,;ilsofth'emai.,cheked«..,.^^^^^^^
Z ^:: ^fi^tt: ■rhrN^JSrl^alue biol„g,ca1 coLerpart of the Notothen.iform
draw between uie r g a r f^.^oP^.n Micropadus znA Micromezistius, Merluccius productus
callorhynchus, Clupea pallasei and C. fuegensis, Microgaaus an Sebcutodes is of
Uckthy. spp.) m S™* Am ™^7*J.i, „„ait ons, and in colder waters the scarcity of flatfishes
:rj:f:Ui:ri:Tn-Uhrrngtn"he southern hem.sphere. Sizeable Salmonidae are also unknown
258 DISCOVERY REPORTS
in the southern hemisphere (apart from introduced trout) so that these two deficiencies are by no
means peculiar to the Patagonian shelf. The most encouraging features, more apparent when the
quantitative data are considered, are that a true hake is common on the Patagonian grounds, that
minor quantities of another merlucciid, a butterfish or pomfret, and lesser numbers of other species
are all good eating, and that there is a Falkland herring which may prove to be of real value. Though
our gear was admittedly unsuitable for sampling this last species, its abundance is placed beyond doubt
by the frequency with which it was observed in the stomachs of larger piscivorous species.
The Patagonian fish fauna has certain species in common with other southern hemisphere localities,
and it may be thought that in attempting to draw descriptive parallels it would have been better to
turn to these rather than to the northern hemisphere. I did not attempt this because the essentially
subtropical conditions of the other southern localities leads to the prevalence of various percoid groups
very different from the Nototheniiformes, so that although New Zealand for example has some species
m common with Patagonia, and a few others closely allied, the general character of the vastly richer
fish fauna is ahogether different from that of the Patagonian one. This remark applies with even
greater force to the fish faunas of Southern Australia and South Africa. It is mainly in the scarcity
or absence of certain groups, as Salmonidae and Gadidae, that the fish faunas of these areas can be
said to show any resemblance to that of Patagonia. The extent to which we are forced back to the
northern hemisphere for closer parallels is a measure of the extent to which meteorological factors,
chiefly temperature, determine the conditions of life in the sea.
It IS a striking fact that a true hake is found in each of the regions we have discussed : Merluccius
merluccius to the west of Great Britain, M. bilinearis in the Gulf of Maine, M. productiis on the Pacific
coast of North America, M. capensis off South Africa, and M. hubbsi on the Patagonian shelf. Three of
these five species are already heavily exploited: prior to the war Hickling's work had shown (1935a)
that the European stock was being overfished, while in the last twenty years M. bilinearis has risen
from the status of 'rubbish' to the New England fishermen, to the highest place among the frozen
fish products of the eastern states. Moreover, only half the catch is frozen. M. capensis forms a third
of the catch of the trawling industry at the Cape, where large quantities of it are salted and dried.
M. productus is not yet sought after. In a region where better fish are still abundant the softness of
Its flesh makes it unpopular. In less favoured localities the other species of hake, probably little
better m this respect, are valued and are successfully marketed, owing to improved methods of pre-
servation and storage. This is especially true of M. bilinearis, half of the catch being sold inland in the
form of frozen fillets, etc. This species is the one most nearly allied to the Patagonian M. hiibbsi.
Such small quantities of the last named as are caught by the small trawlers which operate from the
mouth of the River Plate, well to the north of our area, fetch prices well up to the average in the
Buenos Aires market.
In summarizing the points dealt with in this section of the report, we may say that the fish fauna of
the Patagonian shelf is peculiar in quality and in the small number of species to be found there The
number of potentially valuable forms is small too, and their quantity not encouraging, but hake are
moderately abundant. The nearest parallel among fish faunas of better-known regions is probably
that with the North Pacific, but with important reservations detailed above. There is also some
resemblance to the conditions found on the hake grounds to the west of the British Isles.
259
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
PETROMYZONIDAE
A specimen of Geotria australis Gray was obtained by Dr Hamilton in the Falkland Islands, but none
was taken during the trawling surveys. Possibly the marine phase of the life history does not extend
far beyond the littoral zone, but the general distribution of the species appears to be circumpolar in
sub-Antarctic waters, and even extends to subtropical waters in AustraUa.
MYXINIDAE
Three 'good' species of the genus Myxine were recognized by Norman (1937, pp. 4-7) from the
Patagonian region. One of these, M. tridentigera Garman, was not obtained by the Discovery Com-
mittee's ships. It is known only from the unique holotype from Magellanes. Four specimens of
M affinis Gunther were taken at St. WS582 in a dip-net, but none from the area of the trawlmg
surveys The remaining species, M. australis Jenyns, seems to have a much wider distribution. It is
common at the Falkland Islands, where specimens have been collected by Bennett and HamiUon, on
both coasts of South America and particularly in the Magellan channels. A specimen has also been
collected by Hamilton at the South Shetlands, a fact of exceptional interest, as I am not aware of any
other record of a cyclostome being taken south of the Antarctic convergence. Details of our obser-
vations on the distribution of this species are :
I : in ' Trawl + accessory nets '
Mean depth
Station
WS763I
WS789
WS792 A
WS797C
Myxine australis Jenyns.
Mean depth
m.
84
94
104
112
Numbers
I
7
I
10
Station
WS812 I
WS833
WS834
Numbers
Station
Port Stanley (J. E. H.)
Salvador waters (A. G. B.)
Port Stanley (A. G. B.)
WS835
WS836
II: in 'Other gear'
Mean depth Numbers
m.
?
?
?
15
64
etc.
I
I
I
6 in BTS
I in BTS
Station
WS856
WS871
WS873
WS878
S3
34
32
Mean depth
m.
104
338
93 (-°)
121 (-0)
Numbers
etc.
I in BTS
I in BTS
I inNR
5 inNR
Thus although the majority of specimens were captured in shallow inshore waters the species has
been taken below 300 m., and at four of the remaining fifteen stations the depths slightly exceeded
100 m It also extended much farther north on the east coast than the other species, a considerable
concentration being found off the Golfo san Jorge. The other records are mostly grouped much
farther south, off the mouth of Magellan Straits and the north-east coast of Tierra del Fuego There
is no obvious reason for such a discontinuous distribution, but the gear used is obviously far from
efficient for the capture of such slender organisms, and as it was not obtained in l^'-f ^.^^ers any-
where, insufficient sampling is quite possibly the sole cause. Apart from the probability that M
australis will be found to attack useful fish, especially those that come well inshore, it is of no potential
economic significance. LAMNIDAE
Cetorhinus maxirmu (Gunner). No basking sharks were seen or captured by our ships within the
trawling area. Norman (1937, P- V) gives the details known concerning one washed up in East Falkland
and quotes HamiUon's observations on another seen off Cape Dolfin m 1936. He notes that it may
prove to be distinct from the common species of the northern hemisphere.
1 See pp. 250-1 for distinction between ' northern', ' intermediate', and ' southern ' regions.
26o DISCOVERY REPORTS
SCYLIORHINIDAE
Scyliorhinus {Halaelurus) bivius (Smith) was not obtained at the trawUng stations. Five specimens
were obtained with ' Other gear ' at three stations in the western channels.
SQUALIDAE
Squalus lebruni (Vaillant) is very close to the familiar S. acanthias of the northern hemisphere. From
the material at Norman's disposal up to 1937 he judged that 'there appears to be only one species of
Spotted Spiny Dogfish in the southern hemisphere' (Norman, 1937, pp. 9-10). Dogfish are rare on
the Patagonian grounds, in strong contrast to their abundance in European seas. This would be a
helpful feature if long-lining for any of the more useful species were found practicable in the future.
They could of course be utilized in the same way as the northern species, but our records suggest
that they are far too scarce ever to form an appreciable part of the catch of a commercial fishery.
S. lebruni was not taken in ' Other gear ' ; records of its occurrence in the trawl are :
WS90 I of 67 cm. length. WS791 B i of 66 cm., 1 150 g.
WSg4 3 of 61, 62 and 64 cm. WS797C i of 66 cm., 1300 g.
WS218 I of 64 cm. WS853 I of 65 cm., 1200 g.
These scanty records suggest that the species probably ranges over the whole of the shelf in summer,
and that it commonly attains a length of 2 ft. and a weight of some 2\ lb.
SQUATINIDAE
Squatina armata Philippi was not taken by the expedition, but Norman (1937, pp. lo-i i) provisionally
identified two Argentine specimens with this Chilean species, so that it may possibly occur within
the area of the trawling surveys.
TORPEDINIDAE
Discopyge tschudu Heckel. This species is known to range far north on both sides of South America,
but was only once taken by us :
WS776 103 m. 4 $? specimens
RAJIDAE
The general distribution and depth relations of the species of this family taken by us have been
summarized, after the notes on individual species, in the form of tables and a figure dealing with all
the Elasmobranchs taken in the trawl (Fig. 18, p. 276). Measurements of disk width and many weights
were recorded, but owing to the scarcity of most of the species they are not sufficiently numerous to
repay statistical treatment. The individual records therefore refer only to numbers of specimens
taken. It is noteworthy that disk-width measurements made by Gunther and his colleagues in the
field show discrepancy with some of Norman's (1937) records, owing no doubt to shrinkage of pre-
served material.
Rajaflavirostris Philippi. This species is morphologically very closely related to R. batis of British
seas and R. stabalifons of New England. It was widely distributed over the shelf in small numbers
rather more frequently in the northern region than elsewhere. In common with R. magellanica \t
favoured shallower waters than other members of the genus (Fig. 18), in spite of our difficulties in
sampling such a large area we can claim to have revisited most parts of the shelf at least three times
so that It is a remarkable fact that this skate was never found in the same place twice. Evidently it
must roam widely over the plain of the shelf (Fig. 8), and such behaviour is precisely the opposite of
that which Steven (1936) found in R. clavata. The young and adolescent stages of the thornback seem
nearly stationary, while even the migration of adults (Steven, 1932) is probably on a small scale
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
261
R. flavirostris was taken in the ' Trawl + accessory nets' at twenty-four stations as shown below.
None was captured with ' Other gear ' :
WS77
WSyg
WS9S
WS214
WS217
WS233
WS236
WS245
WS763
WS76S
WS789
WS790A
WS790B
WS791B
WS792B
WS793
WS796B
WS797C
WS810
WS815
WS816
WS817B
WS834
WS857
I
8
I
3
I
I
2
I
Of the twenty-eight individuals which were sexed, there were fifteen males and thirteen females;
sex ratio 53 "5% <S'S-
Fig. 8. Distribution of Raja flavirostris. First survey: triangles; second survey: squares; third survey:
circles; negative observations left blank.
Raja doello-juradoi Pozzi. Our captures of this species were most frequent in the southern region,
and it was there that the only rich haul (of fifty-two specimens) was obtained, but isolated captures
were recorded far to the northward all along the edge of the shelf. It is a deep-water species found
262
DISCOVERY REPORTS
almost exclusively at or over the shelf edge (Fig. 18). Fig. 9, which shows the spatial distribution,
also brings this point out very clearly. R. doello-juradoi is the most numerous of a small group of rays
which we found only in deep water, especially in the Falkland trough. The other three species were
all new to science. R. doello-juradoi was taken only in ' Trawl + accessory nets ', never in ' Other gear ' :
WS98
I
WS245
52
WS795
2
WS2I5
I
WS246
I
WS817B
I
WS2I8
4
WS783A
I
WS820
I
WS237
I
WS794
I
WS851
2
Of the sixty-eight specimens thirty-eight were males and thirty females; sex ratio 55-9% SS-
Fig. 9. Distribution of Raja doello-juradoi and the other deep-water rays. First survey: triangles; second survey: squares;
third survey: circles; negative observations left blank. Numbers refer only to R. doello-juradoi. g = R. griseocauda;
s = R. scaphiops ; a = R. albomaculata ; positive records only.
Raja macloviana Norman. The general spatial distribution of this species (Fig. 10) is very similar
to that of R. doello-juradoi, but a consideration of its distribution with depth shows that a higher
proportion of R. macloviana was taken just on the shelf, instead of just ' over the edge ' ; hence the
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 263
dumbbell-shaped distribution shown in Fig. 18. It is probable that this is caused by a migration to
deeper water with increasing size (and age), but our data are too few to demonstrate this clearly.
A more pronounced depth relation of the same type shown by R. brachyurops, a more common species
presently to be discussed, is almost certainly due to this cause, and when plotted in similar fashion
shows a more sharply angled polygon of the same basic type. R. macloviana was taken only in the
' Trawl + accessory nets', never in 'Other gear':
WS80
WS87
WS9S
WS109
WS217
WS218
WS225
WS236
3
4
4
3
WS24S
WS813
WS817A
WS817B
I
I
I
2
Of twenty-two specimens ten were males and twelve females; sex ratio 45-4% SS-
Fig. 10. Distribution of i?.;a«ac/o..a„«. Spring: diamonds; summer: circles; autunm:tria^^^^^^^^
squares; negative obser;ations left blank. Note. Symbols here refer to seasons, and not to surveys.
Raja magellanica Steindachner. All our captures of this species were made in ^hej-terrneto^^^^^
southern regions, and all but two of them definitely on the shelf. The spatial distribution is shown
264
DISCOVERY REPORTS
Fig. II, and the characteristically shallow depth distribution can be seen in Fig. 18. One specimen
was secured with the small beam trawl, all the others with ' Trawl + accessory nets'. It will be seen
that the distribution is almost co-extensive with that of another shallow-water species, Psammobatis
scobina, but Raja magellanica was found slightly farther south at some points, and tends to occur
O
o C3)
GO°
D
,□
2*40 D CO
'Vl3
A3 D
^h (^
°*^XD?.^^a^^°
%
P'AO A
Sb
"^v- ^<^i
D
Fig. II. Distribution oi Raja magellanica. First survey: triangles; second survey: squares; third
survey: circles; negative observations left blank.
With greater relative frequency in the southern region. Thus of the totals of the two species, less than
30% of the Psammobatis were from the southern region, but 48% of the Raja magellanica were taken
there. Although the observed sex ratio of this species was practically 'normal', there was a hint of
segregation mto a unisexual shoal at one station. This is a marked feature in some of the more abun-
dant elasmobranchs of British seas and may then lead to a most anomalous apparent sex ratio
especially in commercial landings, which may be further distorted by greater value and ease of capture
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 265
of the (usually) larger females, and by close packing of the schools of the latter, especially when
gravid (Ford, 1921, pp. 483-5 i Hickling, 1930a, pp. 537-8 ; Steven, 1933) :
2
WS72 I WS96 2 WS810
WS77 2 WS108 3 WS811II I
WS78 I WS223
WS90 5 WS245
WS91 I WS246
WS92 2 WS787
WS94 5 WS797C
WS95 13 WS802A
WS834 4
WS837 I
WS862 I
WS86i(BTS) I
Of fifty-three specimens, twenty-six were males and twenty-seven females; sex ratio 49-1% 33-
Raja multispinis Norman. The holotype of this new species was taken in the trawl at St. WS851 m
the southern region, depth 221-197 m. < u j '
Raja scaphiops Norman. Two specimens were obtained in the north, in deep water over the edge
in winter. Five specimens were subsequently secured from four southern stations, one on the edge
and three in deep water (Figs. 9, 18). This species was taken only by the trawl : WS218 (2), WS250 (i),
WS8i8A(i),WS8i9B{2\WS824{i). . ,, • u
Raja albomaadata Norman. Seven specimens of this new species were trawled at six southern
stations Two of these were on the edge of the shelf, the others all in deep water. None was taken
in 'Other gear'. Only the positive records are indicated in Fig. 9, along with the other rare deep-
water species. The negative records for the more common R. doello-juradoi, shown in the same figure,
can of course be taken as negatives for the others provided that they are sufficiently remote from the
appropriate positive symbols. It has not been easy to achieve this where more than one species
occurred at one of several closely spaced stations. Specimens of R. albomaculata were obtained at the
following stations :
WS-'d^ I WS824 I WS868 I
WS817B I WS839 2 WS875 I
Raja brachyurops Fowler. This was the commonest ray of the trawling surveys It occurred with
moderate frequency in all three regions; most frequently, in relation to the total of hauls m which it
might have been taken, in the intermediate region. Most of the richer hauls were in the southern
region, and the total of numbers taken in each region showed marked and progressive diminution
towards the north (Fig. 12, Tables 7 and 8). j . 1 .^^ ^Uct
When the data relating to the depth distribution of this species are treated as a whole, we see that
it had a wide range extending into deep water over the edge of the shel ; but the ^^P^^f^^^^^
polygon shows two maxima, one over the edge and the other in much shallower water on the shelf
(Fig 18). The distribution charts (Fig. 12) show that this may be due to seasonal moveoient on to the
hJf in summer and down into deep water m winter. Such a movement -"l^^^e ^logons to t^^^^
known in some European species, e.g. R. cla.ata, R. radiata and R. >/W« which show ^ le in
the aduh stages) a movement into the North Sea in summer, but are found chiefly - deeper wat r to
the north and west at other seasons (Meek, 1916, pp. 41-3)- More recently Steven (1932, P- ^o) ha
shown that R. clavata appears to hatch out in shallow water, that the dispersal of young and adolescent
fishi greater depths is'Lw (Steven, 1936), but that in the English Channel adults are rarely found
inshore except in spring just prior to the deposition of eggs. , , .
If an an logous movement takes place in R. brackyurop. we should expect our deep hauls to contain
a hig'r proportion of large ind.viduals than the shallow hauls, taking the year as a whole, though the
re^d might be masked by a few good catches of adults m shallow water durmg then m.grat.on.
266
DISCOVERY REPORTS
Disk-width frequencies of all records from the two depth categories, shown as percentages in Fig. 13,
clearly support the possibility of such a migration : individuals of the greater widths were more frequent
in the deep hauls and the smaller individuals in the shallower hauls. About the main mode, however,
the deep hauls show considerably greater frequencies, so that the difference is less well defined than
could be wished, while the adverse factor of some migrating adults being taken on the shelf is
unavoidable by this method of array.
Fig. 12. Distribution of Raja brachyurops. Spring: diamonds; summer: circles; autumn: triangles- winter- squares-
negative observations left blank. Note. Symbols here refer to seasons, and not fo s^™" ^
Fig. 13 gives just a hint of possible submodes at roughly 4-5 cm. intervals that may indicate year
classes. Irial freehand curves drawn from grouped data by Buchanan Wollaston's (1929) method
show some support for this view, which seems reasonable from the annual increment of 6 cm.
established by Steven (1936, p. 614) for R. clavata, which is a larger species. Our data are notnumerous
enough for us to regard the 4-5 cm. increment for R. brachyurops as more than a possibility.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES ^67
, ;^ r»„r nrpfl there were only 14^ measured individuals (of
"" Iirer r sul The w d di persion. leading ,o large values for c. renders .he difference between
g,ve abetter resull he WP ^^^ ,^^ ^^^^^ility of migrat.on was estab, shed,
means of some 2 5 ^■''"If "" J ^*^„,di ,„ reason could be undertaken without fear of waste
^ff:": rr: sTetttmejrLSlish b^eyond al, reasonable doubt the fact of migratron on to
the shelf in summer and over the edge into deep water m wmter.
II I >
A. HAULS OF LESS THAN ZOOM MEAN DEPTH
B. HAULS OF MORE THAN 200 M. MEAN DEPTH
DISC WIDTHS IN CMS
Fie X. Size-frequency distnbution (percentage at each en. of disk-width) of Raja brachyurops
Fig. H shows that whether we consider total abundance of ^^^^^^ ^^^"t.^
depth categories, or frequency of occurrence, the l^'^^^^:^ i:i'J':i''^,^:%,,,.^ appears to
more abundant on the shelf in summer and -'^T'^-T'^^nXt^i conformity with our thesis ;
268
DISCOVERY REPORTS
80-
>
liJ 60
<
t; 40
20
Fig. 14. A.
(<2O0
Seasonal variation in total abundance of Raja brachyiirops 'over the edge' (> 200 m ^
m.) white; also seasonal variation in its frequency of occurrence B on the shelf, and
black, and on the shelf
C 'over the edge'.
Season
Spring
Summer
Autumn
Winter
Spring
Summer
Autumn
Winter
Table 5. Full data relating to Raja brachyurops /)/o«g^ in Fig. 14
On the shelf, depths less than 200 m.
No. of
individuals
6
62
74
17
% of total
caught in
each season
60
95
97
14
Over the edge, depths more than 200 m.
No. of
individuals
No. of
hauls
14
65
42
17
No.
positive
3
17
18
3
positive
21-4
26-2
42-9
17-6
4
3
2
106
% of total
caught in
each season
40
5
3
86
No. of
hauls
2
10
8
13
No.
positive
positive
SO
10
12-5
38-5
Separate depth-frequency polygons of the type used for summarizing the main features of depth
dj nbutton of all elasmobranchs may also be plotted for R. brachyr.ops at each season, as m Fig '5
This presents the evidence of seasonal migration even more clearly ^'
Very good evidence that the young of R. brachyurops are hatched on the shelf can be obtained bv
considering our records of captures of minute post-embryos and yearlings less than zo cm wide'
The r seasonal distribution also is so circumscribed as to establish the approximate hatching peT od
and they are accordingly tabulated in full (Table 6) naicning period,
3oMaThaLTMlv'lltt' 't '"" V"' T'"''''^'' ^^^ ^"^^" ^^^^^^^^ -- -"g^^ between
xcfp ons I'o cm Id f T'n . ' '""'"' observations were made at other seasons. Of the
exceptions, the 9 cm. individual in October was presumably a late-hatched survivor of the previous
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
269
in
UJ
an
U)
Q-
Fig :5. Seasonal variation .n depth distribution of Raja Irachyurops. The widths oj the Polygons a^e P-portional to the
^ ^ percentage depth-frequency distribution when the observations are graphed m 50 m. classes.
Table 6. Captures of Raja brachyurops less than 10 cm. wide
Station
WS87
WS90
WS94
WS9S
WS109
WS76S
WS797
48
51
51
WS867
Date
3. iv. 27
7. iv. 27
16. iv. 27
17. iv. 27
26. iv. 27
17. X. 31
20. xii. 31
3. v. 26
4. V. 26
4. V. 26
30. iii. 32
Numbers, size
I of 4 cm.
I of 5 cm.
I of <S cm.
I of < 5 cm.
3 of <5 cm.
I of 9 cm.
1 of 5 cm.
2 of 7-5 cm.
I of 7 cm.
9 of 4(5)-9 cm.
I of 5 cm.
Depth, gear
mm.
81 m.
118 m.
108 m.
145 m.
115 m.
no m.J
no m., OTL
115 m., DLH
no m., OTL
148 m., BTS
year's brood, leaving a single anomalous observation in December. It seems atao« "^"'^ *^' *^
Iggs are deposited on the sbelf in summer, and hatch m ^^^-^-^^X^^JuLtZ,.
mean depth of these records .s i ,6 m.. w.h a= 15 f/'^ J"™X^„,, ^ J„een these means is
which R. brachyurops was taken IS 194 <"■ with (j=78-2598- 'he d.tterence
78 m., and ,,.j^;^..5.,644.-
* A^i = 274-
270 DISCOVERY REPORTS
Hence i/oi = 13-5, and the difference, which indicates that the smallest individuals were found almost
entirely near the upper Hmits of the depth range of the species, is very strongly significant. From this
it appears that the summer migration of R. brachyurops on to the shelf is a breeding migration,
particularly when we remember that nearly all the very young specimens have been secured in the
autumn, when the movement towards shoal water of the species as a whole appears to be at its height.
The sex ratio of R. brachyurops seemed to be remarkably constant whatever grouping of the data
was adopted. For all specimens whose sex was recorded it worked out at 477% males. There are
indications that shoaling is more marked towards the end of the year, but large schools segregated
according to sex, such as are known to occur in some European species, were not encountered:
WS73
5
WS225
4
WS792A
8
WS77
I
WS234
8
WS793
20
WS78
2
WS236
6
WS797B
I
WS79
32
WS23g
5
WS797C
4
WS80
3
WS245
78
WS800B
6
WS87
2
WS250
7
WS801
I
WSgo
5
WS765
3
WS806
2
WS92
2
WS772
4
WS811II
3
WS94
2
WS774
I
WS813
I
WS95
I
WS776
2
WS815
3
WS97
I
WS782A
I
WS816
I
WS98
6
WS783A
5
WS817B
3
WS108
I
WS783B
I
WS864
I
WSiog
6
WS784
2
WS866
I
WS214
2
WS785B
2
WS868
I
WS217
8
WS791B
I
WS874
2
WS218
7
' Other gear ' :
48 2 (in
OTL)
51
12
WS867
I
51 2 (in
DLH)
WS865
2
Raja griseocauda Norman. Eight specimens of this new species were trawled : two at a deep northern
station, two at a deep intermediate station, and one at each of four southern stations, three of which
were in deep water and one on the shelf edge. Its distribution is much like that of R. scaphiops. It
seems to be one of a group of uncommon deep-water rays : R. doello-juradoi, R. scaphiops, R. griseocauda
and R. albomacidata, that occupy an ecological position in the Fatagonian fauna similar to that of
R. falsavela, R. hyperborea and R. oxyrhyticha to the north-west of the British Isles (Meek, 1916).
The scanty distributional data for R. griseocauda are shown in Fig. 9, and the depth records are sum-
marized along with those for the other species. None was taken in ' Other gear ' :
WS218
WS236
IVS245
WS250
WS817B I
WS824 I
Psammohatis extenta (Carman). A single specimen trawled at St. WS788B in the northern region
is assigned by Norman (1937, pp. 28-9) to this species. Previous records were mostly from the
Brazilian coast south of Rio de Janeiro, suggesting that the area we surveyed is south of the normal
range of the species.
Psammobatis scobina (Philippi). Except Raja brachyurops, this was the most abundant elasmo-
branch we found during the trawling surveys. It was most definitely a species of the shelf, and was
found in fair numbers closer inshore than any of the other shallow-water rays. (The slightly shallower
'effective mean depth' of R. magellanica is not significant, and doubtless is due merely to insufficient
sampling of that less common species.) Psammobatis scobina was rare in the northern region and
relatively most abundant in the intermediate region, especially close inshore off Puerto Deseado and
Puerto san Julian. It was fairly plentiful in the southern region also, but was only twice recorded
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 271
south Of the entrance to Magellan Straits, well offshore near the shelf edge. None was found to the
south of the Falkland Islands. Fig. 16 shows the regional distribution, while the depth data are sum-
marized along with those of all the elasmobranchs in Fig. 18 and Table 9.
.... . A, 0/ „ai.._is abnormal. It seems to be due to a tendency
The sex ratio observed m th,s spec.es-67 % ""'"^^ '^ ."„„„„ of the sesregated males than
towards unisexual shoahng. and mere *-« ' ^kT w''„™fitTs rfem* sho^^^^^^^ -= --"y
females. Among European rays in wh.ch th.s hab.t .s ^^°^"^l'™Zr^tciy rieh hauls of this
272
DISCOVERY REPORTS
from ' Other gear ', and two others at which there was some doubt as to the depth logged, have been
excluded from the former:
WS73
2: I (J, 1$
WS243
2??
I^5,Sij
2 6i
WS77
3: !<?, 2??
WS765
iS
JF55i^
3: !<?, 2??
WS79
6:3c?<?, 3??
WS775
Ic?
Pr5&5
5:2<?rf^3?$
WS80
4: I cJ, 3 ??
WS782A
2??
W^5,Si6
IS
WSgi
I?
WS786
I ?
WS817A
4??
WS92
2??
WS787
2??
WS817B
2??
WS94
5-^<S<S, I ?
WS792A
I(J
WS855
icJ
WS95
5, all 3<S
WS796B
Ic?
WS857
I(?
WS96
7:6<Jc?, I?
WS797B
4??
WS862
!<?
WS108
4: 3 <?<?, I ?
WS797C
37. all
00
WS864
I (?
WS109
I?
WS806
i(J
WS866
2: I 0, I ?
WS222
2<S<S
WS808
Ic?
WS868
2<SS
WS223
I 0
WS809A
2: I 0*,
I?
WS874
icJ
WS225
Ic?
WS809B
I ?
WS856
I (J (in BTS)
WS22g
i<S
WS810
20: 10
cJcJ, 10 ??
WS861
I 3 (in BTS)
WS239
3 •■ 2 ,36, I ?
WS811II
2: I (?,
I ?
WS865
I 3 (in BTS)
Psammobatis microps (Gunther) was not taken by the expedition's ships, but it is known from the
mouth of the Plate and may occasionally range as far south as the northern part of the area we sur-
veyed.
CHIMAERIDAE
Callorhynclms callorhynchiis (Linnaeus). Our records of this species show that it was very definitely
confined to the warmer inshore waters in our area. Most of our specimens were taken in the northern
region, in spring. Single records in the intermediate region (autumn) and in the southern region
(late summer, coincident with maximum temperatures for the year) suggest the possibility of a
N-» S^ N migration along the coast. The distribution is shown in Fig. 17, and the depth relations
in Fig. 18.. Callorhynchiis occurred in the trawl only, none being taken in 'Other gear'.
Norman (1937, pp. 35-6) has shown that Carman's (1904, 191 1) distinctions between two 'species'
of Callorhynchiis on either side of South America cannot be maintained. C. srnythii Bennett, 1839
should be regarded as a synonym of C. callorhynchus (Linnaeus), 1758. Norman further stated that
' It is probable that the examination of an adequate series of specimens would show that the nominal
species capensis, from South Africa and milu, from Australia, Tasmania and New Zealand, are nothing
more than varieties of C. callorhynchus'. Careful study of Carman's keys inclines one to agree. All the
external characters, such as the relative extent of various fins and so forth, given as diagnostic by
Carman, show complete overlapping. Most of them were exhibited within the limits of our own
series of C. callorhynchus. They would be better described as agnostic characters. One is left with the
modifications of the palatine teeth as the only character on which a distinction can be based, and
Carman himself (191 1, p. 97) remarks on the danger of such a practice on account of the extent to
which the teeth may change with age and use.
Barnard (1925, p. 96) maintained that Carman's recognition of C. capensis Dumeril as distinct from
C. callorhynchus was correct 'on the form of the dental plates'. He examined many specimens, but
does not say how many or of what size, though from two maximal measurements given it is certain
that he saw at least two large specimens.
Is the form of the tritors of the palatine lamellae a good specific character? They are essentially
the same m the young of all these 'species', and no author has yet defined the stage at which the
differentiation becomes apparent. Moreover, the fossil C. hectori Newton, 1876 is said by Carman to
agree with C. capensis in respect of this character. This fossil was found in M^hat are probably cretaceous
rocks m New Zealand. One can recall teleosts in which teeth regarded (by some) as diagnostic in the
adults are absent m the young and 'occasionally absent' in large specimens.
273
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
We may say of the species of Callorhynchus that smythii Bennett, 1839 is a synonym of callorhynchus
(L^naTuJ) 17 8, whik mUii Bory, 18.3 may proye to be another synonym. C capens. Dumenl,
mlZX^^^^^^ but the whole problem appears insoluble by purely verbal descriptive methods,
and must wait upon the collection of long series of adequate biometnc data.
I. „„, be seen from .he d«a recorded below ,ba. we were not ''^^ ^;^ZZ<^^X^^^^^
seeured, so that our records do not tally exartly w„h those B-- ^-V ^"^J^^'fJ;^ S, Vs*,?-
ntistake has occurred over the spechnen recorded ^Y Norman as a fema of 80 cm ^ ^^^^^^ ^^
The depth given by Norman is wrong for that ^'^-"^ ">;"=""=: J™ .probably the specimen
274 DISCOVERY REPORTS
of C. smythii and C. callorhynchtis . The particulars for the specimens not seen by Norman are given
separately :
WS96
WS762B
WS763
WS788
16
2
WS847B
WS853
Specimens not seen by Norman : WS762B i male of 48 cm. ; WS763 unsexed specimens of 25, 30,
31 and 33 cm. and one unmeasured; WS788 unsexed specimens of 32 and 40 cm.; WS84JB one
female of 90+ cm. (tail damaged), weight 4750 g.
SUMMARY OF OBSERVATIONS ON ELASMOBRANCHII
The main features of the distributional trends of the various species- of elasmobranchs found on the
shelf are summarized in Tables 7 and 8. These show frequency of occurrence relative to the total
number of hauls made in each region, and abundance of individuals in each region relative to the
total for each species for all regions. With these and the data given in the Appendix any desired com-
putation as to fish per hour's trawling, or per hour's positive hauls, could also be made ; but as the
group is not sufficiently abundant to encourage commercial exploitation, I have not done so here.
Table 7. Distribution of Elasmobranchii: occurrence in total of roughly comparable
hauls in each region
Species
Northern region
Intermediate region
Southern region
No.
% of hauls +
No.
% of hauls +
No.
% of hauls +
Squalus lebruni
Raja flavirostris
R. doello-juradoi
R. macloviana
R. magellanica
R. scaphiops
R. albotnaculata
R. brachyurops
R. griseocauda
Psammobatis scobina
Callorhynchus callorhynchus
3
8
2
I
I
I
0
6
I
4
4
ii-i
29-63
7-41
370
370
37°
0
22-22
370
14-81
14-81
I
9
4
3
7
0
0
17
I
17
I
1-89
16-98
7-55
5-66
13-21
0
0
32-08
1-89
32-08
1-89
2
7
6
8
13
4
6
26
4
24
I
2-o8
7-29
6-25
8-33
13-54
4-17
6-25
27-08
4-17
25-00
1-04
Number of roughly com-
parable hauls
27
53
96
Table 8. Distribution of Elasmobranchii: regional abundance of individuals
of the trawled species
Species
Squalus lebruni
Raja flavirostris
R. doello-juradoi
R. macloviana
R. magellanica
R. scaphiops
R. alburnaculata
R. brachyurops
R. griseocauda
Psammobatis scobina
Callorhynchus callorhynchus
Total Elasmobranchii
Total no. of
individuals
39
68
23
54
7
7
292*
8
iS7t
28
697II
Northern region
No.
3
10
5
4
I
2
o
43
2
5
26
77f
/o
37-5
25-6
7-4'
17-4
1-9
28-6
o
14-7
25-0
3-2
92-8
ii-o
Intermediate region
No.
I
18
5
7
27
o
o
53
2
93 1
/o
211'
* Includes eighteen southern specimens taken in
X Includes two taken in 'Other gear'.
II Includes six rare specimens not tabulated.
** Includes four Discopyge tschudii not tabulated.
12-5
46-2
7-4
30-4
50-0
o
o
i8-2
25-0
59-2
3-6
Southern region
No.
30-3
Other
4
II
58
12
26
5
7
196*
4
59§
/o
409tt
50-0
28-2
85-2
52-2
48-1
71-4
1 00-0
67-1
50-0
37-6
3-6
587
gear
f Includes three taken in 'Other gear'.
§ Includes one taken in 'Other gear',
f Includes one Psatnmobatis extenta not tabulated.
ft Includes one Raja multispinis not tabulated.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 275
Fie i8 to which frequent reference has been made in the notes on individual species, summarizes
the main 'features of their depth distribution. The statistical significance of the diiferences between
the effective mean depths observed for the several species, which are plotted m the figure, are given
'"^ iT tm be seen that the species may be grouped into three classes according to their depth relations,
'^l' Shallow, species of the shelf: Squalus lebruni, Raja flavirostris, R. magellanica, Psammobatis
'''".^t^.: Raja maclo.i^a and R. brachyurops. The peculiar depth distribution of
the last named has been shown to be due to a seasonal migration over the shelf edge. The less common
R. macloviana shows a general depth relation of the same type (an hour-glass or dumbbell-shaped
polygon), and there are other grounds for supposing that a similar migration in tha species is po sible
m. Deep-water species found almost exclusively over the shelf edge: R. doello-juradot, with the
rarer forms R. scaphiops, R. alhomaculata and R. griseocauda.
CLUPEIDAE
Clupea fueoensis (Jenyns). This is one of the most numerous and important species of our area, but
^C^l main y pela ic habits and small (' Pochard ') size, it could not be adequately sampled wi h
ol ge For this reason I give only the list of stations where it was captured. The numbers mean
3i tie. The Falkland herring is a most important forage fish for l-g-/P--;. ^^f^^^^.^tr'
anZi also extensively eaten by birds and seals. It is possible that it could be utilized directly for
human ood-it is most excellent eating, as the Falkland islanders well know-but prospects of
regTr a^- m numbers sufficient to support, say, a small canning factory are not good. Any
fofm of gm net would be continually threatened by seals and penguins anywhere wi hin reasonable
SllLce of the Falkland Islands, while the herring trawl would almost certainly fail on the rough
bottom Some specialized form of purse seining might possibly provide an answer.
Some other clupeoids recorded from the Patagonian region may be mentioned here, though we d.d
not cTotu e any of them and they are probably not normal inhabitants of the area surveyed. The
; c es rr^ownTchU "; ' sardina' was Lnd by Norman (:937, P- 38) to be quite di^mct from the
vTa ly ealrn C.fuegensis. Norman has named it C. berrtincki, after Mr Cavendish Bentinck, who
sIntHm a tne series Lm Talcahuano harbour. From Col. Tenison's drawings it can at once be
Te^n Z I is much deeper in profile than the species so common on the continental shelf to the east.
It resembles typical northern hemisphere herrings much more closely.
Gunther has a note that Clupea sp. (? maderen.s) was on sale at the -/^^ ^ ^^^ t"; f Itfs
u f oAor^T^rnv This soecics he savs, s represented farther south by C. /z^c^ewm. n is
smaller than C. fuegensis, of ' sprat ' as compared with ' pilchard size. ^
We obtained specimens of C.friegensts at the following stations, mostly in the accessory
S wis WS^So A TS8%B WS^3S
^fi4 ^St Ws4b WSS^S S^f/,(N:ooB)
Also: Stanley Harbour, 2. .i. zj and Chartres River, West Falkland Islands, 13. iii- 33-
276
DISCOVERY REPORTS
fe 'So t^
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
277
S
•5^
s
a
c
o
<^
'S
E
.£
S 3
-4- c
O
+ 1 +-c«
Cfi +CAJ
+ Ci5 ^C« +Cfi +CJ5 ^.Cfi +'cfi
2 a
oi-
e«)'
0)
^ e
V
+ o
■^f. ^
03 .
O tjo
;«5
rr, 6J)
"^i . 2. • o
f^ 60 "^ o« -
"~>.oo
^ 7^ +z '^ 1
^ +
00 »)
1^
ro 6a fv o<|
7^5 7<^
A
CO
-a «J
^ e
(U ^4-H
C
a
<u
e
DS S
■o .
I ^
00 .Sp CO 5f ^^ o
1 00 +02 +2
CO
3^ <u SS 11 00
"I O "-I ^O 7 O I jO
t^ 5i 00 g
7^0
1 s«
(i;
oi,
1^
00 o
&0
'CO 7 CO
.!»«
^
1^ 1^
1^
•O 60 On bo
"i« 7^
I
00
+1
t^_M ON.SP
•z +
Cfi +Cfi
+ (/3 +C«
bb c» Ml ^.y>
?s
+ 1 1^
tQ.a^ t^-^ f^.^
1 Co I CO +cfi
1^ +z +1 '^ +z
I CO 7^
.'^
s
"fe
se
s
a
-S
■^^
a
«
3
■^
O^
CO
^
Qi Q>
278 DISCOVERY REPORTS
Cliipea arcuata Jenyns. Our gear was even less suitable for sampling this small coastal but mainly
pelagic species, which is very closely related to the European C. sprattus, and which is known to be
abundant at times within the area surveyed. In life the muscles show up yellow through the skin,
permitting rapid distinction from small fry of C. fuegensis of similar size. Doubtless a frequent food
of larger fishes, we only happened on it in numbers at St. WS8g, where thirty-nine were taken in the
N7-T attached to the back of the trawl. This station was worked in very shoal water close in to the
north-east coast of Tierra del Fuego. Quite possibly the normal habitat of the species is too coastal
for it to have been taken by the fine ' accessory nets ' on other occasions.
GALAXIIDAE
Galaxios attemiatiis (Jenyns). This fish was not trawled by us— it is improbable that they ever depart
far enough from the coast to be sampled by a trawl— but a Falkland specimen was readily obtained
for Norman's report of 1937 by Mr Bennett. The majority of the galaxiids are fresh-water species,
though an increasing number have been found in the sea. Most of them can probably be regarded as
anadromous. G. attenuatus is catadromous, descending to brackish water or to the sea itself to spawn.
It is possible that much of the spawning takes place in the lower reaches of estuaries rather than in
the sea itself (Phillips, 1924), but it is certain that the larvae must be widely distributed in the sea,
or the species could never have populated the smaller brooks in which it is found, for many of these
have no estuarine transition area at their mouths. Moreover, there is an interval between the time of
known spawnings and the records of upstream movements of larvae. A species of Galaxias unknown
m fresh waters, G. bollansi Hutton, was described in 1899. Very recently, Scott (1941) has shown
that apart from G. attenuatus (which he studied in very great detail in an earlier paper there quoted)
the type species of the genus, G. truttaceous Cuvier, is euryhaline, and that 'it is not improbably
facultatively catadromous, and that when not confined in land-locked waters it may retain the pre-
sumably primitive spawning habit of the family'. Since Scott had already shown (1938, pp. 125-6)
that young G. truttaceous may arrive with the upstream spring immigration of G. attenuatus, that
adults descend (1941, pp. 57 et seq.) to almost completely saline sea water, and that when ripe
they are found at or near the coast (p. 68), his conclusion seems most cautiously worded.
I cannot understand why the catadromous migration of G. attenuatus was ever doubted Beginning
with Hutton in 1872I the facts had been described by several observers in New Zealand, Tasmania
and Australia, who were familiar with the fish in fife. It is true that McCulloch (1915) had had to
improvise a very primitive experiment in the endeavour to prove his point, but there is nothing to
show that It was not effective ; and Meek's assertion (1916, p. 147) that the 3 J-4 cm. larvae are denatant
IS at variance with McCulloch's direct observations. It is, indeed, probable that the catadromous
migration of G. attenuatus was well known and understood by the Maori— who still eat them— before
white men ever came to New Zealand. Phillipps (1919, quoting Best, 1903) was doubtful of one tradi-
tional Maori account that described spawning at the mouths of rivers. He then thought that the
spawning was entirely marine, but within five years his own observations (Phillipps, 1924) had shown
that the Maori account was substantially correct. It may seem unnecessary to labour the point now
but G attenuatus is an important fish in New Zealand, where the ascending fry are captured and
canned as whitebait', at one time the only fishery product exported from the Dominion Moreover
visiting the Dommions twenty years afterwards in a research ship, one found that biologists 'from
home still tended to be weighed on their willingness to accept enlightenment on the subject
The general distribution of G. attenuatus is of particular interest; it is known from southern
Australia, Tasmania, New Zealand, both coasts of Patagonia and the Falkland Islands. The family
^ I have not seen this paper.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 279
Galaxiidae has a marked ' circumpolar sub -Antarctic ' distribution, but the other, mostly more truly
lacustrine, species have localized distributions, and only G. attenuatus, known to be catadromous,
extends across the great ocean barriers. r ^ .. , .«
According to PhiUipps and Hodgkinson (1922) the older lacustrme stages of G. attenuatus are
generally known as 'minnows' in New Zealand. The young fry eaten as whitebait are very good. All
the whitebait examined from Auckland market were found to belong to this species.^ They were on
sale in August, September and October.
In the Falkland Islands G. attenuatus is one of several fishes locally termed ' smelt '. They sometm.es
shoal in littoral waters, probably when spawning, or later when re-entering the rivers. Their excellence
as food is already appreciated. Adults are said to attain a length of 6 in. which is consistent with
Scott's voluminous biometric data from Tasmania. This is undoubtedly one of the most promising
species for small-scale local exploitation. Owing to the small size of the rivers m the Falkland Island
he scale of operations would have to be kept down, or rapid depletion would result. The stock cannot
be a large one for there is not enough fresh water to maintain large quantities of the maturing adults.
' 1 :r .^c./... (Jenyns), which occurs m the Falkland Islands and on both coasts of Patagonia,
was not obtained by us. It is known chiefly in streams and brooks, but also occurs in the sea (Norman,
",7 p 40 In the Falklands it is known as ' trout', but this local name is also given to AplochUon
II; 'it' noteworthy that the larger galaxiids of southern Australia, Tasmania and New Zealand
were al o called 'trout' by the settlers until the artificial introduction of true salmonids from he
were also calleat y observation in the application of the
ra^::t ' 'o3' nle^oTui:: rtTtely related fishes m some instances ; while m others, also dating
from the great period of human expansion, no discretion at all has been used.
APLOCHITONIDAE
AMuton ^ebra Jenyns. No specimens of this fish were obtained by the expedition, but Bennett
Aplocfuton .ebra jenyns p forwarded it to Norman when the latter s
found a specimen which had been '^^ ^^^ ^^/^^ ^9 ^^^^^.^^ maculatus,
report was being written, ^^^^^l^^ .^J^^^l ^^^^^ might indicate. Indeed.
but there is no ^^^^^ ^^l ^^^^^^^^ ,he F^land Islands, Norman (x937, P- ^37)
It was probably one of the first fishes ever o ^^^il Darwin brought
Jenyns in 1842.
SYNGNATHIDAE
Upto^,us min.:iUanu. (Eydoux a„a Gerva,s). Th,s pipefish -Z:^^;:':^l\tr t:^::::^Z
l^^^XXX^^^ 'ee„ Unown for J. cen.u^, .he • WUia™ Scoresby • also ob.a.ned
a specimen in a tow-net, at St. WSS93.
MACRURIDAE
"'" TT'ZZZ'C^Z^^^^^ in co„siae«b,e depths is possible. I. ™y
™:fd fl:: nonh, bu. *::: *e s.eep„ess of .he slope p.ecluaes aae,ua.e exp,o.a.,on a. su,.ab,e
. T™.,, wh„*.i, c„„ls, ,.gd, of Voc.i.o„ld.e wi,K G. .,-™-.. .nd »,n...™. O. ..-« (ScC, ,,36).
28o DISCOVERY REPORTS
depths. Out of ninety specimens, two only could have been taken in less than 150 fm., and none in
less than 272 m. The only large catch showed a high proportion of males, but at WS821A (ten
individuals) the entire catch consisted of females, which generally preponderated in the small catches:
2??
WS82O
I?
2$?
WS821A
10 ??
3??
WS839
69: 54cJiJ, IS??
2??
WS84O
i^
Coelorhxtichus fasciatiis (Giinther). All our specimens of this species were obtained in summer and
autumn (third survey) in deep water over the shelf edge, mainly in the Falkland trough. Out of 140
specimens captured one only could have been taken in less than 200 m. of water. A notable prepon-
derance of females was observed in three out of four catches of fourteen or more individuals. This
may be due to the higher escape ratio of the smaller males, but abnormal sex ratios were also observed
in Coryphaenoides, and it may be that there is a tendency towards sexual segregation among shoals of
fishes belonging to this family at certain seasons. It is possible that the geographical range of Coelo-
rhyfichus fasciatus extends considerably farther north. Owing to the steepness of the descent from the
shelf there our chances of trawling in suitable depths were extremely limited. It is noteworthy that
this smaller species, though found exclusively over the edge like members of the family everywhere,
favoured slightly shallower depths than did Coryphaenoides.
From Phillipps (1921) we learn that Coelorhynchiis australis, called 'javelin-fish' in New Zealand, is
occasionally taken by trawlers in Golden Bay, and highly esteemed as a food fish. It may therefore
be that the usual dumping of macrurids, as ' rattails ' among the rubbish of trawl catches, as is general
off the British Isles and off" South Africa, is a needless waste. If Coryphaenoides is similarly edible it
would be the more valuable of the two Patagonian species on account of its larger size — up to 87 cm.
as against 38 cm. in our catches:
WS817A I
WS819A
3 (2 3o)
WS829
7 (all 66)
IVS817B 57 (4 o^.^)
WS819B
5 (4 o'^)
WS840
I
IVS818A 15(20^0^)
WS820
14(11 'SS)
WS870
8
IVS818B 8 (6 ,^0")
WS821 A
H (I S)
WS875
7
MERLUCCIIDAE
Merluccius hubbsi Marini
INTRODUCTION: ECONOMIC IMPORTANCE OF ALLIED SPECIES
Merluccius hubbsi Marini, the Patagonian hake, is the most important and one of the most numerous
fishes of the region. In our trawling catches it ranked third in numbers to Notothenia ramsayi and
Macruronus magellaniciis, but the weight of Merhicciiis captured exceeded that of the other two species
combined, and formed 47-3 % of the weight of all the fishes caught at eighty-three stations worked
during the third survey for which weight data are available.
All the true hakes are very closely related ; there is, indeed, still room for doubt as to whether some
are specifically distinct. Their distribution raises many problems of great general biological interest
which I hope some day to discuss at length elsewhere. For our present purpose it is sufficient to note
that the Patagonian species is known to range from the southern coast of Brazil (probably from about
the point where the Brazil current begins to swing off'shore) to the neighbourhood of the eastern
entrance to Magellan Straits. From Norman's (1937, p. 46) diagnoses there appears to be no doubt
that it is specifically distinct from M. gayi (Guichenot), the common species of the west coast of South
America, with which it was for long confused. Indeed, M. hubbsi in its general bodily proportions
resembles the silver hake or 'whiting' of New England and the north-west Atlantic, M. bilinearis
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 281
(Mitchell), more closely than it does any other member of the genus; but here again Norman's scale
counts point to the specific distmction being justified. (Compare Col. Venison s figures of M^^a>.
and M hMsi (Norman, 1937, P- 46), with the figures of M. productus and M. bthneam by H. L.
Todd in Goode's Atlas, 1884, pi. 65.) ,,,.,. • r 1 ^u
All the true hakes are edible : M. merluccius of Europe and North Africa, M. bibneans of the north-
west Atlantic and M. capensis (Castelnau) of South Africa are already the staple of important fisheries^
M productus (Ayres) of the north-east Pacific has been little utilized, and the same may be said of
M australis (Hutton) of New Zealand; but so long ago as 1907 the British Columbia Fisheries Com-
mission reported that M. productus was not inferior to the Atlantic species, and it would seem that
Therneglecr^ due merely to the plenitude of better food fishes in these favoured regions. Mgayr
of Chile is captured and eaten locally, but its exploitation by modern large-scale trawling methods
is rendered impossible by the absence of any continental shelf on the west coast of South America
FinaUy the Pa^agoman species M. hubbsr is captured by the small trawling mdustry operating from
the mouth of the River Plate. This originated with Don Pedro Galceran in Montevideo, but the
U urayan enterprise failed and the small-scale industry was then carried out from Buenos A. re
(DevTncenzi 1926). In 1932, when Gunther visited the fish market at Buenos Aires, he found tha
'^eriuv^'ere selling at o 50 pesos per kg. The Buenos Aires trawlers are not known to have operated
Zt^Z lile so far afi'eld as th'e area we surveyed. Up to the time of our last survey i^,^^)r. was said
that thev rarely proceeded out of sight of the land at the mouth of the Plate.
In orL to IsLs the potential value of Patagonian hake we may briefly consider the history of th
expbitatn of the three'species of Merlucaus that already provide the raw material for considerable
The European hake, M. nrerlucaus (Linnaeus), ranges from the Norwegian Rinne southwards along
the edge ofThe continental shelf as far as Dakar on the Mauntanian coast, and perhaps even farthe
south I local race has been reported off Cape Verde (Belloc, 1937)- Belloc's observations on vertebral
rlrt !how that, while the previously known stocks of European hal^e/-^"-^ --
m the west of Ireland, had 'vertebral numbers' regularly increasing from 50-48 (±029) o 51 ^5
lo6rthe number or the large sample he obtained off Cape Verde was 54-09 (±o-3) Ooc cit.,
fi^^TWsT of exceptional interest because increased numbers towards the northern end of the
fig. 3)- ^his IS ot exceptioi .0 . ^;j, .o^^a and herring (numerous workers, quoted by
1 r.;^c Tf Belloc is right n maintaining that the Cape Verde haKe are uui d , , • •
'^^ spe -t-ana L . e.pHa.c *a. a,, ^ ^^^^^^^^Z:;^!::::^
-"^ '^^::X::t:z:^^t!:^"^^-^ a.f ... a„a *.
' T ' L r o b Mlowed by the other stoeks of European hake for which data are ava.lable.
r rrrknutth detailed observations on vertebral numbers of Merluum wuh the pre-
LZs aXfinle^'that Ford (,,38) has shown to be desirable, cou.d not fa„ .0 be of exceptronal
'"xhThake is contnron in the Mediterranean especUl, on t^.enorth«n shore whe.^^^^
. nrentioned ,n the literature' from the t.nre »' Ar.s.o.^ Cou h .86 . v ^ ■ ■ • P^ J ;__^ l^ ^^^ ^^^
rpo;tL':tcid;:—i^^^
282 DISCOVERY REPORTS
principal fish captured at Fiume in 1879-80, ibid. p. 167), also mentions that the superiority of hooked
over net-caught hake was so marked that they could command an appreciably higher price.
In the north records of M. merliicciiis from Iceland have been confirmed, but some from south-
west Greenland are more doubtful. Even at Iceland such stragglers are rare and represent the extreme
range of the species. Hake are not common in the North Sea, though there is a regular small-scale
immigration into the north-eastern portion of it. In fact the fish, which lives for a good part of the
year over deep water oflF the edge of the continental shelf, is essentially an inhabitant of warm tem-
perate waters from the west of Scotland southwards. Its southern limits in all probability are normally
defined by the limits of influence of the Canary current, which is cold relative to the tropical surface
waters south and west of it, and relatively rich in nutrient salts and all the larger forms of life (cf.
Hentschel, 1936, p. 243 and Beilage ix).
M. merlucciiis, the merluce of heraldry, has been an important constituent of the fish food of the
western European nations throughout historic times. In Britain it has been the subject of various
commercial treaties from the time of King John to Queen Mary. Much was eaten during Lent.
Latterly it fell into disfavour, partly perhaps because the disestablishment of the Church Ted to a less
rigorous insistence upon traditional Lenten fare, but chiefly because improved boats and gear enabled
fishermen to catch far greater quantities of the choicer fish than before. In France, Spain and Portugal
It must always have been relatively important, owing to the lack of the colder-water gadoids near at
hand; but as recently as the second half of the last century we find leading ichthyologists in this
country dismissing it as a poor, coarse fish, of inferior table qualities. It is said, indeed, that the
German and Dutch names for hake, Stockfische or Stokvische (there are others), derive from the
habit of lettmg smack's boys keep them for ' stocker ', which makes it certain that they were practically
unsaleable.
The rapid development of otter trawling, especially the introduction of steam trawling, at the turn
of the century saw a rapid decline in the proportion of prime fish landed (though of course the actual
quantity was at first increased), and the development of fried-fish shops greatly stimulated hake
trawling. Indeed, Hickling (1935a, pp. 70-1) has been able to show that there was serious depletion
of the stock through overfishing before the war of 1914-18. During that war all fish stocks recovered
to some extent, and with improved boats and gear (the V.D. trawl) and the development of deep-sea
traw mg on distant grounds throughout the whole latitudinal range of the species, British hake
trawling showed increasingly heavy catches until well into the 1920's. But the catch was only main-
tained by increased fishing efl^ort, as Hickling (1935., pp. 74-5) has so clearly demonstrated. Since
1925 catches declined and depletion continued until 1939.
The importance of hake to the modern British trawling industry is very great. It is the staple
catch of our great west coast trawling ports, Fleetwood, Milford, Cardifl^ and Swansea. Even in 19^0
when decline of the stock was beginning to show itself, it ranked third of all our trawled fish whether
reckoned by quantity or by value. Only cod and haddock were more important. During 1920-5 the
British landings averaged 38,500 tons, and some of the fleet were working as far south as North Africa
sZ^r' ^ V r' r'""^" P'"''^ '^' ''''''''^^ °f fi^h^^g by French and (latterly) modern
Spanish vessels, in the southern part of the range of the species, also greatly increased. Though the
c^;ra"fi::::::rtrpeX^^^^^
rp, ., ,,,<,. y : 920 33 was ^1,532,000 per annum (Hickhng, 193 5 «, p. 74, fig 7)
to No^th cllt T 7' ' t ""T ^r^^^'^'^' ""■ ''''''^'' ("^^^^^''^)' -"g- f-- Newfoundland
to No th Ca olina. To the south it has been recorded (from deep water only) as far as the Bahamas
and Florida (Longley and Hildebrand, 1941, p. 38). Panamas
The history of the commercial exploitation of M. bilinearis aflfords a close parallel to that of the
■ DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 283
European hake, but with a characteristic acceleration due to the American fishermen having a totally
undepleted stock to deal with when modern methods of capture and treatment of the catch were
introduced. In dealing with this species it is important to realize that much of the earlier American
and Canadian literature is vitiated by confusion between it and several species of Urophysis (Gadidae
which are called ' hake ' with or without distinguishing prefixes along those coasts. The once important
bv-product trade in ' hake-sounds ' for isinglass was based on these, and not on Merluccius.
In the early days silver hake were regarded as rubbish by the New England fishermen, and also
as a great nuisance when large numbers were caught in mackerel nets (Goode et ai, 1884 pp. 240-3) and
had to be discarded. Their inshore migration in summer appears to be even more marked than that ot the
European species, and they frequently become stranded in pursuit of prey. At times they were used for
manure (Bigelow and Welsh, ig-S, PP- 386-96), and as recently as 1895 only 37,000 lb. were marketed
from Massachusetts and Maine. By 1919 more than 14 million lb. were sold, and even a that da e
most were still caught in traps and weirs, the price being too low for the offshore fishermen to bring m
those they captured. Since then the demand has increased enormously with the development of large
cold stores, especially in the mid-western states. From the Statistical Digest No. 4 of the U.S. Bureau
of Fisheries (i94o)-a mine of fascinating information-it can be seen that nearly 50 million lb. of
M.^SLn. were marketed m that year, more than 80O/. of the total coming from the New England
States. Nearly half this catch was frozen (forming 11% of the total frozen fishery produce of the
country). It has, indeed, become the most important single species of the frozen fish trade. This rapid
develo ment ha; synchronized with a big change in fishing methods, more being ^^^^^^^ZZ^r
than in pound-nets now, except in New Jersey and Rhode Island. One wonders whether the over
fishing problem may not soon become evident with the New England whiting ;
M capensis (Castelnau), which is known in South Africa as stockfish or stokvisch has been the
stapk of the modern trawling industry developed in that country. In the absence of adequate data
^ets stm some room for doubt as to whether the distinction between this and the Eu-p^anha -s
sufficient for M. capensis to be regarded as a separate species (Barnard, 1925, P- 3^0). The bouth
Afric n form ranges from Angola to Natal, in deeper waters towards the equatona limits of its range
on dthe side of L continent. It may be noted that this seems to be a general rule for all species o
rltU and for very many other fish besides. There are no statistics ^J^^^/ ^f ^^^ ^^^^
l^nrlinps to enable one to trace the growth of the modern fishery at the Cape, but the figures tor
Tat; uH^he^ by von Bonde (^934) leave one in no doubt as to the i-P^^-;^^^^^^^^^^^^^^
5n that period the catch was more than one-third of the total weight of fish landed: some 7850 tons
nf ctnrkfish worth ovcr / 1 ^0,000 per annum. r ^^ 7
Sorrjes relating .0 the siz^ and value of the fisheries for these three speces of S^^-^"-"-"
sumrrized n Table ,0 below. They reveal truly astonishing difTerences tn the value of the fi h ,n
hecTu 1 concerned. The British catch used to be nearly double that 0^*= An^encan wh,,.„g
fishery but its value at first sale was nearly eight times as great, weight for we.ght. The South A rtcan
ctchw- about 40% of the weight of the American catch, with the value at first ^k micrn,.d,.v.
bettrthTprici obtained in the other two countries: about 3! times the value °f *e A^- and
tr,hrhai?the value of hake landed in Britain. While the difference between Bnt.sh and South
A i an prict m yTn part be explained by the employment of cheap coloured labour tn South Afnca
fhr"ch chlper Amertcan prices point to profound differences in eeonomtc cond.t.ons wh.ch
cannot be understood without first-hand knowledge of the ' whnmg trade.
Twould seem therefore, that while it is obviously desirable to exam.ne our b.ologtcal dat con-
It would seem, me Knowledge accumulated concernmg the speces already
284 DISCOVERY REPORTS
by direct biological comparison alone. Peculiar economic and even political considerations — in fact,
the bionomics of the prospective human producers and consumers — play at least an equal part in
the determination of industrial possibilities. At the same time, some knowledge of the natural history
of the Patagonian species is obviously a basic necessity in any attempt to deal with the problem.
Our data are therefore considered here from the biological point of view, while suggestions as to com-
mercial prospects are deferred to a later section of this report where the summarized weight data
allow one to take into account the possible value of other less important species.
Table lo. Figures illustrating the size and relative values of the principal
fisheries for Merluccius spp.
Country, period, species,
source of information
Approximate average annual
catch in tons, cwt. or lb.
Average price at first sale per unit of
weight, in sterling or in %
Great Britain 1920-33
Merluccius merluccius
Hickling (1935)
United States 1939 and 1940
Merluccius bilinear is
Statistical Digests, i and 4
Union of South Africa 1929-32
Merluccius capensis
von Bonde (1934)
36,904 tons or 738,088 cwt.
or 82,666,000 lb.
19,767 tons or 395,335 cwt.
or 44,277,500 lb.
7845 tons or 156,900 cwt.
or 17,572,800 lb.
£4,1. 6s. i^d. or I171.49; £2. IS. 6d. or
I8.62 ; 4|^. or nearly 8 c.
^4. i()s. 2ld. or I20.62; 45. \i\d. or
$1.03; \d. or just under ic.
l\(). gs. 2,d. or $79.15; igs. o\d. or
$3.95; zd. or 3|c.
THE SIZES OF PATAGONIAN AND EUROPEAN HAKE COMPARED
First, it is desirable to establish the sizes of the hake we captured and to see how they compare with
the sizes of better-known species captured with similar gear.
If it were possible to obtain ' ideal samples ' of a slow-growing fish like hake, whose length increases
almost as a Imear function of age, the length frequencies, when plotted graphically, would approximate
to the Ix curve of a life table (a curve like a left-handed ogive, but with the curvature reversed and
enormously produced at the very beginning, because of the high infant mortality rate). An imaginary
curve of this nature is shown by the solid Ime in Fig. 19. It would begin with astronomical numbers
of newly-hatched larvae, and end with the largest hake caught. Departures from this curve would, in
part, be due to slight changes in growth rate. Such fish grow rather faster than usual early in life and
slower near the end of their lives. The straight line of the age-length relationship becomes slightly
bent at the ends, approaching a parabola.
But the ideal length-frequency distribution would show other more important deviations about the
smoothed curve if considered with small class intervals (say i-ocm.): there would be modes or
shoulders whose magnitude would reflect the difl^erential survival of hake hatched in successive years
the resultant of all the environmental factors, animate and inanimate, that influenced their lives An
imaginary distribution of this type is shown by the pecked line in Fig. 19. Hake have a prolonged
spawning season, and the changing environmental conditions would not have favoured early- and late-
hatched fry equally in successive years. This would be a further source of variation in the modes or
shoulders in our imaginary curve, for it is their dispersion that helps in age determination, and this
even more than their magnitude, may thus be distorted. In actual hake samples this factor so com-
plicates the length-frequency distribution that Pettersen's method of age determination is rendered
unsatisfactory except for the younger (smaller) fish.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
28s
50:-
40^
202-
\oZ-
0 r
0
10
In practice samples of a hake population are modified by the catching power
of the trawl used. An ' idealized ' trawl, that captured all of the fish above a
certain size in its path and allowed all the others to escape, can be imagined to
catch samples showing a length-frequency curve of the same pattern as our
' ideal ' sample, but with the point of origin at (say) 40-0 cm. instead of the
length of the newly-hatched larvae. Unfortunately such ideal conditions do not
exist, and the selective action of a sampling instrument such as a commercial
otter trawl of normal mesh retains a varying proportion of the smaller fish
below the size at which all are captured. The proportions of these smaller fish
retained varies in regular fashion according to their size, and follows yet another
ogive curve, the ' selection ogive ' of the trawl used. Selection ogives for trawls
of diflFerent mesh taking European hake are given by Hickling (1933, p. 71,
fig. 38). It will be seen that as the size of the mesh is increased and the
selection ogives become centred higher in the length scale the curves slope less
steeply. Theoretically it would be possible to obtain length-frequency samples
approximating to a normal distribution by increasing the size of the mesh.
By covering an ordinary trawl with shrimp netting, Hickling has been able to
obtain samples which, when treated cumulatively over five annual surveys of
the grounds to the south and west of Ireland, show length frequencies approxi-
mating to the smoothed curve one would expect in ' ideal ' samples for hake
upwards of two years old. No practical gear could sample the younger length
classes simultaneously, for the minute fry are pelagic, and the yearlings do not
inhabit the same grounds as the older fish. Curves of percentage length fre-
quency of European hake, derived from Hickling (1933, table IXa), are shown
in Fig. 20, where they may be compared with the age distribution (a partial
'life curve') upon which that particular series of mean lengths was based.
These data of Hickling's also give us a valuable demonstration that the numbers
of the two sexes of European hake are nearly equal, the actual ratio from that
set of figures showing a slight preponderance of males. If samples such as can
be obtained with a commercial trawl are considered, it is found that females
preponderate to a considerable extent, owing to the higher escape ratio of the
smaller males.
Unfortunately, such data as we have for Patagonian hake are not directly
comparable with those for the European hake set out in Hickling's table IXa,
but they do compare very well with mean values of the figures set out in his
series of measurements
on commercial trawlers,
of European hake from
areas 1-5 (Hickling, 1933,
tables XI a, XI b). These
areas cover that part
of the European hake's
geographical range most
nearly comparable to the
areas best sampled by the
' William. Scoresby ' when
\\. ^^
80
90 100 110
Fig. 19. Imaginary length frequency curves for an 'ideal sample' of hake.
Solid line: smoothed curve; pecked line: with small class intervals.
286 DISCOVERY REPORTS
catching hake off Patagonia, and direct comparison of the two sets of length frequencies seems highly
instructive.
These data of Hickling's are given in the form of percentage length frequencies for each month,
irrespective of sex, over a period of more than two years. To facilitate direct comparison with our
southern data the results have been recalculated to exclude the very small number of fish less than
20-0 cm. long, and then meaned. For the southern species we have 4704 measurements of hake
caught with a commercial trawl at all seasons, between 42 and 53° S, the latter probably being the
normal southern limit of the species. In addition to the calculation of the mean, standard deviation, etc. ,
by the long method, they have been secondarily grouped into 5-0 cm. length classes, and the percentage
length frequencies computed, so that curves could be drawn for direct comparison with Hickling's
data.
-1
30
so 60 70
LENGTH _ CMS
u In IbtItI ytIyhImiI kIx Ixr Ixnl
YEAR CLASSES
Fig. 20. Percentage length frequencies, and a partial life curve, for Merluccius merhicciiis. From
Hickling, 1933, table IX a.
Before this comparison is made, it is necessary to make clear one most important difference between
the two stocks : the males of the Patagonian hake are relatively much smaller than those of the European
hake. It will be shown that Patagonian hake are some 5-0 cm. smaller than European hake, on the
average, when the sexes are lumped together; but although directly comparable data for the separate
sexes are not available, it can be shown that males of the Patagonian stock are relatively much smaller
than this difference would indicate. Means for each sex of European hake may be obtained from
Hickling's table IX a. While the means for the Patagonian stock are not directly comparable with
these, the difference between the mean lengths for each sex, in each set of data, may readily be com-
puted; and the difference between these differences is probably significant, though this cannot be
established statistically. For Patagonian hake the difference is 49-9 cm. for females less 36-4 cm. for
males, i.e. 13-5 cm. For European hake, with data including a much higher proportion of the smaller
individuals of both sexes, it was 34-9 cm. for females less 31-0 cm. for males, i.e. 3-9 cm. Moreover,
if we take the largest decile of the 1396 Patagonian males, we find that they show a mean length of
46-1 cm., the largest individual was 64 cm. long, and only i % of the total of 50-0 cm. and over. The
largest decile of the European males has a mean length of 56-1 cm., while the largest individual age
group had a mean length of 767 cm. ; and some 7% of the total were 50-0 cm. long and over, in spite
of the much greater proportion of small individuals in this series of measurements.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 28?
From these considerations it would seem that if we assume the disparity in size beUveen the sexes
to be at least twice as great in Patagonian hake as it is in European hake, we shall not be far wrong.
Turning back to direct comparison of the lengths of the two stocks irrespective of sex, the figures
for European hake being derived from Hickling's (1933) tables XIa and Xlb as previously stated, we
obtain the curves of percentage length frequency shown in Fig. 21 a. This shows higher frequency ot
the European species in the higher length classes; but since all frequencies m these higher length
classes are low this difference is much better illustrated if the results are plotted on an anthlog scale,
as in Fig 2ib\t must also be remembered that Hickling has shown that the figures for the European
species were derived from a heavily overfished stock, in which the proportion of large fish had been
seriously depleted, so that the smaller size of the virgin Patagonian stock is even more marked than
can be shown from these figures. • r ^u
Proceeding to direct comparison of mean lengths it was found that the secondary grouping ot the
southern data leads to an error of +0-5 cm. Further, the secondary grouping m itself reduces the
significance of the difference between the means, for using percentage length frequency as the basis,
iV becomes 100, and the value of cr, is grossly exaggerated. It is not surprising, therefore, that on
comparing the means of the two sets of percentage length frequencies and applymg the test rf/a,>3,
their difference cannot be shown to be significant. Thus :
European: Mi = 5i-4, ffi=i5-475> (?f = 239-476, Ni=ioo;
Patagonian: M2= 46-4, (72=12-535, al=i57-i26, A^2=ioo;
J=Mi-M2=5i-4-46-4 = 5-o.
d 5-0
. — = ^ — = 250.
Orf 1-99
If, however, we take the true values for the southern species, calculated by the long method from
all the individual observations, we have :
European (M. merluccius) values as before ;
Patagonian (M./?M6fc«):M2= 45-9. 0^=12-^2, al= 154-33. iV2 = 4704;
J=Mi-M2=5i-4-45-9 = 5-5.
^ __ /(4+4U /(219j476i54l3_3\ ^2-42757= 1-56,
°^~VliVi N2' V\ 100 ^4704/
Oa 1-56 ^^^
Here the difference is clearly significant. Moreover, if the number of individual observations upon
wHchtrln of Hickling's means depends were known. N, would be much larger, and the signifi-
rnnre of the difference even more marked as Oa diminished. i„„„tl. nf
Even ift assume that the unknown grouping error introduced by ^^'^^'^'^f^^l^TJ^^^^^
the European species from the secondarily grouped percentage frequencies was of h s^me magm^^^^^^^^
and of opposite sign, from that which was found for the Patagonian data so treated, we
Mi=5o-9, Patagonian values as before, ^=5-° and ad=i-63,
1=. 1^ = 3.07.
Oa 1-63 -*
j88
DISCOVERY REPORTS
50
50 70 ao
LENGTHS _ CMS
90
100
110
'i';:cSrc£Sx-ijs='t^t3J:sr,iSi^^^^^
60 70 80
LENGTH5_CM5
Fig. 21 J. The same as Fig. 2i«, on an arithlog scale.
90
100
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 289
It therefore seems quite clear that Patagonian hake really are some 5-0 cm. shorter than European
hake, and that the disparity would be even more marked if it were possible to compare the males
separately.
Another interesting point is brought out by comparison of the length-frequency curves: Patagonian
hake, despite their smaller ' average' size, are relatively more numerous in the 50-60 cm. length classes
(to the right of the mode) than are European hake. The sharp inflexion above the mode is doubtless
due to scarcity of males over that length; but it can be seen that the succeeding flattened portion of
the curve is on a distinctly higher level than the corresponding part of the curve for European hake of
about 5 cm. greater length. The 50-60 cm. hubbsi—^hont 'ordinary chat' size by Fleetwood market
standards of the period-probably correspond physiologically, if not in actual age, to European hake
one category larger; so that it seems to me that we have here a striking indirect confirmation of
Hickling's proof that the European stock was heavily overfished.
THE DISTRIBUTION AND RELATIVE ABUNDANCE OF MERLUCCIUS HUBBSI WITHIN
THE AREA SURVEYED, AND THE EFFECT OF LATITUDE ON
NUMBERS, SIZE AND SEX RATIO
The general distribution and abundance of this species cannot satisfactorily be shown on charts of
such scale as could be reproduced here, because of its widespread occurrence and pronounced
migratory movements. These are essentially similar to the migrations of the European and north-west
Atlantic species, as will be shown in a later section. The occurrences of the species at our trawling
stations have been tabulated hi exienso in Appendices II A, B and c. From these it can be seen that
the most important feature of the general distribution is a marked decrease in relative abundance
from north to south. Local concentrations were encountered inshore in autumn (St. WS853, Wb855),
offshore in winter (St. WS216, WS217), and at intermediate distances from the coast m early summer
(St WS7Q0 WS791). There were, of course, numerous less pronounced concentrations, most ot
which tended to conform to the general pattern of migration suggested by the extreme examples
quoted. . , ....
Now since hake tend to be more closely congregated when on their inshore spawning migration,
and the larger fish tend to move inshore first (cf. Hickling, 1927, P- 59. on the European species), a
series of observations in early summer might give an erroneous impression of the effect of atitude on
the size of the fish. Spawning takes place earlier in the year in the more equatorial part of the range
of the species, and so considerable concentrations of individuals larger than the average for their
latitude may be sampled when fish in higher latitudes are not so concentrated. At the same time,
moderate numbers of the smaller hake, which do not seem to migrate so far or so fast, can nearly
always be found in relatively shallow water throughout the year. In Merlucaus Imbbstj. found that
the resultant of these factors completely masked the effect of latitude upon size of fish in December,
but this effect was quite clear when the data for all seasons were considered together. A small series
of observations in a single longitude taken over a narrow time interval later in the year, when the sma ler
individuals were at the peak of their inshore movement, also showed the effect of latitude q-^e clearly^
The general decrease in abundance of hake from north to south of our area is deinonstrated by the
figures in Table 11, which are taken from eighty-three hauls spread over the whole of the third
survey. Earlier resvdts are m agreement with these, but are not considered here because comparable
weight data are lacking : ^
290
DISCOVERY REPORTS
Table ii. Decrease in relative abundance o/ Merluccius hubbsi with increasing latitude
Northern
region
42-46° S
Intermediate
region
46-50° S
Southern
region
50-54° s
Number of hauls
Hours trawling
Mean number of hake
per hour's trawling
Mean weight of hake
per hour's trawling
14
23
100-48*
54-817 kg.
29
44
20-59
19-399 kg.
40
60
6-67
7-464 kg.
* Obviously fractional hake could not long exist in nature, and it is mathematically indefensible to treat fish as indiscrete
objects— but I feel that the fractional expression is less misleading than giving results as hake per 100 hr. trawling in order
to get whole numbers, because these results are based on less than 100 hr. of comparable hauls-nn each region.
It will be seen that these figures provide evidence of the second feature of the influence of latitude
already mentioned, namely, the increase in size (and weight) of the individual fish as one proceeds
southwards. This can better be demonstrated by considering the mean lengths and length frequencies
of the fish caught in the three regions, including data from the earlier surveys (Table 12).
Table 12. Variation in size 0/ Merluccius hubbsi, as shown by the differences in mean lengths for all
comparable hauls in the three regions here surveyed, regardless of season. The sexes considered
separately
Note. The numbers of individuals do not indicate the relative abundance in the respective regions because of the very
difl^erent number of hauls made m each of them. They show merely the number of individual measurements upon which
these mean lengths are based. The sex ratios shown at the head of the table are based on a different array of the data in-
cludmg some specimens sexed but not measured. '
Northern region
Intermediate region
Southern region
Sex ratio, % males
54-1
24-0
19-3
Males:
No. measured
Mean length
oMl
<yyN
901
35-9 cm.
5-5740
0-0345
475
37-7 cm.
6-1837
0-0805
77
38-9 cm.
5-4056
0-3795
Difference in Ml ( = d)
Significance
1-8 cm. 1-2 cm.
0-3391 0-6783
5-3 1-77
Strong Not significant
Females :
No. measured
Mean length
aMl
oUN
1260
47-3 cm.
12-5634
0-1253
1616
50-2 cm.
i 1-2877
0-0788
535
55-5 cm.
12-0040
0-2693
Difference in M/ ( = d)
.-. dlca =
Significance
2-9 cm. 5.3 cm.
0-4518 0-5900
6-4188 8-983
Strong Strong
Table 12 shows the mean lengths of males and females of M. hubbsi, based on all the comparable
data, for the three latitudinal regions, the difl^erences in mean lengths between adjacent regions and
their statistical significance. It can be seen that there is a clear increase in mean length with increase
in latitude m both sexes, and that this is strongly significant except as between the males of the inter-
mediate and southern regions. It can also be seen from the sex ratios given at the head of the table
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 291
that males become increasingly scarce towards the south, and it is probable that this exception is due
merely to insufficient sampling of the more southerly males caused by this scarcity.
Table 13. Observed differences in populations 0/ Merluccius hubbsi sampled in different latitudes
between 21 March and 30 March, 1932, in long. 64° 15' W
Stations considered
Mean latitude
Hake per hr. trawling
Sex ratios, % males
Males:
No. measured
Mean length
Northern region
WS853 and WS855
45° 39' S
702
76-4
544
35-2 cm.
5-1676
0-0491
Intermediate region
WS857, WS862 and WS864
48° 21J' S
37
37-8
Southern region
WS866 and WS868
Si-iol'S
27
16-7
42
39-3 cm.
5-6454
0-7588
9
42-4 cm.
1-5720
0-2746
Difference mMl{ = d)
.'■ djaa =
Significance
Females:
No. measured
Mean length
aMl
Difference in Ml (
<^d
.'. djaa =
Significance
--d)
4-1 cm.
0-8988
4-56
Clear
3-1 cm.
1-0166
3-p4?
Just significant
332
39-9 cm.
10-3011
0-3196
69
48-3 cm.
11-0042
1-7550
45
56-0 cm.
9-1894
1-8765
8-4 cm.
1-4404
5-832
Strong
7-7 cm.
1-9057
4-041
Clear
Table 13 shows a similar consideration of more limited data from each of the three regions m the
last ten days of March 1932 in one longitude. This array of the data also permitted comparable figures
for relative abundance (hake per hour's trawling) and sex ratios to be given at the head of the table
Both show marked diminution towards the south. This falling off in the proportion o males is the
third important feature associated with increasing latitude in populations of M. hubbsi It would
seem to imply that spawning activity must be much reduced near the southern limits of the range of
*FurthlTdemonstration of the effect of latitude upon size is given by Fig. 22. Here the length
frequencies for either sex, m each of the three regions, have been summed into 5 cm. §-"?«' -^uced
to percentages, and the results plotted graphically. The curves show quite clearly how small fish
become mo?e are, and larger fi^l. commoner, as one proceeds southwards. Incidentally the figure
p vrds a good iUustration'of the unusually large discrepancy in size between the sexes oi MMbbn
L has already been mentioned. Females are significantly larger than males among a large majority
of fishes, but it is unusual to find differences as great as these.
THE RELATION BETWEEN LENGTH AND WEIGHT OF MERLUCCIUS HUBBSI, AND
THE RELATION m.^i ^^ ^^ INDICATOR OF THE SPAWNING SEASON. AND
FOR OTHER PURPOSES
During our third survey the weights of male and female M. hubbsi were reeorded m the form of bulk
Durmg our tmra s J ^ . , ^,^„^^^ b,; 3ed. Hieklmg (.930*, PP- 7. 8) has proved
Zf A^S 1 d remaSly rat'e figurl at sea. even 'when used for much more delicate weighings
tn anTaUempted by us^The lengths of our fishes were known individually, so that a '.rue mean
292
DISCOVERY REPORTS
40 50
LENGTHS _ CMS
Fig. 22 0. Percentage length-frequency distribution of male Merluccius hubbsi in different regions, plotted as 5 cm. groups.
Thick line: northern region; pecked line: intermediate region; thin line: southern region. Note. The fish increase in size from
north to south.
30
40
50
60 70 80
LENGTHS- CMS
r
90
100
1
110
Fig. Z2h. Percentage length-frequency distribution of female Merluccius hubbsi in different regions, plotted as 5 cm. groups,
lorth t^s'ouTh '' ''^'°"' ^ intermediate region; thin line: southern region. Note. The fish increase in size from
These graphs also serve to show the great disparity in size between the sexes of trawl-caught M. hubbsi.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 293
length could be used to compute the ponderal index or condition factor K from each weight recorded.
All the reliable data of this nature are given in Appendix lie. From them it has been found possible
to trace the seasonal variation in K by using mean values from appropriate groupings of the results.
Throughout this work the 'ponderal index', ' condition factor' or ' weight-length coefficient' K has
been calculated from the formula
„ w
K=™Xioo,
where w = mean weight in grams and Z=mean length in cm.
The results suggest important analogies with those Hickling (19306) obtained for the European
species, but there are divergences, some due to the different methods employed, others probably to
inherent differences between the two species. Here again it must be emphasized that Hickling,
dealing with a single species, was able to concentrate his efforts, and weighed individual gutted fish
and organs. Gunther and Rayner, on the other hand, were investigating virgin ground, so that it was
necessary for them to ' do something about' everything that came up in the trawl, including ' rubbish'.
The marvel is that they found time to make as many bulk weighings as they did, in addition to the
vast number of individual measurements.
K values derived from heavily grouped data like these may still give us a good idea of the broad
outline of the seasonal cycle for the species, although they cannot be expected to prove so accurate as
more detailed results, especially if one were to attempt to apply them to individual fishes. They are
mean K values, derived from meati weights and mean lengths. Applied to whole samples of fishes they
permit of surprisingly accurate estimates of weight from known mean length (and the converse) as
will be shown later, and these may be of great practical value. If, however, they are applied to the
study of variations in condition of limited subsamples, they show features in contradiction to what
one would expect from Hickling's more detailed work on the European species.
Apart from the seasonal variation in condition which this method is particularly adapted to show,
there is superimposed upon it a secondary variation related to the length (here the mean length) of the
fish. Older (longer) fishes tend to show a slightly lower level of condition throughout the seasonal
cycle, consequent upon the increased metabolic strain of spawning. The point of inflexion on a curve
showing this diminution of K with increasing length is thus a good approximate indication of the
length at which sexual maturity is attained. Analogous findings in several other species of fishes
could be quoted, but probably the best general exposition of the more important deductions to be
derived from the study of K values is that given by Sir D'Arcy Thompson (1942, p. 194 et seq.).
In M. hubbsi the ' average 'i values of K in relation to length of either sex are shown in Fig. 23.
From this it can be seen that a majority of the males probably mature at a length of about 32 cm.
(some certainly do so when still smaller), while it is probable that most of the females below 42 cm.
length are immature. Considering this in conjunction with our previous findings of a difference of
over 10 cm. between the sexes (in trawl-caught samples), it seems possible either that the males
mature at least one year earUer than the females, or that the early growth rate of females is li times
as fast as that of the males. The former is more probable, but the great scarcity of males over 50 cm.
long while females diminish in numbers more gradually to twice that length, makes it fairly certain
that the growth rate of the older mature males is much lower than that of females of comparable size.
Condition factors for males of 50 cm. and over were too few to include in any 'average' curve, but
those we have (records of five individual fish) are all very low, suggesting complete analogy with
1 I use this expression 'average' to indicate values based on means of means, not on the arithmetic mean that would be
derive^from full^integration of fll the constituent data. This device is necessary here m order to ehmmate the seasonal eflFect
when studying the relation between length and ponderal mdex.
294
DISCOVERY REPORTS
Hickling's (1933, pp. 43-4) observations on the very rapidly increasing metabolic strain of spawning
among older males of European hake.
By selecting data from Appendix lie (samples in which five or more fish were weighed, northern
and intermediate regions only, and working to mean dates) we can draw curves of the variation
in K over the period covered by the observations, for either sex of M. Inibbsi, as shown by the con-
0700-
0-650
30
40
50
LENGTH IN CMS
60
I
70
Fig. 23. Variation of 'average' K with length in male and female Merluccius hubhsi.
0750-
0700
O650
/
^9\
I JULY I AUG|SEPT|OCT|N0V| DEC | JAN | FEB | MAR |APR | MAY |JUNE|JULY| AUG|SEPT|oCT |
Fig. 24. Seasonal variation in the condition factor A' of Merluccius hubbsi, for fish approximating to the
mean length of either sex; for further explanation, see text.
tinuous lines in Fig. 24. Now these curves show a steep fall in the value of K during the summer,
which tends to level out towards autumn. This is an almost certain indication that this hake is a summer
spawner, as other species of hake elsewhere are known to be. All temperate fishes hitherto studied
reach their peak of condition just prior to spawning, and some show very quick recovery from the loss
occasioned by that act. Thus the seasonal cycle takes the form of a harmonic curve, and the relative
steepness of the left- and right-hand sides of the 'wave' depends («) upon the speed of recovery after
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 295
spawning and {b) upon the extent to which spawning is Umited in time. On the assumption that
M. hubbsi conforms to this rule (as European hake are known to do), a purely hypothetical completion
of the harmonic curves has been shown by the pecked lines in Fig. 24, for that part of the seasonal
cycle for which we have no data.
Obviously such curves must regain a similar level for the beginning of the next seasonal cycle,
and there are good grounds for belief that recovery after spawning is rapid, so that they will rise more
steeply in autumn than in winter. The hypothetical portions of the curves are therefore almost cer-
tainly more accurate than might seem probable at first sight.
Before we proceed to a more detailed examination of the evidence provided by our ponderal indices,
and their shortcomings, it may be pointed out that the generalized curves shown in Fig. 24 provide
valid evidence of one important analogy with Hickling's findings on the European species— vahd
because it depends upon the portions of the curves derived from actual data and not from hypothesis.
It can be seen that while the fall in condition of males and of females in early summer is nearly
parallel, the females begin to recover first. The inference is that the spawning season of a majority of
males is longer than that of a majority of the female fish. The agreement with Hickling's (1930*.
pp. 33-4) observations on M. merliiccius seems complete.
The generalized curves show a greater annual variation among males than among females. Since
eggs are larger than sperms it might appear that this is anomalous, but it is probable that it is not so,
for the following reasons: the smallest length classes of males contain a far higher proportion of
mature fish than corresponding length classes of females, and as the male growth rate falls off the
metabolic strain of spawning increases far more rapidly than in females (cf. Hicklmg, if)Zob, p. 36;
1933. PP- 43-4)- Consequently these curves show a greater annual variation among males just because
they are generalized: the data for males include a greater proportion of spawning fish.
When we come to consider the seasonal variation in condition within individual length groups
certain undoubted anomalies appear. In the following table the data used to compute the relation
between K and length are also subdivided to show the means at mean dates within length classes.
Several untoward features are at once apparent. Males of the 21-30 cm. length class showed greater
seasonal variation than males of 31-4° cm., and females of the smaller length classes a greater seasonal
range than larger females.
Table 14. Variation in ' average' K of Merluccius hubbsi in relation to length, and seasonal
variation of K zvithin length classes
(The left-hand columns are the data of Fig. 23.)
Length
class
21-30
31-40
41-50
'True'
mean
length
27-4
35-7
43-1
21-30
26-8
31-40
35-8
41-50
45-9
51-60
55-3
61-70
647
71-80
73-4
'Average' K
in relation
to I
0-705
0-688
0-654
0-678
0-695
0-685
0-649
0-628
o-6i8
Males
Females
Mean K at mean dates within I classes
31 Oct.
0-791
0-731
0-751
0-736
0-767
0-747
0-686
0-662
0-633
16 Dec.
0-705
0-685
0-657
0-677
0-647
0-667
0-636
0-614
0-621
24 Mar.
0-630
0-647
0-555
0-621
0-671
0-641
0-624
0-608
0-600
296 DISCOVERY REPORTS
The anomaly among the males mentioned above is almost certainly due to inclusion of an ex-
ceptionally large proportion of spawners of the smallest length class at one of the rich March hauls,
with inadequate sampling in the earlier periods. The curious anomaly of 31-40 cm. females showing
recovery by March, while smaller and larger fish did not, is probably due to inadequate sampling also.
As it stands one would think it might indicate that some of this small group of mainly immature fish
completed their cycle before their older sisters, and we know that the reverse is more generally true.
However, there were indications of early shoreward movement among females of that length class in
193 1-2, derived quite independently from this data, and it might be that our samples of it happened
to include a majority of some year class which had got out of phase with the norm for the species,
through exceptional conditions affecting them while still younger, or some such cause. In fishes with
such a prolonged spawning season as hake non-conformity of this type is bound to occur from time
to time, but to prove the point is quite another matter, and quite impossible with such limited data
as ours.
The mere fact that mainly immature length classes showed a seasonal variation parallel to that for
older fish need occasion no surprise, for Hickling (19356) has established that a seasonal cycle fore-
shadowing the full sexual cycle of the adult takes place in immature European hake. The point is that
the adult cycle should show the greater range. Here I believe that it is the limitation of our data in
time that is at fault, and not insufficient sampling.
We have considerable evidence that larger females of M. hubbsi tend to move inshore earlier than
smaller ones, and it is probable that they are normally the first to spawn (a feature well known among
European hake). Now a generalized cycle such as the curve for females in Fig. 24 would only hold
strictly for fish of the mean or 'average' length considered, the 41-50 cm. length class. Had we been
able to obtain adequate figures for the whole year, it is probable that the whole cycle for the larger
fish would have been found to be centred earlier in the year. If so, means for the end of October are
not early enough to show the full extent of the annual variation in condition of these larger fish.
It is also possible that the anomaly shown by the females is not entirely due to the unavoidable
limitations of our data, although I think there is little doubt that it was collected too late in the year
to show up the peak period for the 51-60-70 cm. fish. In the first place, some M. hiibbsi under 40 cm.
long are mature. This would tend to increase the range of seasonal variation in mean values for the
smaller length classes far above the range observed in European hake of similar length. In European
hake, an altogether larger species, mature fish of such small size have only rarely been found. These
were among a peculiar localized stock in the Clyde basin (Hickling, 1930^', pp. 52, 53 and Table Via).
A further likely source of discrepancy is that the sexual activity of the largest size groups may be
reduced, or at any rate less regular, than in younger mature females. In European hake a reduction of
sexual activity in the largest length groups, associated with reduced growth rate, was postulated by
Belloc (1922, p. 40). Some concrete evidence in support of this view, relating to female hake of
90-100 cm., is given by Hickling (19306, p. 29).
The condition factors of a few very large female M. hubbsi (over 90 cm. long) are extremely in-
teresting. Unfortunately we only caught fish of this size in the southern region, so that the results
are not strictly comparable with those previously tabulated ; but it can be seen that all but the smallest
of these very large fish showed very high values for i^ at a time of year when all the other females
showed reduced values owing to spawning. These data are insufficient to be conclusive, but I believe
that the oldest and largest female hake have reached a state of suspended sexual activity, and no
longer show the seasonal variation in condition characteristic of younger fish (Table 15).
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
Table 15. Condition factors of the largest Merluccius hubbsi weighed
{Note. All were from deep water in the southern region.)
297
Station
Date
Weight
Length
K
WS817B
WS818A
WS819A
WS819B
WS820
14. i. 32
17.1.32
17-1-32
17. i. 32
18. i. 32
kg.
6- 100
7-300
5-000
7-400
6-650
cm.
94
100
92
96
93
0-734
0-730
0-642
0-836
0-827
A similar phenomenon is suggested among fishes of other species: TSning (1937, pp. 8, 23) has
shown that the largest of the west Greenland cod do not take part in the return spawnmg migration
to Icelandic waters which is characteristic of their slightly smaller sisters. It is certam that such
spawning as occurs in Greenland waters themselves is on quite a minor scale; moreover, the extra
large stationary west Greenland cod tend to be found near the extreme northern limit of the range
of the species, where spawning does not occur. The analogy with our observations on the largest
female M. hnbbsi seems very close, and the suggestion that the latter are past spawning gains strength
from the facts that males are scarce or absent from the most southerly waters where the large females
are found and that at no time have the latter been met with farther north.i
If the figures in the left-hand columns of Table 14, showing variation of K with length, are trans-
posed to a percentage basis, it is possible to read oS the correction to be applied to seasonal values
for K (as plotted in Fig. 24), for fish of given mean length, working to the mid-point of each month. ^
This has been done in Tables 16 (males) and 17 (females) for the length range likely to cover the
mean lengths of any sizeable sample of M. hubbsi.
With the aid of these tables the approximate weight of any sample of fishes of known mean length
can be calculated thus : W= — xi^xf
100
where / is the number of fishes in the sample, the sexes being dealt with separately and the results
summed. Some error is unavoidable, and these mean K values do not apply well to individual fishes.
(Hake weights are too erratic to be dealt with individually, unless they are weighed gutted. Food niay
be snatched up in the trawl or stomachs may be evacuated, and their gorging habits are such that
even ' in nature ' the weights of any two fish of the same length at a given time might be expected to
"^^ An Zttnl source of error is the distance in time from the mid-point of the month in which the
sample is taken, but equally obviously it is impracticable to tabulate K for every day of the year.
Such derived values cannot take account of abnormal seasons. By using the tables given it has been
found that with samples of ten or more fishes, using the formula for each sex separately, and checking
against eighteen samples of which the actual weight was known, the 'theoretical weights mean
. gross error was 6-7%, with a range of -16-4 to +8-4%. Further, it could be shown by summing
fhe squared deviatiot from known weights that the derived values tabulated gave a better estim^e
than other average values for K, such as mean K for the whole year, either umping the sexes or
treating them separately. The range of error is great, but from a purely practical P^^^ f ^^ ^ ^
be extremely valuable to know the probable rmnimum weight o catches where it w s ^^^^^^^'^ ^
get actual weights in the field. For this reason ' theoretical ' and probable minimum weights for the
unweighed catches of the first and second surveys have been calculated wherever the length data
permit and are tabulated along with the numbers of fishes in those catches in Appendices IIa and
1 Of course they may be there at depths greater than we were able to fish.
298
DISCOVERY REPORTS
II B. The 'theoretical' weights are derived direct from the formula as explained above, K values
being used from Tables 16 and 17; the 'probable minimum' weights are 18% less than these, i.e.
I have assumed a probable maximum error of — 18 % .
An individual example of the potential value of K may be given from our richest haul of M. hubbsi.
Here, at St. WS853, 1154 fish were captured in an hour. Of these, 943 were males and 21 1 females;
all the females and 414 of the males were measured, and also weighed in 10 cm. length groups. The
remaining males, 529 in number, were not measured, but were weighed in bulk. First let us see how
far out our estimates of weight for this catch would have been, using the appropriate values for K
Table 16. Table of K for male Merluccius hubbsi througliotit the year, computed as described
in the text. Interpolated values in italics
Length
in cm.
25- 1-26-0
26- 1-27-0
27-1-28-0
28-1-29-0
29-1-30-0
30-1-31-0
31-1-32-0
32-1-33-0
33-I-34-0
34-I-35-0
35-I-36-0
36-1-37-0
37-1-38-0
38-1-39-0
39-1-40-0
40-1-41-0
41-1-42-0
42-1-43-0
43-1-44-0
44-1-45-0
45-1-46-0
46-1-47-0
47-1-48-0
48-1-49-0
49-1-50-0
50-1-51-0
51-1-52-0
52-I-S3-0
53- 1-54-0
54-I-55-0
5S-I-56-0
56-1-57-0
57-1-58-0
July
Aug.
0-744
0-742
0-742
0-742
0-740
0-739
0-735
0-733
0-72J
0-722
o-yi6
0-709
0-701
o-6g2
0-683
0-671
o-66o
o-64g
0-640
0-634
o-62g
0-624
0-623
0-621
0-618
0-617
0-615
0-614
0-614
0-614
0-613
0-613
0-612
o-jyi
O-JJO
o-76g
o-76g
0-768
0-766
0-762
0-760
0-754
0-749
0-743
0-735
0-727
0-718
0-708
o-6g6
0-684
0-673
0-664
0-658
0-653
0-647
0-646
0-643
0-641
0-640
0-638
0-637
0-637
0-637
0-636
0-636
0-635
Sept.
0799
0797
0797
0797
0-795
0-793
o-78g
0-787
0-781
0-775
o-76g
0-761
0-752
0-743
0-733
0-720
o-7og
o-6g7
0-688
0-681
0-676
0-670
o-66g
0-666
0-664
0-662
0-661
0-660
o-65g
o-65g
o-65g
o-65g
0-658
Oct.
0-819
0-818
0-817
0-817
0-815
0-814
o-8io
0-807
o-8oi
0795
0-789
0-781
0-772
0-762
0-752
0739
0-727
0-715
0-705
0-697
0-693
0-687
0-686
0-683
0-681
0-679
0-678
0-677
0-676
0-676
0-675
0-675
0-675
Nov.
0-786
0-784
0-783
0-783
0-782
0-780
0-776
0-774
0-768
0-763
0-756
0-749
0-740
0-731
0-721
0-709
0-697
0-685
0-676
0-670
0-665
0-659
0-658
0-655
0-653
0-651
0-650
0-649
0-648
0-648
0-648
0-648
0-647
Dec.
0-705
0-704
0-703
0-703
0-702
0-700
0-697
0-695
0-689
0-684
0-679
0-672
0-664
0-656
0-647
0-636
0-626
0-615
0-607
0-601
0-597
0-592
0-590
0-588
0-586
0-585
0-583
0-583
0-582
0-582
0-581
0-581
0-581
Jan.
0-678
0-676
0-676
0-676
0-674
0-673
0-670
0-668
0-662
0-658
0-652
0-646
0-638
0-630
0-622
0-61 1
o-6oi
0-591
0-583
0-578
0-573
0-569
0-567
0-565
0-563
0-562
0-561
0-560
0-559
0-559
0-559
0-559
0-558
Feb.
0-662
o-66i
0-660
0-660
0-659
0-658
0-655
0-653
0-647
0-643
0-638
0-631
0-624
0-616
0-608
0-597
0-588
0-578
0-570
0-565
0-560
0-556
0-554
0-553
0-551
0-549
0-548
0-547
0-547
0-547
0-546
0-546
0-545
Mar.
0-658
0-657
0-656
0-656
0-655
0-654
0-651
0-649
0-643
0-639
0-634
0-627
0-620
0-612
0-604
0-594
0-584
0-574
0-567
0-561
0-557
0-552
0-551
0-549
0-547
0-546
0-545
0-544
0-543
0-543
0-543
0-543
0-542
Apr.
o-66g
0-668
0-668
0-668
0-666
0-665
0-662
o-66o
0-654
0-650
0-645
0-638
0-631
0-622
0614
0-604
0-594
0-584
0-576
0-571
0-566
0-562
0-560
0-558
0-556
0-555
0-554
0-553
0-553
0-553
0-552
0-552
0-551
May
June
o-6g3
0-718
o-6g2
0-717
o-6gi
0-716
o-6gi
0-716
o-6go
0-715
0-688
0-713
0-685
0-710
0-683
0-707
0-677
0-702
0-673
o-6g7
0-667
o-6g2
o-66o
0-685
0-653
0-677
0-645
0-668
0636
0-659
0-625
0-648
0-615
0-637
0-605
0-627
0-596
0-618
0-591
0-613
0-586
0-608
0-581
0-603
0-580
0-601
0-578
0-599
0-576
0-597
0-575
0-596
0-573
0-594
0-573
0-594
0-572
0-593
0-572
0-593
0-571
0-592
0-571
0-592
0-571
0-591
given in Tables 16 and 17. 414 males had a mean length of 36-5 cm., and 529 males unmeasured,
assummg the same mean length, give us a total of 943 males, i^for March males of 36-37 cm. (Table 16)
is 0-627. Hence the theoretical weight of the males is given by
W'~
100
X 0-627 X 943 g.
= 486-27 X 0-627 >^ 943 g-
= 304-89129 X 943 g.
= 287,512-48647 g., say 287-5 kg.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 299
21 1 females had a mean length of 41-4 cm. K for March females of 4i-i-42-o cm. (Table 17) is 0-695.
Hence the theoretical weight of the females is given by
W-
41-4'
X 0-695 X 211 g-
Table 17.
100
= 709-56x0-695 X 211 g.
= 493-14420x211 g.
= 104,053-4262 g., say 104-1 kg.
Table of K for female Merluccius hubbsi throughout the year, computed as described
in the text. Interpolated values in italics
Length
in cm.
34-I-35-0
35-1-36-0
36- 1-37-0
37-1-38-0
38-1-39-0
39-1-40-0
40-1-41-0
41-1-42-0
42-1-43-0
43- 1-44-0
44-1-45-0
45-1-46-0
46-1-47-0
47-1-48-0
48-1-49-0
49-1-50-0
50- 1-51-0
51-1-52-0
52-1-53-0
53-I-54-0
54-i-55-°
55-1-56-0
56-1-57-0
57-1-58-0
58-1-59-0
59-1-60-0
60-1-61-0
61-1-62-0
62-1-63-0
63-1-64-0
64-1-65-0
July
0-759
o-y6i
o-y62
0-763
0-763
0-762
0-762
0-760
0-759
0-757
0-755
0-753
0-750
0-749
0-746
0-743
0-740
0-737
0-734
0-731
0-728
0-725
0-723
0-720
0-718
0-715
0-712
o-7og
0-707
0-706
0-704
Aug.
0-777
0-778
0-780
0-781
0-781
0-780
0-779
0-777
0-776
0-774
0-772
0-770
0-768
o-j66
0-763
0-760
0-757
0-754
0-751
0-748
0-745
0-742
0-739
0-736
0-734
0-731
0-72^
0-726
0-723
0-723
0-720
Sept.
Oct.
0-796
0-799
0-798
0-801
0-799
0-802
0-800
0-803
0-800
0-803
0-799
0-802
0-798
o-8oi
0-797
0-800
0-795
0-798
0-793
0-796
0-791
0-794
0-789
0-792
0-787
0-790
0-786
0-789
0-782
0-785
0-779
0-782
0-776
0-779
0-773
0-776
0-769
0-772
0-766
0-769
0-763
0-766
0-760
0-763
0-758
0-761
0-755
0-758
0-752
0-755
0-749
0-752
0-747
0-750
0-744
0-747
0-741
0-745
0-741
0-744
0-737
0-741
Nov.
Dec.
0-789
0-750
0-791
0-752
0-792
0-753
0-793
0-754
0-793
0-754
0-792
0-753
0-791
0-752
0-790
0-751
0-788
0-750
0-786
0-748
0-784
0-746
0-782
0-744
0-780
0-741
0-778
0-740
0-775
0-737
0-772
0-734
0-769
0-731
0-766
0-728
0-763
0-725
0-760
0-722
0-756
0-719
0-753
0-716
0-751
0-714
0-748
0-71 1
0-746
0-709
0-742
0-706
0-740
0-704
0-737
0-701
0-735
0-699
0-734
0-698
0-731
0-695
Jan.
Feb.
0-720
0-700
0-721
0-702
0-723
0-703
0-723
0-704
0-723
0-704
0-723
0-703
0-722
0-702
0-720
0-701
0-719
0-700
0-717
0-698
0-715
0-696
0-713
0-694
0-7II
0-692
0-710
0-691
0-707
0-688
0-704
0-685
0-701
0-682
0-699
0-680
0-696
0-677
0-693
0-674
0-690
0-671
0-687
0-669
0-685
0-667
0-682
0-664
0-680
0-662
0-677
0-659
0-675
0-657
0-672
0-654
0-670
0-652
0-670
0-652
0-667
0-649
Mar.
Apr.
0-694
0-695
0-696
0-700
0-697
0-701
0-698
0-702
0-698
0-702
0-697
0-701
0-696
0-700
0-695
0-699
0-694
0-698
0-692
0-696
0-690
0-694
0-688
0-692
0-686
0-690
0-685
0-689
0-682
0-686
0-679
0-683
0-677
0-681
0-674
0-678
0-671
0-675
0-668
0-672
0-666
0-670
0-663
0-667
0-661
0-665
0-658
0-662
0-656
0-660
0-653
0-659
0-651
0-655
0-649
0-652
0-647
0-650
0-646
0-650
0-643
0-647
May
0-714
0-715
0-716
0-717
0-717
0-716
0-716
0-714
0-713
0-711
0-709
0-707
0-705
0-704
0-701
0-698
0-695
0-693
0-690
0-687
0-684
0-681
0-679
0-676
0-674
0-672
0-669
0-667
0-665
0-664
0-661
June
0-735
0-736
0-738
0-739
0-739
0-738
0-737
0-736
0-734
0-732
0-731
0-728
0-726
0-725
0-722
0-719
0-716
0-713
0-710
0-708
0-705
0-702
0-700
0-697
0-695
0-692
0-690
0-687
0-684
0-684
0-681
Thus the gross weight of the whole sample should ' theoretieally ' be 39i« kg or, fi'^f''^% '» %
tolet our ■ probable minimum- weight, we should sa, that at the very least 3« kg. of hake had been
The actual weights recorded at St. WS853 were:
414 males (weighed in 4 length groups) totalled 121-350 kg.
529 males unmeasured, weighed in bulk , iSS'OOO kg.
211 females (weighed in 6 length groups) totalled 103-290 kg.
Grand total 379-640 kg-
and the total for males alone, for comparison with the first calculation above, was 276-350 kg.
300 DISCOVERY REPORTS
From these it can be seen that the theoretical weights, using mean monthly figures for K from the
tables, gave errors of:
+ ii'i5/276*35 kg., or +4-07% for males,
+ o-8i/i03-29 kg., or +078% for females,
+ ii-96/379-64kg., or +3-15% for the total sample,
but consideration of the ' probable minimum ' weight would have saved us from any over-optimism.
Converting to units used among British practical fishermen, our theoretical 'guess' at the catch would
have been less than 2 stone too big in a catch of 7I cwt., and our cautious ' probable minimum weight '
would have involved the statement that the catch was at least 708 lb. or 6| cwt., instead of the 7I cwt.
that we happen to know that it was.
Another way in which the ponderal index K can assist us is well shown by the figures for this
station: the actual weights of the 414 measured males being known, 'true' mean K for them can be
computed for this individual catch thus :
j^ w 203-116 ,
K = j^ X 100 = -^^ — X 100 = 0-603.
Now we have the bulk weight of the 529 unmeasured males, and if we calculate their theoretical
weight from this figure, assuming the same mean length as the measured subsample, and find good
agreement between theoretical and actual weights, we have a good argument that our assumption as
to length is justified, thus :
P
^"^ loo '^ ^ ^ 529 = 486-27 X 0-603 X 529 g- = IS5-II4 kg.,
in fact we know that they weighed 155 kg., and the agreement is so close that there is little doubt the
unmeasured subsample really had a mean length almost identical with that of the measured one.
MIGRATIONS
We have seen that the seasonal changes in condition of M. hiibbsl correspond to those that occur in
European hake. Variation in condition is primarily connected with the sexual cycle, and since the
mam inshore movement of European hake is a spawning migration, it was expected that the move-
ments of Patagonian hake would also correspond with those of the better-known species. It was by
no means easy to demonstrate this from our scattered data. The necessity for investigating the whole
area as fully as possible prevented us from repeating observations within more localized portions to
the extent desirable when trying to follow the seasonal movements of a single species.
One of the most valuable features discovered by Hickling in his work on European hake was a
direct relation between increase in depth and size of fish, which may be locally reversed at the time
when the larger fish are moving inshore. By watching the seasonal variation in size over limited
ranges of depth one may thus obtain valuable clues as to seasonal movements. But the peculiar
topography of the Patagonian Continental Shelf, with its uniform depths prevailing over vast distances,
was found to defeat this method of attack. Although correlations were found between depth and size
of the fish, and these showed the change of sign with the seasons that one would expect if they moved
in a fashion analogous to that of the European species, the correlations were not large enough to be
considered significant. The slight gradient of the Patagonian shelf thus masks a feature that is beauti-
fully clear off western Europe, where the shelf slopes more steeply and the edge is less abruptly
defined, giving a more continuous depth gradient in most localities.
301
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
However, the distribution of exceptionally rich hauls of M. hubbsi clearly indicated the possibility
of seasonal migrations similar to those of M. merluccius, and eventually it was found that this could
be shown with some certainty when relative abundance and size differences were considered in
relation to distance from the mainland coast. It was necessary further to restrict the data to be con-
sidered, so as to follow the changes within limited areas (the northern and intermediate regions con-
sidered separately), for it is well known that such movements may take place at different times in
different parts of the range of a widely distributed species.
In the northern region, three series of observations at different seasons seem together to give
reasonably good evidence of shoreward movement in summer, and will now be described in some
detail.
In October-November there are data from nine stations that can be arranged according to their
distance from the coast, and which were completed within a reasonably narrow interval of time
(see Table i8).
Table i8. Merluccius hubbsi captures betzveen 15 October and 3 November 1931 in the northern
region, with their distance from the coast
Station
Depth
m.
Distance
offshore
sea miles
Numbers of hake
Total hake,
remarks
Males
Females
WS762A
WS762B
WS777
WS763
WS771
WS764A
WS764B
WS765
WS772
66
66
98
85
90
108
107
116
236
H
10
43
75
130
145 1
140/
212
236
0
0
0
?
41
12
II
0
0
0
0
?
50
15
16
9
0
3 juv. in accessory net
0 (torn)
1 5 not sexed, some juv.
91
27
27
9
Consideration of the lengths of the females, together with the distance intervals, seemed to justify
a further lumping of these data, to the form shown in Table 19.
Table 19. Summary of October-November data bearing on hake movements in the northern region
Distance
offshore
sea miles
Total
hake
Hours
trawling
(no. of
hauls)
Hake
per
hour
Males
Females
Sex
ratio
No.
Mean
length
No.
Mean
length
Less than 100 miles
100-200 miles
Over 200 miles
18
117
36
3(3)
3(3)
Ii(2)
6
39
28
53
II
35-8
32-4
65
25
43'5
50-1
45% ^cJ
30% cJc?
From these two tables it seems fairly clear that at this time of the year the hake were mamly con-
centrated more than 100 miles from the coast. A shoreward movement may have begun. (St. WS771,
the richest of this series, was the innermost of the three between 100 and 200 miles offshore.) The
length data do not help here; one would not expect the smaller females to head the advance, and m
fact the size distribution might well be indicative of a prolongation of winter conditions. The offshore
males were too few for the size difference to be considered significant for that sex. However, the sex
ratios suggest the beginnings of concentration shoreward, and it is possible that bigger females catch
up and pass the smaller ones later. (There is considerable diffuse evidence that the speed and extent
of migratory movement is a function of size of fish.)
302 DISCOVERY REPORTS
In December there were only six stations suitable for this comparison, but at three of them additional
4-hr. hauls were made, so that there is a considerable body of data over the most critical part of the
distance range. The whole series of observations was completed between 13 and 16 December 1931,
so that there is little chance of the comparison being vitiated by the time factor.
Table 20. December data bearing on hake movements in the northern region
Station
Depth
m.
Distance
offshore
sea
miles
Hake nos.
Hours
trawling
(no. of
hauls)
Hake
per
hour
Mean
length
Sex
ratio
%
males
Males
Females
Total and
remarks
Males
Females
WS788
WS789
WS790A + B
WS791A + B
WS792A + B
WS793
8S
94
100
97
107
no
27
53
85
121
147
17s
4
5
58
53
13
0
2
6
319
15°
74
7
6( + 24Juv.)
II
377( + SJuv.)
203
87
7
1(1)
5(2)
5(2)
5(2)
1(1)
6 ( + 24)
II
75( + 0
40
17
7
37-8
30-6
31-8
31-9
36-8
6o-8
617
51-2
43-3
54-6
54-0
66
45
18
26
15
0
I
The results are summarized in Table 20, and show that the greatest concentration was at St. WS790,
85 miles offshore. The length data for males are again unsatisfactory because of the wide dispersion
and smallness of the samples containing the larger fish. I suspect that the males were still widely
dispersed, as is suggested by the low sex ratios at the richer stations. Here the female length data are
very interesting, the mean length at the peak station being considerably higher than at the richest
station of the earlier series, although 45 miles farther inshore. Thus
Difference ^=7-2 cm
Mean length ?, WS790 A + 6 = 51-2,
Mean length ?, WS771 =44-0,
\/i-^i()= 1-19
6-05. Strongly significant.
0-4672,
N'
^=0-9488.
od=\/{o-i\.6']2 + 0-9488) --
d 7-2
Od 1-19
This agrees well with the view that larger females may catch up and pass the smaller ones on their
way inshore. But, considering the December figures alone, we find that at St. WS791 A + B the
females were significantly smaller than those on either side of them ; the relation of length to distance
offshore is discontinuous. The detailed length-frequency distributions showed that this was due to
a much higher proportion of immature fish, especially of about 30-32 cm. length, at St. WS791. It
would seem that large females are definitely heading the shoreward movement at this time, a few
having penetrated right inshore among the juveniles, which are perhaps almost non-migratory. Com-
parison with the October-November results suggests that in the interim (about 7 weeks, taking mean
dates) the large fish have travelled shorewards some 100 miles, while the smaller fish advanced some
15 miles only.
In March also there were only six stations providing comparable data from the northern region,
but fortunately five of these yielded rich hauls. The remaining station, WS860, presents some
anomalies that spoil an agreement with our theory that is otherwise complete ; for a small number of
large hake were found less than 100 miles from the mainland, where a good haul of small hake should
have been taken. It is probable, however, that the net did not fish properly at this station. A note in
Gunther's hand in the original rough log reads : ' Haul disappointingly small. A few hake escaped.
Those present of large size— majority caught in after wing which opens suspicion that net may have
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
303
been fishing foul.' Possibly, therefore, we should be justified in disregarding the results from this
anomalous station as 'not strictly comparable'. The data from the series as a whole, however, appear
to fall quite naturally into inshore and offshore groups, and when so lumped the doubtful result from
St. WS860 is completely swamped by the more abundant data from the other two inshore stations.
Table 21. March data bearing on hake movements in the northern region (21-25 March 1932)
Station
Depth
m.
Distance
ofFshore
sea
miles
Hake nos.
Hake
per
hour
Mean lengths
Sex
ratio
%
males
Males
Females
Total, and remarks
Males Females
WS853
WS860
WS855
WS859B
■ WS859A
WS8s8
90
102
112
108
108
127
44
77
80
142
146
212
943
5
130
64
113
28
211
16
120
42
82
62
1154
(only 414 a measured)
21 (? net foul)
250
106
19s
90
^"54
21
250
106
195
90
36-4
38-0
33-8
38-5
38-9
40-9
41-5
54-6
37-8
45-9
45-3
50-3
8i-7
29-4
52-0
60-4
S7-9
31-1
The data from individual stations are shown in Table 21. The very large catch at St. WS853, the
station nearest the land, was the best we obtained at any time. The hake here were of small size, but
some were ripe, and the high proportion of males was also noteworthy. It would seem clear that the
smallest mature females are the last to spawn in any given season.
Table 22. The March data combined into inshore and offshore groupings
Distance
ofFshore
sea miles
Less than 100
More than 140
Total
hake
1425
391
Hours
trawling
(no. of
hauls)
3(3)
3(3)
Hake
per
hour
475
130
Males
Nos.
549'
205
Mean
length
35-8
39-0
Females
Nos.
349
186
Mean
length
407
47-1
Sex
ratio
75
52
* =29 males were not measured at St. WS853, the number given refers only to the measured specimens upon which the
mean length given is based ; hence the discrepancy with the ' total ' column.
The combined data (Table 22) show very clearly that smaller fish predominated in the inshore
catches and that the proportion of males was higher inshore. The differences m mean lengths are
strongly significant by the usual statistical tests. Comparison with the December figures (Table 20)
indicates that a complete change in the hake population had taken place; in March at the inshore
stations, the females were much smaller and the males larger than in December. Another striking
feature of the March results is the high sex ratios-even at the offshore stations the proportion
of males was much higher than earlier in the year. This is probably due to the main concentration for
spawning being later in the summer than exact correspondence with the habits of the European
species would demand. Thus the larger oflfshore fishes were probably only just beginning to disperse
after spawning. In this respect M. hubbsi may come closer to M biMnearrs (where the correlation
between shoreward concentration and the peak of annual temperature is very strong) than to M
merUcciMs, where the first wave of larger spawners moves inshore at least two months b fore^he
maximum temperatures are reached, and the second wave contains ";7;'^77^^,^^^^^^^^^^^
be borne in mind that the range of temperature is much greater in the habitat of M hdrneans than m
ha o^rJ./.^.) The timing of the cycle of movement of M. hubbsi ^^Pf ^^^^ --^J^;^^^^^
that lateness of the whole 'plankton calendar' of these southern waters, which was described m the
304
DISCOVERY REPORTS
introductory section of this paper. It is not likely to be a direct effect of temperature (as seems
possible with M. bilinearis), for the range of temperature here is small, approaching British conditions
more closely than New England conditions.
The broad fact that smaller fish are still closely congregated inshore, while large fish are beginning
to disperse offshore, seems sufficiently clear for us to state that the March results are consistent with
the view that the seasonal movement of M. hiibbsi is essentially similar to that of other species of hake
elsewhere. Ponderal indices also support this view: the small inshore population, with a probable
majority of late spawners and immatures, showed average K 20% higher than that of the larger
offshore population sampled at the same time. The latter must have completed spawning, for they
showed the lowest average K values (around 0-550) recorded at any season.
In the Intermediate Region data covering more of the annual cycle are available, if we may con-
sider results obtained in different years as roughly comparable for our present purpose, but they are
scantier and less satisfactory than those from the northern region. We have already seen that the hake
diminish in numbers towards the south, so that this difficulty was only to be expected.
In October and November no satisfactory series of observations was obtained here, but the size
and abundance of the hake at two stations, WS773 and 775, 206 and 82 miles off the land respectively,
were consistent with the view that a winter type of distribution still prevailed. Hake were nearly
twice as numerous at the offshore station where they were very much larger than those found farther
in. The proportion of males was greater inshore. Twenty-two juvenile hake of indeterminate sex,
less than 20 cm. long, were also captured during this period. This was at St. WS776, 60 miles from
the land.
In December numerous observations were obtained during the third survey, but they were too
scattered, and the samples too small, to warrant individual treatment of the results. When the stations
are grouped according to their distance from the coast, the picture of frequency distribution obtained
IS m accordance with more conclusive results from farther north at the same season. It therefore
seemed legitimate to use similarly grouped data in studying the size distribution, etc., since although
either chain of evidence may appear bald and unconvincing by itself, they corroborate each other.
Table 23. December data bearing on hake movements in the intermediate region, 5 to 22 December 1931
Distance grouping
sea miles from
mainland coast
Actual
mean
distance
Hours
trawling
Hake nos.
Hake
per
hour
Mean lengths
Sex
ratio
% males
Males
Females
Total
Males
Females
I, 0-49
II, 50-99
III, 100-149
IV, 150-199
V, over 200
35
75
129
161
219
6
6
12
3
2
34
17
20
I
0
77
134
219
25
17
III
151
239
26
17
18
25
20
9
8
33-1
31-9
?*
42-0
41-9
457
56-8
58-6
55-5
30-6
II-3
8-4
3-8
0-0
* The mean would be misleading here (see text).
These data are summarized in Table 23. It will be seen that the relative abundance was greatest
between 50 and 99 miles from the coast, and that the proportion of males was highest close in to the
land. The length data for males are unsatisfactory owing to wide dispersion in the small samples. In
distance grouping III a mean length for males would be meaningless, for the sample was composed
of some very small and a few large fish with intermediate lengths entirely unrepresented. The more
abundant length data for females tell a consistent story: the differences in mean lengths of the two
inshore groups from all the offshore groups are significant. The inshore fish were smaller, and the
offshore population apparently very homogeneous at this time.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
305
Here, then, we have the suggestion that shoreward concentration is beginning, but is not yet so
well defined as in the northern region at the same period.
A few January observations, quite insufficient by themselves, fit in so well with the general theory
that I give them here in full (Table 24) and have included them in the general diagrammatic summary
of the observations on hake movement in Fig. 25. Following on the December observations these few
January stations clearly suggest an increase in the tendency to shoreward concentration. This seems
mainly due to an incursion of larger females (just as we should expect), for the difference in mean
length of over 5 cm. between the peak station WS809 and the inshore figures for distance grouping II
of the previous month is strongly significant.
Table 24. Observations from -] to () January 1932, to be considered with the December data
bearing on hake movements in the intermediate region
Station
Distance
offshore
sea miles
Hours
trawling
(no. of
hauls)
Hake nos.
Hake
per
hour
Mean lengths
Sex
ratio
Males
Females
Total
Males
Females
WS810
WS809A + B
WS808
WS807
20
52
77
104
1(1)
5(2)
1(1)
1(1)
0
7
0
0
2
69
3
0
2
76
3
0
2
15
3
0
30-7
37-0
51-4
58-0
0
9-2
0
Some observations made in April during the first survey are also in very good agreement with the
theory Catches were more uniform than they were during the peak of shoreward movement (m
March only northern region data available), and the distribution in relation to distance from the land
was bi-modal The hake were most abundant at two distinct points over the range observed, at 94
and 238 miles from the land, the former being the richer haul and composed of significantly smaller
fishes A distribution of this type is just what one would expect if there had been a double wave of
shoreward movement, and if in April (autumn) the fish were again moving offshore towards their
winter quarters. We have already seen that the larger fish get farther out than the small ones, except
at midsummer, and they must therefore lead in the offshore movement, just as they seem to catch up
and pass the smaller fishes during the season of shoreward movement. The April data are shown in
detail in Table 25.
April data bearing on hake movements in the intermediate region, 17 to 25 April 1927
Table 25.
Station
WS96
WS95
WS108
WS97
WS98
WS99
Distance
offshore
sea miles
34
68
94
1 65
238
285
Depth
m.
96
104
119
146
172
237
Hake nos.
Males
14
10
62
14
II
o
Females
16
29
66
20
60
19
Total
per hour
30
39
128
34
71
19
Mean lengths
Males
30-0
39-4
35-6
38-1
39-4
Females
33-7
43-4
43-6
57-8
53-6
63-3
Sex
ratio
% males
46-7
25-6
48-4
41-2
15-5
0-0
Another point brought ou, by this table is the relatively greater abundance of ma es than at any
other season except March, and the.r tendency not to go so far offshore as the larger fcn>ales^
tL dXenees in mean length for either sex at the two peak stations are sigmficant by the usual
rf, Jtsf wh eh has been appM throughout this section wherever differences » nn,ean length have
b'een consTdeL specifically I have not tabulated a or 0I.IN io. all the nreans, because many of
them did not individually assist in the building up of the general picture.
3o6
DISCOVERY REPORTS
Finally, during the second survey, a series of observations in June (midwinter) yielded two rich
hauls at the greatest distance offshore, while farther in there were few hake, and only a singleton
within loo miles of the land. These results also showed that, as at all times except midsummer, the
mean lengths of the more adequate samples increased with their distance from the coast, and at the
same time the proportion of males diminished.
Table 26. June data bearing on hake movements in the intermediate region, i to 8 June 1928
Station
Distance
offshore
sea miles
Depth
m.
Hake nos.
Mean lengths
Sex
ratio
% males
Males
Females
Total
per hour
Males
Females
WS222
WS223
WS220
WS219
WS216
WS2I7
29
95
126
134
197
221
103
114
106
115
176
146
0
0
3
16
87
118
I
0
3
26
181
294
I
0
6
42
268
412
44-7*
36-8
40-1
39-4
59-0*
43-0*
407
47-5
46-5
0
50
38
32
29
* Too few for means to have any significance.
CONCLUSIONS ON MIGRATION
The relative abundance of the hake caught in relation to distance from the coast is the best means of
studying the probable seasonal movements of M. hubbsi, from our unavoidably limited data. For
comparative purposes the catches per hour for each series of observations may be summed, and the
catch at each distance category expressed as a percentage of the figure so obtained. A diagrammatic
summary of the seasonal observations, obtained in this way, is shown in Fig. 25.
We have seen also that comparisons of mean lengths and sex ratios corroborate the general picture
so obtained, wherever the data are adequate.
M. hubbsi seems to migrate towards the coast in summer, and offshore in winter, in much the same
way as do better known species of hake elsewhere. It would seem that as in M. merlucciiis, the larger
females move inshore first, passing their smaller sisters who may, however, begin to shoal somewhat
earlier. There is a strong suggestion that the smaller fish are rarely abundant at the greater distances
from the coast ; probably they do not migrate so far or so fast as the bigger ones. This may be a function
of size, and not only due to the greater proportion of immature fish among the smaller individuals, for
the proportion of males always diminished as one proceeded seawards, and in this species, where' the
disparity in size between the sexes is much more marked than in M. merhiccius, it is certain that many
of the males of even the smallest length class are mature. Probably the movements of males show
some marked differences from those of the females: the proportion of males was noticeably high in
March, at the time of greatest shoreward concentration of the smaller females, still fairly high when
seaward movement had begun in April, and low at all other seasons. This suggests that, except at
the height of the breeding season and for a short period afterwards, the males are more widely dis-
persed, and less inclined to shoal, than are the females.
Although M. hubbsi seems to move shorewards in a double wave rather like M. merhiccius, this
movement begms later m the year, and in this respect perhaps resembles more closely the movement
of M. bibneans. It is thought that this is probably connected with the general lateness of the biological
seasons in this part of the southern hemisphere, which has been described in the introduction to this
report. In M. bihnearis the later timing of the cycle may well be a direct effect of temperature, for off
New England the annual range is great. It is unlikely to be so off Patagonia, where the annual cycle
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 3°7
of temperature, though on a rather low level, shows small annual range. Thus the temperature con-
ditions in the habitat of M. hiibbsi resemble more closely those found on the European seaboard of
the North Atlantic (see also the Hydrological Notes in the Introduction).
>
o
INTERMEDIATE REGION
50 IQO 150 200 250 50 100 150 200 250
MILE5 FROM THE COAST
Fie 2. Diagram showing seasonal variation m relative abundance of Merluccius hMsi at
^' ^ different distances from the coast.
3o8 DISCOVERY REPORTS
THE FOOD AND FEEDING OF MERLUCCIUS HUBBSI
Owing to the wide scope of the survey programme it was not possible to carry out such detailed
observations on this subject as have been found desirable by European specialists working on single
species of fish (e.g. Hardy (1924) on herring, Hickling (1927) on hake). On our first survey 191
stomachs were examined and on the second 186. On the third survey forty-six notes on stomach
contents were made in the field, but owing to pressure of other work the numbers of fish examined in
this way were not always recorded. It was evident, however, that the proportion of fish containing food
was least in summer, which accords with the known habits of European hake (Hickling, 1927, p. 49).
A general loss of appetite just prior to spawning is known among many diverse species of fishes. ^
Our first survey (autumn) and second survey (winter) results show that a slightly higher proportion
of Patagonian hake were found to contain food in winter, but I do not believe this indicates more
extensive feeding during that season. From notes on the size and numbers of food organisms in
individual stomachs it is evident that feeding was heaviest in autumn. The apparent anomaly is due
partly to a seasonal change in diet, partly, no doubt, to the limitations of the data.
The number of times that food of recognizable categories was recorded, and the percentage occur-
rence of each category, during each survey are set out in Table 27. From this it is at once apparent that
the feeding habits of M. hubbsi art essentially similar to those of M. merluccius, and perhaps even closer
to those of M. bilinearis (Bigelow and Welsh, 1925, pp. 389-90). It feeds chiefly upon other fishes,
mcluding even its own species, squids and more or less planktonic Crustacea. This last constituent
was chiefly found in the stomachs of the smaller and younger fish.
The list of fishes eaten includes the commonest species of the area. Falkland herring {Cliipea
fiiegensis) was by far the most important forage species, especially in winter, but other species are
preyed upon with equal voracity when readily available. Thus as many as sixty-seven individuals of
the small scald-fish Thysanopsetta naresi have been taken from the stomach of a hake less than 56 cm.
long. Merluccius hiibbsi, like other species of hake, are known to devour individuals of their own kind
more than half their own length. Had we been able to obtain more food records during the offshore
phase of the seasonal migration there is little doubt that such instances would have been commoner.
There is one peculiar difference from the feeding habits of European hake that makes the paucity
of deep-water observations during winter the more regrettable.^ Although the distribution of M.
hubbsi overlaps that of Micromezistiiis australis to the southward, the latter was not recorded as a
constituent of the hake food. This is extraordinary because M. australis is very closely related to our
own blue whiting M.poutassou (Norman, 1937, p. 51), which Hickhng (1927, p. 42) had shown to be
such an important constituent of the food of European hake. It seems that M. australis keeps more
exclusively to deeper water and higher latitudes than its European counterpart, and therefore its
habitat— during the warmer months of the year— does not coincide with that of the hake to anything
like the same extent. In winter the southern blue whiting may move northwards as well as offshore,
as will be shown in the section dealing with that species. Possibly the hake then feed upon them as we
should have expected ; but the fact remains that there was no evidence of this at the few stations where
we did locate the two species together.
It will be noted that the list of food organisms includes several bottom-living fishes and a little
benthos. The slightly more benthic tendency in choice of food is doubtless occasioned by the uniformly "
hv^H^t^'^T t° this phenomenon in Quinat salmon (Jordan), silver eels (Petersen) and pleuronectids (Todd), are quoted
by H.ckhng (1927), and he was further able 'to suggest that the blue whiting {Micromezistius poutassou) also feeds much le7s
? olmrilerv.Thr'^?" -h'™'" behaviour among Labrador cod ^o^uld seem to belplied ly Harold TZmpson
seasoA^' ^ ^ ^^ "P°" ''P'" ^°' '^^ P'™'^ °^ S'"'' ^""'^'"g ^"'^ recuperation which succeeds the spawning
season .
2
I must reiterate that this was due to the precipitous slope preventing trawling, and not to any lack of endeavour.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
309
moderate depths of the Patagonian shelf, which gives a much sUghter depth gradient than the habitat
of the better known species, until the edge is reached. The few echinoderms, etc., were, moreover,
recorded in autumn, when the fish feed ravenously upon whatever comes their way.
Table 27. Feeding of Merluccius hubbsi
Food categories
Times recorded and percentage occurrence (italics)
Third survey
(summer)
First survey
(autumn)
Second survey
(winter)
Clupea fuegensis
Merluccius hubbsi
Macruronus magellanicus
Salilota australis
Notothenia ramsayi
Notothenia spp.
Stromateus maculatus
Thysanopsetta naresi
Unidentified fish
Post-larval fish
12 26-1
2 4-3
1 2-2
7 15-2
2 4-3
I 2-2
19 i6-5
1 o-g
3 2-6
2 1-7
2 1-7
7 6-1
8 7-0
I o-g
34 28-1
I 0-8
6 5-0
5 V-!
9 7-<^
Total fish
25 54-2
43 37-4
55 45-4
Large squid
Squid
10 21-y
7 6-1
14 12-2
4 JJ
Total squid
10 21-J
21 i8-3
4 3-3
Decapoda
Munida subrugosa
M. gregaria
Munida unidentified
Euphausians
Parathemisto gaudichaudii
Hyperid amphipods*
Amphipods
Serolis sp.
6 13-0
I 2-2
3 6-5
I 2-2
7 6-1
13 11-3
1 o-g
18 15-7
4 3-5
2 1-7
2 i7
56 46-j
2 i7
Total Crustacea
II 23-g
45 39-2
60 .^97
Holothurians
Asteroids
Ophiuroids
Sponge fragments
— —
I o-g
1 o-g
2 1-7
2 1-7
2 1-7
Total echinoderms, etc.
— —
6 5-2
2 i7
Total separate records
46 gg-8
115 100-1
121 lOO-i
* Doubtless mainly Parathemisto.
Hints of some interesting seasonal changes in dietary may be gathered from Table 27, although the
data are not quantitative. Squids were obviously an important food in summer and especially in
autumn, but were rarely recorded in winter-caught hake. Of the crustacean constituents, chiefly
important to the smaller hake, Munida were eaten largely in summer, when there were no records of
euphausians in the stomachs. In autumn there was still some Munida, considerable quantities of
euphausians and many hyperid amphipods. (There is little doubt that nearly all of these were Para-
themisto gaudichaudii, a very common species over most of the southern ocean, but I give the categories
as stated in the original records.) In winter euphausians were the most frequently recorded food, and
few other Crustacea were present. It is probably significant that the widest variety of food was re-
corded in autumn, when the hake were feeding intensively after spawning.
The relative importance of the different constituents of the food of Merluccius hubbsi cannot be
3IO
DISCOVERY REPORTS
accurately gauged without weight or volume records and altogether more extensive data, but a system
of arbitrary' weighting (kept well on the ' safe side ') permits a diagrammatic presentation of seasonal
changes in the major (lumped) categories, that gives a useful general picture. It may be regarded as a
cautious understatement of the predominance of fishes and cephalopods in the diet, and of the
obviously great importance of Chipea, especially in winter. The 'weighting' employed, after due con-
sideration of known weights of most of the food organisms was : regarding Crustacea and benthos as
unity, cephalopod records were multiplied by four and fish records by five. The results are shown in
Fig. 26. Doubtless this picture would be altered by more detailed results, especially if the size of the
fish could be taken into account, but such work will only be possible when a naturalist can devote his
whole time to the one problem.
A B C
Fig. 26. Diagrams showing crude relative proportions (arbitrarily weighted) of the main food categories of Merlucdus huhbsi
at different seasons. Weighting: fish x 5, squids x 4, Crustacea x i, and echinoderms, etc., x i. Echinoderms, etc., which are
rarely eaten are left white in the diagrams. A, third survey, summer. B, first survey, autumn. C, second survey, winter.
PARASITES
Like most other sea fishes M. hubbsi were observed to be very commonly infested with nematode,
cestode and trematode worms. On the first survey these three classes of parasites were observed to be
present in (roughly) the order of frequency in which they are named above, and it was noted that they
seemed particularly abundant in the larger (older) fish. Almost all the specimens examined had
nematodes in some part of the digestive tract or in the body cavity. Copepodan parasites were
evidently less frequent but not uncommon. Chondracanthidae were more than once recorded as
numerous in the mouth, and Miss N. G. Sproston informs me that members of this family frequently
infest European hake also.
There aje numerous references to Lernea and Lerneidae in the log-books, which introduce an un-
fortunate element of doubt into some carefully collected statistics of the incidence of this form of
parasitism during the third survey, when the hake were sorted into length classes for weighing. I
believe that these records all refer to a lernaeocerid either identical with our own Lernaeocera branchialis
or very closely allied to it, but earlier references to ' Lernea ' on external situations (L. branchialis is
strictly a blood-vascular parasite, and has been recorded from European hake) leaves some element of
doubt. The situation is clouded by the unfortunate change of status of the genus Lernea so justly
deplored by Gurney (1933, p. 336). To be quite safe, these parasites may all be referred to the family
Lernaeoceridae as proposed by him.
The bulk of the figures were obtained in December, in the northern and intermediate regions, that
is, in the most favourable part of our area for hake, at a time when the seasonal shoreward migration
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 3ii
had begun. There is thus a good chance that the figures may be fairly representative, though the
chances of infection by these parasites might vary with the seasons. A few data obtained farther
south, at other times, do not suggest this, and in the main confirm the picture given by the good
December figures. The latter only are considered here. The results are given in summarized form m
Table 28, and the increase in percentage of female fish infected with increasing length is shown
graphically in Fig. 27.
Table 28. Incidence of infection by Lernaeoceridoe, Merluccius hubbsi,
north of 50° S, December 1931
Length
classes
cm.
No.
females
examined
No.
females
infected
Females
0/
/o
infection
No.
males
examined
No.
males
infected
Males
/o
infection
21-30
31-40
41-50
51-60
61-70
71-80
81-90
63
107
216
334
164
37
I
I
4
13
24
15
8
0
1-59
374
6-02
7-19
9-15
21-62
64
70
29
I
4
0
I
0
6-25
0
3-44
S922
26s
M7-05
5:164
S5
M3-05
It will be seen that female Merluccius hubbsi showed a steady increase of ' percentage mfection' with
these parasites as they grew longer (and older), and marked increase in the highest length group for
which adequate figures are available. Females appear to be more than twice as often mfected as are
males.
Zl-30| 31-40 I 4-1 -50 I 51-60 | 61 "70 1 71-80
LENGTH CLASSES _ CMS
Fig. 27. Increase in infection by Lernaeoceridae with increase in length (and age) of female
Merluccius hubbsi.
312 DISCOVERY REPORTS
The males do not show any correlation between increase in length and increase in liability to
infection. The smallest length class showed the highest percentage infection, which was also con-
siderably higher than that of females of the same length. I think this discrepancy is in some way
bound up with the marked differences between the sexes that became apparent in the study of the
general bionomics of this fish. Apart from the size difference, and almost certainly slower later growth
rate, male M. hiibbsi reach maturity at a much smaller length than females, they seem less given to
shoaling during the ' off' season, and do not migrate so far as the older females. All these factors may
influence their chances of infection by Lernaeoceridae.
Macruronus magellanicus Lonnberg
This is a long, slender fish, with close superficial resemblances to the Macruridae, with which it
was formerly classified. Norman (1937, p. 49) has shown, on osteological grounds, that it should be
placed in the family Merlucciidae. The tail, tapering to a point and without a separate caudal fin, is
the most noticeable point of similarity to the macrurids, but Macruronus lacks their projecting snout,
and its distribution and habits are markedly different. Macruronus is found in numbers only on the
shelf, in relatively shallow water during at least nine months of the year, whereas the macrurids are
essentially a deep-water group inhabiting the slope beyond the shelf edge and even greater depths.
The coloration of Macruronus reflects this difference in habitat. In Plate XVI a water-colour sketch of
a living specimen taken at St. WS99, by E. R. Gunther, is reproduced. At an earlier station he had
described its coloration thus : ' laterally a pale lustrous blue, becoming more intense dorsally into
tones of sapphire and turquoise, ventrally losing colour becoming silvery white.' A colour pattern
such as this is normally characteristic of mid-water fishes inhabiting moderate depths over a well-
illuminated sandy bottom, and this would apply fairly to that part of the plain of the shelf where we
found the smaller individuals most abundant.
It IS interesting to note that the correct taxonomic position of Macruronus is reflected in the local
Spanish-American name ' Merluza de cola ' which Norman (loc. cit.) states is applied to it. This might
be freely translated into 'long-tailed hake' with advantage, for English-speaking fishermen tend to
apply the names ' rat-fish ' or ' rat-tail ' to anything remotely resembling a macrurid, consequently
confusmg them with fishes as gsnetically remote as chimaerids in some parts of the world.
M. magellanicus is the second most important fish of the Patagonian Continental Shelf. In our
catches it was outnumbered only by Notothenia ramsayi, but although slightly more numerous than
hake in the aggregate it was less widely distributed, a few exceptionally rich hauls augmenting the
total unduly. A much more slender fish than the hake, it is about half as heavy at a given length, and
a larger proportion of the smaller individuals escaped through our normal cod-end mesh. There are
numerous references in our rough logs to 'Macruronus seen escaping'. In the eighty-six summer
hauls of the third survey, for which roughly comparable weight data are available, Macruronus yielded
29-5% by weight of the total fish taken (rubbish excluded) as against the 47-3% of hake. The relation
to other less important categories can be seen from the tables in the concluding section of this report.
Macruronus was the most important species in the southern region, where, as we have already seen,
the hake diminished greatly in numbers. Of the weight of fish caught here, Macruronus provided
4^2 % •
M. magellanicus is most excellent eating. In the third (unpublished) scientific report' on the work
of the ' William Scoresby ' Dr Mackintosh wrote : ' It is generally agreed that Macruronus is superior
to any of the other common fishes. The flesh is reasonably firm and free from too many small bones.'
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
Our records of the total catch of this species are as follows :
313
WS77
WS79
WS91
WS92
WS94
WS95
WS99
WS108
WS216
WS218
WS762A
WS762B
WS763
WS764B
WS765
WS771
WS774
WS775
WS776
WS786
WS788
WS789
12. 111. 27
13. iii. 27
7. iv. 27
8. iv. 27
16. iv. 27
17. iv. 27
19. iv. 27
25. iv. 27
1. vi. 28
2. vi. 28
16. X. 31
16. X. 31
16. X. 31
17. X. 31
17. X. 31
29.x. 31
i.xi. 31
2. xi. 31
3.xi. 31
7. xii. 31
I3.xii. 31
13. xii. 31
I
4
3
24
49
2
4
32
I
3
I
123
6
4
25
12
368
43
162
34
I
I
WS790A
WS790B
WS791B
WS792A
WS792B
WS793
WS796A
WS796B
WS797B
WS797C
WS799A
WS799B
WS800B
WS805
WS806
WS807
WS810
WS811I
WS811II
WS812I
W 881211
WS813
14. xii. 31
14. xii. 31
14. xii. 31
15. xii. 31
15. xii. 31
15/16. xii. 31
19. xii. 31
21. xii. 31
20. xii. 31
20. xii. 31
21. xii. 31
21. xii. 31
22. xii. 31
6. i. 32
7- •• 32
7- i- 32
9. i. 32
10. i. 32
12. i. 32
10. i. 32
12. i. 32
13. i. 32
107
23
30
9
932
5
56
3
26
37
37
12
I
I
I
12
3
162
75
33
5
80
WS814
WS815
WS816
WS817A
WS817B
WS818A
WS818B
WS838
WS848
WS8S3
WS8s5
WS857
WS858
WS859A
WS859B
WS864
WS866
WS868
WS870
WS874
WS875
13.1.32
13.1.32
14. 1. 32
14. i. 32
14. i. 32
17. 1. 32
17.1.32
5. ii. 32
10. ii. 32
21. iii. 32
22. iii. 32
23. iii. 32
24. in. 32
25. iii. 32
25. iii. 32
28. iii. 32
29. iii. 32
30. iii. 32
31. iii. 32
3- iv. 32
3. iv. 32
981
2
56
5
2
2
I
33
3
26
I
48
5
12
II
I
1180
98
227
II
78
Two main features of the distribution already mentioned become quite clear from these figures:
the greater relative abundance of the species in the southern region, and the tendency to form dense
local shoals, so that a small minority of the catches are vastly bigger than the others This latter feature
is the probable reason for lack of a clear north to south gradient in abundance of Macruronus m our
records A single extra large catch (of small immature individuals) was made in the northern region,
but no corresponding shoal happened to be encountered while we were sampling the intermediate
region. It is clearly necessary to consider other lines of evidence, bearing on the probable spawning
time and movements of the fish, that may help to explain the observed distribution. The most fruitfu
studies possible from existing data appeared to be considerations of seasonal variations in ponderal
index regional variation in mean length, and relation of relative abundance and mean length with
depth • but before we pass on to these one most important feature, quite clear from the catch records,
must be emphasized: Macruronus was present at only two stations worked in winter, and these were
among the most northerly of thirty more or less comparable hauls worked along the shelf edge, and
over the shelf, in the main summer haunts of the species. This strongly suggests that the fish move
north in winter. The repeated attempts made by Mr John to trawl in deep water over the shelf edge
during the winter survey sufficed to show the offshore movement of hake at that season and should
also have revealed the presence of Macrr^ronus offshore if there were not also a considerable mendiona
component in the direction of movement of the latter species. We have no proof, of course, tha
Macruronus does not move eastwards offshore-it probably does, though not to the same ex^nt as th
hake-but it seems certain that it moves north as well, whereas any meridional component in the
direction of movement of the hake would seem to be too small to be demonstrable from existing
"^Tn Table 29 the data have been grouped at mean dates, from stations selected according to their
tine distribution, so as to show the seasonal variation in 'average' ponderal index of Macruronus in
each o the three regions. The chief stress was laid upon inclusion of stations within a narrow interval
ot: about the m'ean date and therefore regional differences in the other futures shown-reative
abundance mean lengths and sex ratios from the same groups of data-are not fully illustrated by this
array The' Tata suffice to indicate three main points, however: greater abundance and size of in-
d "duals in the south, and the constancy of the sex ratios, showing a slight preponderance of females
314
DISCOVERY REPORTS
in nearly all groupings. The seasonal variation in ponderal index shows a rise in all three regions from
spring and summer to autumn. These results are also shown graphically in Fig. 28.
Table 29. Data selected over short-time intervals in each region, to show the increase in ponderal index of
Macruronus magellanicus during the season, with corresponding figures showing relative abundance,
mean lengths and sex ratios, and the mean sex ratio for each region
i
Mean date
Total
Macruronus
Hours
positive
hauls
Fish per
hour's
+ haul
Sex
ratio
i, males
Mean
length
cm.
Average
K
Sex ratio
mean for
region
23. X. 31
15. xii. 31
23. iii. 32
37
1089
45
Northern region
2
IS
3
18
73
15
41-7
40-0
6o-o
417
37-6
33-8
0-290
0-312
°-354
40-8%
2. XI. 31
20. xii. 31
23. iii. 32
570
153
48
3
15
I
Intermediate region
39-9
190
10
48
49-0
43-8
38-9
37-6
40-6
0-270
0-345
0-358
41-9%
<S3
16. i. 32
8. ii. 32
2. iv. 32
1321
36
1212
24
2
5
Southern region
45-8
44-4
55
18
242
38-0
47-1
55-9
49-0
0-311
0-322
0-343
42-1%
36
The figures for 'average' K are means of means. The fish were weighed in length groups, their
individual lengths being known, but of course the numbers included in the weighings differed widely.
Little difference between ponderal index of males and of females could be detected. The averaging of
the ii: values obtained within the time limits stated is an attempt to obtain a general indication of the
direction or trend of the seasonal variation, although the figures have no precise significance. The
fact that they indicate a rise in ponderal index in each of the three latitudinal regions over the period
studied seems, however, strongly significant in the general sense, though not demonstrable mathe-
matically. The implication is that Macruronus spawns in spring.
0-350-
2
^ 0-300 ■
5
OCTOBER NOVEMBErIdECEMBERI JANUARY IfEBRUARY I MARCH I APRIL
Fig. 28. Seasonal variation in mean ponderal index of Macruronus magellanicus: • northern
region, ■ southern region, A intermediate region.
We have seen that in summer the fish is most numerous in the southern region, from which it was
absent m winter. Thus the direction of movement in spring and summer must include a considerable
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
315
LENGTHS -CMS
10-
5-
.jj
Ul
W5B70
52"49'5
Fig. 29. Percentage length frequencies of Macruronus magellanicus at stations worked in long. 64° 15' W, between
21 and 31. iii. 32, showing increase in length with increasing latitude.
3i6 DISCOVERY REPORTS
southerly component. Further, it is highly probable that Macruromis spawns in spring, for the
ponderal index rises during the summer. Many fishes feed most heavily just after spawning and it
would seem that the movements of Macruromis during summer are essentially a feeding migration.
The search for food may be expected to lead to considerable dispersion, and to local concentration
where food is plentiful. This would help to explain the greater variation in size of Macruromis catches
when compared with the catches of hake, whose summer movements are more in the nature of a
breeding migration. At two of the large southern hauls of Macruromis the fish were observed to be
glutted with clupeoids — probably the most important food of the larger individuals — and I think it
probable that the southern concentrations oi Macruronus and Clupea fuegensis will be found to coincide
in summer.
Small immature individuals of Macruromis were found mainly to the north of our area, in com-
paratively shallow water. The southward movement in summer seems mainly confined to the larger
mature individuals. Although a clear gradient in relative abundance and increasing size of individuals
cannot be shown in lumped data as we proceed southwards from the northern region through the
intermediate region, the greater abundance and size of individuals of the southern region population
is clear however the data are arrayed. Over a short-time interval in March, the length frequencies
clearly show the increase of size with increase of latitude (Fig. 29), and there is little doubt that this
would have been apparent at other times had it been practicable to work more north and south lines.
The distribution of Macruronus in relation to depth has been found to have an important bearing
upon the problems of its movements. The depth relations from all data, regardless of region, are
shown in Table 30, where the catches have been grouped in 50 m. depth classes. From this it is at
once apparent that it is very much an inhabitant of the plain of the shelf, showing a marked falling off
in relative abundance at depths greater than 150 m. It was never taken below 300 m. and, as can be
seen by comparing the individual depths given in the Appendices with the catch records already listed,
nearly all the deeper records were obtained in autumn. This may imply some offshore movement
superimposed upon the meridional movement of which evidence has already been given.
Table 30. Data summarizing the depth relations of Macruronus magellanicus
Depth
grouping
m.
No. of
hauls
Hours
trawling
Total
Macru-
ronus
Hauls
present
/o
occur-
rence
Fish
per
hour
Sex
ratio
/o
males
No.
measured
{ = N)
Mean
length
cm.
a
mean
length
'y^N
1-50
3
2I
0
—
—
—
—
—
—
—
51-100
33
48
606
15
45-S
13
447
598
40-1
15-2759
0-3909
101-150
80
98
4271
38
47-5
43
41-3
3907
42-8
8-4497
0-0183
151-200
24
24
140
5
20-8
6
44-9
139
53-5
11-8251
I -0060
201-250
IS
15
84
3
20-0
6
363
84
53-7
10-9372
I -424 1
Over 25 1
14
21
232
4
28-6
II
33-9
232
53-3
8-8461
0-3373
The depth distribution is particularly interesting in relation to the mean lengths. In each of the
three deeper depth categories in which Macruronus was found these slightly exceeded 53 cm., and no
significance attaches to the slight differences between them. But in slightly shallower water, where
the vast majority of our specimens were secured, the fish were much smaller, of mean length 42-8 cm.,
the difference from the means for all three deeper groupings being strongly significant. In the next
shallower depth category there was a marked falling off in relative abundance and the fish were even
smaller. Their mean length was 40-1 cm., and the further small decrease of 2-7 cm. in mean length is
statistically significant. These data show that larger, mainly mature, fish preponderate in the smaller
catches of Macruronus from the greater depths in which the species was found, while smaller mainly
immature fish predominate in shallower waters. We have already seen that the smaller fish tend to be
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 317
commoner to the north, so that it would seem that we have here another instance of the influence of
size of fish upon movement. The larger fishes seem to range farther afield than the smaller ones. This
is in good agreement with our unavoidably scanty observations upon the feeding habits of Macruronus,
presently to be described.
Numerous computations of mean lengths of Macruronus have been made besides those tabulated
here. From these it appears that the difi^erences in size between the sexes is small. Females are usually
larger by 1-2-7 ^^"^- ^^ mean length than males, but the diff^erence can only be shown to be significant
in large samples. The fairly constant sex ratios, with slightly fewer males in the deeper catches, and
the general slight preponderance of females in the trawl, suggest {a) that the slightly smaller males
may tend to migrate a little less than the females, (b) that the sex ratio is probably nearly normal near
the main locus of the species, the discrepancy being due to a higher escape ratio of the smaller males.
We have already seen how the much greater difference in size between the sexes seems to aflFect the
distribution of the hake population.
Detailed notes upon the food of M. magellaniciis were made at eight stations only. At four of these
the fishes were small, the mean lengths being 34-4, 34-9, 35-0 and 44-5 cm. These fish were almost
entirely carcinophagous and had been feeding heavily upon euphausians and Parathemisto; one was
crammed with fish larvae, and a single specimen had taken a Munida. The other four records are from
stations where the fish were considerably larger, of mean lengths 50-4, 51-8, 52-5 and 55-6 cm. These
had been feeding exclusively and heavily upon Clupea fuegensis and Notothenia spp. (most being
undoubtedly A^. ramsayi). Squid have not yet been observed in the stomachs of Macruronus. It is
therefore probable that Macrtirofius exhibits, perhaps to a marked degree, that change over from a
carcinophagous to a fish diet that seems characteristic of many demersal fishes as they grow older (e.g.
hake). The point should not be stressed unduly in the absence of data as to possible seasonal changes
in diet.
The facts that only the larger fishes were found to be ichthyophagous in summer, and that they
appear to migrate much more extensively than the smaller fishes, seem to find a close parallel in some
of Harold Thompson's observations on Newfoundland cod (1943, p. 86). He says: 'It is seen that
the volume [of food] increases rapidly with the size of cod, and it is probably this need for large
quantities of food (unobtainable all the year round in any one locality) which leads to cod making
increasingly great migrations as it becomes older; whereas small cod apparently find it possible,
within a limited inshore area for example, to find sustenance in the smaller crustacean species for the
first few years of their lives.'
There is further evidence that the age at which M. magellaniciis attain a length of some 50 cm.
marks a critical period in their lives, in the relation between length and ponderal index. Among
females from the southern region values for K were highest among fishes of the 41-50 cm. length
class (weighted mean 7^=0-352). Fish of the 51-60 cm. class gave values some 4% lower (weighted
mean 7^=0-340) and two higher length classes gave lower values. If we had sufficient information to
plot a curve for this relationship it seems fairly certain that the point of inflexion would lie at around
50 cm. The fall was slight because the reliable data were collected in March, and this is too long after
the spawning period to give the best demonstration of this phenomenon. Earlier data are roughly in
agreement, but inadequate to show the efi^ect within convincingly narrow limits of time and space.
The implication is that it is at a length of about 50 cm. that sexual maturity is first attained.
Thus it seems possible that sexual maturity, increased migratory movement (in search of food)
and a change in the nature of the diet, follow closely upon one another in that year when the fish reach
a length of some 50 cm.^
1 Possibly their sixth year of life.
3i8
DISCOVERY REPORTS
LENGTHS_CMS
A
B
Fig. 30. Actual length frequencies of Macruronus magellaniais at selected stations, including probable year-classes at about
18, 27 and 35 cm. A, St. \VS762 A + B, 16. x.^i; B, St. WS776, 3. xi. 31, and C, St. WS790 A, 14. xii. 31.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
319
Several series of length frequencies of Macruronus showed strong and moderately consistent in-
dications of year classes, but the wide dispersion of our data in time as well as in space preclude any
useful pooling of these results. As Petersen found long ago it was the younger fishes that tended to
show this feature best. A few unmanipulated frequencies from stations worked early in the season,
in the northern region, are shown in Fig. 30. These indicate possible year classes at 18-19 cm- (prob-
ably just two years old) at 27 cm. and at 35 cm. (.'' three and four years old respectively) while the
third confirms the probable existence of a year class 35 cm. or so in length at that season. Later
0-700
0 650-
^ 0 600-
z
U 0-350-
MERLUCCIUS
0CT05ER NOVEMBER I DECEMBER JANUARY FEBRUARY MARCH
MACRUKONUS
0-3QO-
0 250
OCTOBER InovEMBER I DECEMBER I JANUARY [FEBRUARY | MARCH
Fig. 31. Seasonal variation in ponderal index of Merluccius, a summer spavvner, and of
Macruronus, which would seem to spawn in spring.
results suggest that growth is most rapid in summer, and that the yearly increments diminish so that
fish of over 70 cm. may only grow some 5 cm. in the course of a year, but the data are insufficient to be
conclusive.
Norman (1937, p. 50) has already pointed out the extremely close relationship between M. magel-
lanicus (Lonnberg) and M. novae-zelaiidiae (Hector) from Tasmania and New Zealand, with which
it was at first identified. This afiinity extends beyond morphological features to coloration and
habits. From Waite (1911, pp. 180-1) we learn that M. novae-zelandiae attains much the same sizes
as the Patagonian species, that its colour may be described as a deep iridescent purple with the fins
smokey and the lower parts silvery, and that it has been observed to feed upon Clupea neopilchardiis.
During the New Zealand trawling experiments of which Waite was writing several considerable
catches of the species were made, all between 16 and 28 fm., and it was not observed in deeper water.
320
DISCOVERY REPORTS
Reference was made to Dr Albert Gunther's opinion (1887, p. 157) that 'it is not probable that it
descends to the same great depths as the other Macriin, and to several strandings of the species in
Cook Straits, which have also been referred to by other New Zealand writers. It is interesting to find
that the essentially shoal-water habitat of the genus was recognized so early, in the days when it was
still classed with the macrurids.
At the time of which Waite was writing it was thought that M. novae-zelandiae would be unmarket-
able in New Zealand, and I have not found any recent reference to its being utilized. It is evident,
however, that New Zealanders suffer less from traditional inhibitions as to what constitutes a good
food fish than some older communities, for we learn from Phillipps (1921) that Coelorhynchus australis,
occasionally trawled in deep water in Golden Bay, is highly esteemed under the name of 'javelin-fish ',
and that large Callorhynchus sell well in Christchurch as 'silver trumpeter'. The nearest equivalents
of both these species would certainly be discarded as rubbish by British trawlers. Since we found the
Patagonian species of Macruronus most excellent eating I therefore venture to suggest that an attempt
to market the New Zealand one might prove well worth while.
By way of summarizing our findings as to the bionomics of M. magellanicus, it seems both interesting
and profitable to compare and contrast it with its near ally, the Patagonian hake :
Comparison and contrast of the main features in the bionomics of Merluccius
hubbsi and Macruronus magellanicus
Merluccius hubbsi
Macruronus magellanicus
Spawning season
Extreme range of K and seasonal
variation
Size difference between sexes
Depth relations
Size/latitude relation
Migration
Growth (speculative)
Potentiality as human food
Summer
From about 0-540 to 0-830. Falls during
summer (see Fig. 31)
Males much smaller than females (differ-
ence 13-5 cm. between grand mean
lengths)
Vary with migration, but larger fish are
found in deeper water except at the
height of summer. Extended down to
the greatest depths fished
Mean lengths increased with latitude
significantly except at height of summer
Strong inshore movement in summer,
little north to south movement if any,
offshore movement from autumn to
winter
Perhaps similar to that of European hake,
with length increments smaller in
parallel with the smaller size of this
species
Flesh rather soft but not inferior to that
of European hake of similar sizes
Almost certainly spring
From about 0-250 to 0-390. Rises during
summer (see Fig. 31)
Males only 1-2 cm. smaller than females,
but the difference is nearly always signi-
ficant in adequate samples
The larger fish in deeper water but the
species is mainly confined to the plain
of the shelf. Not found in hauls ex-
ceeding 300 m. mean depth (we made
seven such hauls, at four of which hake
were present)
Mean lengths increased with latitude
significantly whenever comparable data
available
Considerable southward movement spring
and summer. Some offshore move-
ment in autumn? Northward movement
in winter
Earlier length increments probably
slightly greater than those of European
hake, later ones diminishing sooner,
more regularly, and markedly
Flesh firmer and better flavoured, but
less readily available as the fish are so
much more slender
GADIDAE
Micromesistiiis australis Norman was discovered during the first of the surveys described here.
Norman (1937, pp. 51-2) has pointed out its close relationship to M. poutassou of the Mediterranean
and north-eastern Atlantic. So far as I have been able to determine, it is the only ' typical ' gadid
with three dorsal fins known to occur in south temperate and subpolar waters. At St. WS 80, E. R.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
321
Gunther noted its colour as 'slatey blue, darker above, a purer blue laterally, white ventrally'. Its
general distribution in our catches (markedly southern and in deep water) is shown below:
IVS80
WS99
WS216
WS217
WS218
WS816
WS817A
154 m.
238 m.
176 m.
146 m.
279 m.
150 m.
196 m.
4
10
I
I
I
I
58
WS817B
WS818A
WS818B
WS819A
WS819B
WS820
WSS21
220 m.
275 m.
281 m.
320 m.
359 m-
464 m.
52
52
222
27
72
22
I
IVSS25
WS839
WS840
WS850
WS868
IVS870
IVS875
140 m.
418 m.
415 m.
161 m.
164 m.
272 m.
243 m.
2
I
5
12
7
I
5
We have already had occasion to note that Micromesistius australis seems to inhabit higher latitudes
(for the most part) than its European relative. We never found Merluccius hubbsi to have been feeding
upon it, although as a result of Hickling's work
LENGTHS-CMS. it has long been known that Micromesistius pou-
35 40 _ 45 50 ^^^(^ is one of the most important forage species
for the larger European hake. The two European
species overlap almost throughout their lati-
tudinal range, while off Patagonia Micromesistius
occurs mainly to the south of the range of all
but a very few of the hake.
It can be seen above that all our specimens
of M. australis— excepting one haul of 10 in
49° 42' S and three singletons farther north
in autumn and winter — were obtained in the
southern region south of 50° S. It is also very
clear that it is a deep-water species, very rarely
to be taken on the shelf in numbers. All but
one of our richer hauls were obtained in deep
water beyond the shelf edge. It was present in
all three of the deepest trawlings made, in depths
of over 400 m.
A study of the length frequencies showed
some remarkable features. All were very strongly
unimodal and of very narrow dispersion. The
pooled frequencies for the January stations,
worked fairly close together in the southern
region, over a brief period, are shown in Fig. 32.
These data yielded mean lengths of 41-5 cm.
((7=2-6257) for males and 43-2 cm. ((T= 3-2392)
for females. The difference of 17 cm. is clearly
significant. The slightly larger size of the females
(a well-known feature in many diverse species
of fish) might be expected to increase their
chances of capture with the result of a slight
'spurious' preponderance of females in the
catches The sex ratio observed however was strikingly ' abnormal ' in the unexpected direction-77 %
the J':^^^^^ males. This suggests the Po-biUty of a tendency tow.^^^^^^^^
ir. this species during the non-breeding period. (The ponderal indices show a steady rise from January
Fig. 32. Percentage length frequencies of Micromesistius
australis, January 1932.
322 DISCOVERY REPORTS
to April— suggesting spring or early summer spawning— but the data are inadequate for detailed
consideration.)
The very narrow dispersion of the observed length frequencies (Fig. 32) shows that a majority of
any modal class less than about 38 cm. long escaped through the normal cod-end mesh, as would be
expected with such a small slender species. This was seen to occur and was duly recorded in the log-
books. Above the very pronounced mode, however, such narrow dispersion (the biggest fish taken
were only 49 cm. long) can only mean that little growth takes place after the fish have'' attained such
modal length. The absence of all trace of a higher secondary mode shows that only a very small
mmonty of the 41-43 cm. fish normally survive another year. It is possible, therefore that
M. austrahs is a short-lived fish of comparatively rapid growth. Fine-meshed trawlings to sample
the smaller year classes would be essential to establish the point with certainty.
The food oi Micromesistius was noted at two stations. It consisted oi Parathemisto and Euphausians
In this respect, as in their deep-water habitat along the edge of the shelf, they resemble their European
relative (cf. Hicklmg, 1927, pp. 53 et seq.). Hickling's findings also show that the European species
spawns late in spring (in agreement with the season suggested as probable for the Patagonian species
from the seasonal shift in ponderal index) in the nearest corresponding latitudes. It is in the restricted
distribution in colder waters than those frequented by the local hake that the Patagonian species shows
the most striking difference from Micromesistius poiitassou.
Saltlota australis (Giinther). Norman (1937, pp. 52-3) has explained how this genus is barely
separable from the Physiadus of Kaup, with which our species was identified in the field He has also
a footnote concerning the circular, unsealed, pigmented area between the bases of the pelvic fins-
This IS associated with a luminous gland', and gives references to Hickling's work on the subject'
From the latest of these (Hickling, 1931) it would seem that similar organs had long been known
among a variety of Macruridae, and in Physiadus japonicus Hilgendorf alone among Gadidae. Hickling
concluded that the gland is essentially a larval organ which may remain functional throughout life
in some species (loc. cit. pp. 863^5). It functions strongly in the adults of Malacocephalus laevis
(Lowe) (Hicklmg, 1925), but though functional in the young of Coelorhynchus coelorhynchus (Risso)
It becomes vestigeal in the older fish. I have found a note of E. R. Gunther's stating that with Saldota
... m/« luminescence was not observed in the field, so that it is possible that the gland is vestigeal
or less developed from its primitive condition in the common ancestral Anacanthini) in this spedes
also. Gunthers first colour note (St. WS73) on the species reads: 'Evenly grey, slightly darker
towards the back. Ventral surface slightly violaceous, becoming almost black'at'one spot anterior to
he anus (this spot probably luminous). The scales suggest a faint brazen glitter.' It is therefore clear
that our observers knew of the possibility of lummescence m this species'from the first, but no uch
phenomenon was seen although numerous specimens both old and young were subsequ ntly secured
The gland may very probably be functional in the larvae, but it would seem to be undeveloped or
vestigeal in the older stages of S. australis. uiiueveiopea or
Our records of this species, set out below, show that its seasonal and regional distribution resemble
those of Macrnronus It is, however, much less numerous and more widely spread, wh"h suZs
that it IS much less given to shoaling. suggests
Salilota austraUs was also taken in seventeen hauls during the first autumnal survey. Only four of
these were situated in the intermediate region; all the rest, including the four richest hauls of ten or
more individuals, were in the southern region. St. WS99 provided the only record of he specie
m deep water over the edge of the shelf at this season ^
During the second survey, made in winter, the species was taken in fifteen hauls. Of these one was
m the northern region, eight were m the intermediate region and only six m the southern region Th"
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
323
Species was recorded seven times in depths greater than 200 m., but four of these records refer to
single individuals. There is here a strong suggestion of northward movement in winter, coupled
perhaps with some offshore movement (mainly confined to the larger fish).
WS73
3
WS216
II
WS785A
3
WS818A
1
WS75
34
WS217
39
WS785B
3
WS838
3
WS78
I
WS218
8
WS791B
I
WS847A
I
WS79
10
WS219
3
WS792A
2
WS848
28
WS80
25
WS222
2
WS792B
7
WS849
3
WS81
3
WS225
I
WS794
3
WS8S5
2
WS83
2
WS234
I
WS799B
3
WS857
2
WS84
5
WS237
I
\VS803
8
WS864
3
WS8g
2
WS2 3g
I
WS804B
I
WS866
15
WSgo
32
WS243
I
WS805
2
WS868
30
WS92
2
WS244
I
WS810
57
WS93
7
WS245
10
WS811II
9
WS586
2 on LH
WS94
I
WS2^0
I
WS812I
3
51
3 juv. in OTL
WS97
6
WS764B
2
WS812II
7
WS863
2 in BTS
WS98
I
WS773
8
WS813
2
WS861
2 in BTS
WS99
4
WS774
I
IVSSJ4
12
WS86j
6 in BTS
WS108
2
WS775
3
WS815
I
WS867
2 in BTS
WS213
I
WS776
I
WS817A
1 1
I inNR
WS214
7
WS781
I
WS817B
15
WS869
I in BTS
During the third survey, worked from late spring through summer and autumn, the species was
recorded from thirty-five trawling stations : nineteen in the southern, eleven in the intermediate and
five in the northern regions. There was only one northern record after midsummer, and only two
records in deep water, both near the southern limits of the range of the species. One rich haul was
made in the intermediate region, six others were all southern with a hint of increasing abundance in
late summer and autumn. Southerly movement during summer seems fairly certain, and this evidently
takes place over the plain of the shelf (cf. depths at the relevant stations, recorded in the Appendices).
I 050
^ 1000-
'uJ
^0 950
or
UJ
^ 0 900-
0 850-
20
25
30
35 40
LENGTHS _CMS
45
50
55
60
Fig. 33. 'Average' K of Salilota aiistialis plotted against the true mean lengths in length
groups, January 1932.
A general examination of length measurements showed that juvenile S. mistralis were nearly always
found in shoal water, and near the northern limits of that part of the range of the species covered
during any given period. There was a tendency for larger fish to be found in deeper water than
smaller ones. The sex ratios appeared to be roughly normal, and females significantly larger than
males, but our data are less complete than for more important species (there was not always time
to sex Salilota) and are not given in full here.
The weight records are sufficient to yield evidence on two important points. The ponderal indices
of Salilota in January, plotted against length, are shown in Fig. 33. The form of the curve (pecked
324
DISCOVERY REPORTS
I-200-I
HOO-
LU
<S)
<
LU
•000-
line) joining the points has no real meaning. It is useful as a guide to the eye, and is a freehand
approximation to the curve expected, judging by the way this relationship varies in other fishes for
which we have better data. It is quite clear, however, that whatever curve were fitted it would show a
point of inflexion between a length of 25 and 30 cm. It is therefore probable that these fish first
attain sexual maturity at about that length.
The seasonal variation in ponderal index of Salilota more than 25 cm. long, for the period December
to the end of March 1932, is shown in Fig. 34. This shows a steady rise, steeper latterly. We have
just seen that most of these fishes are probably mature,
so that we have here a strong suggestion that spawning
takes place in late spring or early summer.
Several series of length frequencies of this species gave
strong modal indications of the younger year classes.
The percentage frequencies of the pooled results, in the
southern and intermediate regions, over stated periods,
are shown in Fig. 35. The implications of these are fairly
clear: the autumn results are most helpful, for we were
then fortunate enough to capture sufficient of the young
fry (about 5 cm. long) in the accessory nets attached to
the back of the trawl, for these to appear as a mode of
equal strength to that formed by the (presumably) I-group
fish at 16 cm. This indicates a growth of some 11 cm.
during the first year of life, and there is also just a hint
of a possible submode at 26 cm. suggesting 10 cm. as the
second annual increment. The main mode of the January
figures, and of the winter (June-July) figures is clearly
due to I-group fish as represented by the 16 cm. mode
in autumn. It will be seen that the shift to the right of
6 cm. of the I-group mode between midsummer and
midwinter is compatible with the growth-rate suggested
above, for growth is usually more rapid during the second
half of the year in fishes living in cold or temperate lati-
tudes. In conjunction with the evidence afforded by the ponderal indices we may therefore say that
S. australis is a fish of rapid growth, probably reaching maturity in the third year of its life. It may
be well to state that rapid growth rates are known in other small Gadidae, notably in Gadus merlangus
(Hartley, 1940, p. 48, where an even higher growth rate than that postulated for Salilota is explained
by the absence of a winter fast in young fish frequenting estuaries).
No time could be spared for investigation of the feeding of S. australis. One was observed to have
eaten large isopods, and it may be permissible to guess (from its colour pattern) that it is more of a
bottom feeder than Micromesistius.
Physiculus marginatus (Giinther). This small species was taken at the following trawling stations,
all in our southern region :
0900-
DECEMBER JANUARY FEBRUARY MARCH
Fig. 34. Seasonal variation of ponderal index observed
in Salilota australis, 1932.
WSys
WS817A
18 of 5-7-3 cm.
I of 18 cm.
WS820
WS821A
1 of 16 cm.
2 of 16-3 and 17 cm.
Norman (1937, p. 54) gives 22-5 cm. as the length of the largest specimen known to him, and
southern Chile and Magellan Straits as other known localities. As far as we know, therefore, the
species is too small and too scarce to be of practical use to man.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
325
MURAENOLEPIDAE
Muraenolepis microps Lonnberg. A single specimen of this Antarctic species, 19 cm. long, was taken
on the Burdwood Bank at St. WS82, in the extreme south of the area covered by the trawling surveys.
Muraenolepis orangiensis Vaillant. A single specimen, 19 cm. long, was taken at St. WS825, near
the edge of the shelf to the north-east of the Falkland Islands. It is also known from the Magellan
Channels.
LENGTHS -CMS _
40
50
60
15-
10
5-
20-
15-
10-
5-
JLIJjJLJB-^HHBl
JANUARY 1932
MARCH -APRIL 1927
JUNE -JULY 1928
Fig. 35. Salilota australis. Percentage length frequencies, southern and intermediate regions,
at stated periods.
CARANGIDAE
Parana signata Jenyns was taken by us only at St. WS847, close inshore in the southern region,
where we secured six specimens between 46-5 and 60 cm. in length. The list of localities given by
Norman (1937, p. 60), southern Brazil, Uruguay, Buenos Aires, Rio Grande do Sul and Bahia
Blanca, indicates that these were probably stragglers. The normal habitat of the species is probably
to the north of our area.
326
DISCOVERY REPORTS
BOVICHTHYIDAE
Cottoperca gobio (Giinther). The range of this species extends to the west coast of southern Chile,
outside our area (Norman, 1937, p. 64). Our records show that most of our specimens were captured
in the southern region, though one rich haul was taken in the intermediate region at St. WS97.
Cottoperca was captured in the northern region twice only :
WS71 41
WS73
WS77
WS79
WS80
WS81 8
WS83 57
WS85 22
WS86 27
WSgo I
WSg2 2
WS93 31
WSg4 4
WS95 13
WS97 79
WS98 I
WS108 I
WS217 I
WS218 2
WS221
WS225
WS237
WS243
WS244
WS245
WS246
WS247
WS248
WS781
WS787 2
WS792A I
WS 795 I
WS797C I
WS803 I
WS804A 3
WS804B 8
WS805 I
WS809 2
5
2
23
4
5
22
I
7
I
WS814
WS815
WS817
WS818A
WS837
WS847A
WS847B
WS848
WS849
WS850
lVS8-^i
WS866
WS872
WS874
Port Stanley
Puerto Acero
WS583
WS836
WS867
I
2
3
I
2
3
I
2
2
I
I
I
I
I
3 with LH (A. G. B.)
I with LH
3 in BTS
49 in BTS
I in BTS
The depth records (Fig. 42) show that Cottoperca ranges from shallow coastal waters right out to
the shelf edge and, rarely, beyond ; but it is mainly an inhabitant of the plain of the shelf. There does
not seem to be any definite migration over the shelf edge ; although the fish showed some tendency
to occur in deeper water in winter than at other times, this movement did not appear to be extensive.
The widespread occurrence of this species in small numbers leaves us with measurements that
cannot usefully be pooled. The largest samples give a hint that the annual length increments over
the main growing period are around 7 cm. Somewhat larger than most of the nototheniiform fishes
of the area, Cottoperca commonly attains a length of some 35 cm. at a weight of about 500 g. Our
largest specimen was 61 cm. long, and we had several over 40 cm. The larger fishes usually occurred
in deeper water, but from our material the females could not be shown to be larger than males. They
were in fact rather smaller, which is a most unusual feature. Our few rich hauls of Cottoperca were
made in autumn, which suggests that schooling may take place at that season. The sex ratio appears
to be normal. Weight records suggest that maximum condition coincides with maximum temperatures
for the year in late summer. As maximum schooling seems to take place in autumn, this strengthens
the suggestion that spawning may take place at that season.
Although edible and of better size than most members of the group in our area, Cottoperca is
unfortunately the most tasteless and undesirable of the nototheniiformes (not a very palatable group)
when used as human food.
The stomach contents of thirty individuals show that Cliipea was the main food, but apart from
these It seems that Cottoperca gobio is mainly a bottom feeder. The other recognizable constituents
were: the small flatfish Thysanopsetta maresii, the Atelecyclid crab Peltarion spimdosum, and other
Brachyura.
Bovichtiis argentinus MacDonagh. We took no specimens of this apparently coastal species, but a
young example from Puerto Madryn was given to Norman by Mr MacDonagh, and since the holotype
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 327
came from the Golfo san Jorge it is evident that it occasionally penetrates southward to the area
of the trawling surveys. Norman (1937, p. 65) points out that it may prove to be identical with
B. chilensis Regan, from the west coast, when comparison of specimens of similar size is possible.
NOTOTHENIIDAE
Notothenia macrophthalma Norman. The holotype of this new species was trawled in deep water
(368-463 m.) at St. WS840, near the Burdwood Bank. Norman (1937, p. 68) states that it is very
closely related to N. squamifrons Giinther, from Kerguelen. Unfortunately no other specimens have
been secured.
Notothenia trigramma Regan. We did not capture any examples of this distinctive species, known
only from the holotype from Port Stanley in the Bruce collection (Norman, 1937, p. 69).
Notothenia canina Smitt. This small coastal species showed a very restricted distribution in our
catches, being taken only near the eastern entrance to the Magellan Straits, and in Grande Bay. All
the records come within our southern region, and it is noteworthy that the types also are from Puerto
Gallegos in Grande Bay (Norman, 1937, p. 70). Norman also states that some specimens from Tierra
del Fuego, undoubtedly referable to this species, were wrongly attributed to N. acuta Gunther by
Steindachner. Our records of N. canina are :
WS8Q 2 WS834 8 WS847A I WS835 45(BTS)
WS833 18 WS837 I WS812 7 WS836 4(BTS)
All but the first of these were obtained in late summer, so that it is perhaps permissible to take some
note of the pooled measurements. The species is a small one, these scanty data showing a mean
length of lo-i cm. with ohjN^o-iogg. A well-defined mode at 10-5 cm. may indicate the I-group
year class. The range of sizes observed was post-larvae (< 5 cm.) up to 19 cm. (see Fig. 36).
The depth relations of this species are extremely interesting: although taken in shallow coastal
waters, it did not seem to come quite so far inshore as the extreme littoral members of the group,
and the range was very restricted. A single specimen was taken at a depth of 100 m., but the other
records were grouped so closely around the effective mean depth of 26 m. that that figure is signifi-
cantly different from the mean depths recorded for all the other species of nototheniiformes withm
our area (Fig. 42 and Table 36). Fig. 3 shows that the slope from the coast down to 80 or 100 m.,
where the plain of the shelf may be said to begin, is steep in the north, and though more moderately
inclined in the south it still shows a far more obvious gradient than is to be found on the plam itself.
It would seem that N. canina is confined to this ' first slope', and mainly to the upper portion thereof.
It is further noteworthy that the species is not known from the other side of the deep water of the
Falkland trough. None has as yet been recorded from the Falkland Islands. The probable ecological
significance of this depth distribution becomes apparent when we come to consider the depth relations
of the two species next to be discussed.
Notothenia jordani Thompson. In our catches this species was even more closely restricted to a
southern area off the eastern entrance to Magellan Straits than was A^. canina. Some of Thompson s
types, however, came from farther north, in the Golfo san Jorge. He found them most abundantly
in the same place as we did, off Cape Virgins, and also within the eastern end of the Straits themselves
as far as the first narrows. It will be noted that we did not find the species very plentiful; forty-three
in the small beam trawl was the only large catch :
WS90 2 WS834 14
WS833 9 WS836 43 (in BTS)
N. jordani is a small species, the mean length of our specimens being 13-8 cm. with (t1,/N= 0-2386.
328
DISCOVERY REPORTS
The pooled length frequencies show one strong mode at i6 cm. which probably represents Il-group
fish. Hints of submodes at 7 and at 12 cm. may indicate the o- and I-groups. The extreme range in
length observed was 6-21 cm.
The mean depth at which we captured N. jordani was 54 m., and the difference of 28 m. deeper
than the mean for TV. canina is statistically significant. The extreme depth range observed was 27-82 m.
Thus the depth distribution seems to be very restricted, the species being confined almost entirely
LENGTH . CMS
Fig. 36. Percentage length frequencies of Notothenia canina and N. jordani^hte summer
stations pooled.
to the lower half of the ' first slope ', not spreading out on to the plain of the shelf or overlapping very
much on to the territory of N. canina which, as we have seen, frequents the upper half of the 'first
slope'. In our catches the difl^erence in mean length oiN. jordani (13-8 cm.) and N. canina (lo-i cm.)
is statistically significant. Possibly we have here an example of the general rule ' larger fish in deeper
water '-so familiar within the limits of single species having wide depth distributions-operating
interspecifically, for the regional distribution of these two species is so nearly coincident that the
difl^erence in depth distribution will be the main factor tending to prevent territorial overlapping
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES J,2C)
between them. The ecological advantage of a system tending to lessen competition between two species
so nearly related in size and habits may be considerable. The size difference, small as it is, is also
apparent in the position of the modes that probably signalize the I-group year classes of the two
species. From Fig. 36 it can be seen that these occur at 10-5 cm. in N. canina and at 12 cm. in
N.jordani. It is also noteworthy that, like N. canina, N.jordani is not known from the Falkland Islands
themselves, being apparently unable to cross the comparatively deep water of the Falkland trough.
Notothenia tessellata Richardson. This also is a small coastal species, and an inhabitant of the ' first
slope' in the southern region, but it was not taken in the same localities as N.jordani and N. canina.
It seems to favour more exposed coasts where the slope is steeper. In our catches it was found most
frequently to the north of the Falkland Islands and around Cape Horn, but it is also known from the
Magellan Channels and southern Chile (Norman, 1937, P- 73)- Bennett's notes, quoted by Norman,
show that it is quite common inshore at the Falkland Islands during the summer, but it is unpopular
as food, though easily caught :
WS72 8 WS8r2 I WS582 ii(onLH)
WS73 6 51 6(inOTL) WS583 8 m B TS)
j^^yr re 55 I (in BTS) Puerto Bueno ? (on LH)
WS8i 28 222 I (in TNL) B. s. Nicholas 2 (on LH)
WS84 I 223 3 Field Anchorage 3 on LH)
WS756B I 724 10 (in seine) ^""^ ^tanle^ ^ by A. G. B )
WS802B I WS576 I New Island 8 (by J. L. H.)
The sizes of N tessellata taken by us show complete overlap with N.jordani with modes, possibly
indicative of age groups, at 11 and at 15 cm. They seem to run a little larger than N. jordam (our
largest specimen of N. tessellata was 29 cm. long), but our samples are insufficient to show whether
this is a constant feature. The effective mean depth at which A^. tessellata was captured (73 m.) is,
however significantly different from that at which N. jordani was taken (Table 36). Indeed, this
figure differs significantly from those for all the other nototheniiformes here studied. Frequenting
more exposed coasts with a steeper slope, this species ranges more widely (into littoral waters on the
one hand and down to the plain of the shelf on the other) than the 'first slope' dwellers previously
mentioned The resulting narrowing of the polygon of depth frequency shows this (Fig. 42).
Notothenia hrevicauda Lonnberg. This small species was never taken in the trawl and appeared to
be restricted to very shallow littoral waters at the Falkland Islands and in the Magellan Channels.
Here the deep water of the Falkland trough does not seem to have limited dispersal as in N. canina
and N jordani. Possibly detached bodies of floating kelp, such as are frequently met with at great
distances from land in the southern ocean, may provide means of dispersal for habitually littoral species,
though they would be less readily available to fishes normally living at slightly greater depths on the
' first slope ' :
Port Stanley , (by A. G. B.) New Island ■ (by J. E. H^ s6 a (in BTS)
Norman (.937. pp. 74-5) bought that two of the types of JV. longicauda Thompson, from shallow
water on the mainland coast, were referable ,0 N. hrevicauda Lonnberg, and that seven others from
Albatross St. 277. at a depth of over 100 m. (of which he saw one that was too decomposed for com-
p r^son probaJy belong to the species he descnbes as N. gunlkeri Norman. The very restr.cted
S per depth distribution of the latter, unhesitatingly identified by Norman m our catches^makes ,t
seem probable that his view is correct (cf. F,g. 42). The two are clearly dtst.ngmshable (from Norman s
desTriclitns) by the extent of the posterior rays of the dorsal and anal fins. These overlap the caudal
in JV. hrevicauda Lonnberg, while in N. guntheri Norman they do not.
14-2
330 DISCOVERY REPORTS
Notothenia guntheri Norman. This small species, named after E. R. Gunther by Norman, is mainly
an inhabitant of the plain of the shelf, in the southern region. It was recorded from the intermediate
region twice, but is not yet known to occur north of 49° S :
WS86
232
WS98 2
WS814 I
WS87
I
WS225 I
WS82'i 2
WS93
21
WS781 I
WS841 I
WS97
4
WS804B I
652 2 (in OTL)
The depth distribution (Fig. 42, Table 36) is remarkably constant, suggesting that A^. gutitheri
frequents the plain of the shelf at all seasons, with little migratory movement. There is no evidence
of any migration into deeper water in winter, such as can clearly be shown for N. ramsayi. The effective
mean depths for A^. ramsayi and A^. guntheri (151 and 147 m. respectively) are significantly different,
but their close similarity masks an extremely wide difference in the dispersion of the depth frequencies
from which they are derived. This is demonstrated by the shapes of the depth-frequency polygons
in Fig. 42, and becomes fully apparent when the seasonal migration of A^. rajnsayi is shown in detail,
when we remember that no hint of any such movement is given by our records of A", guntheri. It is
interesting to note that the seasonal migration of A^. ramsayi is reflected in a dumbbell-shaped polygon
when the results of all seasons are lumped as in Fig. 42, just as was found with Raja brachyurops, a
ray which also migrates over the shelf edge in winter (Fig. 18). By contrast the non-migrating Noto-
thenia guntheri shows a very squat kite-shaped polygon reflecting the narrowness of the depth range
over which that species was observed.
A^. guntheri has not yet been recorded outside the area of our trawling surveys. Little can be deduced
from our measurements of the species beyond the fact that it seems to be a small one ; our largest
specimens were 20 cm. long. Unlike the small species inhabiting shallow depths it was captured
almost entirely in the trawl, so that all but the largest specimens were presumably outside the selective
action of the gear used. A strong mode at 16 17 cm. in the only large catch may indicate a year class,
probably the largest well-defined one (perhaps Il-group). There is just a hint of another at 12 cm.
(? I-group), which suggests a growth rate similar to that of other Notothenia spp. of comparable size.
There is also a slight suggestion of conformity with the behaviour of several other members of the
group in the fact that the only considerable concentrations of A^. guntheri were met with in autumn,
whde at several summer stations it was taken singly, suggesting wide dispersal at that season.
Notothenia ramsayi Regan. This was the commonest fish trawled by us in the area investigated:
9665( + ) individuals were recorded, a number which just exceeds the combined totals for Macruronus
(4953) and Merhiccius (4704). As a result of the smaller size of the Notothenia, however, the weight
of hake taken was five times as great, while even the slender Macruronus weighed 2^ times as much.
It is of course highly probable that the Falkland herring is present in vastly greater numbers than
even Notothenia ramsayi, but we have no means of assessing the relative abundance of such small
semi-pelagic clupeoids vis-d-vis demersal fishes. Captures of A^. ramsayi in the ' Trawl + accessory
nets ' were as shown on opposite page.
It can be seen that A^. ramsayi was very widely distributed over the plain of the shelf, and (in
autumn and winter) beyond the shelf edge. Unlike most of the nototheniiformes caught in the trawl,
which show a strong preponderance to the southward, this species appeared to be almost equally
abundant in the northern and intermediate regions.
The smallest individuals were found mainly to the north in relatively shallow water, but fish almost
certamly less than two years old were plentiful at times in the southern region. The larvae, presumably
denatant, must be carried northwards by the prevailing current, and it is thus probable that the fry,
by tendmg to work inshore, are assisted southwards by the counter-current. Without some such
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
331
mechanism it is difficult to see how the species is maintained within its ecological norm, for the contra-
natant abilities of the young fish cannot be great during their second year of life (cp. E. S. Russell,
1937. P- 321)-
WS71
WS72
WS73
WS76
WS77
WS78
IVS79
WS80
WS81
WS83
WS84
WS85
WS86
WS87
WSgo
WSgi
WS92
WS93
WS94
WS95
WS96
WS97
WS98
WS99
WS108
WS109
WS210
WS211
WS'ziz
WS213
WS214
WS215
WS'216
WS'217
125
4
90
42
3
2
125
42
74
136
I
5
lOIO
I
4
72
191
18
270
92
I
1615
43
18
166
10
27
63( + )
I
7
1362
I
117
358
WS218
WS219
WS220
WS222
WS223
WS225
WS226
WS234
WS235
WS237
WS23g
WS242
WS243
WS244
WS245
WS246
WS248
WS250
WS756A
WS756B
WS764A
WS764B
WS765
WS771
WS772
WS773
WS774
WS775
WS776
WS781
WS784
WS785A
WS785B
WS785C
232
8
26
19
I
27
I
22
4
47
24
I
25
30
35
31
13
I
2
14
9
41
17
221
2
8
II
I
2
29
13
6
10
4
WS787
WS788
WS789
WS790A
WS790B
WS79iB
WS792A
WS792B
WS793
WS794
WS795
WS796A
WS796B
WS797B
WS797C
WS798
WS799A
WS799B
WS800A
WS800B
WS801
WS802A
WS802B
WS804A
WS804B
WS8o^
WS806
WS807
WS808
WS809A
WS809B
WS810
WS811II
WS812II
140
I
2
4
3
57
432
432(+;
220
85
70
247
22
I
7
2
48
49
12
42
I
3
I
16
41
6
104
20
36
64
3
16
14
15
WS813
WS814
WS815
WS816
WS817A
WS817B
WS819A
WS823
WS824
WS82S
WS837
WS838
WS839
WS847B
WS848
WS849
WS850
WS851
WS853
WS855
WS8s7
WS858
WS859A
WS859B
WS860
WS862
WS864
WS866
WS868
WS870
WS874
22
I
23
14
4
7
I
I
2
4
3
7
5
5
17
4
5
18
82
4
42
15
7
35
8
16
We also captured N. ramsayi with 'Other gear' at the following stations:
51
652
WS750
WS752
WS754
35 (in OTL)
2 (in OTL)
I (in NR)
I (in NR)
I (in NR)
WS755
WS767
WS779
WS832
WS852
5 (in NR)
10 (in NR)
I (in NR)
6 (in NR)
12 (in BTS)
WS856
WS861
WS863
WS865
WS867
3 (in BTS)
I (in BTS)
16 (in BTS)
23 (in BTS)
15 (in BTS)
WS871 I (in BTS)
WS874 I (in NR)
N. ramsayi was rarely abundant in depths of less than 100 m., and never observed by us in depths
of less than 50 m. The 'effective mean depth' was found to be 151 m., but the dispersion was very
wide as can be seen from the diagram in Fig. 42. It will also be noted that the depth-frequency
polygon tends towards the dumbbell shape, though the upper mode is by far the greater. In view of
our findings with Raja brachyurops it was therefore obviously desirable to test the possibilities of a
seasonal migration over the shelf edge in order to account for this. The data were arrayed by seasons
in 50 m. depth categories and relative abundance, as indicated by numbers of N. ramsayi per hour's
trawling, was computed for each.
The resuhs shown in Table 31 are also indicated diagrammatically in Fig. 37. The graphs in the
top half of this figure seem sufficient to show that the very varied amount of trawling within individual
depth categories does not affect the main result ; namely that A^. ramsayi shows a well-marked migration
into deep water over the shelf edge in winter and is relatively most numerous on the shelf in summer
Actual numbers of A^. ramsayi were highest in autumn, at all but the greatest depths. Nearly half
the rich hauls of more than 100 individuals and all three exceptional hauls of more than 1000 m-
332
DISCOVERY REPORTS
dividuals were obtained at that season. From this it seems safe to infer that schooHng of A'^. ramsayi
is most marked during autumn.
Table 31. Seasonal variation in relative abundance of Notothenia ramsayi at different deptlis
Season
Depth groupings, m.
1-50
51-100
101-150
151-200
201-250
251-300
301-350
351-400
401-450
>45o
Spring
Hours trawling
0
6
8
0
1
4
1
0
0
0
0
N. ramsayi per hour
0
39
14
0
8
8
0
0
0
0
Summer
Hours trawling
2
32
57
7
5
5
5
i
2
i
TV. ramsayi per hour
0
8
yi
16
5
0
1
5
0
1
0
Autumn
Hours trawling
\
8
25
7
6
1
0
0
0
0
A'^. ramsayi per hour
0
II
157
40
238
8
0
0
0
0
Winter
Hours trawling
0
1
9
6
7
•/
i
0
0
0
N. ramsayi per hour
0
0
49
29
20
67
0
0
0
0
An investigation of the mean lengths within depth categories at each season showed that there was
a significant increase of size with depth. The fish caught between 201 and 250 m., and 251 and 300 m.,
in winter, do not show this ; but all the other observations, taken in pairs successively (twelve pairs in
all) show the deeper sample of the pair to contain significantly longer fish (Table 32). It is therefore
clear that A^. ramsayi conforms to the general rule 'larger fish in deeper water'.
Table 32. Variation in mean length of Notothenia ramsayi at different depths during
each of the four seasons of the year
Depth
range
m.
Spring
Summer
Autumn
Winter
Mean
length
cm.
oUN
Mean
length
cm.
ol^lN
Mean
length
cm.
<^l,IN
Mean
length
cm.
o^N
1-50
51-100
101-150
151-200
201-250
251-300
None
8-9
II-8
None
20-5
26-9
0-0270
o-i8gi
0-1250
2-0763
None
13-0
13-9
17-8
26-9
None
0-2107
o-oigo
0-5113
1-3578
None
i6-7
219
22-3
27-5
31-0
0-2221
0-0045
o-og28
o-oo8g
1-1250
None
None
23-4
24-8
24-2
26-1
0-0348
o-og27
0-1386
0-0513
These results also indicate that the vast majority of that part of the N. ramsayi population found
beyond the shelf edge (i.e. in depths greater than 200 m.) are more than 20 cm. long. Reference to
the actual frequencies shows that only sixty-three out of 1894, or 3-3 % , of the fish captured over the
edge were less than 20 cm. long. The winter migration to deeper water is thus almost entirely confined
to the larger fish that, as will presently be shown, are almost certainly more than two years old.
The numerous length-frequency data available for N. ramsayi can only be used to test the probable
ages of the younger fish, by Pettersen's method. It was rarely possible to sex the individuals, owing to
pressure of work upon the more obviously useful species, and, when sexing was achieved, it was
found that a very large majority of the larger fish (of more than 25 cm.) were females. It is therefore
almost certain that after maturity is reached the male and female growth rates diverge, the females
growing the faster. Consequently one cannot base any conclusions as to age groups or growth rate
upon the length frequencies of the larger fish in unsexed data. If the mature males are considerably
smaller than the females, the apparent excess of the latter will in the main be accounted for by the
selective action of the net. For the younger stages, however, the rich autumn hauls (many of them
taken in the ' accessory nets') furnish length-frequency data that seem to show age groups with some
certainty, judging by the consistency with which modes recurred at the same lengths.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
333
Fig. 37. Depth relations of Notothenia ramsayi. Above: relative abundance of fish caught within each depth-grouping at each
season (circles), and the relative amount of time spent trawling within the same limits (crosses). Below: widths of polygons
are proportional to the relative abundance of fish within the limits shown. (Data from Table 31 transposed to a percentage
basis for comparison.)
334
DISCOVERY REPORTS
30-
;20-
10
B
Fig. 38. Percentage length
frequencies of A'^. ramsayi
taken (A) between 51 and
100 m., and (B) between loi
and 150 m., in spring.
We have no extensive autumn data for the smallest Notothenia ramsayi, but the pooled length
frequencies of spring samples taken in 51-100 and 101-150 m. show such strong modes about
8-10 cm. (with the larger fish in the deeper water) that it is reasonable to assume that they represent
a year class. These fish are thought to have been almost one year old — o-group becoming I-group
(Fig. 38).
In autumn we obtained eleven rich hauls of between eighty and 161 6 individuals of this species,
and on plotting the percentage length frequencies (Fig. 39) it appeared that modes at around 14-16 and
22-23 cm. recurred with such consistency that there can be little doubt
that they represented year classes. They are thought to indicate I-group and
I I-group fish respectively. The scale of the figure is necessarily much reduced,
in order to permit comparison of all the samples on one page. Table 33,
summarizing the important points arising from Fig. 39, and giving relevant
geographical data, has been prepared to cover any loss of information due
to the unavoidably small scale of the figure.
It will be seen that clear evidence of either or both of the two year classes
mentioned is provided by all but three of these samples notwithstanding
their diverse locations. At St. WS83 there was a strong mode at 17 cm. —
considerably higher than the modal values for most of the presumed I-group
fish captured around that time, which were at 14 and 15 cm. It is believed
that this is explained by the geographical position of St. WS83 — close in to,
but on the southern side of, the Falkland Islands, considerably farther south
than any other station at which such small A^. ramsayi have been taken in
quantity. From the general distribution of the species as already described
it seems at least highly probable that only the largest members of the I-group
would be likely to penetrate so far south. This notion is perhaps strengthened by the strong 14 cm.
mode shown by the sample from St. WS73, almost equally close in to the islands, but to the north
of them.
The very rich sample from St. WS97 yielded length frequencies which are not incompatible with
the idea that the Il-group predominated, but the mode is ill-defined, as stated in Table 33. The
suggestion that this may be due to the slower growth rate of mature males as compared with females
is strongly supported by the fact that such sexually differentiated grovi^h is known to take place in
various other fishes (e.g. hake).
Finally, at the only deep station at which a large haul of N. ramsayi was secured in autumn,
St. WS214, it seemed that we were dealing with an altogether larger age group (? Ill-group) with the
mode at 27 cm. This would be in full accordance with our findings as to the general relation between
depth and size of fish where N. ramsayi (and many other species) are concerned.
Considering these results in conjunction with the general distributional data, it seems probable
that early growth of N. ramsayi takes place somewhat as follows. The fish probably hatch in early
summer and grow rather more than 10 cm. in their first year of life, at least 8 cm. in their second year
and 6 cm. in their third. At this point maturity is probably reached and no reliable conclusions can
be drawn from unsexed data, for it is probable that mature females grow considerably faster than
males, but there is some evidence suggestive of a 4-5 cm. increment during the fourth year.
I have included this brief and admittedly speculative suggestion as to growth rate in N. ramsayi
because it will at worst provide a working hypothesis if any future work on the shelf is possible, and
we know so little of the growth of any of the fishes in southern temperate waters.
Several records of the stomach contents of A^. ramsayi were made in autumn and winter, when some
75 % of the fish contained recognizable food. The number of times food of each category was recorded,
10
10
lo-
ws 73
10-
10-
10-
10-
10-
10-
10-
20-
10
WSS3
WS97
30-
20-
10-
WS2I7
W5859B
Fig. 39. Percentage length
frequencies of TV. ramsayi at
autumn stations with more
than eighty individuals, show-
ing probable year classes.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 335
0 and an arbitrary weighting system (based on some contemporary volumetric
records) which provides some approximation to relative values of major food
categories, are shown in Table 34.
Clearly N. ramsayi is much more of a bottom-feeder than most of the other
Patagonian species for which we have any data. Benthic Crustacea were the
largest item in the dietary observed, and a considerable proportion of poly-
chaetes were eaten. A'^. ramsayi, however, feeds heavily upon Falkland herring
whenever opportunity offers, though the same might be said of almost every
animal in the area capable of swallowing them, including such other typical
bottom-feeders as Cottoperca and some of the rays. In winter some plank-
tonic food was taken by some of the larger A^. ramsayi far offshore, and this
brings us to a point that the existing data are not adequate to solve, namely,
that it is highly probable that in this species, as in so many other larger
demersal fishes (cf. hake, cod, Macruroniis, etc.), increased size and mobility
is accompanied by a change-over from a carcinophagous to a fish diet. It is
therefore probable that our approximate diagram really represents a sort of
summation effect of the dietary of Notoihenia ramsayi over the whole of its
life, and that had it been possible to examine stomach contents of samples
of small fish and larger fish separately, we should find the organisms occupy-
ing (roughly) the upper and lower halves of the diagram in very different
proportions. I should expect the benthic invertebrates to be relatively more
important in the smaller Notothenia, and fishes to predominate in the larger.
Another notable feature is that no Miiiiida were recorded from the stomachs
of Notothenia ramsayi, but it is highly probable that they would have been
had it been possible to make summer observations on this point.
Enough has been said to show that, if only on account of its small size,
A'^. ramsayi is not likely to be of much value as human food in spite of its
abundance within our area. The largest individuals of 35 cm. and upwards
may attain a weight of a pound, and are not less palatable than the (rather
insipid) larger Antarctic members of the genus. The average weight is little
more than J lb. however, and the majority of the trawled fish are just too
large to be fried and eaten as ' sprats ', a process that the crew of the ' William
Scoresby' found most efficacious with the young fry some 10-12 cm. long.
These were said to be '. . .exceedingly good, though the bones are rather
hard. The flesh resembles that of whiting'.
A^. ramsayi is of the first importance as a forage fish for larger and more
useful species on the Patagonian Continental Shelf, being one of the main
sources of food supply for hake and Macruronus, as we have already seen.
Notothenia wiltoni Regan. This species seems to be one of the more extremely
littoral Nototheniiformes, though it may depart from the shallowest waters
in winter. In our collections with ' Other gear' it was taken between 2 and
35 m. of water at the Falkland Islands and in the Magellan region, the mean
depth being but 5 m.:
Port Stanley 45 (by A. G. B.) 56 2 (in BTS)
Field Anchorage i (on LH) 222 i (m TNL)
55 I (in BTS)
None was taken at the regular trawling stations.
15
336
DISCOVERY REPORTS
Table 33. Summary of the observatiom on length frequencies of Notothenia ramsayi at autumn
stations shown in Fig. 39, with relevant geographical data
Station
WS73
WS79
WS83
WS86
WS92
WS94
WS95
WS97
WS108
WS214
WS859B
Date
6. iii. 27 51° 02'
Position
Lat. S
13. ni. 27
24. iii. 27
3. iv. 27
8. iv. 27
16. iv. 27
17. IV. 27
18. iv. 27
25. IV. 27
31.V. 28
25. iii. 31
oi|'
52 29
53° 53^
51° 58i'
50° ooi'
48° 58'
49° ooV
48-31'
48° 25'
45° 14'
Long. W
58° 55'
64° 59r
60° 07I'
60° 34I'
65° 01'
64° 57l'
64° 45'
61° 58'
63° 34'
60° 40'
61° 56'
Distance in sq. m.
From
Main-
land
226
126
232
176
125
107
68
From
Falk-
lands
13
78
> 100
o> 10
> 100
165 > 100
94
199
Depth
m.
142
132
133
149
144
118
109
146
119
214
108
No. of
N. ramsayi
measured
82
124
131
lOIO
183
269
92
1616
166
1362
81
Remarks
Very strong mode about 14 cm., hint of
a submode at 22 cm.
Fairly strong submode about 14 cm.,
mode at 22 cm.
Strong mode at 17 cm.
Strong mode about 23 cm.
Strong mode about 22-23 cm.
Hint of a submode at 16 cm., mode at
23 cm. Higher values tail off rather
gradually, owing perhaps to larger
males growing more slowly than females
of similar age
Submode at 14 cm., mode at 22 cm.
Ill-defined mode — the values at 19-22
and at 24 cm. all high. This is readily
understandable if the hypothesis as to
slower growth of mature males is ac-
cepted
Two well-defined modes at 14-15 and
at 22-23 ^^■
Strong mode at 27 cm. Here we seem
to have a later year class (? Hl-group)
dominant than at any of the other
stations. Note that this is the only rich
autumnal haul in deeper water
Very strong mode at 15 cm.
Table 34. Observations of stomach contents of Notothenia ramsayi in autumn and winter, and an
approximate evaluation of the relative importance of the main food categories by arbitrary zveighting
{data of Fig. 40)
Food category
Times
recorded
Food category
Times
recorded
Food category
Times
recorded
Sagitta sp.
Nemertinea
Nereidae
Sabellidae
Terebellidae
Other Polychaeta
I
I
2
8
4
5
Mysidacea
Serolis sp.
Other Isopoda
Parathcmisto
Hyperiidae
Other Amphipoda
Euphausiidae
Paralomis granulosa
Eurypodius latreilli
Other Brachyura
2
II
3
I
I
15
I
I
7
I
Cephalopoda
Ophiuroidea
Holothuria
Clupea
Notothenia sp.
Other fish
I
2
I
13
2
2
Major (grouped)
Times
Weighting
Relative
Relative impor-
food categories
recorded
factor
importance
tance as %
Polychaeta
20
X I
20
6-7
Benthic Crustacea
41
x3
123
41-1
Other benthic invertebrates
4
X2
8
2-7
Planktonic invertebrates
4
X I
4
1-3
Cephalopoda
I
x8
8
27
Clupea
13
x8 .
104
34-8
Other fish
4
x8
32
107
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 337
N. wiltotti is somewhat larger than the shallow-water nototheniids discussed previously, attaining
a length of 34 cm. with individuals of more than 20 cm. fairly common. It is very easily confused
with small specimens of N. ramsayi, and resembles A^. longipes Steindachner so closely that Norman
(1937. PP- 8°~2) found it impossible to give complete synonymies for the two species, and believed
that they might yet prove to be identical. However, Norman was able to give characters that should
prevent confusion with the deeper-water A^. ramsayi in the future.
Bennett's interesting notes on the occurrence of this species in shallow harbour waters at the
Falkland Islands are also quoted by Norman. It is common inshore during summer, becoming scarce
in autumn, when the gonads are enlarged. Hence Bennett concludes that its departure may be for
Fig. 40. Approximate relative importance of the main food categories of N. ramsayi, the data
being arbitrarily weighted as shown in Table 34.
breeding purposes. N. wiltoni is known as ' rock-cod ' locally, but this vernacular name, perhaps the
most promiscuously applied of all fish names m English-speaking countries throughout the world, is
also eiven to several other Notothenia spp. in the Falkland Islands.
Notothenia longipes Steindachner. We have just mentioned the possibility of confusion between this
snecies and N. wiltoni. Such scanty distributional evidence as we possess favours the view that they
really are distinct. All the specimens referred with any confidence to A^. lorigipes seem to come from
the west coast of southern Chile or the western end of the Magellan channels, where they show a
depth distribution rather similar to that of A^. canina at the eastern end of the straits (Fig. 42) and very
different from that of the more exclusively littoral distribution shown by A^. wrltont and various other
Notothenia spp. We took no N. longipes at regular trawling stations, but specimens obtained as shown
338 DISCOVERY REPORTS
below gave an 'effective mean depth' of 43 m. with extreme range 12-78 m. Our largest specimen
was 17 cm. long, but most of them were very small:
WS582 I (on LH) WS583 20 (in BTS) Ringdove Inlet i (on LH)
Notothenia sqiiamiceps Peters. The 'William Scoresby' did not capture any specimens of this small
species, which seems to have an extremely littoral distribution at the Falkland Islands and in the
Magellan region. The 'Discovery', however, obtained a few in the autumn of 1926, all from East
Falkland and some with ripe eggs. None of these can have come from depths greater than 16 m. The
largest specimen was but 12 cm. long:
53 3 (in RM) 55 I (in BTS) 56 3 (in BTS)
Notothenia sima Richardson. This is yet another exclusively littoral species, taken plentifully by
Mr A. G. Bennett with shore seines and traps in Port Stanley, and (rarely) by the ' Discovery' using
the small beam trawl in very shallow water (10-16 m.); but it was never encountered at the regular
trawling stations. Norman (1937, p. 85) quotes interesting observations by R. Vallentin that seem to
show that this species spawns in littoral waters in spring or early summer, i.e. at quite a different time
of year from that at which A^. squamiceps was taken with eggs. A^. sima is another small species.
Bennett's largest specimen was 14 cm. long:
Port Stanley (Nov. and Feb.) 24 (in seine and trap, A. G. B. co//.) cc 2 (in BTS)
56 I (in BTS) ^:> K I
Notothenia cormicola Richardson. This species has been recorded at depths down to 35 m., but in
the main it seems almost as exclusively littoral as the last two. It was never taken in the trawl, and
with one exception (from Cape Horn) our few specimens were all obtained at the Falkland Islands.
A^. cormicola resembles A^. sima very closely, but may be distinguished by the absence of scales on the
lower part of the operculum (Norman, 1937, p. 85). Norman (p. 87) also quotes references from the
literature that point to the possibility of an extended breeding season in A^. cornucola. The record of
this species from New Zealand is extremely doubtful, as Norman has shown, but it is known from
southern Chile (northwards to Chiloe) and the Magellan channels in addition to the localities where
we obtained specimens. The largest A^. cornucola obtained by us was only 13 cm. long:
52 I (on LH) 222 I (in NRL)
53 I (in RM) Port Stanley Several (A. G. B. coll.)
55 I (in BTS) New Island (West Falkland) 6 (J. E. H. coll.)
56 I (in BTS) '
Notothenia elegans Giinther. This little species with its slender body and proportionately large fins,
so well figured by Col. Tenison (in Norman, 1937, fig. 42), cannot easily be confused with any of the
other Patagonian Nototheniidae. It was the smallest nototheniid that seemed regularly to inhabit
moderately deep water on the plain of the shelf and, rarely, beyond. This is shown by the depth-
frequency distribution (Fig. 42). N. elegans seems, moreover, to have a more northerly regional dis-
tribution than most of the other species, having been recorded twice in the northern region, and more
frequently in the intermediate than in the southern region in our catches. Too much stress should not
be laid on this point, however, because on account of its shape and small size (we took none more
than 12 cm. long) it is certain that our gear could not sample this species adequately. It was captured
chiefly in ' Other gear ' or in the accessory nets attached to the back of the trawl. A^. elegafis is doubtless
eaten by larger fishes, but is probably not sufficiently common to rank high in importance as a forage
species :
WS83 I WS808 3 WS861 I (in BTS) WS878 ii(inNR)
|f^9J I 51 4(inOTL) WS863 26 (in BTS) '
WS237 I WS767 5(inNR) WS867 I (in BTS)
WS795 1 WS836 6 (in BTS) WS873 I (in NR)
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 339
Notothenia macrocephala Gunther. Norman (1937, pp. 89-90) has shown that there seems to be no
reason to doubt the identity of this species with examples from Kerguelen and New Zealand. It is one
of two Nototheniids that seem to have a truly circumpolar «/6-Antarctic distribution. The silvery
young seem to live pelagically for a much longer period than those of most other members of the
group, and this may have favoured wide dispersal of the species.^ A striking illustration of this was
afforded by the capture of nine specimens (up to 9 cm. long) at St. 63 in 48° 50' S : 53° 56' W, right
out in the open South Atlantic some 300 miles NE x N of the Falkland Islands. These specimens were
caught at the surface with a dip-net, and not by hand-lines as erroneously stated in Norman's report.
We got no A'^. macrocephala in the trawl^the sizeable adults seem to be mainly littoral in their
habits — but some were captured with 'Other gear', and Mr Bennett secured numerous specimens
with seine and hand-lines in Port Stanley. Bennett's notes, quoted by Norman (loc. cit.), show that
the larger individuals are common inshore at the Falkland Islands, where they are known as 'yellow-
bellies ', and stay close inshore later in the year than most other Nototheniidae. He found them good
eating though they are rarely used as food. At the Falklands they attain a length of over a foot and
I lb. weight. The depth distribution shown in Fig. 42 refers to the larger individuals, taken only in
littoral waters, and excludes the pelagic young which (as we have seen) may at times be met with far
out at sea over oceanic depths.
We obtained specimens of Nototheftia macrocephala as follows :
63 9 (in dip-net) 222 i (in TNL)
Port Stanley 8 ( + several not preserved, A. G. B. fo//.) 229 i (in NiooH)
From Phillipps (1921, p. 123) we learn that in New Zealand this fish goes by the name of 'Maori
chief, but in Wellington, where a few were marketed in autumn, the fishermen know it as 'More-
pork'. The fish should not be confused with the hairy owl {Ninox novae-selandiae) or the Tasmanian
night-jar [Podargus ciivieri) that go by the same vernacular name.
Notothenia microlepidota Hutton. We did not obtain any specimens of this species, which is, how-
ever, of special interest because Norman (1937, pp. 90-1) believed that some Patagonian specimens,
variously described, and authentic New Zealand ones, were identical. ' There is, thus, a second species
common to the Patagonian and Antipodes regions.'
Dissostichtis eleginoides Smitt. This is the largest of the Patagonian Nototheniidae. Superficially it
bears a very strong resemblance to a hake, but Norman (1937, p. 92) found the skeletal relationship to
Notothenia very close. It seems to be a rare fish in the Patagonian region, and we captured nine
specimens only. Dissostichtis was always taken in the trawl, never in 'Other gear'. It occurred at
such widely divergent depths that the depth relations could not be expressed by the methods used for
other Nototheniiformes, and individual occurrences are plotted against depth in Fig. 42. The largest
specimen, which was 90 cm. long and weighed 7710 g. (nearly 17 lb.), was taken in 297 m. At
shallower depths the specimens showed regular increase in size with depth, ranging from a specimen
of 13 cm. in 84 m. to one of 33 cm. in 172 m., but two five-pounders (about 64 cm.) were taken m
one of the deepest hauls made (418 m.). It would seem that at most seasons only immature individuals
of this species are to be found on the shelf, the larger fish ranging the deep water beyond the edge.
They may migrate to shallower water to spawn, but we lack any direct evidence on the matter. Our
larger specimens showed a steady increase in ponderal index with increasing length, from about 07
at 47 cm. to a value exceeding unity for the largest fish caught:
WS75 I WS97 I WS98 3 WS245 2 WSSjg 2
1 It is quite probable that adequate search would reveal the presence of iV. macrocephala at Gough Island, the Crozets,
Marion and Prince Edward Islands, and perhaps even at St Paul's Island and New Amsterdam. Our knowledge of the fish
faunas of the isolated sub-Antarctic islands is deplorably fragmentary, especially in the Indian Ocean sector.
340 DISCOVERY REPORTS
D. eleginoides is one of the few Patagonian fishes known also from the Antarctic Zone. It has been
taken among the islands off Graham Land (Vaillant, 1906, pp. 36-9).
Eleginops maclovinus (Cuvier and Valenciennes). This genus may at once be distinguished from
others of the family Nototheniidae by the entire absence of a lower lateral line. E. maclovinus is a
sizable fish, and appears to be exclusively littoral in its habits. We never captured any in the trawl.
Specimens were secured with ' Other gear ' as follows :
724 10 (in seine) Connor Inlet 2 or 3 (on LH) WS586 i (on LH)
In addition to these, Mr A. G. Bennett collected six specimens for Norman's report by seine-netting
in Port Stanley harbour, mostly at Weir Creek.
Bennett also provided valuable notes on the habits of the fish which Norman (1937, pp. 93, 94)
quotes at length. The species is known locally as 'mullet', and this vernacular name is justified not
only by its strong superficial resemblance to true mullets {Mtigil spp.), which are absent from the
region, but also by its habits. Notable among these is its tendency to run right up into the mouths of
fresh-water streams on the last of the rising tide. Bennett tells us that it may grow up to 2 ft. long,
but unless this length is at times considerably exceeded, his figure of 15 lb. for the maximum weight,
quoted by Norman, must be a mistake. A fish 2 ft. long and 15 lb. in weight would have a ponderal
index just over 3-0 (calculated from K=w (g.)// (cm.)=^x 100),^ and it is clear that no fish approaching
the proportions of Eleginops could give even half this value. Authentic weight records of smaller
specimens give ponderal indices from 0-69 to 0-85, and if we assumed an index of i-o for a 2 ft.
specimen its weight would be just 5 lb. Conversely, even if we assumed an index as high as 1-25, a
15 lb. fish would be no less than 32 in. long. I believe that in all probability Bennett actually wrote
5 lb. and that some error crept in subsequently.
Eleginops is eaten quite frequently at the Falkland Islands, but often has a muddy taste. Otherwise
it would be a promising subject for small-scale local exploitation by seine-netting. It extends round
both coasts of the mainland of South America from the River Plate in the east to northern Chile in
the west — much farther towards the equator than most other Nototheniidae.
HARPAGIFERIDAE
In his later report on the coast fishes of the Antarctic Zone Norman (1938, p. 43) places Harpogifer,
with four other (exclusively Antarctic) genera in this separate family and not, as heretofore, in the
Nototheniidae. The chief characteristic of the family is absence of scales on the body.
Harpagifer bispinis (Schneider). This is the only member of the family found in the sub-Antarctic
Zone. To the southward it has a wide distribution in the northern part of the Antarctic Zone, having
been recorded from Graham Land and almost all the isolated island groups (Norman, 1938, pp. 52-3).
Norman's description of it as mainly littoral (and frequently intertidal) in habit, applies accurately
enough in the Patagonian region, though even there we have taken it down to 95 m. ; and the depth
relations, shown in Fig. 42, show it closer to the 'first-slope' dwellers than to the exclusively littoral
species of Nototheniidae. Farther south, however, where the intertidal zone is usually small and
subject to ice action, Harpagifer usually occurs at greater depths, although it is true enough that it
' The high 'condition factors' quoted in some salmon literature (of the order 36-40 or more) are obtained from the formula
w (lb. & fractions)// (inches)^ x 10,000. Menzies' Scottish 'coefficients' make use of the same heterogeneous British units, but
get rid of the unwieldy decimal ciphers resulting from the first term of the formula by dividing by 0-00036, a figure just below
the mean for 'normal' east of Scotland salmon. This has the effect of bringing all the values close to unity (and close to those
obtained by direct use of the metric formula). His system is perhaps ideal so long as we wish to consider salmon only (and
only east of Scotland salmon !) but the principle of dividing by the mean value implies that the factors for any given species
(or local race) of fish will be grouped close around unity. If we wish to visualize the difference in the ratio weight to cube
of length between fish of diverse form it is not possible to use his method, whereas direct application of the metric formula
permits this, and with less heavy arithmetic. A propos the immediate problem above, Menzies' figures (1925, p. 190) show
that a 24 in. salmon should weigh about 5 lb. 2 oz.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 34i
keeps close in to the land. In the Patagonian region we captured specimens of this very small species
at one trawling station only: WS89, in shallow water, where three were taken in one of the ' accessory
nets'. Other specimens were obtained for Norman's 1937 report as follows:
Port Stanley 24 (under stones, A. G. B. coll.)
New Island (West Falkland) 3 (J. E. H. coll.)
WS749 3 (in NR)
WS832 I (in NR)
None of these exceeded a length of 10 cm. If readily available between tidemarks at the Falkland
Islands, Harpagifer might be useful as bait.
CHAENICHTHYIDAE
Champsocephalus esox (Gunther). This species occurred with some frequency in trawl catches on the
rough ground near the Falkland Islands in the southern region. A single large specimen was taken
farther north, at St. WS97 in 49° S:
WS71 30
WS73 6
WS75 29
WS76 I
IVS81 I
WS83 41
WS84 2
WSgs I
WSgj 2
WS97 I
WS756B 2
WS802B I
WS823 5
WS834
WS837
51
724
Port Stanley
22 (in OTL)
2 (in seine)
3 (A. G. B. coll.)
Most were taken in autumn, and none during the winter survey when a large proportion of the
stations were worked in deep water beyond the shelf edge. The species is, indeed, very much an
inhabitant of the plain of the shelf, judging by the depth distribution shown in Fig. 42. From this it
can be seen that while a few have been taken in shallow
LENGTH. CMS littoral waters there was only a single specimen from
beyond the shelf edge. From Bennett's notes, quoted
by Norman (1937, p- 96), it would seem that if there is
any inshore migration of Champsocephalus, it will take
place in late summer or autumn, for his records of the
infrequent capture of the species in littoral waters at the
Falkland Islands all date from that season.
The length frequencies of our autumn-caught speci-
mens (Fig. 41) show two very strong modes at 15 and
at 28 cm. If these indicate year classes it follows either
that this species is of extremely rapid growth,^ or that
''"■"■ r;£r^;l*?;iXTi.°r""'"- a„ intermediate year class ts entirely lacking in our
samples. The presence of such an mtermediate year class
would indeed, bring the growth rate into line with that observed for Notothenia ramsayi, a related
fish of much the same size, but it is extremely difficult to see why such an intermediate age group
should be absent from the grounds frequented by both younger and older fish of the same species.
Moreover in view of our extensive observations, coupled with those of Hamilton and Bennett in
littoral waters, we should certainly expect to have found evidence of such a group somewhere if it
actually exists. . „ • u
Champsocephalus esox, which has near relatives in the Antarctic Zone, is excellent eating: much
firmer and of better flavour than most Nototheniiformes. Unfortunately, we rarely captured it in any
great quantity, and our largest specimen was only 36 cm. long, so that it is unlikely that the 'pike',
as it is called in the Falkland Islands, could be exploited profitably.
1 14 cm. in its second year. Among better-known fishes of similar size such a rate is approached by estuarine pollack and
whiting, which later descend to the sea (Hartley, 1940, pp. 47-5°)-
342
DISCOVERY REPORTS
SUMMARY OF OBSERVATIONS ON PATAGONIAN NOTOTHENIIFORMESi
The survey of our observations on the distribution and bionomics of the individual species of Pata-
gonian Nototheniiformes enables us to present some of the main features regarding the group as a
whole in more concentrated form. Table 35 gives a list of all the species known from the area up to
the time of the publication of Norman's report, and shows which were captured in our trawls or with
'Other gear', with brief notes on records of occurrence outside the area investigated.
Table 35. List of the Patagonian Nototheniiformes, their occurrence in our material and their
distribution outside the area surveyed
Page
references
Family
Species
in Norman
(1937)
A
B
C
Distribution outside the area
Bovichthyidae
Cottoperca gohio (Giinther)
63-65
+
+
—
Southern Chile
Bovichtus argentinus MacDonagh
65
—
—
+
Northwards to La Plata
Nototheniidae
Notothenia macrophthalma Norman
68-69
+
—
—
Holotype only, not yet known else-
where
N. trigramma Regan
69
—
—
+
Holotype only, not yet known else-
where
Magellan Straits
A'^. canina Smitt
69-70
+
+
—
A^. jordani Thompson
71-72
+
+
—
Magellan Straits
A^. tessellata Richardson
72-73
+
+
—
Magellan Straits, and southern
Chile north to Chiloe
N. brevicaiida Lonnberg
74-75
—
+
—
Magellan Channels
N. giintheri Norman
75-76
+
+
—
Not yet known elsewhere
A'^. ramsayi Regan
76-80
+
+
—
Not yet known elsewhere
A'^. wiltoni Regan
80-81
—
+
—
Magellan Straits
A^. longipes Steindachner
81-82
—
+
—
Magellan Straits, southern Chile
A^. squamiceps Peters
82-83
—
+
—
Magellan Straits
N. sima Richardson
84-85
—
+
—
Magellan Straits
A'^. cormicola Richardson
85-87
—
+
—
Magellan Straits, and southern
Chile north to Chiloe, ? N.Z.*
N. elegans Giinther
87-88
+
+
—
Magellan Straits
N. tnacrocephala Giinther
88-90
—
+
—
Circumpolar sub-Antarctic
N. microlepidola Hutton
90-91
—
—
+
Circumpolar, but not at Kerguelen
Dissostichus eleginoides Smitt
91-92
+
—
—
Magellan Straits, Graham Land
Eleginops maclovitius (C. and V.)
92-94
—
+
—
Up to River Plate on the east, and to
northern Chile on west coast
Harpagiferidae
Harpagifer bispinis (Schneider)
94-95
+
+
—
Circumpolar northern Antarctic
Chaenichthyidae
Champsocephahis esox (Giinther)
95-96
+
+
"
Magellan Straits (near relatives
Antarctic)
A = taken in ' Trawl + accessory nets'; B = taken in 'Other gear' or by shore parties; C = not taken by the Expedition.
* See Norman (loc. cit.) for the doubtful records of this species in New Zealand.
In their regional distribution within our area, the group as a whole is a southern one. Of the
deeper water species only the dominant Notothenia ramsayi, with Cottoperca gohio and Notothenia
elegans were recorded in the northern region ; and Cottoperca was much more plentiful in the southern
region, while Notothenia elegans seemed to find its optimum in the intermediate region. N. guntheri
appeared twice in the intermediate region, but 49° S was its northern limit among our observations,
which was true also of Dissostichus and Champsocephahis, characteristically southern genera. Little
can be said concerning the probable northern limits of the littoral species, owing to unavoidable lack
of observations in Argentine territorial waters. Eleginops and Bovichtus are known to range far to
1 This supra-family grouping is convenient for purposes of this summary. It includes the families shown in the Table
with other exclusively Antarctic ones. Elsewhere in this report I have avoided using such groupings, as many of them
involve questions of classification that are still unsettled.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 343
the north, and W. F. Thompson took Notothenia jordani in Grande Bay. We took young A^. macro-
cephala in 48° 50' S in the open ocean, but large individuals were obtained by us only in southern
coastal waters. In New Zealand, however, adults of this species are known to range as far north as
the southern coasts of North Island. All the other shallow-water Nototheniiformes were taken (by us)
only in the southern region.
The Nototheniiformes are preeminently an Antarctic group, completely dominant over all other
fishes in such small areas of the vast oceans south of the Antarctic convergence as are sufficiently
shallow to support any coastal fish fauna. Their abundance and variety in the Patagonian region and
their tendency to be distributed mainly near its southern limits, are consistent with the view that
they have spread northwards, but we have seen that only Harpagifer and Dissostichus have species
common to both zones. It thus seems probable that the Patagonian fish fauna has been insulated
against invasion from the south for a very long time, and that the hydrological barrier of the Antarctic
convergence provides too big a contrast in environment for the majority of such fishes to overcome.
Quite early in these investigations, when attention was focused on large species (notably hake) that
migrate over long distances in a comparatively short space of time, we found it impossible to gain
much by the study of depth distribution because of the very slight gradient on the plain of the shelf.
This necessitated laborious calculations of the distance of each observation from the coast before we
could attempt to follow the movements of such fishes. With the Nototheniiformes it is quite otherwise.
Here we have a group consisting for the most part of fairly small bottom-living fish with limited powers
of movement, and the study of depth distribution has helped a great deal in our attempts to gain some
insight into their probable way of life.
The depth relations of the Patagonian species are summarized in Fig. 42, and the differences m the
mean depths recorded are given in Table 36, with data sufficient to determine their statistical signffi-
cance. The figure shows the effective mean depths and extreme ranges observed for all the species
except Dissostichus eleginoides, our few specimens of which were so widely dispersed as to demand
individual plotting. The distribution of all the species descending below the 50 m. level is also in-
dicated by the black polygons. The widths of these are proportional to the relative abundance of each
species within each 50 m. depth grouping. The littoral species are further indicated by stippling of
the rectangle covering the whole of the observed depth range of each.
It will be seen that the species can be divided into three main groups according to their depth
distributions :
I Deep-water species found mainly on the plain of the shelf and rarely beyond the shelf edge:
Cottoperca gobio, Notothenia guntheri, N. ramsayi, N. elegans, Dissostichus eleginoides and Champso-
cephaliis esox. ,
II ' First-slope' dwellers, rarely descending below 100 m., where the plain of the shelf may be
said to begin: Notothenia canina, N. jordani, N. tessellata, N. hngipes and perhaps Harpagifer hsprnis.
III Exclusively littoral species: Notothenia brevicauda, N. zviltoni, N. squamtceps, N. stma,
N. cornucola, N. macrocephala (adults) and Eleginops maclovinus {Bovichtus argentinus may fit m here).
Within the first two groupings it proved possible to recognize further distributional trends, either
regional or bathymetric, which serve to differentiate the species still further. As a result of this (always
excepting the dominant and ubiquitous Notothenia ramsayi) most of the species show a distinctive
distributional pattern that tends to minimize territorial overlapping between them. The possible
ecological ' advantage ' of this in lessening competition between species of similar size, may be one of
the factors that has led to the slight modifications in food requirements and general habits that it must
have entailed.
344
DISCOVERY REPORTS
1=
9
L
3
IN
t
p
II "
•
•
ii
•
•
« •
•
i
5 i
==
Pi
^
*
^
m
-
1^ V
NOTO-
THENIA
L0N6IPE5
>
NOTO-
THENIA
WILTONI
NOTO-
THENIA
RAM SAY 1
w
K-
r=
(
>
1^1 i
1
^
% H "^
^
rfi
s9
il^
H
ft-
m
9
W
DEPTH
GROUPS
IN
METRES
s
o
o
o
1
O
LH
(\i
1
O
o
o
in
o
in
hi
)
o
o
o
1
in
o
f
o
o
o
o
o
Ml
4-1
<u
Q
N
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
X0S3
sniDi(<i33os4uivi{j
siuiistq xafiSvixvfj
snuia,
-opvui S<f0UtS3]g
1^
o
k
CO
sapioutS
-ap smpusossiQ
N.^ t^.- 'O,- ^•.- Y^^ ^^ _^^ ^^ 1^ ^^ 1^3 iCfi+Cfi |Cn
+ ^ I ^ \
m
!£.ei) ^.M
_l_a! +c« +0! +0:1
7z
'I
2 +Cfi
7z
+ -cfi
■- +Cfi + C/3 +Cfi 4-Z +C/2 -l-Cfi
«| ?;.^ ^1
+
60
c?5
vjvtftfajoxjvm -/^
suvSip -jY
vjojmaoj -jY
»z«M 'jV
s433iiuvnbs -^
2 bo
+ cfi +cJ5 +cfi +cfi +
Z +^3 +C/3 I /C +ai ^ 1 /i +^ +
« .2f^ 00 .i
.en
Cfi
!.^ ^.Sf S.
m:3
c/:
C/2 I C/2 +C/2 +
I
I
biD ° 60
cfi
^.S' vo .i:
' 1^ i2".M ^
^(n +cfi +cfi +
Cfi
Z +ai +!/3
J S.23
I Z +c«
\0
o
+
7^
2.£P N.2f
2 60
CO +Cfi +Cfi +C« +r^ +CO +
c/: Z +cfi +Z
« 2 •
-^•^ "-.S" ^^
I ti '=£ Ml ^.ci ^.60 „ o
.cJ5 +cJ5 +c« +cr! +Z +cn +
"" 60 2" S^ (N-) O J^
be
'Z +Z.+CO Z
S3(flSuO] ■J\^
+ c«
o.w) °.W) °^,ae
+
.a^ ?).^ ^.2° "- "- • -7:
in +CC I cfi +cfi +c« I en
jq M "^ M ^.M f?,.M) jq.W)
I Cfi I Cfi I Cfi +Cfi I c«
mofjm '^
N
« bO
+ a5
+ cfi
g^bb^bb^o S"") » .2°
^'(n +cfi +Z +cfi +C/3
+ C/D
r^^ o| ^1 ".^ ^7
+ Z Z +Z +cfi +Z
^^
iiCvswvx -jY
uaiffunS 'jy
Dpnvjiiisxq -fq
V}VJpSS3f -^
!^ 60
f:;: bb^.M ^.M
0\.
Cfi I C/D I
2". 23 ^.a'
C£l
m
fo —
'I- bb '^
M
fn
I
■-'' CO I CO
CO I CO 1 CO
op
CO I
S.2= S._ .,.^
CO I CO
T)- 2^ ^.25 S .^
+ CO I CO I CO
:? 2> 4".2° ?.2° ?^.y=
I CO I CO I CO I CO
CO
:2.2>
+ CO
^2" -0.2°
_|.CO +CO
00 p) .
"^ bVi "^ bO
-CO
bo
+ 'co
~t-° S.2>
1 z +^
Z fZ
vO
2 6b
Z +«^ I
o
z
C;.23 ^.25 °.2f
CO I CO
.:? tb ^ M ^.2) S .2> ^.25 ^.2' S .2) ^.2' ^.2° - y"
+ CO I
I CO +CO +C0 I CO
CO
vo'.S" vd -XT' vd .s^ ■*." "o ._
I CO I CO I CO +CO I CO
luvpxoC -jV
vumvo viu3i^tofopj
OtqoS V3X3<j0}f0J
^.23'§.2=
+ CO I CO
S 2° ^.23 S.2» 0^.23 ^.2°
J.CO I CO 4-CO +CO I CO
1 60 ^ 6b ^ M ^.M ^ M ^.60
CO iCO |CO |CO+CO |CO
2 60
+ CO +CO I
CO
+ CO +(
00
-CO
CO
CO
00 .
+ CO
"=Jt> "^ 2°
CO
S.£P
"^.2° {0.2' ~
I CO I CO I
S.2°
CO
+ 03
^, bo
+ ^ I
^ <^.2°
CO I CO
I cio I CO I CO
Af/1^
vO
OS +
00 00
0
Ov
00
Ov
VO
Ov
00
VO
vO
b
o
00
ov
00
o
-*-
o
m
vo
ov
m
ov
34S
1.2= S.2>
I CO I CO
oJ.y
+ CO
-I 2.2=
+ Z +CO
•+K
-CO
VO
o
::;.2' S.sp 2.£P
I CO I CO I CO
>^^
s;.M
Z +CO
Tz
+ Z +CO
I
+ Z +CO
^.2= ^.2= vS.23
I CO I CO 4- CO
Z 4-Z +CO
CO I CO I CO
00
I CO I CO I CO
Ov.ir
+ CO
vo'.a' V0.2" ^.2=
I CO I CO +CO
?,.23 ^.2° S>.£f
I CO I CO +CO
J] bb J\ 60 <i^ M
I CO Y ^ +^
VO
I CO \ a
ov
00
00
Too
idely
persed
0
0
?.2
1.^
N
•a
vO
N
6
p
UC3UI aAI}D3JJ3
pUE ajqBJIBAE •SOJy[
vO
00,
tn
-0
u
0
0
6<)
u
«
a.
V
CO
1
§^
0
0
00 H <^
~^ 1- "2
"1
fi tI- 6 OvS 00 I
t^ T ov
a nS
00
,-k vD ,_
^a
<3
•S
<3
"<3
«
•0,
«
^
■|
0
-2
1
-0
a
a
S
2
'§
g'
^
^
^
^
^
^
'<
.3
o
a
s
<3 -2J
■Oh ^
a
S
c
?«
<3
S
to
•S
So
a
-S
53
16-2
346 DISCOVERY REPORTS
This feature is best demonstrated by the 'first-slope' dwellers. Thus while A^. canina and N.jordani
inhabit much the same area off the mainland coast in the southern region, the latter was consistently
found in slightly deeper water than the former. A^. tessellata, with a depth distribution overlapping
that of both the last-named species, was not found along the same coasts, but mostly in more exposed
positions where the 'first slope' was steeper (e.g. round the Falkland Islands, where the other two
species did not occur at all). A^. longipes, if it is indeed a distinct species, seems limited to the western
end of the Magellan Channels, while the others were found to the eastward, mostly along open coasts.
Lastly, Harpagifer bispinis might almost have been placed in group III, for it is common intertidally
at times, but we have taken it down to 95 m. in the Patagonian region ; and in the Antarctic Zone,
where it has a far more extensive distribution regionally, it would certainly be more accurately
described as a ' first-slope ' dweller.
The deep-water group consists mainly of larger fish, so that, partly owing to increased powers of
movement, there is considerably more overlapping between the species. Considering the Notothenia
spp. first, the two smallest, A'^. elegans and N. giintheri, seem to move but little and to remain on the
plain of the shelf throughout the year. A^. giintheri has its centre of distribution well to the south of
that of A'^. elegans. N. ramsayi, dominant throughout and the largest of the genus in the region, was
found to have a well-marked seasonal migration to deeper water beyond the shelf edge in winter.
Dissostichus eleginoides was too rare to be studied on these lines. It is the largest by far of all the sub-
Antarctic Nototheniiformes. Our large specimens were captured beyond the edge in very deep water,
but smaller ones were taken on the shelf, so it may move inshore to spawn. Cottoperca gobio and
Champsocephalus esox were mainly inhabitants of the plain of the shelf in the southern region, showing
considerable overlap with Notothenia ramsayi, but although a few Cottoperca were taken beyond the
shelf edge, neither of them seemed to show anything like the definite movement to deeper water of
Notothenia ramsayi. Moreover, both Cottoperca and Champsocephalus were frequently taken in lesser
depths than those to which Notothenia ramsayi normally penetrates— sometimes, indeed, in littoral
waters.
The exclusively littoral group naturally show almost complete territorial overlapping, but their size
differences may serve to lessen the competition among them. A^. brevicaiida, N. squamiceps, N. sima
and A^. cornucola are all very small species and no doubt compete for small invertebrate food ; but
A^. macrocephala and A^. wiltoni run to a fair size (30 cm. or more) as Patagonian Nototheniidae go! and
here, doubtless, ' the great ones eat up the lesser ones '. Eleginops maclovinus may run up to 5 lb. weight,
but its diet is unknown. It would be extremely interesting to see whether the convergent evolution
evident in its close superficial resemblance to true mullets, and emulation of their powers of ascent
mto shallows and fresh water, extends also to the adoption of a vegetarian diet.
In the matter of size generally, the 'first-slope' dwellers provide a majority of the intermediates
between the extremes shown by the smallest littoral species and the deep-water group. Thus, with
certain obvious exceptions, the division by depth distribution is broadly reflected in a corresponding
gradation in size, just as we have so often found within the limits of individual species.
It is not yet possible to say much concerning the growth of Patagonian Nototheniiformes. Length
frequencies of Cottoperca gobio suggest approximately 7 cm. annual increments during the main
growmg period of that species, and in Notothenia ramsayi it seems fairly certain that increments of the
order of 10, 8 and 6 cm. accrue during the first three years of life, With perhaps a 4 cm. addition in the
fourth year. It is probable that in this species, as in hake, the males grow much more slowly than the
females once maturity is reached. Among species of intermediate size, 7 (? + ), 5 and 4 cm. increments
durmg the first three years of life are suggested for Notothenia jordani. There is a possibility that
further work might show Champsocephalus esox to be a fish of exceptionally rapid growth.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 347
Most of the shallow-water Nototheniidae are abundant intertidally in summer, moving to slightly
deeper water in winter. Thus they parallel the movements of the offshore species Notothenia ramsayi
on to the shelf in summer and over the edge in winter. The larger, deeper- water species show increased
tendency towards schooling in autumn, and it may be that some of them spawn then. There is a little
direct evidence as to spawning time among the small inshore species, but this shows all possible
ranges from spring spawning in N. sima to autumn spawning in N. squamiceps, with the possibility
of an extended breeding season in N. cornucola. Among larger inshore species, A', wiltoni almost
certainly spawns in the autumn at the Falkland Islands (Bennett) but in New Zealand A^ macrocephala,
the principal species common to both the Antipodean and Patagonian regions, is said by PhiUipps
(1921, p. 123) to spawn in September (spring).
GEMPYLIDAE
Thyrsites atun (Euphrasen). During the trawling surveys forty-five adult specimens of this large fish
were captured. They occurred at six stations only, and twenty-nine of them at a single one of these
(WS847B). All these captures were made fairly close in to the mainland coast in moderately shallow
water, in late summer and autumn, and all but one of them in the southern region:
WS96 I WS833 4
WS812I I WS847A 6
WS812II 4 WS847B 29
The lengths of these fish were very consistent, with the means 97-3 cm. for males (range 87-104 cm.)
and 97-8 cm. for females (range 89-112 cm.). Females were more numerous than males, but the
numbers are too small to support the suggestion that this is a constant feature, as it is in so many other
fishes Similarly, no significance can be attached to the slightly greater lengths attained by the females.
At two stations Thyrsites was observed to have been feeding heavily upon Clupea fuegensis and
Thysanopsetta naresi. .
These fish were in prime condition: the values for K averaged 0-380, a high figure for a species of
such slender proportions. The extreme range of K observed was o-320-o-449- We should expect high
values for K in autumn if the stock of Thyrsites off the east coast of Southern America spawn at the
end of winter or in early spring, as the species is known to do off New Zealand (Philhpps and Hodg-
kinson 1922, p. 94) and oflF South Africa (Gilchrist, 1916, p. 8).
After the conclusion of the third trawling survey, when the 'William Scoresby' was on passage to
Europe, young Thyrsttes were captured in the young fish trawl at St. WS881. This station was worked
in the area to the" north-east of the Falkland Islands, which is periodically influenced by mixing of
warmer water from the Brazil current to the north with sub- Antarctic water.
In the regions where Thyrsites is abundant (off South-West Africa and south-eastern Australia)
nearly all are caught by trolling, jigging or other forms of line fishing. Very few have been taken m
trawls Our records cannot therefore be considered as conclusive evidence of its distribution. Never-
theless they show a pattern in time and space so consistent with what is known of the habits of the
species elsewhere, that it seems worth while to draw certain tentative conclusions from them.
It is believed that the part of the Patagonian Continental Shelf with which we are chiefly concerned
provides a habitat too cold for Thyrsites throughout most of the year, and that such adults as were
captured in the warmer inshore waters in autumn indicated the probable southern limits of the feeding
migration of the species. The distribution of Thyrsites off South Africa and New Zealand shows that
it favours warmer waters than the hakes (Merluccius) of these locaUties; but although its tolerance of
348 DISCOVERY REPORTS
cold seems less than that of Merluccius, its tolerance of heat seems similarly restricted, for Thyrsites
does not penetrate so far up the east coast of South Africa, into the region of the warm Mozambique
current, as does Merluccius capensis. Thus Thyrsites seems limited to cool southern subtropical (rarely
sub-Antarctic) waters where the surface temperatures range from about lo to about 20° C. ; whereas
species of Merluccius range from areas with surface temperatures below 8° C. to the tropical con-
vergences (temperature about 23° C. at the surface). The difference in thermal tolerance may be more
precisely expressed by the statement that the optimal range for Merluccius is centred lower than that
for Thyrsites, but the total range of Merluccius is the wider. Thyrsites is much more exclusively pelagic
in habit than are any of the Merluccius spp., which may be regarded as demersal fishes during the
daylight hours, and whose distribution will thus be affected by the cooler subsurface temperatures to
a much greater extent. For this reason I believe that Thyrsites will only be found to work southwards
into our area close to the mainland coast, in the warmer counter-current of ' old shelf water', or right
offshore, where the influence of the warm Brazil current sometimes extends to about 49° S. I do not
think that the species would normally cross the colder portions of the Falkland current, and such few
stragglers as occasionally reach the Falkland Islands themselves (Norman, 1937, p. 96) have probably
come from the north-east.
T. atun is most excellent eating, with exceptionally firm white flesh of good flavour and (in due
season) rather high fat content. In November 1933, when the ' Discovery II ' was outward bound on
her third commission, we paid a brief visit to Tristan da Cunha, and there secured a plentiful supply
of these fine fish by somewhat novel methods. While lying at anchor in about 7 fm., near the edge of
the kelp off the main landing place, the fish were observed ' hovering ' round the gangway lights at
night. Fishing for 'five-fingers', etc., had been proceeding all day, and the school of Thyrsites were
doubtless (primarily) attracted by some of the resulting offal. A cargo-cluster was lowered over the
side (the night being calm) and a few volunteers, fishing with variously improvised jigs (I found an
artificial squid very killing), landed close upon three-quarters of a ton of the fish in about 3 hr. The
fish were placed in the ship's cold-store as soon as we finished cleaning them next morning, and
provided at least one course per day for all who cared for them until we reached New Zealand (and a
copious supply of mullet) some z\ months later. It is thus evident that when chilled soon after capture
the keeping qualities of Thyrsites are excellent. It is preserved by smoking in Australia, and is both
salted and smoked in South Africa. Very large quantities are eaten fresh in both these countries.
Since Thyrsites is one of the most valuable food fishes in the southern hemisphere, it would certainly
merit further attention if any commercial fishery is developed near the area we surveyed, and a
discussion of its importance in other regions therefore seems worth while. Any captured incidentally
in trawling for more plentiful species would augment the value of the catch, and if the suggestion of
an autumnal invasion of the warmer inshore waters should prove a constant feature, it might even be
worth while to try trolling for it.
In South Africa T. atun (locally ' Snoek') is a very important fish. Returns for three years 1929-32
(von Bonde, 1934) show it second only to Merluccius capensis (stockfish, hake) in weight of landings
and in their value. The quantities landed fluctuate much more violently than those of most of the other
important fishes. This is partly explained by the fact that the majority of the snoek are taken by small
line-fishing vessels whose activities are much more subject to the vagaries of the weather than are
those of the trawlers, which operate from the best harbours, and take the majority of the other fishes.
Of the fishes captured by the line-boats Thyrsites is by far the most important. Figures given for two
recent years (Director of Fisheries, 1938, 1939) show that Thyrsites yielded 57 and 35% by weight,
52 and 46% by value, of the total landings by vessels other than trawlers at the Cape. As these figures
include the valuable crawfish landings the importance of the snoek is even greater than it at first
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 349
appears. 1 The annual catch of snoek at the Cape has varied from some four million to over nine
million lb. weight, and between about ^(^23,000 and ^90,000 in value, over a long period between the
two wars. The price has shown the usual tendency to inverse relationship with supply, but has
maintained a slight superiority over the average for all fish in each individual year for which data are
available.
Some of the main features of the biology of T. atiin in South African seas are succinctly dealt with
by Gilchrist (1914, 1916). They are especially abundant in the colder waters off the south-west coast
of Africa in late summer and autumn (the peak usually in May). At these times they feed heavily,
especially on 'sardines' (Clupea sagax) when available, and become notably fat. They are in full roe
towards the end of winter and are found with ripe eggs, and in poor condition, in September (early
spring). The seasonal variation in condition is evidently most marked, for Gilchrist has told us how
the fishermen used to sign-on for the 'poor snoek' season on 17 September and how they had learnt
to look for the first appearance of the ' fat snoek ' about the middle of December. From then until
winter (June) is the best period for the fishery. The relative scarcity of Thyrsites in winter and early
spring (the spawning period) is attested by returns from all the countries where it is regularly fished.
This is doubtless partly due to a slackening of feeding intensity, well known among most fishes when
spawning, that will obviously tend to reduce catches of line-caught species ; but some definite migration
for spawning purposes may also be involved.
In Australia T. otiin (locally 'barracouta') is nearly as important as it is in South Africa. The in-
formation concerning it has recently been summarized (anonymously) in Fisheries Newsletter (1944,
vol. Ill, part 5, p. 2) and there have been previous interesting references to it in that most stimulating
publication. Thyrsites ranks third in importance among the individual species of Australian food-fishes,
the annual catch averaging some five million lb. weight over the period 1930-44. Large fluctuations
in supply are experienced with consequent variations in price. Normally the fish fetch from 8^. to
155. per box, but in 1941 the supply fell until the price reached an ' all-time high' of 35. lod. per fish
[Fisheries Newsletter, 1941, vol. i, part i, p. 10). Although there had been a steady decline since 1938
there was not thought to be any immediate fear of over-exploitation. Previous periodic scarcity of the
species had been known, and there is some hint that it may recur with a seven or nine year cycle. It
was also observed that the fish are much more difficult to catch, by the prevailing trolling or jig-stick
methods, when their natural food is abundant. This was so in 1941, when it was observed that the
boats frequently worked through large schools of barracouta, milling among the clupeoids with which
they distend themselves, without catching any. This is thought to have been a subsidiary cause of the
scarcity. If the tentative suggestion of a seven or nine year cycle of abundance in the Australian
stocks of Thyrsites is substantiated, the 'couta' boats should be obtaining peak catches in 1945-6
and 1947-8.
Thyrsites is limited to the most southerly of Australian seas, about 90% of the catch being taken in
Tasmania, and most of the remainder in Victoria. In New South Wales it is rare, and farther north
it is apparently unknown. I have not been able to find any account of the occurrence of Thyrsites in
south-western Australia, but it is reasonably certain that it is found there, for Australian writers add
the isolated islands of St Paul in the southern Indian Ocean to the list of localities from which it has
been recorded.^
In Victoria trolling with a crude development of the native lure, from auxiliary sailing boats of
1 In the same two years the total trawler landings were from nearly three times to nearly four times as great as
those of the other vessels by weight, and from more than three times to nearly seven times as great in value.
2 I have not yet traced the origin of this record, hut have no doubts as to its validity, for I believe that the
distribution of Thyrsites atun is continuous round the world in southern subtropical seas. Local stocks may, however, become
distinguishable when more intensively studied, as they will need to be in the future.
3S0 • DISCOVERY REPORTS
20-45 ft-, is the principal method of capturing Thyr sites. Fishing is chiefly performed under sail at
speeds not exceeding 4 knots, after a school of the fish have been located. In Tasmania similar methods
are employed, but rank second in importance to the 'jig-stick', by which the fish are swung directly
inboard on to a fore-and-aft chute, after snatching at a lure trailed on a very short trace attached
directly to the end of a flexible pole. This interesting method of fishing bears an obvious relation to
trolling, and also to the methods employed in hooking tunny and albacore off the Californian coast.
(Most of the operatives in this latter fishery were Japanese.) In Fisheries Newsletter {ig^z, vol. i, part 2,
p. I ) we are told that netting was being supplied to Port Fairy with the object of testing the possibilities of
a gill-net method of capturing barracouta. The results of this experiment should be most instructive.
Although the Australian fishery has hitherto operated exclusively upon the surface schools of Thyrsites,
the occasional occurrence of the species in otter-trawlings, in areas far removed from those fished at
present, has been noted. These catches were made in 60-70 fm., i.e. nearly twice the depths in which
we occasionally trawled the species off Patagonia.
Spawning of Thyrsites may take place to the north and east of Tasmania. This is certainly a nursery
ground for young fry up to 3 in. in length, which have been abundantly found in stomachs of various
large predaceous fishes (including adult Thyrsites) taken in this area. Young Thyrsites of 10-12 in.
(.? I or Il-group) have also been observed not far away.
In Australian waters Thyrsites are often heavily infested with muscle-worms (? nematodes), and
these may possibly be one cause of the emaciation that leads to afflicted fish being described as * axe-
handles '. I should be inclined to suspect the extreme seasonal fluctuation in condition, so well known
in South Africa, as the main factor, for repeated attempts to correlate parasitic infection with loss of
condition in numerous species of fishes have broken down when strict tests are applied. A disease of
Thyrsites known as 'milkiness', due to protozoan infection, presents a more serious problem, and is
now causing some concern in Australia. In New Zealand it is stated that Thyrsites occurring in northern
waters are much subject to disease (Phillipps, i92i,p. 118). Now the fish are less common in the north,
but the whole stock will tend to move northwards in winter, and, moreover, they are believed to
spawn even earlier off New Zealand than they do off South Africa (August rather than September:
Phillips and Hodgkinson, 1922, p. 94). Hence it is probable that natural seasonal loss of condition is
at least partly responsible for reports of disease.
In the New Zealand fisheries ' barracouta ' are not so important as in Australia and South Africa,
doubtless owing to the good supply oi Jordanidia solandri (' southern kingfish', ' hake ' !). Comparable
figures for the whole country are not available, but separate returns for Thyrsites atun have been made
at a few of the individual ports. Wellington shows the highest of these returns, and from the Fisheries
Reports of the Marine Department it can be seen that Thyrsites formed from 2 to 4% of the total
catch here over several years between 1931 and 1938. During the best years this figure represents over
1000 cwt.
The common names applied to Thyrsites are used for other fishes so promiscuously as to lead to the
possibility of endless confusion to anyone not personally acquainted with these other fishes as well,
so I have thought it best to treat this matter in detail :
T. atun (Euphrasen), a gempylid, has three main vernacular names in different parts of its wide
range, that extends round the world in the cooler parts of the southern sub-Tropical Zone. These
are snoek (from the Dutch word first applied to the European fresh-water pike, Esox lucius) in South
Africa excepting Natal; 'barracouta' (from barracuda, first applied to the marine Sphyraenidae in
Europe and the West Indies) in Australia and New Zealand ; and sierra (from Spanish, lit. a saw,
applied chiefly to ' king- ' or ' spanish-mackerels ' elsewhere) in Chile. These names are all descriptive:
snoek relating to the superficial resemblance between body-form and pointed, formidably toothed,
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 351
jaws of Esox and those of Thyrsites ; barracouta relating to a similar resemblance between the marine
Sphyraenidae and Thyrsites; and sierra presumably referring to the rows of spiny finlets between
the posterior median fins and the caudal of Thyrsites.
The first objection to widespread use of these names is already clear: they are applied to other
fishes elsewhere, many of which bear no close relationship to Thyrsites. But there are two other equally
important objections. First, some of the other fishes showing such overlapping of common names
may inhabit the same areas as Thyrsites; this is especially true of Australian seas. Secondly, these
other fish have themselves been given a further range of vernacular names, which again overlap, and
increase the possibility of confusion twice confounded with yet other fishes (or groups of fish).
In Natal ' snoek ' refers to Scomberomorm commersoni (Cuvier and Valenciennes) (Scombridae :
Gilchrist, 1914, p. 118) which is the 'narrow-barred spanish-mackerel' of Australia, and is also known
as 'kingfish', 'king-mackerel' and even (in western Australia) as 'albacore' (Munro, 1943, p- 74).
In Puerto Rico (West Indies) ' snook ' refers to Centropomus parallelus Poey, the ' robalo ' (Jarvis,
1932, p. 4). The Centropomidae are a percoid family closely allied to the Serranidae. In Australia,
on the other hand, 'snook' sometimes refers to the 'short-finned pike', Sphyraena novaehoUatidiae
Giinther, which a European or an American would call a 'true' barracuda (Waite, 1921, p. 85).
Barracuda or ' barracouta ' refer to Sphyraenidae in Europe, the West Indies, North America and
South Africa, and it was to this group of fishes that the name was first applied. But this group is also
represented in Australia and New Zealand and other haunts of Thyrsites, where they are known as
'pikes' or 'sea-pikes', with or without such specific qualifications as 'short-finned' { = Sphyraena
novaehoUatidiae Giinther (Waite, 1899, p. 132)), and also, more rarely, as 'snook'. According to
Evermann and Radcliffe (1917, p. 51) Sphyraena idiastes Heller and Snodgrass is known as 'aguja'
(Spanish, lit. a needle or bodkin) in Peru. This name is applied to a very wide variety of slender fishes,
including 'pipe-fishes', in various parts of the world.
The original ichthyological connotation of 'sierra' in Spain seems to have been Pristis sp., a 'saw-
fish', one of the elasmobranchs ! The name is also applied to various Scomberomorns spp. ('king-' or
' Spanish-mackerels', Scombridae) in Spanish-speaking parts of the Americas north of the range of
Thyrsites.
Let us consider only two examples of the secondary complications that can arise among the common
names applied to fishes which share the main vernacular appellations of Thyrsites :
The 'king-' or 'spanish-mackerels', Scomberomorus spp., Scombridae, are often referred to as
' kingfish '. This name is also extensively used for certain Gempylidae other than Thyrsites, notably for
Jordanidia (Rexea) solandri (Cuvier and Valenciennes) in Tasmania and New Zealand. This fish is in
its turn sometimes called 'hake' in New Zealand, even though there is a 'true' hake, Merhiccius
australis (Hutton), on the spot. ' Kingfish ' is also (rarely) applied to the ' king- whiting ' {Menticirrhus
spp., Sciaenidae) in parts of the United States.
'Pike' is applied in the Falkland Islands to Champsocephalus esox (Chaenichthyidae, Notothenii-
formes), and in Australia and New Zealand to the Sphyraenidae. In Australia also ' long-finned pike'
refers to Dinolestes leivini Griffith (Waite, 192 1, p. 99). This is one of the Apogonidae, a family equally
remote from the Sphyraenidae and the Gempylidae.
SCOMBRIDAE
Gasterochisma melampus Richardson. Norman (1937, p. 97) has shown that although we did not
obtain any specimens of this interesting oceanic species during the trawling surveys, there is evidence
of three specimens reaching the Falkland Islands. Some portions of a damaged skeleton from West
Point Island were secured by Dr J. E. Hamilton. Norman adds an interesting note recording his
D XXIII
352 DISCOVERY REPORTS
opinion that one species hitherto considered as distinct, and two others previously described in
different genera, are but growth stages of G. melampus, and quotes close parallels in yet other genera.
Such difficulties are bound to occur with rare species known from very few specimens, and can only
be rectified as larger series are collected.
ZOARCIDAE
Ophthalmolyciis macrops (Giinther). We took no specimens of this species, which is known only from
the holotype obtained by H.M.S. 'Challenger', in the Magellan Strait.
Iluocoetes fimbriattis Jenyns. This was the commonest member of the family captured during the
trawling surveys. It was found to attain considerable size (more than 40 cm. long), and was found
mostly in moderately deep water near the outer margin of the shelf, and more rarely beyond the shelf
edge, down to the greatest depths fished. It was found in all three regions within the area, with
perhaps a tendency to be most numerous in the intermediate region :
WS71
I
WS244
2
WS812I
I
WS92
I
WS246
I
WS821
I
WS98
I
WS76S
I
WS825
I
WS99
I
WS784
6
WS85S
I
WS210
5
WS792B
I
WS2I3
2
WS795
I
51
4 (in OTL)
WS2I4
I
WS801
I
WS856
I (in BTS)
WS2I6
5
WS809
I
WS869
I (in BTS)
WS2I8
I
WS811II
I
WS82g
I (in NR)
The spotted brown and white colour pattern so well shown by E. R. Gunther's sketch (Norman,
1937, pi. i, fig. 4) suggests concealment value among the bracken-like forests of coralline hydroids
and Polyzoa frequented by the species. The sketch also shows how in its general appearance, the beast
without background is extraordinarily reminiscent of a wet hen. I. fimbriatus is known to range as far
as southern Chile, outside our area.
Iluocoetes elongatus (Smitt). This species was taken three times only, once in the trawl and twice
with 'Other gear', in shallow water ('first slope') in the southern region. The colour pattern is
exceptionally variable but is always barred rather than spotted, and the species may always be dis-
tinguished from the last named by the entire absence of scales (/. fimbriatus has small scales embedded
m the skin): WS834 6 WS835 14 (in BTS) WS749 17 (in NR)
Austrolycus depressiceps Regan. No specimens of this littoral southern species were secured during
trawling operations, but series for Norman's report were readily obtained from East Falkland
(Mr A. G. Bennett) and West Falkland (Dr J. E. Hamilton). The species extends to the Chonos
Archipelago (between 46 and 44° S) on the west coast of southern Chile, outside our area. One of
Bennett's specimens was 48 cm. long, and he tells us that the species has been known to attain a
weight of 3 lb. at the Falkland Islands, where they are sometimes known as ' rock-eels '.
New Island 6 (J. E. H. coll.) Port Stanley 13 (A. G. B. coll.)
Austrolycus laticinctus (Berg). We obtained four small specimens of this species in the rectangular
net at a single station in the southern region. From Norman's account (1937, p. 104) it appears that
the synonymy has been much confused, and that it may be found to range much farther up the main-
land coast than most of the Zoarcids with which we have to deal :
WS749 4(inNR)
Phiicocoetes latitans Jenyns. A few specimens of this tiny species were obtained at two southern
stations in shallow water, and others were collected by Bennett from kelp holdfasts in Stanley harbour.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 353
This appears to be the normal habitat of the species (Norman, 1937, p. 105), and the brown and yellow
coloration would there have 'concealment value'. Ripe ova were present in two specimens taken at
St. WS 847 in February ; this suggests autumn spawning, and shows that the species really is very
small, for none more than 14 cm. long was taken, and yet two of these were mature:
WS847B 4 WSy4g 2 (in NR) Port Stanley 4 (A. G. B. coll)
Crossostomus chilensis (Regan). We got no specimens of this species which, despite its name, is
known only from the holotype taken off the east coast of Tierra del Fuego.
Crossostomus fmciatiis (Lonnberg) is only known from the holotype, taken at the Falkland Islands.
Norman (1937, p. 106) thought that it may really have been a young example of Austrolycus de-
pressiceps.
Pogonolycus elegans Norman. Five examples of this new species were obtained at four stations
ranging in depth from the ' first slope' to beyond the shelf edge. Three of these were in the southern
region, the other in the intermediate region. The colour sketch by E. R. Gunther (Norman, 1937,
pi. i, fig. 3) shows a pattern that probably has concealment value among the heavy sessile invertebrate
bottom fauna with which it was found. It was especially noted that its appearance coincided with
large catches of Cephalodiscus colonies :
WS97 I WS749 I (in NR)
WS246 I WS878 2(inNR)
Platea insignis Steindachner. No specimens of this species were taken in the trawl, but a series was
obtained with ' Other gear' from two southern stations worked in shallow water. The colour scheme
is of the basic type so characteristic of most of the group in our area, consisting of more or less regularly
disposed bars, blotches or spots, contrasting in shade with the ground colour. In this species it is
the ground colour which is the paler :
WS749 6 WS835 8 (in BTS, haul B)
Maynea patagonica Cunningham. No specimens of this small, rare, probably coastal species were
obtained by us, but it is known to occur within the trawling survey area (Norman, 1937, p. 108).
Maynea brevis Norman. Four specimens of this new species were taken in the trawl, two m the
intermediate region and two farther south. Three were in moderately deep water near the edge of the
shelf, and one in deeper water beyond the edge. This last was the largest specimen (length 16 cm.).
Norman notes that it may eventually prove necessary to place this species in a distinct genus, and that
its dentition approaches that found in the genus Melanostigma:
WS216 I WS784 I
WS244 I WS825 I
Melanostigma gelatinosum Gunther. This species is known only from the unique holotype taken in
the Magellan Strait; it may yet be found more widely distributed in our area.
Melanostigma microphthalmus Norman. Two specimens of this new species were trawled, both m
deep water south of the Falkland Islands :
WS246 I WS248 I
SUMMARY OF OBSERVATIONS ON ZOARCIDAE
The eel-like shape and small size of most of the members of this family render it certain that they
cannot be adequately sampled by ordinary trawling. Moreover, many of the species are littoral in
habit while those dwelling at greater depths seem to favour very rough ground, with the dense sessile
17-2
354 DISCOVERY REPORTS
fauna of corraline hydroids and Polyzoa that E. R. Gunther aptly likened to bracken. In such con-
ditions all types of collecting gear have their efficiency much reduced. Nevertheless, I believe that
the group really are comparatively scarce and unimportant ecologically, as our scanty collections
would seem to imply, for a very comprehensive series of hauls with several types of ' Other gear ' was
carried out, and the ' William Scoresby ' was more successful in capturing Zoarcidae than any previous
expedition to the area had been. Of the total of fourteen species now known from the region, specimens
of nine were obtained ; and of the five missed, three are known only from their unique holotypes,
while a fourth is still a doubtful species.
Two of the species living in deeper water, Maynea brevis and Melanostigma microphthalmus, were
taken only in the ' Trawl + accessory nets'. Three probably littoral species were obtained only with
'Other gear', Austrolycus depressiceps, A. laticinctiis and Platea insignis. The remainder, taken with
both main types of gear, showed a wider depth distribution, the deeper dwelling larger species
(notably Iluocoetes fimbriatus, the commonest in our collections) being relatively more numerous in
the trawl as one would expect. Only /. fimbriatus and possibly Austrolycus laticinctus extended to the
northern region of our area, and the group as a whole were definitely most numerous to the south.
The Patagonian Zoarcidae seem broadly divisible into shallow-water or littoral species, and deep-
water species, thus :
Shallow- water or littoral species: Iluocoetes elongatus, Austrolycus depressiceps, A. laticinctus, Phuco-
coetes latitans and Platea insignis.
Deep-water species : Iluocoetes fimbriatus, Pogonolycus elegans, Maynea brevis and Melanostigma
microphthalmus.
Of the latter only Pogonolycus elegans has shown a single small specimen in shallow water, while
most were taken quite deep down. It is also noteworthy that the deep-water species showed a strong
tendency to occur on the deepest portions of the shelf near its south-eastern boundary (and beyond).
They were not found at the slightly lesser depths of the level plain of the shelf that covers such a large
proportion of the area surveyed. It is not possible to consider this depth distribution in detail on the
basis of such small numbers of zoarcids as were obtained, but there is some indication that it will
eventually be found to be correlated with the distribution of certain types of sessile benthic fauna. The
markedly ' patterned ' colour schemes so prevalent throughout the group, with bars and stripes, or
spots and blotches of contrasted shades, and often with brown and yellow or white tints, even upon
the upper surface of the body, strongly suggest that camouflage is a necessity among them ; and just
as the shallow- water species (e.g. Phucocoetes latitatis) are known to frequent kelp, so is it probable
that the deep-water species find their optimum on the rough ground with bracken-like fauna of
coralline hydroids and Polyzoa, known to prevail over some of the deeper parts of the shelf.
The large variety of species in this group is one of the peculiar features of the Patagonian fish fauna,
as already explained, but it does not seem likely that they are sufficiently abundant to play an important
part in the ecology of the region.
LYCODAPODIDAE
Norman (1937, p. 1 10) has pointed out that it is not yet certain whether this family can be maintained
as distinct from the closely allied Zoarcidae.
Lycodapus australis Norman. We obtained four specimens of this new species at a single haul of the
rectangular net at St. WS748, in one of the deepest parts of the Magellan Strait. Its occurrence is
of particular interest, for all previously known members of the genus hail from the Pacific coast of
North America, and we have already seen that there are other resemblances between the fish fauna of
that region and that found off Patagonia.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 355
OPHIDIIDAE
Genypterus blacodes (Schneider). This species is also found in New Zealand and Tasmania, where it
is known as ' ling', and in southern Australia where it is sometimes called ' rockling'. In New Zealand
(and to a lesser extent in Tasmania) it is esteemed for the table, and considerable quantities are
marketed. It is too rare off southern Australia to attract much attention from professional fishermen.
We learn from Norman(i937, p. i i2)that off Patagonia G. blacodes got^hy tht Spanish name ' abadejo'.
In Spain this word is often loosely applied to cod, but dictionaries state that it is more strictly applicable
to pollack. Two allied species of Genypterus known only from the west coast of South America are
given the names ' congrio. . . ' with suitable adjectival qualifications. All the species of Genypterus are
very similar in form, bearing a strong superficial resemblance to a true conger, so that this vernacular
distinction of G. blacodes in South America, where it is known from both east and west coasts, argues
close observation on the part of the local fishermen. r> lz j / f
A very closely allied form, G. capensis, may yet prove specifically identical with G. blacodes (ct.
Norman, 1937, p. 113). G. capensis, known as 'king-klip', is an important fish m South Africa
Most of them are obtained by trawling in rather deep water, and aUhough they form only about 2 /o
by weight of the total landings, they fetch prices considerably above the average. Thus they repre-
sented 2-o-4-5% of the value of the fish landed during three years 1929-32. and the sums realized at
first sale ranged from ^C 11,000 to just over £20,000 (S.A.). Evidently the edible qualities of the genus
are appreciated wherever they are to be found, but they cannot be a cheap fish to catch, for they are
not much given to shoaling. Numerous trawling records off South Africa (von Bonde, 1934, pp. 42-
63) show 'king-klip' in consistently small numbers where they occurred at all, though large catches
of shoaling species like Merluccius capensis were being made.
In the area of the trawling surveys Genypterns blacodes was taken in very small numbers throughout
the year but the data are far too scanty to permit of any detailed consideration of the bionomics of the
species The distributional data are, however, very interesting and sufficient to suggest two main
trends of movement, supporting the view that most of the G. blacodes taken in the area are seasonal
southern stragglers from a stock inhabiting warmer waters, farther to the north. The records are:
WS78 T, WS783 I WS811II 3
WS7Q T. WS785A I WS812II I
WSos I WS789 3 WS816 I
WSq8 I WS792A I WS817A I
WS108 I WS792B I WS819B I
WS214 2 WS794 3 WS^4S I
WS216 S WS795 6 seen to escape
WS217 4 WS797B I ^SSjo 1
WS218 13 WS809A I WS875 2
WS772 I WS8IO 2 , TU\
WS773 4 WS8nI I WS586 3(onLH)
•^G,!jj(, I Connor inlet i (on LH)
It can be seen that apart from four specimens caught on hand-lines in the western channels, outside
the area of the trawling survey, all the specimens were taken in the trawl. . r ,t.
In Fig 43 the records are shown on three seasonal charts with a schematic representation of the
trends of movement they are believed to indicate. In winter and spring (a) the species seerned to be
confined to a small area of deep water near the edge of the shelf, and near the northern boundary
assigned to our intermediate region. In summer (b) the records suggest that these fish tended to
move into slightly shallower water, and also southwards. At the same time it seemed that other
Genypterus blacodes invaded the area from the north, working south along the 80 m. 1^-e m^he
relaUvely warm inshore counter-current. Where this current peters out, in the northern part of our
3S6
DISCOVERY REPORTS
r
'10
O
€%""-■■■■
s
c
o
3
Q
M
E
O
O
u
O)
c
■t.r^.
^ c
in (J
QJ (- L.
IT -?
bo
c
3
bo
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
357
c
o
CL.
a.
<
I
^
O
3
.0
U'
,v>Q°?
O.'^/^'^'^-
if-p-t^i^-
o
■'"6
O./
;0\
a
B
3
*^
3
•«
p
I
3
CiO
358 DISCOVERY REPORTS
southern region, there was in late summer a strong indication of offshore movement, that continued
into autumn (c) as one would expect if Genypterus moves offshore in winter like so many other
demersal fishes.
The measurements — too few for detailed consideration — show extremes of size in the north, while
individuals caught in the south were more uniform. Possibly the southward movement down the
coast and out again is mainly confined to adolescent individuals. Really small young individuals were
taken at one northern station only.
The interesting point that seems fairly certain from the distribution observed is that Genypterus
seemed to avoid the central plain of the shelf, being found mainly along the lines of steepest gradient,
i.e. the ' first slope ' inshore, and the edge of the shelf to the north. Our only captures of the species
on the central plain were near its southern limits in autumn, when those that have worked down the
coast are believed to be seeking deeper water. This they seemed to succeed in reaching in the deeper
southern portion of the Falkland trough, and since they have not been seen at any of the numerous
stations worked in the immediate neighbourhood of the Falkland Islands I conclude that if they
further succeed in regaining a more normal, northern habitat, it is by way of the Falkland trough
that they do so, as indicated by the dotted portion of the line on the diagram (Fig. 43 d).
BROTULIDAE
Cataetyx messieri (Giinther). Two small specimens of this rare fish were taken in deep water beyond
the edge of the shelf, one in the intermediate region and one farther south. It has also been recorded
from Messier Channel, Chile, and South Africa. A specimen 63 cm. long has been taken in very deep
water off Cape Point. E. R. Gunther considered that the normal habitat of the species lies below the
depth hmit of ordinary trawling :
WS248 I WS773 I
CENTROLOPHIDAE
Seriolella porosa Guichenot. This species was only captured at a single northern inshore station,
WS853 (8), worked early in autumn. The French found it common in Orange Bay, near the eastern
end of the Magellan Strait in 1883 (Vaillant, 1888, p. C30), and it might therefore be expected to
occur throughout the length of the shelf. Endeavouring to explain our lack of evidence of such
extended distribution, E. R. Gunther noted that its shape and colouring suggest fast swimming near
the surface, which might help to account for its being missed by the trawl. Alternatively, it may be
a strictly coastal species rarely moving offshore so far as the main trawling grounds.
The species is also known from New Zealand, and from Tasmania, where Neptonemus dobiila
Gunther was the synonym used by Johnston (1891), and Seriolella douhla^ Gunther the synonym
used by Lord (1923, p. 66). These writers tell us that it is known as 'trevally' or ' mackerel-trevally '
in Tasmania. The closely allied Seriolella brama Gunther, the 'snotgall' or ' snotgall-trevally ' of
Tasmania, is also sold under the name of 'trevally' in Christchurch and Dunedin, New Zealand, but
in that country the name is more usually (and correctly) applied to Caranx platessa Cuvier and
Valenciennes. It is also noteworthy that in Tasmania a true carangid (C. georgianus, Cuvier and
Valenciennes) is called ' silver-trevally ' (Lord, 1923, p. 67), while the usual New Zealand name for
Seriolella is 'silverfish', which can of course lead to immediate confusion with atherines in that and
most other countries,
1 A misprint.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 359
The etymology of the name ' trevally ' is exceedingly interesting. It appears to be derived from the
Portuguese cavalha, a horse or horse-mackerel (spanish caballa), names which are still applied to
Caranx in their countries of origin. In the Gulf of Mexico and eastern Florida this name applies to
Caranx hippos (Linnaeus) but has become corrupted to 'cavally', and even to 'Horse crevalle'^ in
South Carolina. As a further complication on the ichthyological side, ' crevalle ' is sometimes applied
to Trachynotus carolimis (Linnaeus), although this species is widely known as the ' common pompano'
in the eastern states. I suggest that 'crevalle' became readily corrupted to 'trevally' and, as applied
somewhat loosely to carangid or closely allied fishes, reached the southern hemisphere with the New
England whalers and sealers early in the last century. Whether the final corruption took place en route
or after arrival must remain a matter for speculation.
Palinurichthys coeruleus (Guichenot). Norman (1937, pp. 116-18) tentatively refers both this and
the next species to the genus Palinurichthys, remarking that the genera of Centrolophidae need revisioi]L.
It was with some doubt that Norman identified our two specimens from the centre of the shelf with
Guichenot 's Seriolella coenilea from Juan Fernandez off the west coast, owing to vagueness of the
original description : ^^g^^ ^ ^^^^6 ^
Palinurichthys griseolineatus Norman. Our specimens of this new species, which would seem to be
rare, all came from the centre of the shelf. I have found a note of E. R. Gunther's which states that:
' Its colour in life is of a delicate blueish and silvery gray : the gray running along the sides in horizontal
undulating bands which divide and merge. The reference by Norman (1937, p. 117) to brownish and
yellow has doubtless resulted from staining by teak and alcohol used in storage and transport.'
WS75 I WS97 I WS108 I
STROMATEIDAE
Stromateiis maculatus Cuvier and Valenciennes. This fish is more likely to become of value as food for
man than most of the others found within the area of the trawling surveys. Its flesh is rich and well
flavoured (other members of the family being also noted for their pleasant taste, and the high fat and
mineral content of their flesh) and free from too many small bones. In South America, where Norman
(1937, p. 118) tells us that it is called 'pampanito', the species seems to have acquired an undeserved
reputation for causing gastric disorders when eaten, and some hint of this prejudice may be found in
the Falkland Islands.'^ The crew of the 'William Scoresby' ate Stromateus with relish and without
any ill effects, and from what is known of closely allied species that have long been exploited on a
large scale in the eastern U.S.A. and in China, I have no doubt that its ill repute is due merely to more
rapid decomposition (when stored too long under primitive conditions) than species of lower food
value. There seems to be no English vernacular name for this fish, and I would venture to suggest
'spotted pomfret' as most suitable; alternatives based on other common names applied to members
of the family elsewhere would probably leave more scope for confusion, as will be shown later.
Stromateus ranked third in total weight of the fishes captured during the third survey, and fifth in
total numbers during all three surveys. Although of uniformly small size they are nearly all potentially
saleable, owing to their compact bream-like shape. They commonly range from 10 to 13 in. (25-34 cm.)
in length, and from 7 oz. to i lb. (200-470 g.) in weight. A very high proportion of the fish (almost
certainly over 60%) is edible. Analysis of the closely allied Chinese species Parnpus (Stromateoides)
argetiteus (Euphrasen) shows 64% edible (Read, 1939, p. 45). Our largest specimens of Stromateus
maculatus measured 38 cm.
1 Truly a work of supererogation ! . , ^ „ , , 1 u . c
2 Stromateus, however, seems rarely to penetrate to the immediate vicinity of the Falklands: we took them at tour
only of the many stations worked within 100 miles of the islands, and the nearest was 75 miles away.
18
360
DISCOVERY REPORTS
Although we took only ahout a thousand of this species during the three surveys, we have strong
evidence that many more large hauls of it could have been obtained in summer (when it tends to shoal
inshore), had we been concentrating upon it as a commercial fishing vessel would have been able to
do, instead of sampling the whole area.
During the three trawling surveys, S. maciilatus was captured in the ' Trawl + accessory nets' at
fifty-two stations, as shown below. It was never taken by us with ' Other gear ' :^
WS78
20
WS792B
13
WS817A
4
WS97
8
WS794
2
WS817B
2
WSI08
22
WS797B
2
WS833
203
WS2I4
I
WS797C
8
WS838
2
WS2I7
4
WS798
10
WS847A
146
WS762A
9
WS799B
6
WS847B
100
WS763
I
WS800A
32
WS848
27
WS764B
4
WS800B
7
WS84g
26
WS77I
2
WS806
I
WS850
2
WS784
4
WS809B
92
WS8S3
3
WS785A
6
WS810
3
WS8s8
I
WS785B
7
WS811I
21
WS8S9B
I
WS786
3
WS811II
58
WS862
I
WS788B
I
WS812I
19
WS864
5
WS790A
14
WS812II
40
WS866
18
WS790B
19
WS813
52
WS868
I
WS79IA
I
WS814
I
WS79IB
7
WS815
2
The distribution of these catches at different seasons of the year is charted in Fig. 44. From this
it can be seen that in spring (a) a few Stromateus were taken on the shelf in the northern region, and
that there a few were already right inshore. The species was not observed on the single line of stations
worked in the intermediate region, nor at a couple of odd stations worked farther south at that season.
From this it may be concluded that in spring some shoreward movement has begun, but probably
little southward movement.
In summer {b) the species was much more frequently taken and in very much larger numbers,
especially to the southward. Several really good hauls of Stromateus were made inshore, from Cape
Virgins northwards to a point off Puerto san Julian. Moderate numbers were taken at several stations
on the plain of the shelf, but the species was not taken in the deeper water to the southward. It was
also absent from all but one of the catches at the outermost stations (those worked nearest to the shelf-
edge) on the northern and intermediate lines, and from the immediate vicinity of the Falkland Islands.
It would seem clear that the species works southwards inshore during summer, and reaches maximum
concentration (shoaling) at that season.
In autumn {c) some Stromateus were caught still inshore in Grande Bay in the southern region,
but this was in only one of the two years for which we have autumn data, and the location of most of
the catch was farther offshore (though still on the plain of the shelf) and to the north. The distribution
IS clearly compatible with the hypothesis of dispersal of the southern summer inshore shoals in that
direction.
Stromateus was captured at two stations only during the winter survey {d). These were right offshore
on the shelf edge in the intermediate region, and the numbers of individuals were but one and four
respectively. None was taken at any of the numerous stations worked on the plain of the shelf and
to the southward at this season. These observations clearly point to continued dispersal offshore and
to the north.
Perhaps the ' spotted pomfret ' should be regarded as a semi-pelagic wanderer in winter. Its seasonal
1 Cf. Goode's remark of Poronotiis triacantlius: 'it has never been known to take a hook.' Quite recently, however, some
Ime-caught Poronotus were reported from Florida.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
361
B
B
S
e
&0
3
Xi
c
o
en
a
CO
$
'id
CO
3
XI
«
c
o
S3
v
en
a
$
l8-2
362
DISCOVERY REPORTS
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 363
movements find a close parallel in those of its near relative Poronotus (Stromateus) triacanthus (Peck)
on the north-eastern seaboard of the United States. This 'butterfish' finds its optimum from New
Jersey to Chesapeake, and its shoreward congregations in summer may be even more pronounced
than those of the Patagonian species, for pound nets are the main means of capture. Near the northern
limit of the butterfish, considerable quantities are trawled in Massachusetts ; and in more southerly
waters off North Carolina, where it was previously only known from the summer inshore fishery,
•400H
•350-
1-300-
^ l'250-
i-zoo-
M50
DECEMBER j JANUARY | FEBRUARY | MARCH
Fig. 45. Stromateus maculatus: seasonal variation in ponderal index, and corresponding mean lengths.
small quantities have been trawled offshore in winter (Pearson, 1932, Table 2, p. 18). Thus the
butterfish compares with our spotted pomfret not only in shoaling shorewards in summer, as so many
useful fishes do, but also in being trawlable near the colder limits of its geographical range, and in
wide dispersal to warmer and deeper waters in winter.
A first step in further efforts to elucidate the bionomics of Stromateus maculatus was to test the
seasonal variation in ponderal indices for indications of the spawning season. Weight data were
available only from a limited period during the third survey, and average K for either sex, plotted at
mean dates, yielded the values joined by the heavy lines in Fig. 45.
364
DISCOVERY REPORTS
This shows, first, a 27% rise in average K values between December and the end of March, with
httle discrepancy between the sexes. Secondly, an apparent drop in K values in February, to a level
which was, however, still well above the December figures. This was found to be associated with a
drop in mean lengths (thin lines in Fig. 45), i.e. with an increased proportion of smaller fish in the
population sampled in February over that of the population sampled in January.
It seems fair to conclude that most spawning takes place in early summer, and further (though at
this stage only as a working hypothesis) that larger fish spawn first and therefore recover condition
earlier than smaller ones. The smaller fish shoal inshore later than the larger ones, as will presently
be shown ; and any mature individuals among them are presumably correspondingly later in spawning.
But it is also quite possible that a proportion of the later, smaller shoalers are immature. If so they
may still tend to decrease the mean February K if, like adolescent hake, they show seasonal harmonic
variation in condition like mature fish, but on a lower level.
1350-
2
hi 1-300-
2
1-250 ■
--0
[8:37]
\^:46o]
— \ —
25
50 75 iOO
MILES FROM THE COAST
125
150
Fig. 46. Stromateus maculattis: variation in average K with distance from the coast; southern region, summer.
Figures in brackets denote numbers of weighings, with the numbers of fishes measured in iUilics.
The further testing of these hypotheses as to the movements of S. maculattis, from the available
data, depends mainly on considerations of size distribution (length frequencies) in relation to distance
from the coast and at diflFerent periods; but one further application of the use of K seems to help,
and may be described before we proceed to the other evidence. In Fig. 46 the mean K values of
Stromateus taken at different distances from the coast in the southern region, in summer, are shown
joined by a line (merely to guide the eye). It will be seen that the values increased sharply with
increasing distance from the land. This shows that the ofl'shore fish had had longer to recover after
spawning, for it is reasonably certain that the converse explanation (inshore population spawning
while that off'shore had yet to do so) could not hold, for we have already seen that the general rapid
increase of K values throughout the summer suggests that most spawning takes place early. The
point becomes even clearer when the length-frequency data are considered.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
365
In Fig. 47 the length-frequency distributions of Stromateus taken inshore in the southern region,
in successive summer months January and February, are shown as histograms. Here we see that in
January the distributions were strongly unimodal, with the mode at 31 cm. (males) or slightly higher
(females). In February the distributions were clearly bimodal for both sexes, with the modes at about
LENGTH5_CM5
Fig. 47. Stromateus maculatus: seasonal variation in percentage length frequency at inshore
stations in the southern region.
25 and 30 cm. for males, 26 and 31 cm. for females, the smaller fish being more numerous than the
larger. This proves the point made above, that the smaller fish moved inshore later than larger ones.
It will also be shown later (Fig. 50) that in the intermediate region the length frequency of inshore
Stromateus around midsummer resembled the February distribution in the southern region. The
smaller fish in particular are thus mainly confined to the warmer inshore counter-current in their
movement down the coast, but also the whole of the seasonal cycle is probably centred earlier m the
366 DISCOVERY REPORTS
year farther north. Some of the northern and intermediate fish probably do not proceed so far south
before they begin to move offshore again ; and indeed it can be shown that the mean lengths of southern
samples were significantly greater than those of fish from the intermediate region. We cannot deter-
mine the relative extent of meridional and on- and offshore trends of movement, owing mainly to
dearth of inshore data from the intermediate and northern regions early in summer, particularly from
the Golfo san Jorge; but it seems certain from the general distribution observed that both trends
exist together.
The disposition of our catches of S. maculatiis in the southern region in summer, the time of
maximum concentration, is shown diagrammatically in Fig. 48, in relation to distance from the main-
land coast. This arrangement of the data gives some idea of what a practical fisherman in quest of the
species could expect, in addition to amplifying our ideas as to its general bionomics.
It is very evident that the species was most plentiful close inshore, but well distributed up to
100 miles from the land. It was rare farther offshore at this season. The sex ratio was highest inshore
('normal') and the proportion of males diminished as one proceeded seawards. The males of this
species are slightly (but in large samples significantly) smaller than the females, and it is just possible
that this difference has some small effect on their mobility, for we know that in other fishes where the
difference in size between the sexes is more pronounced, the larger females travel farther and faster
than the males. Here, however, the sexual dimorphism is so slight that one would expect that some
factor not yet determined, such as more rapid dispersal after spawning among males, must be mainly
responsible for the diminished sex ratios in samples of the offshore population.^
In Fig. 49 the percentage length frequencies of both sexes of Stromateus for the same area and
period are considered in relation to distance from the coast, the observations being grouped according
to the same distance intervals as were used in constructing Fig. 48. The inshore grouping showed
bimodal distribution, with modes at about 25 and 29 cm. for males, 26 and 3 1 cm. for females. These
probably indicate year-classes.
There was an increased tendency towards suppression of the smaller mode as one proceeded off-
shore, until, at distances of more than 100 miles from the land, such few fish as were caught belonged
almost exclusively to the older (larger) year class.
Within a brief period around midsummer 193 1-2, a series of observations were obtained that fall
naturally into two groups, serving to show the strong contrast in the population of Stromateus inshore
and that found offshore in the intermediate region, in respect of length-frequency distribution
(Fig. 50). The inshore grouping showed bimodality with the smaller length-class dominant; the off-
shore grouping was unimodal, with almost complete suppression of the smaller class.
There are still some big gaps in our knowledge of S. maculatiis, for pressure of work in the field
upon other, more important species, made it impossible to collect routine data on the condition of the
gonads, for example. We have very little idea of its diet, though some were found to have fed upon
Parathemisto. Hake have twice been found to have preyed upon Stromateus, but we can only guess at
its other natural enemies in this region.
The main features in the bionomics of S. maculatus that seem clear from our data are : a double
trend of seasonal movement, inshore in spring and summer with maximum concentration, and offshore
in autumn and winter with maximum dispersal ; and, superimposed upon this, a meridional move-
ment southwards in summer and northwards in winter. A schematic representation of these move-
ments is given in Fig. 51.
1 Probably females have somewhat greater food requirements than males during the recovery period, with a resultant
tendency towards greater local concentration where food is plentiful (though such concentrations would not be comparable
with the spawning shoals). This would lead to the same result.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
367
LESS THAN 25miles
FROM LAND
26 -50mile5
FROM LAND
51-IOOmiles
FROM LAND
MORE THAN
IOOmiles
FROM LAND
Fig. 48. Stromateus maculatus: relation to distance from coast in the soutiiern region in summer. A. Total fish taken m each
distance grouping. B. Number of hauls (positive black, negative white). C. Number of fish per positive haul, showing also
the sex ratio. White segments indicate the percentage of males. D. Hours trawling within each distance groupmg. E. Number
of fish per hour's trawling.
D XXIII
19
368
DISCOVERY REPORTS
LENGTH5 CMS
5
O
en
o
o
o
H
Fig. 49. Stromateus maculatus: percentage length frequencies at different distances from the
land, southern region, summer.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
369
Fig. 50. Stromateus
maculattts: variation in length frequency with distance from the coast,
intermediate region, 17. xii. 31 to 9. i. 32.
19-2
370 DISCOVERY REPORTS
Stromateus probably spawns early in summer, but we do not yet know how prolonged the spawning
season may be. It seems clear, however, that the larger, older fish lead in the migratory movements,
and spawn earlier than any mature individuals among the smaller ones.
The summer shoreward movement of the spotted pomfret finds a close parallel in the behaviour
of the most valuable of its northern relatives, Poronotus triacanthus (Peck), off the southern New
England, middle Atlantic and Chesapeake Bay States of America. With Stromateus the superimposition
of a marked meridional trend of movement is probably occasioned by the necessity for maintaining a
species with a prolonged pelagic post-larval phase within its ecological norm (cf. E. S. Russell, 1937,
p. 321) and directly correlated with the current system of our area (Fig. 4).^ Any similar tendency in
Poronotus will be less obvious because of the more complex hydrological conditions of its habitat.
Only the most northern parts of its range show resemblance to conditions off Patagonia. The main
locus of Poronotus is influenced by the warm gulf-stream flowing towards higher latitudes, which is
the converse of the Patagonian conditions. Moreover, the coastline in the habitat of Poronotus runs
more east and west than the coasts frequented by Stromateus, and therefore any meridional component
of movement will be acting more nearly in the same direction as that of the primary on- and offshore
movements of Poronotus.
Stromateus maculatus has an extensive distribution up the west coast of South America as well as
off eastern Patagonia, and Norman (1937, p. 1 19) has pointed out that large series of specimens from
the more distant localities might reveal the existence of two or more races or subspecies. He also
stated that there is some doubt as to the specific identity of the specimens reported as far north as
Peru by Valenciennes. However, the narrower Peru coastal current brings relatively temperate con-
ditions very much farther north up the west coast, than does the Falkland current off the south-east
coast of South America; and since these Peruvian specimens were recorded as most abundant in the
Lima market in winter, they might have resulted from seasonal meridional movement of the most
northerly of the west coast stocks, in view of the known behaviour of the species elsewhere. It is well
known that the extreme northward extent of the Peru coastal current leads to a more northerly dis-
tribution of other temperate types, including such fish as hake {Merlucchis gayi), than is to be found
anywhere else south of the equator.
In order fully to appreciate the potential value of the spotted pomfret in our area, it is instructive
to consider the Stromateidae already exploited elsewhere. The family has been much subject to
taxonomic changes, and further systematic revision — impossible without further widespread col-
lecting— is still needed. It is hoped that Table 37 gives sufficient synonymy to leave no doubt of
the identity of the species referred to. There remains the bewildering tangle of common names, many
raising most interesting etymological problems, which I give with their localities in the second column.
To quote individual authorities for all these would take up too much space. The object of the table
is to give some idea of the range and relative importance of the Stromateidae already exploited, in
compact form, before proceeding to some further consideration of their common names.
The etymology of the name ' pampano ' or ' pompano ' is extremely interesting. Its literal meaning
in Spanish is 'a young vine branch or tendril', but the ichthyological connotation is very old, and
may possibly have first applied to Sparus salpa. Of the ultimate origin of the word from the root
'pomum', a fruit, there can be little doubt. The Dutch word ' pampelmoes ', applied to Stromateus
fiatola in South Africa, derives from the French ' pampelmousse ', the grape-fruit or shaddock (compare
'pomelo' and 'pompoleon', names occasionally applied to this fruit formerly in England). Perhaps
the likeness of a deeply compressed fish seen in profile to a grape-fruit viewed in the same way accounts
^ Such very few Stromateus juv. as we captured were found only in the north.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
371
' ' I I J'-l ' ' ■ ■ ■ ' ' ' '~^~'
Spring" • —
Summer ■
Autumn ...—..--
Winter
1 1 1 1 1 1 ' I I
h-^ I I I I 1 I I I I 1 I I I I I i=c=\
45
50"
55
Fig. 51. Schematic representation of the probable main trends of movement of Stromateus maculatus.
372
DISCOVERY REPORTS
r
o
.8
'9
o
o
e
o
<3
S
S
t;
o
a,
e
o
c
o
o
W
j2 ^o't;
O T)
9 c
T)- Q. 9"
o S o ^ c ^
a^rS U-;
60
■as
°0 «
I ^
XI o d o -r;^.3
*~» *-> C r3 ■*-'
(X o
■S-a
« rt CD 3 ^ M
-^■2 ^ ^ f2 S
U-. r2 T3 ~
-.o O S " . h
°~- M O ^ >, fi
, , C u > rt °
3T3 •£ ^ g m
2 E t*- M C il
;5— O « « C
-a d
« ex
-5 '
M .™_>-.
to «
.S -S
■" o
D '-' 3
^ 3 —
^ C
— u ^
4J CO I
-.£; 6
O 73 uS
« C
U I) ■ —
•s-s &■
>r^ C/?
lU
rt '
o c P
M rt 5
-P -C
60
O 60
2 .2
CI X -3
Mug
pa
1-3
-a
c;
a
a
u
M
o
U
J, '^ '^ 1i.
< ^ s
^ -^S
^ t3 &■ . «
CQ ' — > 5
c/3 ^"^ • ^
to « -. Z O
rt rt ^ h
C '^ 1) u o
« J3 ^ 2
<^ *^.S
r^ O P «
6 C ° !3 60
c
3 C Q
c8 ^
> "^
^ c
9 rt
cc;^
o
u. o
Id
a
a
a
C
IS
U
— Ph
S2
^^ Crt
6
O
fct,
o PJ o
W
M
G
C
o
6
e
o
U
o
V
Oh
is -H
.. o
a,
O
o
U
d 1— .
1) c
« s
c o
« V 2
TJ
C
c
=3
O
a
CO
Vi
e
E
OJ
-r)
6
n
a
W
,
•^ ,
-^
"a
«
^
•^
M
It^
iS^t'
■« -«
Jt
c>3CoaH
« 6
>--s
o
O rt
^ en
n g
<u
C CA
^ 3
3 ^
"J
u • — •
2 S
-a
e
«
60
o 2
-a
c
13
b J5
«
■3 3
« E ■;;
60
Icn
g u E o
o
I i: £
■ex, a "
2 "3
-a a
<u
ns
X
0^
0)
«J
tfl
T3
tn
d
<u
91
SoJ
•a
c
«j
rt
fn
c
00
ro
-C
OS
o __
5 c
B
ifi
<n
i> o
■5 c
ex o
-a
P
e -P
ex g
>. h o
60 8 X
W a.^
„ 3
« -S ^-o
u So
.- -IN
< :s-£ o
•S 2 p -*
,p
3 <(< « .
ex
_p —
-in
manize
y have
Philo (
,^ -«
S ■« o «
pane
erfis
fish
pelrr
iLa
e (r
is m
that
i: C
^ S ^^ j= _
« a
0,05
O ^ T3
rp i> ■ —
•-^ P rxi
60
P
o
be '
P
o
X
« J!
S:^-^t3 «^
«J N «
O r^ =^
P . U
-a; -a;
■ct,
o s
^
J3 «3
60
P
a
j=
P
■Ci,
S
Co
a
s
CO
2-^
1^
1-2 <
U W - I t
<^ CT] }-(
a lu o
S >. ii b
3 '5-«
«a
-a
G
C/3
J3
<3 3
)^ P
~ p
"« p
"J
p
P <n
P*
_ P
P •<
M .5 o
" a -S n S
-^ !< 3 >
VI
3
u
P
P
« ft; < XI -^
Sl
CO
a
^
S
c
-s:
G
G
ft5
J2
^
<u
u
H
o
2.
§
^
Si
CO
«
u
« ^-,
s
«
<
S 60
■1
S
■ex u
"^
^
-^
d:^
373
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
for this usage. 'Pampano' is applied to S.fiatola in Spain to-day (Navarro et al. 1943, p. 132) and,
as we have seen, the diminutive form 'pampanito' has been applied to our S. mactilatiis in South
America, while Palotneta simillima is sometimes called the 'California pompano'. Both words are
much more widely used to describe Trachinotus spp., and apart from encouraging possibilities with
Stromateus maculatus, the Stromateidae to which they have been applied are unimportant species.
If there is any choice of vernacular names, therefore, I feel that ' pampano ', ' pompano ' or ' pampanito '
should be confined to Trachinotus spp., clearly distinguishable from Stromateidae by the presence of
pelvic fins in the adult.
' Butterfishes ' has strong claims as the best alternative for Stromateidae, as the most widely accepted
common name of the most important species, Poronotus triacanthiis. Unfortunately the habitat of this
species overlaps that of Peprilus alepidotus and, as shown in Table 37, both species share a wide range
of vernacular names in the eastern U.S.A. It would be undesirable to take any step that might extend
the sphere of confusion that may result from this, for as can be seen from Table 38, the name ' butter-
fish ' is also in current use for very widely different fishes in other English-speaking countries.
Table 38. Other usages of the chief common
names applied to Stromateidae
Chief common
names applied
to Stroma-
Fishes other than Stromateidae to which these
names are also applied
Localities in which such usage prevails
teidae
Pampano or
Sparus salpa (Linnaeus)
Sometimes, in Spain
pompano
Trachinotus spp., especially T. carolinus Linnaeus
the common pompano
West Indies and eastern U.S.A.
Trachinotus ovatus Linnaeus
In South Africa
Trachinotus blochii (Lacepede)
Sometimes, in Hong Kong
Pampanito
Trachinotus rhodopus Gill
Gulf of California southwards to Panama
Butterfish
Pholis gunellus Linnaeus
In Britain
Coriodax pullus Forster
In New Zealand
Pseudolabrus celidotus Forster
In New Zealand, when large
Pseudolabrus cinctus Hutton
In New Zealand, 'deep-sea' butterfish
Johnius (Sciaena) hololepidotus (Lacepede)
In South Australia [this is the well-known Kabel-
jauw of South Africa]
Pomfrets
Bramidae
In Bermuda and U.S.A., notably in Jordan's writings
Drepane punctata (Richardson)
In Hong-Kong, 'chicken-basket pomfret'
Trachinotus blochii (Lacepede)
Sometimes in Hong Kong, 'yellow-wax pomfret'
There remains the possibility of using the name ' pomfret ' for Stromateidae, with suitable prefixes.
In favour of this one may urge that the usage is already widely current in eastern waters where the
family attains an economic importance second only to that of Poronotus triacanthus} Drepane punctata,
the ' chicken-basket pomfret ' of Hong Kong, need not deter us ; for that is merely an attempted literal
translation of a Chinese name, the fish is relatively unimportant and is already well named the ' con-
certina fish' in Natal. (Here we see, however, that confusion of fishes of different families under the
same vernacular names takes place in Chinese as well as in the European languages.) ' Yellow-wax
pomfret' for Trachinotus blochii is similarly an attempted literal rendition from the Chinese, and it
appears that there is already a tendency to replace it by ' pompano'. (With advantage, as I think; cf.
Herklots and Lin, 1938, p. 21.)
A far more serious objection lies in the use of ' pomfret ' for Bramidae by a great ichthyologist like
Jordan, but I cannot discover that the Bramidae (of which the type species is widely known as ' Ray's
bream') are sufficiently common to be of much economic importance anywhere, so that there is not
1 It is even probable that the far eastern pomfrets may be relatively more important than Poronotus, but in con-
sidering far eastern fishes we have only limited local data to go by, in contrast to the superb U.S. statistics.
374 DISCOVERY REPORTS
the same urgency in the matter of deciding the best common name for them. Now 'pomfret' in its
ichthyological connotation is stated to be of uncertain etymology by most EngHsh dictionaries, but is
I believe derived from a German vi^ord 'palmfett' — palm (oil) butter (De Vries, German-English
Dictionary for Scientists, 1939) ; at least the well-known oiliness of the fishes most widely called by the
name supports such a view. The use of ' pomfret ' for the Stromateidae would thus preserve the literal
significance of 'butterfishes', already current in America, while avoiding the confusion that may
result from widespread use of ' butterfish' for other groups in other English-speaking countries, shown
in Table 38. I therefore tentatively suggest ' spotted pomfret' as a suitable English common name for
Stromateus maculatiis.
ATHERINIDAE
Austromenidia smitti (Lahille). The ' William Scoresby ' secured three specimens of this littoral species
on hand-lines in the Golfo Nuevo, off the jetty at Puerto Madryn. Specimens from Port Stanley in
the Falkland Islands, where it is not uncommon, were sent to Norman (1937, p. 120) by Mr A. G.
Bennett.
Austromenidia nigricans (Richardson). No specimens of this littoral species were obtained during
trawling operations, but series from East and West Falkland were sent to Norman by Mr A. G.
Bennett and Dr J. E. Hamilton.
A note of E. R. Gunther's, referring to both these species, reads : ' . . .known in the Falkland Islands
as " smelt", were never taken by nets fished outside the littoral zone. They are among the most prized
as food, growing to a length of twenty-two inches (56 cm.).' Norman (1937, p. 122) quotes Mr A. G.
Bennett's notes on the erratic movements and shoaling habits of smelts, and their spring spawning in
sheltered shallow water. From the series of measurements given by Norman it seems probable that
A. smitti is the larger of the two species.
Already in good repute as food, Falkland smelts would appear to be among the most promising
subjects for the development of any small-scale fishing industry to supply local needs.
SCORPAENIDAE
Sebastvdes oculatus Cuvier and Valenciennes. We obtained more specimens of this species by casual
fishing in the Magellan channels than by the systematic trawling oflF the east coast of Patagonia, and
this suggests that the Pacific may be the real home of this fish, as it is of most other Scorpaenidae.
As we have already had occasion to note, the ecological niche filled by the ' rock-fishes ' in the north
Pacific seems to be occupied by the characteristically southern Nototheniidae in our area.
Norman (1937, pp- 123-4) 'preferred to use the name Sebastodes in the wider sense of Jordan and
Evermann (1898)', finding recent further subdivisions of the genus unreliable. He points out that
the Patagonian species is doubtfully distinct from Sebastodes chilensis Steindachner, found on the west
coast of South America; and is barely separable from Sebastichthes capensis Gmelin from South
Africa, the Tristan Group and Gough Island (Steindachner himself regarded this last species as
identical with the Patagonian one).
The most familiar relative of Sebastodes oculatus is of course Sebastes marinus Linnaeus, common
on both sides of the north Atlantic, where it is known as the ' rose-fish', ' Norway haddock', and even
(at Halifax, N.S.) as the ' John Dory ' ! I mention this last because I can find no record of ' John Dory '^
referring to any fish other than Zeidae in any other part of the world, and confusion of common names
is one of the greatest bugbears all fisheries workers have to face. The rose-fish has become very im-
portant commercially in recent years ; it is one of the staples of the frozen fish trade in eastern U.S.A.,
^ Though the 'dory' part of it, if derived from the Spanish dorado, is a different fish altogether!
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 37S
and large quantities are also taken by European trawlers off north Norway. Its Patagonian relative
could be utilized, but seems to be so rare that it could never form an important part of a commercial
catch.
Norman tells us that in South America Sebastodes is known as 'cabrilla'. This name is in much
more general use for the Serranidae in warmer waters of the Caribbean Sea and elsewhere.
Specimens of S. oculatus were obtained by the ' William Scoresby ' as shown below :
WS800B I, length 30 cm., weight 500 g. {K = 1-852)1
WSSii II I, length 40 cm., weight 1200 g. (A' = 1-875) " I" ' Trawl + accessory nets'
WS823 I, length 23 cm. J
Fortune Bay, Baverstock Island, 5. v. 31, 2 on LH at 22 m. depth, lengths 25 and 27 cm., weights 233 and 283 g. (giving
i^= 1-491 and 1-438)
Puerto Acero, 9. v. 31, i on LH at 23 m. depth.
WS742A I in BTS at 58 m. depth.
Helicolenus lahillei Norman. We did not obtain any specimens of this species, but as Norman (1937,
p. 124) gives 'coasts of Uruguay and northern Argentina' as its habitat, it is possible that it may
occasionally reach the northern part of our area. A specimen of the closely allied H. lengerichi Norman
was taken by the ' William Scoresby ' during the Peru Current investigation, but this is known only
from the west coast of South America.
CONGIOPODIDAE
Congiopodus penwiamis Cuvier and Valenciennes. This heavy little fish of grotesque appearance was
well figured by Lt.-Col. Tenison (Norman, 1937, p. 127). It was never abundant, and is most unlikely
to have any economic value, but the known facts of its distribution are of considerable biological
interest.
Congiopodus has a crest-like dorsal fin supported by strong spiky fin rays. The three recurved anterior
rays project over the head. This armature is supplemented by downwardly directed spines on the
pelvic fins and by a most unusually tough skin. From its appearance the fish seems in no way adapted
to swimming fast, but rather to hovering. The olive-coloured skin is variously spotted and marbled
with dark brown or black. It matches the colour of much of the bottom deposit and may be supposed
to have concealment value.
Congiopodus was found within limited areas on the plain of the shelf, mostly in the northern and
intermediate regions, at different seasons and even in different years, as the records show. It was
never taken at less than 63 miles from land, and only once at more than 200 miles. Further, it showed
a very restricted depth distribution, all the records falling within the range 97-146 m. These data,
coupled with the characteristics making for a stationary disposition already described, strongly suggest
that Congiopodus does not wander about, but that these areas on the plain of the shelf are more or less
permanent haunts. The extreme geographical range of the species is, however, wide, extending
throughout the temperate waters of South America, on both coasts from Uruguay to Peru.
Our specimens of Congiopodus, all taken in the ' Trawl + accessory nets', were obtained at the
following stations :
WS97 I WS792B
WS217 2 WS793
WS790A I WS794
WS791A I WS800B
WS791B 2 WS807
WS792A 9 WS85S
4
WS8S9A
6
WS8S9B
2
WS860
5
WS862
I
WS866
2^6 DISCOVERY REPORTS
PSYCHROLUTIDAE
Neophrynichlhys marmoratus Gill. This creature of hideous aspect is locally known as ' el Gran sapo
de Mar'— the great big-toad of the sea. A glance at Lt.-Col. Tenison's drawing (Norman, 1937,
p. 128) shows why. It was obtained in small numbers at a few stations in the southern part of the area
investigated. Although the depth range was very wide, it is perhaps noteworthy that the only haul
in which it was at all plentiful was made in very shallow water just south of the entrance to Magellan
Strait. The distribution appears to be correlated with prevalence of rough ground, which is perhaps
a necessity for survival of such an obviously slow-swimming gelatinous creature, lacking all armament
(the dermal excrescences are not spinose). Thus we found it only inshore, or out on the edge of the
main slope, on ground so foul that little trawling could be done there. This distribution may be con-
trasted with that of Congiopodus, a member of the nearest related Scorpaenoid family, which, with its
heavy dermal armour, inhabits the centre of the shelf:
WS93 2
WS8J4
>i7
WS583
I in BTS
WS97 I
WS847A
I
WS832
I in NR
WS244 2
WS851
I
WS877
I juv. in NR
WS825 I
AGONIDAE
Agonopsis chiloensis (Jenyns). To the south Agonopsis was found chiefly on the shallower coastal
grounds. Farther north it penetrated to the centre of the shelf also, and there some 30% of our
specimens were secured. Its distribution is of great interest, for it seems to show a strong tendency
to remain in localized areas, roughly intermediate between the contrasted types of distribution shown
by the two species last mentioned, and in conformity with what one might infer of its habits.
The fish is small, seldom exceeding a length of 15 cm., and the body is, in Boulenger's words, 'com-
pletely cuirassed with osseous scutes', giving it a lizard-like appearance. The body is marked by a
number of brown cross-bars on a sandy buff background. Beyond a very evident lack of agility there
is little direct evidence as to its habits.
The colour of Agonopsis was often matched by hauls of buff-coloured invertebrates such as the
hydrocaulus of tubularians and the reef-like polyzoan Smittina among which it was found ; or the
sand, shell and coral fragments of the bottom deposit.
Agonopsis was often caught in the Russell bottom-net. Unless these were individuals behaving in
an exceptional manner, it shows that the fish rise when disturbed, or alternatively are in the habit
of swimming several inches above the sea-floor. The small flatfish Thysanopsetta lives right on the
bottom and was very commonly caught in the fine nets attached to the back of the trawl, but was
taken in the Russell net only once.
As can be seen from Fig. 52, Agonopsis was caught in five separate localities. The conclusion that this
indicates a stationary habit is borne out by the fact that within each group catches were secured in
different months. In the most northern of these localities the fish were met with in October, December
and March; in the most eastern (Falklands area) in February, May and September; while in two
others the fish were met with in different years :
WS7I 20
WS243 I
WS583
5 (in BTS)
WS81 2
WS787 4
WS754
3 (in NR)
WS83 4
WS791B I
WS767
I (in NR)
WSgj I
WS799A I
WS832
3 (in NiooB)
WS95 s
WS800B I
6 (in NR)
WS216 I
WS847A I
WS836
2 (in BTS)
WS219 I
WS860 I
WS852
I (in BTS)
WS221 I
WS873
Post-larvae present (in NR)
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
377
Fie ,2 Distribution of Agonopsis chiloensis: positive records only, roughly contoured to show their localization in differerit
seLns'. oLonds, springfcircles, summer ; triangles, autumn ; squares, winter. Solid symbols represent captures m Trawl +
accessory nets'; cross symbols with 'Other gear'.
LIPARIDAE
Careproctus falklandica (Lonnberg). Six specimens of this fish were trawled at St. WS89 off Cape
Virgins in April 1927. It was previously known only from the Falkland Islands and the Burdwood
Bank. This suggests a distribution similar to that of Neophrynkhthys, which it resembles m havmg a
soft and gelatinous body. Its colour in life was pale orange.
Unidentified liparids were obtained at two stations in the Magellan Strait, outside the area of the
trawling surveys.
BOTHIDAE
Thysanopsetta naresi Gunther. This small flatfish, resembling the 'scald-fish', Arnoglossus laterna
(Walbaum), of British seas, was the only member of the family at all numerous m catches obtamed
during the trawling surveys. Its small size renders it unsuitable for human consumption, the largest
378
DISCOVERY REPORTS
..<l
■••■■■ < 2.
<\
4 /
8
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES
379
>.
>i
C
O
3
05
O
O
J3
<u
V o
J= Cu
*->
rt -
>, 6*
•^ u<
<L>
■? -^
'A O
0^ -
c
U= J3
11 .t;
•o ^
-« M
C U
rt u.
M 3
C D.
^ n
O
J3 -^
»-»-. • -
O tfl
1)
M >
OJ -jS
b ca
t« bO
V
"3 c
1-1
c-S
S-^
« . '
-C M
OJ
o c
c fc^
o o
ca
i2 +
<n
o g
C
^
« H
60-
cn
II
a
CO
f «
•!S"S
C
S R
a
« u
3
C l-(
3
a ac
c«
Tin
e
&>"«
E-1 -
)-«
>-.
u
C 3
o ^
6
6
3
05
^-R
C/3
n .3
aj
•C-fl
o
^ ^
o
CO —
a.
rt 3
03
^
'^^-a
05
. c
C
en o
o
e
M ="
rt
ta <rQ
38o DISCOVERY REPORTS
captured being only 14I cm. long. As forage for larger species, of greater potential economic value,
Thysanopsetta is evidently of considerable local importance, for, as we have already seen, it figured
largely in the stomach contents of hake [Merluccius hubbsi) and Thyrsites atun, as well as in more
typical bottom-feeders such as Cottoperca gobio and some of the rays.
The distribution of Thysanopsetta is very interesting. It seems to be peculiar among the fish fauna
of the shelf in being influenced more by the nature of the bottom deposit than by any other distri-
butional factor.
Some seasonal movement there is, for when the catches are charted (Fig. 53) it can be seen that the
fish were farthest south in autumn, farthest north in winter, and nearest to the mainland coast in
summer. These trends agree in time and direction with those followed by most other fish in the region
that show seasonal movements, but were much smaller in extent, so that Thysanopsetta was rarely
found far from its ' metropolis ' on the central area of brown sand that occupies most of the plain of
the shelf in the intermediate region. This was delineated by Matthews (1934, pi. xiii), when he was
working on the bottom deposits obtained with the dredge during the trawling surveys, long before
the fish data had been worked up. The close correlation between this deposit and the distribution of
Thysanopsetta was first seen by E. R. Gunther, and was very clearly described by him in a preliminary
(verbal) account of these surveys. He had subsequently altered the manuscript so drastically that I
have had to attempt an independent presentation of the data (Table 39) and can only hope that this
will prove adequate, in conjunction with the charts.
Table 39. The relation betzveen a localised type of bottom deposit and catches of the flatfish
Thysanopsetta naresi
Figures in heavy type relate to observations made within the central area of brown sand, delineated by Matthews
(1934, pi. xiii); figures in italics to observations made elsewhere, on other deposits.
First survey
Second survey
Third survey
No. of hauls
No. of hauls positive
% occurrence
No. of fish caught
Hours' trawling
Hours' positive hauls
4
3
75
21
4
3
25
5
20
16
24
5
7
4
57
10
7
4
27
0
0
0
27
0
19
12
63
1712 +
34
21
85
9
10-5
35
logl
15
Rate of capture : Per hour's trawling
Per hour's positive hauls
S-2
70
0-7
3-2
1-4
2-5
0
0
500
81 0
0-3
2-3
It also seems clear that Thysanopsetta is most concentrated (i.e. most given to shoaling) in summer.
Its absence from our few spring hauls should not be considered as evidence of the converse (greatest
dispersal at that season) for they were nearly all worked to the north of the normal limits of distri-
bution of the species. It is more likely that Thysanopsetta is most widely dispersed during winter.
Thysanopsetta naresi is evidently a true bottom-dweller, and even more so than some other species
of flatfish inhabiting similar depths, for the latter often show by the nature of their food that they
must spend a considerable part of their time swimming clear of the bottom. We know that Thysanop-
setta must lie exceptionally close, because although it was so frequently captured in the accessory nets
attached to the back of the trawl, it was only once taken in the Russell bottom net. This is designed
to fish a few inches clear of the bottom, and very many hauls were made with it in regions where
Thysanopsetta was known to be plentiful. It is the more remarkable that such (normally) semi-pelagic
feeders as hake and Thyrsites were found to have fed extensively upon Thysanopsetta. I believe that
the explanation partly lies in the uniformly moderate depths over the plain of the shelf, for this brings
the bottom fauna well within reach of larger predators which are given to diurnal vertical migration.
DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 381
and which have to be satisfied with a mainly pelagic diet when in deeper waters. Further, the hake
are to be found in the Thysanopsetta area just after they have spawned, and at that time, as is well
known, they eat everything within reach.
The greatest concentrations of Thysanopsetta were found near the southern and inshore limits of
its main habitat, in late summer. Its concentration there may be correlated with the relatively warm
and intermittently southward flowing inshore surface water. One would suppose it to be a prime
requisite for survival that the post larvae should attain the bottom-living stage before being swept
north of their ecological norm in the main path of the Falkland current. On these premises it seems
a fair hypothesis that spawning will be found to take place near the southern limits of the species'
range. It is also possible that the period of the pelagic phase may be shorter than is usual in allied
species, thus making for quick settlement upon suitable terrain, but the only factual support for this
suggestion comes from a single rich haul of post-larvae secured with the Russell bottom-net, slightly
to the south of the main locus of the species, in the late summer of 1930-1 (St. WS832). Here also
larvae were taken in plankton nets fished nearer to the surface. Adults have never been taken (in
numbers) much farther south than this, so it seems certain that these larvae must have resulted from
a very recent spawning:
WS77
4
WS795
1 1
WS814
4
WS80
I
WS796A
333
WS848
I
WSgo
8
WS796B
6
WS857
10
WSgi
I
WS797B
43
WS862
7
WS92
2
WS797C
12
WS874
2
WS94
I
WS804A
I
WS96
19
WS805
I
WS742B
I (in BTS, Mocha Island, west coast)
WS97
I
WS806
6
WS832
20 larvae (in plankton nets)
WS2i6
I
WS807
5
95 post-larvae (in NR)
W.S219
3
WS808
547
WS861
2 (in BTS)
WS220
2
WS809A
713
WS863
6 (in BTS)
WS222
5
WS809B
34
51
I (in OTL)
WS787
36
WS810
I
WS791B
2
WS811II
3
Hippoglossina mystacium Ginsburg. Norman (1937, p. 132) records this species (which is closely
related to the more northerly H. macrops Steindachner, of the west coast) from Magellan Strait, so
it may occasionally penetrate to the eastward, although we did not find any.
Paralichthys patagonicus Jordan and Goss. This is an east coast species, but its main habitat lies
far to the north of the area of the surveys, and we did not capture any. Several specimens of
P. microps (Gunther) were obtained by the ' William Scoresby ' off the west coast, but this species is
not known to extend 'round the corner'. According to Norman it is sometimes called 'lenguado' in
South America, but this common name for flatfish (used for Solea in Spain) is applied so promiscuously
that it is safer to regard it as a family name to be applied to any small Heterosomata (just as the ancient
word ' butt ' is still used in parts of East Anglia, usually for flounders, but without any intention of
specific distinction).
Paralichthys isosceles Jordan. This is one of two flatfish of the area surveyed that are big enough to
eat ; it is sometimes over a foot in length and half a pound in weight. We found very small numbers
not infrequently in shallow water (90 m. and under) in the extreme north. None was taken south of
lat. 46" S :
WS762 I WS788B I
WS762B I WS853 2
WS763 3 WS852 I (in BTS)
Xystreurys rasile (Jordan). We found this species also in the extreme north of the area surveyed, but
it was very rare. Norman tells us (1937, p. 135) that it, too, is called 'lenguado', which suggests that
382 DISCOVERY REPORTS
somewhere— perhaps farther inshore, or ahogether to the north of our area— it is more common.
Xystreurys also reaches ' pan-size '.
Mancopsetta maculata (Giinther). A single specimen of this rare species was obtained during the
trawling surveys at St. WS218, in deep water over the edge of the shelf, just north of the southern
boundary of our northern region. It was 24 cm. long. Norman (1937, p. 136) refers to this specimen
as taken 'north of the Falkland -islands'. This is true, of course, but might well obscure the fact that
it was well over three hundred miles to the north of them. This capture off the Patagonian shelf is
interesting because the species was previously known only from the holotype collected by H.M.S.
' Challenger' in 310 fm. (567 m.) off Prince Edward Island in the Indian Ocean. Although such wide
distribution is not unknown among more versatile pelagic fishes, such as Thyrsites, this is the only
example known to me in a group so specialized as the Heterosomata. E. R. Gunther noted 'it is
evidently a species inhabiting deep water and is unlikely to be of commercial value'.
Achiropsetta tricholepis Norman. This genus was discovered during the investigations here described.
Two specimens are known. One, 10 cm. long, was taken at St. WS89 off Tierra del Fuego. It was
noted as 'having a grey colour, but highly transparent so that heavy print could be read through
most parts of the body'. In 1932 further hauls were made round the position of St. WS89 in hopes
of finding more, but none was taken.
The other specimen had been found at the Falkland Islands some years previously, but was not
reported until the publication of Norman's paper in 1937 (p. 136). It was seen from the jetty in
Stanley Harbour swimming near the surface, and was caught in a bucket by Mr Bert Radclif. Dr J. E.
Hamilton into whose hands the specimen passed had independently noted the transparency of the
living fish. E. R. Gunther suggested that the transparency may be related to a surface swimming
habit, accounting for the virtual absence of the fish from our bottom nets, and its occurrence on both
sides of the Falkland trough :
WS89 I, of 10 cm. in the trawl
Port Stanley Pigzz i, of 10-5 cm. (J. E. H.) in a bucket at the surface
FEATURES OF GENERAL BIOLOGICAL INTEREST
To draw up a useful formal summary of a work of this kind would be difficult, and perhaps even
impracticable. It is a long report, but it is in itself little more than a collection of summaries, for it
deals only with the more important aspects of a body of data too vast to be analysed in full at present ;
and for some species the text is already compressed to only a few lines. The present section, however,
may serve a useful purpose in focusing attention on some of the features of general biological interest
that have emerged. These have already been discussed as fully as the data allow, in the notes on the
several species. First, however, I take the opportunity to repeat, in as near to summarized form as
possible, the essential features of the environment described in the introductory sections. These must
constantly be borne in mind if the other features are to be properly appreciated.
The Patagonian Continental Shelf is the largest expanse of sea shallow enough to support a con-
siderable population of demersal fishes, in the cold temperate zone of the southern hemisphere. That
part of it which we surveyed, between 42 and 52° S lat., is rather larger than the North Sea. The main
physical features of this area are :
(i) Uniformly moderate depths over the ' plain of the shelf with very slight depth gradient as one
proceeds offshore from the 80 m. to the 200 m. contour. Abrupt descent to oceanic depths ' over the
edge'. (This means that there is but a very small area below 200 m. accessible for trawling, despite
the great width of the shelf.)
FEATURES OF GENERAL BIOLOGICAL INTEREST 383
(2) Rough ground with coralUne hydroids and coarse deposits prevails to the south, and in many
places along the shelf edge. Finer deposits are found to the northward, with better conditions for
trawling.
(3) The main current affecting the area is the relatively cold Falkland current flowing from south
to north, coldest along its offshore margin. The warm Brazil current impinges upon this offshore in
the extreme north-east of the area, where the hydrological conditions may be very complex, but this
is too far offshore normally to affect conditions on the shelf. On the inshore flank of the Falkland
current (which is a movement of sub-Antarctic surface water) the flow is not so strong, so that ' old
shelf water ' is warmed and sometimes even flows southwards as a small intermittent counter-current
close in to the land. This seems to have an important effect on the distribution and movements of
some of the fishes.
(4) The annual cycle of surface temperatures is centred lower than in corresponding latitudes in the
northern hemisphere, with the peak after mid-summer; and the annual range is small. At greater
depths, while the range of temperature is of course even smaller, the time lag between sea and air
temperatures is even greater, so that bottom temperatures are highest in autumn and lowest in spring
or early summer.
Some preliminary observations on plankton conditions have been discussed, but the large collections
obtained will require many years working up by specialists before one could fully consider the inter-
actions of this part of the biological environment with the fish fauna. The same must be said of the
copious collections of benthos. However, with regard to the plankton conditions, at least, some
important features are already fairly clear. While the general facies of the plankton is closely similar
to that found in cold temperate waters in other parts of the world, two major differences are already
discernible.
First, with regard to the phytoplankton conditions (the basic element in the life cycles of any sea
area) it seems reasonably certain that the onset of the main increase falls later in the year than in
corresponding latitudes in the northern hemisphere, especially along the outer edge of the shelf.
Production probably begins earliest inshore, owing to more favourable conditions for the establish-
ment of a thermocline, but even so the cycle is later than (say) on the hake grounds south of Ireland.
Dependent cycles of higher organisms must therefore also be centred later in the year, the whole
' plankton-calendar' (Bogorov, 1941) being later. Even the larger nekton (fish and squids), all of which
must be affected directly or indirectly by the plankton conditions, must be expected to show a similar
' lateness ' in their biological seasons. Direct evidence of this is already forthcoming in a few instances,
as in the apparent spawning times of some of the fishes. Species apparently corresponding to spring-
spawners elsewhere here seem to spawn around mid-summer, and whereas hake may generally be
regarded as summer spawners the local species has been found ripe chiefly in early autumn, although
some had certainly spawned earlier.
Secondly, with regard to the zooplankton, although euphausians, Parathemisto, and calanoid cope-
pods are prominent among the larger Crustacea, as elsewhere, there is a notable local abundance of
' lobster-krill ', the pelagic post-larvae of two species of Munida. This is paralleled elsewhere only in
New Zealand waters (where one of the same species occurs) and on some parts of the Pacific coast of
North and South America. Off Patagonia lobster-krill are an important food of whales and seals, birds
and fishes, especially the younger individuals of the larger fishes, that later become almost exclusively
ichthyophagous, like the hake. (This is equally true of Munida in New Zealand waters, where they
are commonly referred to as 'whale-feed'. See Matthews (1932), and Rayner (1935).)
Proceeding to general consideration of the fish fauna, as a whole : its chief characteristics have been
illustrated by comparison and contrast with the faunas of better known fishing-grounds elsewhere.
384 DISCOVERY REPORTS
Clupeoids are important, but our methods, aimed at investigating the ground-fish, could not demon-
strate their relative quantity, even roughly. It is very clear however that they are a staple food of the
larger mid-water and ground-fish, and, indeed, of all larger creatures in the region capable of swallowing
them.
Systematically, the two most prominent groups of ground-fishes were the Nototheniiformes and
Zoarcidae. The Nototheniiformes are a large group whose metropolis is the Antarctic Zone, but the
Patagonian species, although concentrated mainly on the southern part of the shelf, show only two
examples common to both regions. This suggests that dispersal between the Antarctic and sub-
Antarctic Zones has for long been difficult, and the hydrological barrier presented by the Antarctic
convergence may be the prime factor involved. The Zoarcidae (eel-pouts) prominent elsewhere in the
North Pacific and in the Antarctic Zone, provide a large number of species, but were taken in such
small numbers that they are thought to be of little ecological importance.
Two Merlucciidae, a true hake and the ' long-tailed hake ' (Macruronus), were dominant among the
larger fishes, and were surpassed in numbers only by the most widespread of the Notothenias
(N. ramsayi), a much smaller species of little potential value to man. Gadidae and Heterosomata were few
and small. Elasmobranchs also were less important than on most other roughly comparable grounds,
dogfish being particularly scarce. The Rajidae provided a profusion of local species difficult to deter-
mine, but their numbers were few and sizes small. In general the fish fauna resembled that found in
the North Pacific more closely than that of other known fishing grounds, if we regard the Notothenii-
formes as filling an ecological niche similar to that occupied by the smaller ' rock-fishes ' of the North
Pacific. Jordan (1905, il, p. 501) had pointed to this last similarity. There are two outstanding and
discouraging differences between the two areas: the North Pacific abounds in salmon and sizeable
Heterosomata, but on the Patagonian shelf, salmon are absent and the only common flatfish too small
to be of use. However, the absence of salmon is a drawback shared by the whole of the southern
hemisphere, except where small-scale artificial introductions have been made.
Apart from the hakes, another Patagonian fish is to be found in some plenty in summer, and might
eventually be exploited with profit. This is the 'spotted pomfret' (Stromateus maciilatus). Near
relatives of this species, whose commercial value has already been proved, are the butterfish of north-
eastern U.S.A. and the silvery pomfret of China seas.
The fish fauna of the Patagonian Continental Shelf is notably poorer in species than that of roughly
comparable areas elsewhere : nearly twice as many species are recorded from the Gulf of Maine, and
more than three times as many from British seas. Unfortunately, it would seem that there is a corre-
sponding lack of quantity of fish that can be trawled, apart from the three most promising species.
The study of distribution, migrations and general biology of demersal fishes has proceeded apace
wherever commercial fisheries are pursued. A vast body of evidence concerning these subjects has
accrued, but, with notable exceptions (such as Meek, 191 6; and Kyle, 1926), little attempt at synthesis
has been made. Certain broad similarities of behaviour, especially in respect of migration, are common
to many demersal fishes of the most diverse philogenetic origin. One can hardly call these tendencies
'rules', for there are too many obvious exceptions, but they are followed more or less closely by a
majority of the species inhabiting extra-tropical waters, and thus provide an invaluable aid to the
ordering of our thoughts on the bionomics of the several members of a fish fauna. Fisheries workers
will be very familiar with the tendencies I have in mind, and many more beside, but it is desirable to
state a few of them here, because they form a convenient guide to the general biological usefulness of
the data presented in the main body of this report, that is to say the notes on the individual species.
The main value of such work, from the general biological point of view lies, not so much in discoveries
of odd individual deviations from ' normal ' behaviour, as in seeing how far most members of a fauna
FEATURES OF GENERAL BIOLOGICAL INTEREST 385
not previously studied or exploited, conform to tendencies recognizable elsewhere. (By this means
lesser divergencies, such as time-lag, usually become clearly attributable to the differing environmental
conditions). To proceed :
Many demersal fishes show seasonal movement shorewards in summer, and offshore to deeper water in
winter. Usually, but not always, the shoreward movement is a breeding migration.
This is one of the best known tendencies of fish migration. Examples among Patagonian species are
furnished by fishes as diverse as hake {Merluccius hubbsi), spotted pomfret {Stromateus maculatus), a ray
{Raja brachyiirops) and Notothenia ramsayi.
In a migratory species of ground-fish, the larger individuals tend to travel farther and faster than the
smaller ones. {Often the movement of immature individuals is so limited that they never proceed ' over the
edge' to the greater depths reached by the adults during the ' off' season).
Very good examples of this are provided by the four species mentioned above.
A similar effect may be apparent interspecifically among the members of a taxonomic unit. Thus
hake migrate more extensively than the slighter Macruronus; and Notothenia ramsayi, the largest of
the Patagonian members of its genus-, migrates farther than any of the others.
In temperate latitudes many migrating species of demersal fishes extend their range polewards in summer
and towards the equator in winter. This may result directly from the effect of temperature, but is ?iot {as
yet) clearly to be distinguished from secondary effects of the strength and direction of the currents {the factor
most strongly emphasized by Meek). Thus it is essential for fish with denatant pelagic larvae, inhabiting a
region like the Patagonian Continental Shelf, where the main current flows from south to north, to spawn
near the southern limit of their range {or so close inshore that the larvae are drifted by the inshore counter
current to the southward) if the species is to be maintained zvithin its ecological norm {cf. E. S. Russell,
1937. P- 321)- ^ considerable meridional trend of seasonal movement may thus be superimposed upon the
on- and offshore movement, and may even almost completely mask the latter.
Considerable meridional trends of movement are shown by such fishes as Stromateus, Macruronus
and Genypterus in the Patagonian region. With Stromateus such trends are superimposed upon a strong
on- and oflFshore movement of the usual type, and seem almost certainly to be conditioned by the
currents as outlined above. Macruronus gave more indications of meridional than of on- and oflFshore
movement, and this may be partly due to its marked preference^ for moderate depths of water. Some
tendency to oflFshore movement was discernible, but the fish was never plentiful at such great depths
as those in which true hake are sometimes captured. It is very much a fish of the plain of the shelf,
and even the largest individuals rarely seemed to go far 'over the edge'. Since Macruronus sttms to
spawn in early summer, the southward movement cannot be regarded as a breeding migration. It is
perhaps a feeding migration, in which shoals of Falkland herring are the attraction. (Our largest
concentrations of this species were discovered near the southern limits of its range early in autumn.)
With Genypterus diflFerent factors are probably involved. The area investigated is probably south of
the main habitat of the species. Our catches certainly suggest an ellipsoidal path of movement similar
to that found in Stromateus, but the individuals concerned should probably be regarded as straggling
adolescents, and considerations of the necessity for contranatant movement before spawning do not
arise if this view is correct.
The coastal elasmobranch Callorhynchus showed evidence of southward movement in summer. It
keeps so close to the 80 m. line (limit of the 'first slope') that such progress would be aided by the
inshore counter-current, but it may well be that temperature is the more important factor here. We
do not yet know where this species breeds.
1 Using a teleological expression, in the interests of clarity and brevity.
386 DISCOVERY REPORTS
Demersal fishes of extensive depth range tend to show larger individuals in deeper zvater, except when
this tefidency may temporarily be masked by seasonal movements. A similar tendency can sometimes be
discerned interspecifically , but there are many obvious exceptions {such as large, exclusively littoral species)
if purely taxonomic units are considered.
Most of the fish for which we have any adequate data demonstrate the first (intra-specific) part of
this proposition very clearly. The second part is also clear when suflicient data are to hand to permit
ecological rather than taxonomic groupings to be considered, and is very well shown by the depth
relations of Patagonian Nototheniiformes.
Fishes of wide latitudinal range often show a gradation of size with latitude, the individuals being larger
towards the polar limits of the range of the species. This implies either that the stocks are different in different
latitudes {with different spazvning times) as has been shown for European hake; or that only the larger
individuals penetrate towards the polar limits of the range of the species.
The first type of correlation between size and latitude seems to be exemplified by our Patagonian
hake, which consistently showed larger fishes as one proceeded south, but in the movements of which
little meridional trend could be detected. But to the extreme south of the species' range only a very
few outsize female individuals were taken. These can hardly be a breeding stock and are thought to
be exceptionally aged individuals that may have lost the seasonal urge to migration through reduced
reproductive activity.
Macrurotius also showed the largest individuals to the southward at most seasons, but with this
species a considerable meridional trend of movement seems certain, so that the size latitude correlation
is probably a transient one, consequent upon the greater speed and strength of the larger fish, which
may enable them to proceed farther afield in the search for food.
In many fishes a marked change in diet occurs zvith increasing size {age). Often this may coincide with
the attainment of sexual maturity and jor greatly increased migrations (cf. Harold Thompson, 1943, p- 86,
on Nezvfoundland Cod; Hartley, 1945, pp. n, 26, on British fresh-water fish).
This change of diet is very well shown by the larger species of the shelf, Macruronus and hake. The
smaller individuals seem to feed mainly upon macroplanktonic Crustacea, the larger ones upon fish
and squids. Of the more definitely bottom-feeding species it is practically certain that Notothenia
ramsayi will be found to show an analogous change, from a diet of mysids, benthic Crustacea and
polychaetes to one of larger prey — fish, especially Falkland herring. Our data are insufficient to
establish the point beyond question. The probable coincidence of change in feeding habits with onset
of sexual maturity and increased migration is particularly well shown by Macruronus.
In a large majority of fishes, females are larger than males; and this is usually sufficient to account for
the slight abnormalities of sex-ratio {preponderance of females) commonly encountered in net-caught
samples. Marked abnormalities in this or the other directioti are usually due to peculiar seasonal trends
towards unisexual shoaling, and are zvell known in certain elasmobranchs.
Nearly all the Patagonian species for which we have adequate data showed the females to be larger
than the males. Usually the difference was small but strongly significant. The single possible exception
to this ' rule ' was Cottoperca gobio (Bovichthyidae) where the males seemed to be the larger. More
data would probably dispel this apparent anomaly. The Patagonian hake {Merluccius hubbsi) shows
a most unusual discrepancy in size between the sexes, the males being very much smaller than the
females (much more so than in European hake). With a few species strongly abnormal sex-ratios were
encountered that suggest possibilities of unisexual shoaling (Macruridae, Micromesistius australis).
387
PROSPECTS OF COMMERCIAL DEVELOPMENT
THE WEIGHT OF CATCHES
The best basis for any consideration of the possibiUties of developing commercial fisheries in the area
surveyed is the weight data collected in 193(^1. In the endeavour to present this mtelligibly but at
the same time in sufficiently concise form, I have divided the entire contents of all roughly comparable
hauls for which weight data are available into the categories shown below, with brief notes on which
each represents :
Categories employed in
further considerations
of weight data
Notes
Elasmobranchii
Merluccius
Maauronus
Gadidae
Nototheniiformes
Thyrsites
Genypterus
Stromateus
Other fishes
Total fish
Lithodes
Squids
'Rubbish'
Since dogfish are rare, this category consists almost entirely of the small rays and skates of the
region, with a few Callorhynchus at some inshore stations. Not important
Hake (Merluccius hubbd), smaller than the European species and nearest to M. bilinearis, the
'whiting' of New England. Very important, and would be the staple of any trawlmg on the
shelf that could be developed
'Long-tailed hake' Macruronus magellanicus, allied to the true hake. Its 'rat-tailed' appear-
ance might be against it, but it is very good eating and locally abundant at times. Important
Here comprise only Salilota, Micromesistius (roughly equivalent to the 'Scotch haddock' or
' forked hake' and the ' blue whiting' of British seas), and a i^v^imy Physiculus. Not important
The large southern group characteristic of the region, where they correspond ecologically to
the smaller 'rock-fishes' of the eastern North Pacific. Mostly too small to be ot value as
human food. Notothenia ramsayi, the principal trawled species, is sometimes so abundant
that it might be used for fish-meal if a fishery could be established on the more valuable
categories Unimportant, except as forage for hake, etc. Eleginops may prove useful in
small-scale inshore fishing, but was not trawled
Thyrsites atun, the 'snoek' of South Africa, 'barracouta' of Australia. Properly an inhabitant
of warmer seas; such stragglers as are taken oflF Patagonia would always be valuable tor they
are excellent food fish, but are too rare for the species to be considered potentially important
here
Genypterus blacodes, excellent food fish (the 'ling' of New Zealand, and very close to the
'king-klip' of South Africa) but probably too rare to be important here
'Spotted pomfret', Stromateus maculatus, close to the 'butterfish' of eastern U.SJV. and the
' silvery pomfret ' of China. A very good food fish, locally abundant in due season. Important
Unimportant, though the minute 'scald-fish' Thysanopsetta was numerically abundant in the
intermediate region and is eaten by hake. Such odd herrings and rattails (Macruridae) as
are included in this category, chiefly in the southern region, could be utilized
Useful for comparisons of the relative potential of the three regions, and the relative amounts
of rubbish present
Lithodes antarcticus, one of the more abundant of the larger invertebrates, and prized as food
in South America, where they are known as 'Centolla crabs'. Of some potential value
Several species. Some are eaten in South America, and they would be valuable as bait if any
line fishing were developed. Imponant as food for some of the larger fish
The remaining contents of the trawl-invertebrates of no foreseeable value sand stones, etc.
Of considerable negative importance, indicating where the greatest loss through damage to
gear and time spent in sorting may be expected in any fishery that might be developed
It is imperative to bear in mind that a commercial vessel, fishing with similar gear in the same area
as that surveyed by the ' William Scoresby', could reckon on much better catches. A surveying vessel
needs to cover and re-cover as much ground as possible, for from the point of view of the survey it is
iust as important to determine where and when the fish are not to be found, as it is to find out where
they are A commercial vessel, on the other hand, will continue to work as long as possible wherever
the fishing seems best. Consequently it is safe to assume that a commercial vessel would average far
greater weights of fish per hour's trawling-probably at least twice as great-as did the ' Scoresby .
388
DISCOVERY REPORTS
In the northern region we have roughly comparable data from fourteen hauls, totalling 23 hr.
fishing. The abstracted weight analysis is shown in Table 40.
Table 40. Analysis of abstracted weight data from the northern region
Hauls considered: WS788, 789, 790A, 790B (4 hr.), 791 A, 791 B (4 hr.), 792A, 792B (4 hr.), 793, 853, 855, 8S9A,
859 B, 858.
"/
Total
% of total
Weight
per hour's
trawling
kg.
Weight per
Category
Presence
/o
occurrence
weight
kg.
weight (of
fish only)
hour as
%
Elasmobranchii
10/14
71-4
26-498
1-74
1-152
1-63
Merluccius
14/14
loo-o
1260-795
82-84
54-817
77-57
Macruromis
13/14
92-8
188-900
12-41
8-213
11-62
Gadidae
4/14
28-6
0-520
0-04
0-023
0-03
Nototheniiformes
13/14
92-8
19-360
1-27
0-842
1-19
Thyrsites
0/14
0-0
0-000
0-00
o-ooo
000
Genypterus
3/14
22-2
1-800
0-12
0-078
o-ii
Stromateus
10/14
71-4
16-589
1-09
0-721
1-02
Other fishes
10/14
71-4
7-438
0-49
0-323
0-46
Total fishes
14/14
loo-o
1521-916
100-00
66-170
93-63
Lithodes
11/14
78-6
33725
—
1-466
2-07
Squids
10/14
71-4
23-983*
—
1-043
1-48
Rubbish
13/14
92-8
45-858 + t
—
1-994
2-82
* Calculated from six weighing out of ten stations at which squids were caught,
••f Calculated from twelve weighings out of thirteen stations at which rubbish was present.
A hake fishery in this region might just show a profit if markets comparable to those in Europe were
available. The 'WiUiam Scoresby' obtained an average of just over i cwt. (51 kg.) of hake per hour.
A commercial vessel could reasonably expect to catch twice as much, and an experienced skipper,
Capt. Drennan, quoted by Hickling (1927, p. 10) has estimated that some 150 cwt. per 100 hr. was
the minimum necessary to enable a big trawler to pay her way. There was, however, very little in the
rest of the catch here to supplement the hake, and though relative absence of ' rubbish ' would make
for few repairs and easy sorting, we have to remember that these hake are smaller than the European
species. Most of them might correspond to (say) 'inters' at Milford Haven or 'ordinary chats' at
Fleetwood. Very few would equal the larger British trade categories. The big difficulty, of course,
which at present appears insurmountable, is the absence of a suitable market.
From the intermediate region there are data from twenty-nine hauls totalling 44 hr. trawling,
shown in Table 41.
Here the prospects are very poor, for less than half the weight of hake per hour that we had captured
in the northern region was taken, and a slight increase of Stromateus and Genypterus among the better
sorts was far too small to compensate for the shortage of hake. Moreover, a ruinous amount of rubbish
came up in the trawl here.
In the southern region data are available from forty hauls, totalling 60 hr. trawling, shown in
Table 42.
Here there were even fewer hake, but they included all the largest individuals of the species that
we captured, as can be seen from the detailed figures discussed elsewhere. Two of the best of the other
categories, Macruronus and Stromateus, showed their greatest relative abundance in this region. These
were both much more concentrated locally than the hake, so that commercial vessels seeking them
could be sure of far greater quantities than we captured while trying to sample the whole area. Several
categories — Gadidae, Nototheniiformes and Thyrsites — showed greater abundance in this region (and
individuals of greater size). Though the total quantities of these are very small they would all help to
PROSPECTS OF COMMERCIAL DEVELOPMENT
389
eke out catches of the more plentiful species. Even the ' Other fishes ' here include species that could be
utilized, e.g. herring and Macrurids. However, it cannot be said that the trawling prospects were at all
encouraging, taking the southern region as a whole. The proportion of rubbish that came up in the
trawl was nearly as great as in the intermediate region, and prevalence of foul ground leads to much
time lost in mending nets, apart from loss of gear through occasional serious damage. Of this our
field workers had bitter experience.
Table 41. Analysts of abstracted weight data, intermediate region
Hauls considered: WS773, 774, 775, 776, 784, 78SA, 785B, 785C, 786, 794, 795, 796A, 796B (4 hr.), 797B, 797C
(4 hr.), 798, 799A, 799B (4 hr.), 800 A, 800B (4 hr.), 801, 807, 808, 809A, 809B (4 hr.), 810, 857, 862, 864.
Category
Presence
/o
occurrence
Total
weight
kg.
% of total
weight (of
fish only)
Weight
per hour's
trawling
kg-
Weight per
hour as
/o
Elasmobranchii
18/29
62-1
24.186
2-04
0-550
1-28
Merluccius
26/29
897
853-540
71-95
19-399
45-03
Macruronus
15/29
517
162.730
13-72
3-698
8-58
Gadidae
11/29
37-9
11-430
0.96
0-260
o-6o
Nototheniiformes
28/29
96.6
52-478
4-42
1-193
2-77
Thy r sites
0/29
0-0
Q.OOO
O.QO
o-ooo
0-00
Genypterus
7/29
24-1
12-355
1-04
0-280
0-65
Stromateus
15/29
Si-7
55-570
4-68
1-263
2-93
Other fishes
18/29
62-1
14-004
i-iS
0-318
0-74
Total fishes
29/29
—
1186-293
99.99
26-961
62-58
Lithodes
16/29
55-2
39-957*
—
0-908
2- II
Squids
, 27/29
93-1
56-344t
—
1-280
2-97
Rubbish
29/29
lOO-O
613-031 + J
—
13-932
32-44
* Calculated from fifteen weighings out of sixteen hauls containing Lithodes.
f Calculated from fifteen weighings out of twenty-seven hauls containing squids.
X Calculated from twenty-eight weighings out of twenty-nine hauls containing rubbish.
Table 42. Analysis of abstracted weight data, southern region
Hauls considered: WS802A, 802B, 803, 804A, 804B, 805, 806, 811 1, 811 II (4 hr.), 812I, 812II (4 hr.), 813, 814,
!i5, 816, 817A, 817B (4hr.), 818A, 818B (4hr.), 819A, 819B (4 hr.), 820, 824, 825, 833, 834, 838, 839, 847 A, 847B
(4hr.), 848, 849, 850, 851,
S66, 868, 870, 872, 874, 875.
0/
Total
% of total
Weight
per hour's
trawling
kg-
Weight per
Category
Presence
/o
occurrence
weight
kg.
weight (of
fish only)
hour as
%
Elasmobranchii
20/40
50-0
54-115 .
2-00
0-902
1-34
Merluccius
26/40
65-0
447-860
i6-55
7-464
11-07
Macruronus
21/40
52-5
1257-435
46-48
20-957
31-08
Gadidae
29/40
72-5
258-367
9-55
4-306
6-39
Nototheniiformes
34/40
85-0
89-374
3-30
1-484
2.20
Thyrsites
5/40
12-5
158-000
5-84
2-633
3.90
Genypterus
9/40
22-5
14-070
0-52
0-234
0-35
Stromateus
19/40
47-5
256-750
9-49
4-279
6-35
Other fishes
28/40
70.0
169-546
6-27
2.826
4-24
Total fishes
—
—
2705-517
loo-oo
45-085
66-86
Lithodes
21/40
52-5
68-184*
—
1-136
1-68
Squids
21/40
52-S
77-2i5t
—
1-287
1-91
Rubbish
40/40
100. 0
1195-443!
—
19.924
29-55
* Calculated from twenty weighings out of the twenty-one hauls in which Lithodes was taken.
■f Calculated from seventeen weighings out of the twenty-one hauls in which squids were taken,
j Calculated from thirty-eight weighings out of the forty hauls.
390
DISCOVERY REPORTS
A reshuffling of these weight data, with the addition of corresponding numerical values for the fish
categories, is given in Tables 43 and 44. These permit direct comparison and contrast of the three
regions, considered as possible fishing grounds ; but for a full appreciation of the possible implications
of these, the reader must refer to the general account of the region given in the introduction, and to the
individual accounts of the more important species given in the body of the paper. It is hoped that the
table of contents will make it possible to do this without the necessity for reading through the whole
report.
Table 43. Further analysis of zveight data: occurrence, relative numbers and relative zveights of
the main fish categories in each of the three regions
Main categories
of the fish fauna
Frequency of occurrence
(% of hauls positive)
% by numbers of each
category of totals for each region
% by weight of each category
of total weight for each region
N. region
I. region
S. region
N. region
I. region
S. region
N. region
I. region
S. region
Elasmobranchii
Merlucciiis
Macruronus
Gadidae
Nototheniiformes
Thyrsites
Genypterus
Stromateus
Other fish
714
lOO-O
928
286
92-8
22-2
714
71-4
62-1
89-7
51-7
37-9
96-6
24-1
51-7
62-1
50-0
65-0
52-5
72-5
85-0
12-5
22-5
47-5
70-0
0-87
4700
23-61
024
26-01
o-io
114
1-02
2-82
23-21
16-93
2-23
19-44
0-33
4-81
30-23
1-38
7-47
47-32
12-47
9-23
0-82
0-22
13-28
7-81
1-74
82-84
12 41
0 04
1-27
0-12
1 09
049
2-04
71-95
13-72
0-96
4-42
1-04
4-68
1-18
2-00
16-55
46-48
9-55
3-30
5-84
0-52
9-49
6-27
Table 44. Further analysis of weight data: mean numbers, mean weights and relative zveights per
hour's trawling of the main categories, in each of the three regions
Main categories of fish,
extras and rubbish per
hour's trawling
Mean numbers of fish (per
hour's trawling)
Mean weight of fish, etc., per
hour's trawling (kg.)
% by weight per hour's
trawling of fish, extras and
rubbish
N. region
I. region
S. region
N. region
I. region
"S. region
N. region
I. region
S. region
Elasmobranchii
187
2-50
1-23
1152
0-550
o-go2
163
1-28
1-34
Merlucciiis
100-48
20-59
6-67
54817
19-399
7-464
77-57
45-03
11-07
Macruronus
5048
15-02
42-23
8213
3-698
20-957
11 62
8-58
31-08
Gadidae
052
1-98
11-13
0023
0-260
4-306
003
0-60
6-39
Nototheniiformes
55-61 1 17-25
8-23
0 842
I-I93
1-484
1-19
2-77
2-20
Thyrsites
—
—
0-73
—
—
2-633
3-90
Genypterus
0-22
0-30
0-20
0-078
0-280
0-234
O-II
0-65
0-35
Stromateus
243
4-27
11-85
0 721
1-263
4-279
1-02
2-93
6-35
Other fish
2-17*
26-82t
6-97X
0323
0-318
2-826
0 46
0-74
4-24
Total fish
213-78
88-73
89-24
66-170
26-961
45-089
9363
62-58
66-86
Lithodes
—
—
—
I 466
0-908
1-136
2-07
2- 1 1
1-68
Squids
—
—
^
1043
I -208
1-287
1-48
2-97
1-91
Rubbish
—
—
—
1994
13-932
19-924
2-82
32-44
29-55
* Almost negligible in northern region except for Palinuricthys (useful) at one station.
f Mainly large numbers of the tiny flatfish Thysanopsetta, too small to be of value.
I Of some importance in the southern region — sizeable Macrurids, Falkland herring, Parana signata.
CONCLUSIONS
In the exploration of natural resources the primary function of the naturalist is to provide fundamental
information on the nature, quantity and accessibility of the raw material. Thereafter the administrator
and technologist are in a better position to assess the prospects of commercial development. It is felt,
however, that a report such as this would be incomplete without some practical suggestions from those
who have collected and collated the biological data. In the remarks which follow, I am confident that
PROSPECTS OF COMMERCIAL DEVELOPMENT 39'
where our evidence is already good, the opinions expressed are shared by colleagues who were directly
engaged in collection of the data at sea. The more tentative suggestions are my own (T. J. H.).
The primary object of these investigations was to provide information, upon which the prospects
of carrying on any commercial fishery from the Falkland Islands could be assessed. It must be
plainly stated that the results are not encouraging; but this is due to economic and geographical
factors, rather than to lack of suitable fish. The best trawling grounds are not very near to the Falkland
Islands, but it can be shown with reasonable certainty that on the shelf to the northward, roughly
equidistant from the Falkland Islands and the lesser Argentine ports, there is a stock of hake just
suflScient to enable a modern trawler to pay its way if there were markets equivalent to the British ones
within a few hundred miles.
The population of the Falkland Islands is too small and too scattered (with limited means of com-
munication between the settlements) to enable a large trawler to pay its way on local trade alone. If
a considerable part of the catch could be sold in, for example, the Argentine at a reasonable price, a
trawler working from Port Stanley might be able to keep the latter supplied with the results of, say,
one voyage in four. The possibilities of smoking, drying and dehydration would no doubt be taken
into consideration, but it seems doubtful whether the fish could be marketed at an economic price
in the Argentine. However, I venture to suggest three possibilities, on the strength of the knowledge
of the fish-fauna that we have gained.
I Local inshore seining for 'mullet' (Eleginops), 'smelts' {Austromenidia) and such other species
as present themselves. Dr Kemp informed me shortly before his death that our former colleague
Dr J. E. Hamilton was even then trying to establish some inshore fishery in the Falklands. Much
might be done to place such a scheme on a permanent footing if a small-scale canning plant could be
established. This could deal with an occasional glut of 'herring' (Clupea fuegensis) but might aim
primarily at developing a small luxury trade in canned Centolla crab {Lithodes), serving to keep a few
hands permanently employed. It is not yet known for certain that these crabs would be accessible m
sufficient quantity to small coastal craft, but we found encouraging numbers of them in the trawl on
the rough ground round the islands, that would not support ordinary inshore trawhng. The main
part of this scheme would aim at providing some fresh fish for local consumption (a real need). The
canning is a further suggestion to aid in keeping it on a self-supporting basis, which could hardly be
hoped for from small-scale seining alone.
II Exploitation of Clupea fuegensis, possibly by some form of purse-seining, for drift nets or other
forms of gill-nets would almost certainly suffer too much from damage by seals and birds to make
them workable in the Falkland area. Such a scheme would depend upon canning, production of fish-
meal or other means of processing the product. As already explained the trawl could not provide
adequate evidence of the quantities of these small semi-pelagic fish available, so that further, possibly
costly investigations would be needed, before one could form an adequate opinion as to the feasibility
of such a scheme. , -r m ■ i
HI Part-time trawling. If a cold store were available in Port Stanley, and if sufficient employment
could" be found for a suitable vessel (possibly on inter-island communications) during more than half
her time a modern trawler occasionally working the hake grounds we found to the north could easily
keep Port Stanley on a full supply of fish; but it is very doubtful whether she would pay her way
at this.
392 DISCOVERY REPORTS
REFERENCES
Barnard, K. H., 1925. A monograph of the marine fishes of South Africa. Ann. S. Afr. Mus., xxi.
Belloc, G., 1922. Note sur la croissance du Merlu, variations ethniques et sexuelles. Rapports et Proces-Verbaux des Reunions,
Cons. Perm. Internal. Explor. Mer, xxxi, 2, 11, pp. 34-43.
Belloc, G., 1937. Note sur la presence du Merlu dam les eaux de la presqu'ile du Cap Verte. Rev. Trav. Peches Maritimes,
Paris, X, pp. 341-6.
Bigelow, Henry B. and Welsh, William W., 1925. Fishes of the Gulf of Maine. Bull. U.S. Bur. Fish., xl (for 1924).
BoGOROV, B. G., 1941. Biological seasons in the plankton of different seas. Comptes Rendus (Doklady) de I'Academie des
Sciences de I'U.R.S.S., xxxi (4), pp. 404-7, figs. 1-2.
BoNDE, Cecil von, 1934. Report No. 11, for the year ending December 1933. Fisheries and Marine Biological Survey, Union
of South Africa. Pretoria: Govt. Printer.
Bordale, L. F. See Pozzi and Bordale (1935).
Couch, Jonathon, 1864. A History of the Fishes of the British Isles. 4 vols. London: Groombridge and Sons.
Day, Francis, 1880-84. The Fishes of Great Britain and Ireland. 2 vols. London and Edinburgh: Williams and Norgate.
Deacon, G. E. R., 1937. The Hydrology of the Southern Ocean. Discovery Reports, xv, pp. 3-123, pis. i-xliv.
Devincenzi, Garibaldi, J., 1926. Feces del Uruguay. Anales del Museo de Hist. Natural, Serie 2a, T. i, pp. 272-4. Monte-
video, 1924-6.
Director of Fisheries, 1938. Divn. of Fisheries : Fifteenth Annual Rep. for the year ended Dec. 1937. Off. Journ. Dept. Comm.
Ind. I, No. 3. Pretoria; Union of South Africa Govt. Printer.
Director of Fisheries, 1939. Divn. of Fisheries: Sixteenth Annual Rep. for the year ended Dec. 1938. Ibid. 11, No. 3.
Eigenmann, Carl H. See Jordan and Eigenmann (1890).
Evermann, Barton Warren and Radcliffe, Lewis, 1917. The fishes of the west coast of Peru and the Titicaca basin. Bull.
U.S. Nat. Mus., xcv, pp. i-xi, 1-166, pis. 1-14.
Faber, G. L., 1883. The Fisheries of the Adriatic and the Fish thereof. A report of the Austro-Hungarian sea-fisheries, with a
detailed description of the marine fauna of the Adriatic Gulf. Pp. i-xxiv, 1-292, 24 pis., 18 woodcuts. London: Bernard
Quaritch.
Fiedler, R. H., 1943. Fishery Statistics of the United States, 1940. U.S. Dept. of the Int., Fish and Wildlife Service, Statistical
Digest, No. 4. Washington.
Fisheries Newsletter, 1941. The Barracouta. Vol. i, Part i, p. 10. Issued by Divn. of Fisheries, C.S.I. R. Australia.
Fisheries Newsletter, 1942. Vol. i. Part 2, p. i.
Fisheries Newsletter, 1944. Australian Food Fishes. V. The Barracouta. Vol. iii. Part 5, pp. 2, 3; Dept. War Organ. Ind.
C.S.I.R. Australia.
Ford, E., 1921. A contribution to our knowledge of the life histories of the dogfishes landed at Plymouth. Journ. Mar. Biol.
Assoc, N.S. XII, pp. 468-505.
Ford, E., 1938. Vertebral variation in teleostean fishes. Journ. Mar. Biol. Assoc, xxii, pp. 1-60, xvi pis., 18 figs.
Garman, Samuel, 1904. The Chimaeroids (Chismopnea Raf., 1815; Holocephala Midi., 1834), especially Rhinochimaera and
its allies. Bull. Mus. Comp. Zool. Harvard, XLi, No. 2, pp. 245-72, pis. 1-15.
Garman, Samuel, 1911. The Chismopnea {Chimaeroids). Mem. Mus. Comp. Zool. Harvard, xl. No. 3, pp. 81-101.
Gilchrist, J. D. F., 1914. Marine Biological Rep. No. 2, for the year ending t,o June 1914. Pp. 1-167, i, ii; 2 charts. Pro-
vince of the Cape of Good Hope, Union of South Africa.
Gilchrist, J. D. F., 1916. Ibid., No. 3, pp. 1-135, •> "•
Goode, G. Brown and Associates, 1884. The Fisheries and Fishing Industries of the United States. Section I: Natural History
of useful Aquatic Animals. Text. [No. 72. The Silver Hake and the Merluccio, pp. 240-3.] Atlas of 272 plates published
as a separate volume in the same year. Washington.
Gran, H. H., 1929. Investigations of the production of plankton outside the Romsdalsfjord, 1926-7. Rapports et Proces-Verbaux
des Reunions, Cons. Perm. Internat. Explor. Mer, lvi, 112 pp.
GiJNTHER, Albert, 1887. Report on the deep-sea fishes collected by H.M.S. 'Challenger' during the years 1873-6. Rep. Sci.
Res. Challenger, Zool, Vol. xxii.
GURNEY, R., 1933. British Freshwater Copepoda, ill. Ray Soc. Monograph 120. London.
Hamilton, J. E., 1934. The southern sea lion, Otaria byronia [De Blainville). Discovery Reports, viii, pp. 271-318, pis. i-xiii.
Hardy, A. C, 1924. The herring in relation to its animate environment. Part I. The food and feeding of the herring. Min. Agric.
Fish., Fish. Invest. Sen, 11, Vol. vii. No. 3.
Hardy, A. C. See Kemp, Hardy and Mackintosh (1929).
Hart, T. John, 1934. On the phytoplankion of the south-west Atlantic and the Bellingshausen Sea, 1929-31. Discovery Reports,
VIII, pp. 1-268.
REFERENCES 393
Hart T Tohn 1042. Phytoplankton periodicity in Antarctic surface waters. Discovery Reports, xxi, pp. 261-356.
Hartley, PH.' T 1940. t£ saltash tuck-net fishery and the ecology of some estuarine fishes. Journ. Mar. Biol. Assoc, xx.v,
HartTey!'r H.' Tt^!945'^rSfooi of Coarse Fish, being the Interim Rep. on the coarse fish investigation. Freshwater
Biological Association, Sci. Publ. No. 3, pp. 5-33. 5 Ags- xvii tables.
Harvey H W., 1934. Measurement of phytoplankton population. Journ. Mar. Biol. Assoc, N.S., xix, PP- 7f>i-73-
Hentsc^el Ernst 1936. Allgemane B^olog^e des Sildatlantnchen O.eans. Wiss. Ergeb. Deutsch. Atlantischen Exped.
'Meteor', 1925-7, Band xi. [Beilage ix-xlii published under separate cover, same year.] p , ,• u ^ K r A P
Herklots, G. a C and Lin, S. Y., 1938. Common Marine Food Fishes of Hong-Kong. 75 PP- 4i fig^- Published by G. A. C.
Herklots at the University, Hong-Kong. ■ „io ^ fi„c
HiCKLiNG C F., 1925. A new type of luminescence in fishes. Journ. Mar. Biol. Assoc, N.S., xiii pp. 914-3?. iv pis., 7 hgs.
HiCKUNG, C. F., 1927. The Naturai History of the Hake. Parts I and H. Min. Agric. Fish.. Fish. Invest. Ser., 11, Vol. x,
HiCKUNG,' C. F., 1928. The Fleetwood exploratory voyages for hake. Journ. du Conseil Internat. Explor. Mer, iii. No. i,
HiCKUNrc%., i930«. A contribution to the life-history of the spur-dog. Journ. Mar. Biol. Assoc N.S., xvi. pp. 529-76.
HICKUNG, C. F., 19306. The Natural History of the Hake. Part III. Min. Agric. Fish., Fish. Invest. Ser.. 11, Vol. xii,
Hicku°nJ,'c. F., 1931. A new type of luminescence in fishes. III. Journ. Mar. Biol. Assoc, N.S., xvii, pp. 853-67, iv pis.,
HickLn'g'c F..1933. The Natural Hrstory of the Hake. Part IV. Min. Agric Fish., Fish. Invest. Ser.. 11, Vol xni. No. 2^
HICKUNG, C. F., 1935 «. The Hake and the Hake F.hery, being the Buckland lectures for 1934. London: Edwin Arnold
HICKUNG%. F., 19356. Seasonal changes in the ovary of the ^mmature hake, Merluccius merluccius L. Journ. Mar. Biol.
Assoc, N.S., XX, pp. 443-62.
Hildebrand, Samuel F. See Longley and Hildebrand (1941).
HoDGKiNSON, E. R. See PhiUipps and Hodgkinson (1922).
Hodgson W. C. 5ce Wollaston and Hodgson (1929). „ m r .,
JIrvis. SoRMAN D., 193- The Fisheries of Puerto Ruo. U.S. Dept. Comm., Bur. Fish., Invest. Rep. No. 13. pp. t-41,
loHN d'dilwyn, 1936. The southern species of the genus Euphausia. Discovery Reports, xiv. pp 193-324-
JOHNSTON, R. M., 1891. The Fishes and FtsMng Industries of Tasmania, together wUh a revised hst of tndigenous speaes.
Paps.'Proc Roy. Soc Tasmania for 1890, pp. 22-46. , u-u ,^ n . ki ,.^ Po
Jordan David Starr. 1905. A gmde to the study of fishes. 2 Vols. Westminster: Archibald Constable and Co.
ORDAN DAVID Starr and EiGENMANN. Carl H.. 1890. A review of the Genera and Species of Serramdae found rn the waters
oMmmV« ««<^ £«ro/... Bull. U.S. Fish. Comm. VIII (fori888), No. 9, pp. 329-441. pls- Ix-lxix.
Kemp, S. W.. Hardy, AC. and Mackintosh, N. A.. 1929. Disc^ery investigations: objects, equipment and methods. Dis-
K.Zl.^ZTm''le ZZr^mungen .wischen Kap Horn und der La Plata-MUndung. Ann. Hydrogr. Mar. Meteorol..
K.Z:.ZoSo%t H^^^^^^ Band II. Die Bewegungsformen des Meeres. Pp. i-xvi, 1-766, 3 tables.
Kyle'hIrry M.; f926!'?A. Biology of Fishes, xvi + 396 PP- 77 %-. xvii pis. London: Sidgwick and Jackson.
L^ngL' W^Tan^d^'ni^^^^^^ Samuel F.. 194^. Systematic catalogue of the Fishes ofTortugas, Florida. Being
Vol XXXIV oJ Papers from the Tortugas Laboratory. Carnegie Inst. Washington Publ. 535. -» + 33- PP-, 34 pls- Wash-
L0Ri;!'?u;rf923. A List of the Fishes of Tasmania (Read Sept., 1922). Paps. Proc Roy. Soc. Tasmania for 1922,
MCCULLO^! Alan R., 1915- The migration of the jolly-tail or eel-gudgeon, Galaxias attenuatus, from the sea to fresh water.
Australian Zoologist. I (2), pp. 47-9-
Mapkintosh N a See Kemp, Hardy and Mackintosh (1929). , t^. d ..» „
Matthews LhIkison. 1932. Lobster-Krill: Anomuran Crustacea that are the Food of Whales. Discovery Reports, v.
MATTHEfs!L. HARRiloN, 1934- The Marine Deposits of the Patagonian Continental Shelf Discovery Reports, ix. pp. 175-206.
Meek,'1™r, 1916. The Migrations of Fish. Pp. vii-xviii. 1-427; -8 figs., xi pis. London: Edwin Arnold. ^^^
394 DISCOVERY REPORTS
Menzies, W. J. M., 1925. The Salmon, its Life Story, vi + 211 pp., 5 diags., 36 figs., photographs. Edinburgh and London:
William Blackwood.
MuNRO, I. S. R., 1943. Revision of Australian species of Scomberomorus. Mem. Queensland Mus., xii, 2, pp. 65-95, pis.
vi-viii.
Navarro et ah, 1943. La Pesca de Arrastre en las fondos del Cabo Blanco y del Banco Arguin {Africa Sahariana). [F. de P.
Navarro, de F. Lozano, J. M. Navaz, E. Otero, J. Sainz Pardo and others.] Trabajos No. 18, pp. 1-225, Lam. i-xxxviii.
Madrid: Minis. Mar. Inst. Espagnol de Oceangr.
Norman, J. R., 1937. Coast fishes. Part II. The Patagonian region. Discovery Reports, xvi, pp. i- 150, pis. i-v.
Norman, J. R., 1938. Coast fishes. Part III. The Antarctica Zone. Discovery Reports, xviii, pp. 1-105, pi. i.
Pearson, John C, 1932. Winter Trawl Fishery off the Virginia and North Carolina Coasts. U.S. Dept. Comm., Bur. Fish.,
Invest. Rep., No. 10, 31 pp.
Phillipps, W. J., 1919. Life history of the fish Galaxias attenuatus. Australian Zoologist, I (7), pp. 211-13.
Phillipps, W. J., 192 1. Notes on the edible fishes of New Zealand. N.Z. Journ. Sci. Tech., iv, pp. 114-25.
Phillipps, W. J., 1924. Notes on the life history of the New Zealand minnow {the adult whitebait) Galaxias attenuatus. N.Z.
Journ. Sci. Tech., vii, pp. 1 17-19.
Phillipps, W. J. and Hodgkinson, E. R., 1922. Further notes on the edible fishes of New Zealand. N.Z. Journ. Sci. Tech.,
v, pp. 91-7.
Pozzi, A. J. and Bordale, L. F., 1935. Cuadro sistemdtico de los peces fnarinos de la Repiiblica Argentina. Anal. See. cient.
Argent., cxx (4), pp. 145-89, i map.
Radcliffe, Lewis. See Evermann and Radcliffe (1917).
Rayner, G. W., 1935. The Falkland species of the crustacean genus Munida. Discovery Reports, x, pp. 209-45.
Read, Bernard E., 1939. Common Food Fishes of Shanghai. 52 pp., 32 figs. North China Branch, Royal Asiatic Society.
Shanghai: Mercury Press.
Regan, C. T., 1914. British Antarctic {' Terra Nova') Expedition, 1910. Fishes. Terra Nova Reports, Zoology, I (i), pp. 1-54,
13 pis., etc.
RowE, Anne. See Simpson and Rowe (1939).
Russell, E. S., 1937. Fish migrations. Biol. Rev. Vol. xii. No. 3, pp. 320-37, 4 figs. Cambridge.
Schmidt, J., 1930. The Atlantic cod (Gadus callarias L.) and local races of the same. Compt. Rend. Lab. Carlsberg, xviii,
p. 6. Copenhagen.
Scott, E. O. G., 1936. Observations on some Tasmanian fishes. Part III. (Read Sept. 1935.) Paps. Proc. Roy. Soc. Tasmania
for 1935, pp. 113-29.
Scott, E. O. G., 1938. Observations on fishes of the family Galaxiidae. Part II. (Read Nov. 1937.) Paps. Proc. Roy. Soc.
Tasmania for 1937, pp. 1 11-43, P's- xvii-xxvii.
Scott, E. O. G., 1941. Observations on fishes of the family Galaxiidae. Part III. (Read Nov. 1940.) Paps. Proc. Roy. Soc.
Tasmania for 1940, pp. 55-69, pi. ix.
Simpson, George Gaylord and Rowe, Anne, 1939. Quantitative Zoology, xv + 414 pp., 52 figs. New York and London-
Magraw Hill Book Co. Inc.
Steven, G. A., 1932. Rays and skates of Devon and Cornwall. II. A study of the fishery ; with notes on the occurrence, migra-
tions and habits of the species. Journ. Mar. Biol. Assoc, N.S., xviii. No. i, pp. 1-33.
Steven, G. A., 1933. Rays and skates of Devon and Cornwall. III. The proportions of the sexes in nature and in commercial
landings, and their significance to the fishery. Journ. Mar. Biol. Assoc, N.S., xviii. No. 2, pp. 611-25.
Steven, G. A., 1936. Migrations and growth of the thornback ray (Raia clavata L.). Journ. Mar. Biol. Assoc, N.S., xx No 3
pp. 605-14.
Taning, a. Vedel, 1937. Some features in the migration of cod. Journ. du Conseil Internat. Explor. Mer xii No i ic no
8 figs. ' ' • . Ji FF-.
Thompson, Sir D'Arcy Wentworth, 1942. Growth and Form. Revised ed. Camb. Univ. Press.
Thompson, Harold, 1943. A Biological and Economic study of Cod (Gadus callarias L.) in the Newfoundland area including
Labrador. Newfld. Govt. Dept. Nat. Resources. Res. Bull. No. 14, 160 pp., viii charts, 12 figs., Appendices a b c
and A-Z. > > >
Vaillant, Leon, 1888. Poissons. Mission Scientifique du Cap Horn, 1882-83, T. vi, Zoologie, pp. Q-C35, pis. i-iV Paris
Vaillant, LfoN, 1906. Poissons. Exped. Antarct. Fran, commandee par le Dr Jean Charcot, 52 pp., 4 figs Paris- Masson
Waite, E. R., 1899. Fishes. Sci. Res. Trawling Exped. H.M.C.S. 'Thetis'. Part I. With addendum. Mem. iv Vol \
PP- 27-132. Sydney: Australia Museum.
Waite, E. R., 191 i. Pisces, Part II. Sci. Res. N.Z. Govt. Trawling Exped., 1907. Rec Canterbury Mus. i, No 3 pp 1,7-
272, pis. xxiv-lvii. J' ff 3/
Waite, E. R., 192 1. Catalogue of the Fishes of South Australia. Recs. South Australian Mus., 11, i, pp 1-208 ^^2 fies
Welsh, William W. See Bigelow and Welsh (1925).
Wollaston, H. J. Buchanan and Hodgson, W. C, 1929. A new method of treating frequency curves in fishery statistics with
some results. Journ. Cons. Internat. Explor. Mer, iv. No. 2, pp. 207-25.
APPENDICES
396
DISCOVERY REPORTS
o
<
o
I— I
o
o
Pi
<
Ph
I— (
Q
iz;
w
PL,
Ph
o
a,
o
c
o
c
O
H
O
2
2;
i>l
u
^
a.
o
c
S
a,
J3
Cao
bjo
c
, 3
J3
1
1 o
-4-*
O
XI
y3
c
3
»
o
-o
CJ
H
(D
3
O
2
fc
u
s
o
4-*
O
pq
3
V
<L>
, 12
li o (- "^ -
^'S,;;'^ —O"^ r^r^N N O O O N I^OO
t^^o CO ^6 t^ob CO vo v6 CO ^^ob Kob r^ t^cb 6 >o io iovo
r^ t^so CO ioob f^
NCO r<^HH r*-)ONroO '^foa^O^
CO '^ C^ "^ 0^ O^ O^ OnO i-h n N ro
wico li^ u-,co J^ ro C^ tnco N M M
iy-> u-i Lo io to io^b Tf- f^ f^ob ob f^
N lo ro\0 Tj- t^ ro N
O -^ ro rOCO ON J-^ M
C>vD ^ oj r^ O NO CO
OnOO O On On On O On OnOO nO CO CO OD f^NO CO
ONf^N OnOOO O " O Onu^On
N O rooo i-H y^^^-^ONO N N
On ONCO CO 00 CO ON 6 00 ONCO On
IT) lo Ln -ho NO NO O -i-NO ui W o N
NO NO NO nd so NO NO f^NO t^ r^ob cb r^
CO o o o
O OO 1-1 M r<-)NO N CO ro O
O l-^ « On M i-< I^OO O
O O
O O
-i- O
O
OONf^Ot^NNOOOOOOOOOOOO
OnOOOOio"-iij-iOOOOOOOOOOOO
AAA
AAAAA AAAA
AAAAAAAAAAAA
O ON 'i-co NO
^ lO N 00 N
ro fo ro N N
- - -_ ON^r^*-<ON -^nO ■^ On 1^
roOO On-'I-COOO ror^ONONONNNO rot-1 N N On
NN roNNNNuMMNOlMi-''
a. ,
Q
3
■T3
o
a.
o
t.
Ml
C
O
CO
u-i
ro
w o
o
On
o
hH
w CO LO N
1
Tn
M (N)
r^ 1
1
0^ t-H HH oo
1 1 1 1
N
ON tH
t>
o
u-i N N M
r-
o
l-<
0^ rrj looo
t-<
11 i-«
-J On t^ 1^00
'^-N '^ON-i-N w H l-l
►-'t-'i^r^'-HMMojoo
I I I I I I I I I
O t^ioONi-HNOCO roN
^f^f^t^voONt-i MOO
lo ro O NO CO
O -^ r«^ (N) O NO
M W M hH
I I I I
M LO fO O
C7N T^ ro tH
l>^ M lO O
■^ t^ (N) N
I-, On 11 M N M
I I I I I I
OnnO NO r^ 1-1 OO
o ON -^ r^ to w
1-1 w « N "
ON ON OnnO conO t^ -^ '»^ no "^
I^ -^ fO 1-1
1-1 -^ N N N
I r^ I I I
" " M (Nj OO
NO •* ^ O
" N N N
'^
r^
^ rh
)_,
o
1— 1
o
HH
I—I
1— (
^
N
"-I
X
K- 1
H
0^
OO
^n
on
n
-i-
1— 1
^
1— 1
o
r^
o
N
N
ro
i-<
l-H
O O ^roM loLoij-, N o >oio looo O
u-.«rpu^ro;+i-. O O loN w « io« -i O u-i" « u-i^M ""iroioi
•^ >-o o oo
•I "-^ N
r^iONvO r^N r^O r^O u-^O Om
.-hhhhhwi-.O'-hOO'hOO'-h
O O »J-)N ^^-u-iu-iioi
ON On r^ 1-1 r^ t^ 1^00 NO ^ - -
Tt-moNt^^oco o
1-1 M M n o
t^oo
o
LO N -^ O lO 11
LO r^ OnnO lt-^nO nO no NO t^
i-i«„Oi-iOi-iOiH"
tor^ino o o o too r^oo u-i lo w
^ HH LT-jroi-i O roroN i-i O fOio-^
O N O 'I- OnnO COnOOOCOnOOO t^u-j
>-iihi-cmO«Oi-iOO'-iOOi-i
NTt-lOW roi-i ^ONt^N t^r^ONM -+NO i^nO 03 Aw
roroiocoo roO lorO" ^O O" ro'S !o, O ^ P^ O
°r^°i^?^ °/^ °». °;^ 2, °_._° ooooooooooo
to f^ lOOO CO to rj- Tj-00 nOnOOOOOOCOK-iNOOOOOO
tOtOtOOO^^^NO^C^Ng S^^ ^^S nS-N^N^vS N^^nJnS ^ S^^ N8Ngv§N8N§N8v§N8S;NgN2 N^N^V^
!^^ ■'r^ "*. "^ ^^ ■" *■ '~'''^'"- ^ ^ ». «]rl.^|c.lr-*|CJH«^^
(-OOOOOOOOu-iroON c^ O K O §
O Ov "tI-oo 00 O CO t-H
O "-< loioioo LOO
Hlei^ ^ ^ ^
O tJ- N M 00
O u-i '^ m M
r^ !>. N N lO r^ r^oo loo ^O m ro r^
LOlOVOU-llOLOLOlOlj-ilOlO
'^-N M N ro-rf-ThfON N
tototo.o.o^^to^to.0 S>=§-<=^ ???=§- aS>&>??=^q:*J^¥!J^5:^?
>N
M
C
n u <u
-^"^^^---------5^^^f^^^^!?frfrJ^f^fr!:rs-
.^ .£; .^ .^ .£;.£; .5 .>.>.>.> .> > > >
fOrO-^-ONl-i ■n^-'^tocO fONO
wiHwwNNNN
l^ r^OO 00 ONNO r^ t^oo 00 ON ionO
OOCOCOOOQOOO°°°°°0°0°°000000
COCOCOOOOQCOmnnnNNNN
> > > > >
ro ro -^-oo 00
N
N
N
N
N
N
M
>
>
>
>
>
>
>■?
">
ON
N
ON
N
O
O
ro
ro
ro
« w
N
N N
m
»^ eg 0-; l^NO t\Oo 0\ O >N (Ni i-r^ ^ lo NTN l\ Or\ rKf>*^fN, r^v4..^«^ COO\0>HNrO^ ionO t^OO On O
PARTICULARS OF TRAWLING STATIONS
397
otrope
ed well
o
■5 -s
1 1 1 1 1 1 1 1 1 1 1 E 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ^111 Sill
Z
o ^ bjo
•T3T313 ^T3-a "^-^En
333 ih33 C jsS-^^m
Z Z Z Z ZZ Z JfiZ hS
0
s
i-irt-wvO-^-i-i">-<tn r^vO Ol-O^|-0C" ^MMt^r-lMMN O OO^ Nt^O^:^ rOOO t^ N OO Tt- O^
o
-^00 O^ »^ '-' r*^ t^GO sO ^O O O O^ N ^O O 1 C^O^^^Nloo^\0 O 1^ I irio 1 u-^N^or^N^ioiot^'-' 100
o
f^\6 O TJ-ioiotOtotovotowiAt- io\6 ioiol^Otoioiou-iio^ ThlcO IsOO 1 io^O ^ Ti--0 sO f^ t^ ^ io 1 ^
p,
u-j losC 00 -^ 0^ lo trice a^•^fO^^C^•-''-'0 r^Mr^'-''+-roONN O ■^'-' OOi-tONOONO u-tOO O -^
6
^
r^OO OO I^ f^ ro I^ OVOO I^ 0^ N O ^ "^ O^^O 1 N OO 0^N'^0 r^lvO 1 ^u^l t^"-! (^00 O O^ wi I^ Ln ro i O
3
CD
t^O sb ioioioioioioio u-jsD vOO\b^iLol\bioioiototrvto ^lob Icbf^l'ob r^O CO t^OO 6^ ui f^ 1 CO
h
3 ^
2 c iJ c
^-•> o
rt tfl ? o
U -1
N-!hOOOOiH>00000000 m OOOOTt-u-it^vO^OOOOOOOOOOOOOOi-OO 0^\0
un u-1 o Tj- Th u-iO vOOOOOOO 1 N i i OOioO"NN ""OOOOOOOOOOOOO-^-^ -^vO
.2i; =
A AAAAAA AAA AAAAAAAAAAAAA
Q P
V
u
•a
o
C '—
S £
-2 2J
•^1- m u-i M ov -hvO vO t^ t^ LOOO t^I^ivOi |vO«T|-i-<0 irivO 0000 ^O inu-iO N O >-^OCO N O mt^ro f^OO
2 o
ON t^ u^0O 00 00 LO lO -h ^ " t^^O " N OOir,"^"00"-it^t^«i-.t^^-t-«ror^O ^CO vO -1- O " O r^
.2i
NN"r^<^r^roroNNNP-i«Ol'N' •-•N"NNr<-)NN m«NmNNi-i Nr^rOr^,
Q
6
SO M N loO Mt^OvQiot^u-tOO (-^"-"r^ONt-ir^OOONf^Tl- O ^00 N 00 ^ O O sD OS
c
N O »o O^ rn t^vO u-iONI^Oir^O»ri on^^'^*-''^^<^0'^'-h 00»^i^Ni-tO'-'OsOO>'^'-tOsa^ rh-^ui
HHh-tl-lMNN'-'l-HI-HWN>-'rON 1 MM^t-HWMNNNrOtH OSVO \D00MWI-lONt-lN'-''-tON ON*^' hH »-< 1-1
I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I 1 1 1 1 1 1 1 1 1 1 ) 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 ! 1 I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
-^ N r(- O CTs 0 0> r^OO toioinN O ' sO -hroON^ro-^-r^O "-i 00 -^t^r^ t^sO O CO O ON f' ON lo 1^00 CO '-' M U-)
C
3
NnO -ThN N n-isOnO OnoO ON^Ot^i'^ On'^'^-i-' -^loOvO ^ lo *-< OnDnOOO O "-• '-' OnO ONro"-" O O^ ^ -^ -^ ^^
MMi-nroNNi-'MiHi-.i-iMNi-. «««iH«Nr^MNNMi-c mm,-, c^nwmm wmmm
c
I^nO " OnoO O O " m 't- O uiOO Lr)N"'1-^'-'N0r^0"->0r^0"-i0000'J-)00tnro00io00u-,0 0
o
o
roiON ION O iOfO»or^io»oO iom»-< "^N ■Tj-'-i HH >H fi lo-^- m-^r^'-' ror<^tH wir'-j'^-M roN N CON N N *-*
•^
H
o N CNOO On N Tj- ONOO N CO CO r^ " o 00 onno 00 r^ onoo ■-■ oc o " loco o o CO o onoo no loco r- r^ ON onco ■-■ on
CJNOOO'-i'-HOO»-t'-HO»-'»Hi-Hi-<0»-iO'-'0'-''-tON hH>-iO'-'NO'-H'-i'-<'-i'-<'-i>-«0'-i'-iO'-'0
e
2
OOO Tt-00 O O NO lO tJ- O >000 >ON00NN0N0r0toio0^00i000io0ioio00ro00io00>O0>o
E
H
fe
tHNOOOt-iMOO'-'»-<0'-iMO'-<0'-*0»-tO'-''-'0'-i t-<i-HOO*HOO>-'w»-iwi-(MO»-'i-<Oi-iO
^
00 ^O ^O ^O OO ^O ^N ^N ^io'"io'io^O ^O *"to^O O^OOOOOOOOO NONOroOOt^OONOONOOO-^NON OnCO
bi)
tororou-itoN Tt-T)-T)-Tj-N i-H ThO lOM 11 i-c rotOTj-^O roO lOioO O ro>00 rorOO ^ O rOO N u-ii-i i-i O
c
ooooooooooooooooooooooooo ooooooooooooooooooo
o
O^ On N NO NO r^OO OOOnOnQ'-'OOi-'Nio tovO ''t'NrO'-toOt^ ONONtotoro'-'NONOONrO ionO CO 00 GO O
c:
J
lO ionC lo lo lo lo to »o tONO vOnOOOnOnOnOnOnOnOnOnO lOiO to tONO vOnO^nOnOnOnOnOnOnDOnO toio u^nO
C/2
OOOOOOOioOOiOiONNOiOOOOio lONO NO OnO >oO ionOnO "00 "I-nO O no O <o t^ 0^ ^ OnnO On O O cO
■HNNO'oro«i-iNNio>oioOrO"ioioOOOroN-*-^ioioio^"tOTt-0-*-i-iNO'i-"'ON<oroO
jj
rt
ooooooooooooooooooooooooo ooooooooooooooooooo
hJ
OOOnmOOOOOn onoo I^nO nO00ih"'-<>-<i-iNNNN'-' OOrorO'^"^'<^iON»ot^ I^nO nO lo O O O O
lOtOTi-lolotolOtO'^'^'^^^-^'^tolOlOlOtOlOlOlOlOtO totO-^'<t''^"^Tt-Tf-ThTt-rJ--^Tl-Tj--^tolOlOto
,^,^^,^^00000O0O0000000000O0000O00000000000O0OCO >j>.,m-m"""
u
NNNNN rorotororoeotorocororo'^'^'^'^'^. . .
Q
■>'>'>'>'>'>'>'>'>'>'>">>'>'>'>>>>>>>>>> X X X y. '^ X y, y, X y, y, y. X X X ■■< •f< >; y.
onono n d " coTi-toio toNO NO t^oo iono no r^ r^oo oocnoo ooNONONor^r^i^ONOt-tMNro cond -^ ^- to
""„„Ji5 ^ „„„„„i_„„NN«M""<""«MNroro
§^
'^o;<jm-jje3 '^ttJ
^l"i^
-> i<-)nO t^oo On O -^ N ro ^- tONO r^OO gN O 3 J> £2 3: IS'S.'S. ° "R"^ N N r'J-r-^iOM N p;* tONO t^ -H ^ ^ r:^
Pi 0] rl fN) (N) !Ni fn^oro^o^oro<o<»5<0^'1-'+~^•J^^~t•"*~^'0 ^ i^nO nOnOnOnOnO t^J^t^t^t^r^ r»=g '^ "5 ^
|1
!Nl<NlN'MtN)<M<NgO)NNtNINNNNtNi<Nl<MCMM<N)tN)(NlC\lN ts^t\|^t^(^t^J^r».t^|^t^t^t^t^l-Nt^t^t^t^
co<>oc«coc'}'>3'>3^cncfic«Mcr!c«a)fococotocojoto^coai cococ«c«iy3c«aic«c«c«cncADcflMaijojococo
C/3 ^
2^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
398
DISCOVERY REPORTS
o
Q
w
PL,
<
o
Trawl fouled
u
o
3
2
e
e
o
o
%^\ irKS^Kf^l^lll^^^Kl JCI IS^I i 1 1 1 1 1 1 1 1 i 1 1 1 1 1 1 1 1 1
1)
o
3
C/2
moo 00 moo 1- M 0 LoOOO" t^ lot-^o n
or-, ir'^TfT-ri'^irir^r'T^ir^l |9?^|9^ir^| 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ;
U -1
o _
.22 i =
S ^ "- '-
2 c " c
S^-S o
ca M ^ o
V
u-iooooooooooooooooooooooooooo 'i-o o^t>oogoooogggg
t-OOOOOOOOOOOOOOOOOOOOOOOOOOO"" roO vOO^OOOOOOOOOO
AAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAA
2 2
.22 i
Q
III
oo w mo\o or— r^io»'^»-' lo«oa^ Looo '-' o o oo a^f— c^»^o^n t^oo n o^r^c>i^»^-+-r-N i^io t— o^o^r»
O u-> ^ -1- ON r^ N •i->0O 00 M N ^ -)- r-CO "OOOOr-oiONNMi-cro -t-vO -I- -h On r-vO -<-NOt^in-t-Nror^ "
NWIHl-l „„„„„„« „„M„„p)N«««w««
Q
.^r.^r^j-^ONO HH^ MsDoJ^sO^-ifON '-'■^ t— O^t^ONtoN O^O roOOO NvO f— -hO
„„„««m00ON"'-'On""""'-''-'"'-' I mi-.-0«"mi-cm'-'«i-'"'-'""'-i""'-'<:7>ONOn "-)C0
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
O O O O r^.vO N i/-> ON ON t-^C N nO oo r^i t-vO CO ' u-i«ONi-<t^ONr- loCO N r^ O roOO On lo O r- ^h uoco nO r«-) jn-j
t-iniou-imOOOONONONONONOOONu->00 ""^^i-^r^ (^nO N r^t^"-)T^-T^^^ N •-■ O O O^ONO^i^Ti-
o
u-iOOOOmirimOOOOOOOOMOO'^.N"-lNr-«/^OOK-i:^Ou-ir-u->mioO''^"->mLr>i-iOOin
CO" u-ir^mM u-iTt-m'^O K-)iriLou-i«0" O t^ 1 " •^rj-O ^m u-iu-i-|-rO'*--1--i-^" -^mu-ir^ino O N "
GO i-( N '1-nO r— 00 ro r— M t— r^ fOOO roONN r^t— 1 00 mN rot— M NnO -^nO r^NO OO I^I^ror-inO NGO O •<*-r—
„w„«O""NO«"M0000m>-'0 «Ni-imwNOOm"NOO"0>-<OmNOOO"i-i
S
S
u->0000»J^"^r— *^*^00 0000'-'OOOrlONt^u-^0 0»'^i'^0»ot— iou-)»00 00»00»-'OOLn
fO^-^ li-iroN ►-! i^N M roO t/^i^iou^iO"-* O u^O « O -1-0 -+-•-< uiiy-, roro'^'^'^'^'-' ThroiorDO O O N «
I— O " rM "^nO t-^NvOOOvO OnN '^NOO m n rOOO t— ©"MrO-Nuirou-iNu-, I— vO nO N nD tJ-\D N I^ O r^ <••-)
„w«i-iOi-'>-'NOO«i-<OOOOm'hO"'-iN"""MMO""NOO"0'-iO'-iwOON"m
c
]■»-<
O
Oh
c
roTht^M Ln-i-O N N N r^triroM rJ fi ^ O m O N O N O rot-ioo O OO m t^ >-< r^<:^'H O I>mgo mt^'+O
O roro^»ON ON Thrl-u-iThN m -tJ-Q N-^rJ-O N O O ^ON w LomN NO '^'loN m O-^N-^OO O^OtO
oooooooooooooooooooooooooooooooooooooooooooo
•-H NNNrniOU-^T^-r^r^NN NNhh h-i O roroiO'+»OLONNrl hh •-( i-(i-h NNN ro*^»^ iri\0 O f^OO OO OO OO
0'0'^^0'00'OsOO'O^OOsDOOO'0'00'0'OOO^OvO'sD\DO'0'0'00'OOOsD\00^0^0'>0'0'0
h-1
■^NNNO^O'-^OIN^n'^T^u-llJ-^l-HwlOli-lr^T^ror*^00•-H^-^Nr^■^-^^NN'-lOlO'^NNl-HNN'-«l-H
oooooooooooooooooooooooooooooooooooooooooooo
ON c^ c^ o^ ONOo u-iu-^voii-iiou-iiovn u-i\o ^o r^ t^ i> i^ r^ r^oo ooooooooo o o o o o oo^c^o^c^c^'-' m »-« ^
Q
^
m r-, ro r-. r-. r^ C-. r-. CO r^ r-. c-. r^ m ro CO r-. CO ro r^. r-. r-. r^ ro ro r--._^ ro ^ ^ ^ ^ ^ ^ „ ^ ^ ^ „ ^^ ^ ^ ^ ^
.^ .« .^ .^ .-; .^ .-; .-: .-; .— .-^ .— ..-' •-; --4 .-h ._; .-: .^' ...;._; ..4 .^ .« -^ .-; 'p ._; m ro m ro ro ro ro rn r^. f^- ro ro ro ro fo ro
'>< '>< 'S '>< 'S '?< '>< 'i< '>< 'x '?. 'x "P, 'S 'P 'x 'x 'x 'x 'x 'x 'x 'x 'x 'x 'x '. 'x .-; .« .« .« .- .-: .- .- •-; •-: .« ■- .- •-: .- •-;
N
lONO nOvC I— t— f^c^Tj-Tl-.^.rJ-iy^ u-iO r— CC ONtHONOOOt-'HHMNMLrj*^ u-inO nO nO t— t— 00 00 00 On O N O N
N
o c
1]
1-
Ti- u-i u-i u-iND r-00 a^oOMMNNf<^'t• >^vo NO r- r- t^oo ONONOOi-c<Ni<N)<N-i'T-'^ >o<5 r-oo on on o -i •-< c^i (m
oocooooooocoooeo o^o^ONONONONa^ONa^ONa^ONONONONONONO o o o o o o o o o o o o o « "-< "-i •-( >-(
t-i^i^t— t— r-t^t^i^t^t^c^t^r>.t^t-~t^t^r-~t— i^r-r-:^ t— oo oooooocooooooooocooooooooooooooooooo
PARTICULARS OF TRAWLING STATIONS
399
J3
perly
«c
o
1-1
a.
•a
s
, . . . ">
s -o ^
j:
o
1 1 1 1 u-
1 1 1 1 1 •? M 1 1 1 1 1 1 SI -5) 1 1
1 1 1 1 i 1 ll 1 1 1 1 1 1 1 1
jz;
u
'-• B —
T3 C D. ^
n
o
JS
~™
o
c
>,
a
g- '^ ^ S 2 2
o
P3
Z Hta ZZ
u
o
2
1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
3
*-*
£2
u
a.
6
o
1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
o
v2
1 1 1 1 1 1
1 1 1 1 1 1 1 1 M 1 1 M 1 1 1 1 1 1 1
1 ! 1 1 1 1 1 1 1 1 1 1 M 1 1
H
3
CO
tf)
li-
o o o o o o
N.-.OCOOO^COOwOOOOOOOOOOOO
r^*ooooooooooooooo
o o o o o o
ON ON r^^ ^ ^ ri'i-ooooooooooooo
NOOOOOOOOOOOOOOO
SJS-g o
)-||-«l-HI-ll-H>-> l-ll-ll-IHH
.2i =
A A A A A A
AAAAAA AAAA
AAAAAAAAAAAAAAA
Q t
V
(U
-«
ii
!= ';:r
^sii
C> M o OO O N
O N ■**-0 sO -I H- lo N u^O O NO NGO rou^'-' »>^On
O "<t-0 ONsO N t^ro-^ OnOO « O I> (^
"-> t^ O N u-i U-,
N ThOO »J^ w 'fi- tJ- r^O O m lO N \0 '-' fO
.2i
'S'^e
I-. I-. *-. HI
WHHW.-.HH^Nrl'-^, H-« _►-,
M (VJMW MWWW WW
Q
6
N 00 N O N CO
OO -^ ON N r^OO t^ ^ N ON -1- r^i O r- t^O
t^ O -t- <^- O O -t- u-iO rl- N O w M -^
O^O'-'NN'-'OO'-'N '+0 t~^ ^ m rn
O " O "-1 o m
t^OO N -^O sO iJ-^ro-^i-*©© O "1 r^O N ^C CO '-' r^O
3
M 1- w « N N
(N| N for^mTj-ON'-H i-i rr.ro'-' •-< '^'i-'-i >j-i»h m ^ i-
hhONwwwwwwwwwwN'-'i-'N
1 1 1 1 1 1
\0 " N O « N
1 1 1 1 1 1 I 1 1 1 1 1 1 1 1 1 i 1 I 1 1
NOONON-^OO tnOO t^OO GO coco ON « ^C' u^ I> t^
1 T 1 1 1 1 1 1 1 1 1 i 1 1 1 1
'-' OiONNO O '-' NOO r^O ON ON u-1 N
C
O M CO "-) 0^ O
t^ r^ HH o» i^\0 00 '^rOfON ON'^J-OnO O u-jiO'-i ro^^
NONwNmOwQwM rovO -O ro ro *o !
w l-l M li M N
NNf^>fOrOTl- WW ►-.Tj-row www
M _wwwwwwwwwNMwN|
c
O O »^ O Ln i^i
O O LT-i^OO O O tou->o u~)ir^t^r^LriO toO "J^ lOOO
LOO O »OioO NQiO lOOO O O Lo O ui
O K-i Tl- o o -t-
'i-'^Tl-r*-lU-)T^^^, O r^iO N N O O --< -h- N M ^-r^
w N 'i-roroiororoio»oto"rt-0 O >-• *o
t^ CO '^- Th -fOO O O -1-N rO'^i^*l^0\0
^
H
lo Ti- i-( rn Ov r^
O w ON -^w r^w w ONt^r^r^.r^w Q ON'^r'^wOO
2 ?
3 g
<U o
O M N •-■ ■-■ N
O'-'-'NO'-'ONOwO'-'i-'N'-'NOi-i-'NO
HHWWWWQWMWWWWWWWW
e
O O "^ O Lo lo
O O u-iLou-jQ O O u-iO »^«J--r^r^toO "-*0 u-tLnO
LOG O^otoG N0*0 lOOO O G lO O lO
e
O "-> '^ o o -+
'^'i f- r*^ ^ r^ mo rr.O N N O O •-' rn >-> N N '+r^,
M N H-roroio*rororoio»oH-0 O '-' lO
u
■^ m O N CO ov
u-i r^O COfO'-'^O O OOOO^O N NO ONOO O N O r^\0 N r^i f^> r^i r>. ON ON ro "-I N c*-i •^\0 On u-j
H
ta
O « N " " -
OOwwQwONOwOwwrJwwOwwNO
H-wwwwOOwwwwwwwOW|
t
^
^HM- ^ ^ ^ ^
CO CO vO 00 t^ C^
r^N HH r^r^r^N oncoo r^o n w w coon ^i-^o o >^
NtoOtOOOwOlOtO lOO Lo LO t^ G
tuO
^ r^ -^ lo ^ O
N w -rhNir^toONi-H O O r^O roinNu-ji^N O rr>
Qwwwr^lOOrOi-iwwwwwwlo
o o o o o o
ooooooooooooooooooooo
oooooooooooooooo
t^vO lO ^ Tt- -)-
f^cDNN'-.OOOO r^OO OO O u^ r^, w Q f^ t^sC tn r^
N -^-^-^O ^ N fO'^'^'^'^'^Thu-iro
_o
hJ
s^ \0 ^ *0 vO 'O
sOOOsOOsO^D u-j LOO sOOnOsOnOsOOO-OOO
^O O ^O ^O O ^O O O ^O "O ^O ^O "O ^O 'O "O
CO
O
CD
H«rH(ci^ i-4c]v
Hs^i ■-.ici ---i-t ^ ■-'Ic-)^ Hci^ ^ ^ ^ wlcJ^ Hci.-<[(M^ *. ^ ^ r-i?]^
r^ TJ- 1/-1 O N N
O T^-w IONIAN w GOO r^OO w ON '-' w u-^oO r^-O 00
ON00\O <-> N Tt-w rorO'-'\0 u-jOnOnioio
•!-•
m rr, "^ -rf- u-^ tn >-> rrju-iN vtitJ-w n u-jw w w r^tow
rorolOw -+W w u-)N roro^'^'^cOfO
hJ
o o o o o o
ooooooooooooooooooooo
oooooooooooooooo
1-c w i-c N N N
NNNNNNNNONNNf^, rocO't-OOGO^-'
w tJ- lo I> u-1 lo u-> t(-00 on G '-' N CO N N
lOTt--^rt-Ti-T^-J-^Tt--^U1lOU->lOlOlO
ir^ lo u^ u-j to lo
iy^u-^iy^Lnu^u^triioiJ^iO»^LniOLOiOU-iLOLoiO»OLri
u
N N N N N N
r^ ro en ro m ro
N
nnnnnnnn'^*^'^^*^^*^^^^'^^''^
mNNNNNNNNNNNNNNN
Q
/4v*-- ••-^ "^ ••H •« '-^ '-^ --4 -^ --^ --^ '-J
.-; :s ::: := :3 :s :=! :S :=! :S :3 :S :s > > >
N ....
ro CO fn -^ "^ *i-
t^ t^ t^ l^oo oo o o^Jg- « pj m lo loo \c o o o o "
w w M, ro-^iOU-)ior^OO OnQ '-' '-1 coco
""""""
i-HNNNNNNNNNNCOro
c c
o o
•^oq
•^aq-^oq ^ oj
< W
7 M —
r-i -^h 'o^ f^ t^
OoOo 0\0\0 -^ ^*J->0'-T-1- t^Co On C -^ 1^ t^^o O, 2
r^ r^ u^ r~~oo On a^ O N ^O CO O N rt- to
ii S y
•-i •-< ^ f-H '-^ -^
►s •>! -< -1 <^l 0) 0| 0) M '^ Po '^ '^ '^ -^ -^ -+- ~h ■> -h lo.
'j~; lo lo ^^' lo »o uosc ooomd r^r^i>r^
2 Si
1-
00 Co Co Co Oo oo
OoCoOoCoOoOoCoOoOoCoCoaoOoOoOoOoOoCoOo>DOo
^■OOOOCOOOOOOOOOOOOOOOOOCOOOOOOO
Co C<J Co Co On Co
CoC/^C/^OOOQC^jCoCoCoCOCoCoCoGnCoCoCoCoCoOQC/:
^c/2aicr!c/2c/:aiaic/:c/Da)c/:cr;c/:cnc/3
C/3 -C
^fe^^fefe
^^^^^^^^^^^^^^^^^^^^^
^^^^^^^^^^^^^^^^
40O
DISCOVERY REPORTS
APPENDIX IlA. HAKE DATA: FIRST SURVEY
Hake nos.
Mean
length
Theoretical
weight
Probable
Station
minimum
5?
Total
o<S
??
kg.
weight, kg.
WS71
0
0
0
—
—
—
—
WS72
0
0
0
—
—
— •
—
WS73
0
0
15 (juv.)
—
—
—
—
WS7S
0
0
0
—
—
—
—
WS76
0
0
2 (JUV.)
—
—
—
—
WS77
0
I
I
—
—
—
—
WS78
4
29
33
—
—
—
—
WS79
I
26
27
—
—
—
—
WS80
0
26
26
—
57-4
32-350
26-500
WS81
0
0
0
—
—
—
—
WS82
0
0
0
—
—
—
—
WS83
0
0
0
—
—
—
—
WS84
0
0
0
—
—
—
—
WS85
0
0
0
—
' "
—
—
WS86
0
0
0
—
— ■
—
—
WS87
0
0
0
—
—
—
WS88
0
0
0
—
—
—
—
WS89
0
0
0
—
—
—
—
WS90
16
71
87
—
—
—
—
WS91
0
2
2
—
—
—
—
WS92
4
10
14
35-8
43-5
9-205
T550
WS93
0
0
0
—
—
—
—
WS94
40
63
103
367
45-4
53-410
43-800
WS9S
10
29
39
39-4
43-4
20-25S
i6-6oo
WS96
14
16
30
30-0
337
6-790
5-570
WS97
13
22
35
38-1
52-5
25-g6o
21 -2go
WS98
II
60
71
37-8
53-6
65-840
53-990
WS99
0
19
19
—
63-3
31-325
25-6go
WS108
62
64
126
35-6
43-3
54-205
44-450
WS109
0
28
28
63-0
45-510
37-320
APPENDIX IlB. HAKE DATA: SECOND SURVEY
401
Hake nos.
Mean length
Theoretical
weight
kg.
Probable
minimum
Station
6<S
??
Total
cJc?
?$
11.111.1X1X1 \^ 111
weight, kg.
WS210
WS211
3
28
31
37-3
45-8
20-035
16-430
WS212
WS213
WS214
WS215
WS216
WS217
WS218
0
18
0
87
118
I
93
5
181
293
I
III
5
268
411
46-3
40-1
39-4
54-0
57-9
54-8
47-5
46-4
r-113
132-770
5-630
176-ggo
260-060
0-913
108-870
4-620
145-130
213-250
4
72
76
38-8
62-1
iig-500
97-990
WS219
WS220
16
26
42
36-8
39-4
17-200
14-100
2-780
3
3
6
447
43-0
3-390
WS221
WS222
0
I
I
—
59-0
1-430
1-172
WS223
—
—
—
~
WS224
WS22S
0
3
3
—
647
1-845
1-513
WS226
WS227
0
I
I
—
59'°
1-430
1-172
WS228
WS229
0
2
2
—
54-5
2-355
1-930
WS230
—
—
WS231
—
—
WS232
WS233
WS234
WS235
WS236
WS237
I
2
I
0
7
4
5
6
4
12
5
7
7
4
19
59-0
44-5
43-°
41-0
47-0
42-4
477
59-0
48-3
4-375
4-010
5-390
5-910
13-325
3-590
3-290
4-420
4-850
10-925
WS238
—
—
WS239
—
—
WS240
—
—
WS24I
—
—
WS242
—
—
WS243
WS244
WS245
WS246
9
0
I
34
26
0
43
26
I
41-3
40-0
50-0
59-8
35-760
39-755
0-437
29-325
32-600
0-358
WS248
— -
—
WS250
—
^
23-2
402
DISCOVERY REPORTS
w
>
CO
Q
Pi
X
<
Q
W
<
X
u
Q
w
-a
c
11
60
J3
.60
'S
a.
s
60
X ^
B
o
o
c
60
C
s
n
J2
3
1^
No hake
No hake
27 + 2 juv.
26
14-500
90
47-800
7-500
68
102-650
No hake
37
22 juv.
ig-g
I -000
No hake
No hake
No hake
No hake
8
1 1 -000
8 8 8. 1 ^§
1 1 1 1 ltnlir5llo^lMO^IKvol"lo'6lll 1 1 1 looob"
wM^ro >OOf>N low
H B
0 0
1 1 1 1 1 ^ 1 „ 1 1 « 1 ^: 1 1 1 ,, 1 ^ 1 r. 1 - 1 1 1 1 1 1 i 111
MM ■^ HH
a.
3
2
60
^
60
C
o
1— (
.3
Of
Of
1)
o
o
1 1 1 M II 1 1 1 1 II II 1 1 1 1 1 1 1 1 II II M 1 1 1 1
o
o
ft
1
1 1 1 1 II 1 II 1 1 1 1 1 1 1 1 1 1 II 1 1 1 1 1 1 1 1 1 1 1 1
o
r
CO
1 1 1 1 1 1 1 1 1 II 1 1 1 II 1 1 1 1 1 1 1 1 II 1 1 1 1 1 1 1
o
oo
1
N „? 1 1 ^^1 1 *.?! 1 1 1 1 1 1 i 1 1 1
o
000 0
llll 0 f-^i >o^ oSioo8| '^"^lllll '^1
1
000 0
ll||| 0 r^ \ r^^ -mSi -,81 «^^||lllll ° \
to <>-> 'Tvo ro'orotoin^vovovo r^-^ob >o v6
u->w-)K-l>ON'ON<0 U-)
o
1
5-
000 0
Mill "-^111 ^p ??- ^?^l ?p|lllllllll
o
1
CO
1 1 1 1 1 «^ mt 1 m?:l III III 1 ^^1 1 1 1 1 1 1 1 1 II
0^ r^ M OTi (Vi
o
1
N
1 1 1 1 1 m^ .§.1 1 1 i 1 i 1 1 1 1 1 0.?.^ 1 1 1 i 1 1 1 1 1 1
C\l 0.) M Cvj
D.
3
2
60
a
3
6
o
o
c
'o
'o
<u
.ii!
c«
o
vO
1
u-i
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 M 1 1 1
o
to
1
5-
1 N 1 ^J;l 1 N ,?| 1 ,9| 1 1 1 1 1 1 1 1 M
o
1
m
1 1 1 1 1 mS^clSl .ll 1 1 1 1 1 1 1 .!?l 1 N 1 1 1 1 1 1 1
O
?
1 1 1 1 1 ..^ H^l .g.f 1 1 1 III 1 ^ll 1 1 1 1 1 1 1 III
^1 fvi (M i>j
a
_ < m <; PQ "^ cq
JOvOvOvOvOvO vO t^ !>. t^ t^t^ t^ t^OQOoCooo
^ tn in vi Ui Ui tn xn vi tn cncn ixi mcootjcooq
HAKE DATA: THIRD SURVEY
403
O rt
o
o
I °
so I O
o
o
I °
O
3
+
>
>
0 3
0
3
U-, •— >
00
t^ ro
ON
N
ib +
vO
+
t^
-1-
0
so
H
o
o
00
N
O
N
O
O
O
ON
CM
O
o
\0
■tt "I
o C
o
o
o
o
o
o
■^
^
o
On
CO
00 •^
m 00
ON to
o
o
0
O rt
I
o
o
o
I
ON
o
ON
I
3
2
bo
e>o
c
s
o
o
o
00
o
I
o
NO
I
o
o
3
o
a
6
u
o
o
o
NO
I
o
I
o
r
o
o
o
o
o
o
ON
r^ nO
o
Vi
N IN)
O
9
o
o
9 T'
n <>-) t^
o
o
o <r) t^
M t^ N
O
o
9 '?
<N1 N
NO
NO
00
NO
N -I
VO
NO
o -too
N NO -*•
O
O
•> m
ON --t- io
r»5No NO
o
o
00 vD
c povb
NO
On »1- 10
NO >-i
0-1
N NO
10 fn
iNa
o
o
o
o
o
o
o
^1 «
o
o
1- o
^
N^OO On irjOO
ir-i -^ 10 "t^nO
O ^ O 00 10 O f
1-1 10 1-1 N "O rn
10
N NO
N 00
o
o
o
•^NO 00
000
000
o NO ON onno "^ ^No r^
o
o r--
< ob 6
o
o
NO f^
"ONO "1 N<om ^*1-0
M C^ M CO <N-5
o
o
CO o
CO
o
Co
<-( NO
<N1
o
o
M M
o
O NO
O
O
ro f^ CO
N C4
M 00
!N1
Co 1
c^
M NO
"%
(Ni 1
o
o
• IN) N
1-
O
Co o
Oq u-1
o
On
N 0\
CO
10
Co
rO CO r<0 N CO N
M Co H Co
o
o
ro '^ 6
Co
o
o
On n
(N) io
CO
Co ON
r»5 CO 6
Co
00
00
en
PQ
00
CO
O
00
NO
00
en
00
en
M NO
M
00
00
M -1-
(Nl
ON
SO
r>.
CO
o 0
o °
rr i\ « ON ij^ b
M (Nj ^l
o
o
o
o
o
O NO
<
o
ON
en
n
o
ON
en
On
en
M (Nl
M
ON
en
<
N
ON
en
23-3
404
DISCOVERY REPORTS
s
I— I
Q
w
0 0 m 0 0 ,2 "-I Ou*+ "1 0 0 0
|t^ I "^ \ ^ I'"- l"^ 1^ 1" I'^W" rl 1'^ |0 1*^ 1^
CO
00^0000*^0 U-1U1000
OONOOK1MO MNunioO
.r^ |W-j i^tl- it-^ lO |a^ |io rroi |CT> i'^' iN ir^ lOO
«/il«volv6 olvo Ml6>clNoolr^vol6ool'o 1 lol^fl^o nIoo wlr^wlN
vOt^ \Oio •^'*-r>.u-iO\t^ mOMroNNMi-
r\.. M tH
Total
males
J- 0 0000^_i_00
"T 0 LoroiOLn'^OO
Os|o.l 1 1 nIm 1 1 1 Mlbf^l^-^lavr^iN 1 r^l^f^l^nii 1 1 1 1 1 1
g
a
Ji
B
o
o
M
a
o^
Of
u
0
T
0
III III III III III III III III 1 III III III III III
0
0
M
1
HI
0
III III III III III III III III 1 III III III III III
0
1
00
III III III III III III III III 1 „ ? t 1 M III III III
00
0
00
1
M
9?^ 9^ ?=?llllllllllll 9?^! ?r^°??>lllllllll
0
0
00 000 0000
10 0 0 0 0, 0 0 0 "O,,,
C^^. Oro Oom t^u-1 CMu-> 00 ~^0 *l-K-i tvjio MM
M'O 0 M'O" M^Oi-iVO MVON ^"NVO-4-^ VO
0
-0
1
IH
>0
000 000 00000
0 0 Oiiiii, 0 0 0, 0 0 0 0 >o
0"-> U-joO '^"^ "^lio ^tn w~)0 >~iO >-iLri 0\0 i~iw-i VOO
0
iri
1
'I-
000 000 00000
0 w-1 0,., 0 0 0, 0 0 lo 0 10
^rt. OMD t-ii-i ot^ o^" "oo 1-0 oio 'n "^ t^ t^ cjvo oi^
NJs Mt-^oroK^MC^b 0 t>vob ooioNOO^N ot^t^ a^'o " r^6\N m>-o6 moo
oo»o 0000 10
lOiot^ 0^000 N
0100 Ofo ^111 ""1° '^°° -^o '0"->| '^'-^111 'II III III
0
1
0 ir> 0 ir>
^ ^ 1 1 1 1 1 1 1 1 1 1 II 1 1 1 '-^ v'^ 'r* ?^ 1 1 1 1 ?? 1 1 1 1 1 1 1 1 1
woo 0''' ''' '' ''vo'-^ooo'oo' '''mi^o''' ''' '''
a.
i
BO
•£
60
C
•u
o
_c
»o
»o
u
.8
III III III III Ml III III III 1 III III III III III
0
1
M
0 00 000
cv.. "-• 1 1 1 \n 0 1 1 1 III 0 w-1 iN,vo 1 1 1 1 1^^ i^<>o 0 00 1 1 1 III
°r
^^? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ^?;|^>f 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
0
1
N
0 0000
■^^111 III III ??^ 9?=P^ ^"^1 III Ml III III III
a
o
■M
CO
N ro'i-io^^o i^t^ooc^<^o 0 ^
C^C^C?^C7^0^a^a^O^O^O^CT^0 O O
t>l>« t^ r^ t^ :^ i^ t^ i^t^ r-»oo 00 00
ai «3 CO CO CO CO CO-CO coco co co co co
u
j:
.3P
-33
&
■M
O
G
e
u
N
HAKE DATA: THIRD SURVEY
40s
4>W4i 0 0 0 Oq> 0 0^0 0 0 0 0
MMj^ 0 0 0 ^^t *A "■) OS^ 0 »o 0 »o
««« ico 1^- |t^ ir'« 1°° |<=p— »ir^ ir- 1^ ic- i=p
000 0 «+t-. NNWco
OOOO ooooooo
OOOw-j u-»uiNO»oO»o
III |°P 1:^ |f^ 1^1 1=? i:*-os|? ir- 1^ 1?^ |°P
1 1 1 MloNlNC»5lro^lcnl wlco©lt-«ol"Nlor»jlcoolo^OOlo
M t-- N N N CO
0-.
^1
0 > 0 0
1 1 1 Ml III III III 1 1 1 1 N 1 ?■; 1 r 1 1 1 III II 1 ,, 1 2
S
M
B
u
0
_c
ot
Of
1
0
1
0
1 1 1 1 1 1 1 1 1 1 1 1 II 1 M 1 1 1 1 1 1 1 1 1 II II 1 1 1 1 1 1 1
0
0
0^
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 II II 1 1 1 1
r
00
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 M 1 1 1 II 1 1 1 1 1 1 1 1 1 1 1
0
1 1 M 1 1 1 1 1 1 1 1 1 II II 1 1 1 1 1 9^ III 1 1 1 ?| 1 1 1
f*i ^00 M C^ CO
0
0 00 000
llllll 91 ^^lllllll 00 v^S^iii 0^ 00 o)^
0
1
00 000 000
III III III Ot*- >no| 00? oSv ^o||| oJo <^ S-, chS^
mo-n-^n' r»job co^<>)co'«->oco ' MO" 100b 00 00 in N
0
T
0 00 0.0 0 0
III 000 o| or^ "^"^llll °^ "^olll ^"^lll °°^
HpOMOv' MC^6N^v" ' ' ' ' MrvON^oob ' ' ' wobb ' ' '00^6^0
•O ^ ^ ^ ^ H '^t*^. "^ ^
?■
M
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ^^L^L^I 1 1 1 i 1 1 Ǥ.!
(V) 0-) (V-) CV)
0
CO
1 1 1 III III III III 1 III 1 1 1 ^?| 1 1 1 III III III
Oh
1
1
0
M
_s
u
0
vO
1
M
to
1 1 1 III III III III 1 III III III III III III Ml
0
5-
1 1 1 III III III III 1 III II 1 ? 1 III III II 1 ?|
M
CO
1 1 1 II 1 II 1 1 II 1 II 1 1 II II 1 ^^l II 1 1 II M 1 „?|
0
T
M
N
1 1 1 III III III III 1 M 1 „^^ II 1 III III III III
■^1
c
0
■*-»
CO
WS802A
WS802B
WS803
WS804A
WS804B
WS805
WS806
WS807
WS808
WS809A
WS809B
WS810
WS811I
WS811II
WS812I
4o6
DISCOVERY REPORTS
Q
w
Ph
Cl,
97
I
No hake
No hake
1-700
6-100
14-050
10
5-000
12-400
6-650
No hake
No hake
No hake
No hake
No hake
000000
0 0 "^ 0 0 >/i
nI mI hI 1 1 HlMMlvor^l^ol hI^nInhIso 1 1 1 1 1
^ IH M IH
O rt
^1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
M
Oh
3
s
00
1
i
o
G.
Of
o
CD
o
1
o
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1? 1 M 1 1 1 1 1 1 1 1 M 1
o
o
1
00 000
lllllllllll 9r 9^*^ 9? 9?" 9^11111
On 0 0\ 0\ 0\ 0
o
1
00
M II II 1 1 III 1 1 1 H?,!^^ i 1 1 H§l 1 1 1 1 1 1 1 1
00 00 Os
o
,? „9 1 i 1 1 1 1 1 1 1 1 „?L? 1 1 1 1 1 1 1 1 1 1 1 1 1 1
In. t^, tx «>,
o
1
vO
^5. 1 1 II 1 1 ??■ 1 1 1 III II III III III 1 1 1 1 1
1
^^ 1 1 ■ ? 1 1 III III III II III III III 1 1 1 1 1
o
1
IH
^ ?; 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 II 1 1 1 1 1 1 1 1 1 1
M \h
1
IH
^^11 II 1 1 III III III II III III III 1 1 1 1 1
o
i
1 1 II 1 1 1 1 1 1 1 II 1 1 1 1 1 1 1 1 II M M 1 1 1 1 M
to
Q.
3
e
1
6
tj
o
IH
O
^? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 II 1 1 1 1 1 1 1 1 1 1 1 1 1
o
1
IH
^^: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
o
„? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 i M 1 1 1 1 1 1
Co
O
M
N
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
a
.2
S
CO
WS812II*
WS813*
WS814
WS815
WS816
WS817A
WS817B
WS818A
WS818B
WS819A
WS819B
WS820
WS821
WS823
WS824
WS825
WS833
Xi
J3
'S
HAKE DATA: THIRD SURVEY
407
-a
■4-J
e2
I)
a> a>Q>VQ> 0 0 0 Guv 0 0 »o 0 »^
M »^_id_;d_^ 0 u-i u-^ o_M_»^ 0 0 r^ 0 r^
« |«««M ir* 1^ IP |r<«•«T^|^ ir> 1° 1 1^ ir
Ji nI J3 Ji J3 Ji oviN mIm ovIm oInj5J3 lnl0^l-llo^Nl-^-O^I lolOvfil'l-
_ __«_H MN N MM M__M r^moofiMOO o^oo »«
0 0000 OOMmN MMM
en
0 S
H g
0000 OO"^ 0«rt
0"-)"-)0 inON Ot^
1 nI i 1 1 1 f^loMlNOvlMolf) 1 1 HlrnMlro>olo>iHl Nlc^Nl^^
N MM
--- CO
0 CJ
*
0 ^000 00
1 II 1 1 1 1 ^, 1 r 1 M III III 1 1 + if 0 il^ 1 ?« 1 r^il ^li
'«-i^ror--)M N M'l->on
H N M M
CO
CI.
3
2
—
2
CH-
u
0
1
HI
0
II II 1 1 1 1 1 1 II 1 1 1 1 II 1 1 1 1 1 1 1 II 1 II II 1 1 II 1 1
8
»-i
1
1 1 1 1 1 1 1 1 II 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 II 1 1 II 1 1 1
0
T
00
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
0
00
1
0 0 ■ 00
lllllll ??'lll ?°lllll '.^? 9^111 °P||||||
0
0
0000 000 00
00"^0,, 000 QUI
0 T' 9 T^ 9 97*" 9"^ T' 9 9? T' 9 T ";'P V^r^
\0 <5yD^O vO^mvo^vO^
0
1
OOOO GOO GO
, , , , , , GG»00,, OOG GO
r^ r^ (MO 7100 "^ M T- G tV) "1 P-lOO 'i~i >0 O '^ in
' ' ' o'<i-CTNN>oNwSN>ou-if-- oo\t-avt«^>>5N^'^rof^u-)m <rivo o^>n b
H in MU-JiH U-l U-) mU-JMHW-jm U-) MU-JMU-JM Irj"
0
5-
GOGO GOO GO
m m m O , , GOO OG
O >Orn t^r^ OvO 0"-> ~t-0 T-tn OG ^ n-"-i l-G
' Mt^ N0bMr5fs,MMVo6N6\M lO-^MlO^MNirjMOOioMOrjOl^rxN
0
1
IH
* G G G "^ G G
||||||||l|l|| o8||||| -oSi Qo8 oSl O ^o!? voS
' NO"''' ' 'oovovoo-^f-Mobb ONob « >o v6 wvo ;^
0
0
1
N
lllllllllllllllllllll 9^ ^? ^?^|| ^r^ 9?
a.
3
2
J5
1°
6
0
M
?■
u
0
u-1
1 II 1 1 1 i III III III 1 II 1 1 „?| 1 II 1 M „? 1 1 1 1 1 1
M ».i tj M ">!
'n "-^
0
i
o o o o o o
lllllll 9?^||lllllllll ^? 9? ^r^ ^ 9?> ?^?'
^ t^^l-rON-^-M •^M^Tf-q-MN'tt-
CO
O O G O O
lllllllllllllllllllll 9? oS>op2 o opS, ^&.
0
1
O G O GO
lllllllllllllllllllll ^v^ '^'^ r'?^ 1 1 v^"? r^^
' ' ' ''' ''' ''' ' 'l«t^n-N t^O rOOo O ' ' N oo O -^00 G
c
0
o^jo-jnorj-j.-^-.). i^ ■^ >). in 'n in m to mm lo
OOOoOoOOOoOoOO 00 Oo 00 cooooo 00 00 OOOO 00
CoOnOriCntOCOC/3 JO fr) Cd CrjCOcfl c« CC MM M
^s ^fefe^fe ^ ^ ^ ^S^ ^ ^ ^ ^ ^
o
c
3
J3
bJO
N
4o8
DISCOVERY REPORTS
Si
X!
Q
w
<
O ra
o o o o o o ^ o o
00 O vnu-iO Om ^5
.gv ,00 ,u-» |0p IN ly-'-S IP 1?
M loo fi ' •'iv© 1 r--loloO Osl " N 1 ""••C ">! t^'^lvD
N"f»j"Trror^mMN ^
!z;
^1
OOOOOO oo
0u-jir-»000 tno
1? |0 IN |t-. |- IV^I 1? 1?
'>olr^irjlooool^ool"->t-.iONl"-il lAlt^^lvO
M MM^. wmN fOM N
e21
o o o o o
^ifoo 1 £00 1 ?;,^i £n 1 ^ 1 1 1 1 III III
H
1
M
C
6
u
o
CH-
o
T
o
II II II 1 1 1 1 1 1 1 1 II II 1 1 1 1 1 1 1
o
o
T
1 1 1 1 1 1 1 1 1 1 II 1 1 1 1 1 1 1 1 1 II 1 1
o
f
M
CO
III III III III III III 1 III III
o
00
1
p.
0000 0
III III ON 00 000 Om| OvO|||
t^ «^ <X t-N t^
o
0 0000 00
0 , , , 0 0 0 0 , U-, 0
OiO f<^v<-i Onk-1 00 00 O-t- i/-lt^
\0 \0V0"^0^ VOVO
o
1
in
00000 00
in 10 0 0 in , , , 0 0
r^N 00 '^0 000 VON 00 'nm
>0 iosb rovo rn m ■^ rn 0 \h N r^^o 00 W o-i en N K N
o
1
*-*
00000
00000
Com M« Mm <N)N '=0|llllllll
•^ <0 N locfe fn 00 t-x N »vv6 in t>'^ in ' III III
Ngl^|f 1 1 1 III 1 1 1 Mi 1 1
O
1
M
N
1 1 1 ^^1 1 1 1 1 1 1 II 1 1 1 1 1 1 1 1 1 1 1
a,
3
2
60
S
o
o
M
O
vO
1
0
1 1 1 ??^ 1 1 1 III III ill 1 III III
o
r
00000
00 ui 0 0 0
O'*- -(t^ ON OOTj- 0"||||||||||
H g_0 O^;*- -i-^N .O^Sn f,^:, 1 1 1 III III
0000
opS, .08 -p2 o^Milllllllill
■<fl-^« >0 txN -"TOO "NOqII' III 1 III III
O
1
N
1 1 1 H^S 1 1 1 III III ill 1 III III
c
WS860
WS862
WS864
WS866
WS868
WS870
WS872
WS874
WS875
C/2
W
I— I
J
Q
<
W
OH
0 ^
J
<I1
fa
ffi
0
0
r/-)
H
Z
0
0
Z
I— 1
CO
P3
0
w
w
H
H
<
w
X
H
0
Z
Pi
hH
Ph
Ph
Q
<
X!
Q
w
fa
fa
o
W
e
,0
a
B
'P.
0
u
Cu
<X
<
c
0
~t^^ V'V' , ^ ^ ^ ^
ONU-iTt-moo uiN r^-^
^minN hh-j-mON
0000 00000
I-!
in in
'm N *ON f^ 00 *0 '0 "-" 00
m-^mN ^-"J-oinN
0000 00000
" m moo 0 t-> 1-1 m 0
m m m -^ tJ- -^ m w^ '^
Date
""«« ■-■■-.■-■I...'-'
'mm r^ c*^ m f^
>> > > > > > > -d
t~- 6 pi 00 r~- 6^^ N M
Mm M
*^
c
.0
CS
«5
WS576
WS582
WS583
WS586
Ringdove Inlet
Puerto Acero
Puerto Bueno
Bahia san Nicholas
Connor Inlet
ni mail
Ito aaiziij'^
nioi'i) .ybt
PLATE XVI
Macruronus magellanicus ( x \). Sketch by Mr E. R. Gunther from a
damaged specimen in which the blues were less vivid than in
individuals from previous stations. Note that the extension of
the fins alters the apparent proportions of the body. (From
St. WS99. OTC.)
DISCOVERY REPORTS, VOL. XXIII
PLATE XVI
W:^
E^.
X
m
'P
a
I— f
•<
W
O
<
m
Z
O
OS
P
Pi
U
DISCOVERY
REPORTS
Vol. XXIII, pp. 1-18
Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands
THE GUT OF NEBALIACEA
by
Helen G. Q. Rowett
CAMBRIDGE
AT THE UNIVERSITY PRESS
1943
Price Three shillings net
CAMBRIDGE
UNIFERSITT PRESS
LONDON: BENTLET HOUSE
NEW rORK, TORONTO, BOMBAT
CALCUTTA, MADRAS: MACMILLAN
All rights reserved
PRINTED
IN GREAT BRITAIN
BY
WALTER
LEWIS MA
AT
THE CAMBRIDGE
UNIVERSITY
PRESS
DISCOVERY
REPORTS
Vol. XXIII, pp. 19-36
Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands
ON A SPECIMEN OF THE SOUTHERN BOTTLE-
NOSED WHALE, HYPEROODON PLANIFRONS
by
F. C. Fraser, D.Sc
CAMBRIDGE
AT THE UNIVERSITY PRESS
1945
Price three shillings net
I
CAMBRIDGE
UNIVERSITT PRESS
LONDON: BENTLEY HOUSE
NEJV rORK, TORONTO, BOMBAY
CALCUTTA, MADRAS: MACMILLAN
All rights reser'ved
PRINTED
IN GREAT BRITAI N
BY
b^ LEWIS M A
AT
THE CAMBRIDGE
UNIVERSITY
PRESS
DISCOVERY
REPORTS
Vol. XXIII, pp. 37-102
Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovemment of the Dependencies of the Falkland Islands
REPORT ON ROCKS FROM WEST ANTARCTICA
AND THE SCOTIA ARC
by
G. W. Tyrrell, A.R.C.Sc, D.Sc, F.G.S., F.R.S.E.
CAMBRIDGE
AT THE UNIVERSITY PRESS
1945
Price eleven shillings net
CAMBRIDGE
UNIFERSITT PRESS
LONDON: BENTLEY HOUSE
NEW rORK, TORONTO, BOMBAT
CALCUTTA, MADRAS: MACMILLAN
All rights reserved
PRINTED
IN GREAT BRITAIN
BY
WALTER
LEWIS M A
AT
THE CAMBRIDGE
UNIVERSITY
PRESS
DISCOVERY
REPORTS
Vol. XXIII, pp. 103-176
Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovemment of the Dependencies of the Falkland Islands
THE DEVELOPMENT AND LIFE-HISTORY OF
ADOLESCENT AND ADULT KRILL,
EVPHAVSIA SUPERBA
HELENE E BARGMANN, Ph.D. i -.r
CAMBRIDGE
AT THE UNIVERSITY PRESS
1945
Price fourteen shillings net
CAMBRIDGE
UNIFERSITr PRESS
LONDON: BENTLEY HOUSE
NEfV rORK, TORONTO, BOMBAT
CALCUTTA, MADRAS: MACMILLAN
AH rights reterved
PRINTED
IN GREAT BRITAIN
BY
WALTER
LEWIS M A
AT
THE CAMBRIDGE
UNIVERSITY
PRESS
DISCOVERY
REPORTS
Vol. XXIII, pp. 177-212
Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands
THE ANTARCTIC CONVERGENCE AND THE
DISTRIBUTION OF SURFACE TEMPERATURES
IN ANTARCTIC WATERS
N. A. Mackintosh, D.Sc.
\
CAMBRIDGE
AT THE UNIVERSITY PRESS
1946
Vrice ten shillings and sixpence net
CAMBRIDGE
UNIFER^SITT PRESS
LONDON: BENTLEY HOUSE
NEW YORK, TORONTO, BOMBAY
CALCUTTA, MADRAS: MACMILLAN
All rights reserved
PRINTED
IN GREAT BRITAIN
AT
THE CAMBRIDGE
UNIVERSITY
PRESS
DISCOVERY
REPORTS
Vol. XXIII, pp. 213-222
Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands
CAMBRIDGE
AT THE UNIVERSITY PRESS
1946
Price two shillings and sixpence
f->^
hlEBALlOPSlS TYPICA \L ,
H. Graham Cannon, Sc.D., F.R.S.
Printed in Great Britain at the University Press, Cambridge
(Brooke Cnitchley, University Printer)
and published by the Cambridge University Press
Cambridge, and Bentley House, London
Agents for U.S.A., Canada, and India: Macmillan
DISCOVERY
REPORTS
Vol. XXIII, pp. 223-408
Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands
REPORT ON TRAWLING SURVEYS ON THE
PATAGONIAN CONTINENTAL SHELF
Compiled mainly from manuscripts left by the late E. R. Qunther, M .A.
T. John Hart, D.Sc.
CAMBRIDGE
AT THE UNIVERSITY PRESS
1946
Price thirty'five shillings net
Printed in Great Britain at the University Press, Cambridge
{Brooke Crutchley, University Printer)
and published by the Cambridge University Press
Cambridge, and Bentley House, London
Agents for U.S.A., Canada, and India: Macmillan
DISCOVERY
REPORTS
Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovemment of the Dependencies of the Falkland Islands
Vol. XXIII, pp. i-vi
TITLE-PAGE AND LIST OF CONTENTS
\' L
CAMBRIDGE
AT THE UNIVERSITY PRESS
1947
Price one shilling net
Printed in Great Britain at the University Press, Cambridge
(Brooke Crutchley, University Printer)
and published by the Cambridge University Press
{Cambridge, and Bentley House, London)
Agents for U.S.A., Canada, and India: Macmillan
iiiiip
iiiiiii
' :^r
liil
ill
iiiiiiiii