^^^■' '"

'*'.*-

CP

^

DEPARTMENT OF THE NAVAL SERVICE

.■^

'-^■•Col

CANADIAN FISHERIES EXPEDITION, 1914-1915

nvestigations in the Gulf of St. Lawrence
and Atlantic Waters of Canada

UNDER THE DIRECTION OF

DR. JOHAN HJORT,

Head of the Expedition

Director of Fisheries for Norway.

OTTAWA

J. DE LABROQUERUE TACH6

PRINTER TO THE KING'S MOST EXCELLENT MA.TESTT

6551— A

LIST OF ERKATA TO BE NOTED BY THE READER.

P, 28, Fig. 20. Merluccius not Merlucius.
P. 30, Fig. 21. Mallotus not Malotus.
P. 32, Fig. 22. Mallotus not Malotus.

P. 118, Fig. 29. The second figure 6 at the bottom of the figure should be 8.

Coloured Plate 1 (Sandstrom's Report). 83 not 33 between 67 and 31 NE. of Station
VIII.

" " 2 (Sandstrom's Eeport). The figure 4 below V is reversed in error.

" " 3 (Sandstrom's Report). In the explanation X — XX should be

XIII — XX, and the small triangular area close to 65 should
be yellow not white. Also the figures 82-88 should be 83-89 and
79-81 consecutively.

" " 4 (Sandstrom's Report). Omit minus sign ( — ) in legend. It should

be <0°C not <— 0'°C.

" " 4 (Sandstrom's Report). The triangular space with V in the centre

should be white, not coloured red.

" " 5 (Sandstrom's Report). Midway between XVIII pi.d XIX the figures

should be 89 not 79.

DEPARTMENT OF THE NAVAL SERVICE.

CANADIAN FISHERIES EXPEDITION, 1914-1915, IN THE GULF OF ST.
LAWRENCE AND ATLANTIC^ WATERS OF CANADA,

Under the Direction of Dr. Johan TJjort, Director of Fhheries for Norway.

Scientific results embodied in nine memoirs as follows :—

Johan Hjort: Introduction to the Canadian Fisheries Expedition, 1914-1915.

Alf. Dannevig: Canadian Fish-Eggs and Larvae.

Einar Lea: Age and Growth of the Herring in Canadian "Waters.

A. G. Huntsman: Growth of the Young Herring (so-called sardines) of the Bay of
Fundy.

Arthur Willey: Eeport on the Copepoda obtained in the Gulf of St. Lawrence and
adjacent waters, 1915.

fT. W. Sandstrom: The Hydrodynamics of the Canadian Atlantic Waters.

Paul Bjerkan : Eesults of the Hydrographical Observations made by Dr. Johan Hjort
in the Canadian Atlantic Waters during the year 1915.

A. G. Huntsman: Some Quantitative and Qualitative Plankton Studies of the Eastern
Canadian Plankton.

1. Introduction.

2. Quantity of Plankton.

3. A Special Study of the Canadian Chsetognaths, their Distribution, etc., in

the Waters of the Eastern Coast of Canada.

H. H. Gran : Quantitative Investigations as to Phytoplankton and Pelagic Protozoa
in the Gulf of St. Lawrence and outside the same.

0551 — Aj

-*.#^

. L. I ■ R A R ^ , ^

PREFACE.

J3y Professor Edward E. Prince, LL.D., Dominion Commissioner of Fi^herie»
and Chairman of the Biological Board of Canada.

The series of reports on the results of the Canadian Fisheries Expedition, 1914-15,
witli the important introductory memoir by Dr. Iljort, Director of Fisheries under
the Government of Norway, and head of the expedition, are weighty contributions to
our knowledge of fish-life in the sea. They inform us, also, as to the complex condi-
tions of marine life generally in our Atlantic waters, and a fe^v preliminary words
:5eem necessary in explanation of the origin, character, and scope of the expedition.
About ten years ago the Biological Board of Canada, of which I am chairman, decided
to ask Dr. Hjort if he could take charge of a comprehensive fishery investigation in
Canadian waters on the lines of the researches which, under his direction, had proved
so beneficial to the fisheries of Norway. In spite of the fact that the great cod and
other fisheries off our Atlantic coast had been carried on for centuries, it was felt
that there were doubtless hidden possibilities of development and expansion that
awaited only a basis of accurate knowledge to turn them to account. Certain urgent
problems affecting the herring resources of the gulf of St. Lawrence and the Atlantic
coast adjacent seemed to call for adequate scientific investigation. After some corre-
spondence with Dr. Hjort, carried on by myself and by Professor McBride (then of
McGill University, and a member of the Biological Board), the matter was allowed to
remain in abeyance. Dr. Hjort could not at that time visit Canada, though he sug-
gested the name of an able young Norwegian scientist who might undertake the work.
The proposal was revived by the Biological Board five years ago, with the result that
after much correspondence. Dr. Hjort came to Canada, and arrangements for the
expedition were completed in Ottawa, after conferences between the Deputy ^linister
(Mr. Desbarats), the Biological Board, and Dr. Hjort.

As the expenditure involved in carrying out a well-planned df cp-sea investigation
was too considerable to be readily met by the Biological Board, the Deputy Minister
of the Naval Service cordially agreed to an arrangement whereby the department
would provide the amount necessary. Dr. Hjort and I met in conference a number
of times and, as a result, a plan of work was outlined, which is summarized in the
Introductory Report. The plan embraced the necessary physical, hydrographical,
chemical, and biological researches, including collection of water samples, etc., and
the determination of the plankton collections, distribution and varied abundance of
the young fry, as well as the eggs of the cod, haddock, flat-fishos, and other species,
in addition to the thorough study of the varieties, distribution, migrations, and breed-
ing of the herring. The survey, in short, was to be as complete as possible in order to
afford a basis for the future development of the fishing industries.

Eight years ago Dr. Hjort had carried on investigations in Newfoundland v.aterp.
in conjunction with the late Sir John Murray, on the well-known cruise of the Nor-
wegian Government steamer Michael Sars. The Canadian investigations now planned
would, it was thought, advance some aspects of the important results achieved by
Dr. Hjort and Sir John ^Murray. They certainly show, as demonstrated in the Pre-
liminary Report on the "Natural History of the Herring. 1014," by Dr. Hjort,
published by the Naval Service Department in 1915. in what a unique degree the
Atlantic waters of Canada, especially the gulf of St. Lawrence, a thorough scientific
investigation can aid in the solution of important fishery problems so exhaustively
studied in Norway. These problems centre round the relation between the distribu-
tion and life-cycles of fishes, and other organisms, and the prevailing environments

vi PREFACE

obtaining in successive seasons of the year, the physical, chemical, biological, and
other conditions, which so profoundly affect animal and plant life in the sea.

Dr. Hjort has succinctly summarized in his Introductory Keport his own main
results, and the results obtained by the scientific staff working under him, in connec-
■ tion with the expedition.

The herring resources of the Atlantic coastal waters of Canada and of the gulf
of St. Lawrence offer a field for enormous commercial development, and Dr. Hjort
gave primary attention to the study of the Canadian herring.

As the Preliminary Report on the various types of Canadian herring described
by Dr. Hjort was published three years ago, it is not necessary to repeat its main
points here, and Mr. Einar Lea's further and more elaborate detailed results must be
studied in his report now presented; but it may be pointed out that, on the whole,
four very different groups of herring may be distinguished, viz. : —

(1) The Newfoundland type (with the remarkable exception of the St. George
Bay herring), which grows slowly during the first summer, but from the third summer
on grows at a rapid rate. The samples studied ranged from four to twenty years in
age, and the age-group of 1904 was dominant, recalling Dr. Hjort's discovery of a
dominant age-group in JSTorway — a single year-class maintaining a dominant position
in successive season's catches for several years. The typical Newfoundland herring
in growth and age composition of the schools form a group distinct and apart, as
demonstrated by all the samples studied.

(2) The second type are formed by the herring off cape Gaspe, Magdalen islands,
and Northumberland straits, roughly embracing the west and northwest waters of the
gulf of St. Lawrence, which grow rapidly during the first summer, and slowly in later
years, and are thus smaller later on than Newfoundland herring of the same age.
The 1903 and 1907 year-classes dominated.

(3) A third type from the Cape Breton (Atlantic) coast, from North Sydney
south, which have a moderate first-summer's growth, but a more rapid growth from
the third summer on. The divergent type from bay St. George, Newfoundland, belongs
to this group. The dominant age-group is that of 1903.

(4) A fourth type from Chedabucto bay and southwest along the Nova Scotia
shore, which grows well during the whole of its first five or six years, and outdistances
all the other types of herring. The 1911 year-class was found to prevail, while in the
large, older herring from Nova Scotia, the 1908 year-class dominated, and in the
Halifax and Lockport samples the 1908, 1910, and 1911 year-classes were best repre-
sented.

Mr. Einar Lea makes the important observation in his report that the samples of
herring which resembled each other in age-composition, also exhibited great resem-
blance as regards growth, hence the types signalized would appear to be valid and
unquestionable, though more extended researches are urgent to fvilly confirm these
results.

The special investigations upon the herring yielded conclusions so interesting
that it seemed necessary to extend the studies in hand to other species, and to eluci-
date faunistic conditions generally in all the Gulf and Atlantic waters embraced in
the scheme.

Problems as to the influence of the earth's rotation, the effects of the annual
melting of the great fields of ice, the features of temperature, salinity, specific gravity,
etc., of the sea water, and other matters vital to the vertebrate and invertebrate life,
and especially fish-life, off our Atlantic coast, were included.

Professor Gran's masterly plankton studies show that in the gulf of St. Lawrence
there is a plentitude of minute floating plants of northern types, whose development
each season appears to be annually much later than in the northern European waters.
The prevalence of an Arctic temperature for so large a part of the jeav is the explana-
tion.

PREFACE vii

On the spawning, the eggs, and the young stages of food- and other fishes, import-
ant observations were made. Mr. Danneviig's report on the material obtained is
exceedingly interesting. Thirty-six species were determined, which belong to twenty
families, the cod being most important, and its eggs and larvae occurred during a long
period, May to August. Haddock were more rare, one egg only being secured in the
gulf of St. Lawrence. Nor were mackerel egg numerous, while very few flat-fish eggs,
or young, were found, excepting the Canadian Plaice.

The eggs and young of fishes and other animal forms constitute an important
part of the plankton, but the scarcity of these eggs and young is a striking feature.
Over the northern portions of the gulf, there occurred, in May and June, eggs of cod
and other gadoids, also flatfishes, and in a marked degree the eggs of the Canadian
plaice {Rippoglossoides, or Drepanopsetta) , with a few larval fish of Arctic types such
as the northern wolf -fish (Anarrichas latifrons), capelin (Mallotus), and others.
Mackerel eggs were taken in Cabot straits, and considerable quantities of Norway
haddock or rose fish (Sehastes) over the deep (Laurentian) channel leading from the
gulf into the open ocean. Expectations of greater quantities of floating fish eggs, in
the later summer cruises, were disappointed. Cod eggs occurred, and a very small
number of young fish, while northern forms such as Mallotus, the capelin, occurred,
and the more southerly types such as mackerel and Sehastes, as before. Outside of
the gulf, cod eggs were scanty, but the whiting (Merluccius), sometimes called hake,
were plentiful. For this scarcity of eggs and young fry the usual explanation (as in
European waters) is that they are swept in vast quantities from the spawning and
hatching grounds, and may be scattered over vast distances outside the gulf of St.
Lawrence. There is no reason to suppose that after the huge schools of parent cod
have cast their eggs into the waters inside the gulf, they perish, although the tempera-
ture may be as low as the freezing point, or lower, 31° to 32°F. ( — 1-a to 0°C.). in
Europe, floating eggs such as those of the cod rarely occur in water under 35° or 36° F.,
but the fishing experiments in the gulf of St. Lawrence showed cod to be abundant
in the vast cold water layers immediately above the bottom, and the eggs are deposited
under these frigid conditions, and undergo their early stages of development there.
It is truly a remarkable result of Dr. Hjort's Canadian researches to find (as he
states), " cod spawning on the floor of the banks in water of absolutely Arctic tempera-
ture," while at the surface, immediately overhead, more southern forms, such as the
mackerel, also spawn.

Dr. Huntsman's interesting quantitative and qualitative plankton research con-
firms the observations, made during the expedition, that the minute floating life in
the Canadian Atlantic waters is more abundant in colder water and where the water
is deep, while on the whole, these myriads of small living organisms are at a deeper
level during the day than during the night.

Professor Willey and Dr. Huntsman made a study of special plankton groups,
the former reporting on the Copepods, the latter on the Chaetognatlis, and their reports
demonstrate a remarkable agreement between distribution and the various water layers
so exhaustively treated by Mr. Sandstrom in his elaborate report on the hydrodynamics
of the waters investigated, and by Mr. Paul Bjerkan in his very thorough hydro-
graphic investigations, including the distribution of temperature and salinity. The
former reports on the phyics of the sea — too technical and detailed to summarize,
but the presence of a great cold intermediate layer is a striking feature, and due, as
pointed out in the report, to ice melting in the gulf or even in the more northerly
Arctic areas, probably as distant as Greenland. The earth's rotation causes it to tend
to the east, or right, and forces it out via cape North into the open ocean.

The salinity features of the gulf and coastal waters, as determined by ^Ir.
Bjerkan, are of peculiar interest. The outflow of relatively fresh water from the
superficial layers in the gulf is more marked in summer than in spring, and one of
its effects is to force seawards, the superficial bank water of higher salinity, 33 to 33,
this influence being noted to a depth of 25 m., but the Salter water, deeper down.

viii PREFACE

forces the banlv water back, especially in the Labrador channel (at depths of 75 to
100 in.), where also the Atlantic water of high salinity, 35°, penetrates farther in
towards the channel in the spring at a depth of 100 to 200 m. As to temperature, the
turface of the gulf is warmer in summer than in spring, owing to the (fresher)
coastal water and the Gulf Stream. In the straits of Northumberland it is as high as
18-5° C. (about 65° F.), while the lowest record is 11-65° C. close up to the north
"shore. In the inner parts of the gulf at 10 m. depth the temperature ranges about
10° C, but in the outer parts at 20 m. an extremely low temperature prevails at 50
to 80 m., especially in the northern parts. There must be a constant inflow of frigid
Arctic water through the deeper parts of Belle Isle straits, as the cold water layers
are more extended in summer than in spring, though this cold influence appears, off
the Newfoundland banks, to decrease in summer. The influence of the Labrador
current and the Gulf Stream introduces many complications, as Mr. Bjerkan shows,
an(| the water-masses of continental origin, of low salinity, add to the complexity,
both superficial and at greater depths. The effects potently influence the breeding,
hatching, and larval development of important food-fishes, and throw light on the
migrations and the distribution of the schools of adult-fish on which our great Atlantic
fishing industries depend.

CONTENTS.

Page

Johan Hjort : Introduction to the Canadian Fisheries Expedition, 1914-1."). . i-xv.

Alf Dannevig: Canadian Fish-Eggs and Larva; 1

Einar Lea : Report on The Age and Growth of the Herring in Canadian Waters. . 75

A. G. LIuntsman : Growth of the Young Herring (so-called sardines) of the Bay

of Fundy 165

Arthur Willey : lleport on the Copcpoda obtained in the Gulf of St. Lawrence

and adjacent waters, 1915 173

J. "W. Sandstrum: The Hydrodj'uamics of the Canadian Atlantic Waters .. .. '221

Paul Bjerkan : Results of the LTydrographical Observations made by Dr. Johan

Hjort in the Canadian Atlantic Waters during the year 1915 349

A. G. Huntsman : Some Quantitative and Qualitative Plankton Studies of the

Eastern Canadian Plankton 405

1. Litroduetion 405

2. Quantity of Plankton 407

3. A Special Study of the Canadian Chaetognaths, their Distribution, etc.,

in the Waters of the Eastern Coast 421

H. H. Gran: Quantitative Investigations as to Phytoplankton and Pelagic

Protozoa in the Gulf of St. Lawrence and outside the same 4^6

"; >-^ ''± * '^

INTRODUCTION TO THE CANADIAN FISHERIES EXPEDITION,

1914-15.

By Jo MAX Hjort,
Director of Fisheries for Xorway, and Head of the Expedition.

(With Five Figures.)

In the course of my researches in Xorth European waters, it has frequently
occurred to me th^t many problems of long standing in the sphere of fishery and
marine investigation might perhaps best be solved by making a comparison between
the two separate areas of sea which contain the same forms of animal life, viz., the
waters of northern Europe, and the range of sea from the coast of Labrador and
Canada to that of Maine.

In 1910, I was invited by Sir John Murray, himself Canadian-born, to make
a cruise on board the ss. Michael Sars, belonging to the Norwegian Fisheries Depart-
ment, the voyage in question extending over the greater part of the northern Atlantic.
Here, naturally enough, the same idea once more asserted itself, and we both felt that
it would be desirable to undertake, at any rate, some slight preliminary investigations
in the Canadian waters, and there make test of the same methods of research as have
been developed, during the course of the past generation, in the fishery investigations
of northern Europe, and the International Council for the Exploratiion of the Sea.

As mentioned in my account of this voyage^ it was also our desire " to set
our course from the Azores to the Bermudas, and thence on to Boston, finishing with
a series of short zig-zag sections between the land and the edge of the coast-banks,
till we reached Newfoundland. "We should in that case have been able to study the
remarkable transition that occurs on passing from the almost tropical conditions of
the Sargasso sea to those of the icy Labrador stream, which creeps southward along
the Labrador coast from Baffin's bay to Newfoundland, and even farther south. Tlie
short time at our disposal made this impossible, and we were compelled to cross from
the Azores to the nearest coaling station, namely, Newfoundland, and then make for
home."

On the way from the Sargasso sea to Newfoundland, however, we had occasion,
after all, to make certain observations, the results of which still further convinced mo
of the great and peculiar interest attaching to such a comparison as that mentioned.
Indeed, this last cruise in itself sufficed to show in what unique degree the waters off
the coasts of Canada and Newfoundland were suited to the study of those very prob-
lems which have ranked foremost in the Scandinavian marine researches of the past
generation, to wit, the relation between the distribution and life-cycle of the organisms,
on the one hand, and the prevalent physical conditions in the sea on the other.

On this cruise, from the Sargasso sea to the Newfoundland waters, we encountered
the sharpest and most remarkable transitions between warmer and colder water layers,
and in the closest connection therewith, a constantly coinciding occurrence of plant
and animal communities, which were now of tropical, now of distinctly northern
boreal character. It was these observations of ours which furnished me with the lead-
ing principles on which the investigations described in the following pages were subse-
quently based. And as, moreover, these earliest discoveries of ours afford a kind of

1 Sir John Murray and Johan Hjort : The Depths of the Ocean. Macmillan, I-ondon.
1912. Chapter III, pages 99 and following.

xi

xii DEPARTMENT OF TEE NAVAL SERVICE

rough introductory survey of those Canadian waters which I was later able to study
in closer detail, it may perhaps not be out of place to give some of the leading features
here.

Fig. 1 shows a temperature and salinity section from the Sargasso sea to New-
foundland. " At stations 64 and 65^ we see the vast layer, with a salinity of over 35
per thousand and high temperature down to considerable depths, the same as found
bj us over the whole distance from away beyond the Canary islands.

" On our way north from station 64 on 2'8th June we saw patches of Sargasso weed
all the morning, and numbers of flying fish, about 10 centimetres long, started up
in front of our boAvs. This led us to believe that we should capture the same forms
as before, when we lowered our pelagic appliances in the evening at station 66. Great
was our astonishment, therefore, to discover next morning on hauling in our appli-
ances that the catches mainly consisted of true " boreal ". plankton, that is to say,
animal forms which we were accustomed to get in the so-called extension of the

Newfoundland bank

atat7Z 71 70

Cargasso Sea

3-88-34 96

3 7 2-3^ "-i

Fig. 1. Hydrographical Section from the Sargasso Sea to the Newfoundland Bank.
Gulf Stream in the ISTorweg^an sea right up to the very shores of Sitsbergen. There
was the amphipod Euthemisto, the copepod Eucliceta, and " whale's food " (the ptero-
pod Clione limacina) , large quantities of which are met with from time to time in the
waters between Spitsbergen and the north of Norway. This last is not an " arctic "
form, that is, it is not associated with polar water in the Norwegian sea, but on the
contrary is found in Atlantic water to the south of Iceland, according to Danish
observations. It seems, however, to be associated with the northern portion of the
Atlantic and the Atlantic water that enters the Norwegian sea. These animal forms
were entirely absent during the whole of our cruise from the Canary islands to
station 64, so that their occurrence at station 66, where lower temperatures were
recorded at no great depth beneath the surface, is very significant.

1 hoc. cit.

iC'.-1-A.i/'/ 1 \ ri>iii:h'ii:s i:.\I'i:iuti(>\. r.ii',-ir, xiii

'• We fancied now that we had said farewell to the Sargasso sea aiixi its interest-
ing animal life, but at stations 67 and 69, in close accordance with the hydrographical
conditions depicted in fig. 93 (fig. 1), we came once more across more southerly forms.
In the upper layers there were the same young fish, many of them with .stalk-eyes, and
Leptocephali, while flying fish, Sargasso weed, and the familiar Sargas.so animals
were all once more in evidence.

" We found a large cluster of eggs, weighing approximately a kilo., drifting about
at station 69, belonging to the common angler-fish (Lophius piscatorius) , the develop-
ment of which was studied by Alexander Agassiz; we hatched out the eggs and
obtained the stages depicted by him. Angler-fish only inhabit the coast banks, so
that our find of slightly developed eggs, that could not have been drifting many days,
indicated that we were now in the neighbourhood of the American coast bank.

" In deep water we found once more at stations 67 and 69 the deep-sea animals
of the Sargasso sea, that i-S to say, all the black fishes and red crustaceans which we
have so often mentioned already. There were not merely the commonest kinds of
small fish, but also large ones (such as three examples of Gastrostomns), and fishes
which are caught in other oceans (Aremtia.'^. Serri vomer).

60

fLEMliH CAP

.BANK OF ; :
\ST PIERRE :-G.».'h

^

'3iABI.E 1.

B A \ N K

70

Fig. 2. " Michael Sars ' Stations 69 to 8u.

'•' At station 70, on the edge of the coast bank, where the depth was 1,100 metres,
we discovered that we had for the second time left purely oceanic conditions behind,
and once more the true boreal plankton appeared in the surface layers. There was
the little copepod Ualanus finmai'chicus, the commonest crustacean in the Norwegian
sea, and we also now met with Euthemisto, Nyctiphanes, Krohnia liamata, Limacina
helicina^ and Clione limacina^ all species that are regarded as specially characteristic
of the Norwegian sea. Still in the deep water from 350 metres down to 1,100 metres
we continued to get the familiar pelagic deep-sea fish Cyclothone signata and C.

1 Limacina was taken in numbers by Haeckel and Murray off Scourie in Scotland.

xiv DEPARTME'NT OF TEE NATAL SERVICE

microdon, as well as the medusa Atolla and other forms; so that the area of distribu-
tion of these animals extends from Africa to North America, that is to say, in all the
water from one continental slope to the other."

"The coast bank itself (fig. 94 [2]) offered us a totally different field for study,
which no doubt would have proved very interesting, but unfortunately our time was too
short to attempt systematic researches ; we had to steam for our coaling station, con-
tenting ourselves with one or two shallow stations on the way.

" Fig. 95 (3) shows the hydrographieal conditions from our last true oceanic station
(69) to a station (74) just off St. Johns. It is extraordinary what a sudden change
there is from the warm salt oceanic water to the cold coast water. The curves of tem-
perature and salinity between stations 69 and 70 go down straight like a wall — the
well-known " cold wall " of oceanographers. Over the bank there is a surface layer,
about 40 metres in depth, with a temperature of over 6°C., similar to what we get
in the boreal portion of the Norwegian sea along the coast of Norway. Below that,
however, the temperatures are under 2°C., and even as low as - 1-5°C., that is to say.

Fig. 3. Hydrographieal Section across the Great Newfoundland Bank.

the water may be as cold as what Nansen found near the North Pole. Probably at no
other part of the globe are there such peculiar temperature conditions — conditions
comparable with those in the Arctic regions, though the latitude is the same as that
of Paris. It would have been an agreeable task to trace these conditions by follow-
ing up the currents and animal life, both northwards and southwards. Still, even our
random investigations furnished interesting results. Thus we discovered that from
station 70 to St. Johns there was the same northerly plankton already mentioned, and
an examination of the young fish showed that they accorded with what had previously
been found by Norwegian naturalists off the coast of Norway, and by the Danes south
of Iceland.

" On the outer side of the coast bank, at station 71, we met with larvae of red-fish
(Sehastes). At station 72 there were cod-eggs and numbers of little cod-fry, besides
fvdly developed eggs of haddock (Gadus ceglefinus) and haddock larvae, 3-| millimetres
in length and upwards, and also young fish of the boreal long rough dab (Drepanop-
seita). At station 73 we came across eggs of this dab (besides a number of eggs that
we have not yet determined), and the shallow- water form Ammodytes. At station 74
there were neither eggs nor young fish.

" Similar catches are taken off the coasts of Norway and Iceland ; near and just
beyond the continental edge there are larvae of red-fish, and on the bank, in 30 or 40
fathoms of water, there are larvae and eggs of cod and haddock. It was interesting to

CANADIAN FISHERIES EXPEDITION, 191/rlo xv

find the eggs and larvae of these fish at station 72, where the bottom-temperature was
between 2°C. and4-6°C., whereas nearer land, where the bottom-temperature was 0°C,
or even less, they were absent."^

In 1915, I had the honour of being invited by the Biological Board of Canada to
make a stay of some months' duration in that country in order to study the Atlantic
herring fisheries of Canada; an invitation which I was extremely happy to accept. I
could not, of course, hope to accomplish very much in so short a time, especially as I
had no prospect of being able to procure the necessary means for practical research
work at sea. The investigations of recent years, however, with regard to the growth of
various species of fish — and particularly of the herring — had shown that it was pos-
sible, by studying the growth of the fish and age composition of the stock, as expressed
in the annual rings of the scales, to form a remarkably close estimate, not only of the
biology of the separate species, but also of the conditions in the sea wherein they
occurred. It seemed to me, therefore, worth while to see whether such investigations,
albeit here necessarily of an occasional and by no means final character, might not open
up fresh points of view, and lead to further and more detailed study in the same field.

The main objects of such scale investigations should then, it seemed to me, be for-
mulated as follows : —

1. Do the herring that visit the Atlantic coast of Canada all belong to a single
race or type, or is it possible to distinguish several races in these waters?

2. Does the rate of growth vary according to the conditions of the waters along
the coast? Can types of different growth be distinguished and defined?

3. Is the renewal of the stock of herring of a constant character, or are there the
same great fluctuations in the stock, i. e., in the number of individuals in the different
year-classes, as in European waters ?^ '

The first two problems, or groups of problems, are of course identical with the
problems of the distribution and migrations of the herring. If the Atlantic stock of
herring can be shown to belong to several different races, then of course the area of
distribution and migration of each race or type may be defined by a study of samples
of herring taken from different localities along the whole coast.

The third problem is of the greatest importance for any elucidation of the old
riddle — the fluctuations in the yield of the fisheries — this being to a very great extent
dependent upon the fluctuations in the number of herrimgs living in the sea at the
time.

On arriving at Halifax, therefore, in October, 1914, I first of all endeavoured to
organize a collection of material. I had no other means at my disposal than such as
could be contained in the not very extensive luggage of an ordinary traveller, and it
was thus useless to think of anything beyond samples drawn from the catches brought
in by the fishermen' themselves. In other words, my material would have to consist
of salted herrings purchased from various localities. Thanks to the very kind assist-
ance afforded me by the Biological Board, especially by the Dominion Commissioner
of Fisheries, Prof. E. E. Prince, and by Prof. A. B. Macallum, Secretary-Treasurer of
the Biological Board, who endeavoured by every means in their power to facilitate my
researches, I succeeded in obtaining samples from various parts of the gulf of St.
Lawrence, and from Newfoundland. It was necessary, however, to form some idea
as to the representative value of the samples thus obtained, and with this end in view
I made a .iourney along the coast, in the autumn of 1915, and over to Newfound-
land, visiting the fishing stations, and taking every opportunity of ascertaining, by
conversation with the fishermen, what kind of implements were employed for the
capture of herring, and what experience the fishermen themselves had acquired as
to the occurrence of the fish. In some places, I was able myself to study the fishing
in progress, and examine the implements used. The fishermen everywhere, almost

1 Loc. cit. p. 106, and following.

2 See my paper : Fluctuations in the Great Fisheries of Northern Europe. Rapports et
Proces-Veirbaux, Vol. XX, Copenhagen, 1914.

xvi DEPARTMENT OF THE NAVAL SERVICE

without exception, use gill-nets with a certain fixed size of mesh (2| to 2| inches).
The nets are placed along the sea-bottom on the coast or in the bays or inlets along
the shore. At no point is fishing carried on far out from the coast in deep water, or
on the surface (by drift-nets or by purse-seines).

As pointed out in a preliminary report of my journey'^ " this particular method
of fishing " naturally " has great disadvantages for the study of the life-history of
the herring. The big meshes of the fishermen's nets can procure samples of the large
mature herring only, and it is further quite uncertain whether the samples are in any
way representative of even the mature shoals or not. It may be that the fishermen,
through a long experience of fishing in these waters, have been able to adopt a size of
mesh which takes practically all the sizes of mature herring visiting the coast, but
only by means of experiments, carried out with gear taking all the sizes probably
occurring, can this question be satisfactorily answered."

In the course of the winter 1914-15, I had now occasion to study the samples col-
lected, primarily with a view to ascertaining how far it might be possible, with such
material, to arrive at an understanding of, at any rate, some points in the natural
history of the herring. The biological laboratory at the University of Toronto afforded
me excellent facilities for this work, and it was there that I can-ied out the investiga-
tions referred to in my preliminary report above referred to.^

As will be seen from this, the samples collected distinctly showed that there are
on the coasts of Canada types of herring differing widely one from another. The
scales were eminently suited for investigation purposes in the case of all samples
from the northern parts of the waters concerned, as, for instance, those
from Newfoundland and gulf of St. Lawrence (e.g. Magdalen islands), whereas the
samples from the west coast of Nova Scotia were far more difficult to deal with.

It was very interesting, at the outset, to find that samples from so small and
restricted an area as the gulf of St. Lawrence should present such clear and definite
differences in the manner of growth as those here found. The herring from the west
coast of Newfoundland, for instance, were distinguished by a poor growth during the
first years, and a long continued later growth, while those from the waters around
Prince Edward island and the Magdalen islands showed considerable growth for the
first years, followd by a more rapid decline in the annual increment. This difference
Avas, moreover, strongly supported by the results arrived at on studying the age com-
position of the samples. The Newfoundland saniples, both those from the autumn of
1914 and those from the spring of 1915, all showed a distinctly marked abundance of
fish from the year-class 1904, whereas in the samples from the southern parts, a very
different composition was apparent.

Some preliminary investigations as to the " racial " characters of the herring
(number of vertebrae, keel-scales, etc.) revealed a state of things such as, in European
waters, has only been observed in the Baltic and the White seas ; that is, from enclosed
waters with a very low winter temperature and low salinities. These racial investiga-
tions likewise tended to support the view obtained by growth investigations and study
of the age composition, to wit, that real differences were discernible.

I had now succeeded in making clear, even with the primitive methods here
employed, tliat the herring from the coastal waters of Canada differed widely as between
one part and another of the region concerned. This result in itself must be regarded
as of great importance, since it threw light upon the area of distribution (migration)
of the herring, and paved the way for closer investigations in the future.

It seemed to me, therefore, desirable to endeavour to carry out more comprehen-
sive investigations on a larger scale, throughout the whole of the waters in question,

1 Investigations into the Natural History of the Herring in the Atlantic Waters of Canada,
1914. Supplement to the fifth annual Report of the Department of the Naval Service for the
fiscal year ending March 31, 1915, Ottawa, 1915.

2 Loc. cit.

CANADIAN FISHERIHi^i EXl'EDlTKtX, 191.',-15 xvii

and instead of restricting the work to a single species, and a single method of operation,
to employ, as far as possible, most of the methods which have been developed in fisherj-
investigations of recent years.

Tliis proposal was most kindly received, by the Biological Board, in the first
instance, and subsequently also by the Department of Naval Service, in particular by
the Deputy Minister, Mr. G. J. Desbarars, himself keenly interested in scientific
research. I therefore conferred with the Dominion Commissioner of Fisheries, Prof.
E. E. Prince, and drew up, in collaboration with him, a programme of work. The idea
was to collect, by short cruises made with vessels belonging to the Canadian Govern-
ment, a quantity of material such as would serve to elucidate both the conditions
with regard to marine currents, and the character of the fauna in the sea from New-
foundland to Halifax, and in the gulf of St. Lawrence itself.

There being no vessel in Canada specially built and equipped for fishery investiga-
tions, and the general tonnage available being much in demand for other purposes, it
was necessary to confine operations to the carrying out of veiy simple investigations,
comprising short cruises at times previously determined, and along certain definite
routes. It was evident, moreover, that only the simpler forms of implements and
apparatus could be used; for the hydrographical work, for instance, water bottles and
thermometers, and for the fishery work, silk nets for studying the distribution of fish
eggs and plankton. I considered it highly desirable to cover the same routes at least
twice; once in the spring, at the time when the most important species of fish would
presumably be spawning, and once later on in the summer, when we might hope to pro-
cure fish lavvce in the more easily recognizable stages.

Apart from the work on these routes, it seemed to me that we should also endeavour
to carry out fishing experiments with a small steam drifter, the No. 33, belonging to
the Government, and which had previously beeu employed for practical experiments.
I hoped thus to obtain samples of another and more valuable sort than those taken
from the fishermen's hauls, and also to ascertain whether herring could be taken with
drift nets of different sizes, especially in the gulf of St. Lawrence.

The plan thus formed was most cordially and liberally adopted by the authorities
concerned, and in the spring of 1915 we set to work getting together the requisite
implements and apparatus. I was here fortunate in being able to procure the assist-
ance of two of my former colleagues, Capt. Thor Iversen and Mr. Paul Bjerkan, with
whom I had worked together for years past, and who now found time to come over for
some months in 1915 and help in the work. With their aid, the following instruments
and apparatus were collected.

(a) For the hydrographical work:

Six Nansen water bottles, made in Norway.

Ten Richter reversing thermometers from Schmidt and Vossbcrg, calibrated to
0-1° Centigrade, delivered by the International Hydrographical Laboratory at Copen-
hagen, which also provided us with two liand-winchos, with wire and reels taking about
500 metres of wire of about 4 millimetres diameter. A meter wheel of the usual pat-
tern for oceanographical research was used for determining the depths. Several water
bottles could be used at the same time.

(6) For the hiological ivorh:

A large number of Michael Sars plankton nets of 1-meter diameter, and oi the
type found most suitable by the Norwegian Fishery Investigations. A description of
this form of net will be found on p. 46 of my account of the cruise of the Michael Sars
in the Atlantic in 1910 ^. Several of the nets were also furnished with closing
mechanism, according to the model described by Nansen.

1 " Depths of the Ocean."
G551— B

xviii DEPARTMENT OF THE NATAL SERVICE

(c) For use on hoard the fishing steamer No. 33:

A number of drift nets with different widths of mesh, with cable and buoys. A
number of cod lines and other implements for capture of cod and other fish, as aiso a
small shore seine.

At the commencement of May, 1915, I made a reconnoitering tour to Prince
Edward island, together with my friend. Prof. Arthur Willey, of McGill Univer-
sity, Montreal, who, to my great satisfaction, had agreed to take part in the expedition.
On our arrival there, on the first of May, the sea all around the island was still full of
ice as far as one could see, and it was stated that the pack ice lay north of the island
and round the Magdalen islands. This state of things lasted all through the week,
until the 8th of May. We learned that the ice had set southward over toward Prince
Edward island from the northern parts of the gulf of St. Lawrence. Northumberland
Ftrait and the Pictou coast were also blocked by ice, even the steamer connection hav-
ing been stopped. The C.G.S. Princess twice attempted to force a passage through the
ice in order to take Professor Willey and myself out on a preliminary survey cruise.
-At last she reached Charlotteto^vn, and we started on the 10th of May for the Magdalen
islands. We succeeded in taking some few stations, which gave some interesting
material for the study of temperature conditions and the incipient development of the
plankton, but the cruise was unfortunately interrupted, as the commander of the
vessel, Mr. Wakeham, was taken very seriously ill, and had to be brought home.

By the middle of May a sudden change took place in the state of the ice. About
the 20th of that month, the Halifax newspapers stated that the ice had been driven by
northwesterly winds towards the Gut of Canso, and about the 25th, the remainder had
either melted, or been carried out of the gvilf. This state of things was generally
regarded as unusual. Both on land and at sea the general opinion was that the ice
should have heen gone six weeks before. As it was, the delay produced its effect upon
the fishing industry ; the herring fishery round the Magdalen island was a failure, and
the herrings taken came unusually late.

After this preliminary survey of the ground, a definite plan was drawn up for the
work, the Canadian Government placing at our disposal the two cruisers Princess and
Acadia, and the fishing steamer No. 33. The Princess, now under the command of
Capt. J. Chalifour, vice Cormnander Wakeham deceased, was to make two cruises in
the gulf of St. Lawrence : the first in June and the second in August.

The Acadia, under Commander F. Andersen, was to make two cruises, between
the south coast of Newfoundland and Halifax, as nearly as possible coincident with
the cruises of the Princess.

The No. 33, under Capt. Thor Iversen, was to carry out fishing experiments in
the gulf of St. Lawrence, and, in addition, to make occasional hydrographical observa-
tions and collections of plankton. At Souris, P.E.I., a temporary laboratory was estab-
lished, where the material from the cruises could be collected and subjected to a
preliminary examination.

The scientific work on board the two cruisers was carried out by myself, with
the assistance of Prof. Ai-thur Willey, Dr. A. G.. Huntsman (University of Toronto),
Curator of the Atlantic Biological Station of Canada, and, for a shorter period, of
Mr. Nightingale of the United States Bureau of Fisheries.

All members of the expedition worked at the laboratory at Souris between cruises,
and Mr. Paul Bjerkan, with Prof. James W. Mavor, now of Union College, Schenectady,
U.S.A., were occupied there throughout.

The first voyage was made by the C.G.S. Acadia, which cruised from the
29th of May to the 4th of Oune from Halifax over the continental edge towards New-
foundland and the entrance to the gulf of St. Lawrence. In the course of this cruise,
tlnrty-six stations were taken {Acadia stations 1-36).

From the 9th to the 15th of June, the Princess made her first cruise in the gulf,
in the course of which twenty-three stations were taken (Princess stations 3-26).

C'-LV.l/J/.l \ rislli:h'/KS EXl'EDITiOX, I'Jl'i-ir, xix

These tun cruises are charted in fig. 4, wliieh tlius represents the first investiga-
tions made in spring or early sunnner.

From the 21st to the 29th of July, the C.G.S. Acadm made her second cruise in
the waters off the coasts of Nova Scotia and Newfoundland. This gave the Acadia
stations 37-91

Immediately after this, the ('.(i.S. Princess made her second cruise, from the
yrd to the 12tli of August, in the gulf of St. Lawrence, with stations 27-50.

Fig. 4.

These two cruises are charted on tig. 5, which thus shows the investigations
of the late surnraer period.

Throughout the whole of this time, the No. 33 was working in the gulf of St. Law-
rence, and collected there a considerable quantity of material.

Subsequently^ also, it was found highly desirable to procure hydrographical
observations from a later season of the year, and it was therefore a source of great
satisfaction to me when Commander F. Andersen, who had taken the keenest interest
in the investigations carried out on board, undertook to make a cruise off the coast

XX DEPARTMENT OF THE NAVAL 8ERYICE

of Nova Scotia later in the year. This was effected in Novemher, the voyage occupy-
inig the time from the 14th to the 22nd of that month. The route followed will be
found charted on fig. 8, p. 374.

These cruises yielded altogether a very lai-ge amount of material, the treatment
of which naturally called for the services of several experts. Needless to say, I had
looked forward to taking part in this work myself, but as it turned out I was pre-

Fiff. 5.

vented from so doing, as on my return to Norway in September, lOl.'J, 1 found myself
compelled to devote my whole time and energy thenceforward to the handling of
important questions in connection with the Norwegian fishing industry, which in the
autumn of that year was in a critical state. Under these circumstances, it was a
great satisfaction to me to know that the material was to be dealt with by pre-emin-
ently competent hands. A complete and exhaustive treatment of the whole of the

CANAiuw Fisiii:nfi:s i:\i'i:nn loy, iui',-i.j xxi

material collected would, of course, be out of the question here, but the papers
included in the present volume will yet suffice to give a survey of the most important
results attained. The reader will here find: —

The hydrographiical material dealt with in two separate papers. Mr. Paul Bjerkan
gives (pp. 349-403) a survey of the distribution of salinity, temperature, and density
in the waters covered by all cruises.

Then, on the basis of the data furnished by ^fr. Bjerkan, ]\rr. J. W. Sandstrom has
(pp. 221-341) subjected the entire question of dynanrc (onditions in itip Canadian
waters to a thorough and most valuable investigation.

The great mass of the plankton material could not, of course, be dealt with exhaus-
tively here; to do so would, at any rate, have required a far longer time than that which
has actually elapsed since its collection. It is the more fortunate, then, that Dr. A. G.
Huntsman, in addition to his record of the hauls made (pp. 405-420) has found time
and occasion to give an interesting example showing the occurrence of a single animal
group in the material (Chsstognaths : pp. 421-485). We have, furthermore, an extremely
valuable contribution by Professor Willey, on the important group of the copepods, the
distribution of which is here dealt with by a method now applied for the first time
(pp. 173-220). On most of the cruises, in addition to the net hauls, water samples
were preserved for subsequent study of the pelagic plants, which are dealt with by H.
II. Gran, according to methods developed by himself. Professor Gran's paper will be
found on pp. 489-495.

Under the heading of plankton, also, we must of course reckon the pelagic fish
eggs, which form the most important grouji of all from a fishery x)oint of view. These
are dealt with by Mr. Alf. Dannevig, and the distribution of these forms in the Cana-
dian waters is here described for the first time.

Of the considerable material dealing with the biology of the fishes concerned, we
have up to the present only been able to complete the report concerning the growth and
age of the herring, composition of the stocks, migrations, etc. Mr. Einar Lea has in
his paper (pp. 75-164) not only treated the whole of the available material — and tliis
far more exhaustively than could be done in my preliminary report above mentioned —
but has also furnished a general introduction to the methodical aspects of heiTing
investigations on the whole, and to the study of the problems which can now be dealt
with thereby. Mr. Lea's paper, like that of Mr. Sandstrom, will be found to afford a
guide to thestudy of these questions applicable in itself to far more than the restricted
and particular sphere embraced by the actual investigations concerned in each case.

A contribution to the study of the younger year-classes of herring from the
southern part of the Canadian waters, the Pay of Fundy. is given by Dr. A. G.
Ilnntsman (pp. 165-171).

I venture to hope that the work thus produced may, albeit by no means exhaustive
or altogether comprehensive in itself, yet serve in principle and by example to pave
the way for further research, and awaken new interest in the study of these Canadian
waters, which offer such remarkable and valuable features for investigation. I trust.
also, that the scheme of work laid down, and the nature of the material thereby pro-
cured, for which I am of course responsible, may prove to be justified by the results
attained. It might seem tempting in various ways to endeavour to collect the various
separate investigations here given into a single whole. I have refrained, however, from
any attempt at so doing, partly because it is perhaps too early as yet to think of this,
and partly also because I myself would infinitely rather that the reader should have the
advantage of consulting the excellent reports themselves, not a mere extract of the same.

On the other hand, it will doubtless be advisable to give here, in the form of intro-
ductory remarks, some general idea of the principles followed thi'oughout the cruise,
with some reference, also, to the experience gained in the course of the actual work on
board, which will best explaim the particular manner in which the researches were
carried out.

DEPAKl'MENT OF THE NAVAL SERVICE

THE HYDROGRAPHICAL INVESTIGATIONS.

A description of the submarine physiography of the Canadian waters lias been
given by J. W. Spencer (seo chapter IX of Sub-oceanic Physiography of the North
Atlantic Ocean, by E. Hull, London, 1912, and Bull Geol. See. Amer. vol. XIV, 1903,
1>. 207), of vphich a brief summary is given by Dr. Huntsman (p. 479 and following
pages) which latter I will draw upon here, as it affords an excellent introductory view
of the waters investigated.

As will be seen from Mr. Sandstrom's chart (pi. 1), the continuation of the St.
Lawrence river forms " the submerged Laurentian valley cutting across the middle of
the St. Lawrence gulf, and passing through Cabot strait and between St. Pierre bank
and Banquereau. We have referred to this as the Laurentian channel. Another
channel, the Cansan, cuts through between Sable Island bank and Banquereau. Far-
ther to the south is the Fundian channel, i)assing out from the Bay of Fundy and
through the gulf of Maine. These three channels delimit two portions of the continen-
tal shelf off Nova Scotia. That to the north, between the Laurentian and Cansan
channels, includes the Banquereau, Misaine, and Cansan banks, and may be called the
Breton portion of the shelf, or the Breton bank, since it lies off Cape Breton island.
The southern part lies between the Cansan and Fundian channels, and includes La
Have and Sable Island banks. It may be called the Scotian bank, since it lies against
the main portion of the province of Nova Scotia.

" In the St. Lawrence gulf we have, to the north of Auticosti island, the Anti-
costian channel, and running north towards the straits of Belle Isle, the Esquimau
channel. To the south of the Laurentian channel, in the gulf, is an extensive sub-
marine plateau, with, for the most part, less than 30 fathoms of water covering it.
Cropping up from it are the Magdalen islands and Prince Edward island ....
We have referred to it as the Lower Gulf region. It might be called the Magdalen
bay."

Prior to the cruises described in this volume there existed a series of excellent
hydrographical investigations carried out by Dr. W. Bell Dawson, p^^blished in the
reports of the Tidal and Current Survey of Canada for the years 1894 to 1913. Dr.
Dawson made determinations of specific gravity (density), and also some temperature
measurements. He has described the current of fresher (lighter) water layers Avhich
spreads from the St. Lawrence river along the Gaspe coast and out over the Magdalen
bay, as above defined. This current, called the Gaspe current, runs farther out along
the south side of Cabot strait as a Cape Breton current. ^

From the sea outside another current makes its way into tlie gulf along the north
side of Cabot strait off cape Ray, spreading out along the north side of the gulf. Out-
side the gulf, the water was far less known. As pointed out by Dr. Huntsman, there
was known to exist " a slight westward tendency on the southern coast of Newfound-
land, and a southwestward drift along the outer coast of Nova Scotia."

In the gulf of Maine, on the other hand, the currents have been more thoroughly
investigated; there are, as we know, the excellent investigations of many years carried
out by Dr. H. B. Bigelow in these waters. Before commencing the present researches,
as also during the progress of the same, and afterwards, I had frequent opportunities
of diiscussing with Dr. Bigelow himself, the questions involved, and I may say that the
present work has greatly profited thereby.

The fir.st cruise made in connection with the present investigations was, as men-
tioned above, that of the C.G.S. Acadia, from the 29th of May to the 4th of June,
1915. The vessel started from Halifax, the primary object being to take a section of
the westerly drift off the east coast of Nova Scotia, and determine the volume of the
mixed layers between the coast and the warmer water of the open ocean. The reader
will best be able to follow the course of this cruise by referring to Sandstroms figs, on

^ For a closer study of the hydrographical features in detail, it will be better to consult Mr.
Bjerkan's sections and tables (pp. 379-403).

CANADIAN FISHERIES EXPEDITION, /9/.}-/.J xxiii

pis. 2, 4, and 6, where the sections of the spring hydrographical cruises are drawn in
situ. On examination of these, it will be seen that the first observations immediately
revealed the existence of a distinctly marked light coastal layer with low tempera-
ture. At the outermost station (section V, 10) the warm and salt ocean water was
encountered. In the course of the cruise two pronounced temperature minima (below
0°C) were met with, one off Halifax and one off the Continental Shelf. In the latter
case, the temperature at 75 m. depth reached the extremely low figiire of -1-7° C.
(Acadia station 12).

It was obvious, of course, that the former of these two minima must be due to a
cold (and fresher) current from the gulf of St. Lawrence, while the other would pre-
sumably be the last outpost of the cold Labrador current, already encountered by the
Michael ^Sars in 1910, and here shown in the section, fig. 1, as mentioned in the fore-
going.

It therefore seemed to me imperative to shape a course up towards the banks south
of Newfoundland, in order to rediscover this water layer, if possible. This we suc-
ceeded in doing (see Sandstrom's pi. 4, sections VI and VII, stations 21-25), find-
ing the sea floor on the banks covered with a cold layer, th6 temperatures going right
down to -1-4° C (station 24), or about the same figure as found ])y the Michael Sar.s
in July. 1910, off St. John's Newfoundland.

This point heing thus disposed of, the sections VIII and IX were then taken, with
a view to obtaining good sections both of the water pouring into and out of Cabot
strait, and of the connection between the outward current from the gulf of St. Law-
rence and the westerly drift off Halifax. Sandstrom's pi. 4 shows at a glance how the
temperatures alone suffice to reveal the connection between the cold water layers from
the gulf round cape Breton and along the coast of Nova Scotia. Furthermore, we
notice the inflow from the Newfoundland area into the gulf, along the north side of
(Jabot strait; and finally, the sections also indicate the connection between the cold
Labrador water and the temi^erature minimum off the Continental Shelf on the outer
side of the Sable Island bank.

From the 9th to the 15th June, or only five days after the conclusion of the
Acadia's cruise, the C.G.S. Princess set out to continue the investigations in the gulf
of St. Lawrence. Here, two large sections were made across the well-known Gaspe
current, as also one section from the north coast of the gulf to Bay of Islands. PI.
4 shows distinctly the fresher surface layers in Magdalen bay, and along the north
coast, while the eastern part of the gulf has a higher salinity at the surface. It is
interesting to follow the high salinity from the open sea all through the Laurentian
channel in towards the mouth of the St. Lawrence river. The temperatures (Sand-
strom's pi. 6) show that the lower figures, under 0° C represent a great intermediate
layer covering the chief banks in Magdalen bay, off the north coast, and off the wes-
tern shores of Newfoundland. These points are of the highest importance to the study
of all biological questions.

These cruises had thus showed that it was possible to obtain good sections of the
principal currents, and the plan here followed was therefore, in essentials, taken as
a basis for the later summer cruises.

A similar rapid survey of these is best obtained by referring to Sandstrom's pis. 3,
5, and 7, and comparing the same with those for the previous cruises (pis. 2, 4, and 6).

It will soon be apparent that the quantity of fresh water has now considerably
increased, especially on the Gasj^e coast, and in the mixed layers, which are essentially
fresher even far out at sea. A comparison of the two plates 6 and 7 very clearly
■shows the same thing. On the other hand, the saltest water which lies
deeper down seems to have riisen considerably, a feature which calls to mind the funda-
mental investigations of Patterson and Ekman in the Skagerak on the European side,
resembling in many respects these Canadian waters. The figures for temperature
naturally exhibit a marked increase on the surface, and a reduction in the thickness of

xxiv DEPARTMEXT OF THE NATAL 8ERTIGE

the cold intermediate layer. Even in the first half of August, however, considerable
areas on the banks in the gulf of St. Lawrence were found to 'be covered with water
colder than 0° C.

As will be seen from these brief remarks, we had now succeeded in procuring a
material which enabled us to measure the thickness of the various water layers by sec-
tions taken transversely to their direction of movement, and this in the shortest pos-
sible time. We had furthermore been able to repeat such investigations at a later
period. This furnished us with the basis for a theoretical treatment of the material,
and an analysis of the influence exerted by the various factors; the earth's rotation,
melting of the ice, specific gravity, temperature, etc. In Mr. Bjerkan's 'paper, the
reader will find all the precise data concerning the values for salinity, temperature,
and density. And Mr. Sandstrom has endeavoured, on a wide scale, to give a thorough
analysis of the causes conducive to the circulation of the water as a whole, and its
dynamics generally. This is, as iar as my experience goes, the most thorough treat-
ment of these questions which has yet appeared.

We are now brought face to face with a great number of most interesting problems
for future hydrographical investigations.

In the first place, it would be well to ascertain what fluctuations may occur from
the conditions found to prevail in 1915. What variation can take place, for instance,
in the amount of fresh water discharged by the St. Lawrence river, in the Gaspe cur-
rent, in the interchange of water between the gulf of St. Lawrence and the area out-
side, in the great cold intermediate water layer, in the Labrador current, and in the
distance of the warm ocean water from the costal banks? All these questions will
naturally be of the highest importance in th& study of biological problems, chiefly, per-
haps, the varying distribution of the cold water layers. A particularly interesting
feature is the remarkable wedge of the very coldest salt water off the Continental
Shelf (Acadia station 12). Can this current penetrate still further southward in the
spring, and can this be the great cause of the death of multitudes of fish, occasionally
observed in the sea off the east coast of America ? Bearing in mind the state of things
in the North-European waters, as revealed more especially by the Swedish investiga-
tions, there is good reason to believe that the fluctuatiions in the Canadian waters will
likewise prove to be of very considerable extent.

Sandstroom's investigations revealed various possibilities for a further comprehen-
sion of hydrodynamics in the sea. I would here more particularly call attention to his
suggestions as to the study of submarine waves, where fresher layers encounter the
mighty wall of the saltest Atlantic water, or in the intermediate layers in the gulf of
St. Lawrence, or in the current movements of the deeper layers, as, for instance, those
in the Laurentiian channel.

PLANKTON' INVESTIGATIONS.

The scientific, and particularly the quantitative study of plankton in the sea has
long been a subject of discussion giving rise to widely divergent views. Hensen and
his followers have emphatically maintaied that only quantitative investigations could
lead to results of any scientific value, and that the methods developed by Hensen him-
self afforded satisfactory means of ascertaining the quantitative occurrence not only
of separate species, but also of the total plankton in a column of water corresponding
to the range filtered by the Hensen net from bottom to surface. Other investigators,
again, were indisposed to bind themselves to such methods of research, or to accept the
given formula'tion of the problem. It has been pointed out, for instance, that the
method in .question was not altogether satisfactory in itself, as the nets did not by any
means take the entire plankton content— i.e., all the forms represented— in the column
of water through which they were drawn. And it has also been demonstrated beyond
(luestion that many forms passed .through the nets, while the larger ones managed to
avoid them. Moreover, it has been maintained that both the term " plankton," and the

CA\AniA\ risiiiiiniis h\i'i:i)i i lox, i'ji'i-Ij xxv

idea of a vertical haul were calculated to obscure the true nature of the proposition.
The conception of vertical hauls through the whole column of sea water, under one
S(iuare meter of surface, is undoubtedly derived from the simpler process of soil valu-
ation on land, but the knowledge gained as to the circulation of sea water, wirth its
surface, intermediate, and bottom currents, and the highb' fluctuating velocity of the
same, militates ver\- strongly against the acceptance of the vertical haul as a univer-
sal means of estimntimg quantitatively the production of floating organisms in a given
area.

For the present, then, as long as we still lack a clear and indisputal)lc analysis of
the great group of problems actually invoh'ed by the questions raised, and have yet to
find satisfactory methods of reseax'ch for approaching the same, it would surely seem
far better to content ourseh^es with studying the geographical ((jualitative and quan-
titative) occurrence of cer'tain definite forms, applying in eacli particular case the
methods best suited thereto. This view has previously been advanced by the present
writer ^ and by Prof. H. H. Gran. "We pointed out .that a quantitative estimate of the
plankton in a water layer should at any rate be based upon a combination of samples
drawn from horizontal layers which must themselves be ylefined as closely as possible,
and their i^hysical and chemical conditions investigated at the same time.
In the case of the animal forms, several methods have 'been tried in connection with
such horizontal liauls. .From time to time, there have been consti-ucted more or less
adequate plankton nets, which could be opened and closed at certain depths, and indi-
cate what water they had fished, but the methods hitherto employed can hardly be said
to fulfil the claims of absolute accuracy in these respects. My own personal view has
always been, that it is unjustifiable to attempt a task which the methods of work
available do not suffice to accomplish, and that it is therefore better to recognize
the restrictions inaposed by the fact, and aim advisedly at results which shall be
approximately valid for certain selected forms. On the Michael Sars expedition in
1910, 1 found it most practical, and therefore most effective, in the case of the fishes,
to tow nets through the water at different depths, and then endeavour to ascertain the
catch made at a given depth by statistic treatment of the yield. (Vide "Depths of the
Ocean," pp. 615-617.)

Only in the case of the vegetable plankton have we an altogether satisfactory
method of work, viz., that of H. H. Gran. Professor Gr'an was able to show, that by
preserving water samples (with Flemming's liquid) and subsequently centrifuging
them, we can obtain material sufficient for quantitative determination of the entire
vegetable plankton in the sample. Gran has by this means, as we know, succeeded in
obtaining the first real view of the true plant production of the sea,- his experiments
being carried out in European waters (the Skagerak) which in so many respects
resemble the Canadian. At most of our stations, plankton samples were collected
according to Professor Gran's method, and the material is dealt with by Gran in his
paper here given (p. 489.)

, The paper in question throws light upon some extremely important sides of the
natural history of plankton. As will be apparent from the work itself, the Canadian
plankton reveals marked resemblances to that of the European waters. It differs, how-
ever, in the common occurrence of typical arctic forms, corresponding of course, to the
very low temperatures prevailing in the gulf of St. Lawrence during winter in all the
upper water layers, and in summer throughout the great intermediate layers. A point
of great importance for all conditions of growth in the Canadian waters is the fact
demonstrated by Gran, that the development takes place much later in the year there
than in the waters of northern Europe. This naturally agrees with the fact that the
gulf of St. Lawrence was full of ice until nearly the middle of May — and it is interest-
ing to note, in this respect, that while the growth of the herring in the far more north-

1 See, for instance, "Depths of the Ocean," pp. 771-785.

2 H. H. Gran. The Plankton Production of the North European Waters in the spring, 1912.
Bulletin Planktonique pour I'annfee 1912 public par le bureau du Conseil Permanent Inter-
national pour I'exploration de la mer. Copenhague, 1915.

6551 — c

xxvi DEPARTMENT OF TEE NAYAL SERVICE

erly Norwegian waters commences in April, that of the herring in the gulf of St. Law-
rence does not begin until June; in the case of a single sample, indeed, not until July.
In Lea's paper, also, pp. 158-159, it is pointed out that a similar late growth is kno^vn
from the Baltic, near the coast of Finland — a further example of the similarity
between conditions in the gulf of St. Lawrence and those of European bays.

In considering the collection of animal plankton, there are certain points which
should be borne in mind.

For carrying out the cruises above mentioned, I could only reckon on having the
two vessels, PHncess and Acadia at my disposal for about a month, or, to be precise,
for thirty-three days in all. In order to cover, within this short period, a sufficiently
large expanse of water to give a real survey, it was necessary to arrange beforehand
for the quickest possible method of work, making only a short stay, for instance, at
each station. This was the more imperative, since the frequently boisterous or foggy
weather compelled us to allow a margin for delay, though, as it turned out, we met
with no serious difficulties in this respect.

Moreover, the two vessels, not being built specially for the purpose, could naturally
not offer ideal facilities for plankton work. The Michael Sars, as described in the
" Depths of the Ocean," lies low in the water, and can be easily manceuvred against
the wind, so as to keep the line of a net vertical in the water; our vessels here, on the
other hand, with their higher freeboard, could only take their stations by lying trans-
versely to the direction of the wind, which often rendered it difficult to get vertical
hauls.

Owing to these circumstances, the hauls made are by no means all that could be
desired, especially from the point of view of the Hensen school. The critical remarks
proffered by Dr. Huntsman in his notes on the list of hauls in the present volume (pp.
407-4:20) are therefore entirely justified from this point of view, and the quantitative
figures for volume are really only of value to a limited degree — more limited, in the
present case, than such figures otherwise would be in themselves. They are given here,
nevertheless — albeit, as mentioned, with all reserve — because, to those who are them-
selves acquainted with the fluctuating quantity and quality of plankton generally, they
will be of interest as affording some idea of the order of maguitvide in the present
samples. And in connection with subsequent collections, taken for instance at other
seasons and in other years, as also for further treatment of the material, the figures in
question will doubtless be of some importance after all. They also serve to indicate
approximately the extent of the collections from which the different groups of material
treated separately in the special sections were drawn.

In the case of larger, rarer, and more conspicuous forms, such as fish eggs, fish
larvse, Chaetognaths, treated in the present volume, we endeavoured as far as possible to
count, treat, and examine all individuals in the samples. In such treatment it is
always more or less interesting to note the size of the samples, and the depth in which
the numbers of individuals found were obtained. For forms of less frequent occur-
rence, also, the quantitative approximation is perhaps also satisfactory, but gives, of
course no information as to the interesting question of the precise depth at which the
individuals concerned actually were taken.

The study of fish eggs and fish larvae gave rise, during the progress of the cruises
themselves, to a series of most interesting and important questions.

On the first cruise of the Princess in the gulf of St. Lawrence, (May 29 to June 4),
we encountered a characteristic occurrence of eggs of Gadoid species, in particular
of the cod proper, and of the flatfish Drepanopsetta (Hippoglossoides) all over the
banks, i.e., in what we have called Magdalen bay, at Anticosti, along the north shore,
and off the west coast of Newfoundland. Besides these greatly predominating species,
we found on the banks only eggs and a few larvae of arctic species (Anarrhichas lati-
fron.<?, Mallotus villesus, Icelus, and Agonus decagonus. At the southernmost
stations, we found at Cabot strait the first mackerel eggs, and above the deep Lauren-
tian channel — and nowhere else — considerable quantities of Sebastes.

c.4-\.i/^/.i \ ijsnj:h'ii:s i:.\ri:i>ii los, ini'i-io xxvii

These finds led me to expect that we should, on the later summer cruise, encounter
quantities of larvae of the species in question. To my surprise, however, this cruise
likewise yielded hardly anything beyond cod eggs and only a very few larviv indeed.
A characteristic feature, also, was the fact that we found arctic forms (MuUotus) in
the northern part of the gulf; eggs and larvie of southern forms (Ctenolahrus,
mackerel) in the southern part, with Sehastes, as before, sharply limited to the waters
immediately above the Laurentian channel.

Outside the gulf we found, on both cruises of the Acadia, a scanty occurrence of
cod eggs. Beyond this, I will here only mention a considerable occurrence of Mer-
luccius in this water, and Scopelids out on the edge; these last aflFording sufficient tes-
timony that the investigations had covered the whole breadth of the coastal zone.

The most remarkable point in connection with these hauls was the extreme
paucity of the older egg stages, and of all larval forms. It was therefore necessary
first of all to study the ratio between eggs and larvae in our present material, and to
compare the same with what was known from other waters in this respect. Mr.
Dannevig's paper gives a critical treatment of this question, and it will be noted, that,
as he expresses it (p. 44) " the gulf of St. Lawrence i's considerably behind the other
localities with respect to the occurrence of later stages, both in the case of the earlier
investigation and those made subsequently (Princess I, Princess II, and No. 33).
The ova have evidently a far poorer chance of being developed and hatched than in
other places." The few experimental hauls made by the No. S3, and the information
gleaned from fishermen in conversation also confirmed the view that in the gulf of St.
Lawrence, very few young fish of any species are known to occur at all. The question
is, of course, of vital importance for the study of the stock of fish in the water. How,
then, are we to explain the facts as they appear?

The results of the European fishery investigations lead apparently to the conclu-
sion that all boreal forms of food fish are restricted to water with positive temperature,
rarely occurring, indeed, in water under 2° C. In the Arctic ocean, the Greenland
seaj and the deepest part of the Norwegian sea, where the temperature varies from 0°
to - 1-5° C, only arctic forms of fish are found: here, however, in the gulf of St. Law-
rence and on the Newfoundland banks, we encountered masses of cod eggs floating in
and immediately above thick water layers colder than 0°, even below — 1° C. Our
fishing experiments, with hand lines, for instance, had showed beyond any possibility
of doubt that the cod themselves really were to be found in this cold layer, which
immediately covers the sea floor, while this water contained the very youngest stages of
spawned eggs, but hardly any of the older stages at all. Were we then to suppose that
these millions of fish were here spawning under conditions which doomed their milliards
of eggs to destruction? Naturally, the question could not be answered by means of such
investigations as those upon which we were then engaged : to do so would have required
a vessel, say like the Michael Bars, able to devote itself entirely throughout a whole
season to all kinds of work in the water concerned. Hatching and cultivation experi-
meaits would be necessary, as also intensive experimental fishing for young stages, etc.
Failing all this, and wishing to throw some light upon the problem, if possible, I
applied to Prof. August Krogh and Dr. A. C. Johansen, of Copenhagen, with a request
that they would institute experiments in order to ascertain whether cod eggs from
Danish waters could be satisfactorily spawned and hatched out in water of such low tem-
perature. The results of the experiments made by these two gentlemen are quoted in
Mr. Dannevig's paper, and it will be seen that they give no groimds for supposing that
the temperature alone should ofi"er any hindrance to the development of the eggs.

Again, it might be supposed that the cod eggs were carried out from the gulf, en
masse, as is known to be the case in European waters, where they are transported by
the movement of the water to a great distance from the spawning grounds. The hydro-
graphical papers in the present volume, as also Dr. Huntsman's plankton report, con-
tain numerous facts pointing very markedly in this direction. It is impossible, how-

xxviii DEPARTMEyr OF THE J AVAL HEltVICE

ever, from the present investigations, to come to any final decision here, but the work
done has certainly raised questions of the greatest importance for future Canadian
fisherj^ investigations.

These waters are, of course, extremely interesting from the most general biological
jDoint of view. We find here, in one and the same area, and with only a water layer of
a score of fathoms between, cod spawning on the floor of the banks, in water of absolu-
tely arctic temperature, while at the surface, directly above, spawn southern forms such
as the mackerel, wliich have possibly migrated from the water layers just off the Con-
tinental Shelf where the last remains of the Sargasso weed are still to be found. And
a single haul at the surface will be seen to contain pelagic eggs of both these species of
fish.

Of tlie remaining plankton, only two groups have been dealt with up to the present,
viz., the Cha3tognaths, by Dr. A. G. Huntsman, and the Copepods, by Prof. Arthur
Willey. These two studies afi^ord, however, in themselves, excellent examples of what
plankton investigations can yield in the way of results, and of the methods which
must doubtless best be employed until better implements are available for ciuantitative
study. Of the comparatively large and not so very numerous Chsetognaths, Huntsman
has given estimates for the total number of individuals in the samples. In the case of
the small and numerous Copepods, this was impossible, and the method here adopted
v.'as therefore to determine, from a small selected sample, the percentage of com-
position represented by the separate species. In the case of some few more important
forms (especially Calanus finmarchimis) , Willey has given also the percentages of the
various principal stages of development in the sample.

Thus the two papers illustrate the occurrence of various biological types, viz.,
those pertaining to the oceanic water of the Atlantic, those of the colder boreal water
layers, and finally those of the fresher coastal layers And a study of these plankton
reports, compared with those of the hydrographical work, will give an idea'of the
marked agreement between the distribution of the animal forms concerned and that
of the various water layers.

The results of the fishery investigations proper, i.e., investigations as to the actual
life and occurrence of the fish themselves, cannot yet be given in their entirety, as the
treatment of part of the material is not yet concluded. Up to the present, we have
only Mr. Einar Lea's description of all the herring samples; this section is, however,
after all, the most important part of the work from a biological point of view, and the
primary object of the expedition is thereby attained.

Mr. Lea has in his paper given a thorough explanation of the methods employed
in age and growth determinations, based on examination of the scales, and the work
will therefore, it is hoped, prove useful in future Canadian fishery investigations.

For the rest, I would merely point out that Mr. Lea's careful researches have satis-
factorily demonstrated what the results of my preliminary investigations had already
indicated, viz., that there are in the Canadian waters distinct types of herring,
excellently characterized by their maimer of growth and the difference in age com-
position. His thoroug-h treatment of the material likewise shows that the various areas
exhibit the same peculiarity as has been found so markedly apparent in several Euro-
pean (and especially Norwegian) waters, to wit, the pronounced fluctuation in the
increment of young individuals from year to year, whereby the age composition of the
stock reveals an enormous predominance of some few extremely rich year-classes, while
others hardly contribute in any appreciable degTee to the numerical value of the whole.
The conclusions here arrived at, regarded together with the like results which are
becoming, apparently, more and more common in various spheres of biological research,
force us to admit that hitherto prevalent views not only on leading fishery questions,
but also of general biological problems as to the maintenance of species, and all that is
comprised in the old Malthusian ideas, will need to be essentially revised.

CANADIAN FISHERIES' EXPEDITION, 1914-15

BIOLOGY OF ATLANTIC WATERS OF CANADA

CANADIAN FISH-EGGS AND LARV.E

BY

ALF DANNEVIG

Of the Flodevig Sea-Fish Hatchery, Arendal, Norway

6551—1

CANADIAN FISH-EGGS AND LARVJE

I. INTRODUCTION.

When Dr. Hjort invited me to undertake the work of dealing with his valuable
material of fish eggs and young from the Canadian waters. I was fully aware of the
many difficulties which the task would involve. The determination of ova and young^
from a quarter of the globe where many species of fish exist, whose young stages
have not been described, is naturally no easy matter, more especially when only
prieserved material is available, and it will, as a rule, be obvious from the outset
that complete success must be out of the question. The object of the Canadian inves-
tigations was, however, principally to ascertain the biological conditions under which
the most important species of fish lived and multiplied, so that possible errors in the
deterniiiiatiou of some few of the less frecjuently occurring species would be of
minor importance. In view of this fact, and with the promise of every assistance
from Dr. Hjort and Magister Koefoed, I undertook the task, all difficulties not-
withstanding.

As to how far the determination of species has been successful, it is naturally
impossible to say at present; future investigations of the same waters will, in all
essentials, make this clear. I have thought it proper, however, here to make a few
introductory observations as to eertain i)articular difficulties which may be imagined
as having given rise to erroneous determinations.

The transatlantic literature on the subject being somewhat scanty, I have had
recour.ve more especially to descriptions of European species, and it might therefore
seem not unlikely that related western forms could have become confused with or
mistaken for these; but regarding the young of species found on both sides of the
Atlantic, however, and where good descriptions of the .young stages are available, this
source of error does not apply.

A considerable number of forms have been drawn with great accuracy by Mr.
Rasmussen. these including both species previously described and drawn, and also
other lesser-known species. I have not thought it advisable to give any description of
such forms on the basis of the preserved material, more especially since this would
lie outside the scope of the present work.

As regards the determination of the ova, this is naturally more or less uncertain,
since the eggs of many species are indistinguishable one from another until tlie
embryos are well developed. In some few cases, on the other hand, the ova may be
accurately determined from the time of spawning. It will, moreover, be justifiable to
presuppose relation between ova of like appearance in one and the same sample, where
some are newly spawned and others with the embryo sufficiently developed to permit
of certain determination.

The present material of ova and young was collected by means of a silk net (In?.
diameter) and preserved in 4" per cent formol. The duration of the surface haul.>^
varied somewhat, as a rule between ten and fifteen minutes; the depths, as stated for-
tlie vertical hauls, are also in some cases only approximately correct, owing to diffi-
culties in the working of the net.

Similarly, the records as to quantity should not be taken too strictly, these being
for like reasons only approximate figures. The proportional quantitative values, how-
ever, as between eggs at different stages of development, may be taken as fairly
reliable.

Or).'")l— IJ

4 DEPARTMENT OF THE NAVAL SERVICE

The material was collected in the course of five different cruises, of which two
made on board of C.G.S. Acadia covered a zigzag coiirse between the southern point of
Nova Scotia and Newfoundland, the first extending' from May 29 to June 4, the second
from July 21 to July 29.

In the gulf of St. Lawrence, two cruises were made with the C.G.S. Princess from
June 9 to June 15 and August 3 to August 12; here also, the SS. N° 33, in addition
to its ordinary fishery operations, collected a number of samples during the period from
June 1 to August 18.

-/fcwcZ/a //rs^ cruise ^H-i>6 ^/<-36
second ■■ ^/7-2r7 St 3^-9/

Prjncess f/rst cru/se%-'^/6 5t3-26

second ■■ i/s-^^/s 5 1 27- so

Fig. 1.

Before proceding to the treatment of the material, I should mention that a
quantity of ova and young had already been determined by Dr. Hjort in Canada,
■while Mr. Koefoed also has kindly determined a number of Atlantic fish, and has
further assisted me in doubtful cases with other forms.

CAAADJAN FJStn FRIES EXPEDITION, WVrl5

II. THE MATERIAL COLLECTED.

In the plankton samples, ova and young of thirty-six species of fish have been
determined, representing twentj' families in all.

The si)ecies will be dealt with in the following pages in systematic order, having
regard to their occurrence within the area investigated, their geographical distribution
generally, and certain features in their biology and reproduction.

1. FAM. LABRID^.

Ctenoldbrus adspersus (Walbaum).

(Plate II, Figs. 1 and 2; Table lla.)

„ , ,., r- /• I -100 pr Station all hauls
Pnncess 1 bt J-26 ^99^ i^rr^ore than ,oo

Acadia 1 St 2-36

Lan^ael

C/e/70/adrus

30 pr Station
? than 100 ■
- 10 pr StaTion

)more than lo

Fig 2

■*■ Acadia n St 37-91

Ctenolabrus

^^ ( # I - 100 pr station all hauls

+ Princess n St 27-50 '^J'^^ j Ai^gre t han 100 '

I 0'

- 10 pr station

O^

Fig. 3.

CA\An[A\ FISiHERIK^ EXPEDITIOX , 101 ',-15

'55

Eh^t

Ctenolabrus

I # 1 - loo pr Station all hauls
l^fcmore than loo

Fig. 4.

The young stages of this s])ecies have been shown by Agassiz (On the Young
Stages of some Osseous Fishes, Part III, Ctenolahrus coeruleus DeKay. pi. XIII, and
XIV). I have not, however, been able to identify the species with certainty from these
drawings, buti as the larvae agree very well with the young stages of the related
European form Lahrus rupestris, I have had no hesitation in ascribing the Canadian
material to Ctenolahrus adspersus* According to Professor Ehrenbaum's '' Nordisches
Plankton," the larvae of Tautofja onitis are very similar to those of this species, but
as Tautoga is a southern form, there can hardly be any question of confusion with the
present species.

As will be seen from the illustrations, the Canadian form differs from the European
in having a very distinct massing of pigment at the posterior basis of the dorsal fin.

The diameter of the ova is the same in both species, viz., 0-8 — 0-9 mm.

The distribution of Ctenolahrus adspersus ranges from Labrador to Sandy Hook
(Jordan and Evermann) ; it is a distinctly coastal form, especially affecting rocky
bottom. Ctenolahrus adspersus is summer-spawning; on the first cruise of the Princess
(June I) to June 1.5) but not more than twenty-one eggs of this species were taken.
On the second cruise (August 3 to August 12), on the other hand, numerous larvae
were found, and only very few eggs, most of those found being in a very advanced
stage of development.

♦ Tmitogolabni.y ndsperus Walbaum.

8 DEPARTMEl\'r OF THE NAVAL SERVICE

As will be seen from the chart, ova and young were only found at the stations
nearest the coast, especially in the gulf of St. Lawrence along Prince Edward Island
and Cape Breton.

The material of this species from the Acadia amounts to some few eggs and young
taken at Sable island and farther in near the Gut of Canso.

The size of the young varied between 3 and 9 mm.; (see also tables),

2. FAN. CARANGIDJE.

Capros aper (Lacepede).

,0f this species, only one specimen (total length 6 mm.) was taken, this being
from the Acadia, Station 44, where it was brought up in a vertical haul from 150-0
m. ' Although Jordan and Evermann do not mention this species as occurring in
American waters, I have been obliged to ascribe this specimen to the si)ecies in
question. C. aper is also found, by the way, in the eastern Atlantic right up to the
coasts of England, and it also penetrates into the Mediterranean.

3. FAM. SC0MBRID2E.

Scomber scornbrus (Linnaeus).

(Plate I, Fig. 3: Table lib.)

Of this species, quite a considerable quantity both of ova and young were
collected, especially by No. S3, and on the second cruise of the Princess.

Jordan and Evermann give the distribution of the mackerel along the American
coast as from Labrador to cape Hatteras, and it is therefore remarkable that its ova
and young should here have been found almost exclusively in the southern portion of
the gulf of St. Lawrence. The cruises of the Acadia furnished but three eggs from
stations outside Nova Scotia, and neither the Princess nor No. 33 found ova or young
of mackerel in the northern part of the gulf.

Princess I St 3-26
Acadia I St 2-36

^.^'\^

Scomber scombrus

# 1 -100 pr. Station all hauls
more than loo

Fig. 5.

Fig. 6.

C.4,V.4r>//l.V FISHERIES EX fEfHTIOS , 1011,-1'')

11

Fig. 7.

It would seem that the mackerel here when spawning keep to the shallower waters
in towards Prince Edward Island. It is also interesting to note that while the first
stations on the first cruise of the Princess inside and west of Prince Edward Island
furnished purely arctic species, the hauls made a week laten. on the eastern side of the
island, brought up the first mackerel eggs.

The mackerel's spawning season appears to extend over a considerable period,
from middle of June (first cruise of the Pnn-cess) to some way on in August.

The size of the mackerel ova will be seen from the table; it will be noted that the
diameter varies greatly, a phenomenon which is also well-known in European waters.

Princess. Princess. Priticess. " 33."

MM. Station 26. Station 30. Station 31. Station 29.

1-00 • 23

1-05 9 47

1-10 Ifi 34 - 1

1-1.5 1 1 1 4

1-20 8 12

1-25 11 8

1-30 5 ....

The diameter of the oil globules was about 0-03 mm.

12

BEPARTMEJ^T OF THE NAVAL SERYICE

Mackerel larvae were taken only during the second cruise of the Princess, for the
most part from Prince Edward Island, and out towards the edge, where stations 28
to 34 gave 40 larvae between 3 and 8 mm.

In addition, a single larva of 9 mm. was taken at station 45, this being the only
mackerel larva found on the north side of Cabot strait. At Stations 49 and 50 —
between Prince Edward Island and Cai)e Breton — six larvae ranging from 4 to 6 mm.
were found.

4. FAM. PEDICVLATI.

Lophins piscatorius (Linnaeus).

Of this species , only a single specimen, of 11 mm., was found (Acadia, station 47).
L. piscatorius has a very wide distribution on both sides of the Atlantic; according
to Jordan and Evermanm, it ranges from Nova Scotia as far as the Barbadoes; only
^n deep water, however, is it found so far south.

5. FAM. SCOEPAENID^.

Sehastes marinus (Linnaeus).

(Plate I, Figs. 4,5,6,7; Table lie.)

Young of Sehastes were found in all the deeper parts of the areas investigated,
pave at the outermost stations of the Acadia.

Sebastes

O'-io pi" Station all hauls
Prmcess I St 3-26 ^'^^ O'^^'^ *^^«" "^
Acadia I St 2-36

Fig.

Fig. 9.

14

DEPARTMENT OF THE NAVAL SERVICE

j O I -10 pr Station all hauls
Qmore than lo -/ - v

Fig. 10.

It wovild appear to be most numerous above the deep channel south of Anticostl ;
the -A^o. 3S took here 325 young in a single haul. On the slope towards Newfoundland
also, it is numerous. (Princess station 45.)

On the second cruise of the Acadia, also, by the way. it was found in remarkable
quantities on the banks off Nova Scotia ; this fact may be ascribed to the existence
of large depressions in the banl^s, or possibly to the nature of the currents.

SehaMes marinus is found throughout the northern seas between America and
Eurox>e, both on the slopes of the banks and pelagically above the greater ocean depths.
Jordan and Evermann note it as occurring on the American side from Greealand
lalong the coast as far as midway off New Jersey; from Maine and farther south,
however, in deep water only.

These writers affirm that Goode and Bean found spawning Sehastes late in the
summer off the coasts of New England, at a depth of 100 to 180 fathoms and state
that there is no " reason to believe that the young rise to the surface."

According to Dr. Bigelow, adiilt Sehastes are also found in quite shallow water
(about 10 fathoms) in the gulf of Maine — a very remarkable phenomenon, since
Sehasfes is otherwise both south of this (Jordan and Evermann) and farther north,
chiefly encountered at medium depths.

CA^ADIA^' FISHERIES EXPEDITIOy, 1914-15 15

6. FAM. COTTID^.

Cottus scorpius (Linnaeus).
Cottus huhalis (Euphrasen).
Icelus hicornis (Keinhardt).

Three species belonging to this family are represented in the material, numbering
six specimens in all, these being taken, without exception, from the waters near Prince
Edward island and Magdalen island.

C. scorpius^ JS'o. S3, station 1-4, one specimen of 10 mm.; station 17, one of 11 mm.
This species is found on both sides of the Atlantic, and extends some considerable
distance to the northwaid, as far as Spitzlierjien.

According to Jordan and Evermann, it occurs to the southward along the coast as
far as Eastport, Maine. In the European waters, C scorpius deposits its egg capsules
in midwinter, and the young are subsequently encountered as plankton in the spring.
It is noted, however, that in this species, internal fertilization may take place, and
tbat the ova may therefore be in a far advanced stage of development before being
spawned; this occurs especially in the northerly waters.

C. huhalis (Euph.) Two specimens from No. 3S, station 17 — both of 6 mm. —
agree very well with the larvae of C. huhalis, and have been ascribed to this species,
although Jordan and Evermann regard it as doubtful whether C. huhalis occurs on the
western side of the Atlantic.

Icelus hicornis (Reinhardt) was taken at Princess station 7 (14 mm.) and another
of 10 mm. at No. 38 station 15.

This species is an arctic circumpolar form, penetrating, however, southward along
the east coast of America as far as cai)e Cod. Eoth the specimens here taken were
found in comparatively sliallow water.

7. FAM. AGONID^.

Agonus decagonus (Scluieider).

One specimen of 2^ mm. was taken by the Princess at station 7, and strangely
enough, at the surface. A. decagonus is otherwise found for the most part at some con-
siderable depth, down to a couple of hundred fathoms, and in very cold water about
0° C. As to its propagation, little is known.

Aspidophoroides monopterygius (Bloch).

One specimen of 15 mm. taken by the No. 33 at station 21 (C. Gaspe) should
probably be referred to this species. In point of habitus, it is very like Agonus deca-
gonus, but differs from this in having but one dorsal fin. On the other hand, it has
very spinous scales, and differs in this from A. monopterygius; possibly, however, this
may be a larval character.

8 FAM. BLENNIID^.

Chirolophis sp.

(Plate II. Fig. 8; Table lid.)

Two species belonging to the family Blenniidse were foinid, of which the one could
not be determined with certainty.

This was found (no less than sixty-three specimens) throughout the whole of the
gulf St. Lawrence, and out towards the Newfoundland banks. The number of vertebrae,
about 13-1- 42 or a total of about 55, agrees very well with that of Chirolophis galerifa
(L.) Walb. It differs slightly, however, from this in the shape of the head and
intestines, while the pigmentation appears to be of more or less the same character.

16 DEPARTMENT OF THE NAVAL SPRYWE

It was taken in sizes ranging from 8 to 13 mm. and was most numerous during the
time extending to July 15.

Sticliaeus punctatus (Fabricius).

(Plate II, Fig. 9.)

Of this species, two specimens measuring 30 and 40 mm., respectively, were taken.
Stichoeus punctatus is an arctic lish; it is recorded as penetrating as far southward as
Newfoundland. The specimens in question were taken at the surface, station 89
^Acadia) i.e., down towards Nova Scotia.

9. FAM. CRYPTACANTHODIDJE.

Cryptacanthodes maculatus (Storer),

(Plate II, Fig. 10.)

This species was taken in a vertical haul (80 to ni.) at the Princess station 8.
Only one specimen, of 38 mm. Found from Labrador to Long Island sound — but not
common. (Jordan and Evermann.)

10. FAM. ANAREHICHADID^.

Anarrhichds latifrons (Steenstrup).

Two specimens of A. latifrons were taken, one of 21 mm. in a surface haul
(Princess station 3) and one of 25 mm. in a vertical haul 125 to 25 m. (Acadia station
35). These were so far developed as to be distinguishable by the position of the
vomerine teeth. An arctic fish, extending southward along the east coast of America
to Banquereau.

11. FAM. GALLIONYMIDJ^.

Callionj/mus sp.

A fish larva of 6 mm. taken in a vertical haul 150 to m. Acadia station 44,
strongly resembles the European Callionymus species, but lacks the notochord other-
wise so prominent in these species. The formation of the fin rays, however, was so far
advanced that possibly the notochord may have been reduced.

12. FAM. CYCLOPTERID^.

Liparis sp.

Liparis major (Gill).

(Table He.)

Of the Liparidse taken, three specimens from No. 33 station 57 were ascribed
to Liparis major (vide Jordan and Evermann). They had a total length of 25 to 30
mm., but were not in good preservation, and the determination is therefore somewhat
uncertain. L. major is an arctic fish, extending from the White sea to Greenland,
but, has, according to Jordan and Evermann, not been encountered on the coast of
America.

The remaining Liparidse it was found impossible to determine as to species;
in all, sixteen were taken on all cruises.

13. FAM. PLEURONECTID/E.

'Pleuronectidae do not occur in any great number in the material, with the
exception of Drepanopsetta.. There are, however, on the coasts of America, many
species of flounder whose eggs and larval stages have not been described, that on this
point, more especially as regards the ova, errors may have occurred in the determination.

CAXAitf.w risiiEini.s i:\rr:r>rri()\, nii'i-i-') 17

The larva; recorded under the different species are so like the European that, as
V role, there was no difficulty in deciding; several species, moreover, have been figured.
If we consider, houtner, the insiiiniiicaut nuniher ot unknown ri(nni(ler larva' (of which
part could not he determined owing- to accidents in the preservation) it would hardly
seem likely that any fjivat number of unknown eerss would be liable to confusion with
other species. Of unknown Pleuronectida- tlie followin<i- were taken: —

Ijength.

Acadia station 4, 150 — m 1 I'leuronectes? 5 mm.

38.100 — 0ml " "....'.. 6 "

No. " 33 " station 39, vertical 1 " 4 "

No. "33" station 49, vertical 6 " 6-7-6-7-6-7 "

To these shoidd be added one specimen of 4 mm. lenjith from the Princess station
48; probably belonging to the genus Both.us.

P. limanda ( ?) L.

A quantity of ova of this species was found on the first and second cruises f)f the
Acadia (above the banks off Xova' Scotia and Newfoundland).

At Surface. In Vertical Hauls.

Station 29 5 eggs. 5

30 3 " 8

36 1 "

61 ^ "

62 43 "

80 2 "

82

The diameter of the eggs was about 0-9 mm. and the embryos revealed typical
Pleuronectid characters, though none were so far advanced as to ])ermit of their being
iiscribed with certainty to P. limanda; they undoubtedly belong, however, to a closely
related form.

According to Jordan and Everniann, this species also is stated as apparently not
found off the coasts of North America.

P. microcephalus ('Donoyiin)=MirrosfomiisJ,-itt (Walbaum).

At Acadia station 84, a larva of 6 mm., taken in a surface haul, must be referred
to this species.

P. {Gliiploceplialus) ci/nof/Iossits !>.

(Plate IT, Fig. 11.)

Of this species, the following specimens were taken: —

Princess station 34 — Oblique: 1: ca. 6mm.

" " 49 — Surface: 2 : 8-10 mm.

" " 50 — 40-0 : 1 : 7 mm.

Acadm " 83 — Surface : 4 : 6-8 mm.

There were' thus taken eight larv;e in all, but eggs of this speci&s were not found;
these probably did occur in the samples, but must then in some way have been over-
looked. P. cynoglossus spawns in European waters from ]\lay to September; the ova
have a diameter of 1-07 to 1-26 mm., and are otherwise distinguishable by their slightly
striped structure.

P. cf/noglossus belongs to the North Atlantic; on the American side it extends as
far as cape Cod: lives in comparatively deep water, preferably with sandy bottom.

6551—2

18

DEPARTMENT OF THE NAVAL SERVICE

Ancylopsetta (Notosema sp.)

(Plate II, Fig. 12,)

A specimen of 7 mm. from Acadia station 44 (surface). Belongs to the warm seas.
Drepanopsetta (Hippoglossoides) platessoides (Fabricius).

(Plate II, Figs. 13,14,15; Table Ilf.)

Princess 1 5t.3-26
Acadia I St 2-36 ^^^^

flf^S

Drep a nop 5 e it a

• I - 100 pr Station all hauls
^^more than loo "

81 - 10 pr. Station
more than lo

Fig:. 11.

Pig. 12.

0551—2.1

20

DEl'ARTMEST OF THE yiTAL SERVICE

"33'

%i

Drepanopsetta

i-ioopr Station all hauls
kmore than loo -

Fig. 13.

Eggs and larvjE of Ifrepanopsetta have a wide distribution, especially over the
banks. The ova are very often found together with cod eggs, and were especially
numerous on the first cruises.

June 9 to June 15, Prmcess I gave 609 eggs + 5 larvae.

August 3 to August 12, Princess II " 1 " +7

May 29 to June 4, ^carfia I "1,089 " +7 "

July 21 to July 29, Acadia II " 35 " +28 "

To these should be added the hauls made by the ^o. 33, which contain ova of
Drepanopsetta in considerable quantities until July 10. The spawning time must
thus be regarded as over by middle of July, which is somewhat late in comparison with
the European waters. (In the North sea, spawning taken place from January to
May.) The ova of Drepanopsetta are very easily distinguishable by the large peri-
vitelline cavity, and by their considerable size.

Measurements from the No. S3 station 16 give the following numbers of eggs for
■each size (diameter in mm.) : —

2*1 mm.
2-2 "
2-3 "
2-4 "
3-5 '

1
3
10
4
?

CAXADIAX FISHERIES! EXPKDITIOS, 191Jrl5

21

As will be seen from the above, the diameter A'aries up to 4/10 mm.

Larva? of Drepanopsetta were found especially on the cruises of the Acad'ui and
Princess in July-August; they occur but sparsely, save at stations 81 and 8o (New-
foundland banks), where they were more numerous. The length of the larva- varied
between 5 and 20 mm.

Distinction is made between two forms of Drepanopsetta; D. plabessoides, the
more arctic, which also extends down along the coast of America, and D. Umandoides,
the European form.

Drepanopsetta lives preferably on sandy bottom, on the banks, but can also, at
certain times of the year, move down into the deep water of the channels, where it
then lives on soft bottom.

14. FAM. GADIDJE.

Gadus cvglefinus (Linnaeus).
Gadus callarias (Linnaeus).
Onos sp.

Onos cimhrius (Linnaeus).
Merlnccius merluccius (Linnaeus).

(Plate III, Figs. 16 to 22; TableII<7 to Hi.)

The gadoids play the most important part in the material collected; ova and
young of this family were found in all the areas examined and on all the cruises
made. On consulting the chart, however, it will at once be noticed that the different
species have each their own area of distribution, and only occasionally overlap.

The ova of cod and haddock cannot be distinguished one from another with ful]
certainty in the earlier stages. True, those of the haddock are, as a rule, some tenth.s
of a millimetre larger than those of the cod, but both species vary so greatly as to
overlap in this respect. I have, however, always counted such eggs as could be deter-
mined with certainty (i.e. those with developed embryo) in each sample, on the basis
of which it is justifiable to reckon the proportion between the two species, also as?
regards the earlier stages, at any rate with a considerable degree of exactitude.

DiAMKTEKs in millimetres for Eggs belonging to the genus Oadus, from different

Canadian localities.

May 30.

June 9.

■Tune 11.

•June 15.

August 4.

August 5.

—

" Acadia'" Sta. 6.

"33'
Sta. 10.

"Princess"
Sta. 10.

" Princess"
Sta. 2K.

" Princess" Sta. 31.

"Princess''
Sta. m.

Oadits sp.

G.

*fietilefinus

Gadus sp.

Gadus sp.

Gadus sp.

GmIus sp.

*G. ealf (iritis

Gadus ap.

mm.
1.15

i'

14
23
38
33

8

4

1

4

35

43

44

41

K

1

1

1.20

1
4
7

10
2
1

4

6

7
6
2

1

io'

!)
5

2L

1.25
1.30
1.35
1.40
1.45

6

15

1

4
5
4

21

27

• 18

9.

1.50

* With diagnostic pigmentation of the species.

22

DEPARTMENT OF THE NAVAL SERVICE

As will be seen from the table, the ova of cod and haddock have a diameter of
1-15 to 1-50 mm.; a number of safely determinable cod eggs 1-25 to 1-35, and
haddock eggs 1-30 to 1-45 mm.

With regard to the pigmentation, the larvae of cod should most properly be taken
as belonging to the type described by Schmidt from the southern part of the North
sea ; it may perhaps be designed as a warm-water type, with extremely faint pigment.
Vide fig. 17, and, for comparison, also fig. 16, which shows a more northerly type. As
no larva of this type was available suitable for illustration purposes in the material,
one from Norway has here been used.

On looking through the tables, we find a single haddock egg in the gulf of St.
Lawrence, and no haddock larvse; the material from here can therefore, practically
speaking, be taken as belonging to Gadus rallarms.

Fig. 14.

Fig. 15.

24

DEPABTMEXT OF THE NATAL SERTWE

Gadus

I .•i-ioopr Station all hauls

"^^■'l^^more than 100 •'

Q ' - 10 pr. Station
)more than lo

Fig. 16.

Only ou the cruiacc of the Avadia arose any risk of confusion with haddock,
and here, more especially in the case of the southerly stations, where examples of
haddock ova were in the majority.

The charts show very distinctly the distribution of the cod egg's. We find them
always over the banks, and as a rule in greatest numbers where the banks shelve down,
but never distributed over great depths.

Cod larvae occur very sparsely; only on the ^Newfoundland banks have we at
Station 83 a fairly good yield of 216. the sizes here varving between 3 and 10 mm.
and ranging for the greater part between 4 and 6 mm., i.e., comparatively newly
emerged fry.

Haddock larva; were found only on the cruises of the Acadia, especially the
second, off Nova Scotia and ou the Newfoundland banks. They are very few in
number ; only sixteen for both cruises together. The size varied between 3 and 15 mm.

Cod and haddock have, as we know, a very wide area of distribution in the
northern hemisphere. The haddock ranges from the bay of Biscay as far as Spitz-
bergen; on the American side, Avhere, by the way, it is not so numerous, down to cape
Hatteras.

In Euroi^ean waters, the haddock spawn from January to June.

The cod has more or less the same distribution as the haddock, penetrating,
however, also down into tke Pacific. It spawns at the same time. Spawning has,

CA^ADIAy FISnERIlJS EXPEBITIOW lDl',-irj

25

moreover, been recorded in August on a single bank of the North sea. The spawning
takes place for the most part on the bank^, where the temperature keeps at about 4°
C, and it is generally supposed that the cod, for this reason, move down from the
Polar seas to the southward, in order to spawn en the Norwegian coastal banks about
Lofoten, which are washed by the temi^erate waters of the Gulf Steam, f'od held
in captivity seldom spawn in water below 2° C.

The ppawniiifr in Canadian waters evidently extends over a very loniz jx-riod. On
the first cruises of the Acadia and Princess, May 29 to June 16, free larvae of cod were
found, while on the last cruises, July 21 to August 12, ova were still obtained in early
stages. This is doubtless due to the extraordinary conditions of temi)erature in these
waters.

Of the Onus eggs and larva>, most no doubt, belong to 0. cimhrius, Plate III,
figs. 21 to 22, but as pigmentation, especially of the larvae, varies somew^hat, I could not
feel certain that all belonged to this species. A number of Onos eggs from the Acadia
captures certainly belong to another species.

Onos sp

• i-ioopr Station all hauls

Princess I St 3-26 ^^95(^f„ore than loo .

jrjmore than lo

Fig. 17.

Onos sp.

J. ■ H )^ i-ioopr Station all hauls

Princess n St 27-50 ^i'$'5|0more than 100 ■■

»■ Acadran St 37-91 /-^,L„ i Q ' " '° P^ Station
Z-az-nae ^^rno^g thamo

Fig. 1!

C'ANADIAX FIHHERIEf^ EXPEniTlO\, Wl'rlJ

27

"35'

LarM'ae I ><

; \0

Gnos sp.

I -lOo pr Station all hauls

more than lOo

I - 10 pr Station
more than lo

Kig. ly.

The diameter of the egg's is between OS and 0-9 nun. as regards those from the
gulf of St. Lawrence. At the Acadia Station 47, however, the diameter was 0-70 to
0-75 mm.

Onos cimhrius is widely distributed botla in European and American waters. It
does not, however, ])enetrate so far to the northward as the cod; lives mainly on soft
bottom in fairly deep water, near the coast. It spawns more especially in May
(Europe).

As regards its distribution in the gulf of St. Lawrence we notice that it is
especially numerous about Prince Edward Island and towards Anticosti; always close
to land or above the more shallow banks.

Ova of Merluccins were found in considerable quantities on the second cruise of
the Acadia, especially above the banks off Nova Scotia.

'. Cv+

^\

r^^

\

+

'N

\

+

^

_ +

Mer/L/c/us

I - 100 pr Station all hauls
I more than loo ..

Fig. 20.

CANAhfW FISIIKinr.s IIXI'FJHTIOS, 1<)1>,-15 29

A nuniber of measurements made give the following' results : —

0*90 mm 6 eggs.

0-95 " 28 "

1-00 " 5 "

Merluccius merluccius is not noted in Jordan and Everraann's " Fishes of North
America." The ova, however, certainly agree very well with the European form.
Larvae of this species were not found, l)ut some few eggs in a fairly advanced state
of development.

Of the American forms, M. hilinearis (Mitchill) is the species most closely
resembling the European, and the ova should probably be ascribed to this species.

In European waters, Merluccius is of very common occurrence in the Mediter-
ranean and off the coasts of England; it extends, however, up as far as Xorway and
Iceland.

In the Mediterranean, it spawns in tlie spring months; in the more northerly
waters, during summer.

15. FAM. AMMODl'TID.E.

Ammodytes tohianus (Linnaeus).

(Plate III, Figs. 23 and 24; Tab'.e II j.)

A. tohianus was found in greatest numbers on the first cruise of the Acadia,
eighty-nine specimens being secured.

The second cruise of the Acadia, the cruise of A'o. S3, and the first cruise of the
Princess, furnished together only thirteen specimens; the second cruise of the Prin-
cess, none.

It would seem to be found especially above the banks, but has also occasionally
been encountered over the deep channels, and closer in to land. The length varies
from 7 to 25 mm. the diflPerent sizes apparentlj^ occurring together.

A. tohianus has a wide European area of distributii^i, extending from Spain to
Einmarken, the "White sea, Iceland, and Greenland.

It is probably identical with A. americanus, DeKay, which species, according to
Jordan and Everniann, ranges from Newfoundland to cape Hatteras.

The European form spawns in the autumn at about 20 metres depth, where the
ova are attached to grains of sand.

The eggs may be hatched during the course of the winter; not until spring, how-
ever, are the larvae found in any considerable nuaibers among the plankton, where
they are then often encountered in enormous quantities.

IG. FAM. TETRAGONURIDyE.

Tetragonnrus cvrieri (Risso).

(Plate III. Fig. 25.)

One specimen of this Atlantic species, measuring 76 mm. was taken at Acadia
station 56; (210 to 140 fathoms).

T. cuvieri is especially numerous in the Mediterranean and adjacent portions of
the Atlantic; according to Jordan and Evermann, it has only once previously been
taken oif the coast of America. It keeps chiefly to deep water, feeds on medusae, and
its flesh is s.iid fo be highly poisonous.

30

DEPARTMENT OF THE XAVAL SERVICE

17. FAM. GASTEROSTEID^.

Gasterosteus aculeatus (Linnaeus).
= G. hispinosus (Walbaum).

The three specimens of this species taken would seem to be full-grown, living
pelagically in the neighbourhood of land.

Their total length was 65, 44, and 38 mm. they were taken in surface hauls, at
Princess stations 28 and 2'9, and No. 33 station 13.

G. aculeatus has an extraordinarily wide area of distribution in the northern
hemisphere, and is found down along the coast of America as far as New York. It
also penetrates up into the rivers.

]8. FAM. SALMONID^.

Mallotus villosus (Miiller).

(ipiatie III, Fi'gis. 26 land 27 ; Table II k. )

Fig. 21.

CANADIAy /•7.S7/A7i'//;n hWrEDlTIOX. lOl'rlS 31

This arctic cireuniiwlar species ofteu occurs in great numbers down towards
Finmarken and Newfoundland. Tn North America waters it is found as far down as
cape Cod.

In Norwegian waters, it spawns along the coast from April to June, depositing
its eggs in great (juantities, chiefly on sandy bottom, at depths to about 100 metres.

The newly emerged larva is about 7 mm. long. Of M. viUofiiift, a considerable
number were collected, esi)ecially around the coasts of Newfoundland ; it occurs more
sparsely out towards Labrador and down in the direction of Prince Edward Island.

The size of the larvae varies between 7 and 42 mm.,, the great majority, however,
being between 10 and 20 mm.

19. FAM. STOMIATIDJ^.

Stomias hoa (Eisso).
Cyclothone sp.

Stomias hoa (one specimen of 30 mm. taken at the surface, Acadia station 40) is a
cosmopolite, belonging to the warm seas of the world.

Of Cyclothone, five specimens were taken, ranging from 7 to 16 mm.; these were
not determined as to species.

Two specimens from the surface, 8 to 16 mm. Acadia station 16.

Three specimens, vertical haul, 200 to 0, 7-8 and 10 mm., Acadia station 16.

The genus CyclotJione comprises several bathypelagic species from various warm
and temperate seas of the world.

20. FAM. SCOPE LI DyE.

Scopelus sp.

Omosiidius elongatus (A. Brauer).

Myctophum glaciale (Reinhardt).

M. henc'iti (Cocco).

M. punctatum (Eafinesque).

M. humholdti (Risso).

(TABLE III)

32

DKrARTMEXT OF THE NATAL SERVICE

^

+
-3

+

Princess n St 27-50
Acadia n St 37-91

*>

McllOtU3 [////05U5

/ / O I -10 pi" Station al

^^''^'^^lOrnore than 100 ■. -

Scopelus sp.

hauls

a

more than lOO

Fig. 22.

A large number of Scopelida^ were taken at the outermost Acadia stations. It was
found impossible to determine species in the majority of examples, the preservation
being frequently imperfect. Omosudis elongatus was taken at three of the Acadia
stations.

station 16 200 — m. 2 species 25 to 36 mm.-

57 150 — 50 fms. 1 species 105 mm.

75 ISO — fms. 1 species 45 mm.

This species belongs to the warmer portions of the Atlantic and Pacific.

M. glaciale (one specimen of 15 mm. from Acadiu station 45, 150 to m.) This
species is of common occurrence on the^ east coast of Apierica, but is also found
throught the whole of the i^orth Atlantic. Also taken off the coast of Greenland.

M. henoiti. one specimen, 15 mm., from Acadia station 16 (surface haul). Belongs
chiefly to the warm Atlantic waters ; also penetrating into the Mediterranean.

M. punctatum, one .specimen, 23 mm., from A cadia station 16, 200 — m.

This species has frequently been taken up towards Newfoundland. (A. Brauer.)

M. humholdti, one si>ecimen, 31 mm., from Acadia station 11 (surface haul).

Found in all warm seas

CANADIAX FIRHERIEH EXPEDITIOX, lDl.'ri5 33

III. DISTRIBUTION OF EGCiS AND YOUXG IN THE WATERS

INVESTIGATED.

It is a well-known fact that certain species have their more or less definite
spawning grounds, or, at any rate, are to be found during the spawning season, at
places where certain physical conditions prevail, these conditions being, presumably,
such as are best adapted to the propagation of the species in question.

When spawning, our most important food fish assemble upon grounds of restricted
area, where they can naturnllv be tnkeii in the creatoit iiumber: it i=; therefore of
considerable importance, for the fisheries, to ascertain the position and extent of
such grounds. This may be done by making experimental hauls to secure the spawning
fish, or, far more easily, by plankton hauls, from the results of which it will be seen
where the newly spawned eggs and young are to be found, whence we can determine
the locality of the spawning grounds. This latter method has been largely employed
in the international investigations in European waters, and has led to the charting of
the spawning grounds of the dift'erent species.

If we now consider the results of the various cruises in Canadian waters, it
will soon be seen that the eggs and young of each species have one or more principal
areas of distribution, diminishing in frequency to either side of such area; further,
that certain species generally occur together, others again being invariably found apart.

This distribution is dependent upon the diiferent circumstances under which the
fish spawn, and we shall see that each species is always encountered under certain
conditions, wherefore a certain degree of regularity may be shown to exist in their
occurrence. The eggs and young of littoral fish are found close to land, or over the
shallow banks; the banks have their own particular young-fish fauna, and the deeper
parts, again, theirs. To these must be added a fourth group, to wit, that of the
pelagic species, such as the mackerel; in the case- of these, the question of depth is
less important, and their eggs may therefore be found scattered throughout all areas,
where they may be taken together with the ova of the three first groups. An interesting
phenomenon, however, is the fact that both eggs and young of practically all our
principal species of fish have a pelagic stage, during which they are to be found near
the surface, without regard to whether they were spawned there or at greater depths.
In the case of most species, both eggs and young are so balanced as to float in ordinary
sea- water, and when spawned deep down will, owing to their own buoyancy, rise to
the surface, where their development then proceeds. The young will thereafter, as
their growth advances, again move down to greater depths, or in towards land, passing
from a pelagic habit of life to a more or less pronounced bottom stage.

A. The Gulf of St. Lawrence.

The various cruises furnish a number of instances showing how the character of
the fauna changes as we pass from the vicinity of the coast outwards, or from colder
to warmer waters, and vice versa.

As already mentioned, the gulf St. Lawrence was investigated by the Princess,
June 9 to June 15 and August 3 to August 12, with some occasional work done by the
No. S3 from June 1 to August 18; this latter cruise may be disregarded in the present
chapter, both on account of the length of time covered, and also because the voyage was
not planned according to the special lines of our present purpose. The route followed
on such cruises is, it should be noted, of great importance both for the investigation
of the different localities, and also for the collective survey of the waters as a whole.

The cruises of the Princess (vide chart, p. 4) both proceeded from Prince Edward
Island over to the eastern point of Anticosti, continuing up to Labrador, and thence
eastward to Newfoundland, then following the coast southward, past Cape Breton, and
bade to Prince Edward Island. Glancing now at the hauls inade at the different

6551—3

34 DEPARTMENT OF THE NAVAL SERVICE

stations, we find close to land (Stations 3 and 4) very few larvae; only one or two
specimens of the arctic species Anarrliiclias latifrons and MaUotvs villosus; moving
out over the banks, we encounter DrepanopseUa and Gaclus callarias, with some few
ova of Onos cimbrius, besides some arctic fish, viz., Icelus hicornis and Agonus de-
cagonus. At station 9, near the edge, G. callarias and DrepanopseUa are again less
numerous ; we find, however, instead, the young of Sehastes marinus. Stations 10, 11
and 12 above banks east of Anticosti reveal once more a true bank fauna, Drepanop-
seUa and G. callarias.

Stations 13 and 14, up towards Labrador, furnished but a poor yield; the stations
nearer Newfoundland, however, are a^ain seen to be rich in eggs of DrepanopseUa and
G. callarias, with some few Sebastes in the vicinity of the deeper channels. Here, on
the slope towards Newfoundland, lie the richest stations of the whole cruise; stations
17 and 18, for instance, furnished each over 1,000 cod eggs in surface hauls alone,
besides a considerable number of DrepanopseUa ova. Moving southward along the
coast, the hauls are still rich as regards these species, and at station 19, close in to land,
we encounter, for the first time, ova of the southerly form Ctenolahrus adspersus.
Above tlie deep channel in Cabot Strait, Sebastes predominated. Not until station
24 is reached do we again meet with the banks fauna, G. callarias and DrepanopseUa,
which are thenceforward of common occurrence throughout the remainder of the cruise
back to Prince Edward Island. The last stations, south of Cabot strait, also furnished
quite a respectable yield of newly-spawned mackerel eggs, which was not a little sur-
prising in view of the fact that the hauls made but a few days previously on the
opposite side of Prince Edward Island had yielded nothing but more or less arctic
species.

The August cruise of the Princess (vide chart, p. 4) covered, roughly speaking,
the same ground as the first. Close m to land, the yield consisted mainly of eggs and
young of Onos and Ctenolahrus, until we reach the banks, when cod and mackerel
make their appearance. These latter increased in numbers as the outward voyage pro-
ceeded, the Onos and Ctenolahrus gradually disappearing. Above the deep channel,
Sebastes predominates; some few mackerel larvae were also found. On the Anticosti
bank, we find cod eggs once more, and the same alternation of cod above the banks,
Sebastes over the channels, is continued all the rest of the way to Newfoundland, and
thence southward to Prince Edward Island. Station 45, however, near the southwest
point of Newfoundland, also furnished a rich yield of the arctic species Mallotiis
villosus, evidently indicating the existence of a cold current near at hand; on the
southern side of Cabot strait, however, eggs of mackerel and Ctenolabrxis were again
encountered.

As will be seen from the foregoing, the gulf of St. Lawrence may be divided up
into several zones, each with its own characteristic young-fish fauna. In the first
place, a line drawn from Cabot strait to Anticosti will form the northern limit of
occurence for the southern forms found in the gulf, such as mackerel, Onos and Cteno-
lahrus. Only at one or two of the more southerly stations situated near the coast of
Newfoundland were some few specimens of these species found.

Northern species, however, may also be found south of this line, occurring more
particularly in the southwestern portion of the gulf, ^ west of the Magdalen islands.
For the rest, the character of the young fauna varies with^ the depth; we find Ctenola-
brus and Onos near land, and close to quite shallow banks, tlie bank forms consisting
fo G. callarias and DrepanopseUa, while above the deep channels Sebastes is found.
The mackerel, again, is encountered throughout all areas, subject to the restriction
noted above.

B. The waters orxsroE Nova Scotia axd the Newfoundland Banks.

The cruises of the Acadia (vide chart, p. 4) reveal a similar grouping of the
different species, according to depth, distance from land, and geographical position.

VAyADIAy FI.^HKRIi:s EXPEDlTIOy, lOJJf-lS 35

Ou the May cruise, the stations from Halifax out over the banks past Sable
island yielded eprprs of cod and haddock, several species of flounder, Animodiites, and
a very small number of Ctenolahrus above the shallower prounds near Sable island.
Outside the banks we find Sehastes, but farther out again the yield becomes extremely
poor until the outermost station (station IGj, where a number of pelagic Atlantic
forms were found. On reaching the Newfoundland banks, we find once more cod,
Drepanopsetta and Ammodijtes; above the deeper portions towards Cape Breton and
Cabot strait, Sehastes is of more or less frequent occurrence.

Tlie second cruise of the Acadia shows small yields at the start, off the southern
point of Xova Scotia; strangely enough, however, we here find Sehastes at comparatively
slight depths. This is probably connected with the fact that Sehastes in these southern
waters (Bay of Fundy and Xew England waters) spawns in more or less shallow water.
On the slopes, where the depth increases, we find numerous eggs of Merluccius;
farther out, the Scoi)elida^ play a more prominent part. Ova of cod and haddock
are now but sparsely found at all stations outside Nova Scotia ; on the Newfoundland
banks, however, they are once more encountered in great ninnbers, stations 81 to S.3
being particularly rich. Eggs of Drepanopsetia also are fairly numerous here, while
the arctic capelin bears witness to the vicinity of the polar current. From the banks
over to Cape Breton, Sehastes is again more frequently found, Onos (hardly however,
Onos cimhrius) only in very small quantities, Ctenolahrus only at station 91, quite
close to land. We see, then, that in this area, likewise, distinction may be made
between several different kinds of fauna. Up towards Newfoundland, we have —
according to European terms- — eggs and young of boreo-arctic fish, on the banks off
Nova Scotia a boreal bank fauna, while out over the great depths, especially to the
sovithward. the Atlantic species predominate.

On comparing the stock of young in these two areas; the gulf, and the waters
outside Nova Scotia, we find that the gulf should most properly be regarded as a
coastal water, where littoral and coastal forms are found in great numbers; we find
the usual high degree of variation in temperature between summer and winter, with
corresponding variation in the fauna between northern and southern forms; purely
Atlantic species, however, do not appear in the hauls.

An interesting feature, by the wav, is the fact that the mackerel is not represented
at all in the Acadia material (but few eggs on the first cruise); while the greater
portion of the Ammodytes taken during the investigations was furnished by the
Acadia. It is remarkable also, that none of the cruises furnished a single herring larva,
despite the fact that the herring occurs in nearly every part of the areas investigated.

IV. BIOLOGICAL CONDITIONS.

Modern marine research has shown us that life in the sea is to a high degree
dependant upon hydrographical conditions: current, temperature, and salinity. We
should, therefore, in seeking to explain the peculiar phenomena encountered with
regard to the propagation of the fish in Canadian waters, base such endeavours
upon tlie liydrographical data furnished by the ex]>editron. These are, however, not
yet compiled, in order and completeness, and we must for the present content ourselves
with recalling the more prominent features. It will be obvious from the outset that
an area subjected to such markedly contrasted influences as those of the Labrador cur-
rent and the Gulf Stream must present many remarkable features both as regards
fauna and biology, and it is only by bearing in mind the peculiar hydrographical con-
ditions that we can explain the otherwise surprising data furnished by the Canadian
Hydrographical Expedition wifli re.o-ard to the propn<rnt.ion of tlie most important
species. On comparing the Canadian with the European waters in this respect,
several peculiarities will at once be noticed, especially the pronoimced transposition
and extension of the spawning season, which here takes place in the case of several

6.551— 3*

36 DEPARTMENT OF THE NAVAL SERVICE

species, among those pre-eminently which on the eastern side of the Atlantic are
counted as spring-spawners. A prominent example is that of the cod (in addition
to haddock and JJrepanopsetta) which spawns in considerable numbers in the gulf of
St. Lawrence right on into the month of August, whereas in the Norwegian waters,
its spawning- is over in May. As a consequence of this, we may, in the gulf of St.
Lawrence, iind one and the same plankton haul to contain both newly-spawned cod
eggs and newly-hatched mackerel young.

That cod and mackerel should find suitable spawning conditions in the same
water and at the same time is remarkable indeed; in European waters, the spawninjj
of these two species differs widely, both as regards time and place. What is more, it
is only exceptionally that the mackerel come so far north as to touch the principal
spawning grounds of the cod at all. The explanation may not unnaturally be imagined
to lie in the abrupt changes of temperature encountered in the Canadian waters. The
same natural conditions which cause Atlantic warm-water forms to meet, off the
coasts of Newfoundland, with representatives of the arctic fauna, might well account
for the fact that in the gulf of St. Lawrence, the cod of the far north are able to
live and propagate there simultaneously with the more southerly mackerel. If we
take a temperature series from the surface downwards in the gulf of St. Lawrence
during summer, we find, within a vertical range of less than 100 metres, a difference
in temperature equivalent to that encountered outside Newfoundland by sailing from
the Gulf Stream over into the Labrador current, while on the eastern side of the
Atlantic, such difference will involve a distance as from the North Sea to Spitzbergen.

In the gulf St. Lawrence, then, we find the different water layers varying so greatly
in temperature as to afford favourable conditions both for cod and mackerel. That
both species should live in the same stratum . there is no reason to believe; on the
contrary, it must be presumed either that the fish are themselves able to seek out the
more suitable water layers, and avoid those less favourable, or that they are, owing
to certain biological features, more or less passively led to frequent such strata as
are by nature best adapted to their needs.

It is a well-known fact that certain species of fish, such as cod, herring, and
mackerel, vary their habitat according to the temperature of the water. Cod and
herring may at times move in shallow, at others in deeper, water, and in the southern
Norwegian waters the mackerel seems rarely to move in towards the coast until the
surface temperature is at about 10° C, a fact so generally recognized that some of the
drift-net fishermen even carry a thermometer in order to make sure of not laying out
their nets in a cold current.

Among the biological features which might be thought to act in this direction,
we should first of all remember that the cod is a bottom fish, belonging essentially to
the strata nearest the sea floor, where its food is found, and where its spawning takes
place. The mackerel, on the other hand, is a pelagic species, frequenting — at any rate
during the spawning season — the upper-water layers. Indeed, until the roe is spent,
the mackerel hardly seems able to penetrate down beyond some 20 metres depth.

On glancing at a temperature series from the Princess station 30, where ova of
cod and mackerel were found in one and the same sample, the phenomenon will easily
be explained.

Om t= le-l" ot = 20-27

10 " IS-T" 20-33

25 •' 10-S" 21-97

4n " 7-9" 22-99

50 " ; .. 5-65° 23-97

fiO " 0-45° 25-40

If we now apply to these figures what we know of the biology of the cod and
mackerel, it will be justifiable to conclude that the mackerel keep to the upper layers —
down to about 25 m. — where the temperature is favourable to its needs, -while the cod
seek the lower strata, finding there more suitable conditions in this respect.

CA^ APIAN FISHERIES EXPEDITION, 19Vrlo 37

The yield of cod and mackerel eggs at this station gives the following numbers: —

Cod. Mackerel.

At surface. 3 eggs. 172 eggs, 14 larvae.

*• 30-0 metres 12 " 8 " 3 "

" CO-0 " 25 " 1 "

Mackerel eggs and young are by far most numerous at the surface (in the horizon-
tal hauls) the cod eggs, on the other hand, being found in greatest numbers farther
down (vertical hauls). Naturally, however, this cannot be taken as indicating that
the mackerel eggs were spawned near the surface, and the cod eggs lower down. The
vertical position of fish ova in the layers is determined by the specific gravity of the
former as compared with that of the ^vater, and they will, under normal conditions, be
found in water of specific gravity equal to their own. The specific gravity of ova,
however, is not a constant magnitude; that of cod eggs, for instance, which normally
varies but very slightly, may be affected by a low degree of salinity in the water, which
has the effect of gradually reducing the weight of the eggs. If exposed to strong sun-
light, again, or to comparatively high temperature, they will sink, even in water of
high salinity; and the same applies to newly hatched young. This last is a very well-
known phenomenon in marine hatcheries; I am, how'ever, unable to give any definite
figures, having had no opportunity this season of making experiments in this direction.

The specific gravity — and variation of same — in cod eggs at Flodevigen will be
seen from the following experiments (Dahl & Dannevig; Undersokelser over nytten
av Torskutlsekningi Ostlandske fjorde, Bil. I, p. 28).

The roe and young were placed in three separate receptacles in order to determine
their approximate specific gravity.

I. Sp. gr. 1019 at + 5-4° (sp. gr. in situ, 101796).

Ova, about one-third floating, others suspended, the remainder at the bottom.
1 oung, most at bottom, some suspended and at surface.

II. Sp. gr. 1-020 at + 5-4:° (sp. gr. m situ, 1-01897).

Ova, more than half floating, the remainder at the bottom or suspended.
1 ou7ig, about half floating, the rest at bottom or in suspension.

III. Sp. gr. 1-022 at + 5-4° (sp. gr. in situ, 1-020965).
Ova, practically all floating.
Young, practically all floating.

Station 30, above-mentioned, the specific gravity of the water near tlie surface
oscillates about the value for specific gravity of cod egg^ as noted in Norway : we
might therefore have expected to find a greater number of ova at the surface. That
this was not the case should possibly be ascribed to the influence of the high tempera-
tures there prevailing.

^Ve find, then, that the simultaneous spawning of cod and mackerel in one and
the same water may be explained as due to conditions of temperature. And conse-
quently, we cannot, in Canadian waters, distinguish between winter and spring-spaw-
ning fish, as is generally done in Europe; if any such biological distinction is to be
made, it must be based upon the limits of temperature between which spawning takes
place, a method which would of course be equally justifiable on the eastern side of the
Atlantic.

The question arises, however, whether the temperature of the water really does
exert such influence upon the spawning of the fish; whether other causes might not
be imagined, and the explanation above suggested be found insufficient.

In this connection, it will be natural to recall the annual periodicity in the
development of the sexual organs, which might be supposed to proceed independently
of all extf^rnal factors. It has, hovever, repeatedly been found, by investigation in

38

DEPARTMENT OF THE NAVAL SERVICE

other spheres, that where regularity of the seasonally varying physical conditions in
animals or plants ceases, the annual period disappears, or becomes irregular and
indistinct.

The influence of temperature upon the daily intensity of the spawning ma.-v
easily be observed in a hatchery, at any rate whei-e a spawning basin is employed.
Also the degree to which spawning depends upon temperature is in particular most
easily noticeable where the latter factor ranges beween 0° and + 5°, to a lesser extent
-when between + 5° and + 10°. Other influences may make themselves felt, as for
instance of the amount and more or less regular distribution of the food, the fish
spawning largely when well fed. This may, however, ix)ssibly be due to purely phy-
sical conditions, at any rate, as long as the amount of food consumed is sufficient for
nourishment, as the expansion of the stomach will exert a pressure vipon the ovaries.
Furthermore, the salinity of the water may be supposed to have some effect though
this will scarcely be of any importance as long as the variation involved is not too
great.

■ Temperature centigrade]
- Eggs in liters

2% 2Ji ys % '^3 % =% ^i 54 » '5^

Fig. 23.

The curves here shown are based upon daily observations at the hatchery of
Flodevigen during the spring of 1909, and indicate the fluctuations in temperature
and spawning throughout the essential part of the spawning time. In order to present
the clearest possible survey, I have here marked off, not the daily values, w-hich vary
considerably, but a mean value calculated each day from the observations of that date
plus those of the two previous and the two subsequent days, i.e. a movable mean extend-
ing over five days.

"V. — Possibilities of development, axd relative fkequexcv of the different stages.

In the foregoing chapters, we have dealt with the spawning of the different
species in relation to depth, temperature and season. It now remains to consider how
the development of the eggs in the separate species takes place, and what chances
there are for their being hatched.

From the figures in table II, we obtain the following numbers of eggs and larva?

for the six most commonly occurring species : —

Ova. Twarvse.

Ctenolabrus ' 434 318

Scomber 5,575 47

Drepanopsetta 2,369 47

Gadus callarias and G. a-rjlefinn^ 15.988 293

Onos 2,975 92

Merluccius 8,8 21

It will be seen that the gadoids here predominate, and among these, as we have
already seen, the cod is of most frequent occurrence. Ova of Merluccius are also very

CAyADIAX FlHHi:RlEH i:.\ I'i:iHTIO\. 191 ',-13 39

numerous, but the material of these is of a somewhat casual character, being derived
from ouly a few stations, where they must, however, have been present iu great num-
bers.

The proportion between ova and young varies greatly; in the case of Ctenolahrus,
we have nearly as many larvae as eggs while of Merluccius, on the other hand, only eggs
were found.

Such a table cannot, however, be used without reserve as a means of calculating
the percentage hatched for the different species; the period of time required for deve-
lopment of the ova will need to be taken into consideration, as also the age of the
young when captured. With most species it is true these points are insufficiently
known, but in the case of the cod we have knowledge sufficient for a satisfactory dis-
cussion of the question despite the difficulties involved in following the development
of ova under natural conditions.

We know, of course, that the destruction of cod eggs and of the early stages of
young must be enormous ; were this not so, the sea would soon become over-populated
with cod. As to the particular stage of development at which the germ is specially
liable to destruction, however, we know, unfortunately, all too little.

Professor Apstein has, in his work entitled "Die Verbreiting der pelagischen
Fischeier und Larven in der Beltsee und den angrenzenden ^leeresteilen," 1908-09,
dealt with these questions, and calculated that 72 cod eggs are required to produce one
larva. This applies, of course, only to the waters investigated by him, and similarly,
here only for a single year.

THE CAXADIAX WATERS.

During the actual collection of the material. Dr. Hjort already noticed that there
was a very great majority of newly spawned eggs, the later stages of development, and
the larvae, being- far more rare. This was especially noticeable on the cruises of the
Pnncess. Wishing to go closely into this question, I have, in determining the ova,
noted throughout the number of eggs belonging to earlier and later stages; in the
smaller samples all the ova were examined, in the case of the larger, only a represent-
ative portion was measured off and investigated, the figures thus obtained being then
utilized for calculation of values for the whole.

In order to keep the work within a reasonable scope, distinction has here been
made between three stages only; the first being the '' germinative disc" stage, before
the formation of the embryo (figs. 1-2), the second from the early pigmented embryo
to well-developed pigmentation (figs. 3-4), and the third with pigmentation character-
istic of the species (figs. 5-6). These stages correspond to Apstein's figs. 1-0, 10-18
and 19-22, from which these text figures have been taken.

40

DEPARTMENT OF TEE NATAL SERVICE

EiiBRYOxic Development of Egg of Cod.

This system of division is based on no other principle than that of obtaining, as
easily as possible, a survey of the numerical values at the different stages, these last
being so selected as to be distinguishable with no great difficulty when slightly
magnified, and without regard to ^\hether one stage may be of longer duration than
another.

Apstein's tables show, for instance, that my "germinative disc" stage embraces
31 per cent of the total period of development in the ovum; stage 11, 46 -per cent, and
stage III, only 23 per cent.

Taking the material consisting of cod eggs from the cruises of the Princess we
find, for all hauls made, the following figures : —

First cruise. Second cruise.

Stage 1 4.315 1,391

II 799 340

III 13 33

Larvse 6 6

We find then, on both cruises, an overwhelming majority of ova in the early
stages. How is this to be explained? Is it to be taken as entirely due to mortality or
destruction by some means, or do other factors here exert their influence, rendering
the immediate impression produced by the above incorrect?

As we shall see, a table of this sort should be treated with the greatest caution.
The proportion between the different stages is not only dependent, for instance, upon
mortality, but also to a high degree upon the time at which the investigations were
made; in addition to which, we have also to consider the position of the stations
in relation to the spawning grounds.

'When the investigations are carried out early in the spawning season, then
naturally there will be a very high percentage of newly-spawned ova, and no late
stages. Investigations later in the season, on the other hand, will give the opposite
result.

A few rich hauls made on the exact spot where spawning takes place wi)l give
a different result to that obtained from hauls made out over the greater depths where

CAN Am AX FISHERIES EXPEniriOy, 191 'rJo

41

the fish do not spawn; tho few eggcs, found outside the banks, will have been carried
thither by the current, and the longer they have been adrift since leaving the spawning
grounds, the farther will their development have advanced. On the spawning grounds,
therefore, we must expect to find newly-spawned eggs in the majority; outside these
localities, the later stages will predominate.

The investigations in the gulf of St. Lawrence are, however, to a very great
extent free from these sources of error. It will be seen from the table that the
spawning was in full progress during the first cruise of the Princess in June, and that
so few late stages were found, may be explained as due to the fact of the hauls in
question being made comparatively early in the season. On the August cruise of the
Princess, however, this objection is no longer valid; the number of ova has greatly
decreased, the season being now-nearing its close, and a far greater quantity of eggs
in the later stages might have been expected.

'On going through the tables for the cruises of the Princess it will further be
noticed that the proportion between newly-spaAvned ova and those in the later stages
is approximately the same at all stations, only exceptionally do we find the later stages
in the majority. No single station is so rich in newly-spawned ova as to exert a
dominant influence upon the result as a whole, and there is thus no reason to question
the representative value of the material on this head.

As regards the number of larvae, this will always be subject to considerable
error, as the larvae are more or less endowed with power of motion, rendering capture
more difficult, in addition to which, their age is often difficult to determine. Com-
parison between ova and young will therefore be subject to greater uncertainty than
that between ova at different stages.

One thing we can, however, state with certainty : that the number of larvae taken
in the gulf of St. Lawrence was remarkably small.

In the case of the Acadia, the proportion between the different stages is somewhat
better; here, however, we must consider cod and haddock togeliier as one, these two
species being indistinguishable one from another in the early stages of the ova.

Stiigel

Stage II

/ cod

I haddock ....

Lar'

Aeadia II.

\-= 72
= 275

For these cruises, the results are as might have been expected, the May cruise
showing a comparatively large number early stages, with some advanced and a few
larvae, while that of Jvdy has many later stages and still more larvae.

Before proceeding to consider the possible explanation of this great difference
between the occurrence of early and later stages in the gulf of St. Lawrence, it may
be of interest to compare the conditions in Canadian waters with those observed on the
principal spawning grounds of the cod in European waters, viz. : —

LOFOTEN.

The material upon which the following discussion will be based consists of but a
few samples from the otherwise extensive mass of material collected by Dr. Hjort
during the winter and spring of 1913, the remainder having not yet been completely
depth with as regards the stage of development of the ova. (Vide also Johan Hjort:
Fhictuations in the great Fisheries of Northern Europe.)

42

nEPARTMENT OF THE NATAL SERVICE

These investigations, besides rendering possible a comparison between the wastage
per cent in the two waters, further afford an excellent illustration of what has been
mentioned in the foregoing as regards the extent to which the respective numerical
value of the different stages depends on time and place of sampling.^

Number of cod eggs pr s
minutes haut indicated at
each station. ^'A-^^ 1915

rtlLOMCTEO

GEOGR MILES (KunRrMIll

Fig. 25.

By way of illustrating these conditiuns under varying circumstances, I have chosen
a section across the Vestfjord, as marked on the chart, and as the surface hauls (1 m.
net, 5 min.) give the greatest amount of material, I have employed these only.

Six stations are noted on the chart, all from April 23 to April 24, 1913, the total
number of cod eggs per 5 mins. surface haul being appended in each case. It will be
noticed that station 65, on the edge, has a pronounced maximum, with 9,800 eggs in
a single haul, after which the number rapidly decreases to either side, as we leave the
spot where the principal spawning presumably took place. During the actual investi-
gations it was frequently noted that the greatest numbers of eggs were taken on the
fishing grounds themselves, where the fish were most densely congregated. The
number diminishes gradually right over towards station 61, on reaching which we
again find ourselves near the edge. Here we might thus again have expected to find
a maximum ; as a matter of fact, however, there was evidently very little spawning, if
any, on these grounds. The banks here, by the way, are of very restricted extent, the
bottom sloping abruptly down from land to a great depth.

With regard to the proportion between the different stages at each of these stations,
we find the following figures (per cent) : —

station 64. Station 63. Station 66. Station 63. Station 62. Station 61.

1 25 6.3 32 30 40 28

II r.5 30 53 57 30 46

III 20 7 15 13 30 26

From this it will be seen that the earliest egg stages are found in the highest
proportion at the very places where the total number was greatest; the more advanced

1 The inve.stigia.tions extejirled tihrooiglhiout t'be whole of the spawning- seaoon, oomprising- alsio
both the bank grounls and the greater depths. The same places were, moreover, investigated
several times during the season, partly with the closing net, and partly with an ordinary sur-
face net.

CAXADIAX FISiHERIES EXPEDITION, 191 ',-13

43

stages, on the other hand, heing of great relative frequency as the distance froni the
spawning grounds increases.

Geirriinative disk
Cod eggs in late stages

Vo
60

50
40
50
20
10

5t. 64 65 66 63 62 61

Total numbeK: 4050 9800 1384 184 105 89

Fig. 2G.

Station G2 differs slightly from the remainder, inasmuch as we here find a greater
number of newly spawned eggs than was to be expected. Evidently, the current must
have brought down the newly spawned ^gs from the banks southwest of the station.
The current at this spot can at times be very strong, and in this very direction. The
numbers, however, both at stations 62 and 61, are not sufficiently great to furnish an
exact picture of the stock there.

The curve for the later stages (III) shows throughout what might theoretically
have been expected, the later stages themselves being proportionately fewest where
most eggs were found. The curve rises as we move away from the spawning banks,
as far as station 62, falling again over towards the land, thus indicating that we are
once more approaching grounds where, in some slight degree, spawning takes place.

As we have seen, the proportion between the different stages varies greatly at
different places, even for samples taken on one and the same day; the variation is,
however, sufficiently regular to permit of our forming some opinion as to the cause.
The example here shown is taken from a locality where the contrast is marked; great
quantities of cod spawning within a restricted area, close to the very deep water
where cod never spawn, but in which a certain quantity of eggs may be found, having
been carried thither by the current. A different state of things would necessarily be
encountered where the spawning takes place sparsely over areas of wide extent, espe-
cially if forming part of a restricted region of sea; at such places, there would never
be the great difference between the relative frequency of the stages at the various
stations.

Stations 64-65 and ()6 were also previously investigated, viz.: 25 '2-12 '?> and 29/3.
Not until the last-mentioned date however, were cod eggs found in great numbers; the
principal spawning was then just commencing. At these three stations (30-31-32)
(29/3) we find the following yield for the different stages (here stated in %).

No. I. II. III.

station .30 (=Station64) 23.600 93% 7%

Station 31 (zzStation 65) 19.400 50% 50%

Station 32 (=Station66) 2.340 25% 75%

The principal spawning this time takes place farther in on the banks; we find,
however, again the same feature as before noted; where the eggs are found in

44

DEPARTMENT OF TEE NATAL SERVICE

greatest numbers, there also the number of those newly-spawned is at a maximum, its
minimum falling there where the yield as a whole is poorest. We notice, moreover,
that advanced stages are not yet present in the samples, this being doubtless due to
the fact that the season is only just commencing.

If we now consider the results from Canada in conjunction with this sample from
the Lofoten material, we obtain a table for comparison of the stages of development
in the eggs from the different waters.

Stage I . .
Stage IT.
Stage IIT

Lofoten

"^"3

23-24^

a>

«

cS

s

•^

rn

rj-j

(12%

38%

0%

4«%
40%
12%,

Acadia I

43%

2%

35%

49%
16%

Fri

a

St%

16%

0%

>

86%
13%

1%

Acadia II

52%
35%
13%.

6S%

2!)%

3%

Princess II

64%
32%

4%

«4%

15%

1%

"33'

03

!'5%
5%
0%

83%
14%

3%

From this it will be seen that the gulf of St. Lawrence is considerably behind the
other localities with respect to the occurrence of later stages, both in the case of the
earlier investigation and those made subsequently (Princess I, Princess II, and
iV 0. 33). The ova have here evidently a far poorer chance of being developed and
hatched than in the other places; in the gulf of St. Lawrence the eggs must — at any
rate occasionally — be liable to a high degree of mortality, either due to the hydrogra-
phical conditions or to their being preyed upon with unusual intensity.

The task of solving this question must be left to fjiture investigations; I may,
howeven, in conclusion, set forth such material as we have available for the considera-
tion of what factors should be regarded as detrimental to the development of cod eggs
and the growth of the young.

VI. — I^FLUE^cE OF Temperature and Salinity upon the Development of Eggs and

Growth of the Young.

In order to ascertain what effect the low bottom temperatures might possibly have
upon the development of the ova in the gulf of St. Lawrence, Dr. Hjort approached
Dr. Johansen and Dr. Ivrogh, of Copenhagen, who had previously at the Zoo-physiolo-
gical Laboratory there, carried out a whole series of experiments with hatching of
eggs at constant temperatures, with the request that they would repeat their investi-
gations having the special object at present in view.

We are aware that cod in the gulf of St. Lawrence spa%vn near the bottom, i.e. in
water frequently below 0° ; the inherent upward tendency of the ova then lifts them
to the water layers above, where they develop at considerably higher temperatures.

Dr. Johansen and Dr. Krogh then cemmenced experiments with fertilization of
ova at low temperatures, from 0° to -^ 2° C, the eggs being thereafter either main-
tained at various low temperatures or gradually transferred to warmer water, vip to
6.6° C.

According to Dr. Johansen, the resiilts showed : —

1. That fertilization can take place at a temperature between-=-0-6° and
-^2 = C.

2. That full development can be obtained by ova fertilized at such temper-
ature and thereafter exposed for nine days to a temperature between -;- 0-7° C.
and -4- 1-4. mean value -h 1°, and afterwards transferred to warmer water
(e.g. 0° to 3.3° C.)

3. That full development can be attained by ova fertilized at 0° C. and
constantly kppt in water at that temperature.

CANAblAS FlsHKRlf-Js EXPElJlTlOS, lOl'rlS 45

The mortality observed in the course of these experiments was considerable;
as, however, these were carried out under conditions which in several respects must
be less favourable than those prevailing in the sea, we should not, as Dr. Johansen also
points out, attach too much weight to this. Even at the most favourable temperatures,
the mortality was, in the apparatus employed by Dr. Johansen and Dr. Krogh, still
considerable.

Dr. Johansen's experiments show, however, that the low temperature at the
bottom in the gulf of St. Lawrence where the cod are to be found during spawning
time, cannot be regarded as a general hindrance to the development and hatching of
the ova.

With regard to the influence of high temperatures upon the development of the
ova, we may refer to the experiments previously made bj' Johansen and Krogh
(The Influence of Temperature and certain other Factors upon the Rate of Develop-
ment of the Eggs of Fishes; Publications de Circonstance Xo. 08). These show that
cod eggs may be hatched in water up to 12° C. H. C. Dannevig (late Director of
Fisheries for Australia) has even hatched them in water at 14° C, (Scottish Fishery
Board, 13th Report for 1891.)

At Flodevigen, where several hundred million cod eggs are annually hatched, low
temperatures (down to 0° C. ; the lowest figure at which we have worked) never seem
to have any detrimental effect; the most favourable temperature would seem to be
about 3°-5° C, while temperatures up to 8° or 10° occasion a higher degree of mortality.
As the quantity of ova in the apparatus is very great, the direct cause of such mortality
might possibly be lack of oxygen, or poisoning, despite the fact that the water is
constantly renewed.

That the young can thrive at considerably higher temperatures in a greater mass
of water is shown by an experiment which I made in the culture basin at Flodevigen
during May and June, 1909. The basin is situated in the open air; it measures 34 by
22 by 5 m. and is used in the hatching oi^erations as a water reservoir for the apparatus.
It is supplied with sea water by means of a steam pump, and in order that the water
may be constantly renewed, as well as for other reasons, the water for the hatching
apparatus is always drawn from here.

On the 25th of May, when the hatching experiments were brought to a close, and
the circulation of the water consequently ceased, about 100,000 young cod, from one to
two days old, were liberated in the basin. The eggs from which these young were
hatched had been spawned at about 7° C, and the development had taken place in
the course of about nine days at temperatures between 7-6° and 9*5° C. At the time
of liberation, the water in the basin at 3 m. depth had a specific gravity of 1-026 at
9-5° C. The young could now be seen every day in the water, at first near the surface,
and later swimming for the most part in great shoals in the intermediate water layers.

On the 16th of June, I noted in my journal that up to fifty young could be
counted at one time. On that day also, feeding was commenced with finely chopped
Mytilus edulis. The water in the basin, by the way, was extremely rich in plankton,
especially larvae of molluscs and crustaceans, which, with some few copepods. made up
the stomach contents in such ofi the young fish as were examiAed. On the same
day, the water in the basin had reached a temperature of 20° O. at the surface, with
a maximum, at 1 m. of 21-4° C, and a bottom temperature of 21-1^ C.

On the ISth of June, eight fish were taken up and measured; they showed the
following lengths : 25, 25, 27, 24, 24, 23, 23, 22 mm. The fish were by this time moving
nearer the bottom.

The weather now set in colder for a time, with cloudy sky, and the temperature
of the water sank somewhat. On the morning of the 21st, the surface showed 18-9°
C. ; maximum of 19-5° at 2 m. On that day, fresh sea water was pumped in (tem-
perature 8-3° C.) and in the evening, after the surface water had been drawn oil, and
replaced h^ a new supply, the surface temperature was 19 -4". this being the maximum,
while at 3 •4m. a minimum of 15-7° C. was recorded.

46

DEPARTMENT OF THE J^'ATAL SERVICE

On the 29th June, pumping was renewed, but as the young were considerably
fewer, (and always at the bottom), the experiment was brought to a close. The tem-
perature of the air had, during the last part of the time, been very high, with sun-
shine and slight wind.

This experiment shows therefore that the young can thrive — and even grow very
rapidly — at temperatures up to 20°. The increment of growth from May 25, when the
length was 3 to 4 mm. until June IS, amounts to about 20 mm. or about 0-8 mm. per
day.

From the measurements available for growth of young in a free state, the latter
would seem to grow less rapidly, presumably about 0-5 mm. per day.

Among other causes which might be imagined as affecting the development of
ova, we may reckon the salinity of the water.

If cod spawn in water of lower salinity than that which corresponds to the specific
gravity of the eggs, then these latter will necessarily sink to the bottom and be des-
troyed. And again, if eggs should by any means be transferred from water of their
own sx>ecific gravity to considerably fresher water, the mortality here likewise would
be very great.

As an illustration of the manner in which fresh water can affect the ova, the
following data from an experiment which I carried out in March, 1909, at the hatchery
may be quoted.

About twelve hours after fertilization, eggs were distributed in four glasses con-
taining water of different siieeific gravity, and at approximately the same temperature.
The glasses were placed in a cold room — temperature of air about 0° C. The water
was not changed, only aerated from time to time with a large pipette.

The cleavage and the development generally were observed under the microscope.
After the lapse of twenty hours some samples were taken and transferred to fresh sea
water in order to ascertain whether the development could there proceed normally.

In the case of sample IV, this was repeatedly done.

The result of the experiment will be seen from the following table: —

1909.

Number

of

hoiirs.

I
Sp. Gr. 1 0247
T. 2-2"C.

II

Sp. Gr. 1020
T. 2 (TC.

Ill

Sp.Gr. 1015
T. 1-8

IV

Sp.Gr. ICIO
T. IG

March 19, 2 p.m

7 p.m

10 p.m

March 20, 10 a.m. . . .

M 21, 10 a.m

„ 23, 4 p.m

,t 24. 10 a.m

5

8

20

44

98

IK)

All alive .

All alive

All alive

All alive . . .

Devel()|)ment

often abnormal
Some dead.
More dead.
Few alive.

A similar experiment with newly hatched young in water of: —

I Sp. gr. 1-0242 at T = 2-0°C.

II " 1-020 " 2-2 "

III " 1-015 " 2-5 "

IV " 1-010 " 2-5 "

V " I'OOS " 2-7 "

gave the following results: —

I. All the young suspended; after the lapse of an hour some few at the bottom,
after four hours the majority near bottom, but all living. After the lapse of 103
hours, some few were dead, but the remainder stood out well until the experiment was
concluded after 148 hours.

CANADlAy FISHERIES EXPEniTIOy, VJl'rir,

47

II. All oil the bottom by the end of an hour, some endeavouring to rise. Same
course as previous sample, only slight mortality after eighty hours, until about half
died by the end of 142 hours.

III. Some few suspended after a quarter of an hour, but at the end of an hour all
at the bottom. Some few died after the lapse of fourty-six hours, mortality then increas-
ing rapidly until all were dead by the end of 138 hours.

IV. All on the bottom after a quarter of an hour, some trying to work upwards.
Some few died after four hours, more after twelve hours, the remainder with irregular
pulsation of the heart. After thirty hours, many dead, and by the end of fifty-two
hours from the commencement, none were left alive.

V. Some few died in the course of the first seven hours, and all within thirty-four
hours from the start.

From the result of these experiments we may be justified in supposing that a stay
of any duration in brackish water will be detrimental to eggs and young; this will,
however, only occur under quite peculiar conditions, and can be of no importance with
regard to the open sea.

Other factors which might be supposed to have some effect upon the development
of the ova are: lack of fertilization, action of bacteria, and internal causes (natural
death).

Prof. Apstein has, in his investigations in the Belt Sea and the western Baltic,
shown that considerable quantities of dead — or as he considers, unfertilized — ova occur
in the plankton samples. In his work above quoted, " Die Yerbreitung der pelagischen
Fischeier und Larven in der Beltsee und den angrenzenden Meeresteilen 190S-09," we
find the following table for dead ova (p. 231) : —

Per cent dead.

Pleuronectes
platessa.

PI. flesus.

PI. limanda.

Gadus
callarias.

Clupea
sprattus.

Vertical

35

38-5

20-6

219
41

19 '8

27 7

3 7

170

1T3

29

12-7

12 8

31-7
49 6
30-3

Surtace

Deep water

Mean

28 4
i

20 9

20 4

31 9

Dead

i

Jt

i

1

These figures, however, appear abnormally high, for it must be remembered that
dead ova — at any rate those of cod and plaice — are only exceptionally found suspended
in the water, sinking otherwise very soon to the bottom, they would consequently only
appear in plankton samples by mere accident. Unfertilized cod eggs may remain
suspended for about two days; those of the plaice somewhat longer.

As to how far unfertilized or dead eggs appear in the samples from the Canadian
waters, I am unable to say on the basis of the preserved material; the question would
have to be dealt with by means of fresh plankton samples.

That cod eggs should be subject to " natural death " in any considerable degree
is hardly probable; if such were the case, the mortality figures at the hatchery would
be far higher than they are.

When working under normal conditions, solely with a view of obtaining the greatest
possible amount of living young, and without regard to what quantity may be lost,
the mortality i)ercentage at Flodevigen amounts annually to something like 12 per cent
for the entire period of hatching, including unfertilized eggs.

The figures for the three last years are as follows : —

Per cent.

1912 12-5

litis 13-S

1914 11-0

48 DEPARTMENT OF THE NAVAL SERVICE

These percentages are calculated from very large quantities, in 1913 for instance,
1,222 litres, about 550 millions of eggs, and they do not vary greatly from year to year.
When much brackish water occurs on the coast, the mortality may inci*ease considerably ;
this is, however, rarely of long duration. Similarly, if the weather for any length of
time is dark and cloudy, the eggs may become entirely overgrown with bacteria, from
which, however, they may be freed by being placed in a bath with water of high salinity.
As to the degree of influence exerted by these bacteria (Leptothrix?) in a natural
state I can say nothing from experience; possibly the conditions at the hatchery may
be especially favourable to them.

As regards the growing up of the young, we have here to consider another factor,
viz., that of nourishment. That this is of great importance to the growth of the
young is certain, and possibly the most critical period of all in this respect is the
time when the young first commence to take food; they do not live long without
nourishment after the yolk-sac has been consumed.

In the spring of 1914:, a number of newly hatched young were placed under
observation at Flodevigen for the purpose of ascertaining when they commenced to
take food, and what such then consisted of. The investigations were carried out by
Prof. H. H. Gran, and led to the result that the young did not take any food mitil
the yolk-sac had been absorbed, and that on commencing to feed, they appeared from
the first to prefer animal matter, such as mollusc larvse, nauplii, etc., seeming, strangely
enough, to despise the innumerable diatom forms which are likewise present in the
water.

Such food as was pumped in with the sea water was alone available to the young
fish; but as the greater part of the plankton was retained in the apparatus, the
nutritive matter was fairly concentrated there.

As to the influence exerted by conditions of nourishment upon the growing of the
young in a natural state, this question has not yet been investigated, and it will
probably form one of the earliest subjects for future fishery investigations.

It is, we may say, of prime importance in every way to ascertain what factor or
factors normally hinder the germs of our most important fish in their development,
to discover at what stage the destr\iction occurs, and by what means it takes place.

This will have to be done by investigation of the propagation of fish in the sea,
and by experiments and artificial culture, both branches of the work proceeding in
conjunction.

It is from such a point of view that I have in the foregoing pages endeavoured to
treat the material of fish eggs and young collected by the Canadian expedition. Unfor-
tunately, however, we still know far too little of these questions, and my work can
therefore only be regarded as furnishing, in this respect, a preliminary survey of the
complicated, yet highly interesting, conditions which prevail on the Atlantic coast
of Canada.

LITERATURE.
Alexander Agassiz.

1878. I. On the Young Stages of Bony Fishes.

1882. III. On the Young Stages of some Osseous Fishes.

Jordan and (xilbert.

1882. Synopsis Fishes of North America.

E. W. L. Holt.

1891. On the Eggs and Larvae of Teleosteans.

1893. On the Eggs and Larvae and Post-larval Stages of Teleosteans.

(The Scientific Transactions of the Royal Dublin Society, vols. IV
and V.)

CANADIAN FISHERIES EXPEDITION, lOUflS 49

W. Lilljeborg.

1891. Sveriges och Norgea Fiskar I-III.

Jordan and Evermann.

r396-1900. The Fishes of North and Middle America. Vols. I-IV.

C. G. Joh. Petersen.

1904. On the Larval and Post-larval Stages of the Long Rough Dab and the

Genus Pleuronecies.
(Meddelelser fra Kommissionen for Havundersogelser, Bd. 1, Xo. 1)

Johs. Schmidt.

1905. De Atlantiske torskearters pelagiske yngel i de post larvale stadier (Med-

delelser fra Kommissionen for Havundersogelser, Bd. 1, Xo. 4).

1906. The Pelagic Post-larval Stages of the Atlantic species of Gadus (Meddelel-

ser, Br. II No. 2).

E, Ehrenbaum.

1905-09. Eier und Larven von Fischen
(Nordisches Plankton).

A. Brauer.

1906. Die Tiefsee-Fische

(Valdivia Expedition Bd. XV).

Prof. Apstein.

• 1911. Die Verbreitung der pelagischen Fiseheier und Larven in der Beltsee und
den angrenzenden Meeresteilen 1908-09.

David Starr Jordan.

1914. A manual of the Vertebrate Animals of the Northern United States.

6551—4

CANADIAN FISHERIES EXrEDlTION, tOUrlS

Station 2—

20-0 ....

60-25- ..
Station 3-

Snrface .

80-0...

190-60 ...

Station 4 —

30-0 ....

80-40....

150. (

Surface . .

Station 8 —

Surface.

Surface

Station 11—

Surface

Station 12-
Surface . . .
100 0.. ..

Station 13—

Station 15 —

100-0 ."-
Statiun 16—

Station 17—
Surface

Station 19-
Statior 'Ay-
Station 21 —

Station 22—

Surface
Station 23 -

Surface . . .

70-0

Station 34—

Surface . . . .

100-0

Station 25—

Surface

120-0

Station 26—

Surface

100-0.. ..
Station 27-

Surface. . ,
Station 28-

Surface . .

100-25....
Station 29—

Surface . . .

65-0

Station 30—

Surface . . .

Vertical . .
Station 31—

Surface . .

Vertical..
Station 32-

H)0-25 .
Station 33-

Station 34—

100-0. . ;

Station 35—
Surface , .
125-25...

Station 36-
Surface. ,
100-15.-.

* Sample in bad condit

8 except Acadia Sta. 44-Cl.

I.

o...

f
1

s

1

1

1

1

ll

i

a

0„..„.

s
3

t:

I

I

1

1^

1
1

1

station 3-

Jnu a...

■

1
7

Km

2

233
24

7<«
79

3
93

31

10

7

890
60

1

!?
7

si

10

^

Bug
20

162
27

lii

sSi4-

r™. 9...

•

stS'irfc

Suriac*

J»™ 10..

''B'^- ::::■■:

June 10..

Urn 16: ■

"J.

''■

Swtion 8

June .6..

'

Station S-

'"".";

10
2

18

.18
2

3

9

1

369
S

1
■ 2

1

Station 10—
Surface

Station 12—

Jme 11.

Station 13—

June 11,

Station 14-

June 11.

Surface

Sution 16—

June 12.
June 1«:

100-0

Station 17—

June 12;

Surface

June 13.

Station 19-

June is;
June 13:

St*tioD 21-

Stntion 23-

.i'une 13,

1

7

60

17J

June 15,

%

I

^

Station 2I>—

June 15.

i:^

::_

Surface

■'"■!':'';

—

i

1

i 1

■(U«l|iiii«iiuo 1"

■t.«, miiuffiij 1 " 1 "

-.I-.ni.Kloas-l - 1^- ' "■=" _ (3

longdoiaitiv | ■ - - 1

-tuiiiXl^j, 1 ':''■-: • 1

■tlmt«B*l.i|toov 1 ■ : : ■■::-: |

■j-[ii wu-io'i 1 : : : ; :" 1 ; : ■ ■ 1 -

—SJ 1 i ;;::"■:;:- 11 ^^ " " : i S^^' ii S S^ .|i

•muojKwul .iiiq.lo'i 1 ; : ::.::::.:::.;.;. '■'■'_': \ 1"

1

:::::::::;:::: ■ . : : " 1 - 1

.,.,.,«,on.nr, 1

a i : ; ; : : ; : : j i : i::; i :;:::: : : : : ; : Sis'

..l..,.«.,., « : ,-

,pu.u,,,.,,, ; :;:::: :;;:;::;: : :i : : .: ;::::::- ^ ::: W [ - '■ '^ ■ ' ■ ■ : " : : [: i ip

,,„„.l«.»,.„ •,,, - - . : 1 -

■Knnio|3auXa ■(,! :::::: : ::::::::: . . : : : "^ : : . : ' ■ 1 *

-dBi-o]swiiojiiB|a ."::': : ; : : : : : : : ^ ■ 1 "

,™9«,.™ ■»■: »::. -:. - • .= __=.; ,;,«

■.™.-„ ■ . :;:.:: : -i o, s .: g.. .- ; -: ^ ; g

*.„,»„ 1 ; ; ; ; : : ^ i - : : : ; i ; ° - ^ ::;:; = §-::::- ; ; : - i n : : «i S= r = : ; : ; : | ; S

: :::■:: :::■-;: ; . : : : : : : ; : ::::■::■:;:: : . ■ : ::.:::: ; : :

*"""""

s : ii: -i ^ " °- ^ ; "^ :;:■;; S " ;; -: i i:^ ; ; : ;: ;; -: - : :; : Ml

,

.-:: t~ ■ n ■ ^ ^ ■ ■ .:[S

~SZi£a -

1 i ■: -i: : 1 '- ::" U - : : ; | |: ; ;|S

■Hiilv)ouiul . " ; : : : . : ; :::::: : , *^ . ; l "*

■..,»,„. .».„,ms| : ■ M ;:■;■!•:; ^ • ^ ^ : ■ : : :::■:.:;;::;:; : : : : i ■ \ i ■. \:\: :;■:":-:: : . ; : : ; i =

i -, : S i S ■ ■?! ■ ?j ?l •?! ?i ■■■?!••■?!■ S :■ S S ' i Is i SS ' ' 3 S ! .3 S ?, '7. ?; . S Vi tl 8 ' Si i -SSi : ■« S 8 ^ 8 si i sis : SS ^ 8 ' ^ ft R «« ft « i si ^S ^ ^ ^S ' ^ S S :s S i S i S 'S

= 1: r- - - ^ : ?"f |: f li^ J II III II^MI 1 -| ^ ^l f: 1- f ■ " f 1 1 ^ i 2 i:' i ' l^ ^ i?

<

:

ijiit|ilii|Hiiiii

CANADIAN FI8BBRIES EXPEDITION, 19J4-1S

66 Hie 303 47

DEPAUTMENT OF THE KAVAt SpRVlCB

i

1

1

i

i

1

:

Steam Drifter

i

Si

'1

■■N°33".

Dnte.

;

fS

,2

1

•s.

H

OnoH Hp.

i

s-

S

I'

i*

£

1*

3

1

1

1

i

1

:

i

1

<

tj

6

b

i3

o

a

Jt

J

-

C^

Year.

Egg

EgB

%8

1^1

Emr

Kmi

Station 3—

Junel.,

June 2.

ID

60

Surface

.......

3

4

127

"urface

g

Station 9—

'

^.fo

Surface

Station 12—

June 10.

'

*

87

2

Station 16-

June 10,

'

"

350

520

'

''

■i

m

130

'

4

'

«»

■A

Station 22—

June 25.

420

Sution 25-

Jun»»26,

Suriaw

Station 35-

July? .

July 8. .

27

.j

no

»

StatioD 49-

July"27!

^

3

'

1

"

Statioii tm-

Aug. 10.

^

2li

'

Station S9-

Station 70-

->

207

-

"

'

'

'

31

m

2694

'

4479

'

7

6

337

'

4046

CANADIAN FISHERIES EXPEDITION, lOl-i-lo 57

Table II. — vSummary of the six most common species of fish. (All the cruises included.)

(xerminative
disk.

Surface.

Vertical hauls

.'

Pigmented
embryo .

Larvae.

(Jerminative
disk.

70

774

46

1773

208
3252

Pigmented
embryo.

74
392

45
520
116
300

Larvae.

245
2178
1118
10167
1323
4496

61
2231
1160
3528
1328
773

237
16
30

276
J9

81

Scomber

Drepanopsetta

31
17

Gadus

Onos

• 17
53

19527

9081

598

6123

14

199

* Here also included some de<^{) horizontal towings by ss. No. 33.

T.^BLE II a (1). — Egg and Larvse Ctenolahrus adspersus.

Date.

Germinative
disk.

Surface.

Pigmented
embryo.

Larvae.

Ve

Germinative
disk.

5

rtical hauls.

Pigmented
embryo.

Larvae.

"Acadia " I

May 29 to June 4..
•June 9 to June 15.
July 21 to July 2y.
Aug. 3 to Aug. 12.
June 1 to Aug. 18.

1
6

2
15

1

"Princess" I
Stati(»n 19-26

Acadia " II

21
18
26

37
11
26

1

'Princess" II
Station 27-50

"No. 33'
Station 5-48 .

4
234

33
11

237

79

245

61

237

70

74

81

(5551—8

58

DEPARTMENT OF TEE NAVAL SERVICE
Table II a (2). — Egg and larvae, Ctenolahrus adspersus.

—

Surface.

Vertical hauls.

"Acadia" I.
Station 5

Gernxinative
disk.

Pigmented
embryo.

Larvae.

Germinative
disk.

Pigmented
embryo.

Larvae.

1

2
3

1

II 6

1

II 8

1

1

2

5

1

"Princess" I.
Station 19

6

1

15

•1 26

6

15

"Acadia" II.
Station 91

21

37

1

"Princess "II.

Station 27 - . . .

28....

1, 29...

3

1

14
10

8

1

2""

33
23

17
1

8
3

53

5

•1 30 . . .

1

31

2

1. 48

75
54
50

II 49 .

50 ..

18

4

33

237

18

11

79

No. "33"—
Station 5

125

80

26

3

2
2
3

2

1

Vertical and deep horizontal.

II 6

II 7

II 8

11 10

II 35.

12
14

II 39 .

26

II 48

1

234

11

26

26

CA^'ADIA^Sl FI!?nERIES EXPEDITION, 19Vrl5
Table II a (3). — Larvae, Ctenolahrus adspersus.

59

Length in inm.

2

3

4
1

5

6

1

8

9

10

11

12

13

14

15

16

17

More th&n
17.

" Acadia" I

" Acadia "II
Station 91.

1

" Princess " II
Stat ion 28

55
38
24

,1 29

80 ....

31

1

1
75
54

68

.1 48

49

1. 50

Table lib (1). — Egg and Larvae, Scomber scomhrus.

Date.

Surface.

Vertical hauls.

Genninative
disk.

Pigmented
embryo.

Larvae.

Genninative
disk.

Pigmented
embryo.

Larvs.

"Acadia" I
Station 3-9..

May 29 to .June 4 .
June 9 to June 16.
Aug. 8 to Aug. 12.
June 1 to Aug. 18

3

263

80

1,832

2,178

" Princess " I
Station 24-26.

55

10

709

774

"Princess" II
Station 28-50.

"33"
Station 23-49.

204
2,027
2,231

16
16

14
378
392

31
31

6551—8^

^

DEPARTMENT OF THE NAVAL SERVICE
Table lib (2). — Egg and Larv«, Scomber scombrus.

Surface.

Vertical haul.

t' .

(rerrainative
disk.

Pigmented
embryo.

Larvfe.

Germinative
disk.

Pigmented
embryo.

Larvae.

" Acadia " I. —

Station 3

„ 9.

2

1 (?)

" Princess' I. —

Station 24

3

162

21
80

27

23

5

■ 11 25

„ ■ 26

263

2

1
75

55

'Princess" TI.—

Station 28

29

3"

171

30

1
1

,1 30

14

9
5

3

31

32

io

6
5

33

9

1. 34

1

„ 45

1

48

49

11 50

1

2

4

No. "33"-

'= Station 23

29

.1 35

HO

3
1.800

204

3
2,000

16

10

52:)

144

40

709

14

312
06

378

31

36

„ ■ 39

'4'
25

48

24

2,027

...; 49

.,. i

1,832

CANADIAX flf>HERlES E.XPEDITIOy, Wl'rlo

61

Table lib (3). — Lar\-£e, Scomber scomhrus.

Length in n.m.

2

3

4

5

6

1

7
....

8

9

10

11

12

13

14

15

16

17

More
than 17.

Princess II.—
Station 28

b
2
3
2

i'

3
5

29.. ..
30.. . .
31 ... .
32

3

t

1
1

1

1

« •

33

34 . ..

1

45

1

49

2

2

50...

2

Table lie (1). — Sehastes marinus.

Surface.

Vertical Hauk.

Date.

Larvae.

Larvte.

"Acadia " I—

Station 11-36

May 29 to June 4

56 (56)

19 (19)

" PrinceiS I —

Station 9 23

June 9 to June 15

6 ( 6)

102 93)

" Acadia" II —

Station 37-90

July 21 to July 29

73 (73)

155 (155)

" Princess " II —

Station 32-48

Aug. 3 to Aug. 12

50 (50)

170 (165)

No. "33"—

Station 19-58

June 1 to Aug. 18.

337 (37)

522 (222)

446 (43^

Numl^eis mea-sured.

62

DEPARTMENT OF THE NAVAL SERVICE

Table lie (2). — Sehastes marinus.

Surface.

Vertica

hauls.

Total number.

Midle length. 1

Total number.

Midle length.

" Acadia " I

Stat

ion 11

, 26

10

7-8

3

1
5
3
7

8.0

, 31

80

, 32

43

1
1
1

8-7
80

8-0
70

80

, 34

, 35

, 36

7-3
8

56

19

Stat

on 9

- 10

, 16

, 20

, 21

1
1

4

80
9

"" 7-8

34 (25)

7 ( 7)
.33 (33)
10 (10)
18 (18)

8-3

8.3
8 4
8 1

. ?>3

8 1

" Acadia " II —

Station 37

6

7
2

7 4
6 5

102 (93)

52
22
29

3

3

1

3
10
15

9

1

2
4
1

, 38

, 39

7-3
7-5

, 49

, 50

, 51

79

1
40

90

7-6

87

, 52

7.0

, 59

8

, 60

, 63

67

6-5

64

. 65

, 67

, 69

2

6-5
5-3
80

. 70

, 72

- 76

, 84

1
5
5

70
7-6

8-2

7.5

, 86»

10 5

- 88

, 90

?

73

6 3

8

80

155

Stat

on 32 .

1 ( 1)

21 (21)

14 (14)

8 ( 8)

1 ( 1)

11 ( 6)

4 ( 4)
36 (36)

1 ( 1)
24 (24)

2 [ 2)
40 (40)

5 ( 5)
2 ( 2)

90

, 34

9 6

.: 35 :: :::

11-7

„ 36

8.8

, 37

9

, 38

7

, 39

80

, 41

10

„ 42

100

, 43

10 3

, 44

8-5

, 45

38
8
3

1

8-9

&-6

13 3

80

87

. 46

, 47

7-8
120

. 48

50

• 170 (165)

^O. Ai —
Stat

ion 19

< 22

2 ( 2)
1 { 1)
5 ( 5)

325 (25)

3 ( 3)
1(1)

70
80

8
80

9 7
80

, 23

, 26

, 48

" ^'~'

337 (37)

CANADIAN FISHERIES EXPEDITION, 19U-15
Table II d. — Chirolophis sp.

63

Sur

'ace.

Vertical.

Total number.

Length in mm.

Total number.

Length in mm.

"Acadia I " -

Station 23

„ 34

1
5
2

10
ca. 10
ca. 10

„ 35

"Princess " I —

Station 11

8

16

3

1
2

ca. 12

8-10

10

ca. 8

2

,. 13

„ 18

- 24

„ 25

10-11

22

2

No. "33"-

6
1
3

2

ca. 12

12

11-13

ca. 8

18

1

„ 16

„ 17

„ 22

.. 39

8-10

„ 49

ca, 12

12

19

Table II e. — Liparis sp.

Surface.

Vertical hauls.

Total number.

Length in mm.

Total number.

Length in mm.

" Acadia " I —

Station 8

1

4

1

3

„ 24

1

1

" Princess " I —

1

5

1

„ 24

5

"Acadia" II —

Station 81

6
1

4-7

" Princess " II—

Station 36

10

1
1

9

„ 50

8

1
1

2

No. "33"—

Station 17

4

2

^ 1

„ 39

4-3

.. 57

25-30

■64 DEPARTMENT OF TEE NATAL SERVICE

Table II f (1). — Egg and Larvae, Drepanopsetta platessoides.

Date.

Surface,

Vertical hauls.

Germinative
disk.

Pigmented
embryo.

Larvae.

5

23

2

Germinative
disk.

10
29

7

Pigmented
embryo.

23

15
3

1
3

Larvae.

" Acadia " I

Station 8-36 .
"Princess ' I

Station 3-26,
" Acadia " II

Station 49-83
" Princess ■■ If

Station 30-50
"33"

Station 3-17.

May 29 to June 4..
June 6 to June 15
July 21 to July 29.
Aug. 3 to Aug. 12.
June 1 to \ug. 18.

322
260

7

529

734

3C5
25

96

7

5
5

1118

1160

30

46

45

17

Table II f (2). — Egg and Larvae, Drepanopsetta platessoides.

"Ac

St

-

Surface.

Vertical haul.

Gerniinative
disk.

Piginente.1
embryo.

Larvae.

Germinative
disk.

Pigmented
embryo.

Larvae.

adia" I

6

3

! 19

2

2

200

1
3
6

18
2
7

18

20..
21..
22..
23..
24..
25..
29..
30..
31 .
1 32

5
415

2
13
11
114
16
17

5
33
55
24
18

1
2

4

1

12

3

4

7

3

33..
34..

57
2
4

tl w. .

322

731

10

23

7

"Princess" I
Station 5..

3 "

1

7
2

1

2 ""

1
1

T

5 ""

1
2

': i^

3

10 .
11..

12..
15..
1 16

3

11

2

22

15
7
1

25

5

9

180

25

21
1

2

3

1 17

1
2
3
4

1
9

2 ""

6
1
1

, 18..
19..
24..
25

189

20

6

•>fi

1

*.v^ . .

200

305

g

26

15

CA^'ADIA^' FISHERIEfi EXPEDlTIOy, lOL'rJS
Table II f (2). — Egg and Larvae, Drepanopsetta platessoidcs — Con.

65

Surface.

Vertit-al haul.

Genninative

disk.

Pigmented

embryo.

Larva?.

Germinative
disk.

Pigmented
embryo.

Larvw.

Acadia" II
Station 49. .

1

81..

1
2
4

3

9

13

7

1
2

82..

3
1

83. .

16

7

25

23

3

-

" Princess" II
Station 36. .

1

37..

\

39

1
2
1

45..

50..

2

2

1

5

No. "33"
Station 3. .

6

3

4..

2
2
9

7
2
9

1

1

10

8
1

5..

■

6..

7..

8..

9..

1
2
1

10..

1

11..

12 .

13..

1

40 ""

35

7

14..

8

310

95

73

15..

16..

17..

529

96

(

3

66 DEPARTMENT OF THE NAVAL SERVICE

Table II f (3). — ^Larvae, Drepanopsetta platessoides.

Length in mm.

2

3

4

5

2

6

2
1

7
....

8

9

10

11

12

13

14

15

16

17

More than
17.

"Acadia" I —
Station 8

„ 25

"Acadia" II —
Station 49

1

i

81

6

82

3

"7"

83

1

1

"Princess" I —

Station 8

6

"Princess" II —
Station 37

39

1

45

2(20 mm.)

60

2

1

Table II g (1). — Egg and Larvae, Gadus sp.

Date.

Surface.

Vertical hauls.

Egg.

Larvae.

Egg.

Lar

vie.

S .

s.-l

C5

6

1

1
§

.a

a

s
'"5

t

1

5s

>

a ■
■"-«

d

tu
S
bo

.a

1
a

1
1

a

s

!

a

05

"Acadia" I —

Station 5-3G..

May 29 to June 4...

1601

2025

59

31

1

116

158

5

47

3

1

"Princess" I —

Station 3-26. .

June 9 to -Tune 15. . .

3940

743

9

6

375

56

3

1

" Acadia" II —

Station 49-84..
" Princess" II —

Jiily 21 to July 29...

284

187

43

28

250

14

23

10

1

10

1

Station 27-50. .
";-13"—

Aug. 3 to Aug. 12...

310

152

19

4

1081

188

14

2

Station 4-58..

June 1 to Aug. 18. . .

4032

232

1

178

31

6

10167

3339

130

59

262

14

1773

413

2K

49

15

2

CAXADIAX FI.'^HERIES EXPEDITION, 191/rl5

67

Table Jiff (2). — Egg and Larvae, Gadus sp.

i

Surface.

Vertical hauli*.

Egg- 1

Larvae.

Kgg- j

Larvae.

"

s'-S

'at

T3 d

It

i

1
1

6

C3

.2

-J

s >

u •—
1) -^

is

6

i

6

is

"Acadia" I.—

Station 5

6

7

8

61
94

146

1

55

47

850
70
60
4
20
86
60
19

4

1,250

137

1

"15

ca. 10(

11

' i

)dama

ged eg|

12

66

's, prol

23

1

1

40
6

2

1

••ii7:;:;i

)ably haddock

24 1 3
1

1

1

1. 9

„ 19

76

32

440

1

1

13
42

5

1

19
2

2

""44

1

3
2

„ 20

„ 21

„ 22

.. 23

„ 24

II 25

6

6

1

,. 29

30

7

T
1

116

12
2

1

2

. 31

„ 32

„ 33

32
20
26

„ 34

n 36

".Princess" I. —
No. 33.—
Station 4

1,601

2,025

59

31

1

158

5

47

3

1

60

4

1,700

600

150

100

153

150

30

7

83

500

115

124

73

21

33

3

1

1
2
4

20

35
6

17
5

27

60

1

9

108

2

3
13
13

5

1

5

6

7

8

9

10

„ 11

„ 12

„ 13

„ 14

„ 15

1, 16

.. 17

„ 19

M 21

,- 22

„ 26

36

„ 39

„ 48

2
116

96

1

n 49

57

„ 58

n

15

4,032

4

232

1

178

31

6

"Acadia" II. —

Station 49

11 59

5

3

1

61

62

63

66

M 68

90

105

1

20
14

1
2

1

6
1
1
2

"ii

9
3

1

1

3

5

1

10

80

81

82

II 83

6
37
40

1

13

65

71

4

3
11
13

8

' 11

17

12
20

3
206

9

84

1 284

187

43

28

250

14

2.T

10

1

in

1

68

DEPARTMENT OF THE' NAVAL SERVICE

Table II g (2). — Egg and Larvae, Gadus sp. — Con.

'Princess " II.
Station 30.
31.
32.
36
37.
40.
42.
45.
48
49.
50.

" Princess " I.
Station 5.

6.

7.

8.

9.
10,
11.
12.
15.
16.
17.
18.
19.
20.
22.
24.
25.
26.

Surface.

Egg.

1

148

48
113

310

45

152

206

618

69

" 49

29

9

5

34

l,07v

855

750

51
37

15^

3,940

26

82

10

1

44

3

1

18

15

54

307

135

9
21
17

743

19

C5

Larvae.

Vertical hauls.

Egg.

c-c

24

124

687

8'

41

104

2

1,081

c 0^

188

14

Larvae.

1

1

22

2

48

7

19

2

1

3

2

1

8

5

2

4

5

1

1

1

9

114

12

1

29

5

47

3

1

4

5

30

3

3!'

4

375

56

3

1

"Acadia" 7 —

Station 3-29..
"Princes.s " I--

Station (U17..
"Acadia" II —

Station 47-62
"Princess "II —

Station 27-50
" 33 "

Station 19-39

CANADFAN FIf?HERIEH EXf'EDlTlON, 1911,-15
Table lie: (3). — Larva?, Gadus callarias.

69

Le

ngth in mm.

2

3

1

4

5

6

7

8

9

10

11

12

13

14

15

15

17

More than
17.

" Acadia " I —
Station 6 . . . .

„ 8

1
1

2
2

3

1

25

" Princess " I—
Station 8

M 20 ... .

4
20

1

1
1

1

" Acadia " II—
Station 80

4

81

82 ... .
83 ... .

2

2i6'

n 84

'9

1

2

" Princess" II —

1

1

1

"33"
Station 57

1(21 mm)

1

Tabiii: II g (4). — Larvae, G. aeglefinus.

Length

in mm.

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

More than
17.

"Acadia " I—
Station fi

1

Acadia" II —

i

1

2

59

1

68

.. 81

3
1
1

3

82

83

S4

2

I'abi.k nil (1). — Egg and Larvi^, Onos.

Surface.

V

ertical hau

s.

Date.

Germ-
inative
disk.

Pigmen-
ted
embryo.

Larv«.

Germ-

inative

di.sk.

Pigmen-
ted
embryo.

Larvae.

Ma^ 29 to .Tune 4..
June 9 to June 15
July 21 to July 29.
.Aug. 3 to Aug. 12.
June 1 to Aug. 18.

5

1.")

15S

23

1122

1

2

35

21

1269

1

11

2

194

2

5
109

1

35
4

52

1323

1328

39

208

IKi

53

70

DEPARTMENT OF THE NATAL SERTICE
Table II h (2). — Egg and Larvse, Onos.

Surface,

Vertical hauls

Gerrainative
disk.

1
3

Pigmented
embryo.

Larvae.

Gerniinative
disk.

Pigmented
embryo .

Larvae.

" Acadi
Sta

a" I-

.ion 3

5

6

1

19

29

1

1
1

5

1

" Princ

Sta

888 " I —

bion fi

8

9

10

10
4

1

1

1

1

1

16

, 18

2

i5

1

2

" Acadi
Sta

a " II-

tion 47

, 49

20
10

20

10

1

56

61

120

8

15

62

158

35

11

" Princ

Sta

BSS" II —

tion 27

4

15

4

4

8
7
2

1
1

2

2

28

35

29

30

25

5

8

1

31

3

32 .

1

39

1

49

1
4

1

, 50

3

23

21

35

2

5

52

No. "2
Sta

3"—

tion 19

21

22

800

3

9

220

15

75

1000

3

31

200

10

25

110
84

108
1

23

25

26 . .

29

4

35

39

1122

1269

4

194

109

CATfADIAN FISHERIES EXPEDITION, 19U-15
Table II h (3). — Larvae, Onos.

71

Length in mm.

2

3

1

3
3
6

i'

4

5

6

7

8

9

10

11

12

13

14

15

16

17

More
than 17.

"Acadia" I —
Station 6

"Piincess" II—
Station 28

5
3

29

30

31

3
1

1

7

4

32

49

50

No. "33"—
Station 29

Table II i. — Egg and Larvae, Merluccius merluccius.

Surface.

Vertical haul

Germinative
disk.

Pigmented
embryo.

Larvfe.

Germinative
disk.

Pigmented
embryo.

LarviB.

' Acadia " II —
Station 44

2

7

2650

176

1

"266"'

750

400

300

10

1^
1400

1

50

1800

100
200

47

53

350

12

1

5

250

85

70

54

55

56

58

59

60

61

62

4496

773

3252

300

72

DEPARTMENT OF THE N iVAL FiERVIC^
Table ITj. — Ammodytes tohianus.

Surface.

Vertical hauls.

Total number.

Length in nun.

Total number.

Length in mm.

" Acadia" I—

2

5
1
7

9
3

8
1
2

6
1

4

ca 15

„ 5

1

ca 25

ca 15

M 8 . .

15-22

19

21

5
4
2

15-20
ca 1!")
12-15

- 22

23

6-10

24

25

26

31

17
3

10-18

io-15"""

8-14

10-13

10

12-20

32

7

8-22

.1 34

ca 15

36

5

12-17

12

46

43

"Princess " I —
Station 9

1
2

1

15

7-7
18

15

16

4

" Acadia " II —
Station 82

1
3

15
12-20

1

16

„ 83

4

1

No. " 33 "—

2
4
2

10-20
ca 8
12-13

— —

M 17

21

8

CA\.\nr\\' /7N///; /.'//•>• E\f'i:niTio\, wj',-i5

73

Tari.k Ilk. —^fallofllf! villosns.

Surface.

Vertical liauls.

Total number.

1
12

ca 75

Lenprth in mm.

1
Total number. '

Length in niiii.

"Acadia" I—

Station 30 '. . . .

"Princess" I —

Station 4

"Acadia" II —

Station P3

"Princess" II —

Station 34

1
42

10-15
7-25

1

1

15

1

1

ca 325

15
12
22

M 41

„ 42

16

12

,■ 45

7-25

ca 75

ca 342

No. "33"-

Station 17

3
3

ca 8

» 58 ..

6

Table II 1. — Scopelus sp.

'Acadia" I —

Station 16
.. 17
'Acadia" II —

Station 3^i
„ 42
.. 44
., 45
M 46

M 51

.. 74
M 75

37
4
2
1
1
5
1

56

8-11

C-9

14-16

5-6

8

6

6-12

9

13
3

6
33

7-14
6-10

7
8

6-16
S-21

6551—9

74

DEPARTMENT OF THE 2s'A7AL SERVICE

EXPLANATION OF PLATES

Plate I.

Fig.

Ctenoluhrus adspersus, 4'2 mm., Princess Station, 50 (surface).

Ctenolabrus adspersuSj 8 mm., Princess Station, 50 (surface).

Scomber scombrus, 6"2 mm.. Princess Station, 33 (obi. haul).

Sebastes marimis, 5'6 mm., Kristianiafjorden July 1913 (taken from the adult fish).

a and b, Sebastes murinus 8 mm., No. " 33 " Station, 26.

a and b. Sebastes viarinus, 10*4 mm., Princess Station, 46 (surface).

Sebastes niai'inns, 20 mm., PiHncess Station 41 (30-0).

Plate II.

Fig. 8. Chirolophis ?, 11mm., Princess Station 13 (100-0).

9. Stichatts punctatus, 30 mm., Acadia Station, 89 (surface).

" 10. Cryptacanthodes nChculatus, 38 mm.. Princess Station, 8 (80-0)

" 11. Pleuroncctes cynoglossus, 16 mm., details lost.

" 12. Ancylopsetta sp., 7 mm., Acadia Station, 44 (surface).

" 13. Drepanopsetta platessoides, Acadia Station, 25 (120-0).

" 14. Drepanopsetta platessoides, 7 mm., Acadia Station, 82 (30-0 fms).

" 15. Drepanopsetta platessoides, 13 mm., Acadia Station, 81 (surface).

Plate III.

Fig. 16. Gadus callarias, 3*8 mm.. The Flodevig- Sea Fish Hatchery.

" 17. Gadus callarias, 4'5 mm., Acadia Station, 8 (vertical).

" IS. Gadus callarias, 11 mm.. Princess Station, 20 (surface).

" 19. Gadus aeglefinus, 3"7 mm., Acadia Station, 81 (surface).

" 20. Gadus aeglefijius, 7 mm., Acadia Station, 59 (25-0 fms.)

" 21. Onos cimbrius (?), 4*4 m.. Princess Station, 50 (40-0).

" 22. Onos cimbrius (?), 4*6 mm., Princess Station, 31 (oblique).

" 23. Ammodytes tobianus, 7*2 mm., Acadia Station, 23 (70-0).

" 24. Ammodytes tobianus, ca., 15 m., Acadia Station, 82 (30-0 fms.)

" 25. Tetragonurus cuvieri, 76 mm., Acadia Station, 56 (210-140 fms.)

" 26. Mallotus villosus, 8 mm.. Princess Station, 45 (oblique).

" 27. Mallotus villosus, 19 mm., Princess Station, 45 (130-0).

The figures, except those of macroscopical fishes, are drawn from specimens
imbedded in glycerine-gelatine, transferred 'directly from formaline 4 per cent with-
out being cleared or stained. The originals of figs. 3, 4, and 10, however, are borax-
stained specimens mounted in Canada balsam.

All drawings from original material by Mr. Thorolv Rasmussen, draughtsman to
the Fishery Board of Norway.

All depths in meters, when not otherwise stated.

PLATE I.

CANADIAN FISH EGGS AND LARV^. DANNEVIG.

^.

'-^.

5.6 m.fn.

5a

•

#

6b

>-

da

^^

^

5?

10.*t tM.m

2o fTi.fn

PLATE II.

CANADIAN FISH EGGS AND LARV^. DANNEVIG.

.././.

■ t:

.y~^- '. '"7 'X'

Xf ■ • •■■ ., ' .^ . . ,, ■ . ,

15

^

;::^

I3m."i-

PLATE III.

38 m

CANADIAN FISH EGGS AND LARV^. DANNEVIG.
16 19

3 7 Ml

18

26

^

CANADIAN FISHERIES' EXPEDITION, 1914-15

BIOLOGY OF ATLANTIC WATERS OF CAiNADA

REPORT ON "AGE AND GROWTH OF THE
HERRING IN CANADIAN WATERS"

BY

EINAR LEA,

Bergen, Norway

6551—9.1

AGE \XD GROWTH OF HERRING IN CANADIAN WATERS
BY EINAR LEA, BERGEN, NORWAY.

CONTENTS.

Page.

"A " (Jknkual.

81
I Introduction

II Methods of determining age and growth in herring 81

III Structure of herring scales

IV Scales as an indication of age. Sources of error 9:3

V The scales as an indication of growth. Sources of error 100

VI Age determination and growth measurement in practice 101

VII Preparation of scales for preservation 101

VIII Definitions and abbreviations lO"!

IX Collecting of material

"B" Spkcial.

The Canadian Material.

X Description and preliminary grouping of material 115

XI Age (a) Waters about Prince Edward Island 116

(&) Waters about Magdalen Islands 1^1

(c) Newfoundland waters 122

(d) Cape North southwards along Atlantic Coast 126

(e) Comparison of the different areas 130

XII Growth.

Comparison of growth in samples of similar age composition . . . . 132

1. Samples from Newfoundland 132

2. " " Magdalen Islands 13"

3. " " Northumberland Straits 139

4. " " North Sydney compared with exceptional

sample from St. George's Bay, Nfd.. . 143

5. Samples differing in age composition 146

6. Comparison with Magdalen Islands samples 147

7. Comparison with exceptional sample from St. George's Bay. 148

8. Comparison with sample from North Sydney 149

9. Comparison with old fish from West Ardoise, etc 150

10. Graphical illustrations of results obtained 150

11. Samples from Magdalen Islands compared with remainder. 153

12. Graphical illustration of the results obtained 155

13. Samples from Northumberland Straits, North Sydney, and

the Atlantic coast compared 156

14. Summary 1"^ '

XII Appendix. Observations as to seasonal growtJi in young herring. . . . 157

XIV Results of age and growth studies viewed in the light of observations

as to the racial characters 1"*9

77

LIST OF PLATES AND TEXT FIGURES.

Plate I.

Fig. 1. Herring' scale on a dark background: the winter rings appear darker than
the summer zones. xl4. (Lea phot.)

Plate IT.

Fig. 2. Herring scale on light background: the winter rings appear brighter than the
the summer zones. xl4. (Lea phot.)

Plate HI.

Fig. 3. Left half: Impression in a collodion film of surface of portion of a herring
scale. Right half: Same portion of the actual scale itself. xl5.

Fig. 4. Scale split, the detached part showing a contour similar to that of the whole
scale; pattern formed by fibrils on the surface of distension. x9.

(Lea phot.)

Plate IV.

Fig. 6. Part of herring scale, the outer covering removed, showing margins of
fibrillar plates with alternating tangential and radial course of the
fibrils. x75. (Lea phot.)

Fig. S. Part of cross-section through a herring scale, stained with thionine. x380.

(Lea phot.)

Fig. 13. Effect of silver impregnation acting from the outer surface of a scale. xl70.

(Lea phot.)
Plvi'eV.

Fig. 14. Total views of a herring scale, impregnated with silver, demonstrating the
inner system of annual zones. Oblique illumination. xl3.

Fig. 15. Half of a scale with normal structure. xl7.

Fig. 16. Half of a scale with abnormal structure from the same herring as the scale
shown in Fig. 15. xl7. (Lea phot.)

Plate VL

Fig. 17. Part of an abnormal scale, the outer covering layer of which is removed,
showing irregular course of fibrils. xl()7. Compare with fig. 6, Plate IV.

!Pig. 18. a. b. c, d. The appearance of scales situated near an abnormal one. Central
portion of four scales shown, taken in successive order from a lateral
scale-row :

a. Ninth scale in the row being normal,

h. Tenth scale in the row being slightly modified,

c. Eleventh scale, being greatly modified,

d. Twelfth scale is abnormal. The next scales in the row (on the other side

of the abnormal one) show similar feature^.

Fig. 23. Pa't of a scale, showing faint concentric shadows, caused by the inner
fibiilbar plates (corresponding to Fig. 14, Plate V). x9. (Lea phol.)

78

CAXADfAX FisnEinr:s; expeditiow loi.'i-irj 79

Pr.ATR VII.

Fig. 19. Herring scale from iiortliernmost Xorway, end of April, showing the com-
meneenient of a new .summer's growth. xlO. (Lka phot.)

Fi(;uui:s IN THE ti;\t:

Fig. 5. a. and o. Showing pattern formed hy fil)rils in the inner surface of small
(young) herring's scale,
c. Pattern a superposed over pattern h. Diagrammatic.

Fig. 7. Herring scale in section. Highl.y diagrammatic.

Fig. !). Edge of a herring scale in section, indicating true proportions.

Fig. 10. More central portion of a herring scale in section, showing the appearance
of a radial furrow : true pro]X)rtions.

Fig. 11. Part of a section of a herring scale, the position of two winter rings indicated.
True proportions.

Fig. 12. Effect of silver impregnation of the inner surface of a scale, the course of
the fibrils being indicated by small dark spindle-shaped particles. Dia-
grammatic.

Fig. 20. Showing average increase of the last (new) summer belt during the sunnner
of 1910 on the scales of young fish caught near Bergen.

Fig. 21. Composition in point of age of the stock of Norwegian mature herring, during
a series of years as shown by the samples investigated. (From Hjort.)

Fig. 22. a. Shows a scale with normal arrangement of winter rings,

h. Shows a scale with abnormal arrangement of winter rings, the distance
between second and third rings being small.

Fig. 24. Marking off the growth of scale on a paper sliii.

Fig. 25. Multiplying the distances marked off on the paper slips by help of a triangle
with a moveable side.

Fig. 26. Composition in point of age of all the seine samples of mature Norwegian
herringi, spring of 191 (i.

Fig. 27. Frequency cui"A'es for growth-dimension /, in Norwegian herring of the year
group 1904, during a series of years.

Fig. 28. Chart showing localities mentioned in the following: the black numbered
dots show the positions of the stations of " Steam Drifter No. 33,"
during summer 1915.

Fig. 29. Age composition of five samples from various places in Northumberland
Strait, spring of 1914 and 1915, each sample kept .separate: the indivi-
duals arranged according to age.

Fig. 30. Composition in point of year-groups of five samples from various places in
Northumberland Strait, spring of 1914 and 1915; samples from each
season taken together.

Fig. 31. Age-composition in two samples from iNfagdalen Islands, spring of 1914 and
1915.

Fig. 33. Age composition in samples from the coast of Newfoundland, spring 191:1

,80 DEPARTMENT OF THE NATAL SERVICE

Fig. 33. Age composition in samples from the coast of Newfoundland, autumn 1914.

Fig. H-i. Age composition in samples from the coast of Xewfoundland, spring 1915.

Fig. 35. Composition in point of year-groups in samples from the coast of Newfound-
land, samples from each season taken together. Spring 1914, autumn
1914, spring 1915.

Fig. 36. Age composition in North Sydney sample, June 3i, 1915, compared with that
of a divergent sample, St. George's Bay, May 27, 1915.

Fig. 37. Age composition in two samples from West Ardoise (July and August 1914)
taken together, compared with that of a sample from west of Port Hood
(station 42), July, 1915.

Fig. 38. Age composition in samples from different waters, 1914 and 1915.

Fig. 40. Growth of Newfoundland herring compared with that of herring from Mag-
dalen islands and Northumberland Strait, Curves in upper part for
increments (t), curves in lower part for lengths )(1).

Fig. 41. Growth of Newfoundland herring compared with that of herring from North
Sydney and Atlantic coast. 1

Fig. 42. Growth of Magdalen islands herring compared with that of herring from
North Sydney and Atlantic coast. i

Fig. 43. Graph indicating the course of seasonal growth in the West Bay of Port au
Port.

Fig. 44. Graph indicating the course of seasonal growth in the waters between the
coast of Gaspe and Prince Edward Island.

I'ig. 45. Graph indicating the course of seasonal growth in the waters from Souris,
P. E. I., southwards.

REPORT ON ''AGE AND GROWTH OF THE HERRING IN CANADIAN WATERS."

I.— INTRODUCTION.

Dr. Johau Hjort has requested me to ^work up the material of scale samples and
observations bearing on the biology of the herring, collected in Canadian waters
during the Dominion Government expedition, 1914 and 1915. The points especially
taken into consideration ,'when collecting the material — and which naturally also
furnish the main problems to be dealt with in the present work — have been formulated
by Dr. Hjort, in his preliminary report (VI) as follows: —

1. Do the herring that visit the Atlantic coast of Canada all belong to a
single race or^type, or is it possible to distinguish several races in these waters?

2. Does the rate of growth vary (according to the conditions of the waters
along the coast) ? Can types of different growth be distinguished and defined ?

3. Is the renewal of the stock of herring of a constant character, or are
there the same great fluctuations in the stock (in the number of individuals
belonging to the different year-classes) as in European waters?

A small portion of the material collected consists of biometrical observations.,
i. e., determinations of number of vertebrae, number of fin rays, and number of keel
scales, these being characters which, as Heincke has shown, may serve as a basis for
morphological distinctions between different tribes or races of herring. The greater
part of the material consists of observations as to length, weight, sex, state of sexual
organs, and fat, accompanied by scale samples serving for determinations as to age
and growth.

The material embraces a considerable geographical area, including, as it does,
samples from southern waters of Canada (including a series from Gloucester, Mass.,
U.S.A.), Bay of Fundy, Nova Scotia, Cape Breton island, Newfoundland, Magdalen
islands, Northumberland strait and as far north as the Gaspe waters. In some of these
localities, moreover, the collection of material extended over several seasons, and there
are also various samples taken with different fishing implements.

The results arrived at on examination of this material will be describetl in the
following pages. From the nature of the material itself, the greatest weight will
necessarily be attached to the discussion of that portion which deals with the age and
growth of the fish, as indicated by the state of the scales. A brief preliminary des-
cription and explanation will therefore be given as to the methods employed in scale
investigations, with observations as to the practicability of the method for dealing
with the herring in transatlantic waters.

II.— METHODS OF AGE DETERMINATION AND GROWTH MEASURE-
MENT IN HERRING.

It was early discovered that t3ie bony parts of fish were built up in layers,
resembling the annual rings visible in the trunk of a tree, and as far back as the
eighteenth century, the suggestion was advanced that it might prove possible, by

SI

82 DEPARTMENT OF THE NATAL SERVICE

counting these layers, to arrive at the age of the specimen. This theory was formu-
lated in 1759 by a Swede, Pastor Hederstrom, in the following manner : " Anyone
taking the trouble to examine a vertebra from a boiled fish, will observe certain rings
thereon. And as many rings as there may be, so many years will be the age of the
fish."

This suggestion, however, was afterwards lost sight of, and it was not until about
1900 that the skeletal parts of fish were called into requisition for the purpose of age
determinations. The same observation was then revived by two scientists, Hoffbauer
(VIII) and Reibisch (XIII) and the question has since been further dealt with by
many others.^.

It has been found that the different bony parts are not all equally well suited for
the purpose of age determinations ; in the case of the plaice, for instance, the scales
are difiicult to deal with from this point of view, being small, and with indistinct ring
formation. In this species, the otoliths and the opercular bones furnish the best
means of ascertaining the age. In the case of the herring, the opposite has been found
to be the case, the otoliths are here small, and awkward in shape, whereas the scales
are large, easily collected, and distinctly ringed. Although the otoliths and vertebra?
of the herring can be, and have been, used for age determination, the scales offer so
many advantages that there can, in my opinion, be no question as to choice. In the
following pages, the structure of these scales will be briefly described, with special
reference to the pattern of the annual rings. A resume will likewise be given of the
exact facts adducible in proof that the ring formation actually does consist of annual
rings, and that the scales may therefore be employed for the purpose of age deter-
mination. Finally, mention will be made of the methods of preparation and investiga-
tion which the writer's experience has shown to be most convenient.

III.— STRUCTURE OF HERRING SCALES.

On examining a number of herring scales, it will be found that these are thin,
pliable plates, differing both in size and shape according to the position on the body
of the fish. The scales from the forepart of the body are larger than those from the
caudal region, and those on the back smaller than those situate farther down, near
the lateral line, etc. The shape, or outer contour, differs between scales set, for
instance, immediately behind the gill cover and those farther back. Most scales are,
however, more or less, of the shape, shown in fig. 1, which is reproduced from a photo-
graph of a scale taken from near the lateral line, almost straight above the pectoral
fin. A feature common to almost all scales is the fact that the anterior portion, which
lies embedded in the scale pocket, presents an entirely different appearance to that
of the posterior portion. The former is of even contour, and appears to be finely
striped{, whereas the latter has a fringed contour, and lacks the fine striped pattern.
The two portions are divided by a line, termed the basal line (h — h in plate I, fig. 1).
The stripes of the anterior portion appear to fall into two distinct systems, with a
zigzag boundary between, this being as a rule ahr.ost perpendicular to the basal line
(z — c in plate I, fig. 1). At the point of intersection between this zigzag line and the
tasal line lies the centre of the scale (c) ; this is in most cases distinctly marked, but
may scarcely be seen on this figure. In addition to the fine stripes, a number of very
pronounced lines are seen running from the margin some distance in (r) ; these are
called the radial lines, from the fact that in the scales of many species of fish, the
corresponding lines radiate out from the centre of the scale. In plate I, fig. 1 will
be seen eight dark narrow lines, arranged concentrically about the middle of the

lA list of the most important works on tliis subject will be found in Dahl (I) The assess-
ment of ane ajid growth in fish, Intern. Revue d. gesamt. Hydrobiologie u. Hydrographie 1909.
Bd. 11.

CAXADr.W Flf^flEJilEH EXPEDITIOX, 191 ',-1}

83

scale; these divide tlie striped iiortiou of the scale into nine zones, each havin/r the
outline of a horseshoe; in plate I, fij;. 1 these appear lighter than the narrow lines.
By altering the light in which they are viewed, these broad zones can be made to
appear darker than the narrow lines (plate 11, fig. 2). These latter are the winter
rings of the scale, and the broader zones representing summer growth; it is by count-

Plate I.

ing these that the age of the tish is determined, and by measuring the distance between
them, it is possible to calculate the growth.

It is an easy matter to show that all the details above referred to pertain to the
outer surface of the scale, i. e., that surface of the scale which faces outward when
the scale itself is in its normal position on the body. The optical effect is a result
of reflection and refraction in this external surface. The demonstration may most
easily be made by producing an impression of the outer surface of the scale in a

84 DEPARTMENT OF THE l^ATAL SERVICE

transparent plastic mass, as, for instance, collodium solution. The scale is glued to
a glass plate, with its inner sidd next the glass, a small quantity of collodium is
poured over the whole, and the glass set aside in a slanting position, to dry. After a
short time, when still soft, the collodium may be removed in the shape of a thin film,
in which will be found an impress of the surface of the scale. Plate III, fig. 3, shows

Plate II.

,.'/?;u'^W'7^f^^n'^f-u

'<■■ /

FIG, 2

a photographic reproduction of a part of such an impression, together with the corres-
ponding part of a photograph of the scale itself. All details are distinctly visible in
the plastic impression. From this experiment we may with perfect certainty conclude
that the picture presented to the eye when observing a herring scale through a low-
power lens, is nothing but the play of light on the surface of the scale, which is thus
found to be moulded in delicate and detailed relief. The visible winter rings and
summer zones, like the fine stripes, the basal line, and the centrepoint, all belong
solely to the surface of the scale.

CAXADIAX FlHllElilES FXPEDITIOX. 191 ',-15

85

The inner structure of the scale tlius plays no part in the formation of the pic-
ture presented by the visible system of annual rings. It is possible, however, by par-

Plate III.

IG. 3

ticular means, to bring out another system of annual rings, located in close relation
to those of the outer surface, yet altogether different from these. In order to com-
prehend the genesis of this inner system, it will be necessary to have some idea as to
the internal structure of the scale as a whole. This may be arrived at by various

86

DEPARTMENT OF THE NATAL SERVICE

simple methods, the results of which will be found to agree with those obtained by
more complicated experimental processes.

With the aid of a needle and tweezers, a herring scale may be split up into
several thin plates, all exhibiting a contour identical in form with that of the scale
itself. The scale can only be thus split by inserting the point of the needle in the
outer surface, and the size of the flakes thrown off will depend upon whether the cleav-
age is commenced near the edge of the scale or farther in towards the centre. By
commencing near the margin, two large plates will be obtained, starting from a point
farther it will give one large and one small. Plate III, fig. 4 shows the two portions
of a scale thus split. The central portion here thrown otf presents the appearance of
a small scale, with the margin somewhat torn, the remainder, constituting the larger
part, exhibits a transparent " sore " in its central portion corresponding to the smaller
flake detached. From such experiments it may be seen that the scale is built up of
thin plates, similarly formed, each somewhat smaller than the next. The layers must
have been deposited in order of size, forming something approaching a very low cone,
the base of which is represented by the inner surface of the scale, and the apex by the
centre. The exact number of layers of which a sca^e is composed cannot be deter-
mined by mechanical means, as it is impossible to say whether each flake detached
comprises but a single layer, or is itself composed of several adhering together.

In a scale thus split, the surface exposed reveals a kind of pattern formed by fine
fibrils curving in prcs and whorls (vide plate III, fig. 4.) In most cases it will be
found that the fibrils nearest the edge of the exposed portion run parallel
with the same, while those farther in, on the anterior part, form a kind of elUptical
figure, and those in rear, on the posterior part (viewing the scale as in its normal
position on the body) form irregular whorls. From the pictures thus presented,
which are almost always to be found in such exposed portions where flakes have been
detached, it might well seem that the fibrils in each elementary plate were arranged
in this manner, with tangential marginal fibrils. This is, however, not the case, for
on studying the arrangement of the fibrils in scale preparations obtained by other
methods one may sometimes find the pattern above described, and in other cases pat-
terns of altogether different character.

The inner side of the scale is formed by a fibrillar plate, and, with a fairly
strong lens, the fibrils in this may easily be seen, especially if the scales are first

a

Fig. 5.

treated with nitric acid, and, in mounting, placed with the outer side down in a drop
of glycerine, leaving the inner side upwards, exposed to the air. On examining a
number of such preparations, not a few will be found in which the fibrils, instead of
running tangentially to the edge of the scale, fall perpendicularly to this, at any rate

CANADIAN FISUERIES EXPEDITION, lOl'rlo

87

along the greater part of the margin. In the case of hirge scales, the fibrils form a
highly complicated pattern farther in towards the centre. With small scales thus pre-
pared the patterns formed by the fibrils may be divided into two types, schematically
shown in fig. 5. a and h. The one of these types is characterized by a more or less
elliptical figure, and by the fact that the marginal fibrils run tangentially, the other by
a hyperbolar figure and radial marginal fibrils. There is thus no doubt that in addition
to elementary plates with tangential marginal fibrils, there must also be others with
radial fibrils, and the fact that these latter are rarely if ever found in the exposed sur-
face of a split scale must be due to the method employed. It is also obvious, that a
needle, thrust in with the point directed towards the centre, will find no hold until in-
comes 'into contact with transverse fibrils, i. e. until it reaches a plate with tangential

fibrils. ^ £ iix n

The scale may thus be considered as a greatly flattened cone composed ol tihnll-
ary plates, of which some have tangential fibrils, others radial. This cone is evidently
covered entirely by a non-fibrillary layer, on the upper side of which, however, is
found finely marked relief which gives the scales its characteristic appearance. If
this covering layer could be removed, the margin of each fibrillary plate would then
be visible, since'^each plate is larger than the next. A method by which large portions
at least of the covering layer can be removed is as follows : A drop of fish glue is
placed upon an object glass, and a scale set thereon, with the outer side downwards
in the glue, leaving, however, a small corner of the scale free, just large enough to
afl'ord a hold for the tweezers. As soon as the glue has dried the scale is damped very
slightly with an almost dry brush, the free corner is then grasped with the tweezers,
and the scale torn away. If the operator is fortunate, the covering layer of the scale
will then be found adhering to the glue on the glass, and the part detached by the
tweezers may be mounted for observation. Plate IV, fig. 6 gives a photographic repro-
duction of part of such a preparation. The curved boundary lines of the elementary
plates are distinctly visible, and it will also be seen that plates with tangential fibrils
alternate with those having the fibrils radially arranged.

It will be easily understood that the arrangement of the fibrils in the elementary
plates as here described imparts to the scale a high degree of firmness in all direc-
tions, the fibrils in one lamella will, roughly speaking, form a considerable angle with
those of the two adjacent, so that the scale may, in a way, be compared with the com-
posite wooden plates which are made by gluing several thin sheets together, with the
grain of each perpendicular to that of the two adjacent. Fig. 5 c gives, purely schem-
atically, the arrangement of the fibrils in two plates, one with radial, the other with
tangential fibrils, showing the manner in which the direction of the fibrils in the one
lies transversely to that of those in the other, when two such plates are placed in
juxtaposition.

Fig.

Judging from what we have hitherto learned, the transverse section of a scale
should pl-eseut more or less the appearance shown in fig. 7. This is also, roughly
speaking, found to be case. The upper covering layer can be distinctly seen when the
section is stained with thionine, as shown in fig. 8, the layer in question then assuming

Plate IV.

! T 7" ?' rjggg"

?1G.8

-^t^A^^

FIG. 13

C.\\.\rn\\ FISflFRfF.^ F.Xrr.DITloy, 101 'r15

89

an intense blue colour, while all the remainder is unaffected.^ Sections show that
this upper covering layer is of almost equal thickness at the edge and near the centre
of the scale, and evidently does not grow thicker; it is thus easy to understand that the
winter rirbg-s, for instance, upon the surface of this layer, continue eiiually distinct
many years after formation, in contrast to what is found be to the case witli otoliths,
where the earliest annual rings become entirely concealed, and are only discernible
after the otolith has been ground down so as to expose its inner structure.

An examination of sections also reveals a number of breaks in the upper layer;
these will be found to be the so-called radial furrows {vide plate I, fig. 1), which run like
channels in the surface layer. Finally, it will be seen that the exterior covering layer
extends out beyond the edge of the fibrillar plates and forms by itself, independently
of these, the margin of the scale. The winter rings are not particularly conspicuous
in a section ; despite the most careful orientation by means of the fine ridges on the
surface, it has not been found possible to discern any conspicuous peculiarity in those
parts of the outer layer where the winter rings lie. It is noticeable that the ridges,
which resemble the teeth of a saw, are set somewhat irregularly; possibly, also, the
outer layer itself is slightly thinner in the inner than on the outer side of the point
where a winter ring is situated. These features are, however, so inconspicuous, that
without exact orientation it would be impossible to demonstrate where the winter
rings actually lie in a section. Figs. 9 to 11 give diagrams of various details from
sections of herring scales.

Fig. 9.

Fig. 10.

'rr^^^TTf^^^^r^^

Fig. 11.

Below the (juter ccnering layer may be seen, viewing the scale in section, a
thicker stratum, divided into zones of varying appearance. If the section be embedded
in a medium of low refraction, it may be distinctly seen that the fibrils in some of the
zones have been cut transversely across, whereas other zones are clearer, with less
apparent structure, suggesting that the fibrils here are not transversely severed. This
layer formaticm may be strikingly shown ])y placing the sections in polarized light,
when, in certain positions, a series of bright bands, with darker zones between, will b"
apparent. It will then be distinctly seen that the thickness of the lamella? varies, and,

1 This staining did not succeed, when the scales cut had been treated with acids in order to
remove inorganic matter. I have, by microchemical reactions, satisfied myself of the fact that
practically all inorganic matter (principally phosphate of lime) is in the herring scales deposited
in the outer covering layer, while the fibrillar plates are devoid of it. I suspect that the
differential staining is due to the presence of metallic salts in the one layer only, the salts acting,
as it were, as a mordant.

fi551— 10

90

DEPARTMENT OF THE NAVAL SERVICE

as will shortly be shown, this is,, in all probability, to be regarded as in connection
•with the fact that the lamellse themselves form a system of annual rings, which may
be rendered discernible by preparing the scales in a special manner. If we take, for
instance, fresh herring scales, treated, however, with nitric acid, and impregnate them
with bichromate of potassium, thereafter placing them in a solution of nitrate of
silver, a dark precipitate of some silver compound will be formed between the fibrils.
This precipitate is not amorphous, but is deposited in small spindle-shaped particles,
each with its axis in the direction of the fibrils, so that a preparation of this kind
distinctly shows the course of the fibrils themselves just as the course of currents
is shown on nautical maps. As the impregnation takes place only, or mainly, in
those layers which lie nearest the surface of the scale, it is possible, by gluing the
scale to be impregnated, on to a glass plate, to exclude one of its two surfaces from the
action of the silver solution and thus to impregnate either the innermost lamella?
alone, or only that portion of each, which lies immediately adjacent to the outer layer
of the scale. The operator can thus obtain at will either views as in fig. 12, giving

Fig. 12.

schematically a portion of the innermost lamellse, or as in plate IV, fig. 13, showing a
portion of a scale impregnated from the outer side. This latter shows, in another
manner, the same feature as seen in plate IV, fig. 6, while the former corresponds to
plate III, fig. 4.

On examining a scale thus impregnated, with a low-power lens and in oblique
light, the small spindle shaped bodies will not be discernible. They produce, however,
by reflection an effect as shown in plate V, fig. 14. In this manner, an excellent view
is obtained as to the extent of the elementary lamellse, and it is at once strikingly
noticeable that the breadth of the zones exhibits an irregular progression, broader belts
suddenly appearing after a series of narrow zones. Closer comparison reveals the fact
that the transition from narrow to broader zones takes place just where the surface of
the scale shojvs a winter ring. Thus the elementary plates are seen to form their own
.system of annual rings, corresponding to that of the surface layer, but otherwise
differing' greatly from this, and more resembling that found in the scales of many
salmonoids and gadoids, etc, where the winter rings are not so sharply marked, but
a gradual transition from summer to winter is seen.

The foregoing description as to structure of the scales does not apply to all the
scales found on the body"/ of a herring. On looking through a collection of herring
scales, some will be found to differ from the rest, being most easily distinguishable
from these by the fact that the zigzag boundary line between the two systems of stripes
(vide plate I, fi^g. 1) does not reach right down to the basal line, nnd that the stripes
in a more or less considerable central portion of the scale form irregular patterns, in
contrast to what is normally the case. Winter rings, again, are invariably lacking in
this abnormal central portion, although present outside it, excepting, of course, cases

Plate V.

>/■'

FIG. 15

FIG. 16

-7^

r)551--10.^

92

DEPARTMENT OF THE NAVAL SERVICE

where the entire surface of the scale is abnormal and no winter rings are to be seen.
Plate V, figs. 1.5 and 16 give photographic reproductions of the central portions in a

Plate VI.

'/v/- _..<.

-::■'■ «^^-

n

normal and an abnormal scale from the same fish. Ihe occurrence of these abnormal
scales is highly irregular; on one specimen, numbers may be found, while in another,
long and carrfnl search will be required to discover a single one. It is furtlier remark-

CAyADT.iX FIS^HERIES EXPEDiriOX, lOl'rlo 93

able, that in one and the same specimen may be found scales where the abnormal cen-
tral portion is (luite small, and others in which a larg-e portion, or the entire surface,
is of abnormal appearance.

The same confusion noticed iii the relief of the outer surface is likewise encoun-
tered in the inner structure of such abnormal scales. It will be found impossible to
split them, for instance, as long' as the needle jxtint is introduced within the abnormal
central part, and if the covering' layer be removed, we do not find the regular alterna-
tion of radial fibrillar zones with those having- tangential fibrils; on the contrary,
the fibrils will be seen to intertwine, forming patterns similar to that shown in plate
VI, fig-. 17. If such abnormal scales be treated with silver solution as described be-
fore, it will be noticed that the division into zones, which otherwise makes itself so
distinctly apparent during- this process, is not discernible within a central portion
corresponding to the visibly abnormal portion of the surface layer.

The most probable explanation as to the origin of these abnormal scales would
seem to be as follows. The scale originally set in the scale pocket where now the
abnormal one is found, must on some occasion have fallen out, this' taking place at a
time when the original scale was of a size corresponding to the abnormal portion of its
successor. Within a short time after the loss of the original scale, a new scale (the
abnormal portion) grows out, and then, having filled the vacant place, the scale pro-
ceeds to grow normal wise.

Closer investigations of the scales of a herring, with this question in view, will
soon reveal the fact that the scales innnediately adjacent to an abnormal scale are, as
a rule, themselves abnormal, forming together a group. And outside these
again, or rather round about them, may frequently be found scales presenting another
form of abnormal structure than the first mentioned, the central portion correspond-
ing to the abnormal centre in the first being here either dislocated in relation to the
peripheral part, or at least separated from this by a line concentric with the peri-
phery, and more or less distinctly marked. Outside these scales lie the wholly nor-
mal scales, so that transition stages are seen to lie between the two extremes. In
order to make this clear, a group of entirely abnormal scales was sought for on the
body of a herring, and all the scales surrounding this group w^ere then picked off and
prepared, careful note being kept as to their respective positions. Plate VI, figs. 18a
to d are from photographs showing the central portions in a horizontal series of these
scales, the first giving a view of the last normal scale before the abnormal formation
begins, and the last illustration being that of the first abnormal scale. This series of
views, and similar series which I have had no difficulty in finding, distinctly suggest
that the abnormal scales are nothing but supplementary growths intended to replace
scales previously lost by accident. The scales immediately adjacent have become
slightly displaced from their natural position in the scale pockets, as shown by the
fact that the central portion forms, as it were, a scale upon the scale. It will also be
seen that the transition forms may well give rise to erroneous determinations of age,
the boundary line between the central portion and the periphery frequently bearing
more or less resemblance to a winter ring. This will be further referred to later on.

IV.— THE SCALES AS Al^ INDICATIOX OF ACE. SOURCES OF ERROR

IN ACE-DETERMI X.\ TIOX.

We have no experimental proof of the fact that the so-called winter rings on the
scales of the herring actually are annual rings, i. e., that one such winter ring is
formed each year. On the other hand, statistical observations point so emphatically
to the correctness of this supposition, that it will hardly seem possible to otherwise
explain the regularity revealed by the observations. The observations in question
deal primarily with the Norwegian stock of herring, which have been under considera-
tion in this respect for several years. The proofs thus furnished can, strictly speak-

94

DEPARTMENT OF THE NATAL SERVICE

ing, only be regarded as entirely applicable to the stock in question, and it is of course
perfectly justifiable to demand that in dealing with other races of herring, similar
data should be sought for before assuming that the conclusions arrived at are equally
valid for these. As we shall see later on, the Canadian material, like the Norwegian,
does in fact point to the same conclusion, that the winter rings really are an indica-

Plate VII.

tion of the age of the fish. At present it will suffice to mention, as briefly as possible,
the manner in which the observations were carried out in the case of the Norwegian
herring; it should be noted, however, that corresponding features have been observed
in the case of other stocks of herring, e. g., those of the Faeroes.

In the course of the year 1910, the young herring which were continually taken
in the neighbourhood of Bergen were subjected to observation (vide Lea, X). It was
then found that the outermost summer zone on the scales, which in May was extremely

CAXADIAX FISHERIES EXPEDITION, lUl'rlo

95

narrow, grew broader and broader as the summer progressed, until September, when
a period of stagnation set in. Supplementary investigations have shown that at somo
time or another during the months of ^farch or April, small herring are found, some
of which have only broad summer zones (one or two),, while others have either a broad
inner zone and a very narrow outer one, or two broad ones innermost, and one very
narrow beyond {vide plate VII, fig. 19). Fig. 20 shows schematically the manner in
which the annual rigs were seen to appear at different times of the year. Observa-
tions extending over several years have shown that the small herring taken near
Bergen only exhibit this narrow outer summer zone in the spring, and that the outer
summer zone is in autumn invariably found to be broad.

Continuous observations of this nature will be analogous to the observation of
herring kept in tanks, with periodic examination of the scales. If an experiment of

December 1909

UUimo April 1910

Ultimo May

Ultimo June

Ultimo July

Ultimo August

Medio September

Medio October

Fig. 20.

this kind, with herrings in tanks, were found to give the same results, we should then
certainly be justified in concluding that the summer zones of the scales were formed
and developed during the period, April to September, and that the winter rings in
consequence represent the time from October to March. In the present instance, this
absolute certainty is not attained, since the observations with captured herring
naturally to do not represent different states of the same fish at different times.

During a period of years, from 1907 to 1916, samples of so-called spring herring
were collected, these being the fish which early in the spring move in towards the west
coast of Norway to spawn. Scales of each fish in the samples collected were examined,
and the herring grouped according to the number of rings on the scales. It was then
found that for the greater part of the i)eriod embraced, a remarkable regularity
prevailed, as indicated by fig. 21. This has been lent from Hjort (V) and
shows the percentage of fish (in the samples investigated) falling to each group accord-
ing to the number of rings. It will be seen that in 1908, there were many with 4
rings ; in 1909, 5 ; in 1910, 6 ; and so on until 1914. The investigations of 1915 and
1916 revealed the presence of a large number of herring with eleven and twelve rings,
respectively. This regularity, which, by the way, is also encountered in samples of

96

DEPARTMENT OF THE NAVAL SERVICE

another " kind " of x^orwegian herring, the so-called large herring, can hardly be
explained save by the supposition that a certain year-class of herring have throughout
the whole of the time from 1908 been more numerously represented in the stock than
others, and that as time went on, and the fish grew older, one ring was formed on the
scales for each successive year.

These observations exhibit great similarity to the results in the following experi-
ment, which may be easily carried out with fresh-water fish : A number of fish are
caught, all those with a certain number of rings on the scales sorted out and liberated
in a pond where no fish of the same species are' found. A year later they are recap-
tured, scale samples taken and examined, to ascertain whether the scales now exhibit
one ring more than at time of liberation. They are then again set free, to be taken
up once more a year after for renewed examination. The investigations actually made

3456769 10 1

1 12 13 14 15 16 17 18

1907

'7 A/t-"V_

1908

'0 / /V--^

1909

- Jm

1910

10 ^ /stiA

1911

io___y^ ylj\

1912

^° __— >^ /u

1913

"^^. ^,^<—^ ,^

^ 1914

5 4 5 6 7

Age groups

10 II 12 13 14 15 16 17 18

Fig. 21.

lack the absolute certainly as proof which such a test experiment would have; on the
other hand, the length of the period here embraced, and the absolute uniformity of
the results from all samples, are points of no little weight. Throughout the whole ^f
the time from April, 1908, to February, 1916, not a single sample was found to furnish
any exception to the general regularity. Not until February, 1916, was a sample
brought in which lacked the 12-ringed fish present in such great numbers in
the remaining twenty-four samples from the winter in question. This single instance,
however, does not impair the proof-value of the material as a whole, for the regularity
observed can naturally not be expected to continue indefinitely. On the contrary, it
would be natural to expect that the character of the stock in this respect will in a
short time exhibit noticeable change.

The samples of grown Norwegian herring, from the year 1910, inclusive, contain
a considerable quantity of specimens exhibiting a remarkable arrangement of the
winter rings. As a general rule, the distance between the rincs decreases from the

t'A.\AI>f.[\ Jl^HKRlEi^ EXrEniTIOX, lOl/f-lo

97

innermost outwards to the margin of the scale, as shown in fiji-. 22 a; in the specimens
here in question, however, the arrangement was similar to that shown in fig. 22 h,
where, as will be seen, the third summer zone is narrower than the fourth, reckoning
from the centre outwards. In 1(110, this peculiar feature was especiall.v marked

Fig. 22.

among the 6-ringed fish; in the following year, it was chiefly apparent in the scales of
the 7-ringed specimens; and so onwards. The table prepared shows how these parti-
cular fisklieep together in one group, with a gradual progression of the group as a
whole towards the right of the table.*

Tablk 1. — Distribution of herring with abnormal arrangement of rings on the scales.
Material from grown Norwegian herring, seine caught.

Total

Percent of

ndividuals in

tlie different

ring-classes.

number t)f

individuals
with abnor-
mal arrange-
ment.

^p be

a

'V

o be

a

Year.

t3

CO bo

c

•73

00 ic

a

C5 bo

a

c

^1

si

;.*

'"

u

»

h

u

s-

'"

'-

u

U

^

1910

134
372
322

2-2
11
2-5

91-0

4-8
1-2

4-5

86 3

4-3

'4 C

84-5

0-7
1-3
3 4

0-7
11
0-3

0-7
5

"6-3

0-9

litll

191-J

1-9

0-9

1913

217

()-9

0-9

2 3

1-4

3-7

86-7

2-3

0-5

9

5

1914

518

1-9

3 1

1-9

21

2-7

5-4

79 2

2-1

6

0-6

4

1915 . .

1291

0-2

23-8

2-8

3 3

1-8

0-7

30

61 5

1-4

6

0-3

01

These observations may best be compared with an experiment whicli has already
been carried out in the case of the cod (Winge, XIV). The fish, when caught, are
marked with numbered labels, scale samples taken and tbe marked specimens
liberated, care being taken to select only fish of unimpaired vitalit.v.
After the lapse of some time, some of them are recaptured, and the scales examined,
to see whether new annual rings have appeared in number corresponding with the
lapse of time between first examination and recapture. Instead of the artificial mark
affixed wo have here the peculiar arrangement of the rings as the distinguishing fea-
ture of a group whose scales in a given year (1910) numbered six rings in all. It is
liardl.v likely that an approximately equal percentage of this group would then, in
1911, be fcuind among the 7-ringcd fish, in 1912 among those with eight rings, and so
on, up to to the 12-ringed in 1916, if it were not that a new ring made its appearance
each year.

Scales having the appearance shown in fig. 2*2 h were not found, or found in but
niiiiiuial (|uantities, among the samples of mature fish until 1910. On the other hand.

98 DEPARTMENT OF THE NATAL SERVICE

in 1907-09, these were the more numerous among the samples of immature herring
from the north of Norway, and as might be expected (in view of the fact that the
samples consist of autumn-caught iish)' they are found in 1907 only among the herring
with four summer belts; in the following year, 1908, they appear in large numbers (95
per cent)^ among the fish with five summer belts, and in 1909 among those with six
(79 per cent). In the autumn of 1910, only a very few fish with seven summer belts
were found (seventy-eight in all, out of 1614 specimens examined). Of these, how-
ever, fifty-five ihad scales of abnormal appearance, and these fifty-five specimens
amount, after all, to 40 per cent of the total found.

Consideration of fig. 21 and table 1, leads us directly to the question of possible
error in counting the rings of the scale. What would be the results, for instance, if,
owing to various difficulties, the investigator were unable to note with certainty the
number of rings on a considerable proportion of the scales? Would such inaccuracy
by any means be able to account for a filial result revealing so distinctly marked a
regularity as that here found in the case of the Norwegian herring?

It is no difiicult matter to see that erroneous age determinations will tend to pro-
duce an age curve which, where it should exhibit marked differences in the number of
individuals in the various groups, assumes, instead, a more even contour,, with less
pronounced contrast and more gradual transition between the age-groups. The groups
most numerously represented will give ofi^ a greater number of individuals to the
groups adjacent than they receive from these.

The experience gained in the course of the investigations with Norwegian herring
tends to show that the probability of error is greater when dealing with older fish
than with young specimens, and further, that the error more frequently falls to the
One side, thus placing the fish in a younger year-class than that to which it actually
belongs. And finally, it has been found that the majority of such errors are uniform
in degree, the fish being reckoned as one year younger than is actually the case.

The effect of errors of this nature and degree will then be that, where a certain
sample contains one dominant age-group, the predominance of the group in question
will appear less than it should, and the younger group next following be credited
with greater number of individuals than is its due. We are consequently led to the
conclusion that the figures arrived at in the case of a dominant group are minimal
values, i. e., that the diagram shown in fig. 21i, for instance, would have presented
essentially the same features, but in an even more marked degree, if all the deter-
minations of age on which it is based had been correct.

That errors and uncertainty are unavoidable in investigations of this kind will
be admitted by all who have had any experience of such work. The material may be
handled with the highest possible degree of care and attention, so as to warrant the
hope that a repetition of the determinations must give exactly the same results, yet on
going through the whole once more, discrepancies will nevertheless be found. As n
matter of fact, we are hardly justified in using the term " age-determination " when
dealing with scales ; " estimate^ " would be more correct, for there will always be
found, whatever may be the material under consideration, a greater or less niimber of
individuals whose scales must be classed as doubtful, and where the decision must be
based more or less upon personal judgment.

In this respect, the herring from different localities will be found to vary,, and it
is therefore impossible to formulate any generally valid rule as to how great the
probnbility of error will be. And, indeed, any such estimate would always be a matter
of difficulty. Eepetition, of course, affords a certain guide in this respect, and this
method has also frequently been employed, with satisfactory results, in dealing with
the Norwegian herring. Another method is to examine the actual results arrived at
in the investigation of a certain stock. It would be altogether unreasonable to
suppose, for instance, that the age determination in the case of the Norwegian herring

1 Excluding all fish with less than four summer belts.

CANADIAN FISHERIES EXPEDITION, lOl.'i-lo 99

could ever have been brought to exhibit pecularities so inarked, if the number of
erroneous determinations, a neutralizing influence, had been considerable. As will
be shown later on, the Canadian material furnishes several excellent instances of
consistency in the results arrived at. For the present, it will sutKce to point out that
the material from the Canadian investigations presents, in the case of some localities,
no particular points of difficulty, whereas samples taken elsewhere have included fish
with scales by no means easy to decipher. On the whole it may be said that the
Newfoundland material from the gulf of St. Lawrence is to be regarded as well suited
for the purpose of age determinations, whereas some part of that from the open coast
presents many difficulties. It is of great importance, however, in such investigations,
to familiarize oneself, by long-continued observation, with the peculiarities of the
particular herrings to be dealt with. Valuable aid is also furnished where oppor-
tunity occurs of following the progress of a single age-group from year to year. There
is probably no single circumstance which has so largely contributed to tlie firmness of
conviction, now i)revailing among Norwegian investigators, than the fact of their
having been in a fortunate position to watch the growth of a single rich year-class
throushout an extended period of time.

The greater or less probability of error, or uncertainty, may depend upon various
factors. The sources of error may be divided into two categories : —

1. Circumstances of a purely technical nature. The technical methods
employed in dealing with the material are of considerable importance, and will
therefore be described at greater length later on.

2. Conditions independent of the technical treatment, i. e., such as will
make themsdlves apparent even when the most adequate technical methods are
employed. In this case, it is generally a question of fish whose scales have the
winter rings indistinctly marked, or which exhibit fainter intermediate mark-
ings between those normally legible, or again, as with older specimens, the
outermost rings so close together as to render them difficult to distinguish.

Where the winter rings are not sharply defined, they frequently present the
appearance of several very thin lines, one outside the other, in the form of a faint
band. Such double or manifold rings would seem to be of most frequent occurrence
among those earliest formed; a type of scale very commonly met with is that where
the outermost rings are rather clearly mai'ked and easily distinguishable, while the
inner one, and possibly the next few, will be vague and double. How far this may be
connected with the spawning, as tending to render the rings more sharply defined,
cannot be stated with icertainty,, but it is not unlikely that such is the case.

The fainter rings occasionally found between the true ones have been termed by
Dahl (I) " secondary rings ", and are so distinguished in the present report, albeit
the term might well be taken to embrace various kinds of rings. I was at first of the
opinion that the position of all kinds of these '* secondary rings " varied from scale
to scale, and that their disturbing influence might therefore be eliminated by examin-
ing a sufficiently large number of scales from each fish. Objections to this have,
however, been raised by Hellevaara (IV), who considers that secondary rings may be
found, the position of which corresponds on the different scales of a fish — being, how-
ever, in some scales almost indistinguishable.

As to the origin of the secondary rings, nothing ean be said with certainty. The
dislocated scales described on p. 93 show, however, that a slight shifting of the scale
from its normal position may occasion the formation of secondary rings. In other
cases, faint shadows, produced by the inner fibrillar plates, may be seen. Plate VI,
fig. 23, reproduces a photograph of a scale exhibiting such shadows. Where the winter
rings are faint or doubled, it may be conceived that these shadows may become of
some importance as sources of errors.

With regard to the close-set outer rings in older fish, there is little to be said,
save that as the rings lie closer and closer with increasing age, we have here a limit

100 DEPARTMENT OF THE ^^ATAL SERTICE

to the possibility of approximately certain estimates of age. In the Canadian mate-
rial, some herrings were found which must certainly have been older than any which
I have examined among those from European waters, their age being over 20 years.
Where a definite age is assigned to them in the tables, this must expressly be noted as
approximate; in cases of this sort accurate determination is out of the question.

Finally, mention should be made of a particular source of error which needs to
be guarded against by special precautions.

In a sample of herring collected just at the time when the fish are developing
the first stages of a new summer zone on the scales, some specimens of a certain year-
class may be found where this has already visibly commenced, whereas others of the
same group have not yet reached so far. If, then, the rings be counted quite schema-
tically, and the observations recorded accordingly, the results will be that fish which
have already commenced their summer growth will appear a year older than those
which, albeit somewhat behindhand in this respect, are in reality contemporary with
the former, and the greatest confusion will ensue.

In many cases, e. g., in dealing with young herring, the change is more or less
conspicuous and the error may be avoided ; as in such specimens, a narrow summer
zone is visible outside a broader one, or several such. It is a different matter, however,
when the fish are older, and have already one or more narrow zones near the margin
of the scale. A new narrow zone on such a scale does not occasion any conspicuous and
easily disting-uishable alteration. In this latter case, it is difficult to say what precau-
tions should be taken unless it be to avoid collecting samples during this transition
period, or, possibly better, to supplement such samples by others taken before and after.
The grown Norwegian herring, it may be noted, are not regularly fished for during
the transition period.

V. THE SCALE OF THE HERRING AS AN INDICATION OF GROWTH.

SOURCES OF ERROR.

If, in addition to counting the winter rings on the scales, we measure the dis-
tance between them, these measurements will enable us to calculate how each fish has
grown from year to year. This, of course, presupposes that the growth of the scale
takes place at a rate simply proportionate to the rate of growth of the fish. Judging
from the investigations already made (vide Lea IX), the scales from various parts of
the body differ somewhat in this respect. On the whole, however, we may say that such
proportion does exist. ♦

It has been found, however, that in dealing with the material obtained from such
measurements of growth, the values arrived at in the case of j'oung fish appear to be
higher than the corresponding figures for the older groups (vide Lea XI and Lee XII),
so that possibly there may be a systematic error in the growth measurements.

This is, in my opinion, only true to a certain extent. Moreover, some part of the
apparently systematic error will, as a matter of fact, be found to arise from other
causes, to wit, as the results of a biological process. Nevertheless, we may, in prac-
tice, until these problems are fully solved, do well to handle our material as if there
were a systematic error, of the nature shown in the following taW?:^ —

CAXADiAX Flail i:nn:>i kxi'fditiox, loi't-io

101

T.\ni.i: 2. — Showing the decrease with increasing age of the corresponding calculated
average values for the yearly inoremciit in loiiiith. (Table reprinted from
T.ea (V) tab. 7.)

Year of capture.

Age.

No. of
samples.

No. of
indivi-
duals.

h

ti

5-8
5-2
5
51

4-8
3 7
3 5
3 5

U

h

U

ti

1907

1908

1909

191U

4^

2
1
3
6

69

.58

331

78

7-2
7 1
71
6-9

4 7*
4 2
3 9
3 9

"39*
3 4
2 9

32*

28

'2 b*

* l>eiiote incompleted increments.

In the table above, t, indicates increment of leii°^th during first summer of life,
t.. tlie corres])onding increment for second summer, etc. It will be noticed that the
figures in the vertical columns exhibit decreasing values, and that a growth curve
drawn on the basis of the figures for fish at 3* years would differ from one based on
the values for fish 6J years old. This feature, which is very freiiuently encountered,
may render immediate comparison between the growth values for young fish and those
for older specimens precarious ; therefore, where possible, samples of herrings of more
or less equal age should be compared.

As regards the degree of accuracy obtainable in such individual measurements of
growth, this is fairly high, and it is always possible, where greater accuracy is needed,
to measure several scales from the same fish and take their average. This branch of
the question need not therefore be further dealt witb here. It should, however, be
noted that all the sources of error which make themselves apparent in age determin-
ations also apply to measurements of growth. If a ring be overlooked, then two years
growth will be taken together as that of one, and the growth of the succeeding years
will be erroneously reckoned, etc. If, on the other hand, a secondary ring be mistaken
as a true one, the year's growth will be divided into two, and that of the following
years again incorrectly recorded.

Save in the case of particularly diffii-ult scales, however, errors of this nature
will rarel.v occur in the values for growth during the first few years.

VI. AGE DETEmnXATIOX AND (JROWTM MEASUKEMEXT TX PR.VCTICE.

Where the winter rings are distinctly marked, and the fish young, it is of little
importance from what part of the body the scales are selected; in the reverse case,
however, this may be of the utmost imi>ortance. In all the scales of a herring the
winter rings are by no means equally distinct. Where possible, therefore, the scales
where these are most pronounced should always be chosen, i. e., the large scales from
the middle forepart of the body.

VII. PREPARATION OF SCALES FOR PRESERVATION.

In the earlier years of the Norwegian investigations, the herring scales were
scraped from the body with a knife, and placed in small envelopes, where they gra-
duallv dried. Before being used, they were soaked in a soap solution to which gl.vce-
rine was added, which afforded a good enough means of cleansing them. This is a
practicable method, and even presents certain advantages, when working on the fishing
grounds, or on board a vessel under unfavourable conditions. Of late years, another
method has been employed; three scales are plucked out with a pair of tweezers, clean-

102 DEPARTMENT OF THE NAYAL SERVICE

ed while still fresh, then dipped in clean water, and laid in their wet state on an
object glass, upon which have been placed three small drops of white of egg and gly-
cerine (half and half), one little drop to each scale. Care must be taken that the
scale lies with its inner side against the glass. When the water has evaporated, the
scales will be found adhering firmly to the glass, and a permanent preparation is thus
cbtained. Notes as to length, weight, etc., may be made on the glass itself with water-
proof ink (India ink to which is added some water-glass).

A third method of extreme simplicity is merely to take a single scale from each
fisli, and place it in a tube of water, to be mounted later. Tubes are kept for each
length-group, or for each length-group and either sex, and the scales are then placed
each in the tube assigned to its particular length and sex. Each scale in a given tube
thus represents a fish of a certain sex and length and may afterwards be mounted on
glass slides. By t3iis means, a large sample of herring specimens may be dealt with
in a very short time; it involves, however, the necessity of dispensing with data as to
weight, state of development of genital organs, etc.. besides lacking the indubitable
advantage conferred by having several scales from the same fish. The advantage of
the method lies in the fact that it enables the operator to procure a large qiiantity of
material for age determinations in a short space of time, and that it may be employed
even under the most unfavovirable conditions. Moreover, where no examination of
the specimens is made as to sex, the fish are entirely unharmed.

Microscopical examination. — As we have seen in the description given in the
foregoing of the structure of herring scales, the impression of the winter rings is
produced by reflection and refraction of light in the outer surface of the ocale. In
accordance with this, it has been found that no colouring, or similar process, will serve
to render the rings more distinct than they naturally appear. The most that can be
done is to alter the conditions of reflection and refraction by embedding the scale in
a more or less highly refracting medium, experience having shown that this does
render the winter rings more easily discernible. For examination of herring scales,
a medium with not too great power of refraction has been found most useful, water
with a little glycerine, or alcohol (90 per cent) is good, the latter being preferable
when the scales have been mounted with the white of egg and glycerine, as aqueous
liquids tend to loosen the mounted scales.

In observing scales through the microscope, a suitable attachment should be
used, preferably an objective so arranged that the power can be altered at will. Leitz'
objective la and Reichert's objective Ih are both very convenient, as with these the
power may be continuously changed within certain limits.

In age determinations, it is best to remove the condenser from the microscope,
and leave the lighting to the plane mirror alone. It is rarely possible to obtain suit-
able lighting of tlie entire scale simultaneously, and the mirror must therefore be fre-
quently moved. The rings are best seen in slightly oblique light, when they show up
darker than the summer zones {vide plate I, fig. 1).

In counting the winter rings, the operator should make it a rule to commence at
the margin of tke scale and work inwards towards the centre, i. e., first counting the
rings which are most difficult to distinguish.

Where several scales have been preserved from each specimen, it is well to make
a preliminary glance at all those which are ready prepared, and choose the one in
which the rings appear most distinct. When the rings on this have been counted,
another scale may be taken as a check, to see if the same number and arrangement of
the rings is also fouiid there.

Instead of counting the rings on the scales, it is nossible, by means of a drawing
mirror, to outline them on paper and count them afterwards. Such a method pre-
sents several advantages a^so as regards the actual determinations of age. and it is
a question whether the method should not be adopted in most cases. It is best to
use a drawing mirror with a series of smoked glasses for regulating the light. In

CA\ADIAX FISHERIES EXPEDITION, 1914-15

103

order to render the picture of the scales as distinct as possible, a background of black
paper should be laid on the table to the right of the microscope. It will not be necess
ary to draw the entire contour of the winter rings; all that is needed is to mark tlu*
position of each ring along the edge of strip of a card, as shown in fig. 24. The can!

Fig. 24.

is laid on the table beside the microscope in such a manner that, to the eye, viewing
both card and scale in the mirror, the corner of the card falls exactly upon the centre
of the scale, the corresponding long side of the strip lying almost at right angles to
the basal line. If then each winter ring, and the margin of the scale, be marked off
along the edge of the card, a graph is obtained, presenting a magnified picture of the
growth of the scale. Multiplying all these dimensions by a factor which renders the
distance from the corner of the card to the mark for the edge of the scale equal to the
length of the fish — this can be easily done grai)hicany as shown in fig. 25 — we have

/

/

/

cm

5h-

^3-

\^

^^

sXMiJ^

2S-

^^^^S.

28-

^^^^fe..

27-

^^^

26 -

^^tj^

^w

2S-

^

"^WV

2t-

^><^^k.

2J-

^^^^.

22-

^^^pfv

21-

^^^^V

20^

^'^^

19-

>;^

■8-

^V^s.

■«;

^&^^^

1 ,

,18-

"^ ^^

^^i^^v

l'?^

^^ ^^

1*-

"•~~.. ^~-

^

13-

^.^

^^

^%^

.

12-

■"■ ^^ ^^ ^^

^•ift

^N,

11 -

^^ — .

"^*^^

10-

^^*

"^ ^

-^^^^

9-

~ -^ ""-^

^^

a-

■-•*» ^"^ .^

"

-^

^iv

7-

-^^^^

■* —

^^' '^s.

6-

^

^ ■

^.^

^^

s.

5-

--.

^ -^

^^v

5-

^"^

''''"^^l^^k.

2-

1 -

"^^%

Fig. 25

104 DEPARTMENT OF THE NAVAL SERTICE

the length of the fish at the times the different winter rings were formed. This entire-
ly graphical method of measuring the growth of each individual specimen will natur-
ally not give such accurate figures as those obtained by the more exact, but less rapid,
methods of measurement and calculation; both the use of the mirror and the subse-
quent graphical multiplication involve the occurrence of errors which can of course,
be diminished by the application of more exact methods. For most purposes, how-
ever, the method will be found sufficiently accurate.

VIII. DEFINITIONS AND ABBREVIATIONS ADOPTED.

Various expressions and signs employed in the following pages will need to be
defined.

1. Age. — Age is expressed by the number of annual rings which have been found
on the scales of the specimen in question, the relation of time of capture to season of
birth being, however, here taken into consideration whenever possible. Thus a herr-
ing taken at Newfoundland in the spring will, if its scales show, say, ten summer belts
be reckoned as 10 years old, whereas one from the same place, but taken in the autumn,
and with ten summer belts, will be regarded as about 9J years eld, having regard ti)
the spawning season at Newfoundland.

2. Age-group. — In assigning a specimen, say, to age-group 10, this is to be under-
stood as meaning that its scales showed ten summer belts, ignoring the fact, when
nothing is stated, as to whether it was taken in early spring or late autunm. The
term is therefore used without regard to season of birth or time of commencing new
summer growth.

3. Year-group. — When a fish is assigned, say, to year-group 1910, this means,
that the specimen in question probably formed its first summer belt in that year. For
all fish spawned in the spring, the date of the year-group will be identical with that
of the year of birth ; in the case of those spawned later, however, towards the autumn,
it will remain open as to whether or not they formed any summer belt during the
remainder of the year in which born. Fish taken during the period of transition to
new summer growth may be difficult to class correctly in their proper year-group, and
personal judgment will here be brought into i)lay iridp \). 113).

4. Growth-dimensions. — On the basis of growth measurements, we obtain for each
specimen figures indicating its calculated length at the time of forming first winter
ring. This calculated fir^t winter length is noted as I,, while similarly, the calculated
lengths or time of formation of second, third, and subsequent winter rings, will be
designated by ?^, I., etc. Subtracting Z, from L, we obtain the calculated increment
of growth during the time falling between the formation of the first and second winter
rings, this increment being denoted by i, and similarly, /,=?, — L t^^^l^-^l„, etc.

f). Length. — By the length of a herring is understood the distance from the point
of the snout to midway between the extreme points of the tail-fin when naturally
extended. The length may be measured in millimetres or centimetres, and a fish is
assigned to its nearest length-group; thus all fish, between the limits 29-5 and 30-5 cm.
length, will belong to the 30 cm. group.

IX. COLLECTING OF MATERIAL.

The methods of individual age-determination and growth-measurement naturally
suggest themselves as aids to the statistical investigation of the biology of the herring.
If, however, they are to be used for this purpose, it will be necessary to formulate pro-

CANADIAN FISHERIES EXPEDITION, 191>rl5 105

blenis suitable for treatment by statistical methods of work, and to procure adequate
observation material for statistical investigations. The question as to what problems
can be dealt with will largely depend upon the possibilities of obtaining material for
observation, in which respect, different waters will be found to vary, as we have here
to reckon both with the technical features (implements used, intensity of the fishery,
etc.) and also with natural conditions, which are not every^vhere alike. In the case
of each area investigated, therefore, it will be necessary to test the possibilities and
limitations of the methods, i. e., to gain experience by actual application of the
methods on the spot, and to utilize the results obtained by the work in determining
the possibilities of continued operations, and in drawing up plans for the same. On the
other hand, such determination will naturally be facilitated when it is possible to
compare the results of preliminary investigations in a new area with the experience
furnished by similar work elsewhere, which renders it easier, for instance, to distin-
guish between generalities and specific local features. In the following pages, there-
fore, an attempt will be made to sum up conclusions as to the possibilities for obtain-
ing material of a typical and representative character acquired during the Norwegian
herring investigations, which have now continued over a period of ten years.

First, as regards the age composition of the Norwegian herring stock, the inves-
tigations distinctly show that the stock in question does not appear as an even mixture
of every possible age-group; on the contrary, it is seen to be divided up into several
more or less markedly separated groups. These groups correspond in some degree to
the various " sorts " of herring, as known and distinguished by the fishermen, and the
appearance of the various groups at different places and times gives rise to various
kinds of herring fisheries. The most distinctly defined group is that containing the
mature fish, known by the fishermen as "large herring" when taken before spawning,
•and "spring herring" when captured on the spawning grounds and ready to deposit
their spawn. Another group is the " fat herring," which may be characterized as
herrings of moderate size still immature, and of excellent quality, whence the name.
A third group is that of the " small herring," i.e., small, young, immature fish, of
poorer quality than the fat herring. The habitat and migration of the various groups
are evidently different. The spring herring, for instance, crowd in to the west coast
in enormous numbers during the first months of the year. The fishermen have a
characteristic name for these close-packed shoals; they call them " sildebjerg "=a
" bjerg " or mountain of herring. It is a rare thing to find a young, immature fish
among these masses of mature, " full " herring. And on the other hand, fish with
genital products already developed are rarely found among the fat herring taken in
the northern Norwegian waters during autumn.

The individuals in a year-class move up, so to speak, from group to group as time
goes on, and as their development proceeds. The movement, however, does not take
place simultaneously in all individuals, so that a particular year-class may become
divided up, and fish of the same age will be thus encountered at different places and
times, in association with those of various other ages. The Norwegian herring mate-
rial offers several instances of this spreading of a year-class over different groups. In
1908, 1909, and 1910, fish of the 1904 year-class were found both among the mature,
full, herring on the west coast, and among the immature fish in the northern waters,
vide Hjort (Y) and Lea (XI). The investigations upon immature herring, in 1915,
showed that a distinct age-group occvirred within a restricted area, and at the same
time, associated partly with younger, partly with older, but still immature fish. This
feature will be seen illustrated in table 3.

It is apparent from table 3, that the fish 2i years old are in two samples
associated with almost exclusively older fish (one year older) whereas in the two
others, they are found in company with others, almost without exception younger
(one year younger). In the one case, they make up about two-thirds of the total
number in the sample; in the other, less than a third.

6551—11

106 DEPARTsMENT OF THE NATAL SERVICE

The recognition of this important phenomenon leads us to the conclusion that
an investigation of the age-distribution in a herring stock should include a study of
the different groups, and that the interchange between them should be most attentively
observed. One of these groups, that comprising the mature fish, has been under
observation for ten years in Norway. The experience gained during the course of
the work distinctly points to the possibility of following the age-composition, and its
variations, within this group. These investigations have, in methodical respects
especially, furnished interesting results; a brief description and discussion will
therefore here be given of the material collected for the study of this age-composition
and renewal of the mature group.

During the first years, from 1907 to 1913, only a small number of samisles, from
two to four, of the true full herring were collected annually, with, in addition, one to
three samples of the so-called large herring, i. e.i, mature fish with genital products
large but firm, which we now know to be very closely related to the actual spawning
herring. Despite the small number of samples, and of specimens investigated, the
samples were found to agree so closely one with another, and the features observable
with regard to age-composition so marked as to convince the investigators at work on
the material that even these few samples, and this small number of specimens, might
yet be taken as representative, so far related to a certain very important point in the
distribution of year-classes in the group of mature Norwegian herring. The main
point in question was the fact that the year-class 1904 exhibited a marked numerical
superiority over all others.

This year-class made its first appearance in any great number in a sample from
April, 1908. It was thereafter found, in every single sample investigated during the
years in question, to be enormously superior in numbers to all other year-classes. The
striking contrast between the numbers of this one year-class and those of the many
others, and the fact that this contrast was maintained for several years, served largely
to confirm the conviction in the minds of the investigators. In order to give the
surest possible foimdation for the observations;, a larger number of samples was
collected during the folio-wing years, from 1914 to 1916. The analysis of this material
confirmed most emphatically the presumption already arrived at, the year-class 1904,
despite its continually increasing age, being still found to occupy a dominant position.
One sample from the southern verge of the spawning grounds (Ivristiansand, February,
1914) contained, besides the 1914 year-class, another numerically strong age-group,
that of 1908 (vide Hjort V), but as this peculiarity was not obsers-ed in subsequent
samples from either the same area or elsewhere it was presumed that the sample in
question only represented a slight local disturbance occasioned by the immigration of
a new year-class, which supposition was later confirmed. The winter season 1914-15
passed, and still the 1904 year-class, with its now 11-year-old fish, was seen to pre-
dominate. The first samples then investigated, however, already indicated that the
1910 year-class would now come to occupy a distinctive position, being more numerously
represented than the adjacent classes, albeit far from equal in this respect to the old
1904 class. At first, also, there was but a slight degree of uniformity between the
different samples as to the numerical value of this new year-class. Not xuitil the
commencement of March, 1915, was it seen to be evident beyond question that this
year-class was decidedly richer than its older neighbour 1909, and in the two last
samples of the season, it even rivalled 1904, the last but one containing 48 per cent
1910 and 25 per cent 1904, the figures for the last sample being 33 per cent and 38 per
cent, respectively.

With these results in mind, the following season would appear to be doubly
interesting; in the first place there was the question, would the 1904 year-class, now
no less than 12 years old, continue to assert itself as heretofore; and, secondly, would
the new year-class, maintain the same distinctive position as in the two last samples
from 1915, or fail to maintain it as had been seen in the case of the 1908 sample from

CA^'ADIAN FISHERIES EXPEDITIOy, WL'rlo 107

Kristiansand. The first nineteen samples of large herring, in the winter of 1015-lt},
presented the same appearance as noted so many times l>efore, the year-class 1904 being
still predominant, and that of 1910 represented by only a few specimens. The twen-
tieth sample, however, the first of the true spring herring, was a great surprise, as it
contained one single specimen only of the 1904 year-class, and was otherwise composed
mainly of young fish. 6. 5 and 4 years old. i.e.. the year classes 1910 (with 26 per
cent), 1911 (18 per cent), and 1912 (43 per cent). Here, then, was the 1910 year-
class, but associated with two younger ones, of which the class 1912 especially was
present in force. The following spring herring samples (from February, 1916)
reverted once to the old style of composition, with 40 to 50 per cent 1904, but in
March this is gradually changed, and we have first the 1910 year-class, and later the
1912 class, accompanied by the less numerously represented intermediate year-class
1911. Fig. 26 shows the entire series of age-analyses for seine-caught spring
herring, the previous series of drift-net-caught large herring samples is here omitted
for want of space, but may be supplied by imagining the nineteen earlier samples of
fairly the same character as Nos. 2 to 4 in the figure.

A glance at the figure will show that the group of mature fish must have under-
gone a thorough change during the course of the season. At the commencement,,
there were evidently two very different sub-groups (represented by samples 1 and 2-4),
As the season goes on, however, these intermingle, so that the curves for age-distribu-
tion exhibit a highly peculiar appearance. Altogether, the different samples agree
very well one with another, and this despite the fact that the situation this year was
highly variable, and not, as hitherto, practically stationary.

The entire material of mature Norwegian herring, it will doubtless be admitted
distinctly indicates as possible the statistical recording of age-distribution and its
variations in this group of the stock. The results arrived at in methodical respects
from analysis of this material is, that the year-classes which have passed over into the
group in question become so thoroughly mixed that it is possible, even with relatively
few and small samples, to keep trace of the condition, when the same is, as during
the years 1910-13, mainly stationary. During periods where marked alterations take
place, as in 1908, 1914, 1915i, and 1916, the number of samples will need to be greater,
in order to furnish a view of the actual changes occurring. Thus, in 1916, no single
sample can be taken as representative of the conditions, and erroneous conclusions
would certainly have resulted had not several samples been available from various
parts of the season, and different fishing grounds.

It is interesting, from a methodical point of view, to consider somewhat more
closely samples from those periods when new year-classes began to make their appear-
ance. We have here, first of all, the immigration of the 1904 year-class, in 1908 ; then
that of the 1908 class in 1914; the 1910 class in 1915; and finally, the year-classes
1910-12 in 1916.

The 1904 year-class was first observed among the mature fish in 1907,. but only
in very small numbers; in 1908, it is found in the first sample from February, although
not numerously represented ; in the next and last sample from this season, however,
it makes up 65 per cent. In the sample from 1909, it amounts to about 40 per cent,
i. e. a decrease in the proportion. This rule, a rise of the percentage to a maximum,
followed by a fall, applies to all cases where investigation has been made. The 1908
year-class appeared on the south coast at first in great numbers, later on in the same
year the percentage was lower. The 1910 year-class reaches a maximum in the penul-
timate sample from 1915, with 48 per cent, in the last sample from that year percen-
tage is only 33 per cent. The immigration of a new year-class, and the intermingling
of the same with the older components of the stock, evidently takes some time, and
wovild appear to commence with the entrance of the immigrants in a body, which
wedges itself into the stock already on the grounds within a restricted jwrtion of the
same. The absence of the 1910 year-class during a great part of the fishing season

6551— Hi

108

DEPARTMENT OF THE NAVAL SERVICE

JCBC gi-oup 1915 12 11 10 09 OS 07 06 05 04 05 02 01 00 18»»
Age group 5 4 3 6 7 * <) 10 11 12 15 14 15 It 17

Age group 3 ^ ^6 7 6 9 10 H 12 li I-* 15 '6 17
.T»«r groi,p 1914 12 11 10 O* 08 07 06 05 04 05 02 01 00 1899

Fig. 26.

CANADIAN FISHERIES EXPEDITION, 191.'rl5

109

1915-16 also seems to suggest that the process of mixing is gradual and may require
some time, in other words, the group comprising the mature fish can at certain
periods be more or less heterogeneously composed, as was the case to a very high degree
in 1916.

Inv'estigations as to age among the Norwegian herring have been carried out
under favourable conditions, inasmuch as it has been possible to obtain samples not
biassed owing to tho methods of capture employed. Obviously, when gill-nets are
used, some doubt may easily arise as to whether the age-distribution, apparent in a
sample, may be more or less a result of the selective effect of the net itself, and a
sample of such netted herring cannot be credited a priori with the same representative
value as one taken with the fine-meshed seine. It was therefore fortunate for the
Norwegian investigations that the herring fishery of Norway happens to a great ex-
tent to be carried on by seine (shore and purse-seines), and these more reliable seine
samples furnished a starting point from which to investigate those taken in other nets.

The best means of carrying out such investigations would probably be to set nets
of different mesh out in a standing seine. This would furnish excellent material for
the purpose. Unfortunately no such experiments have hitherto been carried out, and
for the present, all we can do is to utilize the material available, and compare seine
samples with net samples taken at approximately the same place and time. In so
doing, however, we have to reckon with a disturbing factor, to wit, a certain relation
existing between time and place of capture, on the one hand, and implement on the
other. Thus drift-nets are employed for the capture of herring out at sea, and espe-
cially early in the season ; the whole of the " large herring " fishery is carried on with
drift nets. The shore seine, on the other hand, can only be used when the fish come
close in to land. Stake nets, again, are fixed on the bottom, while the purse-seine is
worked near the surface. The Norwegian material includes a number of paired sam-
ples more or less satisfying the above requirements. These are set out in pairs in
table 4 so as to permit of comparison between the age-composition in the samples.

Table 4. — Comparison between samples from seine and net-hauls. Each net sample
compared with the seine sample nearest adjacent in point of time and place.

Localitj'.

Svinfihavet.
Gursko. . . .
Karnisund. .
Rovccr

Karrnsund

Skudf^nes. .......

Inside Skudeiies.

Date.

Mar. 5, 1914
„ 26, 1914

Feb. 19, 1914
M 13, 1914
., 19, 1914
M 19, 1914
•, 18, 1915
,. 22, 1915

Gear.

Age

—Group.

Net.
Seine

Xet

—

4

0-9
1 1

5

2-3

3-1
1-9
2-5
15
2-0
2-5

(

2
2
5
3
3
3
4
5

1

3
2

5
7
9
5
1
1

7

4-6
7-7
0-2
30
5-4
40
8-2
8-0

8

50
8-4
100
50
5-4
5-4
5 4
4-7

9

13-9
l!l 4
16 6
'2-1
14 7
11 4
8-2
5-9

10

56-9
51-6
50-8
62!
60-5
60 7
17
8-9

11

4?
5-5
3 9
71
3-4
6 ••
53 1
61 5

12

2-8
1-8
1-4
1-6
1-7
1-0
1 4
1-3

13

3- 7
lo

1-8
19

0-

2 5

14

1-9
4

'o':>

811
5

15

9
0-4
7
0-7
0-3
1-5

16

Seine

Net

0-3
1-5

'o'7

Seine

0-8

0-8

4

The general impression given by the table is that the columns of net-samijle fig-
ures resemble very closely those for the seine-caught fish, such discrepancies as occur
being small and without apparent regularity. Particularly interesting are the
columns for the samples for Karmsvuid and KuvaT, where tiie two seine samples are
from hauls made on the same day and quite close together. The impression, furnished
by such paired comparisons between seine and net samples, leads us to the conclusion
that the age-composition of mature Norwegian herring has been very much the same
in the netted samples as in those taken with the seine. The same result is arrived at
if we consider the material as a whole, the whole of the netted material tending in the
same dir<^ction as that furnished by the seines. And bearing in mind the fact that

110

DEPARTMENT OF THE NATAL SERVICE

the samples as a rule include more than ten age-groups, this result cannot but be said
to be surprising. It might easily be imagined, for instance, that the difference in
size between the younger and the older fish would be so considerable as to involve a
degree of net selection far from negligible, the smallest and largest specimens avoiding
capture, whereby the net samples would be seriously biassed. That this proved not
to be the case, with the Norwegian samples of mature herring, is due to an important
biological phenomenon, which we shall now have occasion to examine more closely,
in considering the question of how best to procure representative material for the
study of growth.

As already mentioned, the individuals in a year-class may fall into several sub-
groups, appearing in company with older or younger fish. This subdivision of a year-
class does not appear as merely accidental, but seems on the contrary to be regulated
in a definite manner according to the stage of development at which the fish have
arrived. This is, of course, natural enough in cases where some individuals of a year-
class attain maturity and are ready to spawn, while others are still immature. Noth-
ing could be more reasonable than to suppose that the mature fish should part com-
pany with those of their contemporaries that have not reached that stage, and move
off by themselves to the spawning grounds. The year-class is thus divided up anrl
dissociated according to the degree of maturity of the genital products. We now find,
however, that the mature individuals in a year-class thus divided are of larger — often
considerably larger — size than those still immature. There exists a positive correla-
tion between degree of maturity and size. The year-class thus divided according to
degree of maturity will therefore also be found dissociated according to size, the
larger specimens of such a year-class will be found associated with older, mature fish,
the smaller with younger, immature companions.

This being ascertained, the question then arises as to whether the attainment of
maturity should be regarded as the only phases of development accompanied by such
dissociation in point of size among the individuals of a year-class.

Table 3. — Age distribution (showing percentage) in four samples of immature

herring from northern Norway, autumn 191.5. Group 3 (2§ years old fish) are

in two samples found in association with older herring, and in two samples
with younger herring.

Locality and Date.

Nnmber of
individuals
in sample.

Per cent belonging to different Age-Classe.s.

s

3

group 2

group 3.

group 4.

groui)5.

1

2
3

Kvaefjord, Oct. 15, 191.5

ariittavfer, Oct. 20, 1915

Ibestad, Nov. 7,1915

KvEefjord Nov 11 1915

Ifi9
194
259

205

%

6

2-6
88-0
68-3

%
74-0
03 9
120
31-3

%
24 9
320

4'

%
6
1-5

The answer here must be that the dissociation already mentioned (vide table 3) is
also regulated according to size; we find, for instance, that the average length of the
2J-year-old fish (group 3 in table 3) is considerably greater in the two first samples,
where this age-group was associated with older fish, than in the two last, where it was
found in company with younger fish (22-1, 22 -61, 17-0 and 18-3 cm., respectively).
The four samples in the table are from hauls made within a restricted area, the
greatest distance between places of capture being about 40 km. and within a short
space of lijne, vi?., from 15th October to 11th November. An even more striking

CAXADI.iy FISHERIE^^ EXPEDITIOy, 191',-lo

111

example of this size-regulated dissociation, among immature fish, is furnished by two
samples from hauls made at an interval of eight days and 21 km. apart. Table 5
shows the age-distribution in these two samples, and the difference between the average
sizes of the year-class 1913.

Table 5. — Showing difference in average size of herring belonging to the same year-
group (1913) in two samples. In one sample the herring considered were found
associated with younger herring, in the other with older herring.

Locality and date.

No. of

indiv id. iu

.sam pie.

254
177

Percentual a^e composition.

Average

length of

group 1913.

Year gr.
1914.

Year gr.
1913.

Year gr.
1912.

Year gr.
1911.

Bergsvaag, Trondences, Sept. 10 .
Tennvik, Sept. 18

89
6

11
63-3

" "35-6

o'e

IG .0 cm.
227 ..

Finally, it should be mentioned, that mature, full specimens of the same year-
class were also found among the spring herring in the spring of 1916, the average
size of these was 25-8 cm., or considerably in excess of that found in any sample of
immature herring, where the maximal average size was 22-7 cm. Neither the differ-
ence in size nor in degree of maturity, can easily be explained as a result of growth in
the period between late autumn 1915 and the spring of 1916, the discrepancy being too
great, the time too short, and the season, as has been shown by experience, being a
period of stagnation. It is therefore most reasonable to suppose that the said year-
class was dissociated into at least three groups, one consisting of small fish, associated
with younger ones; another of large, but immature individuals associated with their
seniors; and finally one comprising those of large size and completed maturity,
associated with the adults.

The remaining two important year-classes in the material, viz., those of 1914
and 1912, do not exhibit fluctuations so violent as in that of 1913; there is, however,
also here a suggestion of dissociation according to size, wherefore it would be well to
reckon with the possibility that a year-class may become dissociated and grouped
according to size of individuals throughout the entire period of growth until full
maturity is attained.

What takes place after this time, when all surviving individuals of the year-class
have reached maturity, cannot be stai;ed with certainty as yet; there is, however, reason
to believe that the process of maturing should be regarded as a phase of development
which, when completed, leads the separate components of a divided year-class to
reassemble, i. e., that the sub-groups formed by the varying rate of development will,
after maturity and subsequent spawning, reconsolidate into a whole.

At any rate, it would seem, from the results of the Norwegian growth investiga-
tions, that the components of the rich year-class 1904 — which, from its numerical
importance and pecularities of growth, furnishes excellent material for a study of this
question — did reassemble after their separation during the process of attaining matu-
rity, and were later found mingled and collectively in the samples of mature fish.
The observations made from year to year present, when seen together, the picture of
a process terminating in a fairly stable mixture of the heterogeneous elements which
compose the year-class in question. This is especially noticeable in the case of that
growth-dimension which exhibits the greatest and most peculiar variation, viz., the
increment for the third year of life (^,). Fig. 27 shows curves of frequency for this
dimension, the observation? of each year being noted separately, and subdivide'1 into

112

DEPARTMENT OF THE NAVAL SERVICE

1 ' 2, 3 -<► 5 6 7

Fig. 27. Mature herring. . ,

a 9 10 cm.

Immature herring.

CA:^ADIAy FISHERIES EXPEDITIOy, 19Vrl5 113

two categories, the one embracing a sample from the spawning grounds taken during
the spawning season (the spring fishery of the west coast) and the other including
immature fat herring taken in autumn on the north coast of Norway.

It will be noticed that the individuals found among the grown fish in 1908 have
a considerably greater t^ than the autumn-caught immature fish from the north, a
similar difference is also observable in 1909. In 1910, however, the curve for the ma-
ture fish exhibits a different course, and is here distinctly bimodal ; during the autumn
of the same year, only a few specimens of this year-class were found among the im-
mature fish, and the following year none. After 1910, the curve maintains its pecu-
liar form, with the two modes, the position of which, as will be seen, corresponds to
the two modes in the two pairs of curves for 1908 and 1909.

Similar conditions will be found to apply in the ease of the remaining growth
dimensions (^i t2 ti fs) save that the original difference between the two categories is
not so great as to produce a final bimodal curve, but only a compound curve.

From the foregoing, it will be seen that the biological process which is termed
the dissociation of year-classes is a phenomenon which must be taken into consideration
in statistical investigations of the growth (and age) of herring. The phenomenon
itself would seem, as far as can be judged up to the present, to be of considerable
dimensions, and to be closely related to other phenomena such as maturation, migra-
tion, accumulation of fat, etc. It is therefore, in my opinion, worthy of closer study,
in the prosecution of which growth measurements may be of assistance, in view of the
relation between dissociation and size (growth).

In the case of investigations dealing with the problem of growth during different
years, or in different waters, the phenomenon will appear as a complication, a new
and variable factor. It will therefore be necessary, in such investigations, to procure
material, the elements of which are as far as possible comparable. And as circums-
tances are, the best material would be that consisting of older fish, as offering greater
facilities for the procuring of representative samples.

With regard to the question as to number and size of the samples, it will be seen
from the foregoing that various points need to be taken into consideration here. The
number of samples, and their size (i. e., the number of individuals contained) must
be determined according to the peculiarities of the stock to be dealt with, or of the
water under investigation, or of the problems which it is desired to solve. In the
case of the immature Norwegian herring, for instance, where, as we have seen, the
year-classes occur in several combinations, and dissociated according to size, a large
number of samples will be necessary; owing to the small number of year-classes re-
presented, however the samples need not be particularly large. The mature Norwe-
gian herring, on the other hand, in 1914 might, as we are now able to see, have been
dealt with through fewer samples than were collected, owing to a comparative stabi-
lity in the distribution of the year-classes with a single dominant group, which mark-
ed this year, as compared with the variability of the situation in 1916, where several
new year-classes cropped up.

A water area containing several different tribes of herring will in particular re-
quire a larger number of samples than one where simpler conditions prevail.

In dealing with problems which demand that the values operated with (e. g.. the
mean values for growth dimensions) shall be accurate, i. e.. with the smallest possible
degree of accidental error, it may be necessary to make the sample larger.

It is impossible to lay down any definite rule or system for determining this ques-
tion, as the circumstances to be considered vary from place to place and from time
to time. In most cases, however, we may say that given a certain niimber of speci-
mens to be examined, there will be more chance of obtaining the best material when
these are distributed among a largen number of smaller samples, then if they are
massed in one or a few large ones. In the Norwegian investigations among the
grown herring, the material was at first collected by taking a few samples each year,

114 DEPARTMENT OF THE NAVAL SERVICE

each sample containing a relatively large number of specimens. As the work went on,
however, it was found desirable to collect the material on a more extensive basis, with
more samples, distributed as widely as possible throughout the area and the season
embraced by the fishery. This involved an alteration of the methods employed, and
a number of the samples thus collected were now subjected to a simpler process of
examination (vide p. 133) in addition to which the question as to possibility of reduc-
ing the number of specimens in each sample was considered and tested. With this
new method, which was tried and adopted in 1914, it was found that for the object
in view, and under the conditions then prevailing among the stock, comparatively
small samples might well be used. An average of some 200 specimens per sample was
seen to be sufficient, and it was also realized that the advantage gained by operating
with larger samples would be altogether disproportionate to the extra work involved,
as the discrepancies between the samples could not be essentially minimized thereby,
and may be due to other causes than fluctuations of sampling.

It should be emphatically pointed out, however, that the prevailing situation was
exceptionally favourable for the work of age investigations, the contrast between the
one enonnously rich year-class 1904, and all the others, being so great that the curves
for age distribution would maintain their characteristic appearance even with con-
siderable errors of sampling. In other words, the age composition was so characteris-
tic, that it was advantageous to work with small samples, and consequently greater
accidental fluctuations, as this rendered it possible to deal with a greater total number
of samples.

In cases where the age distribution is less characteristic, e. g., where several
successive year-classes are more or less equally strong, accidental fluctuations of
samples may impair the agreement between the samples, or at any rate, render it less
obvious. In such cases, therefore, it may be desirable to reduce the extent of the
fluctuations of sampling by increasing the number of specimens in each sample. It
was in anticipation of such a possible change in the situation that the number of
specimens per sample was increased to about 400 in the case of the Norwegian spring
herring investigations in 1916.

The situation prevailing during the past few years among the Norwegian spring
herring has not been quite so favourable for growth investigation; when so great a
percentage of the specimens examined is derived from a single year-class, the result
is an almost superfluous quantity of material confined to the group in question, with
a corresponding reduction in the amount for growth observations for the many other
year-classes represented. Under these circumstances, it would require a dispropor-
tionate amount of work to procure material in which every single year-class should be
represented by the full number of specimens desirable for growth investigations.

CAyADIAy FISHERIES EXPEDITIOX, lOl'rlo 115

B. SPECIAL.
THE CANADIAN MATERIAL.

X, Description A^D pkeluiinary grolpixg of the Caxadux materl\l.

The chart, lig. '2>i, ou page 117 will serve to locate the samples placed at my dis-
posal and dealt with in this chapter. The material comprises a large number of
samples from au area embracing the Atlantic coast of Canada, the gulf of St. Law-
rence, and the coast of Newfoundland.

The comparatively large number of samples, and their distribution throughout
different seasons, counterbalances, at any rate in the case of certain waters, the dis-
advantage arising from the fact that the number of individuals in each sample is, as
a rule, but small. It will, however, probably be desirable in the case of future investi-
gations, to secure larger samples, as the stock in these waters has been found to in-
clude an unusually large number of age-groups. Nowhere have fish of so great age
been found in such considerable numbers as here.

Bj' far the gTeater portion of the samples originates from catches made with
implements, viz., gill-nets, which cannot be regarded as particularly suitable for the
purpose ol obtaining representative material. As mentioned in the foregoing, samples
of netted herring must necessarily be less reliable tlian samples taken with non-select-
ive implements, especially where no opportunity occurs of making comparison, as in
Norway, with samples of the latter sort. Some doubt may therefore arise as to whether
the younger and smaller fish are represented in their due proportion in the sam-
ples. There are, however, as will subsequently be shown, certain peculiarities in con-
nection both with age-distribution and growth, which cannot be explained as merely
resulting from the selective effect of the nets.

In addition to the samples of grown herring from various places and taken at
various times, we have also a number of samples of young, immature fish, drawn
partly from catches made by steam drifter No. 33 with the drift-net, during the sum-
mer of 1915, and partly also from trap-catches. The importance of these samples of
immature herring lies not so much in their representative value, which may be oi)en
to considerable doubt, not only on account of the method of capture, but also because
the experience gained in the course of the Norwegian investigations has shown that
some caution is necessary when dealing with samples of quite young herring. They
form, however, a valuable supplement to the material of gro^vn fish, as furnishing an
aid to to the determination of the time when summer growth begins, and also because
it has been found that these young fish exhibit growth variations for the different
waters exactly similar to those noted in the case of the grown individuals. These
young samples, therefore, add to the value of the remaining material, and give the
results a more general character.

From the very first, on going through the scale-samples it was strikingly evident
that the material must embrace several different, in some cases strikingly different,
"sorts" of grown herring. The difference between fish from the different localities
made itself apparent partly in the more or less distinct marking of the annual rings
on the scales, rendering number and distance more or less easy to read, partly in the
fact that the dominant age-groups (year-classes) differed in samples from different
localities, and finally, in the different character of the growth, as indicated by the
scales. It therefore seemed natural, after this first survey, to make a preliminary
division of the material according to locality. This gave four groups of samples, as
follows : —

116 DEPARTMENT OF THE NATAL SERTICE

1. Samples from the waters about Prince Edward Island, i. e., west and
south of a line drawn from cape North, Cape Breton, to the eastern point of
Prince Edward island, and thence on to cape Gaspe in the province of Quebec.

2. Samples from the waters about Magdalen islands.

3. Samples from the coast of Newfoundland (all from the west coast, ex-
cepting one from White bay).

4. Samples from the Atlantic side of Cape Breton island. Grand Narrows,
Ardoise, coast of Nova Scotia, Bay of Fundy, and from Gloucester, Mass.

The following pages will be devoted to a description of the results obtained from
i.ge-determinations and growth measurements, arranged on the basis of the above
group-division. The first question to be dealt with will be, how far fluctuations occur
in the relative numerical value of the year-classes, and if such occur, whether they
become apparent in the same manner throughout the entire area embraced by the
samples, or whether they vary in the different waters. Thereafter, the growth in the
various localities will be discussed. And finally, by summing up the information
obtained from the study of the age and growth, and discussion of the material avail-
able as to number of vertebrae, etc., an attempt will be made to deal with the question
as to the possible existence of different races or tribes of herring in the Canadian
waters.

XL— AGE.

a. The tvaters about Prince Edward Island.

From these waters we have fiv^e samples of grown fish, taken by gill-net at different
places in Northumberland strait; two of these are from May, 1914; the remaining
three from May, 1915. In addition to these, there are also two samples of herring
caught in traps in the neighbourhood of Souris during the summer of 1915, and eight
samples from drift-net catches made during the cruises of the drifter No. 33 at various
places in the gulf, in June and July, 1915. The chart, fig. 28, shows the localities.

The trap-and drift-net catches contain a quantity of young specimens, and on
examining scales from these we find as will subsequently be shown, that the summer
growth for 1915 commenced some time in June in the case of the young fish, probably
during the second half of the month. It will therefore be reasonable to suppose that
the older and grown specimens in the samples from May had then not yet commenced
their new summer growth, and that, consequently, the last summer zone in the scales
would represent the growth of the fish during the summer of the year previous t^
capture.

The other samples, containing older fish, date from the time between the end
of June and the end of July. With regard to these therefore, we cannot be so sure as
to whether summer growth has commenced, or if so, whether it has commenced in all
cases. Some doubt arises in assigning the fish to definite year-classes (vide p. 119),
in addition to which the length of drift-net used by No. 33 was composed of greatly
varying widths of mesh, and thus rather calculated to take herring of all sizes than to
take them in the natural proportion between the different size-groups. The value,
therefore, of the samples in studying the proportions of the different year-classes is
somewhat difficult to determine, and the samples of grown fish from Northumberland
strait have consequently been kept apart from the rest. Fig. 29 shows the percentual
age-distribution in these five samples.

It will immediately be noticed that the implements have here succeeded in
capturing fish of most widely differing age, from specimens 4 years old to individuals
of (apparently) some 20 years, a circumstance resembling that noted in the case of the
drift-net hauls of Norwegian grown herring, but in the present instance perhaps even
more marked. And although we here lack such means of gauging the value of the

CANADIAN FISHERIES EXPEDITION, Idl.'rlS

117

samples as is afforded by comparison with samples taken with non-seleetive imple-
ments, yet the circumstance in question gives us further grounds for hoping that the
fish of medium age at any rate have been taken in something like their natural pro-
portion. This presumption is further confirmed on examination and comparison of
the curves for the different samples. Here, as will be seen, the different age-groups
are represented in by no means equal numbers, and what is more, the samples them-
selves agree in this respect to a degree which, considering the small number of si>eci-
mens and the largo numbers of age-groups, may be taken as very satisfactory.

Fig. 28.

The curves being age-curves, there is, of course, transposition from 1914 to 1915
of the two characteristic modes; in 1914: there are comparatively many fish aged 7
and 11 years, in 1915 many of 8 and 12 years old, whereas the age-groups 6, 8, 9 and
10 in 1914, and 7, 9, 10 and 11 in 1915 contain relatively few specimens.

In the case of the younger fish, we do not find the same agreement in the different
samples. It must in all probability be taken as due to accident that we find, for the.se
ages also, an almost perfect agreement when we take the two 1914 samples! together
as one. and compare the resulting age-curve for the whole with that for the three 1915
samples together. This is shown in fig. 30, where the grouping is arranged according
to year-classes, and not in order of age.

118

DEPARTMElsT OF TEE t^AYAL SERVICE

Age classes

4 5 6 7 8 9 10

II 12 15 14 15 16 17 18 19 20

-I 1 1 1 1 1 I 1 I 1 1 1 I

Northumberl.
5tr.

1914

2°^ 110 Individ
10

Pictou Hr 1915

CAyADIAX FISUEKIEH EXPEDITIOX, Wl-Ho

119

It will be seen that the 3'ear-classes 1903 and 1907 were considerably more nume-
rously represented in the samples than the other year-classes, and especially charac-
teristic is the difference between the " good " year-classes 1907 and 1903, on the ono
hand, and the three intermediate years 1904, 1905, and 1906. The older year-classes
are represented by few specimens, whereas the younger ones, as mentioned, vary from
one sample to another.

The small number of specimens and of samples, and the nature of the implement
of capture, are points which diminish the representative value of the material. On
the other hand, the marked degree of similarity between the samples, the characteris-

1911 10 09 08 07 06 05 04 05 02 01 00 1899 98 97 96 95

■ 30 %

- 10

19U

218 ind ivid

1915

525 Individ.

1911 10 09 03 07 06 05 04 03 02 01 00 1899 98 97 96 95

Fig. 30.

tic transposition of the modes of the curves for 1914-15, and not least, the peculiar
course of the curves themselves, are features which must be taken as favouring the
supposition that the peculiarities noticeable in the material actually reflect typical
conditions. A bimodal or multimodal age curve will probably, as a rule, arise not
on account of, but despite the selective effect of the implement of capture.

We cannot, however, take it for granted that net samples, even though exhibiting
variations in the relative strength of the year-classes, present these variations in their
correct numerical proportion. If we cannot be sure that the nets have taken the
younger year-classes (1911-1910-1909) in their due proportion, neither can we be cer-
tain that the two strong year-classes are correctly represented in this respect, the nets
might well be supposed to have retained a relatively greater number of the one group
than of the other, owing to the difference in size. This question could be decided by
continued investigations, or by procuring material from hauls made with non-solect.ive

120

DEPARTMENT OF THE NATAL SERVICE

implements. A slight step in this direction may be made by considering the samples
taken during the cruise of No. 33 and the samples from Souris. Most of the samples
contain many young and immature fish, but, in five, some old fish were found,
viz., in the samples from stations 27, 28-29, and 42, and also in one of the trap samples
from Souris. The three first stations are situated between the western side of Prince
Edward Island and the Gaspe coast, while station 42 lies north of cape George {vide
chart, fig. 28).

Table 6. — Age distribution in three drift-net samples taken off the Gaspe coast
(stations 27-29), one trap sample from Souris, P.E.I., and one drift-net sample
from west of Port Hood, Cape Breton (station 42), June, 1915.

Locality and Date.

Station 27, 48° 21' N., 63" 57' VV., June

28-29.
Station 28, 47° 56' N., 63" 27' W., June

29-30
Station'29, 47" 34' N., 64° 12' W., June

30July 1.

Souris, June 7 . . . .

Station 42, 46^ 0' N., 61° 56' W., June

15-16.

Number

of
individ.

82
30
95

117
133

Age-groups — Year-groups.

3 4 5 6 7 8 9 10 11 12 More

1912

1-2

3-3

3-2

25-6
07

1911

13 4
40

68-4

36-8
64 7

1910

1909

9 8

67

3 2

60
38

1908

1907

13-4

3 3

h 3

6
15

1906

1905

4-9
6-7

4 :^
07

19f4

6 1
100

26

1903

3 3

Table 6 shows the percentual distribution of year-classes in these samples. As
already mentioned, there may be some difliculty in determining whether new summer
growth has commenced in the case of the older fish. In these samples the scales of
some of the younger specimens distinctly show an incipient new growth, whereas in
others, judging from the breadth of the last summer zone, it would seem that the
new simimer growth had not yet commenced. This being the case with the younger
fish, one would hardly expect the scales of the older fish to exhibit any commencement
of new summer growth, and, as a matter of fact, the last summer zone on these old
scales was found to be of about the same breadth as the previous one.

The age-tables have therefore been drawn up accordingly, and the column farthest
to the left thus includes specimens, some of which exhibit the number of summer
zones indicated in the heading, and others an additional narrow zone beyond this. In
by far the greater number of the older specimens (from group 7 and upwards) the
last summer zone on the scales is taken as representing the summer of 1914. This
method of arrangement should, as regards all that is here essential, be correct enough,
albeit some doubt may exist in the case of fish of a medium age.

It will be seen from the table that the fish assigned to the 1903 year-class are
relatively numerous, especially in the samples from station 27 ; in three of the samples,
the specimens assigned to 1907 are relatively numerous (stations 27-29, and the samples
from Souris). The sample from station 42, on the other hand, exhibits no resem-
blance to the five gill-net samples from Northumberland strait ; this sample we shall
later on have occasion to consider in another connection, and it is merely mentioned
here in order to note its difference from the samples from Northumberland strait.

All samples from these waters point, when taken together, to the correctness of
the svipposition that the 1903 and 1907 year-classes were present in greater numbers
than the intermediate year-classes of 1908, 1909, and 1910.

CAyADIAX FISn FRIES EXPEDITION, lOl'rlo

121

With regard to the younger fish, it is not so easy to arrive at definite conclusions
on the basis of the material available; it is always, moreover, a far more difficult task
to ascertain the relative strength of the different year-classes among immature
herring.

Nevertheless, a glance at tables 6 and 7 will immediately reveal the fact that only
a few specimens of the year 1912 year-class were tiiken, whereas the 1911 and 1913
year-classes were particularly well represented.

Table 7. — Showing age distribution in samples where young and immature fish

predominate.

Locality and date.

>

1

38
73
47
47
33
32

Age -groups.

Ypar-groups.

2

3

4

5

6
1910.

' 55

28- 1

More.

1914.
37 5'

1913.

789
520
830
1000
42 4
3 1

1912.

18 4

8-2

17

57-6
21 9

1911.

' 3o'i

9 4

Station 33, 48" 13' N. 63" 58' W. July 5-6.
.. al, 47° 52' N. 63° 57' W. ., 1-2.

34, AT 13' N. 64° 21' W. „ 6-7.

35, 46" 51' N. 64° 25' W. „ 7-8.
Souria „ 1 1.
Station 53, 45' 46' X. 62= 23' W. .. 30-31.

2 7
4 1

h. The waters about Magdalen Islands.

From this area, we have a gill-net sample taken in May, 1914, one from May,
1915, and further a very small sample (which is of no value in this connection), with
finally, one from a drift-net haul made by Xo. SS about end of July. 1915 (station 49).
It will be as well to deal with this last sample first, as an examination of the scales
here shows that the younger fish (4- and 5-year-olds) therein contained, had doubtless
commenced a new summer's growth but had made only very slight progress therewith
the new summer zone appearing in most eases as a very narrow belt outside the
broader zones for the previous years. This would seem to indicate that the summer
growth in these waters begins late, as is also the case farther to the west and south.
In the case of the two samples from May, 1914, and 1915, therefore, we may take it
that the last summer zone on the scales represents the summer previous to the year of
capture. For the older fish, however, in the July sample mentioned, it will be a matter
of doubt whether these have commenced their new summer growth or not. In table
8, therefore, only the two youngest age-groups have been assigned to year-classes, the
headings for the older groups indicating only the number of rings on the scales.

Table 8. — Age distribution in samples from station 49, 70 specimens. Year-class noted

only for the youngest fish.

Age-groups

Year-groups.

5

6
1910.

7

8

9

10

11

12

13

14

1911.

1909.

1908.

21-4

14-3

7-1

37 1

2 9

2-9

2-9

71

5-7

14

6551—12

122

DEPARTMENT OF THE NAVAL SERVICE

It will be seen from the table that a considerable number of specimens belonged
to the 1911 and 1910 year-classes, and many fish occured with eight rings on the scales,
with finally a more indistinct accumulation of individuals in groups 12 and 13.

If we now turn to the two samples from May;, 1914 and 1915, we find, as shown
in fig. 31 that one from 1914 contained many 11-year-old fish, while in 1915 there were
many of 12-years-old. The 1903 year-class that is to say is relatively numerous in
both samples. In the 1915 sample, moreover, there were many specimens of the 1911
and 1910 year-classes, a feature common to this sample and that from station 42,, and
as regards the 1911 year-class, also to the several of the samples from the waters
between cape Gaspe and Prince Edward Island. On the other hand, in these two

Age classes

5 ^ b 6 7

9 10 11 12 15 1i^ 15 16 17

samples from the Magdalen islands, we do not find that quantity of fish belonging to
the 1907 year-group which was so characteristic of the samples from Northumberland
strait. What signifiance should be attached to this point of difference it is impossible
to say until further material is available; one thing, however, is certain; the samples
from the Magdalen islands exhibit no small likeness to those from Northumberland
strait and from off the Gaspe coast.

It is therefore open to doubt whether the division here made is in accordance
with the actual conditions, which question will be discussed when the growth investi-
gations have been dealt with, and in connection with the treatment of racial characters
(number of vertebrae, etc.)

c. The Ne%vfoundland waters.

The material from these waters comprises five samples from the spring of 1914,
four from the autumn of the same year, and three from the spring of 1915, making

CAyADIAX FISH t: HIES KM'KDiriOS, 191 'rlo

123

twelve samples in all, of grown fish, all caught with the gill-net. In addition to these,
there is also a sample of quite young herring from the cruise of the steamer No. 33 in
the summer of 1915; this sample however, isolated as it is, will be of no importance
as regards the questions here dealt with. The samples are all from the west coast,
with the exception of one from White bay.

Figs 32, 33 and 34, present, in graphical form, the results of the age determina-
tions, the material being here divided into three groups, according to season.

/^ge groups

10 11 12 15 14 15 tfe 17 18 19 20 21 27 21

Spring 1914

White Bay . pnmo May
156 mdivid

Bay of Jslando, spring
■72 indiv.d

George Bay spring
53 'ndivid

Georges Bay primo May

S? Georges Bay orimo '^zy
156 ndivid

Fig. 32.

The Xewfoundland samples, like those from the ^Fagdalen island and Xorthum-
berland strait, include a considerable number of age-groups, the age of the specimens
varying from 4 to over 20 years. Only very few of these groups, however, are at all
numerously represented; in all the samples save one, the fish are found for the greater
part massed in a particular group, to wit, group 10 in the spring samples from 1914,
and group 11 in the later ones. Taking it for granted that summer growth in these
waters, as in those farther to the south in the Gulf, has not eonmienced by May, then
the age-group 10 will in the spring of 1914 answer to the year-class 1904, and similarly
presupposing that the new growth commences some time during the summer, then
age-group 11 in the autumn samples from 1914. and those from the spring of 1915
will be likewise equivalent to year-class 1904. All the samples, with the single excep-
tion named, thus bear witness to a state of things similar to that noted in the case
of the Norwegian stock, i. e., a single year-class taking up a dominant position, which

6551— 12i

124

DEPARTMENT OF THE NAVAL SERVICE

is maintained for several seasons. Fig. 35 shows the distribution of the year-classes
during these three seasons, as revealed in the samples, the single divergent sample
from 1915 is here omitted, but will be referred to later on, as it is of considerable inte-
rest. We have noticed, in the three foregoing figures, that no essential difference was

Fig. 33.

/\ge qroupi

^ 5 6 7 8 9 10 t1 12 I i 14 15 16 17 18 19 ?0 ?1 2? a

■ 50 %

A

■ 40

/ \ Spring 1915

■ iO

/ A\

• 20

/ / \\

1,0 ^

/ / \ \ Bay ot Jslands May £
1 \\ 115 individ

■ 20%

/ \

'° y

/ \ Bonne Bay May 11
\ 106 individ

20% A

10 /^\/ \_^^^

__y/ \ 5' Georges Bay May 27
\ 102 individ

i, 5 6 7 8 9 10 11 12 13 14 15 16 17 16 1« 20 21 22 2S
/\qe groups

Fig. 34.

apparent between the samples from different localities. Here again, in the present
case, the samples from the three seasons are seen to differ but very slightly one from
another, save perhaps for the fact that the autumn samples seem to include a greater
number of jouns; fi.'--h than those from the spring. The difference, however, is by no

CAXADIAX FISHERIES EXPEDITIOX, 1914-15

125

means considerable, and it is possible, moreover, that the part of the curve indicating
the numerical value of the young fish may in reality be not altogether accurate, owing
to the nature of the implements employed. As a matter of fact, therefore, nothing
definite can be said as to the relative frequency of the young fish here.

The age-groups in the samples have, as already indicated, been carried over into
year-classes; this is analytically supported by consideration of the age-curves result-
ing from the spring samples from 1914 and the autumn samples of the year, with the
similarity between these latter and those from the spring of 1915, distinctly pointing
to tlie fact that the new summer growth must have commenced some time after the
month of May. The strong contrast between of 1914 and groups 11 and 12 in the

Year groups

1911 10 OV 0» 07 06 05 0* Oi 02

00 1899 98 97 96 95 9<. 9J 92 91

Spring of 1914
512 individ

Autumn ot 191A
462 mdivia

Spring of 1915
219 mdivid

1911 10 09 08 07 06 05 Ct 05 02 01 00 1899 98 97 96 <»5 <><, 95 <i-^ 91
Year groups

Fig. 35.

spring of 1915 suggests that the summer growth has not made itself apparent on the
scales at that time of the year. Had such been the case, the contrast would certainly
have been effaced by the division of each year-class into two age-groups, according as
they had or had not commenced their new summer growth. The divergent sample might
possibly to some extent be explained by supposing that some of the individuals therein,
contained had commenced new summer growth; the sample in question was also taken
late in May. But as will later be shown, it is more likely that it represents another
group of herring than those in the remaining samples. This sample apart, the mate-
rial presents an apparance totally different from that furnished b.y the samples from
^Magdalen islands and Xorthumberland strait; moreover, as we shall now have occa-
sion to show, it differs likewise in point of age distribution from that collected on the
Atlantic coast. i ' .

126 DEPARTMENT OF THE NATAL SERVICE

d. From Cape North southwards along the Atlantic coast of Canada.

The material from these waters embraces seven samples of mature herring, and
four of immature fish, covering a range from Sydney, Cape Breton, to Gloucester,
Mass. In going- through the scale preparations from these samples, it was strikingly
noticeable that the scales were considerably more difficult to read than those of the
specimens from the gulf of St. Lawrence. The first and second winter rings were often
found to be faint, thus rendering less distinct the contrast between the true winter
rings and the secondary rings, which latter were of frequent occurrence in these
scales. The rings beyond these may be said to be fairly distinct, but the outermost
ones, on old scales, were again difficult to discern owing to the fact that the fine ridges
on the surface of the scales were often irregularly developed, a symptom of senility
in herrings. Another feature, which will be more important in connection with the
growth investigations, was the frequent irregularity in the proportion between longitu-
dinal and lateral growth of the scales. I have made no measurements in this respect,
but it could often be seen that the scales had during the last few years grown more in
breath than in length, thus rendering the distance between the winter rings relatively
greater, at the two sides of the scale, than in the anterior portion.

The pecularities here noted in the scales of herring from these Atlantic waters
tend to give the investigator an impression that the fish in question must have lived
and grown up under conditions widely different from those which obtain among the
herring in the gulf of St. Lawrence; true it is no easy matter to tabulate and sys-
tematize such features, as distinctness of winter rings or frequency of secondary
rings, the appearance actually presented is, however, none the less remarkable on
examination of the scales. In estimates (and growth measurements) this means
increased labour, and greater uncertainty in the results obtained.

With regard to these difficulties, there appears to be some difference between the
samples from the northern parts of the waters in question, and those from Nova
Scotia, the scales of herring from the more southerly localities being less easy to read
than the others. The material is too limited, however, to i>ermit of any definite
statement in this respect, as much depends upon the quality of the scale preparations.

The difficulties are most felt in dealing with older fish, for the younger specimens,
up to 8 years old, it may safely be said, that a fairly accurate age determination can
be arrived at, with continued investigations, also, and by great care to make the scale
preparations as perfect as possible, a similar degree of accuracy may be attained in
the case of the older fish. Fortunately as matters stand, the material from Nova Scotia
includes only one sample with many old fish; in the others, young specimens pre-
dominate.

As regards the northern samples (North Sydney, Main-a-Dieu and Grand Narrows)
we have the further disadvantage that they were taken during May and June, i. e., at
a time when the herring in these waters may be expected to be commencing their
summer growth. And, as we here lack the support afforded by study of the quite young
fish, and as the samples themselves do not form any series in point of time, it is diffi-
cult to assign the specimens to definite year-classes. In table 9, therefore, where the
results of age-determinations for these three northern samples are shown, the speci-
mens have only been arranged in the age-group, or as one might say, " summer-zone
groups ". As far as it is possible to judge, however, the great majority of the speci-
mens in these samples have not yet begun their summer growth.

CAXADIAX FISHERIES EXPEDlTTOy. lOl'rlo

127

Table 9. — Showing age distribution in samples from Cape Breton, north side. New
growth probably not commenced, so that group 4 should correspond to year-class
1910, in the 1914 sample, and to year-class 1911 in the samples from 1915.

Lo ty and date.

Main-a-Dieu, May 1914

North Sydney, June 3, 1915...
Grand Narrows, .Tune 2, 1915.

Number

of

individuals.

54

98
88

Age-groups.

1-9
5 1

18 2

9-3
12 2
27-3

3 7
01
91

HI

24 5

4-5

7 4

10 2
10 2

9 10 11 12 More

7-4

8-2

10 2

14 8
10
3 4

31-5
17 3
12 5

7 4

12-2

2 3

4 6
2
2 2

The sample from Main-a-Dieu, 1914, and that from North Sydney, 1915, have this
point in common, that groups 10 and 11 in 1914 and groups 11 and 12 in 1915 contain
a relatively large number of specimens. These should presumably be the 1904 and
1903 year-classes. The similarity is not, it is true, either here or elsewhere, remarkably
great, but we have to consider the small number of individuals in the one sample.
The sample from Grand Narrows has, like that taken at the same time from North
Sydney, a large number of fish with five rings, but diifers not a little from this ; there
is, however, here again some massing of the 11 and 12 group fish.

% »

■ ^° A

/ \

/\ Bay of 5' George

• 10 /V \

/ \ May 27 1915

^"-^-Z \ 102 indiv

\ .^/•'•^ ■*-*^

0/ A

/o A

•20 / \

/ \

|\ North Sydney

^°/\ / ^

/ \ June 3 1915

/

\ / \ 98 mdiv

4^ S 6
Age group

10 11 12 13 Vt 15 16 17 18 19

Fig. 36.

On the other hand, if we compare the sample from North Sydney with the one
divergent sample from the Newfoundland area, taken a week earlier, we find the most
perfect agreement. The curves for these two samples will be seen in fig. 36. The
striking agreement between these two samples, on the one hand, and the exceptional
position occupied by the Newfoundland sample among the remainder from that area,
on the other, greatly tend to support the presumption that these two samples represent
one and the same group of herring, in which the age-comix)sition is widely ditferent
from that of the Newfoundland fish. The two samples in question will therefore be
more closely compared in the chapter on growth.

128

DEPARTMEXT OF THE NATAL SERVICE

From West Ardoise, farther to the south, we have two samples, both from 1914,
one, however taken in July, and the other in August. In these the scales may pretty
safely be said to exhibit new, and in some eases, fairly advanced summer growth. It
is therefore possible to group the fish in year-classes, as has been done in table 10.

Table 10 — Age distribution in two samples from West Ardoise, Cape Breton. July
and August 1914. Mature and ripening herring.

Locality and date

Number

of

individuals.

104
125

Year-groups and Age-groups.

1912
3

i-G

1911
4

GOG
42-4

1910

5

7"7
19-2

1909
6

1-9
2 4

1908

7

6-7

21G

1907
8

3-8
2-4

1906
9

2-9
40

1905
10

6-7
1-6

1904
11

5-8
3 2

1903

12

10
1-6

More.

West Ardoise, July 1914

Aug. 10, 1914. . . .

2-9

It will be noticed that the samples include a' considerable number of year-classes,
the younger fish, however, predominating. The 1911 year-class in particular is dis-
tinguished by its high numerical value in both samples. In the one from August, the
1908 year-class is also fairly strongly represented, and the same is the case, albeit to
a lesser degree, in the July sample. In the former also the 1910 year-class is like-
wise good. The older fish are too poorly represented to permit of any decision in their
case.

The curves for these two 1914 samples, and in particular that for the two together,
exhibit strong resemblance to the curve for a divergent sample from the area first

^fear g coups

1912 11 10 00 08 07 06 05 04 05-1*93

%

1 1 • > 1 ) > •

.50 .

.0 \

... r

y West Afdoise 1914

■ 20 K

\ 229 Individ

■V/\

\ /\ ^

% /

\

• 30 /

\ St 42. W of Pt Hood 1915

. 20 /

\ 153 individ

• 10/

^'-^'X^

1912 11 10 09

"Seer group

08 07 06 05 04 05-l«9S

Fig. 37.

described, viz., that from station 42 (west of Port Hood and north of cape George)
taken July, 1915. The likeness will be seen from fig. 37. As in the case of the diver-
gent sample from Newfoundland, it mvist be left for future investigations to study
this question more closely.

The next sample in the series is one from the Atlantic coast of Nova Scotia, taken
in August, 191.5. This sample has been dealt with by Dr. Hjort in his preliminary

CAXADIAX FISHERIES EAPEDITIOX, IDl'rlo

129

report. On comparing Dr. Iljort's and my results in the case of this Pample, it was
found that we agree as regards the young fish, but not in respect of the older one.s,
Dr. Iljort having on the whole counted fewer rings on the scales of old fish than I.

T.xBLK 11. — Sample of large and old herring from Nova Scotia, August, 1914, present-
ing difficulties as regards age determination. The table shows the results
arrived at by Dr. Hjort after his preliminary examination, compared with the
results of Lea's two analvses

Age-groups.

5

6

7

8

111
2 9
4-4

9

311
15 8
10 1

10

17-8
16-8
ltj-9

11

11-9
26 6
30-5

More.

Hjort's analysis, 13.5 individ

Lea's analysis in. I, 13!)

II, 138

30
1-4
15

5 9
2 2
1-5

16 3
16 6
15.9

3
19-4
20 3

These different results may be compared, in table 11, which further shows that my
first estimates, when compared with an analysis subsequently made, likewise reveal
some discrepancy in the case of the older fish. The scale preparation for this sample
not being perfect, and the difficulty in the case of older fish being, as already men-
tioned, considerable, all that can be said as to this sample is that there were a large
number of old fish and that age-group 7 (year-class 1908) is well represented as com-
pared with the neighbouring groups.

Table 12.— Age distribution in a sample of immature herring from Halifax, X.S.,
and a sample of mature and ripe herring from Lockeport, N.S. Autumn, 1914.

No. of
indivi-
duals.

82
269

Year-class(

33 and

ige-classes.

Locality and date.

1912.

1911.

1910.

1909.

1908.

7

1907.
8

1906
9

1905
10

1904
11

1903
12

3

4

5

61
27-9

6

More.

Halifax, X.S., Oct. 14, 1914..

9-8
4

84 2
48-3

3 3

'6-3

Lockeport, N.S., Nov. 1914. .

1-5

11

2-2

4 5

1-9

2-6

Besides this sample we have also one from Lockeport, Xova Scotia, taken in Nov-
ember, 1914, consisting of mature fish, and another from Halifax, October 1914, with
immature fish. Table 12 shows the age distribution in these samples. The sample of
mature herring shows an unmistakable likeness to those from Ardoise, its best year-
classes being 1911, 1910, and 1908. In the sample of immature fish, the 1911 year-
class is very numerously represented.

The two samples of small herring from the Bay of Fundy and Gloucester, Mass.,
present no features of interest in this connection; they contain quite young herrings
with one to three summer belts on their scales.

Taking a general survey of all samples from the Atlantic coast, we find, it is
true, a somewhat complicated picture with much varied detail, nevertheless, one
cannot fail to see that these samples reveal certain definite featurei= upon which to
base a further grouping of tlie material. Rave for the samples from Grand Narrows,

130 DEPARTMENT OF THE NAVAL SERVICE

which owing to the isohited locality of capture, situated so to speak, in the middle
of Cape Breton Island, must be considered apart, the samples may naturally be
divided, according to the age-analyses, into two groups, one embracing those from Main-
a-Dieu and North Sydney, the other those from West Ardoise and Nova Scotia.
The best sample in the northern group bears, as we have seen, a strong resemblance
to the single divergent sample from Newfoundland, while the southern samples again
exhibit features in common with a similarly exceptional sample from station 42, west
of Cape Breton island. In the northern group we find, presuming that summer growth
had not commenced in the beginning of June, the year-classes 1910, 1908, 1904 and
1903 as most numerously represented, in the southern, the 1911 year-class especially,
with, to a lesser degree, those of 1910 and 1908, predominating among the younger fish,
while in the case of the older ones, the uncertainty of the age-determination renders
it impossible to say definitely which year-classes are here the richest.

e. Comparison of the different areas.

In the foregoing, an attempt has been made to describe the age distribution in
the samples from the different areas, on the basis of a preliminary group arrangement.
And it was found that samples from dift'erent waters may dift'er altogether in point
of age composition, while those from one and the same locality exhibit a high degree
of similarity. Particularly striking is the likeness observable between most of the
samples from the Newfoundland coast, as also between those from Northumberland
strait. Ha\'ing now compared the separate samples and noted the points of
resemblance for the different areas, with due reference to such exceptions as occur,
it may finally be of interest to draw up as far as possible, a brief survey of the entire
area embraced, on the basis of the details furnished by analysis of the age distribution.

Fig. 38 shows the age distribution for a number of samples from different waters,
chosen from among the 1914 and 1915 samples. The principles upon which such
selection has been based will be found justified by the foregoing.

The figure shows, in its own way, how thoroughly unlike the samples from different
waters proved. Especially characteristic is the difference between the Newfoundland
samples, on the one hand, and those from the southern parts of the gulf on the other.
The samples from the Atlantic coast, however, also exhibit a marked dissimilarity to
the southern gulf samples and also to those from Newfoundland.

On comparing the two parts of the figure, we find, feature for feature,, a consider-
able likeness between them. Particularly marked is the resemblance between the
Newfoundland samples in the two years and those from Northumberland strait.

Having regard to the nature and quantity of the material, it must be admitted
that the analyses distinctly suggest the existence of several types of age composition in
the waters investigated, and it would also seem that in arranging the samples
according to age composition, we shall simultaneously have arranged them, roughly
speaking, according to locality of capture, i.e., that the samples from one and the
same locality exhibit similarity of age composition. That we can speak of such a
thing as a "type of age composition" is due to the fact that in a sample from a cer-
tain water, certain year-classes are found to be more niimerously represented than
the remainder, all year-classes present are not equally rich, and the curves for these
fluctuations thus acquire a typical appearance.

These differences in the samples from the various waters give rise to the sup-
position that the area investigated may include several races or tribes of herring,
each for the present with its own characteristic age composition. Judging from the
age curves, it might possibly he advisable in the investigations by means of growth
measurements, to start from the hypothesis here indicated by the arrangement of
fig. 38.

1. That the Newfoundland herring form a tribe apart, represented through-
out the present material in all samples, save the single exception.

CAXADIAX FISHERIES EXPEDITIOX, lOl'rlo

131

2. That the exception in question, the sample from St. George's Bay, and
the samples from Main-a-Dieu and North Sydney, represent another.

3. That the samples from Magdalen islands and Northumberland strait,
and off cape Gaspe, represent a third.

4. That the samples from West Ardoise and Nova Scotia, together with the
exceptional sample from the mouth of St. George's bay (station 42, west side of
Cape Breton island), represent a fourth.

1912 11 10 09 08 07 06 05 04 03 02 01
J«ar groups

1912 11 10 09 08 07 06 05 04 05 02 01

Fig. 38.

In the following pages, the observations as to growth of the herring will be sub-
jected to analysis, on the basis of the grouping above given. This should, however, as
in the case of the first rough arrangement, only be regarded as preliminary, and
intended but to serve present needs.

132 DEPARTMEyr OF THE J^AYAL SERVICE

XII.— GROWTH.

COMPARISON OF GROWTH IX SAMPLES OF SIMILAR AGE COMPOSITION.

A number of samples were taken from the material available, for further treat-
ment with regard to growth. In making this selection, several points were taken into
consideration. In the first place, it was desirable to include samples from as many
areas possible. In addition to which, it would naturally be interesting to investigate
in this respect such samples as might be said to occupy a peculiar ]X)sition in the
material, e.g., the one exceptional sample from Newfoundland. It was necessary,
moreover, to select those offering the best scale material, the quality of this being of
more importance in growth measurements than in age determinations. And finally,
it was recognized that the growth material should include, as far as possible, material
illustrative of the manner in which growth proceeds during the different seasons of the
year. With regard to this seasonal growth of the herring, there is but little inform-
ation to be gleaned from the present material, some facts may, however, be brought
to light by examination of the samples containing young fish.

The nature and quantity of the material will to a certain extent determine M'hat
problems may be taken up for consideration in these growth investigations: it is
undoubtedly best suited for a study of the growth in the different waters. This pro-
blem, then, will receive the greatest share of attention in the following examination,
and we shall, bearing in mind the results of the investigations as to age composition,
endeavour to arrive at a solution of the following questions: —

1. Are those samples which have been found to be of uniform character, as
regards age composition, likewise uniform or similar with respect to growth
of the individuals ; and

2. Are those samples or groups of the same, which differ in respect of age-
composition from the remainder,, likewise different from these in point of
individual growth?

The two questions taken as one may be formulated thus : Can the grouping accord-
ing to similarity and dissimilarity of age-composition, as drawn up in the foregoing
chapter, be further supported and rendered more distinct by consideration of the
growth of the fish?

In making comparisions of this nature, it would of course be an advantage to
have an opportunity of examining one or more year-classes common to all samples.
This will be done when we come to compare samples of like character as regards age-
composition ; having regard, however, to the great differences in this respect which
exist between the samples from the different waters, a very extensive mass of material,
with very large samples, would have been necessary in order to ensure that each sam-
ple contained a sufficient number of the particular year-classes. In the material as
it stands, we find, for instance, that a year-class which is well represented in tlie sam-
ples from Northumberland strait (the 1903 year-class) appears but very poorly so in
the Newfoundland samples. By proceeding according to this method, we should, in
comparing two samples from different localities in many, or possibly most cases, ac-
tually be comparing, pairs of values of which one only could be taken as accurate,
having been based on a great number of single observations, while the other would be
subject to a higher degree of error, being based on only a few observations. The error
in a difference between two values depends upon the separate errors in the two values
themselves, and will always be greater than the larger of these, with the method
referred to; therefore many comparisons would infallibly give unreliable results from
a statistical point of view. In the following comparisons, the greatest imiiortance

CAXADIAX FISTIKRIEH KXPEDITIOX, 19r',-l.j 133

is attached to the investigation of those year-classes which are best represented, and
for which we have the most accurate average figures, care being taken, however,
throughout to make sure that these good year-classes do not differ essentially with
regard to growth from the inferior ones. We thus avoid the necessity of making
lengtliy calculation for the inferior year-classes, and obtain at the same time some
idea as to how far the growth of the good year-classes is representative of the sample
(or samples, or water).

In the case of two equally strong year-classes the older will generally be employed,
as, in the case of the younger fish, a greater allowance will have to be made for the
effects of the dissociation previously mentioned. And in comparing young fish, there-
fore, the rule as to taking fish of equal age will have to be more strictly observed.

In making these comparisons, the average values (^4) only, have in some cases
been used. In cases requiring closer analysis, however, the standard errors (e) have

8
been calculated from the standard deviation (8) according to the formula e=\/n, where
n is the number of variates. In testing the difference between two averages, the stan-
dard error (d) of this difference calculated according to the formula d:=^e'^ + e\,
where e^ and e.^ are the respective standard errors of the two' averages. The standard
error of the difference compared with the difference itself (D) will then serve to indi-
cate what value should be attached to the difference; if the difference be great in pro-
portion to its error then it will in all probability be significant, i.e., not due to fluc-
tuations of sampling, and vice versa.

This calculation of the standard errors in the averages lias been carried out for
the growth dimensions t^ — t^; for t^ — ^k, the method employed was, save in such cases
where averages only were compared, as follows: When the analysis of and examination
of the averages for the remaining dimensions indicates similarity between a number
of samples, then one of these is selected, and the standard deviation for the dimen-
sions t^ — tj„ calculated for that sample. Presuming then that the standard deviations
for the other samples will be more or less equal to these, the standard errors for all
samples may be calculated from the standard deviations for the one. I have tested
this method, and convinced myself that the presumption is correct, and the consequent
simplification of the calculations thus justifiable.

In all mathematical comparisons, the calculated increment oi gro\\i:.h (0
has been employed, and not the calculated length (I), the former is, as far as I can see,
easier to deal with than the length ; for reasons of economy also it was found neces-
sary to restrict the work to the consideration of one of these dirnensions. In the fol-
lowing comparisons of the results, on the other hand, the calculated lengths have been
used, as being more immediately legible (the length of a herring can be seen, whereas
one can only form an idea as to its growth or increment).

1. Samples from Neivfoundland. — The age investigations led to the results that
all samples of grown fish from the coast of Newfoundland, with a single exception,
were characterized by the marked superiority of the lOO-t year-class. The single
exception (sample from St. Georges bay, jMay 27, 1915, exhibit remarkable likeness
to a sample from Korth Sydney, and will therefore be taken together with this.

Table 13 shows the averages for all increments of the year-class 1004, and the
standard errors of the averages for dimensions t, — t. all each single one of the nine
samples selected for growth measurements ; the lowest series in the table further shows
the total averages with corresponding standard errors, arrived at by taking all specimens
of the said yesr-class (in the nine samples) together as a single larger sample.

134

DEPARTMEXT OF THE yAVAL SERVICE

2: 'iJ

© o c o © o

©©©©©©©©©

£

<

COiOt^MOWlM — 00
i3 CO — . OC O 3i C5 c; -JO

CO

CO

©000i-(©©0®

©

^

<

•— r-li-l.-ll-(.-l,^r-l,H

9

-

<

© O >— 1 O © t^ CO IM O

ooiOM-riococoinio

r-lr-l,— IrHi-li-li— I.-I1-H

-

<

CC-H;Ce<5TC:C0(M©

;:^

(M(Mi-(I^Nr-l,-l(MN

IM

0)

OO©©©©©)©©

©

^

<

O iCl^ ^ -- t^ — C<I crj

X t^ icm C-. L- 1^ x CO

c^i N iM w M IN c^i e<i e<i

IM

0)

©©©©©b©©©

CO

b

<

W l^ — — ' -r ~ — 10 IM

r-^cOl^lco•rcoeoc«5'r
eoeceoeococoeocow

CO
CO

^

Qj

05lO©©'«r<Nt^©r-l

0.-lrH<M^(M©^»-i

b

©©©©©©©©o

<

o: CO r^ 00 r- C-. r^ 3i -^
^C-^^OtCCOP-llOTf

CO

o

^'

OJ

(MCOOCOOO'X©CO-*
— -H-HIM— Oi-I^l-I

g

©©©©©©©©©

©

<

CCiC^IMt^--r ©©CO

00

C5

50«t^t^t^t^:Ol--5C

CO

ID

C<ii^cooa,-*co— 't^
^iMm-rriOlM^lMt-l

©

©coo®©©©©

©

<

iMcoiMoo^cOT. -ri:

?0©0500'»<©t-~©C:

•-o

■o

>

c

CO — -r Iff t-. CO 'S © oc

— . 'J'OCOMCOCOOir

1-1 1-^

-5

" to ^ * '"

: 0) OJ

0±i— I

G. O

^ m ,x 03 CQ ^ r 02

^ IM CO -l- iff tt t- oc 55

CAXADIAX FI.'^HERIES EXPEDITIOX, 191.'rlo

135

By a study of this table, and comparison of the averages with due regard to their
errors, it will be seen that all the samples are very much alike. Among features
especially characteristic may be noted, that f, is on an average less than t.. and that
average values of less than 1 cm. are not reached until t„; further, that #, is more
than half as great as t^.

Some variation is, however, noticeable between the separate samples. It is found
that several of the largest samples exhibit tlie greatest difference, and it will there-
fore be desirable to inquire how far any degree of regularity can be discerned in the
variation, and whether it is so considerable as to suggest the probability of its being
due to other causes than fluctuations of sampling. In order to ascertain this, the
samples were first divided into three season -groups, one comprising the samples from
the spring of 1914, another those from the autumn of the same year, and a third those
from the spring of 1915. Table 14 shows the seasonal averages with corresponding
standard errors .

Table 14. — The samples from Newfoundland arranged according to season of capture.
Table showing averages and errors for growth dimensions t^-t^^ for year class
1904, all samples from each season being taken together.

Season.

1914, Spring . .

1914, Autumn.

1915, Spring..

Total . . .

No. of
individ.

238
240
128

606

6 20
6 08
600

6- 14

Oil
012
13

07

6- 99
6 98
6-99

6 98

07
008
09

05

5-42
5 39

5-51

5 43

006
006
08

00^

3 21
3 34
:V38

3 30

0-04
004

006

03

2-69
2-74

2-8.

274

04
005
006

03

219
200
2-16

2 11

1 54
137
154

1-47

108
106
114

1-08

0-83
92

0-89

0-88

Uo

64
o 75
74

71

It is not easy to discover any essential difference between the seasons here; ^i
falls a little, #2 shows no change, t?. is likewise unchanged, t^ and h rise slightly,
while i^, L and f^ are lowest in the autumn. All these differences are, however,
insignificant and lie within the limits allowable for fluctuations of sampling. Also
each seasonal group presents the same total view of the growth, as seen in each
separate sample or in all samples taken as one.

A somewhat similar result is arrived at by arranging the samples according to
locality, as in table 15.

Table 15. — Newfoundland samples arranged according to locality of capture. Table
showing averages for growth-dimensions t^ — /,„ for year class 1904, all samples
from each locality being taken together.

Locality.

No. of
individ.

ti

t2

t.

U

h

h

<7

U

h

ho

White Bay

Bonne Hay

Bay of Islands

93

95

179

104

5-62
6 35
6 09
6-92

6-98
715
704
701

5-69
5-58
5-48
5- 17

316
3 .32
3-41
3-21

2-80
272
2-84
2 57

2-48
2 12
213
1-96

1 80
1-47
1-51
1-31

1 13
109
1 11
1 06

0-66
0-88
0-91
0-97

0-55
072
0-76

St. George's Bay

73

In this case, the differences arc much more conspicuous. The growth dimension ^i
is lowest for White bay, and highest for St. George's bay. the reverse being the case
with t.^, while the dimensions t.^ — t^ are lowest for St. George's bay.

135

DEPARTMENT OF THE NAVAL SERVICE

7 GEORGE V, A. 1917

Judging from these averages, there might possibly be some slight difference
between the most northerly locality and that farthest to the south.

In order to test the value of the differences thus found between the various samples,
all possible differences between growth dimensions of the same character were first
ascertained, the nine samples giving thirty-six differences for each dimension, making
ISO in all for the five first dimensions.

For each of these differences (-0), the corresponding standard error (d) was then
calculated, and the fraction ^/d formed. This fraction, which expresses the difference
in units of its error may serve as a i^ind of indicator for the importance of the differ-
ence. Where the value of the fraction is small, there is but little probability that the
difference is due to other causes than accidental fluctuations, and vice versa, and in
particular, where its value exceeds certain limits, the probability becomes an empiric
certainty that the difference is not merely due to fluctuations of sampling.

Table 16 shows the manner in which the various values of the fraction ^/d
arrange themselves, firgt for each separate growth dimension, and finally for all
together. It will be noticed that in 91 of the 180 cases, the fraction is less than 1, in
154 less than 2, and in 171 less than 3. Only in 9 out of 180 cases is the fraction
over 3, i.e., in these cases the difference is more than three times as great as the
error. In the last column will be found figures showing the arrangement which
should have resulted had all samples been drawn from an entirely uniform stock and
subject only to fluctuations of sampling.

Table 16. — Showing distribution of values of — for all possible comparisons between
the nine samples from Newfoundland. (Year class 1904, t^ — t^).

Values of — between :
d

18

11

3

2

(■z-

<3-

ti-

<5-

16

14

5

1

Total.

Theoret.

and 1

1 „ 2

18

14

4

18

11

3

2

2

21

13

2

91

63

17

5

2

2

123
49

2 .. 3

3 ,, 4

8

4 1. 5

36

2
36

36

36

36

A comparison of tlu'.-;e theoretic values with the figures actually found, inclines

us to suppose that the nine cases where the value of — — exceeds three are probably

significant. And on examining these nine differences, it will be seen that they are
due to peculiarities in the sample from White Bay and in that from St. George's bay,
only in one instance is there a difference noted where neither of these samples is
included, i.e., between samples 7 and 8 for the growth-dimensions ^3. The small t^ and
the large t^ in the White Bay sample, with the large t^ the small t. and t.^ in that from
St. George's bay, give rise to the differences. It would thus seem that the dissimilarity
noticed in table 15, as between these two extreme samples, is significant. It is
impossible to say with certainty what importance should be attached to these differ-
ences, as several possible explanations may be given. As far as I can see, the most
reasonable supposition would seem to be that the sample from St. George's bay is not
quite " pure," i.e., that it consists of a majority of individuals, belonging to the same
growth type as that of the remaining samples, mixed up with a minority of another
type. From the nature of the dissimilarity noted, the growth of this second type
should be characterized by a large t^ and smaller i.^ t- t^ and t.,.

CA\ADIA\ risHERIE.^ E\I'EniTIU\, HH'i-to

137

On the whole, however, the samples are as nearly as possible equal in rejrard to
growth of the I'.HH year-class, as they were also found to be nearly equal in respect of
age composition.

The question now arises, whether the picture of growth presented by the 1904
year-class can be taken as representative for the remaining year-classes, and espe-
cially for that of 1903. In order to investigate this point, table IT has been drawn
up. showing the growth of the 1904 year class compared with that of the year classes
1903, 1905,, and 1906. as also with all older fish taken together and all younger ones
together.

Table 17. — Averag's for year class 1904 compared with those for other year-classes
in the samples from Newfoundland. Averages based upon all nine samples.

ear class.

19ll-li)06. .

1905 ..

1904. .

190.3. .
1902-1891...

No. of

individuals.

<,.

7 90

'3-

t.,.

'5-

1 31

<7-

<3-

'»•

<.o-

149

6-36

517

3 13

2 10

126

108

0-83

198

52

7 07

5 43

3 36

2 74

217

1 51

1 16

82

0-69

606

6 14

6 9b

5 43

3 30

2 74

2 11

1 47

108

88

71

45

6 67

7 47

5 17

3 24

2 54

2 00

1 45

114

91

75

84

6-88

7-81

4 72

3 47

2 55

1 69

1 37

109

0-86

0-72

'...

61
6-61

It will be seen from the table that the important features are common to the older
fish. The younger ones, and also, to some extent, the 190t) year-class, exhibit greater
deviation, a phenomenon also noted in the great majority of other herring samples
(ride p. 144.). It is impossible to say whether this dissimilarity in the younger tish, in
the present instance, is due to selective effect of the implements used, to the dissocia-
tion of year-classes, or to intermingling with fish of different growth. Probably each
of these features is to some extent responsible. It is at any rate evident that the 1904
year-class does not differ essentially, in point of growth, from the older tish. Also
that the small f,, at least, is probably characteristic for the herring taken on the
coast of Newfoundland as shown by table 18,. the growth of quite young fish from
western bay of Port au Port, August 16-17, 1915. The 1914 year-class, which is well
represented, and that of 1913, which furnished eleven specimens, show exactly similar
values for t^ and t„ to those of the old fish; the 1910 year-class is represented by only
two specimens, so that the figures for this are practically valueless, and are only
included as a matter of form.

T.^BLE 18. — Growth of small, immature herring caught by drift nets in West Bay of

Port au Port, Aug. 1915.

Year class

Age grouji.

No. of
individuals.

'i-

/,.

<3-

t^

<»•

ts-

11)14 .

2
3
6

101

11

2

« 36

6-80
7 45

7 26

760
6 55

4-24
7 45

4 20'

'2-80'

1913

1910

i 05

2. Samples from the Magdalen islands. — The 3903 year-class is the best common
year-class among the older fish in the Magdalen islands two samples, taken in spring
1914 and 1915. In the one sample, there were seventy-four of this year class, in the
other unfortunately but seventeen.

Table 19 shows the average values for the different increments (t) of this year-
class in the two samples. In the 1914 sample, t^ and t^ are less than in that from 1915,
whereas t. — f^ are greater. The remaining growth dimensions are practically alike in
both samples.

6551—13

138

BEPAETME^'T OF TEE ^AYAL SERVICE

c

«4-l

-TJ

o

^

bti

u

Ch

^r

bu

u

to

o

'TS

<-H

s

.0^. ■*

;h

03

QJ

T"!

bn

C3

^
fl

00

Tf"

!M

CD

^

^

t^

00

l-

o

O

^'

^

t^

CO

o ^

<

t-

t~

l^

^

o

o

o

o

.-^

at

<!

o

Ci

to

b

o

b

50

a:

CO

oo

o

tD

o

o

o

o

oc

«

00

1^

I-

t-

o

o

^

-M

CO

-f

(O

^

o

o

o

r-l

^

"

1—1 l^

rH

*

-<

©

©

O

•^ t^ ' rt

CA^AD]A^' FISH ERIE f< EXI'EDITWX, lOl'rlo

139

The standard errors of the averages for dimensions t^ — t^ have been calculated for
each sample separately and for both taken together. With the aid of these, it has
been possible to estimate the importance to be attached to the differences. It was
found that all the differences were between once and twice as large as the correspond-
ing error, and they can thus hardly be considered as noteworthy. And as the remain-
ing growth dimensions have also been seen to exhibit no difference, the two samples
should be regarded as equal, so far as concerns the growth of the 1903 year-class.
Characteristic features common to both samples are: i, is large, distinctly greater
that t„; average values of less than 1 cm. are reached as early as t.; and finally, t^ and
t„ are less than the following dimensions, t^^ and i^^.

If we now turn to the remaining year-classes in these samples, we find, as will
be seen from tabic 20, entirely analogous conditions, all year-classes exhibit an un-
mistakable likeness in respect of growth. There is the great ^i and the low values for
t; and onwards, and even so slight a detail as the fact that the last year's growth is
somewhat greater than that of the few years immediately preceding, is met with now
and again. As regards the 1904 year-class especially, which it would seem desirable
to compare with the same year-class in the samples from Xewfoundland, we find that
it forms no exception, save perhaps that the characteristic features are possibly even
more marked.

Table 20. — Averages for year class 1903 compared with those for other year classes in
the samples from Magdalen Islands.

Year class.

No. of
individuals.

^i

ti

tz

U

^5

h

^7

^8

. tt

<10

<ii

tl2

1910

21

11 2

8'8

4-7

2-6

1909

16
26
26

11
10
10

9
3

1

8
9
9

4
.5
3

3 9

4 3
3-9

28
2-4
2 6

8
8
7

3

4

1908

19('7

10

190t;

22
31
32

11
11
10

2
3

8

8

8

7

7
6
5

4 3
3-9
4 2

2
2 3

2-7

3
3
4

10
0-8
OS

0-9
9
0-7

1905

0-8
08

6-8

1904

1903

91

10

3

7

3

4 6

2-9

5

0-8

7

0-7

0-7

0-8

0-8

1902

9

11

3

8

C

40

2-4

4

1

0-7

(••7

6

0-6

0-5

0-5

1901

6

11

2

8

9

39

2 5

4

1

0-7

0-6

5

4

04

0-4

1900

13

10

R

1-

q

4 2

2-8

1

9

0-7

5

0-5

0-4

OS

We find then, for these two samples also, that the growth of the fish is alike in
both; and that 1903 year-class may be taken as representative for the remainder, at
any rate for the older fish.

3. Samples from Nortlnimherland Strait. — In the five samples of grown fish from
this area, 1903 is the best common year-class among the older fish, and 1907 among
those of medium age. These two year-classes were therefore subjected to closer
examination. Tables 21 and 22 show the averages and standard errors for these year-
classes, in each of the five samples.

6551—13*

140

DEPARTMENT OF THE NAVAL SERVICE

>

9 '

Sj«2

(D

V— '

03

^

3

rill

Id

t:

n1

^a

f2;

fl

<3l

-v->

CK

•n

rt

be
I

n

=:£ r- t^ oc t—

b o b o o

O 5i "O ■-■ iC

» I o t- 1~

b oo o o

> o o o o

o o o o o

CO C: -r -r —

b- 5D » -.O t^

o b b o b

! <

__

■<,

o => b o o

00

©

^

•t

•>J CO — ?0 t^

© O ^ - —

o

© © o © o

.-- — CO — o
b b b b ©

CO I<l IMCO CO

^•^g2;H:

©

a;

© © © © ©

©

""

iC O iC 2- «
oooc t^© -^

-I-

05

<

•*'»•-*< m lO

•^

© 0-. I- O l^
CO r^ -N rj >-l

©

0,

© © © © ©

©

— CO 1-© —
rf -T Tf X CO

tc w ;o © •©

c^ -^ c CO s:

C-l CO CO CO —

b b b b b

© (31 © © ©

Oj T lO "O lO

' o: © 3i oi

>ir-i T-* ^ T^

\

tH ^'cc aa'co'

i>

'^-^ >,^co

«• >, IS >. >^

Qi

n tH «^ *^ — .
-2 .<5*5

-o

cAi -'/) r r

a

1==^-.«

^

^z r;z;Z

cj

i D i O 01

O

XI +j eS *J ^S

H

>5

o-g.Si.ii-S

i?: 02 CL, cd CO

■o^ a[tluitt^

CANADIAN FISIHERIEfi EXPEDITION, lOUf-lS

141

Table 22. — Showing averages for growth dimensions /, — f^ for the year class 1907 in five
samples from Northumberland Strait, 1914-1915. At the foot, total averages and
their standard errors.

Locali'y and Date.

Northd. Strait, May 1914 34

Baie Verte, May 11, 1914 17

Pictou Harbour, May S, 191.5 17

Richibucto, May 19, 1915 2fi

Bale Verte, May 22, 1915 16

Averagfs for the two samples from 1914 together.

Standard errors •< i< u 1914 n

Average><for the three samples from 1915 .■

Standard errors i. „ •■ 1915 h

No. of
individ-
uals.

Total averages.

10 4;

8-70

3 99

2 49

2 54
2 35
2 44
2 74
2 50
OO'
2 50
0<

2 50

•57
•56

69
•57

56
■57

Of;

60
04

1 43
142
1 .30
140
1 39
1 43

1 37

I 59

1-39

t

1

04

1

01

1

15

1

20

1

IS

1

(»3

1

18

1

11

114
124
118

1 19

119

It will be seen from the tables that there is certainly a considerable degree of
similarity between all the samples ; there are, however, also differences which seem,
despite the small number of specimens, to be by no means negligible. Thus, sample
11 differs in its small /, in the 1903 year-class, and Northd. Strait sample in the
same manner as regards the 1907 year-class. In order to test the value of these
deviations, the method adopted in the case of the Newfoundland samples was also
applied here, all possible differences (D) between samples two and two were taken
(for growth dimensions of like character) and the corresponding standard errors of

the differences (d) formed; and finally, the fraction 7 obtained. It was then found.

as will be seen from table 23 that the value of

in one case out of fifty, is

greater than three. This single difference, of considerable magnitude in proportion
to its error, occurs between samples and for the growth dimension /i. Save for this
single instance, all the differences are relatively small, being here, as in the New-
foundland samples, in the majority of cases less than twice their error.

Table 23. — Showing distribution of values of -^ for all possible comparisons between
the five samples from Northumberland Strait. (Year class 1903, t^ — ^5).

Vahus

of —between
(1

ti

h

t3

U

'5

3
6

1

10

Total.

and 1

1 and 2

2 and 3

More than P>. .

5
4

1

10

7
2

1

4
6

4

1

24

22

1

10

10

10

50

As the remaining growth dimensions likewise show considerable similarity between
the various samples, we must conclude that there is no essential difference, as far as
the good year-classes are coiaicerned. With regard to the remainder, these show,
as will l>e seen from table 24, considerable resemblance to the two more closely
examined. These last may therefore be taken as representative of the samples when
comi>aring them with samples from other waters. This will accordingly be done,
the 1903 year-class being used in cases where it is desirable to have older fish, and
1907 where younger are i)referable for purposes of comparison.

142

DEPARTMENT OF THE NATAL 8ERYICE

Table 24. — Averages for year classes 1903 and 1907 compared with those for other
year classes in the five samples from Northumberland Strait.

Year elates.

1011
1910
1909
1908
19!)7
1900
1905
1904
1903
1902

Number

of

individual.s.

5
43
51
23

110
25
35
52

173
13

/l

h

12-6

8-9

12-8

7'7

12 9

7 3

11-3

8-4

10-5

8 8

111

7-6

12 3

6-9

10-4

6-7

10 2

6-5

11-2

7

^6

^6

t^

h

U

<10

tn

tl2

8

3

1-8

9

1-8

6

6

1-7

3

1-4

5

1-6

4

11

1-2

6

1-5

1

10

10

11

G

13

8

0-9

0-8

0-9

1-6

1

0-9

0-9

0-8

0-8

0-8

10

1

0-8

0-7

0-7

0-7

0-8

0-8

8

1-6

10

0-8

6

0-5

0-5

0-6

0-6

hi

0-6

If we are to characterize the growth of the individuals in these samples, w^e can
only repeat what has been said in the case of those from the Magdalen islands; the
similarity existing between these two waters, in respect of age composition, is sup-
plemented by the resemblance noticeable in the growth of the fish. We shall later on
examine the question as to whether there is any essential difference between these
waters as regards growth.

It may not be out of place here to mention the samples containing very young
fish. Some of these are from the waters about Prince Edward Island and the Gaspe
coast, others from the neighbourhood of Souris and cape George. Table 25 shows
the average value for these samples.

Table 25. — Showing averages for growth dimensions t^ — t^ for young herring in the
waters around Prince Edward Island. Values with an asterisk are incomplete.

Locality and date.

Station 33, 48" 13' N.. 63= 58' \V., July 5-C, 1915 -

Station 3.5, 46" 48' N., 64" 32 W., July 7, 1915

Souris, July 14, 1915 -{

Station 53, 45" 46' N., 62" 23' W., July 30-31, 1915... '

Year

Number

of

li

h

^3

U

group.

individuals.

1913

31

10-6

7-9

1-4*

1912

6

7

9

7-9

5-5

10*

1913

47

10

2

7 1

1-8*

1913

14

11

3

8-2

10*

1912

19

10

2

8 8

4-9

10*

1914

12

11

1

3 9*

1912

7

11

5

80

4-5

1-9*

It should be borne in mind that the last growth dimension in most of these sam-
ples is incomplete, and cannot therefore be compared with the same in the older fish.
The table shows that these young fish also are characterised; by a high value for t^,
exactly the reverse of what was the case with the young fish from Port au Port (taken,
it is important to note, with the same set of gear as the present samples from the
Gaspe coast). As regards the differences apparent between the different samples, and
between the different year-classes, it should be remembered that both dissociation of
year-classes and net-selection must doubtless have had some effect. Making allowance
for these considerations, the resemblance between the samples is marked.

CA^fADIAX FLSHERIES EXPEDITIOy, lOl-'f-lo

143

In view of the great similarity between the samples of grown fish from Magdalen
islands and Northumberland strait, it would be of little use to go into the question
as to how far the present samples of young fish more resemble those from the former
or those from the latter waters ; they may fairly be said to exhibit a considerable like-
ness to both.

4. Samples from North Sydney compared with the highly exceptional sample from
St. Georges hay, Newfoundland. — As repeatedly mentioned already, one of the sam-
ples from the Newfoundland coast differed greatly from all the rest in. point of age-
composition, while exhibiting, on the other hand, a very strong resemblance to a sam
pie taken at the same time from Xorth Sydney. It will now be well to compare these
two samples from the point of view of individual growth.

There is no age group which is particularly well represented in these two sam-
ples, group 7 (=year-class 1908 if, as would seem to be the case, new Summer growth
had not commenced at time of capture) is the best, and comprises 23 specimens in
the North Sydney sample, and 25 in the other. After these, come age groups 11 and
12 (presumably corresponding to year-classes 1904 and 1903). In the following com-
parative tables, where the errors of averages have been called into requisition, age-
gi'oup 7 is taken separately, groups 11 and 12 being combined into a single group, in
order to furnish a reasonable number.

Table 26 shows averages for group 7 in the two samples, in addition to which, the

standard errors of the averages and the values of the fraction -^ are also noted.
Table 26. — Showing averages for growth dimensions t^-t. for the divergent sample
from St. Georges Bay and for sample from North Sydney. (Year class 1908,

presumably). At foot the differences and values of fraction -j-.

Sample.

N. Sydney

St. George's Bay .
Difference (D). . .

D

d

10-40

10-28

012

0-29

9 43
22

0-51

41G
4 28
012

52

2-64
2-79
0-15

0-79

1-79
1-89
010

83

It will be noticed that this age group exhibits the most perfect uniformity as
regards growth, and in no single instance, among those investigated, does th*» diffe-
rence between two averages equal the magnitude of their errors.

Very similar results are arrived at in the case of groups 11 and 12. Table 27
shows these two groups compared in the same manner as with group 7.

Table 27. — Showing averages for grovrth dimensions /, — t^ for the divergent sample
from St. George's Bay, and for sample from North Sydney. (Year class 1903 and

1904 presumably). At foot the differences and values of fraction — r .

Sample.

h

h

tz

ti

h

ts

I
ti

i

10-90

818

4-32

2-41

1-Hl

1 14

77

10-93

8-09

4-36

2-65

1-58

1 20

0-94

03

0-09

004

24

003

06

017

0-07

0-26

15

1-41

0-19

46

2 12

N. Sydney

St. George's Bay.

Difference (D). . .

D

78
0-82
004

0-57

144

DEPARTMENT OF THE NATAL SERVICE

Here again the resemblance between the two samples is striking, while the diffe-
rences are insignificant. Only in two instances does the value of the fraction ~j"

exceed 1, which differences, however, also lie well within the limits for fluctuations of
sampling.

Table 28 shows these two samples compared with regard to the remaining age
groups.

Table 28. — Comparison of the divergent sample from St. George's Bay with the sample
from North Sydney by help of averages for all year-classes presenting a reasonable
number of specimens.

Year group.

1911

1910

1909

1908

1907

190G

1905

1904

1903

Sample.

i St. George's Bay
I North Sydney . . .
( St. George'.s Bay
t North Sydney . . .
f St. George's Bay.
I North Sydney . .
( St. George's tiay
\ North Sydney . .
fSt. George's Bay
\ North Sydney .
t St. George's Bay
( North Sydney . . .
I St. George's Bay
\ North Sydney. . .
j St. (xeorge's Bay
t North Sydney
/St. (Teorge's Bay
\ North Sydney. . .

No. of

dividiials.

ti

t2

tz

^4

h

^6

ti

^8

U

tlO

tn

4

11-7

10-6

4-8

3 2

5

11

6

8 2

5 2

3

13

12

7

8

4

4 6

2

5

8

12

13

8

4-5

2

G

G

9

12

1

8

.5

41

2

9

/

1-4

6

11

9

1

4-2

3

G

1-4

23

10

;.)

9

4

4 3

2

8

9

1-3

1-2

2.5

10

4

9

6

4 2

2

G

8

1-3

1-2

9

11

3

8

9

4 4

2

3

6

14

11

11

9

11

1

8

8

3 4

2

5

7

1-6

11

10

7

10

9

8

.5

4 4

2

1

6

11

1-3

0-9

8

8

11

9

8

.5

4 3

2

3

4

11

11

0-8

8

6

12

1

7-9

3-7

2

4

0-9

10

8

8

0-7

1

n

4

8 8

5 4

2

7

8

1-3

0-8

07

6

0-5

17

10

7

8-(i

4-6

2

5

7

13

0-9

0-8

9

08

0-7

17

11

1

8-3

4 3

2

2

1

1 1

Oh

0-8

9

0-8

0-7

12

11

3

8-2

4(1

•)

8

A

11

0-9

0-8

8

0-7

0-7

12

10

G

8

4-4

2

6

5

11

0-8

0-7

7

7

0-7

tn

0-5
6

This table strongly confirms the impression furnished by the previous ones, of
similarity between the two samples. We may safely say that the resemblance is no
less marked as regards growth of the individuals than it was seen to be in respect of
age composition.

Comparison of samples from West Ardoise, Locl-eport, and west of Port Hood
(station Jt2).

Only young fish of the 1911 year-class are here numerous in all four samples, the
calculations as to errors of averages and comparison, on the basis of the same, have
therefore been restricted to this group.

Table 29 shows the averages for this year-class together with the corresponding
standard errors.

Table 29. — Averages (A) and standard errors (e) for the year-class 1911 in three
samples from the Atlantic coast and one from west of Port Hood.

Locality and Date.

W. Ardoise, .July, 1914

August, 191 4 .
Lockeport, November, 1''14
W. of r . Hood, July, 19 1 5

No. of
individuals.

G3

25
57

8fi

12-44
12 99
12 14
12.68

0-21
0-34
0-24
0-21

8-12
872
7-76
S-44

Oil
0-28
14
12

482
443

4 85

5 12

014
0-20
016
010

2 07*
1 45*
2-20*

3 38

06
09
007
09

CANADIAN FIh;n FRIES EXPEDITIOW J9l.'rl5

145

Only the dimensions t^ — t^ will be of interest in this connection, as t^ is incomplete
in the samples from 1914.

A test calculation of the fractioii -j- gave three ditferences of eighteen, which
were over three times as great as the corresponding standard errors. It was the
sample from Lockeport which differed in respect of ^ from the West Ardoise (August)
sample and from station 42, while station 42 again differed in respect of ^3 from the
West Ardoise (August) sample. In the remaining fifteen cases, the differences were
less than three times their standard error,

5 differences being less than once,
11 " " " twice,

15 " " '' three times,

18 " " '' four times,

the corresponding error.

The similarity between the various samples is thus on the whole good, albeit less
so that in the cases previously dealt with. It should, however, be borne in mind that
the comparisons in this case are made with young fish, where the dissociation of year-
classes will be more likely to make itself felt.

Of other year-classes which might be selected for comparison between one sample
and another, that of 1910 is the best; the number of specimens is here, howe\er, so
small, that it would not be worth while to make calculations of the standard errors.
Table 30 shows the averages for this year-class.

Table 30. — Averages for the year class 1910 in the four samples mentioned in

table 29.

Locality and Date.

\V. Ardoise, July, 1914

August, 1!I14
Lockefwrt, Xovembei-, 1!»14
W. of Pt. Hood, July, 1«J15

Ivo. of

In-
divid.

12
18

28

12 2G

12-30

9 82

11 94

8 Ofi
7-87
8-8S
785

u

h

2-69

I 41*

2 ■ S9

1 14*

2-59

1 44*

2 -95

2 01

Save for the incomplete dimension t^ these samples agree, well enough, as will be
seen, always bearing in mind the small number of individuals dealt with. Only the
sample from Lockeport differs in respect of the small t^. The interesting sample from
station 42 (west of Port Hood) does not appear to differ in any essential degree from
the remaining samples.

Finally, table 31 shows comparison for the 1908 year-class.

Table 31. — Averages for the year class 1908 in the four samples mentioned in table 29.

Localitj- and Date.

No. of
individ.

ti

h

^3

^4

h

^6

l7

VV. Ardoise, July, 1914

August. 1914

Locke))ort. November, 1914

W. of Pt. Hood, July 1915

7

7

10

UG

10 6

11 3

11 6

b 3
77
71
7 G

4-5

4 6

5 3
51

.S 2
3 8
3-8
3 5

20
2 6

18

14
1 6
10
16

9*
0-7*
6*
12

Taking into consideration the extremely small number of observations in each
sample, the resemblance here again is noteworthy. With regard to this year class also,
the sample from station 42 differs in no way from the remainder.

146 DEPANTMEXr OF THE NATAL SERVICE

All things considered, these samples must be said to be fairly uniform with regard
to growth, and even with the points of difference noted, each separate sample yet
presents a type of growth unlike those hitherto observed, and characterized by the
high values of all growth dimensions investigated. The same impression is obtained
if we collect the really old fish from all samples and consider their growth as it will
appear in some tables subsequently.

The type of growth may be characterized as approaching, with regard to t^ the
samples from the southern part of the gulf, while as regards the remaining dimensions,
it more resembles the Newfoundland samples. This will be further discussed in the
following section.

5, Samples differing in point of age composition. — It will be seen from the fore-
going, that samples resembling one another in point of age composition likewise
exhibit resemblance as regards growth. This double likeness is very marked between
all samples from Newfoundland save for the single exception, the sample from St.
George's bay; strikingly so between this and the one from North Sydney; good between
the two from the Magdalen islands and the five from Northumberland strait; and still
good, albeit less conspicuous, between the samples from West Ardoise, Lockeport, and
west of Port Hood.

Hitherto, our comparisons have been made between samples which, from their
age composition, appeared at first sight as belonging to the same group or tribe, or
whatever it may be termed. We may now proceed to make comparisons between the
different groups of samples into which the material was provisionally divided.

The object of this is twofold. In the first place, to carry the analysis of the
growth observations a step further, and endeavour to ascertain whether any of the
groups set up in the preliminary arrangement can be taken together. This possibility
will more especially require investigation in the case of the samples from Magdalen
islands and Northumberland strait, the two groups which bear an evident likeness
one to the other in point of age composition, and growth. With the sample from
North Sydney also (and its companion from St. George's bay) it will be desirable to
look into the question of how far combination should be made, for instance, with the
sample from West Ardoise and Lockeport, i.e., whether the former may be taken as
representing the older fish in the group from which are derived those younger spe-
cimens composing the latter; the latter having but recently attained maturity, and
having not yet attached themselves to the shoals of the elders. And in the second
place, it will be necessary to examine the various groups of samples together, and thus,
by pointing out the differences existing in point of growth, to show up more clearly
the resemblances already found. Further, to describe the types of growth as far as
can be done upon the basis of the resvilts obtained from the previous analysis.

In this further analysis, which will to a certain extent take the form of an analysis
of growth in the different waters, we shall first of all make comparison of the total
averages for growth dimensions as calculated on the basis of several similar samples,
and in addition,, examine the differences between pairs of single samples, e^ch from its
own groups of similar material. This latter method of comparison is more especially
intended to show the manner in which examinations of single samples here leads to
the same results as obtained by taking several together. A demonstration for instance,
of the marked and uniform differences between any of the nine Newfoundland samples
and any other, will, by a further test, afford further justification for our regarding
these nine samples as representing a distinct growth type. And in a similar manner,
the correctness of the remaining group divisions will be tested.

For the sake of convenience, the nine similar samples from Newfoundland will
first be compared with the remainder, then those from the Magdalen islands will be
taken, and in like manner compared with the rest, and so on. The comparisons will
be illustrated by means of curves for growth and increment.

The nine similar samples from Newfoundland will now be compared with the
remaining niatorial.

r.jvi/>/i\ FisiniRii.s i:\iT.niTioy\ loi'rio

147

6. Comparison vith the samples from Magdalen Islands. — The following table 32
shows total averafTt's for the nine samples from Xewfoundlaiid, taken together; for the
two Magdalen Islands samples taken together; the differences between gp>\vth
dimensions of the same character with the corresponding standard error, and the

fraction -j-.

a

Tabf^e 32. — Growth of Newfoundland herring compared with that of herring from
Magdalen Islands by help of total averages for year-class 1904 and 1903 res-
pectively.

.\cea of sanijiles

Newfoundland

Magdalen Tsland.-s . .

Difference ( D)

D _„

Standard error of difference — r
ti

h

h

^3

U

h

ti

ti

h

h

6 14

6-98

f>iB

3-30

2 74

2 11

1 47

108

0-88

10-25

7-26

4-62

2-85

lo2

104

0-78

0-6G

0-66

4 11

0-28

0-81

45

1-22

107

69

42

22

22 8

1-9

7 4

6 4

20 3

26 8

17 3

210

11

071
73
002

10

It will be noticed that all differences except /2 ' and tio must be regarded a3
significant, being greater than seven times their error. Judging from the values of

the fraction ;y the growth dimensions ^i, t-,, tr,, t-, and ^s would seem to differ most

in the two sets of samples, h being greatest in the sample from the Magdalen islands,
the remaining dimensions being the smallest. On comparing each of the tw^^ samples
from Magdalen islands with each of the nine samples from Xewfoinidland, exactly
similar results were obtained, even in the case of the 1915 sample from the former
locality, where the number of specimens was so very small (seventeen). This will be

seen from the following table, showing the value of the fraction -^ for each growth

dimension up to t.^, all values exceeding four, however, being for the sake of conveni-
ence placed in one group.

Table 33. — Showing distribution of the values of -, arising by comparison of either

a

of the samples from Magdalen Islands with either of the samples from Xew-

foundland.

Value of — between.
d

h

■""'l8
18

<2

h

U

h

h

U

U

Total.

and 1

1 .. 2

10
3
5

18

10
3

2 .. 3

1

4

13

18

2

10
6

18

8

3 .. 4

14

4 and more.

18
18

18.
18

18
18

18

18

109

144

It will be seen that save for ^, all the growth dimensions exhibit distinct differ-
ences between the Newfoundland samples on the one hand, and those from the Mag-
dalen islands on the other; most of the differences, moreover, are of a very consider-
able magnitude in pro{X)rtion to their errors, and the impression produced by this

148

DEPARTMENT OF THE NATAL SERVICE

table is thus totally different from that given by table IG (p. 136), showing the inter-
nal differences exhibited by the Newfoundland samples when compared one with
another.

Comparison with the sample from Northumberland strait. Table 34 corresponds
to table 32 and shows the total averages of the Newfoundland samples compared with
total averages for the samples from Northumberland strait.

Table 34. — Growth of Newfoundland herring compared with that of herring from
Northumberland Strait by help of total averages for year-classes 1904 and 1903
respectively.

Area of sami)!es.

ti

6 14

10 18

4 04

311

ti

t3

5-4;s

4 94

0-49

6-1

t,

3-3(1
2-98
32

5 3

2-74
1-58
1-10

23 2

^6

2 11
1 18
1 03

46-9

ti

1-47
81
66

30

t<<

1-08
0-72
0-36

25-7

t

0-88
0-70
18

8-2

<10

Newfoundland

Northumberland strait

Difference (D)

698
6-45
53

5- J

0-71
0-72
01

D

_ D

0-7

Standard error of difference

The table here shows that the Newfoundland samples, when compared with those
from Northumberland strait, differ from these in very much the same way as they
were seen to do from the Magdalen islands samples, the difference for t., however, is
greater, and must be considered as significant. In view of the great differences appa-
rent, I have not considered it necessary to show comparison between the separate sam-
ples; such a table would, roughly speaking, be merely a repetition of table 33.

7, Comparison with exceptional sample from St. George's Bay. — Table 35 shows
the total averages of the nine similar samples from Newfoundland compared with the
corresponding averages for the exceptional sample, the 1904 and 1903 classes being
taken together in the latter.

Table 35. — Growth of Newfoundland herring compared with that of the herrings of
the divergent sample from St. George's Bay by help of total averages for year-
class 1904, in Newfoundland samples, and those for year-classes 1904-1903 in
the divergent sample.

Sample.

1

h

^3

U

ti,

<6

t7

«8

Newfoundland (nine
St. George's Bay. . .
Difference (D)

samples) . . .

6-14

10 -9.^

4-79

6-98
8 09
1 11

5-43
4-36

107

3-30
2-65
0-65

2-74
1-58
116

2-11
1-20
91

1-47
94
53

1-08
82
0-20

D

d

14 5

4-5

5-6

5-9

14-5

10 1

8-7

5 2

The table shows the great differences between the exceptional sample and the
remaining ones from Newfoundland; these differences are indeed so marked as to be
distinctly apparent even when making comparison with the few (seventeen) specimens
of the 1904 year-class contained in the former sample as shown in table 36.

CAXADI.W FISHERIES EXPEDITIOS, 191 ',-15

149

Table 36. — Growth of Newfoundland herring compared with that of the herrings in
the divergent sample from St. George's Bay, by help of averages for the year-
class 1904.

Sample.

Newfoundland (nine samples)

St. George's Bay

Difference (D) .'

D

d

ti

<2

h

U

h

te

<7

6 14

10-68

4 54

G 96
806
1 08

5 13
4 62
0-81

3 3f)

2 50

080

2 74
1-62
1 12

2 11
1 25
0-86

1-47
95
52

9 1

33

3 2

2 7

14

7-2

6 5

1 08
84
24

3 4

There can thus be no doubt that the exceptional sample differs strougly both with
regard to age composition and growth, from the remaining Newfoundland samples.
The difference in growth is again of a similar character to that noted between the
Newfoundland samples and those from the Magdalen islands, the greatest differences
are found in the growth dimensions ^,, t., t^ and t.. Here, likewise, t^ is greater in
the exceptional sample, while the other dimensions are smaller.

Table 37 shows the exceptional sample compared with each of the remaining New-
foundland samjiles separately; the order of magnitude of the fraction -, being given
for the differences between the five first growth dimensions.

D

Table 37. — Distribution of values of fraction ~ arising by comparison between the

a

divergent sample from St. George's Bay and each of the nine other samples

from Newfoundland.

Value of fraction — between :
a

U

h

<,

tA

h

Total.

and 1

1 „ 2

1
I
2
5

1

2 .. 3

1

3 M 4

4 „ 16.. ..

9

2
t

1

8

■'"9"

5
38

It will be seen that the exceptional sample differs strongly from each one of the
remainder.

8. Comparison with the sample from North Sydney. — In view of the great resem-
blance which exists between the sample from North Sydney and the exceptional sam-
ple from St. George's bay, it might appear sui>ertluous to make any further comparison
between the former and the Newfoundland samples. For the sake of completeness,
however, this has been done in table 3S, and we find that the Newfoundland samples
differ from this sample exactly as they were seen to do from the exceptional sample
from St. George's bay.

150

DEPARTMENT OF THE yATAL SERVICE

Table 38. — Growth of Newfoundland herring compared with that of the herring in
the sample from North Sydney by help of total averages for year-class 1904 in
the case of Newfoundland herring, averages for the year-classes 1904 and 1903
taken together in the other case.

Sample.

ti

ti

t3

ti

ti

ti

ti

^8

bl4

10-90

4-76

6 98
8-18
1-20

5-43
4-32
111

3-30
2-41

0-89

2-74
1-61
1-13

2 11
1 14
0-97

1-47
0-77

0-70

108

North Sydney . .

Difference (D)

0-78
0-30

D

D

14

4-8

iJ-2

6-8

8-1

10-8

14

Standard error of difference ~

60

9. Comparison with old fish from West Ardoise, West of Port Hood, and Locke-
port. — In order to procure material for comparison of the Newfoundland herring with
fish of old age from the Atlantic coast, it was necessary to collect older specimens
from the different samples. Taking all those of the 1904 and 1903 year-classes toge-
ther we have a total of thirty-four fish, one of which is from the sample taken at sta-
tion 42, west of Port Hood. Averages and standard errors for these thirty-four speci-
mens have been calculated, and comparison made accordingly, with the results shown
in table 39.

Table 39.- — Growth of Ne^^'foundland herring (year-class 1904) compared with that of
herring from Nova Scotia (year-class 1904 and 1903 together).

Locality.

Newfoundland
Nova Scotia
Difference fD)

d

14

11-24

510

15-5

0-99
7-76
0-77

3-1

5-43
4-97
0-46

2 1

3 30
304
0-26

2-74
205
0-69

211
1 36
0-75

12-5

1-47

in

0-.36

7-2

108
0-87
0-21

5-2

There is here, as will be seen, a very marked difference in growth between the
Newfoundland fish and those from the southern portions of the Canadian Atlantic
coast. The differences here again are most distinctly apparent in the growth dimen-
sions /j and ^5 — t^ and resemble those found in the case of the sample from North
Sydney, t^ and t^ being greatest in the fish from the Atlantic coast, the remaining
dimensions less.

10. Graphical illustratic'ns of the results obtained. — Figs. 40 and 41 show, in gra-
phical form, the results of comparison between the growth of the Newfoundland
herring and the growth of those in the remaining samples. In order to avoid a figure
complicated with too many curves, we have here shown, in fig. 40, the growth of the
Newfoundland herring compared with that of those in the samples from the gulf of
St. Lawrence (Magdalen islands and Northumberland strait) and in fig. 41 with that
of the Atlantic fish (North Sydney and southern samples). The upper portion of
each figure presents in graphical form the data already utilized, i.e., the average
calculated increments, while the lower portion showing ordinary growth curves, gives
the actufil length of the herring at the time of the formation of each winter ring on
the scales.

ti tz ^i -t^ . ^S tt> '^7 ^6 ^9 ^}

fo ''//'

>^

^z t, ^

(. i

s 4 ^

7 ta t

9 -t

Si» r»

cm.

iO

\

■

J

1

ncrement

MewFoundland
\iw\h Svdiie>

5,

^

<

\

1

/

>

Nova Scotia

4

9

>

^^^

-

^^

"•^^c

-^

^^===:

=?=•

^-^=^

»^^

'jri.

_^ ^

,..'"

—

^;-:::

30

-

^^ rf

^,^-

20

-

t

/

/

/

Lengths.

/

/

New found land

1

■

/ ,

/

Nor^h Sydney

lA

J

' /

Nova Scoha

I

1

/

5

f

< i 4

'f ■4" 4 ^ <^ ^ Vtf ^/

Fig. 41.

CAXAfnw Fisin:i{n:>i ExrEDniox, lOi'rJS

153

The curves for increment need no further comment, the differences hetween the
Newfoundland samples and the remainder having been sutftciently indicated in the
foregoing. The length curves show how the Xewfoundland herring, starting modestly
enough, ends, owing to the favourable growth of the later years, with an average length
greater even than that of the fish from the Magdalen islands and Xorthumberland
strait (fig. 40). Fig. 41 shows that the somewhat slower growth of the Newfoundland
fish during the first two years enables the Atlantic herring, and, to a lesser degree, also
those from North Sydney, to maintain a part of the distance gained during the first
years, albeit the differences become more and more reduced with increasing age.

11. Samples from Magdalen islands compared with the remainder. — The samples
from Magdalen islands have already been compared with those from Newfoundland,
and were found to differ greatly from these with regard to growth. It now remains to
compare them with the samples from Northumberland strait. North Sydney (includ-
ing the exceptional sample from St. George's bay) and with the Atlantic herring.

Table 40 shows the samples from ^lagdalen islands compared with those from
Northumberland strait, by means of total averages for the 1903 year-class in both
waters. The table is arranged in exactly the same manner as table 32.

Table 40. — Growth of herring from Magdalen Islands and Northumberland Strait
compared by help of total averages for the year-class 1903.

Locality.

Magdalf n Islands . . .
Xiirthnmberland Strait

Difference (D)

D

d

10-

10-181
OO7I

7 2fi
6 4.^
081

0.351 4-77

4 62
4 94
0-32

2-67

^

h

h

ti

h-

U

2-85

1 52

1 04

0:8

60

66

2 9-

1-58

108

81

72

70

13

006

04

0.3

006

004

1-62

100

1 33

100

3 00

1-33

tia

73
72
01

0-50

It will be seen that the growth dimension t„ is the only one exhibiting a difference
which can really be called significant. Possibly 1 and t^ also differ; otherwise the
differences are unimportant in proportion to their errors. In order to test the one
really serious difference, the two samples from the Magdalen islands were compared
with each of the five from Northumberland strait, as regards the first five growth'

dimensions. Table 41 shows the values of — here found.

a

Table 41. — Distribution of values of fraction -j- arising by comparison between each

of the samples from Magdalen Islands and Northumberland Strait (year-class
1903).

Values of fraction — between :

ti

t;

h

ti

h

Total .

and 1

4

5

1

3
1
4
2

5
2
3

4
5

1

16

1 ., 2

2 M 3.

1
4
3
2

14
13

3 „ 4 ....

5

More than 4

2

Sum

10

10

10

10

50

It will be noticed that the differences for t„ are greater than three times their
error in five out o? ten cases and exceed twice the error in nine, whence it would seem
6.5.51— n

154

DEPARTMENT OF THE ?\ATAL SERVICE

likely that the figures express a real difference between the two sets of samples. The
difference cannot, however, be called great, as presented in these comparisons.

Comparison with the sample from North Sydney, and with the exceptional sample
from St. Georges bay. Table 42 shows the total averages for the samples from Mag-
dalen islands compared with the averages for the above-mentioned samples, which, in
contrast to those from the Northumberland strait, both exhibit a greater t^ than the
Magdalen islands samples. Otherwise, there is a not inconsiderable degree of similarity
in regard to growth.

Table 42. — Growth of herring from Magdalen Islands and North Sydney and St.
Georges Bay divergent sample compared. (Year-classes 1903 for Magdalen
Island, 1903 + 1904 for the other localities compared.)

Locality.

tl ti

h

. U

h

h

U

078
0-94
016

2.3

0-77
001

0-2

h

Magdalen Islands

10 25
10 93

68

19

10-90
65

1-8

7-26
809
0-83

3

818
92

3-3

4-62
4-36
26

1-2

4-32
0-30

1-4

2-85
2-65
20

15

2-41
0-44

3 1

1-.52
1-58
06

0-7

1-61
09

0-6

104
1-20
C16

1-8

114
10

11

66

0-82

0.16

D

( D)
(d)

"(W

01)

3-2

Standard error of difference =

North Sydney

Difference = (Dj

D

0-78
012

2-4

Standard error of difference. =

The comparison with the older fish from the Atlantic coast is interesting. Table
43 shows the samples from ]\Iagdalen islands compared with the thirty-four specimens
of the 1904-03 year-classes representing the samples from West Ardoise and Lockeport.

Table 43. — Growth of herring from Magdalen Islands (year-class 1903) and Nova
Scotia herring (year-class 1903 + 1904).

Locality.

t\

^2

iz

?4

U

h

t-,

1%

Magdalen Islands

Nova Scotii)

Difference [D]

10-25

11-24

0-99

7-26
7-76
0-50

4-62
4-97

o-:^5

2-85
3-04
0-19

1-52
205
0.53

104
1-36
32

0-78
1-11
0-33

0-66
0-87
21

D

, d

2-75

1-79

1 46

1-19

4-08

4-57

5 -.50

5-25

The differences here are more marked than those found in the previous comparisons,
four out of the eight amounting to over four times their respective errors. Bearing in
mind, moreover, the fact that all average values are higher in the case of the Atlantic
fish than the corresponding values for the Magdalen islands samples, it will be realized
that the difference in growth is here bv no means inconsiderable.

CA^ADIAy FISHERlEfi EXPEDITWS, 19V,-15

155

12. Graphical iUusf ration of the results obtained.— CuT\e> for the samples from
Magdalen islands and Xorthumberland strait have already been shown together in
one of the previous figures, it has therefore sufficed to give the curves for ^Nfagdalen
islands, l^orth Sydney and the Atln!itic coast ("fig. 42).

Ys ^it \s ^6 ^7 ■^a

It will be seen from this figure and the foregoing remarks, that the growth of
th3 Magdalen islands fish differs distinctly from that of the Atlantic herriug farther
C551— 14J

156

DEPARTMENT OF THE NATAL SERVICE

south, while exhibiting a considerable resemblance to that of the fish from North-
umberland strait, on the one hand, and ISTorth Sydney on the other. Fig. 42 shows
that the growth of the Magdalen Island herring may be more or less aptly charac-
terized as something midway between that of the Northumberland Strait and that of
the North Sydney fish.

13. Samples from Northumherland strait. North Sydney, and the Atlantic coast
compared. — Of the comparisons which still remain to be made, that of the Atlantic
fish (West Ardoise and Lockeport) with the herring from North Sydney, is the most
interesting. We can judge from the foregoing that the rest of the comparisons will
turn out much as with the samples from the Ifagdaleu islands, only with diiferences
more strongly marked. Table 44 shows, that such is the case.

Tablk 44.-^Growth of herring from Northumberland Strait (year-class 1903) com-
pared first with herring from North Sydney (year-class 1903 + 1904), then with
herring from Nova Scotia (year-class 1903 + 1904).

Locality.

tx

h

t:i

ti

^5

h

h

10- IS

6 4.5

4-94

2 98

1-.58

1 08

0-81

10-90

818

4-32

2 41

1 61

1 14

0-77

72

1 73

0-62

57

03

06

04

206

6 6.5

3-26

4 07

0-20

67

80

11 24

776

4-97

3 04

2 05

1 ."6

1 11

106

r;n

03

006

0-47

0-28

30

.3 1

5

1

4

3 6

4 7

6

Northiimberlanri Strait.

North Sydney

Difference (/))

I)

(i

Nova Scotia

Difference (Z))

D

d

72
0-78
06

1-20

0-87
015

3 7

Unfortunately, the observations available for comparison of the herring from North
Sydney with those from more southerly Atlantic waters are rather few. We have
therefore here taken, in addition to the 1904-03 year-class, also the younger fish of
1908, there being at any rate some of these in the samples from both waters.

Table 4.5 shows the comparison for fish of 1904-03.

Table 45. — Growth of herring from Nova Scotia compared with that of herring from
North Sydney. (Year-classes 1903+1904 in both cases.)

Locality.

h

ti

ti

^4

^5

^6

U

h

11-24

10 90

34

7-76
818
42

4 97
4 32
65

3 04

2-41
63

2 05
1-61
44

1 36
1 14
0-22

1 11

77
0-34

87

North Sydney

Difference {!>)

0-78
09

D

d

74

1-24

2-3

3-2

2-4

2-00

4-9

15

All averages save for t.^ are greater in the case of the Atlantic fish. Only in two
instances however, is the difference greater than three times its error.

Table 46 shows the comparison for the 1908 year-class; here also we find one or
"two marked differences, while the averages for all dimensions, save t-i are again
greatest in the case of the Atlantic fish. The unanimous testimony afforded by these
comparisons between the different age groups is, in my opinion, sufficient evidence
of a real diffe7*ence in growth between the fish from North Sydney and those in the

CANADIAN FISHERIES EXPEDITION, lOl'i-lo

157

more southerly samples. This difference is perhaps more pronounced when we look
at the calculated lengths, instead of the calculated increments, as the differences for
dimensions t, — t^ small, it is true, yet tending in the same direction, are then added
together.

Tablk 46. — Growth of herring from North Sydney and Nova Scotia compared.

(Year-class 1908.)

North Sydney.
Nova Scotia. . .
Difference [U).
D

*litv.

h

<2

h

U

U

10 40

11 16
76

9-65
7-72
1 93

416

4-78
62

2-64
3 57
93

l-7i'
2- 16
.37

1 65

4-28

2 58

3 21

218

1-30
1 37
07

64

14. Summary. — On glancing through the comparisons made in the last two
sections, we are led to the following conclusions: Samples resembling one another
in point of age-composition also exhibit great resemblance as regards growth.

The samples from Magdalen islands, Northumberland strait and North Sydney
(with the exceptional sample from St. Georges bay) exhibit a type of growth which may
to a certain extent be characterized as markedly distinct from the Newfoundland type,
while differing also to a lesser degree, albeit still pronouncedly, from that noted
among the herring from West Ardoise and Nova Scotia, which appear to form in thia
resjject a group apart

The growth of the Newfoundland herring is characterized by a modest commence-
ment in the first summer, proceeding well, however, from the third summer onwards.
The herring from the gulf of St. Lawrence have a good first summer's growth,
but owing to the slower growth in their later years do not, when older, exhibit the
samu average length as fish of the same age from Newfoundland and other waters.
The herring from North Sydney show good growth for the first two years; compara-
tively poor, however, later on ; nevertheless, they exhibit, on reaching a considerable
age, a very respectable average length, inferior only to that of the herring from the
southern Atlantic waters. These last may be said to grow well for the first five or
six years, whereby they outdistance the fish from other localities.

XIII. APPENDIX.

OBSERVATIONS AS TO SEASONAL GROWTH IN
YOUNG HERRING.

Some samples of young herring in the material collected afford some evidence as
to the manner in which the summer growth proceeds in the different Atlantic waters.
This, taken together with what we have learned regarding the growth of the older
fish, furnishes some rough idea, as to the main features of seasonal growth in Cana-
dian waters, though a complete survey is not at present feasible.

The sample of small herring from Port au Port, mentioned on p. 137, seems to show
that most of the summer growth in that area has taken place before the middle of
August. Fig. 43 shows the observations taken together, with curves indicated for
seasonal growth during the second and third sunnners, as far as can be calculated from
the scanty data available.

The samples of young fish from the waters between the Gasne coast and Prince
Edward Island exhibit a growth similar to that of the adult fish from ^VFaiidalen
islands and North\nnberland strait albeit it does not seem possible to class tliem defi-
nitely witli either of these two STOups.

153

DEPARTMENT OF THE NAVAL SERVICE

Fig. 44 shows the observations obtained from these samples, arranged in the
same manner as in fig. 48.

cm

7-

Pc

rt

au

Par

I

vr-

^.

'

V"'.

()

/T

1

1

t

/

2-

I

1

Jan. March May July Sept Nov

Fig. 43.

Gaspe -

Pr E.dw I.

%

,*^. ^r."?-^

Jan. March May July ^ Sept Nov.

'■ Fig. 44.

Finally, a couple of trap samples, from Souris, Prince Edward Island, and a
drift-net haul from the vicinity of cape George (station 53) include some young fijsh,
the seasonal growth of which is indicated in fig. 45.

cm
8

-7

So

uri

5 -

C.G

eor

ge

,<

^--

—

._ -

. (.

/^

^/

•ler

• i¥

<v '

-A-

»o«*'

^ .

•

'"

;

/

4kh.

sum

mer

. O

/
/ 1

. I

;

7

■•■''''

/

Jan. March M.3y Julv SepL Nov:

These later samples give a more definite idea as to the time when summer growth
of the younger herring in these waters begins, as in one of the samples (early June)

C'AXADIAN FIS^HERIES EXPEDITION, 191 'rl5 159

the fish had not yet commenced their growth, while the remaining samples revealed a
distinct new summer helt on the scales.

An interesting feature in connection with these fish is the fact that tlic summer
growth commences so late. Off the coast of Norway, the new summer growth com-
mences in April ; but far up in the Baltic, near the coast of Finland, similar conditions
are observed. Hellevaara (IV) who has investigated the herring of these waters,
observes in this connection; " Not imtil the 27th of June did I observe that the scales
had begun to grow on the young fish 1 or 2 years old; but not on those which had
reached maturity.''

XIV. EESULTS OF AGE AND GROWTH STUDIES, VIEWED IN THE LIGHT
OF OBSERVATIONS AS TO RACIAL CHARACTERS.

In the foregoing I have endoavinired, by analysis of the age determinations and
growth measurements made, to arrive at a satisfactory solution of some of those
problems which group themselves about the two more comprehensive questions, as to
the numerical proportions of the different year-classes, and the growth of the fish in
different waters. It became very soon apparent, during these studies, that the area,
within which the samples analysed were collected, presented contrasts often most
pronounced, and it would therefore be natural to consider both age and growth obser-
vations against the background, so to speak, of the third problem, viz., whether several
more or less distinctly different races of herring could be shown to e.xist within
the area investigated.

We may therefore, in conclusion, glance at this question, while considering the
observations as to racial characters, and discussing the possibility of employing growth
observations as a means of distinguishing races.

The results of the age and growth studies bearing on this problem may be briefly
summed up as follows: The different waters exhibit different conditions with regard
to the age-composition in the samples and the growth of the fish.

In seeking foij the cause of the differences observed between samples from
different waters, and of the resemblance between those from the same locality, we are
naturally led to the supposition that several local tribes or races of herring must
exist, each having its own particular area of distribution. Without such hypothesis,
it would appear difficult to understand how the different waters could differ so
remarkably with regard to age and growth of the fish, as seen, for instance, in the
marked dissimilarity between the 'herring of the ^Magdalen islands and those taken off
the coast of Newfoundland, despite the short distance — but a few days' swim — between
the two localities. By presupposing the existence of such local races, we are at least
able to accept without surprise the fact that such differences do exist, albeit we can,
from our present slight knowledge of the subject, form but vague and uncertain ideas
as to how the differences in question may have arisen.

It would therefore be must useful, if other methods of investigation could be
brought to bear upon the problem as to existence of local races in these waters. One
in particular immediately suggests itself, to wit, that formulated by Heincke for the
very purpose of studying the racial distribution of the herring, by arithmetical descrip-
tion of the morphological features.

As mentioned in the introduction, countings of vertebrae, fin-rays, and keel-
scales were carried out with a number of herring from different waters, at the time
the material was collected. These observations should, however, as Dr. Hjort has
pointed out in his preliminary report, only be regarded as of a tentative character,
the samples in question being few and small. Hence, I can, as a matter of fact, add
but little to what Dr. Hjort lias already stated with regard to these observations;
nevertheless, it may yet be of interest to consider them here in connection with what
has been set forth in the foregoing pages respecting age and growth.

160

DE PART M EXT OF THE XATAL SERVICE

The features noted were as follows: Total number of vertebra; (Vert. T.);
numerical order of first vertebra with closed h^mal arch (Vert. H.) ; number of fin-
rays in dorsal (Dors.); and in anal fin (An.); and nxnnber of keeled scales between
the ventral fin and the anus (K.).

I hiave worked out the standard error of the averages for each of these characters
in the several samples, and found that the resulting error of the differences in all
cases may, without risk of serious inaccuracy, be taken as about 0.17, the actual
figures being sometimes slightly higher, sometimes a little lower than this value. In
making comparison between two averages, then, a value of at least 0-5, or thereabouts,
will be required before the differences can with any great degree of probability be
regarded as statistically significant, i.e., due to other causes than accidental fluctuation.
Where the differences of several averages tend in the same direction, smaller values
may naturally be taken as significant.

In the autumn of 1915, however, I made an observation which seems to suggest
that it may, in certain cases, be necessary to demand a higher order of magnitude
in the differences between two averages before deducing therefrom the existence of
variety in race. And as this particular feature has not, as far as I am aware, been
touched upon before, I have thought it advisable to mention it here.

While investigating the number of vertebrae in a sample of young herring from
the north of Norway, in the autumn of 1915, I found that the fish of about 1-| years
old (numbering forty-nine in all) averaged 57-37 vertebra% while those of about 2|
years (241 specimens) showed an average of 57-72. The difference between the two
age groups in this respect was 0-36 z±z 0-11 giving 0-9995 for the probability that it
was not due to accidental fluctuations of sampling. It would thus seem that variation
may occur in the number of vertebrae of the different year-classes (year-class variation
as opposed to age variation). In samples where a large majority is furnished by a
single year-class, therefore, such year-class variation may possibly affect the results
to a certain degree, while on the other hand, this is hardly likely to be the case in
samples consisting of several year-classes, all more or less equally represented. As
samples of young herring are particularly liable to exhibit such majority of a single
year-class, it will be prudent to regard them as less truly representative of the race to
which they belong.

Table 47 shows the average values for the seven samples examined as to number
of vertebrae, etc. The values are here carried to the first decimal point, a value of
about 0-5 being, as we have seen, required to make the difference statistically signifi-
cant. There are one or two slight discrepancies between this table and the correspond-
ing one in Dr. Hjort's preliminary report. One of these (West Ardoise, number of
vertebrae) is due to the omission there of an extreme variate which has been included
here ; another owing to a slight error in the arrangement of the variates ; the remain-
der to the loss of some of the sheets on which the figures had been noted; none arp.
however, of any grave importance.

Table 47. — Eacial characters. Averages for samples from different localities.

Locality and Date.

Characteristics.

Dors.

An.

K.

Vert.
H.

Vert.
T.

Newfoundland, west coast. Autumn '14

Magdalen Islands, May 1914

Northumberland S-trait, May 21, 1914

W. Ardoise. Aug. 10, 1914

Lockeport, Nov. 1914

BayofFuiiHy, Nov.. 1914

(jloucester, Mass. Nov. 1914

Old, mature, developing sex. org.

Old, mature, ripe

Old, mature, ripe

Younger, mature, ripening

Younger, mature, ripe?

Young, immature

Young, immature .

200
18-3
1.S 3
18 5
20-3
18-7
19-9

13 2

25
2.0-4
2.5-2
2.5 4
25-2
25-0
24 9

56-8

dAAADfAX FiSHERIEf; EXPEDITIO\, U)1',-1o 161

From the figures here shown, it would certainly seem very likely that the method
in question might serve to define the differences between the herring from different
parts of the Canadian waters. Thus the sample from Newfoundland, for instance,
differs distinctly from the Magdalen islands and Northumberland strait samples as
regards the number of rays in the dorsal fin, which character likewise seems to indicate
a difference between the herring of the Magdalen Islands and Northumberland strait
on the one hand, and those of the Atlantic coast on the other. This character also,
therefore, might possibly prove of importance if the error of the averages were
reduced by increasing the number of specimens in the samples. Somewhat the same
applies, indeed, to the number of vertebrae; the present samples are too limited to
reveal any absolutely indubitable difference in this character.

On comparing the figures for the sample from Northumberland strait with those
of the one from Magdalen islands, no essential difference is apparent, and one is
reminded of the great resemblance in point of growth exhibited by the samples from
these waters, with the good 1903 year-class common to both. It would be interesting
to investigate the localities in question more closely, with the aid of more extensive
material, in order to determine, if possible, whether the differences in growth and in
age-composition found, warrant our considering the herring of these two waters as
separate groups.

The sample from West Ardoise differs strongly from that taken at Lockeport, by
the smaller number of rays in the dorsal and anal fins. The differences are so great
that it seems almost certain they cannot be accidental. Here again is a ixtint which
might well be investigated further with a larger amount of material. Possibly the
results might then lead to a definite distinction between several forms of herring
along the Atlantic coast, based on the study of morphological characters and growth
measurements. In our present investigations, the paucity of materiaf rendered it
impossible, as already mentioned to demonstrate with certainty the existence of
any difference in point of growth between the herring of Lockeport and those of West
Ardoise. In further investigations it would also be well to include samples from the
north side of Cape Breton island, in order to determine the character of these fish
more accurately than it has here been possible to do ; the same applies to the waters
between Port Hood, Souris, and cape George, as well as the more southerly parts of
the Atlantic waters of Canada.

One special subject for growth investigations which, owing to the favourable
conditions in these waters, have every chance of success, is the distribution of the her-
ring in the gulf of St. Lawrence, studied by means of quite small samjjles of old fish,
examined as to growth. The very remarkable difference in growth between the her-
ring from the southern portions of the gulf, and those from th3 coast of Newfoundland
renders it, as far as I can see, an easy task to determine, even with small samples,
whether the fish in question can be said to belong to one or the other growth type.
Such identifications of the herring could be carried out in a manner similar to the
method of Heincke for the identification of herring by means of morphological char-
acters. Heincke's method may be briefly described as follows : It is desired to ascer-
tain whether a given specimen, of which certain characters have been determined by
count and measurement, more resembles in such respects the one or the other of two
races in which these same characters are known (average and variation). The first
step is to ascertain in what degree the six^cimen differs, as regards each separate char-
acter, from the average values for same in the one race, the difference in each case
being expressed in units of the respective standard deviations. The differences thus
obtained are then squared and added together. Similarly, the deviations of the speci-
men from the average of the other race is then determined, the differences here being
expressed in units of the respective standard deviations for the race in question.
The specimen then differs in the lessor degree — i.e., resembles more — the race for
which the resulting sum of the squares of the differences is the less. Table 4S shows

162

DEPARTMENT OF THE NAVAL SERVICE

the order of the calculations. Tlie example here used was a tish from one of the New-
foundland samples, belonging to the lOO-t year-class ; the values for the different incre-
ments (t) for the specimen in question are shown in the first column of the table.
The second column gives the corresponding average values for all Newfoundland fish
investigated; the third the deviation of the specimen from these averages; the fourth,
the standard deviations for these averages in question; the fifth, deviations of the
specimen divided by the respective standard deviations (i.e., deviations of the speci-
men expressed in units of the corresponding! standard deviations) ; the sixth, these
deviations squared; and finally, the sum of these squares is shown at the foot. The
second half of the table is drawn up in a manner exactly corresponding to this, the
specimen being here compared with the averages for the 1903 year-class from the Mag-
dalen islands. The resulting sums of the squares are, it will be noticed, highly dissi-
milar; that for the comparison with the Magdalen islands averages being more than
ten times the figures resulting from comparison with the averages for the fish to
which the specimen actually belonged.

It will be interesting to refer to table 48 showing the method of comparison
between growth values for a single herring and the corresponding average figures for
the herring of two different waters, according to the principle of " least squares."

Table 48. — Example showing the growth of a single herring compared with average
growth of the herring from Magdalen Islands and Newfoundland by the
method of least squares.

Individual herring, to be tested. .
Average from Newfoundland san

pies

Deviations from Newfoundland

averages 5

Standard deviations for Newfound

land samples cr

8
~~ Newfoundland

Gf-

Averages for Magdelen Islands
samples. ...

Deviations from Magdalen Islands
averages 5

Standard deviations for Magdalen

Islands samples tr

5

c-r

tx

h

(3

5 1

U
31

h
30

h

ti

h

h

71

7

1-7

11

10

11

6 1

7

5 4

3 3

2 7

2 1

15

11

0-9

10

3

2

3

0-4

0-4

01

0-2

1-69

113

95

67

65

0-57

0-47

0-31

29

0-59

000

0-32

0-30

0-46

70

85

32

0-69

0-35

00

010

09

0-21

0-49

0-72

10

0-48

10-3

7-3

4-6

2-9

15

10

0-8

07

0-7

3 2

0-3

0-5

2

1-5

0-7

3

3

4

1 63

1-38

0-98

59

46

0-28

024

20

22

1-96

22

0-51

0-34

3-26

2 -.50

1-25

1 50

1-82

3-84

005

0-26

012

10 63

6-25

1-56

2-25

3 31

tia

8
0-7
1
25
0-40

016

0-7
1
21
0-48

0-23

Resulting s an of fz\ ■ For comparison with Newfoundland averages 27.
Resulting sum of /"_ | • For comparison with Magdalen Islands averages 28'5.

The specimen here dealt with was one of twent.y others selected for examination
by this method. Ten of these were taken from one of the Magdalen islands samples
among the 1903 year-class, the requisite standard deviations for this year-class having
been previously computed; the remaining ten were selected from fish of the 1904 year-
class in one of the Newfoundland samples. In order to ensure the obtaining of an

CA\AIJ1A\ Fh-^HERIhJs EXI'EUITIOX, 191',-lo 163

absolutely independent seleetiou, the task of choosing out specimens was delegated to
another party, who simply noted down twenty of the numbers which had been assigned
in the tables to the fish of the mentioned year-classes in the two samples. The sample
from Newfoundland was, moreover, the one with averages differing most from the
total average for all Newfoundland samples, the deviation tending in the direction
of the averages for the Magdalen islands. It may therefore be admitted that the
results obtained will not present any unjustifiably favourable view of the possibilities
inherent in such comparison. It was found that in seventeen out of twenty instances
the sum of the squares was the less in comparison with the averages for those fish to
which the specimen in question belong. The reverse was found in the case of one
specimen from the Magdalen islands, and two from Newfoundland, these differing
more from the averages of the fish whence they were taken. This result seems to
suggest that it should be possible, by means of the figures obtained from growth
measurements, to determine, in many cases at least, whether an old fish taken in the
gulf of St. Lawrence belongs in point of growth to the type found off the ;^[agdalen
islands, or to that found in the Newfoundland waters. In order to obtain the best
results, good averages would be required, i.e.. it would be necessary to examine large
samples, taken during the spawning season, and on the actual spawning grounds, such
choice of time and place being the best means of procuring " pure " samples, without
admixture of foreign elements. The comparison should be made not as has here been
done with averages for different classes (Magdalen islands 1903 and Newfoundland
190-i) but with averages for the same year-classes in samples from the different
spawning grounds.

As to how far any similar individual distinction may properly be made with regard
to the herring in other Canadian waters, we cannot at present decide anything with
certainty, the material from the Atlantic areas being insufficient. iNfost probably,
however, we should in the majority of cases be able to determine whether a herring
tiken between Cape Breton and Newfoundland should be regarded as a Newfoundland
herring in the strict sense, or not. always pro\'ided that the investigations are so con-
tinued as to furnish a sufiScient fiuantity of material for computing the averages (with
standard deviations).

LITERATURE REFERRED TO.

I Dahl, Knut. The scales of the herring as a means of determining
age. growth and migration. Rep. on Norweg. Fishery and

Marine Investig.. . . .^ Vol. II. No. 6. 1907

II Dahl, Knut. The assessment of age and growth in Fish. Internat.

Rev. d. gesamt. Hydrobiol. u Hydrograph Bd. II. H. 4-5.

III Heincke, Fr. Naturgesehiehte des Ilerings. Abhandl. des Deutsch.

Seefischerei-Vereins Bd. II, 1898

IV Hellevaara. E. Undersiikningar rorande stromingen i sydviistra

Finland. Finlands Fiskeriar Bd. I, 1912.

V Hjort. Johan. Fluctuations in the Great Fisheries of Northern
Europe. Rapp. et Proc.-verb. du Cons. Internat. iwur I'Explor.

de la Mer Vol. XX, 1914.

VT Hjort, Johan. Investigations into the Natural History of the
Herring in the Atlantic Waters of Canada.
Suppt. 5th Ann. Rep. Dep. of Naval Service, Ottawa, 191.o.
VII Hjort. Johan and Lea, Einar. Some results of the Internat.

Herring Investigations, 1907-1911. Pub. de Circonst.. No. 61, 1911.
VIII Hoffliauer, C. Die Altersbestimmung des Karpfens an seiner

Schuppe. Allg. Fischereiztg .No. 19, 1899.

IX Lea. Einar. On the methods used in herrinii- investigations. Publ.

de circonst No. 53, 1910.

164 DEPARTMENT OF TEE NAVAL SERVICE

X Lea, Einar. A study on the Growth of Herrings. Publ. de

Circonst No. 61, 1911.

XI Lea, Einar. Further studies concerning the methods of calculating

the growth of herrings. Publ. de Circonst No. 66, 1913.

XII Lee, Rosa M. An Investigation into the Methods of Growth-Deter-
mination in Fishes. Publ. de Circonst No. 63, 1912.

XIII Reibisch, F. Eizahl bei P. platessa u. die Alterbest. dieser Form

aus den Otolith. Wiss. Meeresunters N.F. Bd. IV, Abt. Kiel, 1899.

XIV Winge P. On the Value of Rings in the Scales of the Cod as a

Means of Age Determination. Meddelels. fra Komm. f. Haver-
undersog., Serie Fiskeri Bd. IV, 1915.

CANADIAN FISHEIMES EXPEDITION, 11)141915

BIOLOGY OF ATLANTIC WATERS OF CANADA

GROWTH OF THE YOUNG HERRING rSOCALLED
8ARDINES) OF THE BAY OF FUNDY

A PEELIMINARY REPORT

BY

A. G. HUNTSMAN. B.A., M.B., of the University of Toronto.
Curator, Dominion Biological Station, St. Andrews, Xeiu Bntnsivicl-.

In the spring of 1915 Dr. Hjort proposed in connection with tlie extended
investigations in 1914-15 that I study the young of the herring (Clupea harengus) or
** sardines " of the Bay of Fundy to determine if possible how hirge they were during
the first winter, and the amount of growth during the year. The numerous Canadian
weirs that are fished throughout the greater part of the year to supply the sardine
factories chiefly in Maine were practically certain to furnish an abundance of material.

Owing to the work that was being prosecuted in tihe gulf of St. Lawrence it was
not possible for me to examine the material in the fresh state except at the beginning
and end of the season. It was necessary to rely upon salted material.

The material has been collected in large part by the engineer of the Biological
Station at St. Andrews, Mr. A. E. Calder. When circumstances permitted, he collected
samples weekly. The material has proved to be far from complete enough to settle the
points in question. This is particularly the case with regard to the smaller fish,
popularly known as " brit," which are for the most part too small to be satisfactorily
taken by the nets used in seining the weirs. Not only will they pass through the nets
in seining, but when present in quantity they will not be taken out, being too small for
canning. Although there are many gaps in the material, the results are not without
interest.

It appeared desirable to use the scale method of determining the age and the
jearly amount of growth; but the material presented such great difiiculties owing to
the indistinctness of the winter rings that this was abandoned and the method of
measurement, instituted by Petersen, alone was used.

The samples were measured on one of the usual boards, divided into centimetres,
with the divisions at the half centimetres so that in each case the measurement was to
the nearest centimetre. This gave centimetre groups for statistical treatment. To
facilitate accurate determination of the length, the measuring board was marked on

166

DEPARTME\' OF THE NATAL SERVICE

either side of the mid-line with a series of parallel diagonal lines, making an angle of
4(r"degrees with the mid-line. By aligning the margins of the tail with these it was
possible to spread the tail to an arbitrary, constant angle of 80 degrees.

Owing to the mixture of herring of different age groups in the samples, it was not
feasible to take the average size in treating the material. The smallness of the numbers
representing certain age groups in many of the samples rendered the results unsuitable
for extensive statistical treatment. The only feasible method was a compromise and
therefore somewhat open to objection.

The relative frequency of the various length groups in a sample indicated whether
the sample consisted of more than one age group and also showed the mean size in each
group. The various groups could in that way be traced through successive samples and
their rates of growth determined.

The following table gives the results of the measurements : —

Length in Centimetres.

Locality.

Date.

f, 14..

7

8

y

10

2

11

3

12
38

13

140

14

89

15
16

16

2

17
2

18
1

19

1

20

21

22

23

24

25

26

27

28

Back Bay

St. Anrlrews

IV, If...

3

5

16

43

79

59

26

3

3

4

.■!

St. Andrews

V, 13..
V. 24..

6

29
4

29
103

12
92

11
32

25
15

49
10

83

45

1

27

3

4

5

2

St. Andrews

St. Andrews

VI, 2..
VI 15

1

32

150

76

46
22
183

32
53
29

18

68

9

10

60

3

5
40

1

5
25

7
12

2
3

1
1

1

St. Andrews

VI, 28..

St. Andrews

VII, 20.
VII, 27..

7

1.55

28

47

68

17

20

10
17

4

18

1
.38

1
36

22

6

^

St. Andrews

2

5

St. Andrews

VIII, 12..

5

9

8

71

77

14

17

29

38

29

13

3

2

St. Andrews

VIII, 1!!..
VIII, 30..

1

9
]

2
13

21

2

60
6

30
11

25
6

26
8

38
24

37

40

26
32

8
15

2
3

2

i

2

1

St. Andrews

St. Andrews

IX, 7..

1

3

10

5

3

6

3

o

12

49

53

30

15

11

1

1

5

i

Jrmesport

IX, 13.
IX, 14..

1

1
4

19
10

62
51

52
98

45
47

42
20

21

2

20

s

6

2

L'Etang

2

2

Lei)reau

IX, 14..

i

15

10

S

1

6

12

14

26

69

58

25

8

3

Pocolopan

IX, 14..
IX. L5

"i

3

8
2

14
1

7

33
2
1

55

15

3

39
51
16

24

(59
53

9

49
37

2
21

10

17

9

4

1

■3

i

Grand Harbour

Boc abec I

IX, 16..

Bocaljec II

IX, 16..

2

23

24

36

60

98

50

13

1

1

Bocabec III

IX, 16..

4

11

^

6

10

9

18

59

76

45

10

Oak Bay

IX, 28..

]

9

10

7

1

1

1

11

55

83

37

7

4

o

2

2

1

Bocabec.

X, 4..

3

10

5

2

1

5

38

47

63

41

17

6

1

St. Andrews I

X, 14..

1

6

23

25

27

23

19

44

64

68

67

47

14

9

5

7

1

1

2

St. Andrews II

X, 14..

12

28

18

17

20

24

45

36

46

35

20

8

7

2

2

2

St. Andrews

X, 25..

1

25

32

19

8

10

23

51

60

19

12

2

St. Andrews

X, 20.
XI, 3..
XI, 10..

20

124

16
9

69
49
32

5
75
49

1
38

28

1

16
19

.38
26

46
43

24
61

6
.37

13

3

i

St. Andrews

St. Andrews

St. Andrews

XI, 23.

12

43

58

26

18

13

31

23

10

3

1

Back Bay

XI, 29..

2
1

i.s

9

67
12

35
32

47
34

.33
21

49
17

30
8

10

7

5

6

6
7

3

8

15

's

4

1

St. Andre w\s

XII, 6..

Bliss Harbour

XII, 21..
XII, 24.

2

13
2

26
4

38
17

91

83

88
91

37

72

24

16

9
6

2
4

7

1
2

Mascarene

Bliss Harbour

XII, 29..

2

5

18

95

111

63

21

10

4

6

4

2

4

Two of the samples may be said to be homogeneous, consisting of only one age
group. They are those of June 28 and October 29. In both the actual range in size is
6 cm. (9-14 and 11-16) and the practical range is only 4 cm. (9-12 and 11-14) or perhaps
3 cm. (9-11 and 11-13). There can be no doubt that in these cases we have to do with
only one age group. The curve for the sample of June 28, obtained by plotting the
lengths against the numbers of individuals is given by the continuous line in fig. 1.
Evidently too few length groups have been taken to give the most satisfactory curve,
but we will not be far astray in taking 10 cm. as representing the mean length of the
herring in the sample of June 28 and 12 cm. for those of October 29.

CANADTAy FisIlERIEfi EXPEDITION, 191 ',-15 167

If we take the sample of October 4 and plot a curve to show the frequencies <if the
various length groups (interrupted line in fig. 1) we see very definitely a bimodal
condition with two age groups represented, for one of which the length of 12 cni. may
be taken as representative and for the other 19 cm. There is, however, a decided
diflFerence in the ranges of the two groups. The smaller one may be considered to have
a range of 4 cm. (11-14) and the larger of 8 cm. (16-23). This might be due to the
phenomenon of dispersion, the older group showing a broad low curve, and the
younger group a narrow high one. I do not believe that this is the full exjjlanation.
The range is too great in the older group. It probably indicates that the older group
is only apparently homogeneous, that it really consists of two age groups so similar in
size as to fuse and give a good unimodal curve. Other considerations to be mentioned
later support this view.

The sample of September 28 shows a similar condition. The practical ranges of
the two groups would be 3 and 5 cm., respectively. The significance of this would seem
to be that by the third year the spring and fall spawned schools have fused into a single
group.

The third sample (III) of September 16, from Bocabec shows imperfectly a tri-
modal curve (fig. 1, dotted line). The sizes representative of the three groups may be
taken as 12, 15 and 19 cm. The ranges are 3 cm. (11-13), 3 cm. (14-16) and 5 cm.,
respectively. The first and third of these groups are evidently identical with the two
groups of the sample of October 4. The second group (15 cm.) was doubtless present
in the latter sample but not in sufficient numbers to appear distinctly.

Let us designate these three groups A (19 em.), B (15 cm.) and C (12 cm.).
B and C give a bimodal curve with a total range of 6 cm. The growth of the smaller
group (C) appears to continue farther into the fall than that of the larger group (B).
This would bring them close together and make them fuse into one group vrith a
range of 5 cm. and a mean size of 14 cm. as seems to be the case in the samples of
November 3 and November 10. (for the latter see the curve in fig. 1 with alternate
dot and dash). In this latter sample the larger group with a mean size of 19 cm. is
evidently A and the smaller group with a mean size of 14 cm. represents (if our inter-
pretation be correct) B and C fused. In the spring of the year group A seems to have
been in the same condition as shown in the sample of April 10, with a range of 5 cm.
and a mean size of 14 cm.

The degree of fusion of B and C and the relative abundance of the two groups in
the various samples give a varying picture as shown in the samples of October,
November, and December from St. Andrews.

In the middle of September samples from widely separated localities along the
coast were examined and also a number of samples from the same locality in order to
determine whether the mean size of an age group varied greatly in the different locali-
ties and in different samples from the same locality. These samples were in great
part obtained through the courtesy of Captain Calder of the Seacoast Canning Co.,
Eastport. The localities were Jonesport (Maine), Grand Harboiir (Grand Manan),
Lepreau, Pocologan, L'Etang, and Bocabec. The samples sht»wed uniformly a great
preponderance of the A group. The mean size varied, being IT cm. (Jonesport,
Lepreau, and Pocologan), 18 cm. (L'Etang), and 19 cm. (Grand Harbour and Boca-
bec). Evidently there is an appreciable difference in the size of the same age gi'oup
from different localities.

Samples were taken from several boats bringing herring from Bocabec on
September 16. These showed uniformly a preponderance of the A group with in each
case a mean size of 19 cm. The same is shown in a sample of September 28 from Oak
Bay. This shows that herring from the inner side of Passamaquoddy bay may be
considered uniform and treated together. Those from points as far away as L'Etang
must be treated separately. The differences shown in the samples of September 16
from Bocabec indicate the amount of uncertainty to be associated with deductions

168 DE PANT M EXT OF THE NAVAL SERVICE

luade from measurements of such small lots of individuals. All three show in their
curves summits at 19 cm. Two show summits at 15 cm., and the third a definite step
in the curve at 15 cm. Only one shows a summit at 12 cm., the other two samples
having no individuals of that or neighbouring sizes. Summits (or steps) are therefore
quite constant for the same age group in the same locality at any one time.

These examiples will serve to show the manner in which the results of the measure-
ments have been interpreted.

To determine the rate and period of growth of each of the different age groups,
the representative length of each group, in each of the samples from Passamaquoddy
bay in which it was represented in sufficient numbers, was determined in the manner
described above. These lengths have been plotted against the dates on which the sam-
ples were obtained, in fig. 2. For each group a curve has been drawn connecting the
circles that indicate the length of the group at different times. Where there are con-
siderable gaps the curves have been continued with interrupted lines. There are occa-
sional points that do not fit into the general scheme. Those of November and
December, intermediate between B and C we have already interpreted as due to more
or less complete fusion of B and C.

The fresh fish measure somewhat larger than the salted. For this reason the
lengths in the samples of May 24 and September 16, seem high compared with the
others. The samples of June 15 and November 3, show groups with mean sizes of
12-5 cm. and 18 cm. respectively. In these cases we may have group C of the preced-
ing year which has failed to fuse with B of the same year, whereas in A these two
groups are constantly fused.

In fig. 2 we have a graphic representation of the growth of the three groups A, B
and C during the year. The period of growth for A and B is from May to the end
of September. For C the beginning of the period is not shown, as at that time the
fish were too small to be taken by the nets. It appears to continue later for this group,
at least well on into October. The rate of growth for B and C is somewhat less than
2 cm. a month in the middle of the summer when the growth is most rapid. The rate
for A is less, about 1-5 cm. a month as the maximum.

In many of the samples larger fish were present, but their numbers are so small
as to be unsatisfactory for. a determination of their Tiiean sizes and growth. The little
evidence there is, shows a group beginning at 1!) cm (April 16 and May l^i!) and grow-
ing to 23 cm. (October 14, October 25, November 10, and December 6). Also a group
reaching 26 cm. by September 7.

The determination of the ages of these groups presents difficulties. Groups B and
C are quite evidently less than one year apart in age. The lack of material less than
7 cm. in length must leave the question in doubt but it is most reasonable to suppose
from the rate of growth shown in 1915 that group B, beginning at a length of 8-5 cm.,
must have already passed through a full season's growth. The well-established fact of
decrease in growth rate with increasing age would necessitate this interpretation,
unless there were still greater growth during the first year. We would then reach the
conclusion that group B was spawned in the spring of 1914, reached a length of 8-5
cm. by winter and in 1915 grew 6-5 cm. to a length of 15 cm. Group C would have
been spawned in the fall of 1914, have reached a doubtful length by winter, perhaps
5-5 cm., and in 1915 grown perhaps 7 cm., reaching a total length of 12-5 em.

Group A is evidently in its third summer and consists of a mixture of both
spring and fall spawned fish. In the third year, therefore, the herring grow from a
length of 14 cm. to a length of 19 cm. The group growing from 19 cm. to 23 cm.
would consist r f hf'Tring in Iheir fourth year and those reaching 26 cm. of perhaps five
year old fish.

CAJSAniAX FISHERIES EXPEDITIOX, 191 ',-15

169

This interpretation maj' be expressed in the folluwinff table:

First Year.

Second Year

Third Yf-ar.

Fourth Year

Fifth Year.

Size,

Increase

Siz

15
12-5

Increase
5

Size
19

Increase
4

Size
23

Increase
3(?)

Si/e

Spring spawned

Fall n

85 cm.
5-5cm.(?)

6-5

7(?)

28-;?)

I have a small quantity of very young herring collected in the tide ripplings in
Passamaquoddy bay in June, 1911. Two small lots were picked up in dip nets at an
interval of one week. Eleven individuals taken on June 19, range from 3-7 to 4-8 cm.
in length, with an average length of 4-4 cm. Twenty-six individuals taken on June
20 range from 4-3 to 5-5 cm., with an average length of 4-9 cm. This gives a growth
of 0-5 cm. for one week. This is higher than the June rate for group B, but nearly
equivalent to the August rate, as shown in fig. 2. A continuance of this rate to
September would give fish averaging about 9 cm. These fish must have been spawned
in the spring of 1911. The fall spaw^iers which spawn at Grand Manan and on the
Nova Scotia shore do not begin until the later part of July. This confirms our inter-
pretation of group B as fish spawned in the spring of 1914.

A comparison of these results with what has been found in Europe with entirely
different methods shows a fairly close agreement. By studying the increase in the
zone on the scale of the herring outside the last winter ring in a series of samples
taken during the years 1910 and 1911 Lea has shown (Publ. de Circonst., No. 61, 1911)
that in the herring off Norway, growth takes place during the summer from April to
September. This growth period is of the same duration but a month earlier than for
our coast.

As concerns the amount of growth, Lea found it to be 7 cm. in the third summer,
which is much higher than what we have found. By calculations based upon the dis-
tances between the winter rings, Hjort (Publ. de Circonst., No. 53, 1910, p. 23) found
that for 246 spring-spawned fish the average growth in successive years was 8.3, 7.1,
5.9, 3.6, 2.4, and 1.7 cm. Our corresponding figures are 8.5, 6.5, 5, 4, and 3 cm. For
80 autumn spawned fish he fovind the following amounts 12.6, 5.1, 3.6, 2.6, 1.6, and
1.1 cm., believing that the first figure really represented two seasons' growi;h. Our
corresponding figures are, 12.5, 5, 4, and 3 cm. The agreement is as close as could be
expected, considering the imperfection in our material. We have also not been able
to separate the spring spawned from the fall spawned after the second summer.

It would have been very valuable to have correlated the positions and number of
the winter rings with this study of the growth from measurements. In the material
examined it has been possible to make out the rings clearly only in a small number of
cases. What has been seen on the whole corroborates the above mentioned interpreta-
tions as to the ages of the various groups.

CONCLUSIONS.

The data, though incomplete, indicate that: (1) there are both spring and fall-
spawned young herring (sardines) in the Bay of Fundy; (2) the spring spawned
schools reach a length of about 9 cm. (3.5 in.) by the first winter and of about 15 cm.
(6 in.) by the second winter; (3) the fall-spawned schools reach a length of about
12-5 cm. (4 in.) by the second winter; (4) the growth during the third season is about
5 cm. (2 in.) ; (5) the growth during the fourth season is about 4 cm. (1-5 in.) ; and
(6) the x>eriod of growth is from May to September.

It is most desirable that this study be continued in order to either confirm or
refute these tentative conchisions and to extend the observations.

6551—15

170

DEPARTMENT OF THE XATAL SERTIOE

CM

in

01

03
03

.-1

CO
o

CVJ
01

1-1 e

1-1 N)

•H

CO

•1
03

r-^

«H

iH

O
•-<

a>

00

1-

1

/.

ll

CD

•I

/

1
1

,.-';•

->

■<*l

...-

^

o

V

- \

c • - . .

» 1

s.

^1

/

,' \

y'

'. 1
■■. 1
\ 1

\

'\

f

2

5

"\^

/

tic

"3;

— —

"

\

2
'3

c

"

■— -

3

"~^~~~"

^^

-^

<D

ik

oino uo oiooiooio oio

OOCDin COWO<3>NCOt*CO.-4

f^

CA.V.1D7.4A' FISHERIES EXPEDITION, 1914-15

171

19
IS

n

16
IS

13

12

11

10

9

8

7

OjG

gy tjO

• IS

A

!

■

— £

^

9

/

9

/

/
/

: CT

_B_

A L

/

99

o

f= — =

1

/

f^

_Q.

1

r

o

/

a<;

3 '-

'

!

/

P

'

y

/

i

' — I

1

o

/

/'

B

-J

/

1

/

/
/

Jan. Feb.

Mar.

Apr.

Kav

JliU.

Jul.

Au(j.

^e,Y

Oct. 1 Nov. iDecl

Fig.

Curves showing growth of Hening in 1915c The numbers
incUcate the length in centimetres.

•

'Noie (December 6, 1916) • — A continuation of the investigations during this year
has given results which agree well with those of last year. This last season differed
from that of 1915 in that the '' sardines " as a whole were small. This was due to
the practical disappearance of the A group by the end of July, the B and C groups
then in turn predominating. The A group was not as homogeneous as in 1915, con-
sisting of varying proportions of its elements (B and C of the preceding year). It
could therefore not be traced with any certainty.

The B gi'oup appeared at the end of May with an average length of about 10 cm.
It was, however, mixed with larger fish until July and could be followed with difficulty.
After August few were obtained. By October it had reached a length of 15 cm.

The C group was first obtained on July 8, with a leng-th of 7 cm. By September
it had become the dominant group and has remained so. During September, October
and Xovember it has continued to grow in length, increasing from 10-5 cm. (Septem-
ber G) to 13 cm. (November 24). The growth was not, however, as rapid as during
July and August.

6551 — 15A

I

CANADIAN FISHERIES EXPEDITION

ATLANTIC VVATEKS OF CANADA

REPORT OS THE COPEPODA OBTAINED IN THE GULF OF
ST. LAWRENCE AND ADJACENT WATERS, 1915.

BY

PROFESSOR ARTHUR WILLEY, D.Sc, F.R.S., F.R.S.C, etc.

Professor of Zoology^ McGill University, Montreal.

PART I.

The pelagic Copepoda are small Crustacea averaging less than five millimetres in
length, whose abundance in the sea is the measure of their importance as a direct
source of food-supply for the young of the commercial fishes. In addition, they are
pre-eminently the food of the herring which in its turn is preyed upon by larger fi.<5he8
such as the cod and halibut. Accordingly, the investigation of their distribution as
governed by depth, season, currents, salinity, and temperature, is generally recognized
as having an economic bearing, for fishes will necessarily assemble in places where
their food is plentiful. Much work has been devoted to this subject in recent years,
especially in European waters, the general purpose of such investigations being to
correlate the movements of these organisms with the seasonal migrations of fishes and
with the seasonal variations of currents. In the IS'aples monograph of the pelagic
Copepods (1892), the author. Dr. W. Giesbrecht, remarks that very little information
was at that time available for the northern part of the Pacific ocean and for that part
of the Atlantic ocean which lies between the 35th and 48th parallels of north latitude.
It is convenient to remember that cape Cod is situated a little above 42° 'N. In Cana-
dian waters there had been no quantitative study of the zooplankton up to the year
1915, but the gulf of Maine and the coastal waters between Nova Scotia and Chesa-
peake bay have quite recently been thoroughly explored during two seasons, the sum-
mers of 1912 and 1913, by the United States fisheries schooner Grampus, under the
direction of Mr. Henry B. Bigelow.

The investigation has to be conducted along two lines : systematic and hydrogra-
phical, which in this case is tantamount to saying qualitative and quantitative. The
object of the first method is to identify the species, in itself a matter of no little diffi-
culty, involving many microscopic examinations. The second method seeks to deter-
mine their relations to the physical and biological conditions of the local environment.
In 1901, Dr. W. ^[. Wheeler published a systematic report on the free-swimming
Copepods of the Woods Hole region. The delimitation of this region was interpreted
in a liberal sense so as to include not only Vineyard sound but also Plymouth harbour,
Mass., and that part of the Gulf Stream which lies 60 to 80 miles due south of Martha's

173

174

DEPARTMENT OF THE NAVAL SERVICE

Vineyard. The last two localities are within a day's journey of Woods Hole, Plymouth
harbour being- a boreal locality, whilst the Gulf Stream carries a tropical and subtro-
pical fauna. As the four localities covered by Wheeler are typical of their respective
districts, it may be desirable to tabulate his records for facility of reference and com-
parison. These records relate primarily to the fauna at or near the surface, a fact
which may partly account for the absence of such representative boreal species as
Calanus hyperboreus, Euchaeta norvegica, and Metridia longa, which frequent the
deeper strata, of water around 40 or 50 metres, although they sometimes ascend into the
surface layers. They are essentially open-water forms, and it is probable that Ply-
mouth harbour and Vineyard sound are too close to the land for them to find the
necessary conditions.

Table A.— Wheeler's Records (July 1899).

Gyninoplesi.

Woods Hole.

Vineyard Sd.
9 XX

Plymouth Hr.

Gulf Stream.

9 xx

9 1

d' 1

9 1

XX

7. Calocalanus pavo

9 XX
9 X

'■ X

X

XX

10. Centropages typicus . .

11. Centropages hamatus

X
X

XX

XX

" X

13. Teniora longicornis

14. Metridia lucen.s ( =M. hibernica)

15. Candacia armata ( =Candace pectinata)..

XX

9 1

"xx

X
X

' XX

XX(d^XX9l)

XX

XX

18. Anonialocera patersuni , . . . .

19. Pontellopsi? regali.-s ( = Monops regalis).. .

20. Acartia tonsa

21. Tortanus discaudatiis ( = Coryiuua buni-

X \

& XX

'"xx

X

Podoplea omitted

X means present.

XX means abundant.

Of the two divisions of the free-living Copepods, namely Gymnoplea or Calanoids
and Podoplea or Harpacticoids, the former is more directly concerned with the purpose
of the present investigation. Of all the species which have passed under review, the
one that is most widely distributed and most abundant in the fishery districts is Cala-
nus finmarcMcus, known to the more observant fishermen as " red feed " or " herring
feed ". It is part of our problem to deal with this and some other species, not only
in the full-grown condition but in such of the stages of early growth as are brought
up in the tow-net. Six stages in the postlarval development of C. finmarchicus have
been specified by Professor Gran (1902) who seems to have been the first to attempt
an analysis and interpretation of the mode of occurrence of this species in the Norwe-
gian North Sea, where its superabundance had been previously signalized by G. O.
Sars. Gran's six stages, adopted with slight inversion by Damas (1905), may be re-
garded as typical for the Calanoids as a whole, although in some cases the procedure
differs remarkably as regards the details of subdivision and fusion of the abdominal
segments Thus, at a certain stage, the young Met'^idia has a four-jointed urosome
(hind-body) in either sex; the adult female has a three-jointed urosome through fu-
sion of the first a?id second segments; the adult male has a five- jointed urosome through

CANADIAN FISHERIES EXPEDITION, 191Jrlo

175

subdivision of the earlier fourth segment. There is a similar sequence of segmenta-
tion in Tortanus. The development consists of three main periods : embryonic period
within the egg; nauplius or larval period comprising six stages; post-larval or copepodid
period comprising six stages, of which the sixth is the adult (see Lebour 1916).

Table B. — Copepodid Stages of Calanus (Damas 1905).

Stages.

Thoracic
segmetits.

Abdominal
segments.

Pairs of legs.

I

2

2

2 and

rudiments of third pair.

TI

3

2

•A

IP fourth j;air.

III

4

2

4

fifth

IV

5

3

V

4

5

VI /^

5
5

4
5

.0
5

At least two other spefies. nf siiinll f^ize. must bo ro"-arded as important ^mirres of
food for young fishes in the gulf of St. Lawrence, viz, Pseudo calanus elongatus which
attains a length of 1-5 mm., and Tortanus discaudatus (ahout 2 mm.). The former
abounds also in the Norwegian sea, the latter is a characteristic Laurentian species.
When these forms and others which are commonly associated with them, e. g. Temora
longicornis (1-5 mm.) and Centro pages haniatus (1-5 mm.), occur in such numbers as
to overshadow the larger forms, they give the plankton a distinctive character both as
to colour and size, the dominant tone after preservation in formalin being a dusky
grey, and the individual dimensions aggregating to give the aspect of a microcalanoid
plankton. A microcalanoid plankton may also be brought about by the predominance
of the young stages of larger species, especially those of C. finmarchieus. When the
individuals are larger and the dominaiut colour, before and after preservation, is red,
owing to the quantity of red oil in their bodies, the plankton wears a megacalanoid
aspect. Such a sample may be composed in varying proportions of Calanus finmar-
chieus ( 3 to 5 mm.) C. hyperhoreus (3 to 7 mm.) and Euchaeta norvegica (3 to 8 mm).
Intermingled with these we find sometimes the white Metvidia longa which according to
Farran (1910) is the "most typically Arctic copepod of whose distribution there is
any accurate knowledge" (quoted by Bigelow, 1915, p. 292). The four last-named
species are as a general rule limited to the waters north of Cape Cod but C. finmar-
chieus ranges to the south of Nantucket, and Euchaeta norvegica was taken from 50-0
fathoms in lat. 40° N., long. 69° 29' W. (Bigelow, 1915).

Giesbrecht tabulated the species of pelagic copepods under three leading regions:
species of the warm region (between 47° N. and 44° S.) ; species of the northern cold
region; species of the southern cold region. He showed that the copepod farmas on
opposite sides of the American continent are more nearly related than those of the
three hydrographical regions named above. Thus the main faunistic differences ap-
pear in following the distribution from the equator to the poles or from the poles to
the equator, not from the eastern to the western hemispheres.

Calanus finmarchieus is not only the commonest copepod in eastern Canadian
waters and in the North Atlantic coastwise waters generally, but it occurs more abun-
dantV" than any other form in the San Diego region where its daily vertical migrations
have been studied by C. O. Esterly (1911). According to O. O. Sars (1901) both
C. finmarchieus and C. hyperhoreus extend tbroiighout the Polar Sea from Greenland
in the west to Behring Strait in the east. He adds that the former species is equally
devoured by herring and mackerel and " in some cases, as stated by Prof. "Robert
Collett, forms almost the exclusive nourishment of one of our greatest whales, Balae-

^76 DEl'ARTMENT OF THE NAVAL SERVICE

noptera borealis. ' On account of the up-and-down movements referred to in the
preceding paragraph, it becomes important to note the time of day when the hauls are
made. The total quantity of Calanus present in the column of water filtered through
the vertical net at a given station is of more practical concern than the quantity at
any particular depth. Esterly found that the maximum abundaaice (" plurimum ")
at the surface occurred during evening twilight. The surface, in a quiet sea, is prac-
tically deserted during the daylight hours, the plurimum between 6 a.m. and 6 p.m.
being located at about 200 fathoms. J. I. Peck (1896) found in Buzzards bay that
from sunrise to sunset the copepods desert the surface almost completely. The fac-
tors which operate in causing this daily rhythm have been analysed in the case of
Lahidorera aestiva by G. H. Parker (1902). Other cases have been discussed by J.
Loeb (1894). At station 48 of the Michael Sai^s, between the Canary islands and the
Azores, on May 31, 1901, Dr. Hjort states that " the tow-net at 40 metres contained
a mass of red copepods, which were not observed at the surface during the daytime,
but suddenly appeared as soon as it grew dark soon after 6 p.m."

In addition to the diurnal there are seasonal migrations. Gran (1902) found off
the Norwegian coast from Romsdal to Lofoten, females swarming in April and May
over the coastal banks. In August and September great quantities of the young stages
(II to IV) are found at the surface. In winter Calanus descends into deep water.
Gran supposes that the autumnal juniors sink into deep water where they slowly com-
plete their growth and rise again to the surface as the spring adults which then spawn,
in Norwegian waters. On the other hand the first haul made by the " Princess " on
May 11, 1915, between Prince Edward Island and the Magdalen Islands contained
both adults and juniors amidst a swarm of Pseu do calanus (see table 1).

According to Giesbrecht's faunistic observations, the distribution of pelagic
copepods does not conform to the oceanic currents although these are factors in their
dispersal. Beyond a certain point the distribution of Calanus -finmarchicus does not
seem to be determined by ordinary physical factors. In the gulf of Maine this species
was taken by the Grampus in water at temperatures ranging from 42° to 76° F., but
was most abundant between 42° and 50° F. (5-5° to 10° C). The density of the water
in which it was living in swarms varied from 1024 to 1-027. It was wholly absent
in pure Gulf Stream water and in the very fresh water at the mouth of Chesapeake
Bay. Bigelow adds that none of the physical constants which were determined in his
exploration of 1913 will account for " the scarcity of Calanus in the waters south of
New York in July, for the subsurface salinities, temperatures, and densities of many
of those stations were well within the range occupied by the species in the gulf of
Maine. What the limiting factor is, is one of the numerous questions raised, but not
answered, hy our cruise." (Bigelow, op. cit. 1915, p. 290-291). That the Gulf Stream
is no barrier to C. -finmarchicus in the proper latitude, is shown by the records of
Acadia station 16, June 1, 1.45 a.m., where the surface copepod haul contained 82 per
cent of this species, the temperature exceeding 12° C, and the salinity 35 per thousand.

The factor which determines the limit of southern dispersion of C. -finmarchicus
is clearly neither a simple physical constant nor a single organic tropism. It can only
be explained at present in terms of endemicity, which includes the biological factors
of food-supply and propagation. The Calani which swarm in and about the gulf of
St, Lawrence have not been brought there by the Labrador current but are endemic in
the Canadian waters. This is shown not only by the presence of the different stages
but by the occasional capture of spawning females, taken in the act of extruding an
egg or before the latter has had time to become detached from the body of the parent.
This is not a frequent observation but was noted in several instances, viz.. Princess
stations 9 and 17; Acadia stations 3, 35, 65, 66, 88; No. S3 stations 13, 14, 25, 26.
Females with spermatophore were seen at Princess station 20; Acadia stations 66, 79,
85, 86, 87, 89 ; No. 3S stations 13, 25, 58, 59, 64.

The endemicity of C. finmarchicus in the gulf of St. Lawrence being thus proved,
it remains to consider its habit of assembling in swarms, in other words its gregarious

CAXADIAX Flf^HERIES EXPEDITIoy, lOl'rlo

177

habit. The records indicate that the Galanus inhaibiting these waters is part of one
vast, continuous community, whose southern frontier is not a straight parallel of lati-
tude but a scalloped border which changes with the seasons; but it is none the less a
definite boundary because the species holds together in virtue of the cohesion of the
individuals. The Calani, with their rich oily bodies, form a floating mass which does
not readily mix with the pure Atlantic water; the line of separation of the calanife-
rous water from the oceanic water is like a line of contact between fluids of different
viscosity which have a slight tendency to mix ; the degree of viscosity being influenced
by the presence of the copepod swarm.

Calanus finmarchicus is both euryhaline and eurythermal, i.e. it is independent of
ordinary diurnal and seasonal fluctuations of temperature and salinity. This fact is
brought out very clearly by the records of No. 33 station 23, which show a gradation
in percentages of this species entirely disconnected with the gradations in temperature
and salinity.

Table C. — Steam Trawler No. 33, station 23, June 25, 1915, between Anticosti and
Gaspe; 49° 31' K, 63° 58' W.; depth 855 metres.

Temperatures :

. +8-09°
. +8-0°
. +3-42°

Centigrade.
75m. .

. . — 0-71''

lOni

100 "

. . — 0-22°

20 " . . .

125 "

. . +0-3°

30 " . . .

. +2-21°

150 "

. . +1-3°

40 " .

. +1-38°
. +0-43"

250 "

. . +3-7°

50 " .

350 "

. . +4-56"'

60 "

. —0-12°

E OF COPEPOD

CONTEXT IX PLAXKTOX.

PERCEXTAG

Species.

Surface 5 mins.
10 CC.

100 .

Vertical 45— Om.
12 CC.

39
44

8

Closing: net

100— f.0m.

15 CC.

33

52

6

Closing net

340— I45ui.

90 CC.

Calanus finmarchicus

Calanus hyperboreus

Pseudocalanus elongatus

20

60

2

5

Eiichffita norvegfica

2

8
1

7

Scolecithrix minor

Metridia longa

<>

100

100

100

100

Salixi
. . 28*59

riES.

75m

. . . 32-51

10m

. . 28-93

100 "

. .. 32-84

20 " .

. . 30-78

150 "

. .. 33-48

30 "

. . 31-31

250 "

. .,. 34-22

50 "

.. 31-97

350 "

. . . 34-62

The tables (I-XII) accompanying this report display the distribution of the prin-
cipal species met with. It is not necessary to continue the tabulation of every station
from Acadia 57 to 90, and I will therefore deal with this portion of the exploration
somewhat more summarily.

Table D. — Percentage of C. finmarchicus of all ages at Acadia stations 57 to 67.

Stations

Percentages

57
3

58
50

59
51

90

fil

80

62
62

63
45

65

47

66
(17

67

28

178

DEPARTMENT OF THE NATAL SERVICE

At stations 68 and 69 the hauls were so scanty as to be negligible, though C. fin-
marchicus from stage III onwards was present at both. At station 74 the subsurface
haul was sparse and contained hardly any copepods; twenty-five were picked out, and
these included C. finmarchicus III (5), IV (3), and ? (1). Here we have an example
of a tongue of northern calaniferous water of salinity 33 per thousand bearing south-
wards over Atlantic water of salinity 35.

In the following table E the vertical hauls were made from 180-0 fathoms except
at station 76, where the haul was from 150-0 fathoms.

Table E. — Copepod content in the vertical net over the Atlantic slope south of Cabot

Strait, July 26-27.

Acadia Stations.

70

72

74

75
2"

76

3
27
10

79

7

19

5

30

V-VI

C. hj'perboreus III- V

2
3

50
X

C. tenuicoriiis . . .

X

Evicalanus elcngatus

2

1

10

2"

3

X

4

■ 2 '
2
5

X-

X

X

24

4

9

Clausocalanus arcuicornis .

1
1

2"

5
X
X

5

Pspudoealanus elongatus ....

Ae.tideus armatus . . ...

X

Gaidius tenuispinus

Undeuchaeta major

11 minor

l'

Euchirella rostrata

6

X

1

11 noivegica

Scolecithrix minor . . . .

13

14

25

21

5

I

1()
6

5

7

II ovata

11 bradvi

X
X

1
1
1

2 '
10

4

8

2

12
X

1
X

100

11 echinata

Centropages bradyi

Temora longicornis

7
X

8
X
33
X
X

10'

X

Metridia longa

75

15
10

12
1

Pleuiomauima abdominalis

11 rf)bufta

5

11 xi|)hia.s .

11 borfalis.

"2"

Heterorhabdus longicornis

1

1

Candacia armata

100

100

100

100

100

Acadia station 80, July 27, depth 168 metres, at the eastern end of St. Pierre
bank, is of interest as lying close to station 24 of June 2 (see table X). On the earlier
date there was a great paucity of copepods but stages II to V of C. finmarchicus were
observed. At the end of July the same early stages were present and, in addition, the
blue copepod, Anomalocera patersoni, had put in its summer appearance at the surface.

CAXADIW ri>iIIFTni:^ EXPEDITIOX, WH-Io

179

Table F. — Percentage of Copepod content in surface and vertical hauls at Acadia

station 80, July 27, 4 p.m.

S|)ecies.

Surface.

Vertical

: 145-0 metres.

C. fininarchiciis II ....

X

Ill

3
10
13

4

2
44
22

2

100

15

IV

V '

2.5
14

9

3

Pseudocalanus elongatu.s.

15

Centropages haiiiatus

25

Temora longieotnis ... .
Anomalocera patersoiii . . .

3

100

The next important station is Acadia 83, 20 miles south of St. Pierre island. This
station, in its depth (172 metres) and proximity to the south coast of Newfoundland
resembles Princess station 41, which was similarly situated with reference to the north
shore of the gulf of St. Lawrence (see table VI). The plankton sample in the vertical
haul (Acadia 83) consisted of about 85 c.c. of material, but tliis was highly gelatinous
and there were only some 1,500 copepods altogether. In the surface haul (15 minutes)
the net was weighted so as to sink it to 5 to 10 fathoms; the amount taken was about
350 c.c, with excessive numbers of Obelia medusae and young schizopods, and a
moderate infiltration of copepods. The plurality of C. finm^rchicus in this tow (65 per
cent) corresponds to its normal midnight plurimvun at the same depth, according to
Esterly's computations. The agreement is not always so close. At station 86, July
28, 11.10 a.m., the subsurface haul yielded 100 per cent of C. finmarchicus, of
which 72 per cent belonged to stege V; at this station the temperature fell rapidly
from 13"S° C. at the surface to 2'6° at 50 metres, and still further to -0*4° at 75 metres.
Perhaps this minimum temperature, in conjunction with the currents, acted tempora-
rily as a false bottom, obstructing the normal daylight descent to the deeper strata.
The effect of currents upon the vertical migrations of Calanus have been little invest-
igated. The area within which Esterly's collections were made was expressly chosen
on account of its freedom from tidal currents and storms.

Table G.— Acadia 83, July 2S, midnight (12.50 a.m.)

Species.

Surface.

C. finmarchicus III . . .

IV ...

V . . .

9 ...
Pseudocalanus elongatus.
Centropage.'* hamatus ...

Temora longicorKis

Anomalocera patersoni . .

2
27
35

1

10

100

Vertical : 160 metres.

16

34

20

3

X

20
6
1

100

Stations 85 to 87 cross the Laurentian Channel and may be compared with Princess
stations 45 to 47 (table VII) and AcaMa stations 25 and 26 and 34 and 35 (table X).

180

DEPARTMENT OF THE NATAL SERVICE

It will be seen that the comparison of the vertical hauls is chiefly demonstrative of
the relative constancy in the character of the Calanoid fauna in the Laurentian
Channel from June to August.

In order to obtain full value from these tables it is necessary to distinguish
between what may be called eurytropic species which are generally distributed through-
out the area covered by the several cruises, and stenotropic species whose distribution
is seemingly limited within a narrower range by physical factors of temperature,
salinity, and depth. Such eurytropic species are Calanus finmarchicus, Pseudocalanus
elongatus, Euchaeta norvegica, and Metridia longa. In consequence of their ready
toleration of slight changes in the temperature and saltness of the water these species
are not reliable indicators as regards the interaction of currents and the stratification
of the water. On the other hand the stenotropic species such as Calanus hyperboreu^,
Euchirella rostrata, Scolecithrix minor, Metridia lucens, and Heterorhahdus norvegicus,
are especially valuable as indicators. Thus the subjoined table shows that a greater
number of these stenotropic species occurs on the eastern side of Cabot strait in the
boreal oceanic water which is in the track of the inflowing Cape Eay current, than on
the western side in the line of the Cape Breton current which conveys water out of the
gulf of St. Lawrence. This agrees with the behaviour, in this region, of certain
species of Sagitta, as I am informed by Dr. A. G. Huntsman, who has made a most
intensive study of the distribution of the Chaetognaths and with whom I have discussed
the bearing of some of the data presented in this report concerning the Copepods.

Table H. — Vertical hauls in Acadia stations 85 to 87, July 28, 1915.

Laurentian Channel.

Across

Species.

C. hyperboreus

Calanus finmarchicus III.
IV .

V .
9 .
cf .

III.
TV

V .
9

Pseudocalanus elongatus .

Aetideus armatus

Gaidius tenuispinus

Kuchirella rostrata

ICuchaeta norvegica

Scolecithrix niinor

M ovata .......

Centropages haniatus

Metridia longa

II lucens

Heterorhahdus norvegicus.

270-0 metres.
85

1

25
15

2

5

Ifi

15

100

270-0 metres.
86

X
10
34
10
10
X

2
X
X

2
X

X

22

2

100

290-0 metres
87

3

10

2S

8

1

X

1

X

X

1

X
X
X
26
1
X
X
20
1

100

Acadia stations 88 and 89, on Misaine bank, both yielded Euthemisto plankton,
including an abundance of C. finmarchicus from stage III to spawning females, with
an excess of stage IV in all the hauls. On this bank at the beginning of June we
encountered an Aglantha plankton, whereas at the end of July there were no medusas
here. As regards the Calanus, the chief change to be noticed was an increase in the
number of adult females, 26 per cent in each of the surface hauls at stations 88 and
89. The latter station may be compared with station 79. Finally, at station 91 in the
Gut of Canso there was an Evadne plankton with a copepod admixture consisting of

OA^'ADIiy FISHIFJRIHS EX I'EDITIoy , 19I',-to

181

51 per cent Torianus, 25 per cent Temora, 15 per cent Pseudocalanus, and per cent
Centropages. The temperature here fell from 12° C. at the surface to 11-45° C. ?.t 45
metres and the salinity was 29-14 per thousand at 20 metres.

Table J.— Data for Acadia station 89, Misaine Bank, 1-32 metres, July 28, 1915.

8.-35 p.m.

Si)ecies.

C. finmarchicus TIT

IV

V . . .

9

cf

C. hyperboreus III

IV

V

9

Pseudocalanus elongatus. .

Euchajta norvegica

Centropage* hainatus

Metridia longa

Anunialocera patertsoni ....
Tortanus discaudatu.s ...

Surface.

2(;
17

2

5
4

X
X

100

30 -Of.

1
44
18
10

3
15

X

O.T-0 f.

3
34

It;
«

X
4

20
1
1

12

X
2
1

100

100

In September, 1915, Dr. A. G. Huntsman made a short cruise in the Bay of
Fundy in the ss. Prince, belonging to the Biological Station at St. Andrews. The data
for the Copepods at the four stations are given in table XII. For the first time in the
course of these investigations we meet with the species Centropages typicus which
occurs regularly in the gulf of Maine and far to the southward of cape Cod, even
reaching the latitude of cape Charles (Bigelow, 1915, p. 293, and fig. 70, p. 294).

At Prince station 1, Dr. Huntsman observed a great stirring up of the water from
the bottom to the surface in consequence of the eddies caused by the tidal currents
surrounding the points of land. The presence en masse of C. finmarchicus at the
surface between 3 and 4 p.m. under a bright sun is unusual, and perhaps the deep-
seated turbulence of the water, with the resulting lack of stratification, was respoTi-
sible for it. It would be worth while to repeat the station, taking samples at intervals
through the twenty-four hours, in order to ascertain whether the diurnal migration of
Calanus is altogether inhibited. The effect of stratification of the water, so far as
temperature is concerned, is seen in an experiment by Dr. G. H. Parker (1902).
Female Lahidocera are negatively geotropic, and remain at the top at all temperatures
between 10° and 26° C. If the temperature is raised to 30° C. they become positively
geotropic and swim to the bottom. The lower half of a large glass tube was filled with
sea-water at 24° C. ; into the upper half sea-water at 30° C. was poured gently. A
female Lahidocera, introduced at the top swam rapidly downward, but stopped at the
plane of separation for the two temperatures.

Besides the diurnal and seasonal migrations of Calanus to which reference has
been made, there is another kind of translation which has been called ontogenetic
migration by Dr. Giesbrecht. Some pelagic copepods, as Clauso calanus, Pseudocalanus,
Eucha^ta, Eurytemora, carry their eggs in an ovisac attached to the genital segment,
but most species discharge their eggs directly into the sea. These eggs, according to
Giesbrecht's observations, have a somewhat higher specific gravity than the water and
consequently sink slowly; whilst they are sinking they accomplish their embryonic
development. As soon as the Nauplius larva hatches out of the egg, it ascends toward.s
the surface and towards the light, all copepod Xauplii being positively heliotropic.

182

DEPARTMENT OF THE yATAL SERVICE

The development of Calanus from the egg upwards was followed by C. Grobben
in 1881, whose memoir I have not had for reference during the preparation of tliis
report. The larval and copepodid stages have recently been described and figured by
Marie C. Lebour (1916). All the larval stages are so small that they pass through the
fine meshes of the silken tow-net, and this commonly happens with the first two cope-
podid stages as well. In order to obtain the earliest stages they must either be reared
under laboratory conditions or the finest procurable silk bolting cloth must be used for
making the tow-nets.

It is not knov^rii exactly when the ontogenetic migrations end and the diurnal
migrations begin. Esterly (1911) confines his enumerations to stages V and VI which
he considered together, rejecting the younger forms. Making allowance for the fact
that stage I rarely comes under observation, we may still unite stages I, II, III, and

IV as a superstage under the term juniores as employed by Gran, in order to compare
it with the superstage of which stage V is the adolescent and stage VI the adult form.
By grouping certain of the data contained in the tables in the manner indicated, a con-
trast appears between the distribution of the juniors and that of the two final stages,
the former being more bound up with the surface layers than the latter. For example,
at No. S3 station 17 there was a copious and typical microcalanoid plankton of stages II
and III, together with Pseudo calanus at the surface (see table VIII). The contrast
which is brought out in the subjoined table K is a partial illustration of the ontoge-
netic migrations of Calanus finmarchicus. If the closing net could have been used
more frequently, and if actual numbers were given instead of percentages, the diffe-
rences in the behaviour of the two superstages would have been rendered much more
manifest. Of course these ratios have no claim to exactness because so many inter-
acting factors disturb the simple relations, but they may serve to bring the problem
into relief.

There is reason to suppose that if stages V and VI were examined separately they
might also exhibit differential behaviour. Indications of inverse behaviour of stage

V and VI (?) are to be found amidst the data recorded in the tables. At Acadia
station 39 the surface ratio of V to VI was as 10:40; in the vertical haul from 25-0
metres as 15:24; in the vertical haul from 100-0 metres as 26:19. Again at No. 3S
station 23, the surface ratio of V to VI was as 0:90; in the vertical haul from 45-0
metres as 2:20; in the closing net from 100-60 metres as 5:23; in the closing net from
340-145 metres as 15 :2.

Table K. — Ratio of juniors (II-IV) to adults (V-VI) of C. finmarchicus in surface
and vertical hauls. Ac. =^ Acadia; Fs.^= Princess; 7. = Prince; ni = metres;
f = fathoms ; jun = juniores ; ad = adolescents and adults.

Station.

Depth.

Time.

Range of
vertical haul.

Ratio of

jvm. to ad.

in vertical

haul.

Ratio of
jim. to ad.
at surface.

Ac. 5

72ni

1 a.m.

60 - Om

44: 38

SO : 20

Ac. 14

2000m

12.50 p.m.

200 -Om

20: 56

nil

Ac. 16

,,

1.45 a.m.

.1

0: 13

12 : 70

Ac. 20

500in

10 p.m.

100 -Om

9 : 53

27 : 47

Ac. 28

,,

6 a.m.

100 - 25m

8: 41

51 : 36

Ac. 49

126m

3.40 p.m.

125 - Om

42 : 46

80: 7

Ac. 50 . .

151m

7.30 p.m.

145-Om

21: 47

40 : 40

Ac. 52

99m

1.45 a.m.

90 - Om

35: 51

CO : 32

Ps. 16

lOOf

1.15 p.m.
6 a.m.

100 -Om

24: 40
23 : 37

66 : 12

P*. 20

81 : 9

Ps. 21

,,

9.45 a.m.

„

26: 43

89 : 2

Ps. 39

284m

7.30 p.m.

1.30 -Om

»J: 15

50 : 5

Ps. 46

400m

6 a.m.

130 - Om

17: 74

47 : 53

P. 1

ISf

3.45 p.m.

18 -Of

6: 68

18: 72

P. 2

60f

4.30 p.m.

55-Of

3: 70

25 : 17

CAyADIAX FISHERIES EXPEDITION, 191. ',-15

183

PART II.

NOTES ON SPECIES.

1. Calanus finmarchicus. — There is a wide range of variation in the size of the
individuals during the several copepodid stages and this is often to be observed at one
and the same station. It is assumed that the transition from one stage- to another is
effected by a single exuviation. In one case I observed the new cuticle of stage V
forming beneath the old cuticle of stage IV, as shown most clearly by the coxal denti-
culation of the fifth foot. The denticulation on the inner margin of the basal joint
of the fifth foot occupies the whole of that margin in both stages V and VI; at stage
IV only the middle third of the coxal joint shows the marginal denticulation (text fig.
I). It is to be noted that the outer denticulation shown in the figure is much more
restricted in certain individuals.

Fie

1. — C. fimiuirchicus stage IV, length 32 mm.; fifth feet showing
transition to succeeding stage. Acadia station 9.

By employing biometrical methods. Gran (190.) found that within the limits of
the Norwegian North Sea, the individuals averaged larger in the north than in the
south. For example at stage III (synonymous with Gran's fourth stage) the maximum
length of the forebody in individuals captured in latitude 60° 43' N. was 1-25 mm.,
while in those taken in latitude 67° 41' N. the maximum length of forebody was 1-4S
mm. Gran also found in general that the stages are smaller in summer than in spring.
It is doubtful whether these biometrical results are applicable to other regions. To
render the history complete, it would be necessary to take stock of the frequency of
exuviations, and up to the present this has not been found practicable.

184

DEPARTMENT OF THE NAVAL SERVICE

The following tables give a few examples of range of size at stages IV, V, and VI.
The measurements were taken from the front of the head to the end of the caudal
fork.

Table L. — Range of C. finmarehicus IV.

Station.

Ac. 9
Ac. 28
Ac. 79

Ac. ><8
Ac. 89

(measurements in niillinietres.)

22; 32
20: 29

1 9 * to 20; small instars from 30-0 fathoms. In the haul from 180-0 fathoms some of

2 35 mm. were taken.
2-75; 3
30; large instars from 30-0 fathoms (compare Ac. 79).

Table M. — Range of C. finmarehicus V.

Ac. 5 ....

2-

Ac. 48

2-

Ac. 79 ...

2'

Ac. 80

4-

Ac. 85

3-

Ac. 86

2-

Ac. 88

2

Ac. 4.
Ac. 5.
Ac. 8
Ac. 79.

Ac. 85.
Ac. 86
Ac. 88
Ac. 89.

•9;

•65:

•25;

•9.

■6;

•65;

■75;

30;

45

2-5;

4 7;
45
4 0;

3-5

30; small instars as with stage IV at this station.
fai and oily.
4 1; 4-5

Table N. — Range of C. finmarehicus VI (?).

0; 4 0: 50

5; 365; 425

35; 3 75

9 with spermatophore ; 3 with spermatophore.

2 with sv>ermatophore ; 4 5

with spermatophore and turgid with pale pink oil.

65; 5

5

2; 5 5 with spermatophore.

Table O. — Range of C. finmarehicus VI (c?).

Ac. 86.

05; 50

CA\AniA\ risllKlilKS EXJ'EDITIOS, Wl.',-!',

185

The prroiitcst contrast in avorafrc size is exhibited between stations 79 and 89.
Table P accordinjirly gives the hydrofrraphieal data for these stations as worked out by
Mr. Paul Bjerkan.

TaijM': p. — Ilydrographical data for Acadia stations 79 and 89.

Metres.

44° 47' N.. 55" 13' W.
Station 7!*.

45° Ifi' N., 59' 4' W.
Station 89.

*

Salinity.

Temperature.

Salinity.

Temperature.

10

25

32 42
32-53
32-85
:;3 06
33-33

33 97

34-40

34 60
34-61
34 72

13 05
11-1
9-5
5-5
2-5
4 9

5-9
6-2
4 65
4 15

30-52

30 90
3192
32 13
32.13
32 24

13 95

50

9 1

75

100

0-75
015

125

15

05

150

200

300

400

2. Calanus hyperhoreus. This Arctic species is nearly as widely distributed in
our area as the preceding, but it is bound up with the deeper layers of water and in
that sense it is stenotropic, rarely appearing in surface hauls in these latitudes, ^[ore-
over it does not range so far south as C. finmarchicus. The Grampus found that, like
Euchaeta norvegica and Metridia longa, it was limited to the waters north of cape
Cod and was taken only at four out of twenty one stations in the gulf of Maine. At
Grampus sta.tion lOlCK) between cape Sable and Penobscot bay, opposite the mouth
of the Bay of Fundy, the vertical net from 90-0 fathoms contained 270 individuals of
C. hyperhoreus to 5400 C. finmarchicus, this being its plurinuim for the gulf of .Nfaine
(Bigelow, 1915, p. 293).

According to Damas and Koefoed (1905) C. hyperhoreus is the commonest form
at the surface in the Greenland sea. In Canadian waters there does not seem to be
any regularity in its occurrence at or near the surface and each case would probably
need to be accounted for by reference to local and temporary conditions. Its presence
to the extent of 5 per cent in the deep surface haul with weighted net at Acadia sta-
tion 85 and not at stations 86 and 87 is perhaps significant in view of what has been
stated regarding this station (see above, table H).

No male was observed in ai\v of the hauls. Sometimes there nuiy be a little
doubt regarding the identification of this species at stage IV with the three-jointed
urosome. The postcro-lateral angles of the forehody are not always so distinctly point-
ed as is usual. In such cases the do\ibt is at once removed by the examination of the'
fifth legs which although possessing coxal dentievdations in the subsequent stages, are,
unlike C. finmarchicns, devoid of them at stage IV in C. hyperhoreus.

3. Calanus vulgaris. — Other species of Calaniis Avere met with at station* in or
near the Gulf Stream. Of these tne most remarkable was C. vulgaris which is not
mentioned in " 'N'ordisches Plankton" (v. Breemen 1908). The female of this species
has the postero-lateral angles of the forebody produced on each side as a ventrally

6551—16

186

DEPARTMEM OF THE NATAL SERYICE

curved hook; in the male these edges are rounded, and the fifth feet have a peculiar
vermiform process on the left side which distinguishes it from all others. It occurred
at Acadia station 44 in the vertical haul, males and females, some of the latter having
a spermatophore attached to the urosome or hind-body.

C. tenuicornis was recorded at Acadia stations 17, 74, and 75; C. gracilis at 42,
44, and 75; C. minor at 44 (IS per cent at the surface), 56 (one male in closing net
from 210-140 fathoms), and 75 (males and females at the surface). C. minor is ano-
ther species not mentioned in " Nordisches Plankton " ; Wheeler observed numerous
females taken in Gulf Stream tow, July 25, 1899, the locality being 60 to 80 miles
south of Martha's Vineyard.

4. Eucalanus and Rhincalanus.- — All the species of these two genera which are
included in " Nordisches Plankton " were encountered in stations tangential to the
Gulf Stream; many of the individuals were immature, especially was this the case
with Rhincalanus. When attempting to differentiate the young stages of RTi. cornu-
tus and nasutus, the conspicuous feature of the forwardly produced head with its ros-
tral filaments is not an unfailing guide. Figs. 2 and 3 show the appearance of the
fifth feet in young females of cornutus and nasutus; figs. 4 and 5 show the fifth feet
in young males of cornutus of two sizes.

Fig. 2. — Rhincalmius ronaitua juv. $. Length 3 nun.
Fifth feet. Acadia station 44. 150-0 fathoms.

Fig. ?. — Rhincalanus natiutu.^ juv. 9 3 mm.
Fifth feet, one shown. Same station.

Fig 4. — Rhincalanvs cornutus juv, cf.
Fifth feet, hinder surface.

Fig. 5. — Rhinra/anus cornutus juv. cf. Fifth
feet, front surface of older stage.

rWADIW /7N// /:/.'//> I.XI'EDITIOW IDlfi-lo 187

5. Clausocalanus arcuiconiis. — Previously recorded by Wheeler from the Gulf
Stream. In the deep surface tow of Acadia station 74 there were very few copepods ;
out of a total of 25 counted, il were of this species, the others being C. finmarchicus
III (5), IV (3), 9 (1); C. minor (2); Scolecifhri.r minor (1), dance (1); Acartia
sp. (1).

6. Pseudocalanus elongatus. — The abundance of this small but rich and oily species
in the gulf of St. Lawrence is paralled by its frequency in the gulf of Maine. In the
intervening stretch of water between the entrance to Cabot strait and to the bay of
Fundy it does not occur in such great numbers. In the May plankton of the gulf of
St. Lawrence between Prince Edward island and the Magdalen islands, it constituted
on the average between eighty and ninety per cent of the copepod content. At several
other stations inside the gulf if reached to 20 per cent and upwards. In none of the
Acadia stations outside the gulf did it attain as much as 20 per cent, the nearest to
this quantity being 17 per cent at station 67; 15 per cent at SO and 90; 14 per cent at
36.- In all the females which have come under my observation in the preserved material
the ovisac was ruptured and the eggs appeared to be attached singly to the genital
segment, sometimes one at a time, frequently two, rarely three. Often the shreds of
the stalks of attachment are left behind after the egg has been liberated or torn away.
When two eggs are present they may be seen to be attached separately side by side.
Sometimes there will be one egg and the shrivelled stalk of another beside it. Sperma-
tophores are sometimes applied to one and the same female in great numbers. In the
example represented in fig. 6, I counted as many as twenty-four spermatophores.

Fig. 6. — Pscudocidnnu^ elomiatux 9, beset with spermatophores, ''Princess"
station 30, August 4th. 30-0 metres.

7. yEtideus armatus and Gaidius tenuispinus. — Neither of these species has been
recorded previously from eastern American waters, that is to say not by Wheeler nor
by Bigelow. They occurred together in Acadia station 75 in the vertical haul (180-0
fathoms or 325-0 metres) and again at station 87 (290-0 metres). Another noteworthy
record for Gaidius was at No. S3 station 23 in the closing net from 340-145 metres.
Besides this it was present in the vertical haul at Acadia station 46 (270-0 metres).

^tideus was not found at any station in the gulf of St. Lawrence. In addition
to the two Acadia stations mentioned above it occured also at stations 17 (200-0 m) ;
25 (120-0 m.) ; 44 (270-0 m.) ; 74 (325-0 m.) ; 79 (325-0 m.) ; 85 (270-0 m.) ; 8G
(290-0 m.).

8. Undeucha'ta major and minor. — These were not found by Wheeler but they are
mentioned by Bigelow (1915, p. 287). They are Gulf Stream species and minor is the
more frequent. They occurred together at Acadia sta^tions 46, 74, and 76. The.y are
found also in the San Diego region whence the male of U. m,ajor has been described
by C. O. Esterly (1905). The male of U. minor has remained hitherto unknown.
Several examples were taken in the Acadia hauls. The structure of the fifth feet of

6551— 16J

188

DEPARTMENT OF THE NAVAL SERVICE

the male of U. minor differs from that of U. major in the absence of the forceps mecha-
nism described by Esterly on the left foot and in some other respects. The chief
differences are displayed in table Q, supplemented by text-fig. 7 and 8. The postero-
lateral angles of the forebody are rounded; rostrum strongly deflexed; anterior
antennas exceed length of forebody.

%.^

Fig. 8. — Undenrha'ta minor cf
Right fifth foot from behind.
i?e = outer branch.

Fig. 7.— .Same. Left fifth foot.

Table Q. — Comparison of fifth feet (p5) of U ndeuchaeta major and minor 6.

Features.

Length

Right p5

Outer branch (Re)

Inner branch (Ri)

Left p5

Re

Re.

Undeuchaeta major cj'
(after Esterly. )

6 ■ 5mm

Biramous

Terminal joint produced into a long stylet.

Siniule .

Biramous

Terminal joint ending in a short style
which carries a pencil of long hairs near
its centre. Second joint bearing a toothed
process " which flares distally " ; " at the
base of this and on the second joint is
aniculated a process, which together
with the terminal joint of the ramus and
the toothed process forms a forceps. "

Vestigial

Undeuchaeta ■miiiord'
("Acadia " material.)

3 5 mm.

Biramous (fig. 8)

Terminal joint with spoon-shaped
e.vtremity.

.Simple.

Biramous (fig 7).

Terminal joint ending in a stylet
with a pencil of long hairs to-
gether with somedenticulations
near its centre, and a stout thorn
at its base. Second and third
joints incompletely separated.

Vestigial

CAXADIAX FISHERIES EX/'EDITIOX, 191^,-15

189

9. Euchirdla rostrata. — Several species of Euchirella occurred in the vicinity of
the Gulf Stream but the most widely distributed was this one. It has a distinctive
appearance with its portly crimson-tinted forebody and short urosoine. It was present
in varying quantity at twenty-four of the Acadia stations but never in the surface
hauls. At stations 85-87 which traversed the Laurentian channel between St. Pierre
and Mrsaine banks, its mode of occurrence is significant in view of what has been said
before. At each of these stations a vertical haul was taken from 30 fathoms to the
surface, and another from 150 or 160 fathoms to the surface. At station 85, Euchirella
rostrata occurred in both of the vertical hauls, two per cent from 30-0 fathoms, five per
cent from 150-0 fathoms; at stations 86 and 87 it was present only in the deeper hauls.
Its distribution within our area is shown on may (fig. 9). At station 54, where it appeasr
with the high plurality of 43 per cent, it should be mentioned that the plankton here
was very scanty, the vertical haul (150-0 fathoms) including only about 140 copepods
in all. Other species of Euchirella were taken as indicated on table XL

©-.©

ySy ®®©

®

(^^

(5)

@

w

Fig. 9.— Distribution of Euihirelhi ro.ttrata. The numbers indicate percentage of cojjepod content.

190

DEPARTMENT OF THE yAYAL SEIiVICE

10. Euchirella acadiana n. sp. — A few examples were taken of the female only of
tins species which I have not been able to fit in with any published description. The
occvirrence was at Acadia stations 41 (200-0 m., one) ; 44 (270-0 m., two) ; 56 (210-
140 fathoms, two). In the following diagnosis Giesbrecht's notation is employed:
Length 6.25 mm. (5 + 1.25) ; rostrum deflexed, pointed (fig. 10) ; genital segment sym-
metrical except for a low oblique brown chitinous ridge on the left side seen from
above near the hinder margin ; postero-lateral angles of f orebody broadly rounded.

Fig. \0— Euchirella acadiana. Profile of head sliowing frontal
organ and rostrum, after removal of right antenna.

Fig. 11.— Same. Proximal joints of right anterior antenna from inner aspect.

Anterior antennae (fig. 11) of normal length, reaching to about the middle of the
urosome.

C.WADI.W FlsnERIi:S KXPEDITIOy. nil ',-15

191

Posterior antennae (fig. 12) with three very long feathery setae at the end of
Re 7; B 1 with plumose Si (as in messviensis) ; B 2 with one Si (as in messinemi'i) ;
Re 1 and 2 incompletely divided, with a crest and spur at inner distal margin of Re 1
(as in curticauda) ; Ri about four-fifths the length of Re 1 and 2; Ri 2 bearing 15
seta\ of which there are six in a row on Le, 7 in a row on Li, the innermost being the
shortest, and in addition a very short Sp on Li near the point where it is continuous
with Le, and a somewhat longer Sp on Le.

^t^

Fig. 12. — Same. Posterior antenna, hinder surface.
The two posterior set* on Ri 2 may be noted.

192

DEPARTMENT OF THE NAVAL SERVICE

Maxilla (fig. 13) : Le 1 with 8 setae (as in messinensis) , the fifth seta about three-
fourths the length of its neighbours; Le 2 small, without seta (as in rostrata) ; Li 1,
hinder surface glabrous (as in rostrata), in the typical group 11-14 there are only
three seta? (as in messinensis) ; Li 2, with 4 setae, Sp 1 and 2 equal, long and stout, Sa
1 much shorter, Sa 2 slender but nearly as long as the Sp; Li 3 with fringe of long
hairs along the length of its anterior surface, at its end which is exactly on a level
with the end of Li 2 there are only two set«e, a distal long and stout seta and near it
a more proximal shorter seta; B 2, setse as in rostrata, surface ciliation as in messi-
nensis: Ri with 5 seta? of which four are subequal and the first, situated at the inner
angle on the iX)sterior surface, is short and slender like the two proximal setae of B 2 ;
Re with eleven setae as in messinensis.

^^.,

Fig. 13. Same. Maxilla, anterior surface. Sette omitted
from Li 1 and Re. Li 2 is partly concealed by Li 'S
whose fringe of hair.s passes over it. S 1 to S 5 are the
five seta? of Ri ; SI is seen with difficulty from this
side, but is very distinct from the hinder aspect.

CANADIAN FISHERIHS EXPEDITION, 191Jrl5

193

Anterior maxillipede (fig. 14) : Distinctive are the close-set fringes of spine-like
hairs on the posterior surface of L 1 to L 4; otherwise as in rostrata; the deep outer
emargination of B 1, a generic character, forms a right angle with the succeeding
portion of the margin which is nearly straight.

Fig. 14. Same, .interior maxillipede
hinder surface, setie omitted.

Posterior maxillipede (fig. 15) : the single seta of the first or proximal group (or
•' lobe ") on B 1 is placed as in Giesbrecht's figure of Chiridius poppei; L 2 with two
setae, one long, one short ; L 3 with three setae (two long, one short) ; L 4 with three
setae, two long, one short; a short longitudinal series of points on anterior surface
of B 2 near its proximal end ; the first seta of B 2 lies very slightly distad of the centre
of the inner margin.

Fig. 15.— Same. Portion of [wsterior maxillipede.

194 DEPARTMENT OF THE yATAL SERVICE

Fourth legs: basal joints with the characters shown in fig. 16.

Fig. 16. Same. Basal joints of left fourth leg.

11. Euchaeta norvegica. — As mentioned this is a eurytropic species occurring in
surface and vertical hauls both inside the Gulf of St. Lawrence and outside. It was
found in different stages corresponding to those detailed for C.finmarchicus and hyper-
horeus. Ovigerous females with single large ovisac full of blue eggs were found,
notably at Acadia station 48 on July 23, at No. 33 station 57 (Bay of Islands) on
August 9, and at Prince station 3 (Bay of Fundy) on September 15. Immature males
were as common as immature females, but the adult males were very rarely captured.
At Acadia station 11 (70-Om.) on May 30 one was found carrying a spermatophore ;
another at station 70 on July 26. The youngest captured was stage II with three pairs
of swimming legs and two-jointed urosome, first noted in No. 33 station 26 on June 26
in the subsurface haul (30-15 metres, towed for 20 minutes).

12. Scolecithrix cuneifrons n. sp. — Of the species of Scolecithrix met with, the
commonest was minor, next to that dame, and then ovata. The appendages of danw,
for a certain length of time after preservation in formalin, have a delicate mauve tint
which enables the species to be recognized amidst a multitude of other Copepods. Sc.
ovata is characterized by the oval fifth legs of the female; in many specimens they
are not to be found.

There were two other species, only one of which I am describing here since it
appears on table XI. I had at first identified it with securifrons but the structure of
the fifth legs of the male seems to differentiate it from that species.

Description of 8c. cuneifrons: Acadia station 46, 150-0 fathoms. Length of
female, 4-5 mm.; of male 4-8 mm.; high frontal crest and acuminate postero-lateral
angles as in securifrons. but in cuneifrons there is an acumination in the male as well
as in the female (fig. 17).

CAN A UI Ay FISHERIES EXPEDITION, IdUflo

195

Fig. 17. Scolecithrix cuneifrons.

A. Frontal profile of female of 4 '5 mm.

B. Frontal profile of female of 3G mm.

C Postero-lateral angle and genital segment 9.
D. Postero-lateral acumination of cf .

Rostrum produced, bluutly bifid at extremity; anterior antennae as long as bod