ogetetstolmpetrecnsotefsI et oxete

Sten
Mibsenirininsodng tli tny tegistogme sin
Geacievenerershenendie’ baie

Division of Fishes,

: U, $. National Museu

DEPARTMENT OF COMMERCE

perl: TIN

OF THE

UNITED STATES
BUREAU OF FISHERIES

VOL, Xxx
1913

HUGH M. SMITH
COMMISSIONER

236631

WASHINGTON
GOVERNMENT PRINTING OFFICE
1915

oo Pe Se

yaeted 8) nassiyd
ruatu lencite® & U

an: wa oi ED, qT ee
reer ee

~~

; ris ik s
tty
we,
ote 1

&

Live

Oe a

~ >

CONTENTS
*

THE EMBRYOLOGY AND LARVAL DEVELOPMENT OF BAIRDIELLA CHRYSURA AND ANCHOVIA
MITCHILLI. By Albert Kuntz. (Document 795, issued May 29, 1914.)..........2-.ee00s
THE SKELETAL MUSCULATURE OF THE KING SALMON. By Charles Wilson Greene and Carl
iartley; Greene, (Document 796, isstied: Augist:8; 194.) en. 012 iecie le oie cie.e/s else sieleleis essiete.e
THE DIRECTIVE INFLUENCE OF THE SENSE OF SMELL IN THE DOGFISH. By G. H. Parker.
(Mocuments7g85 isstiedPAripist (85710 4) sere ois «ys vere eie eicistays sraialcleveres sielclsievere j= eis sresrereteretereree
THE STORAGE OF FAT IN THE MUSCULAR TISSUE OF THE KING SALMON AND ITS RESORPTION DUR-
ING THE FAST OF THE SPAWNING MIGRATION. By Charles W. Greene. (Document 799,
assiieds Se PLemM DELL 30; LO TAs) Mavetcacaystatoensey crore er ohelaney to reiere eke lare ver ooeeicrecaave si teieras sts sialon
CORRELATIONS OF WEIGHT, LENGTH, AND OTHER BODY MEASUREMENTS IN THE WEAKFISH, CyN-
OSCION REGALIS. By William J. Crozier and Selig Hecht. (Document 800, issued Sep-
tember 17) TO TAS) crerererrre revere ern eerste Oe sela hore Sloyors SUVS Sloe seine are Bisie tiers iste iota
THE FAT-ABSORBING FUNCTION OF THE ALIMENTARY TRACT OF THE KING SALMON. By
Charles W. Greene. (Document 802, issued October 28, 1914.) ..-....... see eee e cece
NOTES ON THE HABITS, MORPHOLOGY OF THE REPRODUCTIVE ORGANS, AND EMBRYOLOGY OF THE
VIVIPAROUS FISH GAMBUSIA AFFINIS. By Albert Kuntz. (Document 806, issued October
2), Gb? Baeeeree.c aanoc co OCR OCD CdS CUCU On ATO cOCanec ae Seana aea arr scat ae
SPOROZOGN PARASITES OF CERTAIN FISHES IN THE VICINITY OF Woops HOLE, MASSACHUSETTS.
ByiCaw Haun «(Document )Sro;, issued “April 29; 1915.) cceiseieiaeeeie ieee nelle cies
AN ECOLOGICAL RECONNOISSANCE OF THE FISHES OF DouGLAS LAKE, CHEBOYGAN COUNTY,
MICHIGAN, IN MIDSUMMER. By Jacob Reighard. (Document 814, issued October 11,
TGS) ohn COO BHI RIED AGBA TARE OHORD DURCH On oA ran ate an ocOoeMcC ct aticn cpnoC HOOoH Eom aoT
THE POTAMOGETONS IN RELATION TO POND CULTURE. By Emmeline Moore. (Document 815,
ASSHIEC UU yi ZS RIO LS! ere tators.c rocerelovnsess tere; cicteser visi crereleve efsissevavare reiavecoier e/a svalsvereners aisle ekers io tavexcievaslelere

Page.

I-20
21-60

61-68

69-138

139-148

149-176

177-190

IQI-214

215-250

251-292

—_

‘
‘ ‘
} '
ie
— >
. i
tal
.
i
.
.
‘
.

. -
- 7
- y -
r
‘
a
2
io
.
i
‘aby
7
**
’
a
Pix ¥ se
if Vv _

dba Sb, (Panad

eee is LL) |

1S be yt ee he

ieitvy (ea Ror, uw anni if :

f ih it ah Ar) gk Dg Pitas

=.autu) i | SS OG ee Tee aes

(Toe Gases ina

Hi ule Tees : "
itetaren 2/1. Pa

niger. vt &

ity ' > (fvarga

ein a) ‘ A sama) uf fais

yc eT © ae oe

onG ds AEP RAG IG a Quored

mi hy

yy d Gal SGA eG
Viewty atuleda 5 FOR

i silly Jebus
[CYALss ere LT, Jee

= “eo Tl rts

ie al

THE EMBRYOLOGY AND LARVAL DEVELOPMENT OF
BAIRDIELLA CHRYSURA AND ANCHOVIA MITCHILLI

a

By Albert Kuntz, Ph. D.
School of Medicine, St. Louis University

THE EMBRYOLOGY AND LARVAL DEVELOPMENT OF
BAIRDIELLA CHRYSURA AND ANCHOVIA MITCHILLI.

Bd
By ALBERT KUNTZ, Ph. D.,
School of Medicine, St. Louis University.

ed

INTRODUCTION.

The present paper embodies the results of observations made on the eggs and larve
of two species of teleosts, Bairdiella chrysura and Anchovia mitchilli. The work was
carried on at the United States Fisheries Laboratory at Beaufort, N. C., during the
summer of 1913.

It is not the purpose of this paper to discuss at length any of the merely technically:
interesting points in the development of pelagic fish eggs. Nor does it contribute any-
thing essentially new to our knowledge of the embryology of teleosts. The work was
undertaken for the purpose of securing a record as complete as possible of the time of
spawning and of the embryological and larval development of fishes with pelagic eggs
breeding in these waters during the summer, one of the primary objects being to afford
a ready means of identifying either eggs or larval fishes at any time during embryological
and larval life.

Observations were made as far as possible on living material. The eggs were
collected in the tow net. The larval fishes were taken primarily in the stow net, the
bunt of which was provided with a hood of cheesecloth terminating at its apex in a large
collecting bottle. A small per cent of the larval fishes taken in this manner were brought
into the laboratory alive. The large majority of them, however, were dead before
being taken from the net. .

Eggs collected at the same hour on successive days were found to be in approxi-
mately the same phase of development. Obviously, spawning occurs regularly each
day at approximately the same hour. Observations show that both species under
consideration spawn regularly in the early evening, probably before 8 o'clock.

The eggs of these species are relatively small and contain but little yolk material.
Embryological development, therefore, proceeds very regularly and requires a relatively
short time. The eggs of Anchovia mitchilli require approximately 24 hours for hatching.
Those of Bairdiella chrysura hatch in approximately 18 hours. The time required for
hatching, doubtless, varies somewhat with the temperature of the water. The height
of the spawning season of Baitrdiella chrysura occurs during the last week of June and
the first week of July. Anchovia mitchilli spawns freely during June, July, and August.
The height of the spawning season of this species, doubtless, occurs in July. The
average temperature of the water in the vicinity of the laboratory for the latter half of
June was 27.15°C. The average temperature for the entire month of July was 27.77° (eS

These averages are based on daily readings taken at 5 o’clock p. m.
; 3

4 BULLETIN OF THE BUREAU OF FISHERIES.

The young of Bairdiella chrysura were taken in small numbers at intervals through-
out the latter half of June and the greater part of July. After the spawning season
began to wane very few young of this species were taken. The young of Anchovia
mitchilli were taken in considerable numbers throughout June, July, and August.

BAIRDIELLA CHRYSURA.

Spawning.—The eggs of Bairdiella chrysura were present in the plankton when work
was begun on June 9, and were taken in the tow net nearly every day after that date
until July 18, when they became relatively rare. Individual eggs were taken occasion-
ally as late as August 15. Eggs of this species were at no time abundant. They were
sufficiently numerous, however, to be readily obtained for study. They occurred in
greatest abundance during the last week in June and the first week in July. These two
weeks, doubtless, witnessed the height of the spawning season.

Adult specimens of Bairdiella chrysura were frequently taken in small numbers in
the pound net and in the seine. Nearly all the adult fishes taken during June and July
had already spawned. On June 20 and again on June 27 a single female ripe for strip-
ping was brought into the laboratory. On the former occasion a few eggs were success-
fully fertilized. All of these eggs, however, died during early cleavage.

Egqs.—The eggs of this species are spherical in form and 0.7 to 0.8 mm. in diameter.
The mature unfertilized egg is slightly yellowish in color. The yolk contains a rela-
tively large oil globule. After fertilization has taken place and the blastodisc has
become differentiated, the egg is almost perfectly transparent. The egg membrane is
thin and horny. Between the egg membrane and the delicate vitelline membrane
inclosing the yolk sphere there is a perceptible perivitelline space. The oil globule
normally rests near the upper pole while the blastodise hangs at the lower pole of the
yolk sphere. The spherical form of the egg is maintained until the time of hatching.

Segmentation.—In the mature unfertilized egg the yolk sphere is covered by a thin
layer of protoplasm. After fertilization has taken place the protoplasm of this layer
becomes concentrated at the pole opposite the oil globule, where it forms a lenticular
cap on the surface of the yolk. This lenticular mass of protoplasm is the blastodisc.
The “streaming”? movements which occur in the protoplasm as it becomes concen-
trated to form the blastodisec have been well described and figured by Ryder (1882)
for the cod * and more recently by other investigators for other species of teleosts.

The fully developed blastodise (fig. 1, bd) is circular in outline. Its periphery fades
away almost imperceptibly into the very thin layer of protoplasm which remains at
the surface of the yolk sphere. No protoplasm is noticeable within the yolk except
in the vicinity of the oil globule. Here there is a small amount of protoplasm which
can hardly be detected in the newly fertilized egg, but which, as development advances,
becomes concentrated to form a protoplasmic cap covering about one-third of the
surface of the oil globule.

Just before the first act of cleavage occurs one axis of the blastodise becomes slightly
longer than the other. The first plane of cleavage cuts the blastodise at right angles
to the longer axis (fig. 2). The second cleavage plane cuts the first at right angles
(fig. 3). The first two cleavage furrows are meridional and cut deeply into the

@ Ryder, John A.: Embryography of osseous fishes. Report United States Fish Commission 1882, p. 45s-6os.

BAIRDIELLA CHRYSURA AND ANCHOVIA MITCHILLI. 5

blastodisc. In surface view the early blastomeres appear distinctly outlined periph-
erally (fig. 3). Viewing the early blastoderm in optical section from the side, how-
ever, it is apparent that the blastomeres are not entirely cut off peripherally, but
are continuous with the thin layer of protoplasm at the surface of the yolk. This
condition is illustrated in figures 25 and 27, in eggs of Anchovia mitchilli. The first
four blastomeres are usually quite symmetrical and approximately equal in size. They

BAIRDIELLA CHRYSURA,
Fic. 1.—Egg with fully developed blas- Fic. 2.—Egg with blastoderm of 2 cells.
todisc (bd). X 55. XG5ss

also show a decided tendency to assume a spherical form, as is indicated by the deep
indentations between the cells at the periphery of the blastoderm and the open area at
the center (fig. 3). In the 4-cell stage the two axes of the blastoderm are approximately
equal.

The third cleavage furrows cut the blastoderm approximately parallel with the first.
When the third act of cleavage is completed and the blastoderm is composed of 8 cells,

BAIRDIELLA CHRYSURA.

Fic. 3.—Egg with blastoderm of 4 cells. Fic. 4.—Egg with blastoderm of 16 cells;
sey pb, periblast. X 55.
one axis is again distinctly longer than the other. In the 16-cell stage (fig. 4) the
blastoderm is usually more or less nearly circular in outline.

While blastoderms in the early cleavage stages show considerable variation, cleavage
in these eggs may in general be said to proceed very regularly. The majority of the
blastoderms observed in the 4-cell stage were almost ideally symmetrical. The same
may be said of many of the blastoderms of 8 cells. At this stage irregularities are not
uncommon, however. A marked tendency toward regularity is apparent also in blasto-

6 BULLETIN OF THE BUREAU OF FISHERIES.

derms of 16 and 32 cells. This tendency may still be recognized in blastodorms of
64 cells.

The successive acts of cleavage follow each other in rapid succession. Eggs showing
blastoderms in advanced stages of cleavage may be observed within three or four
hours after the time of spawning. Such eggs were usually observed between 9 and 11
o’clock p. m.

Formation of the periblast.—During the early cleavage stages the marginal cells of
- the blastoderm are not definitely limited peripherally, but are continuous with the thin
layer of protoplasm which remains at the surface of the yolk sphere. At the periphery
of the blastoderm this protoplasmic layer is concentrated to form a low ridge. This
ridge of protoplasm gives rise to the periblast (fig. 4, pb). As segmentation advances
nuclei become apparent in the periblast. These nuclei, as observed by Agassiz and
Whitman (1884), are, doubtless, derived from the marginal cells of the blastoderm.
The cells at the margin of the blastoderm gradually become more definitely limited
peripherally until in the advanced stages of cleavage they are completely cut off from

BAIRDIELLA CHRYSURA.

Fic. 5.—Egg with blastoderm of many cells, late cleavage Fic. 6.—Egg with blastoderm of many cells, late cleavage
stage, surface view; pb, periblast. X 55. stage, lateral view; pb, periblast. X 55.

the periblast (fig. 6). The blastoderm is now more or less dome-shaped and beneath
its central area may be observed a perceptible cleavage cavity. During the later cleavage
stages the periblast becomes somewhat more definitely outlined, increases somewhat
in width, and also sends a thin sheet of protoplasm centripetally beneath the cleavage
cavity.

Formation of the germ ring and differentiation of the embryo.—While the marginal
cells of the blastoderm are becoming cut off from the periblast there appears a slight
thickening at the periphery of the blastoderm. ‘This thickening represents an early
stage in the differentiation of the germ ring. It is caused primarily by the thinning
of the central area of the blastoderm and secondarily by the ingrowth (invagination)
of the marginal cells. The part played by invagination in the formation of the germ
ring and the embryonic shield is discussed at some length by Wilson (1889) in his paper
on the embryology of the sea bass.’ Evidence of invagination first appears at the

@ Agassiz and Whitman: On the development of some pelagic fish eggs. Proceedings of the American Academy of Arts and
Sciences, vol. 20, 1884.

6 Wilson, H. V.: The embryology of the sea bass (Serranus atrartus). Bulletin of the United States Fish Commission, vol.
IX, 1889, P. 209-277, pl. LXXXVIU-CVU.

BAIRDIELLA CHRYSURA AND ANCHOVIA MITCHILLI. FI

posterior, i. e., the embryonic pole of the blastoderm. At this pole a broad tongue of
cells, several layers in depth, may be observed before any evidence of invagination is
apparent around the rest of the periphery of the blastoderm. Figure 7, plate m, illus-
trates an early stage in the differentiation of the germ ring. In this blastoderm invagi-
nation was not yet apparent. The following figure (fig. 8) illustrates a blastoderm
in which the broad tongue of cells is already growing forward from the embryonic pole,
and the entire germ ring is well differentiated. At this stage the central area of the
blastoderm has become materially thinner than the peripheral area. Viewed from the
under side the blastoderm is now distinctly concave. Between its concave surface and
the periblast there is a perceptible subgerminal cavity closed in on all sides by the
germ ring. The blastoderm gradually increases in size by centrifugal growth. The
germ ring, therefore, which in its earlier stages is comparatively narrow, increases in
width both by the invagination of the marginal cells and by the centrifugal growth of
the blastoderm.

BAIRDIELLA CHRYSURA.

Fic. 7.—Egg with blastoderm showing early germ Fic. 8.—Egg with blastoderm showing fully developed
ring (gr). X 55. germ ring (gr) and beginning of embryonic shield (es);
bb, posterior pole of blastoderm. X 55.

While the germ ring is becoming differentiated the cells forming the surface layer
of the blastoderm become thin and flattened. This flattening of the surface cells is
less apparent in the region of the germ ring, especially in the neighborhood of the
embryonic pole, than in the central area of the blastoderm. In the neighborhood of
the embryonic pole the surface cells remain relatively thick and more or less polygonal
in form.

After the germ ring is completely differentiated the blastoderm increases in size
more rapidly than in the earlier stages and advances around the surface of the yolk
sphere. The broad tongue of cells which grows into the subgerminal cavity from
the embryonic pole of the germ ring also increases in size, and the area of the blasto-
derm immediately over this ingrowing tongue of cells becomes differentiated. This
differentiated area represents an early stage in the formation of the embryonic shield
(fig. 9).

Soon after the embryonic shield has become distinctly outlined there occurs a
thickening along its antero-posterior axis. This relatively opaque linear area repre-

8 BULLETIN OF THE BUREAU OF FISHERIES.

sents the axis of the future embryo. We may now distinguish an embryonic and an
extra-embryonic area within the embryonic shield. The differentiation of the embryonic
axis begins in the head region and gradually advances posteriorly until it reaches
the posterior pole of the blastoderm. When the embryonic area becomes distinctly

BAIRDIELLA CHRYSURA.

Fic. 9.—Egg showing later stage in differentia- Fic. 1o.—Egg showing embryonic shield (es) with
tion of embryonic shield; gr, germ ring; es, embryonic area (ea) outlined; eea, extra-embry-
embryonic shield. onic area; gr, germ ring; pp, posterior pole of

blastoderm.

outlined it is somewhat broader in the anterior or head region than in the posterior
region. Observed in surface view (fig. 10) the embryonic area now has a more
or less regular spatulate form. While the embryonic shield is growing forward into
the subgerminal cavity and the embryonic axis is becoming differentiated, the germ
ring is continually advancing around the yolk sphere. By the time the embryonic axis
becomes well differentiated the blastoderm covers
more than three-fourths of the surface of the yolk
(fig. 11).

The further differentiation of the embryo ad-
vances very rapidly and the germ ring continues to
advance round the yolk until the blastoderm covers
the entire surface of the yolk sphere and the blas-
topore is completely closed. In the eggs observed
while the germ ring was advancing round the yolk
sphere the posterior pole of the blastoderm main-
tained approximately the same position with respect
to the oil globule. Inasmuch as the oil globule main-
tains a more or less constant position with respect to
the early blastoderm, it is obvious that the posterior pole of the blastoderm remains at
a relatively fixed point. This Wilson (1889) observed to be the case also in the eggs of
Serranus atrarius. In the eggs under observation the closure of the blastopore occurred
before 1 o’clock a. m. This is probably not more than six hours after fertilization.

At the time of the closure of the blastopore the embryo extends about halfway round
the circumference of the yolk sphere. There is as yet no evidence of pigmentation in
either the egg or the growing embryo. Within one and one-half or two hours after the
closure of the blastopore, yellow chromatophores become sparsely distributed over the

Fic. 11.—Same as figure ro, lateral view. X 55.

BAIRDIELLA CHRYSURA AND ANCHOVIA MITCHILLI. 9

dorsal and dorso-lateral aspects of the embryo. A few yellow chromatophores are appar-
ent also on the surface of the oil globule. The distribution of chromatophores at this
stage is illustrated in figures 12 and 13. Kupffer’s vesicle (fig. 13, Kv) now appears as
a small bubblelike body on the ventral surface near the posterior end of the embryo.

BAIRDIELLA CHRYSURA.

Fic. 12.—Early embryo showing distribution Fic. 13.—Early embryo showing distribution of
of chromatophores, dorsal view. X 55. chromatophores; Kv, Kupffer’s vesicle. X 55.
Lateral view.

An hour later (fig. 14). the chromatophores have become more numerous and are
distributed more or less uniformly over the entire dorsal and lateral surfaces of the
embryo. Kupffer’s vesicle has now reached its maximum development. After this it
gradually decreases in size until it disappears. The length of the embryo now exceeds
half the circumference of the yolk sphere and shows ro to 12 somites.

As development advances and the time of hatching approaches, the distribution of
the chromatophores undergoes a material change. A few hours before hatching the

BAIRDIELLA CHRYSURA.

Fic. 14.—Egg with embryo showing 1o somites; Kv, Fic. 15.—Egg with advanced embryo. X 55.
Kupffer’s vesicle. X 55.

embryo becomes quite active within the egg membrane. The posterior portion of the
body is now free from the yolk sphere and narrow fin folds are apparent both dorsally
and ventrally (fig. 15).

Larval development.—At the time of hatching the larval fishes are 1.5 to 1.8 mm.
in length. The head is slightly deflected at the anterior end of the large oval yolk sac.

10 BULLETIN OF THE BUREAU OF FISHERIES.

The oil globule appears as a yellowish opaque body on the surface of which are scattered
a few yellow chromatophores. It is located in the posterior region of the yolk sac. The
fin folds are continuous. The dorsal fold arises just posterior to the head; the ventral
fold is continuous with the yolk sac. The depth of each fin fold is less than the depth of
the body. The body is
brownish yellow, marked by
five vertical yellow bands.
These vertical bands are
composed of more or less
closely aggregated chromato-
phores. A few scattered
chromatophores occur also
between the vertical bands.
The fin folds and the posterior tip of the body are transparent. Figure 16 illustrates a
larval fish about two hours after hatching.

For some time after hatching the general color of the body remains unchanged.
The distribution of the yellow chromatophores, however, undergoes marked changes.
Five hours after hatching (fig. 17) the vertical bands have become broken up. A.
distinct vertical yellow
band remains located ap-
proximately two - thirds
the distance from the
vent to the posterior end
of the body. Another
less distinct vertical band
occurs just posterior to
the head. Groups of
scattered chromatophores occur in the head region and above the vent. A few more
or less isolated chromatophores occur also on the posterior half of the body.

At one day after hatching (fig. 18) the young fish has grown to a length of 2.4
to 2.6 mm. A small mass of yolk remains unabsorbed. The head is no longer de-
flected, but slightly elevated. The body is distinctly flattened. The greatest depth of the
body occurs posterior to
the head. From this
point the body tapers
gradually toward the
posterior end. The
depth of each fin fold is

Fic. 18.—Barrdiella chrysura. Larval fish 1 day after hatching, actual length 2.5 mm. greater than the depth
of the posterior half of
the body. The general color of the body remains brownish yellow. The fin foldsand the
posterior one-fifth of the body remain transparent. The yellow chromatophores have
become fewer. The posterior vertical band now consists of a dorsal and a ventral group
of chromatophores. There is no distinct vertical band in the anterior region at this
stage, but a few yellow chromatophores remain scattered over the head and the anterior
region of the trunk.

Fic. 16.—Bairdiella chrysura. Newly hatched fish, actual length 1.8 mm.

Fic. 17.—Barrdiella chrysura. Tarval fish 4 to 5 hours after hatching, actual length 2 mm.

BAIRDIELLA CHRYSURA AND ANCHOVIA MITCHILLI. II

During the following day the larval fishes do not increase in size materially. They
undergo material changes in form and color, however. At two days after hatching
(fig. 19) they remain 2.5 to 2.8 mm. in length. The yolk is completely absorbed.
The depth of the head is now greater than the depthof thebody. ‘The fin folds remain
continuous and the depth of each fold remains greater than the depth of the body
posterior to the vent. The general color of the body is light brownish yellow, marked
by two distinct vertical bands. The anterior vertical band is located just posterior to
the head. It is composed of yellow chromatophores on a blackish background. ‘The
general macroscopic
effect of this band is
blackish. The posterior
vertical band is located
approximately two- #
thirds the distance from Si :
the vent to the posterior “Segre
end of the body. It is
composed of a dorsal and
a ventral group of yellow chromatophores on a diffuse blackish background. The macro-
scopic effect of this band is yellowish. Yellow chromatophores no longer appear on other
parts of the body. The fin folds and the posterior end of the body remain transparent.

The critical period for these larve begins during the third day after hatching.
When kept in dishes of sea water they began at this time to die rapidly. Few survived
until the fourth day. Means of keeping the larve alive for a longer period was not
available. Observations on the later larval development, therefore, were made on
larval fishes taken alive in the stow net.

After the critical period is passed the little fishes feed actively and probably grow
comparatively rapidly. Figure 20 illustrates a young fish 3.5 mm. in length. The
relative depth of
the body in fishes
of this size is mate-
rially greater and
the trunk tapers
more rapidly
toward the poste-
rior end than in
larve which have
not yet passed the
critical period. The posterior end of the notochord is slightly elevated. The
posterior end of the body is asymmetrical and betrays an ancestral heterocercal con-
dition of the tail. The fin folds remain continuous. The depth of each fold is now
less than the depth of the body posterior to the vent. The general color of the body is
somewhat lighter than in the earlier larve. Both,vertical bands are distinctly blackish.
Yellow pigment is still present in the vertical bands, but is obscured by the denser
blackish ground color. From the anterior vertical band two blackish bands extend
antero-ventrally. One of these blackish bands terminates in proximity with the eye,
the other extends diagonally over the preopercle and cheek. The posterior vertical

Fic. 19.—Bairdiella chrysura. Warval fish 2 days after hatching, actual length 2.6 mm.

Fic. 20.—Bairdiella chrysura. Warval fish 3.5 mm. in length.

12 BULLETIN OF THE BUREAU OF FISHERIES.

band is composed of a dorsal and a ventral pigmented area. These two areas are now
so widely separated that in lateral view the band no longer appears continuous. Sev-
eral blackish pigment spots occur also along the ventral margin of the body between
the vent and the posterior vertical band.

Larval fishes 5 mm. in length (fig. 21) retain the same general form as the one
3.5 mm. in length above described. The posterior end of the notochord is curved
upward more strongly and the heterocercal character of the tail is more apparent. The
general color of the body has changed to silvery gray. The anterior vertical band and

Fic. 21.—Bairdiella chrysura. Larval fish 5 mm. in length.

the dorsal and ventral pigmented areas in the region in which in the earlier larve the
posterior vertical band is located are distinctly blackish. A small dark area occurs
dorsally opposite the vent. Several small darkly pigmented areas occur also along the
ventral margin of the body posterior to the vent.

As the little fishes grow larger the trunk posterior to the vent becomes relatively
deeper until there is no longer an abrupt break in the ventral contour of the body. The
caudal end of the body gradually becomes symmetrical dorso-ventrally and the tail
assumes its true homocercal character. The general color of the body remains silvery

pe ee Se

ZZ ‘

. ‘

‘
% ‘
‘
‘
H

[SS

SSS

Fic, 22.—Bairdiella chrysura. Tarval fish 7.5 mm. in length.

gray, distinctly darker dorsally than ventrally. The anterior vertical band and the
other darkly pigmented areas are retained until the little fishes have grown to a length
of 8tog mm. After this they gradually disappear. In fishes rr to 12 mm. in length
(fig. 23) there remain only traces of these pigmented areas.

After the little fishes have attained a length of 7 to 8 mm. (fig. 22) they rap-
idly assume the general form and appearance of the adult individuals of the species.
In fishes 10 to 12 mm. in length (fig. 23) the fins are well differentiated and the
full numbers of fin rays are already present. Fishes of this size have the general

BAIRDIELLA CHRYSURA AND ANCHOVIA MITCHILLI. 13

appearance of adult individuals. However, the depth of the body in the thoraxic
region is relatively great and the head is relatively large and blunt. They are also
somewhat lighter in color.

Figure 24 illustrates a young fish 30 mm. in length. The fins are now fully
differentiated and the entire surface of the body is covered with scales. However,
the scales are still small and deeply embedded in the skin. They are, therefore, not

Fic. 23.—Bairdiella chrysura. Larval fish 11 mm. in length.

Fic. 24.—Bairdiella chrysura. Fish 30 mm. in length.

shown in the drawing. In form and color fishes of this size are practically identical
with adult individuals. In short, they show all the diagnostic characters of the species.

ANCHOVIA MITCHILLI.

Spawning.—The eggs of Anchovia mitchilli were present in the plankton when work
was begun on June 9, and were collected in the tow net nearly every day after that date
until August 23, when the work was discontinued. During the second and third weeks
in June the eggs of this species were not abundant, though they were sufficiently numer-
ous to be readily obtained for study. Toward the close of June they became numerous,
and they were much more abundant in the plankton during July and August than the
eggs of any other fishes spawning during these months. The height of the spawning
season is probably reached during July.

As already indicated, this species, like Bairdiella chrysura, spawns regularly in the
early evening, probably before 8 o’clock p.m. Ona few occasions newly spawned eggs
were collected before 6 o’clock p. m. Usually, however, no newly spawned eggs were

19371°—vol 33—15

9
4

I4 BULLETIN OF THE BUREAU OF FISHERIES.

taken before 8 o’clock p. m. Eggs were found occasionally in the early cleavage stages
as late as 9.30 o'clock p. m. Newly spawned eggs were taken in the tow net alike
on the flood and the ebb tides.

Eggs.—The eggs of this species are not spherical, but slightly elongated. The major
axis, which is 0.65 to 0.75 mm. in length, is 0.1 to 0.3 mm. longer than the minor axis.
These eggs are almost perfectly transparent and contain no oil globule. Furthermore,
the yolk is composed of separate masses. It has the appearance under the microscope
of being broken up into large cells. As observed by Wenckebach ¢ in 1886 and later
by other European naturalists, the elongated form of the egg and the segmented char-
acter of the yolk is characteristic also of the European anchovy (Engraulis encrasicholus).
The eggs of this species, however, are somewhat larger than the eggs of Anchovia mitchillr.
The difference in length of the major and the minor axes in the eggs of the former species
also is considerably greater. According to Heincke and Ehrenbaum ? (1900), the greater
diameter of the eggs of the European species is 1.1 to 1.5 mm., and the lesser 0.7 to 0.9
mm. ‘These measurements approximate very closely the dimensions of the eggs of the
American species, Anchovia browni.

Eggs in advanced stages of development and newly hatched larve were rarely
taken in the tow net at the surface of the water. This fact suggests that before the
time of hatching the specific gravity of the eggs is increased sufficiently to cause them
to sink. This conclusion is verified by the results of experimental observations. Eggs
placed in a dish of sea water 12 to 16 hours after fertilization float at the surface for
several hours and then sink to the bottom of the dish. After hatching the larval fishes
may be found at any levelin the dish. The eggs of this species are very delicate. When
placed in a dish of sea water many die before hatching. All the eggs alike, however,
sink to the bottom before any are hatched.

Embryology.—The eggs of Anchovia mitchilli, like those of Bairdiella chrysura, develop
in a manner typical for pelagic teleostean eggs, and the development differs from that
of Bairdiella only in a few unimportant details. The embryological development of
Anchovia mitchilli will therefore be discussed but briefly and with reference to the above
discussion of the embryology of Bairdvella chrysura.

As indicated above, the eggs of Anchovia mitchilli are not spherical, but slightly
elongated. As the thin protoplasmic layer investing the yolk becomes concentrated
to form the blastodisc, the protoplasm “‘streams’’ toward one pole of the major axis.
When fully differentiated the blastodise appears as a lenticular cap of protoplasm lying
on the somewhat flattened lower end of the yolk mass. The periphery of the blastodise
fades away almost imperceptibly into the very thin layer of protoplasm which remains
at the surface of the yolk. Between the thin egg membrane and the delicate vitelline
membrane there is now a perceptible perivitelline space.

Cleavage in these eggs advances with great regularity. It conforms in all essential
details to the process of cleavage, as above recorded, in the eggs of Bairdiella chrysura.
In many instances the early blastoderms in these eggs are even more symmetrical than
in the eggs of the latter species. Early blastoderms which are quite typical of the eggs

@ Wenckebach, K. F.: De embryonale outwikkeling van de ansjovis (Engraulis encrasicholus). Verhandeling der Kaiser-
lichen Akademie van Wetenschappen. 1887.

> Heincke, Fr., und Ehrenbaum, E.: Eier und Larven von Fischen der Deutschen Bucht. II. Die Bestimmung der schwim-
menden Fisheier und die Methodik der Eimessungen. Wissenschaftliche Meeresuntersuchungen, n. f., bd. m, Abteilung Helgo-
land, 1900, p. 127-332, taf. Ix-x.

BAIRDIELLA CHRYSURA AND ANCHOVIA MITCHILLI. Tey

of Anchovia mitchilli are illustrated in figures 25, 26, and 27. Figure 28 illustrates an
egg in an advanced stage of cleavage in which the marginal cells of the blastoderm
are already cut off from the periblast. Eggs in this stage of development were usually
observed between 11 and 12 o’clock p. m.

The germ ring (fig. 29, gr.) and the embryonic shield (fig. 30, es) are differentiated in
the manner described above in the eggs of Bairdiella chrysura. Soon after the germ

ANCHOVIA MITCHILLI.
Fic. 25.—Egg with blastoderm of Fic. 26.—Egg with blastoderm of 4 cells,

2 cells, lateral view. X 60. surface view. X 60.
ring is fully differentiated the blastoderm begins to grow around the yolk more rapidly
than in the earlier stages. The posterior pole of the blastoderm, however, does not
remain at a relatively fixed point, as is the case in many teleostean eggs, but recedes
as the anterior pole advances. As the blastoderm grows around the yolk, therefore, its
center remains at one pole of the major axis of the egg. The blastopore finally closes
at the opposite pole (fig. 34, b/). When the embryo is fully differentiated, therefore,
it lies approximately parallel with the major axis of the egg (fig. 35).

ANCHOVIA MITCHILLI.

Fic. 27.—Egg with blastoderm of 32 Fic. 28.—Egg with blastoderm in advanced
cells, lateral view. X 60. stage of cleavage; pb, periblast. X 60.

In the majority of the eggs observed the blastopore closed between 4 and 5 o’clock
a. m.—i. e., approximately 10 hours after spawning. At this time the length of the
embryo is somewhat greater than half the greater circumference of the egg. Soon after
the closure of the blastopore, Kupffer’s vesicle arises as a bubble-like body on the ventral
aspect of the embryo near its posterior extremity (fig. 35, Kv). The vesicle soon reaches
its maximum development and then gradually decreases in size until it disappears.

16 BULLETIN OF THE BUREAU OF FISHERIES.

After the closure of the blastopore the embryo increases in length until it extends
more than two-thirds around the greater circumference of the yolk (fig. 36). In some
instances, before the time of hatching, the embryo extends entirely around the circum-
ference of the yolk.

Larval development.—Yhe time required for hatching, as already indicated, is
approximately 24 hours. Hatching usually occurs between 6 and 9 o’clock p. m. The

ANCHOVIA MITCHILLI.

Fic. 29.—Egg with blastoderm, showing early germ Fic. 30.—Egg with blastoderm showing fully de-
ting (gr). X 60. veloped germ ring (gr) and beginning of em-
bryonic shield (es). X 60.

newly hatched larve (fig. 37) are 1.8to 2 mm. inlength. The yolk sac, which remains
comparatively large, is greatly elongated and tapers to a point posteriorly. The seg-
mented character of the yolk, already noted in the egg, is still apparent. The head of
the young fish is deflected at the anterior end of the yolk sac. The body is appreciably
flattened and comparatively slender. The fin folds are continuous. The depth of

ANCHOVIA MITCHILLI

Fic. 31.—Egg showing advanced stage in develop- Fic. 32.—Same as figure 7, lateral view; gr, germ
ment of embryonic shield (es), embryonic area (ea) ring. X 60.
outlined. X 60.

each fin fold is less than the depth of the body. The larval fish is almost perfectly
transparent and shows no evidence of pigmentation.

At 12 hours after hatching (fig. 38) the larval fish has grown to a length of 2.6 to
2.8mm. The remaining yolk mass retains its elongated form and its segmented character.
The head of the voung fish is no longer deflected.

The yolk sac decreases in size until at 15 to 18 hours after hatching it is completely
absorbed. For some time after the yolk is absorbed the larval fishes increase in size

BAIRDIELLA CHRYSURA AND ANCHOVIA MITCHILLI. 7,

very slowly. Nor do they undergo any material changes in form or appearance. ‘They
are relatively long and slender and highly transparent. At 36 hours after hatching
(fig. 39) the mouth is apparently functional and soon begins to show the form character-

ANCHOVIA MITCHILLI.

Fic. 33.—Egg showing blastoderm spreading over Fic. 34.—Egg showing blastopore nearly
yolk; gr, germring. X 60. closed; bp, blastopore; gr, germ ring.
X 60.

istic of anchovies. The maxillaries are comparatively long. The lower jaw is long and
narrow. ‘The tip of the head, however, does not as yet extend forward beyond the mouth.

ANCHOVIA MITCHILLI.

Fic. 35.—Egg with embryo showing 18 to 20 somites; Fic. 36.—Egg with advanced embryo. X 60.
Kv, Kupffer’s vesicle. X 60.

The critical period for the larve of this species begins before the close of the second
day after hatching. When kept in dishes of sea water many of them died before reaching
the third day. Observations
on the later larval develop-
ment were made on larval
fishes collected in the stow
net.

Larval fishes 3 to 4 mm.
in length (fig. 41) do not differ
markedly in appearance from
larve in which the yolk sac is just absorbed. They retain the same general form
and remain almost perfectly transparent. The fin folds remain continuous. Their
relative depth, however, has materially decreased.

ic. 37.—Anchovia mitchilli newly hatched, actual length 1.9 mm.

18 BULLETIN OF THE BUREAU OF FISHERIES.

Fishes 5 mm. in length (fig. 42) illustrate an early stage in the differentiation of the
dorsal and the anal fins. In larve of this size the posterior region of the intestine is
already convoluted. In lateral view these convolutions have the appearance of vertical
folds. ‘This character is apparent externally until the little fishes have attained a length
of 15 to 20 mm.

In fishes 7 to 8 mm. in length (fig. 43) the dorsal and anal fins are becoming
definitely outlined. In some instances the full number of fin rays is already present.

Fic. 42.—Larval fish 5 mm. in length.

ANCHOVIA MITCHILLI.

A few small darkly pigmented areas are now apparent along the ventral margin of: the
body in the thoracic region and at the base of the anal fin.

As the young fishes grow larger they become less transparent, but show very little
pigment. They undergo no marked changes in form, but gradually assume the appear-
ance of adult fishes, showing all the diagnostic characters of the species. The silvery,
longitudinal band characteristic of adult anchovies, however, does not appear until
the young fishes have attained a considerable size.

BAIRDIELLA CHRYSURA AND ANCHOVIA MITCHILLI. 19

During the early summer larvee of Anchovia mitchilli and Anchovia brownti were fre-
quently taken together. In this stage the two species are very similar and might readily
be confused, the larve of the latter, however, being somewhat the longer and compara-
tively more slender. ‘The vent is also located correspondingly farther posteriorly in the
latter than in theformer. As soon as the dorsal and anal fins have become fully differen-
tiated, the young of either species may be recognized by the character of the anal fin.
The number of anal fin rays in Anchovia brownii usually does not exceed 20. In Anchovia
mitchils the anal fin rays number 25 to 28. In the latter species the anal fin also is longer
and terminates less abruptly and nearer the base of the caudal fin than in the former.

Fic. 46.—Adult fish 7 cm. in length.

ANCHOVIA MITCHILLI.

Figure 46 illustrates an adult fish. The adult of this species does not differ markedly
in form and appearance from the adult of Anchovia brown. ‘The average length of the
body is somewhat greater and its relative depth is somewhat less in the latter, while
the silvery lateral band of A. mitchilli is narrower and less distinct than in brownit.
More distinctive characters are the anal fin, as indicated above, and the position of
the vent. In the larve of both species the vent is located opposite the middle of the
dorsal fin or farther posteriorly. In the adult of Anchovia mitchilli the vent is located
opposite the origin of the dorsal fin, while in the adult of Anchovia brownii the vent is
located approximately opposite the middle of the dorsal fin.

_

1 7

Nites Binwetis 7 aa Ripa | a my easel
7
if 7 17! ir Sa ie fae i ne 7 nos 7 by aly ae on fre

-

ean, ri 1 pay Poy, co ra) Bilt gall ee
Mita i ie eae nied | ier a4 : (oat 19; 7
Glin. == my Vis fr? hy Tip Ste Tea 7 i lel vi Tie :
On Min antes Theat gt By ne uae | Vis we THOM ia we

oe is yt 4 ibesady SOL! | TS Nace orhtgy!

creel wl tal en eee wna « Walla ae ale tition: Ol fe Ire:

: eters Abu ae ce = are " Pa) See” are, Cae
Seep th ad rey v4 TG te i’ ‘ye < nt Breaay “iN cae 7 ‘Tar ree

ry 7

4 Ye =
yf - 7 : : :
a 4 - a snd 7 oe Pn ~ so
Vil ae a — — a ay ra a Vhs we
f in i Tuy i. aA a aS eas
i fF : - pea Vie ar oe AL : aap neni Sy Nf
: 7 ine " Te 48 Clas WD Roser Te —
_, 7 a= — ” Ar any = 8 ae
- “= 7 = > aan .
iat te |
ey ; - ——
St. i _ =
ant. ba
> \ wi ii
, oy = ee
it j= < : —
_ ~ — 7 i= 7
eee ea
a. 4g a
: 7 oe _
_ _ aii
7 7? a een ae =
ee
- 7 | oe an
: wh ) —_
- Mae aa ;
: 5, ; a9 0 - -
a ar =
é' oh i “ eayt! _ ais
Wied %, :

f ee, ta
‘me ae

“¢f : 7 a S ©. .

ae tas a Oy otk Hina ye? Gi;

7 a Ve ‘ yay 7
PAN kee wy re i ane un on a . hs} Riou) Ted elie me aia Grae och

a wl Wer ten ol & = paints nla. (anos he hee Pilih ay AD a iy nae cteiviy out He ets) \i
Clipin \t)! ne at ' i = aise a} ines Ay jonda 1 > eae nh Lt tee, ri 4

ee Ay die! ne ui vl netit th Satine ‘a “i aoe “yates Bylih an m7. Al
eral Dad eG 4 Ta ; weil. Rg | Mat ey “its hve re a: or pl Cire” 7 cok i

—_ om a r a tts it dag lPA YL? “be aL e * ee || aia ie ne ati rs wee oii
oo me f mn if =) = at * Ae ahs Nhe an a, bay z 1 ny us 3 Ay Pree 3 ip nt a i : teen o, 4
af BR Dler ne oats poet cksly WY at - nt mith Vigs ae ‘aah ins
; > * lu Daigy i ‘al ins bse mre in paige Bal ae
; i 7 ai: ;
q : 7 : * © 7
+1 i i - _ te
: 7 ! A : i j
Hy 7
- - a —
be -_ i e A af 7

THE SKELETAL MUSCULATURE OF THE KING SALMON
ad

By Charles Wilson Greene and Carl Hartley Greene

Department of Physiology and Pharmacology, Laboratory of Physiology
Umversity of Missouri

+

= » ©
- iw oe
i #»
‘=
: , :
a)
STA AR oe
saa A reali i soivant a vez *
. a
iy bee ae
-
<

CONTENTS?
&*

General arrangement of the skeletal musculature
Muscles of the trunk, the longitudinal muscles
Muscles of the head region
Muscleslolithescaticale fingers systesas cs kseierietierr ee rccormsicis & Teicrsks, anc raters byes, ciel vln eter ast eio travis versione
Muscles of the pectoral girdle
Miusclesiofsthevpelvicioird ecm ae. tres stiyster eh iret eerie tej sie create eee maaan
Muscles of the dorsal fin.............
Muscles of the anal fin

Vl sel® il a Hg
, 1 fi ceplaid yeast ‘
—— it by (outa

114 lel? &2

THE SKELETAL MUSCULATURE OF THE KING SALMON.

ad

By CHARLES WILSON GREENE and CARL HARTLEY GREENE,
Department of Physiology and Pharmacology, Laboratory of Physiology, University of Missouri.

ad
GENERAL ARRANGEMENT OF THE SKELETAL MUSCULATURE.

The general muscular structure of the king salmon has not previously been described.
One must therefore be guided by the general descriptions and comparisons as between
the musculature of the different parts of the body of the king salmon and that of other
fishes of related groups.

The major amount of the muscle mass of the salmon is represented by the great
lateral muscles. These extend from the head and pectoral arch on either half of the body
straight along the sides to the base of the tail. Each great mass is grossly divided
longitudinally into dorsal and ventral portions, vertically into the well-known segments
or myomeres. Out of the extreme dorsal portion of the mass certain special longi-
tudinally arranged muscles have been developed. Along the mid-ventral line similar
longitudinal differentiations have occurred. In the head region the usual complex
differentiations of muscles have taken place. These muscles are undoubtedly derived
primarily from the great lateral muscle.

In like manner, the muscle region at the base of the tail, the caudal peduncle, has
been differentiated into a number of definite and special muscles which produce the
complex movements of the caudal fin.

The pectoral girdle and the pelvic girdle, together with the corresponding fins,
have a number of highly specialized muscles. Also the great median fins, the dorsal
fin, and the anal fin, each are supplied with a complex group of muscle filaments.

These great groups form the basis of the subdivisions which are followed in this
paper in describing the muscles in anatomical detail. In gross outline the groups are
as follows:

Muscles of the trunk, the great longitudinal muscles

Muscles of the head region.

Muscles of the caudal fin.

Muscles of the pectoral girdle.

Muscles of the pelvic girdle.

Muscles of the dorsal fin.

Muscles of the anal fin,

25

26 BULLETIN OF THE BUREAU OF FISHERIES.
MUSCLES OF THE TRUNK, THE LONGITUDINAL MUSCLES.

The longitudinal muscles have been developed out of the great lateral muscle and
form the major mass of muscle substance on each side of the body of.the king salmon
as in other fishes. The extreme anterior portion of the embryonic lateral muscle is
differentiated into numerous highly specialized small muscles in the head region, and
it is similarly, though less complexly, differentiated in the caudal region also.

The great lateral muscle is subdivided both longitudinally and transversel .
Longitudinally the subdivisions are indicated superficially by more or less distinct
longitudinal connective tissue areas. The most developed and largest septum how-
ever, is indicated by the connective tissue band lying immediately under the lateral line,
where a thick septum extends from the under surface of the skin directly down to the
lateral-ventral surfaces of the centra of the vertebral column. This septum completely
divides the great lateral muscle into dorsal and ventral portions, the division extending
from the base of the skull to the middle of the base of the caudal fin. Kingsley, in his
Comparative Anatomy of Vertebrates, speaks of these great divisions as the epaxial
and hypaxial muscles.

The extreme dorsal portion of the epaxial muscle on each side has become further
differentiated by the separation of a definite cylindrical bundle extending from the
occiput to the base of the tail, but interrupted at the dorsal fin, and modified at the
soft dorsal fin. This muscle is the homologue of Owen’s supracarinalis.

That portion of the lateral muscle lying below the lateral line, the hypaxial, has its
extreme ventral portion cut off into definite masses which for the most part are cylin-
drical in form. ‘This portion is the homologue of what McMurrich % calls in the cat-
fish “the great ventral muscle,” the infracarinalis of Owen. It extends from the gular
plate to the base of the caudal fin, but is interrupted at the pelvic girdle and at the anal
fin respectively.

The lateral muscles proper are further differentiated into a superficial and a deep
portion. These subdivisions are rather intimately bound together at their surfaces of
approximation. But in gross anatomical features, in minute histological structures, and
in physiological properties, they are so characteristically different that they were con-
sidered as distinct muscles. They have been described and given distinctive names by
the senior author.?

The entire list of longitudinal muscles, including the divisions of the lateral muscles
and special differentiations at the mid-dorsal and mid-ventral regions, is as follows:

I. Divisions of the great lateral muscle proper.
1. Musculus lateralis superficialis.
a. The epaxial division.
b. The hypaxial division.
2. Musculus lateralis profundus.
a. The epaxial division.
b. The hypaxial division.

@ McMurrich, J. P.: The myology of Amiurus. Proceedings of the Canadian Institute, vol. 1, p. 330.
b Greene, Chas. W.: An undescribed longitudinal differentiation of the great lateral muscle of the king salmon. Anatomi-
cal Record, 1913, vol. 7, Pp. 99-101.

SKELETAL MUSCULATURE OF THE KING SALMON. 27

II. Supracarinales, the dorsal longitudinal muscles.
3. Protractor dorsalis.
4. Retractor dorsalis.

III. Infracarinales, the ventral longitudinal muscles.
‘5. Protractor ischii. ;
6. Retractor ischii (protractor analis).
7. Retractor analis.

DIVISIONS OF THE GREAT LATERAL MUSCLE,

The great lateral muscle, as the term is applied to the adult fish, does not include
the dorsal and ventral differentiations given in II and III of the above list. It does,
however, include all the muscle mass extending from the base of the skull and the pec-
toral girdle to the base of the caudal fin except the supracarinales and the infracarinales,
respectively. This mass, as just described, is divided longitudinally into four actual
divisions. An epaxial and a hypaxial portion is formed by the lateral line septum.
Each of these great divisions is differentiated longitudinally into a thin superficial por-
tion and a thick deeper portion as previously indicated. Each of these may now be
described in fuller detail.

MUSCULUS LATERALIS SUPERFICIALIS.

This muscle extends over the surface of the deeper division of the great lateral
muscle, the profundus, for its full extent from the head to the base of the tail. It is
thickest in the mid-lateral line. There are two separate and distinct portions, the
epaxial and hypaxial divisions. Each of these divisions forms a thin sheet, becoming
thinner as it extends out from the lateral line, dorsally in the epaxial and ventrally in
the hypaxial divisions, respectively. The muscle is several millimeters thick in the
king salmon in the lateral line region, while its extreme dorsal and ventral borders are
represented in thickness by only a few fibers. The dorsal limit of the superficialis is
along the line about two-thirds the distance from the lateral line to the mid-dorsal
line of the salmon body. ‘The ventral division varies somewhat in its extent. In the
anterior portion of the body the superficialis extends only about one-half the distance
from the lateral line to the mid-ventral line. In the posterior part of the body the mar-
gin of the superficialis extends two-thirds to three-fourths this distance. These epaxial
and hypaxial divisions of the superficialis muscle are sharply separated from each other
by the lateral line septum.

The muscle as a whole is characterized by a darker appearance than the profundus.
The latter is the usual salmon pink color in the well-conditioned fish, though lighter in
color in the fish of poorer quality. The superficialis is separated from the profundus by
a rather weakly marked sheet of connective tissue. In macerated examples the super-
ficialis can readily be separated from the profundus. On the whole, however, the two
muscles are very intimately connected. Histologically the demarcation line is sharp
and distinct, but by methods of gross anatomy this line is not so readily determined.

The superficialis has been observed by the senior author in a number of other
fishes. In some of these, for example the California sardine, Clupanodon ceruleus,
this muscle is relatively more highly developed than in the king salmon. In the

28 BULLETIN OF THE BUREAU OF FISHERIES.

literature, however, thus far no previous reference to or description of this differentia-
tion of the great lateral muscle in other fishes has been found other than the sentence
of Miescher’s quoted below.* Miescher speaks of ‘‘a thin muscle plate lying along the
side of the body just beneath the skin which degenerates strikingly (cutaneous muscle).”’
I interpret this statement as referring to the lateralis superficialis, though there is
nothing else in the context that suggests that Miescher recognized this portion of the
lateral muscle as a differentiation out of the total mass. The differentiation is described
in part by the papers of the senior author dealing with subjects in salmon anatomy and
physiology.?

Histologically the superficialis is distinguished from the profundus by its strikingly
different type of muscle fibers. The fibers of the superficialis are more compact, more
uniform in diameter, and relatively smaller in size than the fibers of the profundus.
The fat-storing property of this muscle has been specifically described in a previous
paper.© Analyses made of this muscle showing the percentage of fat in the fish from
the mouth of the Columbia River gave the total of the fat in the fresh wet muscle as
high as 30 per cent. In no other muscle of the salmon is such an enormous quantity
of fat stored, and especially nowhere else are such quantities stored within the fibers.

MUSCULUS LATERALIS PROFUNDUS.

This muscle forms the major portion of the great lateral muscle as defined above.
It extends from the occiput and pectoral girdle to the base of the caudal fin. The
muscle is characterized in the fish of first quality especially by its rich pink color. The
body of the profundus fills the entire space between the superficialis and skin on the
one hand, and the skeletal complex on the other. The two divisions, the epaxial and
the hypaxial, are very sharp and distinct for the entire extent of the muscle. The
attachments of the muscle are better understood after a discussion of the arrangement
of its segments.

The profundus is distinguished from the superficialis always by its characteristic
difference in color, as previously referred to. The king salmon in the Columbia River
shows an especially rich color in this muscle, though the color fades as the period of
starvation progresses during the spawning migration. The form and size of the fibers
vary within wide limits while the length of the individual fibers remains more
constant. In contradistinction to the superficialis the fibers of the profundus vary
in diameter from 25 or 30 to as much as 200 or 250 micra. No such variation in size
occurs in the fibers of the superficialis. This characteristic alone is sufficient to diag-
nose the two muscles.

@ Miescher, Friedrich: Statistische und biologische Beitrage zur Kenntniss vom Leben des Rheinlachses im Siisswasser.
Schweizerischer Fischerei-Ausstellung zu Berlin, 1880, p. 186. Also reprinted in Histochemische u. Physiologische Arbeiten
von Friederich Miescher, 1897, p. 145. Miescher’s exact words are: ‘‘Am starksten degenerirt eine gesonderte diinne Muskel-
platte, die an der Seite des Kérpers direct unter der Haut liegt (Hautmuskel).””

> Greene, Chas. W.: The storage of fat in the muscular tissue of the king salmon and its resorption during the fast of the
spawning migration. Bulletin U. S. Bureau of Fisheries, vol. xxx, 1913.

c Greene, Chas. W.: A new type of fat storing muscle in the salmon, Oncorhynchus tschawytscha. American Journal of

Anatomy, vol. 13, 1912, p. 175-178

SKELETAL MUSCULATURE OF THE KING SALMON. 29
MYOMERES OF THE GREAT LATERAL MUSCLES.

The entire lateral muscle mass, including the superficialis and profundus of both
the epaxial and the hypaxial divisions, is subdivided into vertically marked segments,
the myomeres (Wiedersheim). The myomeres are separated by connective tissue
septa, the myocommata. The septa, and hence the myomeres, are not simple vertical
sheaths but are very complexly folded ‘‘so as usually to form semiconical masses”’
(Owen).* ‘The surface markings of the septa, forming the borders between the
myomeres, present zig-zag lines across the sides of the fish. These septa are not so
simple as the surface lines would indicate, as shown in the figure presented (pl. 1).
From this figure of three myomeres taken from about the middle portion of the body
it is obvious that each myocomma as a whole forms a rather complex membrane.
Owen has described the form of the myocomma in Perca fluviatilis and ilustrated the
same with a fair figure. Allis ° figures the surface markings of the anterior portion of
the body of Amza in his figure 33, the deep folds of the myomeres in figure 34, and the
septa after dissecting away the muscles in figure 35, all of the same region. Allis’
figures are splendid artistic reproductions of the anatomical facts. The region figured
by him is near the pectoral girdle where the myomeres and septal folds are relatively
simple.

The form of the myomere and of the septum varies somewhat in different regions
of the body but is always complex and intricate. The variations are from one and
the same type. In that part of the body from which the figure is taken, in fact also
the myomeres of the entire side of the salmon, the surface markings have the general
outline of the letter ‘“W”’ with the bottom of the letter turned toward the tail. The
middle limb of the curve coincides with the lateral line. (See pl. 1.) For the entire
anterior half of the body the myocommata at the mid-line form sweeping curves. At
about the anterior border of the anal fin this curve gives way to a point of gradually
increasing sharpness. On the caudal peduncle at the lateral line each myocomma
makes a sharp pointed union as between the dorsal and ventral halves.

The dorsal or epaxial half of the musculature has the bend in the myocommata
directed posteriorly. That portion of the myocomma on the surface between the lateral
line and the mid-dorsal bend runs in a sweeping curve, almost vertical at the anterior
portion of the body, set at an angle of about 60° under the dorsal fin, about 45° over the
middle portion of the anal fin, and about 30° on the caudal peduncle. From the middle
of the epaxial muscle to the dorsal margin the myocomma forms a sweeping curve toward
the head, at first at an angle of about 50°, then curving until just at the dorsal margin
the angle is about 10° to 15°, measured with reference to the lateral line. The line
marking the union between the dorsal and dorso-median curves of the myocommata
lies about three-fifths the distance from the lateral line to the base of the dorsal fin.

The surface of the ventral half or hypaxial muscle shows similar curves of the
myocommata. ‘The median portion very closely follows the angle formed by the ribs
along the sides of the abdominal wall. Posteriorly the inclination is ever increasing,
reaching its maximum of about 30° at the caudal peduncle.

@ Owen, Richard: Comparative anatomy and physiology of vertebrates. vol.1, p. 203. 1866.
b Allis, Edward Phelps: The cranial muscles and cranial and first spinal nerves in Amiacalva. Journal of Morphology,
1897, Vol. 12, DP. 487-808.

19371°—vol 33—15——3

30 BULLETIN OF THE BUREAU OF FISHERIES.

The ventral limb of the hypaxial portion of the myocomma, like the dorsal limb,
is very oblique, curving anteriorly. Directly under the pectoral fin this angle is about
70°, in the neighborhood of the ventral fins the angle is about 40°, and between the
ventral fin and the caudal fin it varies from 40° to 20° measured with reference to the
lateral line. The myocommata are placed most nearly horizontal just above the base
of the anal fin.

The form of the septum, i. e., the myocomma, is more clearly shown from plate II
if one follows only the outlines of the most anterior of the four myocommata presented,
considering primarily the relations of the superficial margin to the deepest margin. The
deep margin is in contact with the skeleton and continuous with the median septum or
skeletal membrane. Considering the whole septum the superficial zigzag markings are
shallow while the zigzag outlines of the skeletal border are deep. In other words, the
skeletal boundaries of the septum in the mid-lateral line are attached several centimeters
in front of, i. e., cephalad to, the point at which the septum is attached to the skin on
the surface. In a similar manner, the skeletal borders of the mid-dorsal and of the
mid-ventral portions of the septum are attached back of, that is caudal to, the cor-
responding superficial borders. Posteriorly, i. e., over the anal fin (pl. 1), this arrange-
ment of the myocommata and myomeres is much more extended in the longitudi-
nal axis of the salmon. When a given myocomma of the posterior surface of the epaxial
half of the body is exposed it is seen that the segment ends in a slender wedge directed
caudally, the surface in a particular case being 27 mm. farther back, i. e., caudally,
than the surface at the mid-line. The deep or skeletal attachment of the same sep-
tum was 55 mm. behind the mid-line surface point. Just at the lateral line the deep
portion of the septum dips far forward. ‘The septa of the successive myomeres form
long slender conical sheaths extending from the under surface of the skin anteriorly
down to the skeleton. This distance amounts in the above case to 90 mm.

The significance of this arrangement can be explained only when one keeps in mind
that the individual muscle fibers % of the myomeres run in lines closely paralleling the
axis of the fish. There are many variations from this rule; nevertheless, the general
effect is a relation between the muscle fiber and its septa which gives to the latter the
effect of tendons. This relation enormously strengthens the whole mechanism of
myomeres and septa as a power-producing machine. Figure 1 attempts to show this
advantageous arrangement in a diagrammatic way by a somewhat idealized section
through the anterior conelike fold just under the lateral line and of the posterior dorsal
fold above the lateral line.

The alternate contractions of the great lateral muscles accomplish the propelling
of the body forward in the act of swimming. ‘The skeleton is like a great flexible board.
The masses of the myomeres of either side are mechanically so knitted into this support
by the complex attachments of the myocommata to the skeleton that when a contrac-
tion occurs the force of the act is distributed over an unexpected distance along the

@ Measurements of length of fibers in the myocommata at points on the surface: At the anterior margin of the dorsal fin
at the lateral line, 7.2 mm.; at the dorsal mid curve, 6 mm.; in front of the dorsal fin near the dorso-median line, 3.6 mm.;
ventrally 3 cm. below the lateral line, 7 mm.; 6 cm. below, 8.2 mm.; fibers running obliquely down and back just in front of the
pelvic fin, 5.8 mm.

Measurements just over the anal fin: Dorsal, 2.5 mm.; at the bend, 6.5 mm.; at the lateral line, 8 mm.; deep fibers directly
under this region and 1 cm. dorsal to the lateral line the pink fibers measure 2.8mm. On the ventral line of the muscle appar-
ently the same general variation in length of fibers occurs. At the point where the myocommata run most obliquely just above
the base of the anal fin the fibers measure 3 mm.

SKELETAL MUSCULATURE OF THE KING SALMON. 31

length of the fish. In the caudal region, for example, this extent is so great that the
contraction of a single myomere, should it occur, would bend the skeleton toward that
side through an extent of several segments. The longitudinal extent of a myomere in
the caudal region, opposite the anal fin, is 12 centimeters, i. e., 15 myomeres, of the
muscle. The alternate cone-like folds of the septa mutually support each other. It
SOB
SSSSSS
SSS
SS

BO

Fic. 1.—Diagrams to illustrate the mechanical relations of the muscle fibers and tendonous septa of the lateral muscles. The
diagrams should be considered in comparison with the dissections presented in platestandu. A, positionin rest; B, position-
during contraction of the left side.

The figures are drawn to represent a composite view of an idealized transverse plane that world cut the individual myomeres
and septa through the greatest longitudinal extent. This plane cuts the anterior fold in the median line and the posterior fold
through a plane somewhat dorsal to the median line. The posterior folds are less oblique to the skeletal axis than are the anterior
folds. More anteriorly the septa will be less oblique, posteriorly more oblique than shown (see pl.1). ‘This diagram is con-
structed for the region under the dorsal fin. Note that during contraction of one side the individual fibers on the opposite side
are stretched slightly, a condition favorable to the expenditure of contractile energy. Note also that the muscle fibers retain
their relatively parallel position with reference to the adjacent skeletal axis. The anterior folds of the septa of the anterior sur-
faces of the myomeres act as anchors against the posterior folds (dorsal and ventral) of the septa of the posterior surface.
As both are inelastic they serve as admirable tendons. Considering the depth of the septa it is obvious that flexion will increase
the thickness of the mass of muscle slightly. But the anchoring is such that during flexion parallel septa move or shear over
each other in such a way as to produce a maximal amount of movement of the trunk of the salmon by a relatively small amount
of muscle fiber shortening, a most advantageous physiological justification of a complex anatomical mechanical relation.

is obvious that the successive septa are very close together and that the fibers from
one to the other run very obliquely. In other words, when a contraction occurs every
individual fiber is in the best mechanical position to expend all its energy in a much
more direct pull on the septal sheet and on the skeleton than would be the case if the
myocommata were simple vertical septa placed at right angles to the axis of the fish.

32 BULLETIN OF THE BUREAU OF FISHERIES.

Furthermore, as a contraction progresses and the body of the salmon is sharply
curved, i. e., concave, to the side involved, the muscles pull even more directly on the
skeleton than at the beginning of the movement, as the figure shows. When the lateral
muscles on one side thus bend the ends of the body toward that side, the muscles will
pull along the line of oblique attachment of the anterior myocommata on the one hand
and the similar attachments of the posterior myocommata on the other, so that these
two sheaths serve as direct tendons for the muscle fibers. The arrangement is such that
this relation holds for almost every portion of the myomere.

If the septa were simple vertical connective tissue sheaths the mechanical conditions
would be wholly changed. In such a case the power expended by the contraction of
each myomere would result in a pull on the adjacent myomeres only and from segment
to segment and not a direct pull on the skeleton. Only when the great lateral muscles
contracted for their full extent would the individual myomere exercise its greatest
mechanical possibility. Even then the fibers toward the surface of the myomere would
at the time of their maximal contraction soon reach their physiological limit of shortening.
The total effect would be to produce tension drawing the superficial part of the muscle
away from the skeleton in a relatively inefficient pull. The actual and natural arrange-
ment of the structures in the king salmon is far better and forms a wonderfully efficient
and economical mechanical-physiological device.

SUPRACARINALES, THE DORSAL LONGITUDINAL MUSCLES.

Lying along the extreme dorsal margins of the lateral muscles on either side of the
body are separate and well developed muscles, the supracarinales. These paired muscles
are imbedded in distinct and heavy connective tissue sheaths. In describing the supra-
carinales the muscles should be considered as made up of two divisions: (1) That portion
between the scapula and the anterior portion of the spinous dorsal, and (2) that portion
between the posterior margin of the spinous dorsal and the caudal fin. This latter is
sharply divided into an anterior and posterior division by the soft dorsal. These two
muscle divisions acting together tend to flex the body in the dorso-ventral plane, which
in the salmon would seem to be their chief function. Acting separately, each division
may be assumed to move the spinous dorsal fin, the first division forward, i. e., in pro-
traction, the second division backward, in retraction. From this latter function the
homologous muscles in other fishes have received their names and these names are used
here

PROTRACTOR DORSALIS.

This relatively strong muscle extends from the dorsal end of the scapula to the
anterior margin of the spinous dorsal and is about 25 cm. long in an 80 cm. salmon.
Anteriorly the fibers of the muscle are spread out into a relatively broad fan-shaped mass
about 2.5 cm.in width. The mass of the muscle is correspondingly thin in this region.
From the middle to the posterior end of the muscle the outline is almost circular, the
fibers forming a distinct strong cylindrical bundle even up to the point of insertion.
The diameter of this cylindrical mass is from 8 to 10 mm. in an 80 cm. standard fish.
Each muscle lies in a tendinous sheath (one on either side of the mid-line of the body).
The sheath is less strongly developed anteriorly. The different relations of the walls of

SKELETAL MUSCULATURE OF THE KING SALMON. 33

this sheath are as follows: Superficially there is a relatively thin connective tissue
sheath separating the muscle from the skin covering it. This portion is heavily loaded
with fat. On the ventral surface of the muscle is a thick septum extending from the
skin to the median septum into which it is strongly knitted. The median wall is formed
by the superior portion of the median septum, in which are imbedded the interneurals of
the skeleton. The outlines of a cross section of the sheath are irregular though approxi-
mately circular, the outlines being slightly flattened where the septum is strongly
developed.

The protractor dorsalis is segmental in the arrangement of its constituent fibers.
Connective tissue septa, the homologues of the myocommata of the lateralis superficialis
and profundus, extend through the muscle but in an irregular and complexly folded way.
In other words, the septa are not simple transverse membranes, but form cone-like
spirals. The fibers composing them are strongly interlaced producing in effect a ten-
dinous skeletal framework in which the muscle fibers are imbedded.

The attachments of the muscles are as follows: The anterior end is attached into
the posterior margin of the dorsal end of the scapula and by a strong superficial apo-
neurosis into the skin over the scapula and occiput. This fascia extends forward to the
occipital and temporal bones. The tendinous fibers of the posterior end of the pro-
tractor are knitted into the anterior and superior margins of the two or three interneurals
lying under and supporting the most anterior rays of the dorsal fin. However, all along
the median border of the muscle tendinous slips are strongly inserted into the median
septum and the interhemals imbedded in this portion of the median septum.

Contractions of the protractor fibers produce traction not only as between the dorsal
fin and the occiput, but all along the line of the dorsal margin of the median septum.
The whole mechanical effect of the attachments is more favorable for the production of
a strong dorsal flexion of the body of the fish than for a protraction of the dorsal fin.

RETRACTOR DORSALIS.

That portion of the supracarinalis lying between the posterior margin of the spinous
dorsal and the superior margin of the caudal fin receives the name of retractor dorsalis.
This muscle isa cylinderinform. The anterior attachment is by a short tendon inserted
into an irregular shaped vertical plate which forms a joint with the last interneural
spine, the spine lying under the most posterior dorsal ray. The plate is a modified
and enlarged free end of an interneural to receive the tendon of the retractor. The
posteriot tendon of the retractor is rather broadly attached to the connective tissue
enclosing the dorsal ends of the interneural spines of the caudal peduncle which lie
under and support the dorsal rudimentary rays of the caudal fin.

The retractor does not seem to be so intimately knitted into its division of the
median septum as in the case of the protractor; it is, indeed, free for most of its course.
The fact that the muscle is relatively short and smaller in its absolute size than the
protractor is probably associated with a development which has separated it from the
median septum. That portion of the retractor lying between the soft dorsal and the
caudal fin is very slender, 2 or 3 mm. only in diameter. Under the soft dorsal the muscle
is wholly tendinous and is closely attached to the base of the fin. Possibly it would
be better to consider the two divisions as distinct muscles separated by the soft dorsal.

34 BULLETIN OF THE BUREAU OF FISHERIES.
INFRACARINALES, THE VENTRAL LONGITUDINAL MUSCLES.

Longitudinal muscles lie along either side of the mid-ventral line. These muscles
are the homologues of Owen’s infracarinales and of McMurrich’s fifth portion of the
lateral longitudinal musculature. The muscle mass extends from the basibranchiostegal
plate to the base of the caudal fin. It is sharply separated from the surrounding
muscles for all of its extent except the anterior portion for about one-fourth of its extent.

The infracarinales in the king salmon are divided into three portions, by the inter-
position of the pelvic arch and of the anal fin. ‘These portions can be described under
the names of the protractor ischii, the retractor ischii (protractor analis), and the retractor

analis.
PROTRACTOR ISCHII.

This term has been given by Owen to the anterior portion of the infracarinalis.
In the king salmon this muscle division extends from the anterior margin of the pelvic
arch to the posterior margin of the basibranchiostegal plate, the paired muscles lying
on either side the mid-ventral plane. For the greater portion of its length the pro-
tractor ischii is inclosed in a cylindrical connective tissue sheath which contains a rela-
tively large amount of adipose tissue. In the mid-line between the two muscles the
adjacent portions of this connective tissue sheath form a pretty definite ventral median
septum. In the anterior third of the muscle this sheath is less definite and in most speci-
mens scarcely continuous for the full length. In this part of the muscle the form of
the muscle as a whole ceases to be cylindrical. The myomeres are not definitely separated
from those of the lateral muscle, and the septa are more or less continuous with those of
the neighboring lateral muscle. This portion of the protractor is spread out into a
slight spatula-shaped terminal mass in the region ventral and anterior to the pectoral
fin. The protractor ischii is composed of myomeres, relatively simple in arrangement
in the anterior third, and becoming more and more complexly folded into a sort of spiral
toward the posterior end of the muscle. In an 80 cm. salmon the diameter of the most
cylindrical portion of the muscle varies from 8 to 10 mm., i. e., just in front of the sym-
physis of the ischii.

The protractor ends in a conical tip which is inserted into the fascias of the skin
and of the ventral fin muscles, the median septum, and the antero-ventral border of
the ischium itself. The tendons of insertion are formed by the ends of the whorls of
myocommatous connective tissue. These are best exposed by a median incision through
the skin ventral to the protractor ischii itself.

Contractions of this muscle accomplish two functions. If the axis of the body
is rigidly fixed by the action of other muscles then this muscle merely pulls the pelvic
girdle forward. It is from this action that it receives its name. However, it seems
that a more important function is found in a second action, namely, a strong ventral
flexion of the body. Then, too, in the spawning act, if one is to judge by external
appearances, the protractor ischii contributes sharply to the pressure that is brought
upon the abdominal cavity and which produces the extrusion of the eggs.

RETRACTOR ISCHIL (PROTRACTOR ANALIS).

The retractor ischii consists of a cylindrical muscular slip which extends from
the posterior end of the pelvic arch directly caudalward and around the anal aperture
to be inserted with its fellow into a special bony triangle at the base of the anal fin.

SKELETAL MUSCULATURE OF THE KING SALMON. 35

The relation of the pair of muscles and their insertion into this triangle is shown in
figures 2 and 8. This bony triangle in its normal position in the body rests directly
under and indirectly supports the most anterior rays of the anal fin, with which it is
strongly connected by connective tissue fascias. It is a modified interhemal.

The contractions of this division of the infracarinalis contribute to the ventral
flexion of the body. It does this by fixing both the anal fins and the pelvic fins. When
other muscles are relaxed so that these fins are movable the action of the muscle is to
produce retraction of the ventral fins, i. e., the pelvic girdle. If this latter arch is
fixed then protraction of the anal fin results, a movement by which the muscle may
well receive the alternate designation of protractor analis.

RETRACTOR ANALIS.

On either side of the mid-ventral line of the caudal peduncle lies a slender cylindri-
cal muscle, the retractor analis. The muscle is oval in cross section, about 4 mm.
broad by 2.5 mm. thick in an 80cm. salmon. The fibers run up under the tendinous
end of the most posterior erector muscle of the anal fin, to be attached by a broad
tendon into the posterior margin of the modified cartilage which supports the most
posterior rays as previously described and indicated in figure 14. When the skin is
removed and all the muscles are
in place this muscle has the ap-
pearance of running into the
angle formed by the lateral mus-

cles and the posterior margin of

Fic. 2.—An antero-ventral view, somewhat diagrammatic, of the relation of the fin.
the anterior rays of the anal fin, the supporting triangular cartilage, and 3
the insertions of the pair of retractor ischii, i. e., protractor analis muscles. Posteriorly the muscle runs
The figure shows only indistinctly that the three] anal rays appear one under the bases of the ventral

behind the other, the most anterior of course the shorter. neta rays ofthe caudal fntte
be attached into the connective tissue and fascias and the ends of the hemal spines.
The muscle is slightly conical in shape, becoming more slender posteriorly. It is only
3 or 4 mm. in diameter at its thickest part.
The function of the muscle is that of retraction of the anal fin, but the muscle is so
slightly developed that it can not produce extensive motion.

MUSCLES OF THE HEAD REGION.

The muscles of the head region may be grouped and discussed under the following
heads:

A. Superficial dorsal head muscles.
1. Adductor mandibule, (a) cephalic portion; (5) mandibular portion,
2. Levator arcus palatini.
3- Dilatator operculi.
4. Levator operculi.
B. Deep dorsal head muscles.
5. Adductor operculi.
6. Adductor arcus palatini.

36 BULLETIN OF THE BUREAU OF FISHERIES.

C. Dorsal branchial arch muscles.
7. Levatores arcuum branchialium.
8. Interarcualis dorsalis obliquus, posterior.
g. Interarcualis dorsalis obliquus, anterior.
to. Adductor arcuum branchialium, anterior.
tr. Adductor arcuum branchialium, posterior.
12. Transversi dorsalis, anterior.
13. Transversi dorsalis, posterior.
D. Ventral branchial arch muscles.
14. Interarcuales ventrales obliqui.
15. Transversi ventralis, anterior.
16. Transversi ventralis, posterior.
17. Pharyngo-clavicularis externus.
18. Pharyngo-clavicularis internus.
E. Mandibular and hyoid arch muscles.
(xb. Adductor mandibule, mandibular portion.)
1g. Intermandibularis.
20. Geniohyoideus.
21. Hyohyoideus.
22. Sternohyoideus.

SUPERFICIAL DORSAL HEAD MUSCLES.
ADDUCTOR MANDIBULA! (THE MASSETER OF AGASSIZ, OR RETRACTOR ORIS OF OWEN).

This is the largest muscle in the head. It forms the fleshy mass just posterior to
the eye which for its delicacy of flavor the fishermen prize under the name “salmon
cheeks.”’

The adductor mandibule is in two divisions that are almost though not quite dis-
tinct and separate. There is a cephalic division above the angle of the jaw, and a
mandibular portion lying chiefly below and along the inner border of the dentary.

The cephalic division of the adductor is in old specimens often more or less indis-
tinctly separated into three divisions, as described by Allis* for Amia. These divisions
are, however, not bounded by more than the thinnest of endomysial membranes and are
intimately fused toward the tendon of insertion. In fact they are of lesser importance
and scarcely justify the dignity of special designation. The cephalic division will
therefore be described as a whole.

(a) The cephalic division of the adductor mandibule has an extensive surface of origin
which includes (1) the anterior border of the preopercle for most of its extent, (2) the
entire surface of the quadrate, (3) the metapterygoid, and (4) the hyomandibular.
Some fibers arise (5) from the connective tissue sheath covering the levator arcus
palatini and from the post-orbital septum. The muscle fibers converge in a sweeping
curve or general fan-shaped whorl in the dorso-ventral direction to their attachment
in the broad tendon at the angle of the jaw. The extreme posterior fibers run anteriorly
and somewhat downward toward the ventral margin of attachment. This division of
the muscle is attached by a short, heavy, rather broad tendon into the outer margin
of the posterior part of the articulare. The tendon is intimately fused with the con-
nective tissue that binds the articulare with the premaxillary and the quadrate bones.

@ Allis, Edward Phelps, loc. cit.

SKELETAL MUSCULATURE OF THE KING SALMON. 37

The cephalic portion of the muscle in the medium sized fish is about 4 cm. broad in the
anterior posterior extent and about 5.5 cm. in the dorso-ventral dimension. The thick-
ness is from 1 to 1.5 cm.

(b) Mandibular portion of the adductor.—Besides the cephalic portion of the adductor
there is a stout mandibular portion. It arises from the anterior border of the tendon
over the quadrate bone and the angle of the mouth. It extends anteriorly to an attach-
ment along the inner surface of the middle third of the dentary, i. e., from a point directly
below the angle of the mouth forward to a point on the jaw. At the origin of this
portion the fibers are continuous with the fibers of the cephalic portion. From the
origin the fibers diverge slightly as they are distributed to their attachments on the
dentary. ‘The lower margin of the muscle takes a continuous attachment along the
under and inner surface of the
bone. The upper and outer
side of the muscle remains free
from attachments.

The contraction of the
adductor closes the mouth
with great power. In addi-
tion to its function in feeding
it undoubtedly takes part in
the motions of respiration.

LEVATOR ARCUS PALATINI.

This is a short, thick,
comparatively wide muscle
which takes its origin from the
angle formed in the external

A * Fic. 3.—Superficial head muscles after removal of the skin and a part of the jaws.
surface of the sphenotic, filling A. C., adductor mandibulz, cephalic portion: A. m., adductor mandibule,
the space just posterior to the mandibular portion; L. P., levator arcus palatini; D. op., dilator operculi;
L. op., levator operculi; P. d., protractor dorsalis; Lat., lateral muscle,

eyeball. The fibers run ob-

liquely downward and backward to a broad insertion on the anterior surface of the
superior half of the hyomandibular and also into the superior margin of the metaptery-
goid. The muscle at its posterior dorsal margin is intimately associated with and often
inseparable from the fibers at the origin of the dilatator operculi muscle.

DILATATOR OPERCULI.

This muscle has its origin from the anterior margin of the external surface of the
pterotic and the posterior border of the sphenotic. Its fibers converge sharply backward
and downward to an insertion by a small but strong tendon into the upper margin of
the opercle. The attachment is on the knob formed at the junction of the opercle with
the hyomandibular. The muscle lies in the groove between the exposed margin of the
pterotic and the hyomandibular.

38 BULLETIN OF THE BUREAU OF FISHERIES.
LEVATOR OPERCULI.

The levator operculi is a short, triangular muscle which arises from the posterior
spinous border of the pterotic. ‘The fibers converge diagonally downward and back-
ward to an insertion in the upper margin of the opercle. Its contraction leads to an
elevation of the opercle aiding in the act of respiration.

DEEP DORSAL HEAD MUSCLES.

When the eye is removed along with the upper portion of the metapterygoid and
hyomandibular bones a broad curved sheet of muscle consisting of short thick bundles
isexposed. The homologous mass in A mua has been divided by Allis into three portions—
the levator maxillze superioris, the adductor hyomandibularis, and the adductor operculi.
In Oncorhynchus this region can scarcely be divided except for the small group of fibers
at the posterior limit of the region. The two parts are better identified under McMurrich’s
names, the adductor arcus palatini, and the adductor operculi.

ADDUCTOR OPERCULI.

The adductor operculi arises on the ventral surface of the pterotic directly under
the origin of the levator operculi. Its origin is overlapped by the posterior fibers of the
adductor arcus palatini. ‘The fibers form a short but thick bundle, its length being from
8to1omm. It is inserted into the inner surface of the opercle a little above the inser-
tion of the dilator operculi which it opposes in action.

ADDUCTOR ARCUS PALATINI.

This muscle has an extensive origin along a line from the origin of the adductor
operculi to a point in the ventral portion of the eye socket. The fibers are short, thick,
and massive for the position in which they lie, but are not readily separated into distinct
bundles.

The posterior half of this muscle arises just ventral to the articulation of the hyo-
mandibular and from the ventral surface of the pterotic and sphenotic bones. ‘The
fibers are only a few millimeters in length, run directly outward, and are attached into
the inner surfaces of the upper half of the hyomandibular and the posterior portion of
the metapterygoid. The anterior portion of the mass, which is relatively the larger,
arises from the outer surface of the orbitosphenoid. Its fibers extend outward and
downward to a broad attachment on the inner surface of the metapterygoid and the
dorsal surface of the mesopterygoid.

The contractions of the entire mass tend to elevate the angle of the jaw and to
compress the palatine arch.

Allis @ divides this muscle into the levator maxille superioris and the adductor
hyomandibularis. No natural division along these lines can be observed in Oncorhynchus

tschawytscha.
DORSAL BRANCHIAL ARCH MUSCLES

When the palatine arch is removed and the adductor arcus palatini reflected, one can,
by trimming away the gill filaments, readily expose the group of muscles of the dorsal
half of the branchial arches.

@ Allis, Edward Phelps, loc. cit.

SKELETAL MUSCULATURE OF THE KING SALMON. 39

LEVATORES ARCUUM BRANCHIALIUM.

This group consists of five diverging muscle slips which are subdivisions of one
thin broad sheet. The origin of the muscle sheet is on a line immediately ventral
to the origin of the middle portion of the adductor arcus palatini. At the origin the
sheet is continuous and its divisions spread out to their attachments somewhat like a
miniature fan.

Fijth division.—This, the most posterior slip of the group, is a very slender muscle
arising from a point on the skull just in front of the foramen of the tenth nerve. The
fibers run posteriorly downward and backward to an attachment on the dorsal margin
of the flange of the fourth epibranchial. This muscle is about 2 to 2.5 cm. long, the
longest of the group.

Fourth division.—A similar though slightly stouter muscle arises just in front of
the latter and is attached on the crest of the corresponding flange of the third epibran-
chial.

Third division—The next differentiated strip runs under the tendon of the fourth
division and the flange of the third arch and is attached to the dorsal end of the carti-
laginous rod corresponding to a fifth epibranchial.

Second division.—The second division is attached on the flange of the second epi-
branchial. But a tiny slip of this muscle also runs to the dorsal surface of the pharyngo-
branchial of the third arch.

First division.—The most anterior slip is attached superficially to the flange on
the first epibranchial. But its deeper fibers run to the pharyngobranchial of the second
arch. These fibers are only a few millimeters long.

The points of attachment of the first, second, and third muscle slips are also com-
mon points of union for the connective tissue and septa covering the corresponding gill
clefts, i. e., the first muscle is opposite the angle of the second gill cleft, the second oppo-
site the third, and the fourth opposite the fourth.

The pharyngobranchial attachments of the first and second divisions are apparently
the homologues of McMurrich’s levatores interni, while the five divisions described here
would be his externi.

INTERARCUALIS DORSALIS OBLIQUUS, POSTERIOR.

There are two obliqui interarcuales on the dorsal half of the branchial arch. The
most posterior dorsal oblique arises from the posterior dorsal surface of the third pharyngo-
branchial. The fibers run obliquely backward to an insertion on the anterior surface
of the flange of the fourth epibranchial.

INTERARCUALIS DORSALIS OBLIQUUS, ANTERIOR.

The second or anterior oblique arises from the second pharyngobranchial near
its union with the epibranchial. The fibers run obliquely outward to an insertion on
the anterior margin of the flange of the third epibranchial.

ADDUCTOR ARCUUM BRANCHIALIUM, ANTERIOR.

There are two dorsal adductor muscles, an anterior and a posterior. The anterior
adductor arises on the posterior surface of the fourth epibranchial plate and is inserted
into the dorsal surface of the distal end of the corresponding ceratobranchial. Its
contraction approximates the cerato and epibranchials of the fourth arch.

40 BULLETIN OF THE BUREAU OF FISHERIES.
ADDUCTOR ARCUUM BRANCHIALIUM, POSTERIOR.

The posterior adductor muscle arises on the internal surface, i. e., median, of the
bony plate of the fourth epibranchial just within the origin of the preceding. It is
inserted into the external surface of the cartilaginous cap of the fifth ceratobranchial.

TRANSVERSI DORSALIS, ANTERIOR.

This thin and slightly developed muscle arises from the postero-dorsal surface of the
second pharyngobranchial near its junction with the corresponding epibranchial. It
runs to a similar attachment on the other side. It is one of the few unpaired muscles.

TRANSVERSI DORSALIS, POSTERIOR.

This unpaired muscle is much more strongly developed than the preceding. It
arises from the dorsal surfaces of a part of the fourth pharyngobranchial and the dorsal
margin of the central end of the fourth epibranchial. The fibers run to similar attach-
ments on the other side of the body.

Some fibers arise on the dorsal surface of the fifth ceratobranchial and become con-
tinuous with the constrictors of the pharynx.

VENTRAL BRANCHIAL ARCH MUSCLES.

The ventral muscles of the branchial arch consist of three groups, the interarcuales
ventrales, the transversi ventrales, and the pharyngo-claviculares.

INTERARCUALES VENTRALES OBLIQUI (VETTER).

A group of more or less distinct muscles corresponding to the interarcuales dorsales
is present on the ventral side of the branchial basket. In the salmon the anterior three
of these muscles are distinct and separate and not divisions of one sheet as in the dorsal
group. The posterior two are intimately united. Their dissection should follow that
of group E. They are exposed better beginning with the anterior one of the group.

First division—The most anterior or first division belongs to the first arch. It is
a comparatively small slip which has its origin from the ventral surface of the first
basibranchial. It extends along the under surface of the hypobranchial to an attach-
ment into the cartilage and ventral tip of the ceratobranchial near its union with the
hypobranchial.

Second division.—The second division arises from the ventral surface of the second
basibranchial. It runs its course over the second hypobranchial and is attached by a
short strong tendon into the ventral surface of the second ceratobranchial. The first
and second arch muscles are completely separated at their area of origin. A tendinous
band runs over the ventral surface of the basibranchials between the two slips.

Third division—The third division or muscle of the third arch arises from the
third basibranchial and the median portion of the ventral surface of the hypobranchial.
Its attachment on the third arch corresponds to that of the first and second divisions.

Fourth division —This division is continuous with the fifth. They arise from the
third basibranchial on the ventral surface somewhat median to and in close contact

SKELETAL MUSCULATURE OF THE KING SALMON. 41

with the third. The insertion of the fourth is into the extreme ventral portion of the
cartilage of the fourth ceratobranchial.

Fijth division—The fifth division is regarded as a subdivision of the preceding
muscle. It has its origin in a tendinous raphé which is strongly developed at a point
ventral to the insertion of the preceding. Some fibers also arise from the cartilaginous
plate posterior to the insertion of the fourth division. The muscle is relatively short
and thick. It is attached by a short, stout tendon to the fifth ceratobranchial, its
tendon being fused with the anterior border of the tendon of the pharyngo-clavicularis
externus.

TRANSVERSI VENTRALIS, ANTERIOR.

A short thick triangular bundle of fibers arises on the median surface of the ventral
end of the ceratobranchial of the fourth arch. It is an unpaired muscle and runs directly
across to an attachment at the corresponding point on the opposite side.

TRANSVERSI VENTRALIS, POSTERIOR.

This stout unpaired muscle is very much like the preceding, but three times larger.
It runs from the inner surface of the base of the fifth ceratobranchial transversely under
the esophagus to a corresponding insertion on the opposite ceratobranchial.

The transversi ventrales by their contractions approximate the ventral portions
of the fourth and fifth arches of the branchial basket.

PHARYNGO-CLAVICULARIS EXTERNUS.

This is a short broad muscle band extending from the antero-dorsal surface of the
clavicle directly dorsalward to the lower surface of the fifth ceratobranchial. Its length
is only about three times its breadth. Its contractions depress the branchial arch.

PHARYNGO-CLAVICULARIS INTERNUS.

This is a broad thin muscle band arising from the anterior surface of the inner
margin of the clavicle at about the middle of its arch. Its fibers run diagonally forward
and inward to an insertion on the ventral margin of the fifth ceratobranchial just under
the insertion of the pharyngo-clavicularis externus. There is a strong tendinous line
along the upper margin of the muscle.

The internus muscle retracts the branchial basket, i. e., draws it backward toward
the esophagus.

MANDIBULAR AND HYOID ARCH MUSCLES,

INTERMANDIBULARIS.

A short thick unpaired muscle extends transversely from the left dentary to the
right. In cross section it isa rough oval17 by6mm. The muscle is 2 cm. long. It is
attached to the inner surfaces of the two dentaries just back of the symphysis. It serves
to approximate the mandibles.

GENIOHYOIDEUS.

This is a broad flat sheet of muscle arising from the ceratohyal. The origin is
along a diagonal line extending from the postero-ventral border to the antero-dorsal
margin of the bone. The muscle joins with its fellow to form a practically continuous

42 BULLETIN OF THE BUREAU OF FISHERIES.

sheet at the insertion into the inner surface of the anterior portion of the dentary around
the symphysis. At its insertion the tendon is divided into an external and an internal
portion, one passing above, the other below the intermandibularis to its insertion.

HYOHYOIDEUS.

This long thin sheet of muscle arises from the ventral surface of the hypohyal and
passes diagonally outward and backward to insertions over the branchiostegal rays.
The muscle has attachments to the internal margin of each successive ray. It also has
insertions along the ventral margins of the ceratohyal and epihyal. The left hyoideus
somewhat overlaps the right at its origin.

STERNOHYOIDEUS.

The name sternohyoideus is applied to a broad and thick sheet of muscle arising
on the dorsal surface of the anterior end of the clavicle directly in front of the attachment
of the pharyngo-clavicularis externus. Its fibers run forward and are attached to the
ventro-lateral surface of the hypobranchial plate. Its differentiation from the ventral
portion of the great lateral muscle is apparent and probably it would be better to group
it with the longitudinal muscles.

MUSCLES OF THE CAUDAL FIN.

The modifications of the musculature which have come about for the control of the
movements of the caudal fin are associated with striking modifications of the caudal
skeletal structure. In order to present more accurately the form and relations of the
muscles it seems desirable to give the facts concerning the caudal skeletal complex.

CAUDAL SKELETAL COMPLEX.

The caudal fin in the king salmon is regularly bilobed and symmetrical. Externally
it appears of the regular homocercal type. The caudal skeleton, however, still shows the
heterocercal structure as presented by figure 5. The skeleton reveals the fact that the
epichordal component is limited to the rudimentary rays and at most the first two rays
of the dorsal lobe. The remainder of the dorsal lobe and all of the ventral represents the
hypochordal component. This modification rests on a rather complex caudal skeletal
base, as was shown by KoOlliker * for Salmo salar.

The axial region may be considered as composed of those vertebre entering into the
caudal peduncle, and those of the caudal fin proper. Of the three obvious vertebra that
enter into the caudal fin skeleton one only has a well developed centrum. The second
and third centra are very much reduced in size, the latter being only a tiny bony nodule.
The modifications of the vertebrae of the caudal peduncle begin sharply with the last three
vertebra of the group. However, the spines of the fifth and fourth, counting from the
tail, have a median flange on the anterior margin of the neural spines. In the last three
vertebre these flanges are fused each with the spine in front of it. The neural spines
of the first and second caudal vertebre enter into this fusion, the five spines making a
firm mass.

a Kolliker, Albert von: Ueber das Ende der Wirbelsiule der Ganoiden und einiger Teleostier, taf. rv, fig. 1 and 2.
Leipzig, 1860.

SKELETAL MUSCULATURE OF THE KING SALMON. 43

Lying on the dorsal surface of the three centra of the caudal group, and extending
out over the bases of the neural spines is an irregularly fan-shaped bony plate, the Deck-
knochen der Chorda of von Kolliker.¢ This plate is coalesced into the dorsal surface
of the second, and usually the third, centrum. It has a caudally projecting spine extend-
ing in the direction of the axis of the third centrum.

The hemal spines of the last three vertebrae of the peduncle are also sharply modified

Fic. 4.—Caudal skeleton. Five centra of the caudal peduncle with their modified spinesare shown. ‘The three caudal centra
are much reduced, the last quite rudimentary. ‘The hemal spine of the basal caudal vertebra is very stout. It bearsa
transverse spine near its base. ‘The lower one of the five hypurals is marked “‘h."’ The large irregular plate “‘d’’ is von
Kolliker’s ‘‘Deckknochen der Chorda”’ of Salmo salar. A few of the deeper fibers of the musculi interfilamenti (int.)are
shown,

by being greatly thickened and broader. The borders of these plate-like hemal arches
are not fused, though they are intimately bound together by connective tissue.

The hemal spine of the most anterior vertebra of the three that belong to the
caudal fin proper is very strong and bar-like. It is heavier than the caudal peduncle
spines anterior to it, and is especially characterized by its strong and stocky base which

‘ earries a well-developed lateral process. This process stands out sharply for 3 to 5 mm.

a Kolliker, Albert von, op. cit., p. 12.

44 BULLETIN OF THE BUREAU OF FISHERIES.

from the base of the spine. It is directed somewhat posteriorly and serves for the
attachment of a group of the deeper caudal muscles. The last two caudal fin vertebrae
are sharply modified. Ventral to the rudimentary centra there is a series of strong
and broad hypurals. In the king salmon there are five hypurals, the most anterior one
the strongest, and the individuals of the series diminishing in size toward the dorsal
lobe of the fin base. The development of the hypurals is commensurate with that of
the caudal musculature.

Saddled over the ends of the hemal spines of the last two vertebre of the caudal
peduncle, the spine of the first caudal vertebra, the hypurals, and the bony fusion of
neural spines previously described, are the series of paired fin rays constituting the
caudal fin. ‘The fully developed rays are 19 in number, with about 12 rudimentary
rays above and as many below. The middle ray of the fully developed series represents
the axial ray. It is not only in the middle of the series but the interfilamenti caudales
muscles are inserted symmetrically with reference to this axial ray (fig. 4). These
rays form a joint of limited movement over the end of the skeletal complex to which
they are strongly anchored in a firm mass of ensheathing connective tissue.

CAUDAL FIN MUSCLES.

The muscles of the caudal fin are derived from the posterior myotomes of the
embryonic lateral muscles. The only probable exception is the interspinous muscle,
which is very intimately associated with the dermal fin rays and the skin itself. The
muscles are superficial and deep.

SUPERFICIAL MUSCLES.
CAUDAL END OF THE MUSCULUS LATERALIS SUPERFICIALIS.

This is a trunk muscle, but the details of its caudal insertion have been reserved for
description at this point. The lateralis superficialis or dark muscle is continued over
the lateral surface of the caudal peduncle to be inserted into the base of the tail. It
forms a sheath on each side of the mid-line of the caudal peduncle estimated in width
at two-thirds the distance from the mid-line of the peduncle to the dorsal and ventral
borders respectively. The muscle substance ceases posteriorly in the middle line at a
point directly under, i. e., ventral to the base of the first long dorsal caudal ray. The
caudal end of the muscle, i. e., marking the termination of the myomeres in its
tendon and fascias, is distinctively clavate. The dorsal myomeres are narrowed, and
the myocommata run together into a strong tendon that is attached to the bases of the
first, second, and usually the third long dorsal rays just exterior to and in the fascia
of the profundus lateralis. In a Baird specimen (small male) the last three dorsal myo-
meres are modified, the last two into a muscular slip running obliquely dorsalward and
caudalward to end in a delicate flat tendon or fascia. The dorsal lobe of the superficialis
is rendered more prominent by the fact that the dorsal border of the muscle, just at the
base of the caudal peduncle, is attached to fascias which are intimately connected with
the myocommata between the ventral two-thirds and the dorsal third of the epaxial
half of the lateral muscle. There is considerable irregularity in the arrangement of
the muscular fibers of the last two or three myomeres of the dorsal lobes of this
muscle. A rather common irregularity is that shown in figure 5. The ventral lobe of

SKELETAL MUSCULATURE OF THE KING SALMON. 45

the lateralis superficialis forms a similar attachment into the connective tissue over
the bases of the first and second long ventral caudal rays.

THE TERMINAL OR CAUDAL PORTION OF THE MUSCULUS LATERALIS PROFUNDUS.

The terminal or caudal portion of the profundus is characterized by the excessive
proportion of connective tissue of the myocommata. In fact the myocommata are
finally reduced to tendons of insertion.

The epaxial and hypaxial portions are well separated in the region of the caudal
peduncle, partly by the greater development of the superficialis which ensheathes the

Fic. 5.—The superficial muscles of the caudal fin and the caudal peduncle. L. s., lateralis superficialis; D. s., dorsal slip of
lateralis superficialis; V. s., ventral slip of lateralis superficialis; D. ¢., dorsal tendon of lateralis superficialis; V. ¢., ventral ten-
don of lateralis superficialis, L. »., lateralis profundus; ¢., terminal tendons of the lateralis profundus; Jnt. c., interfilamenti
caudalis.

profundus next the septum. The final terminal tendons run straight back under the
clavate margin of the tendon of the superficialis to a very strong insertion into the
aponurosis which covers the bases of the rays of both the dorsal and the ventral lobes
and includes all of the intermediate series (fig. 5).

19371°—vol 33—15——4

46 BULLETIN OF THE BUREAU OF FISHERIES.

The epaxial division covers the deeper muscles presently to be described, but is not
strongly fused with their fascias. The hypaxial division is strongly attached into the
tips of the hemal spines of the last vertebrae of the caudal peduncle as well as into the
bases of the rays. Its superior margin covers the ventral inferior caudal flexor.

The epaxial and hypaxial aponuroses are strongly united across the median line as
shown in the area between the terminal myomeres of the superficialis and the inter-
filamenti muscles.

MUSCULI INTERFILAMENTI CAUDALIS, THE INTRINSIC MUSCLES OF MCMURRICH.

This small superficial double fan-shaped muscle consists of dorsal and ventral
portions. ‘The fibers originate in the fascia lying over, or covering, the termination of
the lateralis profundus muscle on the base of the tail. They are attached into the
admesial margins and the external surfaces of the bases of the seven caudal rays lying
on either side the median ray. This middle ray, so far as the muscular arrangements
indicate, is axial. The muscle fibers attached to its base above run diagonally upward
and caudalward, those below downward and caudalward. The superficial fibers form
a continuous layer, while the deep fibers run from ray to ray, as shown in figures 4, 5,
and 6. When this muscle contracts it draws the caudal rays together, narrowing the
spread of the fin.

The width of the caudal interfilamenti muscles, at the best, is about 10mm. The
dorsal lobe is about 26 mm., the ventral lobe 24 mm. long.

DEEP CAUDAL MUSCLES.

The caudal fin is used by the salmon both as a steering rudder and as a propeller.
The deep ventral muscles move the parts of the fin to set its form for a rudder, but the
musculature which utilizes it as a propeller is limited to the great Jateral muscles acting
on the finas a whole. If, during the movements of the fin as a whole, it is set in some
special position or given a characteristic shape, that shape will be utilized for steering
the forward motion of the salmon. This activity is accomplished by means of the deep
caudal muscles, as can readily be seen by consideration of the effect of the contractions
of the muscles singly or in groups. ‘There are six pairs of these deep muscles. They
vary considerably in detail of size and position but the usual type will now be described.

FLEXOR CAUDALIS VENTRALIS SUPERFICIALIS.

This is a delicate muscle slip which begins by a small flat tendon attached to the
bases of the hemal spines, the third and fourth from the end of the caudal peduncle.
A few fibers also arise from the fascia and tendons of the superficialis in the median line.
It runs posteriorly to a slender tendon attached to the tip of the transverse spine on the
first caudal vertebra. The muscle is continued from this point to an insertion on the
base of the third caudal ray ventral to the axial ray. Its attachments lie over the inter-
filamenti. The ventral margin of the posterior division of this muscle is sometimes
fused with the dorsal margin of the next.

SKELETAL MUSCULATURE OF THE KING SALMON. 47
FLEXOR CAUDALIS VENTRALIS SUPERIOR.

This caudal flexor is a rather broad group of fibers which arises from the ventral
surfaces of the centra and the bases of the hemal spines of the last two vertebrae of the
peduncle, also from the base of the spine and the ventral surface of the lateral process of
the most anterior caudal vertebra. The fibers run slightly ventralward as they proceed
to their insertion into the bases of the fifth to the eighth caudal rays below the axial ray.

The contractions of this muscle lead to a flexion of the lower half of the middle
portion of the caudal fin, and of the ventral caudal lobe. The tension in this case is
brought primarily on the uppermost rays of the ventral lobe. ‘The muscle presumably
acts in conjunction with the next to be described.

FLEXOR CAUDALIS VENTRALIS INFERIOR.

The origin of the inferior ventral flexor is from the surfaces of the last three hemal
spines of the caudal peduncle. The attachment is in a line which begins somewhat
ventral to the anterior limit of origin
of the preceding muscle and runs
posteriorly and toward the trans-
verse process of the first caudal ver-
tebra. The fibers of the border of
the superior muscle arise under the
ventral border of the preceding
muscle.

The fibers of the inferior flexor
run ventrally and caudally to inser-
tions into the bases of the last two
long ventral caudal rays and into the
adjacent series ofrudimentary rays.

Contractions of this muscle
sharply flex the extreme ventral
border of the ventral caudal lobe.
Contraction at the same time with
Fic. 6.—The deep caudal fin muscles. f. v. spf., flexor caudalis ventralis the superior ventral flexor would

superficialis; f. v. spr., flexor caudalis ventralis superioris; /. v. ty sharply flex the whole ventral half
flexor caudalis ventralis inferioris; a. v., adductor caudalis ventralis; of the caudal fin toward that side on

f. d. s., flexor caudalis dorsalis superior; f. d.i., flexor caudalis dorsalis ,
inferior; int., interfilamenti caudalis. which the contraction occurred.

ADDUCTOR CAUDALIS VENTRALIS, THE ADDUCTOR OF THE DORSAL CAUDAL, LOBE.

This muscle is a relatively thin and broad sheet of muscle fibers lying below but
in its body closely parallel with the caudal vertebral axis. The origin of the muscle is
revealed by cutting away the posterior portion of the superficialis, the major portion
of the interfilamenti caudalis, and the superior border of the superior ventral flexor.
It has a rather broad line of origin extending from the lateral spine of the first caudal
vertebra directly posterior to the second caudal ray below the axial ray. The tendon
of origin is rather thickened at the spine. The line of origin is along the dorsal margin

48 BULLETIN OF THE BUREAU OF FISHERIES.

and the surface of the lowermost hypural bone, and from the connective tissue ove1
the bases of the corresponding caudal spines. ‘The tendon is somewhat stronger at the
ends of this broad line of origin. ‘The fibers of the muscle run dorsally and somewhat
caudally, converging as they go and ending in a broad and strong tendinous attachment
into the fourth, fifth, sixth, and sometimes seventh, long caudal rays dorsal to the axial
ray.

The belly of this muscle is 16 mm. wide by 3 to 3.5 mm. thick and the muscular
portion is 22 mm. long, i. e., in the 80 cm. fish used as a standard.

The contractions of the adductor lead to sharp flexion and adduction of the dorsal
caudal lobe. Since the tension on the lobe is almost directly from the point of attach-
ment of the tendon toward the median axial ray, this would naturally lead to an
approximation of the rays and a decrease in the spread of the fin.

FLEXOR CAUDALIS DORSALIS SUPERIOR.

The dorsal flexor is the longest and strongest muscle of the deep caudal series. It
lies almost in the axial plane of the fish. It takes its origin from the median septum
over the fifth and fourth neural spines and from the third and fourth vertebra, counting
from the posterior end of the caudal peduncle. Some fibers of origin are found along
the tips of the second and third neural spines. Fascias of the muscle are more or less
intimately attached to the median septum as far back as the first true caudal vertebra.
The insertion, in conjunction with the attachment of the adductor caudalis ventralis,
is by a thin flat but strong tendon ending on the lateral surface of the bases of the rays
of the most superior portion of the hypaxial division of the caudal fin.

Occasionally the muscle is more strongly developed, in which case it has an origin
anterior to that described.

Contractions of the superior flexor produce strong flexion of the dorsal caudal lobe
toward the side on which the contractions occur. Undoubtedly this muscle and the
one preceding it exert the most powerful influence in the control of the rudder-like
function of the caudal fin.

FLEXOR CAUDALIS DORSALIS INFERIOR.

The inferior dorsal flexor is a much more slender muscle than the preceding one. Its
origin is directly ventral to and lies parallel with the superior flexor. ‘The fibers of
origin are from the connective tissue over the basal part of the bony plate formed by
the fusion of the neural spines of the last three vertebre of the caudal peduncle. Some
fibers are also attached into the myocommata at its most anterior margin. ‘The muscle
belly extends caudally in a line parallel with the general axis of the fish, running under
the adductor to a flat tendinous insertion into the bases of the median two or three
caudal rays next above the axial ray.

Contractions of this muscle produce simple flexion of the middle portion of the
caudal fin.

From the descriptions presented and the accompanying figures it is now more clear
that these muscles are the ones concerned in shaping the position and form of the caudal
fin during the active movements of forward swimming. The great lateral muscles
must be supposed to act on the caudal fin as a whole in the alternate propelling move-

SKELETAL MUSCULATURE OF THE KING SALMON. 49

ments. If, during this general propulsive motion, the form and shape of the caudal
fin is adjusted as it would be by graduated contractions of the deep caudal muscles, it
is obvious that the fin will be the guiding rudder controlling the exact direction of the
forward movement. In closing the discussion of this phase of our subject we may
reiterate once more the statement previously made, that these deep caudal muscles
control the positions of the caudal fin which will adapt it to the purposes of a rudder.
The great lateral muscles furnish the power which acts on the caudal fin as a whole,
furnishing a piscine propeller seldom equaled and never excelled in the aquatic world.

MUSCLES OF THE PECTORAL GIRDLE.

The pectoral muscles of the European salmon, Salmo salar, have been briefly
described by Harrison,* and more recently described and figured by Pychlau.? The
development is given by Harrison, and by Vogel® for the trout, Trutta fario. Many
instructive comparative points in the myology of fishes are to be had from the exhaustive
papers by Allis ? on Amaia calva and Scromber scromber.

In Oncorhynchus tschawytscha the pectoral fin has 14 rays, the basal or external
one being markedly heavier and the others successively more slender. The base of
each half ray is curved sharply toward the median border of the fin. The two halves of
each ray are widely separated at the base. The series of rays is seated like a saddle
across the skeletal ridge of the basalia, forming a very mobile joint, as described by
Pychlau for Salmo salar. This type of joint is also found in all the salmon fins, but
with modifications.

ABDUCTOR PECTORALIS SUPERFICIALIS.

This muscle arises from the anterior ventral border and the inner or median ventral
margin of the coracoid as far back as the base of the fin. Its surface of origin along the
coracoid is widest about one-third the distance from the anterior end of the coracoid,
where it covers a surface of about 9 mm. wide in a standard fish. The median line of
origin is along the ventral ridge on the coracoid, covering this ridge for one-third its
length. The fibers of the muscles run back over the deep abductor to a tendinous
insertion in the tips of the processes of the ventral half rays. The ventral surface of the
muscle near its origin has its tendon joined by the fibers of the protractor ischii. These
occasionally spread fan-shaped over the surface of the angle between the ventral ends
of the clavicles. The external fibers of the pectoralis superficialis are in close approxi-
mation to, and have tendons intimately fused with, the internal portion of the profundus.

The action of the superficialis is to bend the fin downward and forward and to
close the rays.

@ Harrison, Ross G.; Die Entwicklung der unpaaren und paarigen Flossen der Teleostier. Archiv fiir Mikroskopische
Anatomie, bd. 46, 1895, p. 500-578.

> Pychlau, Waldemar: Untersuchungen an den Brustflossen einiger Teleostier. Jenaische Zeitschrift, bd. 43, 1908, p. 692-728.

¢ Vogel, Richard: Die Entwickelung des Schultergiirtels und des Brustflossenskelettes der Forelle (Trutta fario). Jenaische
Zeitschrift, bd. 45, 1909, D. 499-544-

d Allis, Edward Phelps: The skull and the cranial and first spinal muscles and nerves in Scomber scomber. Journal of Mor-
phology, 1903, vol. 18, 1903, p. 45-328.

Same author: The cranial muscles and cranial and first spinal nerves in Amia calva. Journal of Morphology, vol. 12, 1897,
Pp. 487-808.

50 BULLETIN OF THE BUREAU OF FISHERIES.
ABDUCTOR PECTORALIS PROFUNDUS.

When the superficialis muscle fibers are reflected the abductor profundus is exposed.
It arises from the ventral surface of the coracoid. Beginning at a point one-third the
distance from the anterior end there is a thick muscular mass intimately attached into
the surface of a triangular area on the ventral face of the coracoid. The base of this
triangle is marked by a line parallel with the base of the pectoral fin over the union of
the coracoid with the basalia. The profundus has a short, heavy tendon divided into
slips corresponding with, and inserted into, the inner margins and tips of the curved
bases of the ventral half rays of the pectoral fin under the tendons of the superficialis.

The contractions of this muscle draw the fin downward, helping to balance or
support the body when quietly resting on the bottom.

EXTENSOR PECTORALIS.

There is a rather thick muscular bundle which arises under the anterior origin of
the abductor superficialis and along the margin of the ventral portion of the clavicle.

Fic. 7.—Ventral view of the pectoral fin muscles. A segment is cut out of the anterior end of the protractor ischii, pr. i., together
with the anterior ventral portion of the lateral muscle. This uncovers the abductor superficialis, ab. s., and its attachments
to the ventral half-rays of the fin. The end of the extensor, Ex., with its insertion into the base of the first fin ray is shown.

This muscle lies close within the angle formed in the ventral surface of the clavicle. It
is inserted by a short thick tendon into the external surface of the base of the first fin ray.
Contraction of this muscle spreads the fin out in the horizontal position. When
the fin is folded back against the body the external ray forms the upper margin of the
fin. From this position the extensor pectoralis throws the fin forward, bends it slightly
downward and spreads the rays. The muscle tends to support the abductor.

The great lateral muscles are attached by strong slips into the clavicle just dorsal
to the insertion of the pectoral fin. There is also a muscular slip from the great lateral
muscle running just ventral to the base of the fin and inserted into the fascia of the
dorsal wall of the pericardium. This fascia is closely attached to the internal or ventral
margin of the coracoid. Undoubtedly contractions of the great lateral muscle would
tend to draw the pectoral girdle posteriorly. When the fascia is dissected off a rather

SKELETAL MUSCULATURE OF THE KING SALMON. 51

thick triangular muscular mass of the adductors comes into view on the inner and the poste-
rior surface of the coracoid. —

ADDUCTOR PECTORALIS SUPERFICIALIS.

This is a short thick muscle lying in the angle between the coracoid and the clavicle
and the base of the pectoral fin. It takes its origin from the posterior ventral surface
of the coracoid, and the spine on the superior margin of this bone, also from the thin
bony plate lying between the superior margin of the coracoid and the clavicle. The
muscle fibers converge into a broad tendon which is inserted into the posterior surface
of all the dorsal half-fin rays except the first five. When this muscle is reflected a
deeper muscle is exposed.

ADDUCTOR PECTORALIS PROFUNDUS.

This muscle arises from the dorsal margin of the extreme ventral portion of the
clavicle and the surface included in the angle between this margin and the bony ridge
projecting on the side of the clavicle, also from the connective tissue septum joining
the clavicle and coracoid.and from the dorsal margin of the coracoid including the upper
surface of the median spine. The muscle is divided into two parts. Those fibers
arising in the angle between the anterior end of the coracoid and the clavicle form a
stout tendon inserted into the stout base of the marginal ray. The remaining fibers
converge to strong tendons inserted into the bases of the dorsal half rays of the pectoral
from the second to the last.

Contractions of the profundus support the contractions of the superficial muscle
in throwing the fin back against the side of the body.

INTERFILAMENTI PECTORALIS.

When the skin is removed from the basal half of the ventral surface of the pectoral
fin there is exposed a series of very delicate muscle fibers running across the bases of
the fin rays. These fibers run from ray to ray, being arranged diagonally so that when
they contract they tend to close the rays. No fibers were observed on the dorsal surface.

MUSCLES OF THE PELVIC GIRDLE.
ABDUCTOR VENTRALIS SUPERFICIALIS.

A slender slip of muscle, the abductor ventralis superficialis, arises from the median
longitudinal septum of the pelvis beginning at the ventral border of the anterior end
of the ischium, also from the adjacent cutaneous fascias. It is surrounded by a strong
aponurosis continuous anteriorly with that into which the tendon of the protractor
ischii is partially inserted.

The superficialis runs as a slender wedge of muscle to a strong tendinous inser-
tion into the tips of the ventral half rays of the ventral fin. A cross section of the
middle of the muscle presents a wedge-shaped surface, the base of the wedge in approxi-
mation to the skin, the surface shown in figure 8, and the side in contact with the median
(vertical) septum.

BULLETIN OF THE BUREAU OF FISHERIES.

on
to

Contractions of the abductor ventralis superficialis produce ventral flexions of the
ventral fin. It tends to bend the fin downward, i. e., away from the body. If the fin
rays are at the time spread then approximation of the rays also occurs.

ABDUCTOR VENTRALIS PROFUNDUS.

This is a large and strong muscle which lies external and dorsal to the superficialis,
It takes its origin from the entire ventral surface of the ischial plate, from the septum
which connects the external margin of this bone with the skin, and from the similar
septum that runs from the internal or median border to the mid-ventral line. This
last septum joins the median longitudinal pelvic septum just at the mid-ventral line
of the abdominal cavity; hence the peritoneum, the dorsal border of the median longi-
tudinal septum and the internal border of the ischial septum are fused.

The muscle fibers from this extended origin converge in the general caudal direction
toward the base of the anal fin. The insertion is by very short tendinous slips into the

Fic. 8.—Ventral view of the superficial muscles of the ventral fins and of the pelvic arch. ab. s., abductor ventralis superficialis;
ab. pr., abductor ventralis profundus; ad. pr., adductor ventralis profundus; Pr. #., protractor ischii; re. ., retractor ischii;
Lat., lateral muscle, ventral margin.

ventral half rays of the fin. The tendons run dorsal to the tendons of the superficialis
and are inserted into the inner border of the ventral half ray, i. e., the border next the
median plane of the fin itself. The fibers arising most anteriorly are inserted into the
rays of the external border of the fin. The fibers of the extreme posterior portion of
the muscle, arising in the deep angle in front of the ischial thickening, run almost dor-
sally to the tips of the bases of the rays.

The two halves of the individual rays are more widely separated in the ventral fins
than in the unpaired fins, and a distinct synovial joint is formed here as described for
the pectoral by Pychlau.¢ The manner of insertion of the abductor profundus and the
presence of this very efficient joint insures a strong abduction of the ventral fin on its
contraction. ‘There is only a minimal approximation of the fin rays, if any at all, accom-
plished by this muscle. This latter function seems limited to the superficialis.

@ Pychlau, loc. cit.

SKELETAL MUSCULATURE OF THE KING SALMON. 53
ADDUCTOR VENTRALIS SUPERFICIALIS.

This muscle represents the most dorsal portion of the pelvic musculature. The
muscle is separated into two divisions, as described by Harrison for Salmo salar.

The pars anterior arises from the dorsal surface of the anterior end of the ischium and
along the line of aponurosis in which the median longitudinal septum, the abdominal
peritoneum, and the ventral attachments of the great lateral muscles meet. ‘The pars
anterior division of the muscle runs as a flat band to end in a broad tendinous sheet
which covers the dorsal surface of the caudal end of the profundus just over the posterior
end of the ischium. It is inserted into the dorsal half rays of the external six or seven
rays. The insertion is into the curved bases of the rays at about the middle of the
curve.

The pars posterior consists of the short and relatively thick mass of muscle fibers
which arise from the fascias along the lower border of the lateral muscle immediately
dorsal to the posterior part of the ischium and the insertion of the fin. These fibers
are well separated from the pars anterior. The fibers run abruptly downward and
posteriorly to an insertion into the bases of the last three or four anal rays, i. e., the rays
on the median border of the fin. The tendinous insertion does not seem to be sharply
subdivided into slips and is extremely short.

ADDUCTOR VENTRALIS PROFUNDUS.

The profundus arises from the dorsal surface of the illium for its full extent, and from
the horizontal septum extending from the inner margin of the illium to the mid-ventral
line, also from a similar septum extending from the external margin to the skin. The
fibers constitute the largest muscle of the pelvic series. In the middle of the belly the
muscle is broad and rather thick (22 mm. broad by 6 mm. thick, in an 80 cm. fish).
Posteriorly the muscle is somewhat heavier on its median border, the fibers extending
transversely out and back over the posterior thickened margin of the illium, and under
the tips of the dorsal half rays. When the adductor superficialis is removed the ten-
don of the profundus is revealed as a broad and relatively long sheath. The insertions
of the tendon are on the inner, that is median, borders of all the dorsal half rays. This
tendon enters also into the formation of the capsule of the movable joint by which the
fin is attached to the posterior end of the illium.

Contractions of the adductor profundus lead to two motions, first, rotation of the
fin over the illium at this point, i. e., throwing the fin up against the body of the fish,
and second, the spreading of the rays, by throwing the outer fin margin in a lateral
direction with reference to the median plane of the fish. These last motions, it will be
seen, are directly the opposite of those produced by the abductor group.

The ventral margin of the lateral muscle is strongly attached into the supporting
connective tissue of the posterior part of the illium by means of the myocommata. It
is evident that contractions of the lower borders of the myomeres lying immediately
posterior to the pelvic girdle will have a tendency to draw the pelvic arch as a whole
backward. The muscular development does not seem to be of an extent which would
lead one to infer that this is a chief function of the muscle. It justifies only the infer-
ence that the movement is an incidental but possible one.

54 BULLETIN OF THE BUREAU OF FISHERIES.
MUSCLES OF THE DORSAL FIN.

Harrison @ has briefly described the muscles of the dorsal fin of Salmo salar in con-
nection with his study of the development of the fins of teleosts. The muscles in the
king salmon are similar in character and arrangement. ‘The number of dorsal fin rays
is greater in Oncorhynchus tschawytscha than in Salmo salar. The muscles of the fin
have a correspondingly greater number of divisions, one for each finray. A typical fin
ray is moved by three pairs of muscles, (1) an inclinator, (2) an erector, and (3) a depres-
sor. Beside, the fin as a whole is moved forward by the pair of protractors and back-
ward by the pair of retractors described with the group of longitudinal muscles. The
specific fin muscles may be described more fully as follows:

INCLINATOR DORSALIS (THE SUPERFICIAL LATERAL MUSCLE OF MCMURRICH).

This muscle in reality consists of a series of short muscles, i. e., independent slips,
corresponding in number with the dorsal fin rays. Lach tiny slip has its origin in a
fascia which is strongly attached
to the skin and which covers the
dorsal margin of the great lat-
eral muscle. The fibers of each
muscle slip converge as a conical
mass ending in a short tendon
inserted into the postero-lateral
margin of the base of each fin

sp

‘\y ray. These muscle slips are
1 about 20 mm. Jong in the ante-
HN rior members of the series and
My 15 mm. in the posterior. The

Was Ay WY :
BYU NN YIN NNUOD NSN NAVARA SANA UUIRANA NAN ATAXIA Cae extreme anterior three or four

slips are very rudimentary and
Fic. 9.—Superficial muscles of the dorsal fin after removal of theskin and subcu- F f
taneous fat, left side. Jnc. d., inclinator dorsalis, one slip for each fin ray; May readily be overlooked in the
p.d., protractor dorsalis; 7. d., retractor dorsalis; Lat., great lateral muscle, dissection. There are 1 3 free
epaxial portion. z =
slips, 16or17inall. In the sea

form the spaces between muscle slips are filled with subcutaneous fat. The ends of the
deep fin muscles are to be seen just between the ends or insertions of the inclinator slips.
Harrison was the first to describe these muscles carefully, and to him we owe the
name‘ ‘inclinator.”’
The contractions of the divisions of the inclinator muscles tend to bend or incline
the dorsal fin toward the side.

ERECTOR DORSALIS.

The erector muscle divisions lie between, hence alternate with, the depressor slips
of the double series. Each of these arises from the fascia between its interneural spine
and the one next in front of it, and from the posterior border of the latter. The mus-

—

@ Harrison, Ross G., loc. cit.

SKELETAL MUSCULATURE OF THE KING SALMON. 55

cular divisions of the erector dorsalis are separated from those of the depressor by very
thin connective tissue septa. But the whole group of muscle slips is encased in a much
thicker and tougher sheath. When the muscles are uncovered by removing the lateral
muscles the connective tissue sheath is more evident. This is seen to be intimately
attached along the line where the neural spines and interneurals are interlocked and
embedded in the median longitudinal septum. This sheath, the median septum, and
the partitions between muscle divisions serve to form a series of slender glove-finger-
like cavities enclosing the pairs of muscle slips on each side.

Each muscle division of the erector is attached by a very short tendon into the
anterior margin of the base of the dorsal halfray. The largest erector divisions are 40
mm. in length. At the posterior and shorter margin of the muscle they are about 32
mm. in length. The anterior two or three muscle slips are rudimentary, very slender,
and more or less fused.

The contractions of the erector muscle elevate the dorsal fin rays as the name
implies. The point of attach-
ment of the tendon of insertion
above or distal to the center of
movement of the joint favors
the erection.

DEPRESSOR DORSALIS.

The depressor muscle of the
dorsal fin is intimately associated
in position and attachments with
the erector dorsalis. The depres-
sor divisions are also segmental
in arrangement. They are very
slender slips of muscle which

Fic. 10.—Deep muscles of the dorsal fin after removal of the great lateral muscle;

arise each along the anterior bor- e. d., erector dorsalis; d. d., depressor dorsalis p. d., protractor dorsalis; r.d.,

der of the corresponding inter- retractor dorsalis. The skeleton is shown embedded in the median longitu-
dinal septum.

neural spine, and from the fascia

separating this muscle from the erector muscle in front of it. The fibers pass across the
end of the interneurals to insertions on the posterior border of the base of each dorsal ray.
The last muscle division of the posterior border of this series is very strongly developed.
It is somewhat broader than its mates and is attached into the bony plate previously de-
scribed for the retractor dorsalis muscle.

MUSCLES OF THE ANAL FIN.

The musculature of the anal fin is built on the same plan as that of the dorsal fin.
The modifications are slight and more or less unimportant. The fin has a protractor, the
retractor of the pelvis, and a retractor—both of which are divisions of the infracarinales
previously described as the most ventral longitudinal differentiation of the lateral muscu-
lature. The proper muscles of the fin are (1) the inclinator analis, (2) the erector analis,
(3) the depressor analis, and (4) the interfilamenti analis. These muscles are in divisions

56 BULLETIN OF THE BUREAU OF FISHERIES.

corresponding to the skeletal divisions of the fin itself. Their relations are shown by a
consideration of the anal fin skeletal complex.

SKELETON OF THE ANAL FIN.

The anal fin of the king salmon consists of 16 well-developed fin rays, with three
rudimentary rays at the anterior margin. There are no spines. A pair of typical rays,
say from the middle of the series, serves to show the general skeletal plan (fig. 11).

The ray itself consists of two half rays, the so-called dermal plates, intimately bound
to each other except at the base. Near the base the plates diverge slightly and curve
sharply posteriorly, not unlike the curve of a hockey stick. The ray is seated cross-
saddle fashion over a median cartilage to which it is strongly
bound by connective tissue ligaments but with a movable joint.

The median cartilage is bound by a movable joint to the
head of the supporting interhemal. The interhemal is firmly
imbedded in the median longitudinal septum. The position of
the interhemals alternates with the hemals but forms an acute
angle with the axis of the body, the inclinations being directed
caudally.

This skeleton is modified at two points. Anteriorly the
whole complex just described is represented by a triangular
plate, apparently the homologue of either the median cartilage
or more probably of the interhemal. This plate is strongly
bound to the anterior margin of the fin base. Its most dorsal
angles receive the tendons of the retractor ischii (fig. 2).

Posteriorly there is also a sharp skeletal modification. The
last two interhemal spines at the posterior end of the series are
fused at their ventral ends forming an irregular club-shaped knob
Fic. 11—Two segments of the Under the second from the last fin ray. This interhemal knob

skeleton of the anal fin. The js larger and stronger than the ones immediately in front of it.
lettering is on the cephalic bor- 5 :
Ger tetra Wanieunterenal Just dorsal to the last rays of the anal fin, and in series
spine; ¢.¢., cartilages; o.,ossicle with and bound by the enlarged interhemal, is a specially mod-
forming movable joints be , 6 9 :
tween the interhemals andthe ified cartilaginous plate. It is rather strong and laterally com-
ENE pressed. This plate receives the attachment of the retractor
analis. The plate is very strongly bound to the ones in front of it and to the inter-
spinous septum by bands of fibrous connective tissue. Doubtless this modified cartilage
is the homologue of one or more intermediate plates or of the interhemals.

INCLINATOR ANALIS.

The constituent serial divisions of this muscle are exposed by removing the skin
from along the base of the anal fin and the ventral surface of the adjacent part of the
body (fig. 13). There are muscle divisions for each ray including the rudimentary rays
at the anterior margin of the fin. The largest and longest divisions are opposite the
anterior full rays of the fin. The muscle slips become progressively smaller posteriorly.
‘The muscles of the rudimentary rays are small and imperfectly separated.

SKELETAL MUSCULATURE OF THE KING SALMON. 57

The longer muscle slips of the inclinator are about 12 to 15 mm. long and from 4 to 5
mm. wide. They are broad at their origin and the fibers converge to the point of insertion
into the rays.

The origin of the inclinator fasciculi is from the skin and the fascia covering the
ventral border of the great lateral muscle. This broad origin gives to each slip a base
which is seated on the cylindrical border of the lateral muscle. From this broad origin
the fibers converge toward a short slender tendon which is inserted into the base of the
corresponding ray on its lateral surface, and between the tendons of the erector and
depressor respectively. The insertion is in the plane of the axis
of the ray joint. The inclinator muscle slips fill the triangular
space between the skin, the lateral muscle border, and the
erector-depressor group of muscles (see fig. 13). The divisions
are strongly embedded in connective tissue sheaths as best
shown in formalin-preserved specimens.

Contractions of the inclinator muscle strongly bend the fin
toward the corresponding side. This motion is most pro-
nounced at the anterior margin of the fin where the muscle
slips are longer and larger. The pull of the muscle is at an
angle of about 70°, an angle that decreases with the flexion of
the fin in that direction.

ERECTOR-DEPRESSOR MUSCLE COMPLEX.

When the great lateral muscles are removed from the region
of the anal fin a muscular mass is exposed lying under the
superficial and deep lateral muscles and covering the interhema
spines of the anal fin. This mass consists of alternate slips of F!S- 12-—Section across the anal

| z fin in the plane of the interhe-
muscles constituting the erector and depressor muscles of the mat spines, the fin rays, and

aero fen respectively. the erector-depressor group of
wu " o muscles; int., interhemal

The whole group of muscle divisions is, like that of the spine: d.a., depressor analis;
dorsal fin, covered with a fibrous connective tissue sheath of — *¢» inclinator analis muscle;

a.7r., anal ray.

considerable thickness. This sheath is continuous with that
between the interhemal spines and is especially well developed in the longitudinal line
marking the border between the hemal spines and the interhemals.

ERECTOR ANALIS.

The erector muscle of the anal fin is composed of the larger of the alternate divisions
mentioned above as constituting the deep muscle complex. There is a muscle slip for
each fin ray.

Each erector slip arises from the posterior margin of the interhemal spine in front
of the one to which the ray is attached, and from the entire surface of the connective
tissue septum between the two interhemals in question. Each muscle division is spindle
shaped. It tapers at its ventral end into a short tendon, which runs to an attachment
in the anterior basal margin of the corresponding fin ray.

58 BULLETIN OF THE BUREAU OF FISHERIES.

The posterior slips of the erector analis are somewhat modified from the regular
arrangement. The last pair of erector and depressor muscle slips is greatly enlarged, or
rather the erector is greatly enlarged and the depressor moderately so. The fibers near
the tendon of insertion pass over the modified interhemal cartilage to which the retractor
ischii is attached, to insertions into the posterior border of the last anal fin ray. This
fin ray is itself very small. The tendinous end of the muscle slides over a groove formed
by the modified cartilages supporting the ray.

At the anterior margin of the series of erector divisions there is a muscular slip which
seems to belong to the series, judging by its origin, but the insertion of which passes into
the skin in front of the fin and near the base of the anal papilla.

The contractions of the erector muscles tend to elevate the anal fin and in con-
tinued contraction to hold it in the erect position. This is favorable in increasing the

Fic. 13.—Superficial muscles of the anal fin. inc., inclinator analis showing divisions for each ray; re. i., retractor ischii (pro-
tractor analis); r. a., retractor analis; Lat., great lateral muscles. The dotted line indicates the ventral limit of the lateralis
superficialis.

efficiency of the lateral movements which this fin contributes in balancing the fish in
the water.

The contractions of the larger posterior muscle slip described tend to draw the pos-
terior end of the anal fin sharply against the body and to some extent to antagonize the
retractor analis. ‘The function of the most anterior slip which has an insertion into the
skin in front of the fin would seem to be in connection with the movements of the anal
papilla.

DEPRESSOR ANALIS.

The depressor muscle consists of a series of slips which arise from the anterior surface
and lateral margin of the interhemal of the segment to which each belongs and from the
fascia separating it from the erector muscle attached to the same ray. The fibers of
each slip run as a slender ribbon over the shaft of its interhemal to an insertion into the

SKELETAL MUSCULATURE OF THE KING SALMON. 59

posterior border and tip of the corresponding fin ray. Each muscle slip is somewhat
thicker at its external border. j

The muscle slips are from 30 to 35 mm. long anteriorly and 20 mm. at the posterior
border of the series. They are about 4 mm. broad by 1 mm. thick in the middle of the
muscle belly. They are closely wedged in between the bellies of the erector divisions.
The anterior slips attached to the rudimentary rays are very small and slender.

The contraction of these muscles depresses the rays of the fin, tending to close it up
against the body.

Fic. 14.—Deep muscles of the anal fin. e. a., erector analis; d. a., depressor analis; inc., four reflected tendons of the inclinator
analis; re. 7., retractor ischii (protractor analis); r. a., retractor analis.

INTERFILAMENTI ANALIS.

Between the fin rays of the anal fin, especially at the base and more strongly devel-
oped anteriorly, are found delicate muscle fiber bundles. These muscles are attached
to adjacent fin rays, being attached nearer the base of the anterior ray, and more distally
to the posterior ray, carrying them in an oblique direction from ray to ray.

Undoubtedly these slips aid in the elevation of the fin rays. They are very delicate
and consist of only a very few individual fibers, a fact which easily leads to their being
overlooked.

The protractor analis and the retractor analis are longitudinal muscles of the anal
fin which have been described on page 34.

7 = id - toe

u [wy , i, ob a
PT ON eS aes ae
426 : vy
+ ae - an) a 4 ‘i
Pig vii “a he : ie Td) me a7 Y ley 7 qa’ vi, &
i; > n . = , .s . 7 ee — _
tora han Fy AD IEE. 0 is > ‘tee ee deal lin s
Bure ye ail ay te hil Maal’ ai yeh ats ange Arey 7 bees:
7 i Se on ee ~ a ye A il » a fi Sagat eer x ie
i. ; a i ny; Se J Pens At hi wae gee
ia ed VCP Er Me oe) re
n° : ; i me J a (Om ba 7
: i all " th oe
he’, rl Amt ca a ie
1 us ae ae a ‘il
oy “| pr:
A art! f a) i
aa ae Paha” ae
7 Sek, oe
‘ 5 eee ce £, a
ae pan < a ey
-_ P ~"2 on a a > Tae
nd ; a we a= mn ~ om i; A vet i)
a =a < - : ¥
! : = a rod «A Agen yy
_ ee we
a ei ry Tee 7
: a 4 La
ea e-" 4 a ig eo
- -* ot, eee Sm
7 y) i's
; Paar
a a 2
; a
tr
. : a -
cata: © jit Uvwegy ai) hte ma db, hata giv in oe ae Ae +7
7 -_ ey st SAP ot} f it, Le w/m aa ie Mae he ia : 7 i=
Oo Mu 7 ~~ = i
7 dat = ; Pree tts tieaueti fe aii hes ng ervey. 7
= ce Sn al
; Aaya ae SME Tes 1] mag Stas raya A Lisle 7 Vi ocak nil @ as yew fl
deste Lab alr ® STi DN nd” tbitest is ora ‘thal Wish rt mi 7 shy win hot a
; Betti anal aay i fig lagrinsi Bib af 20) Shh Tilley sip) ath a Jeni 21 Poli
_ ek UA iret iii Oe fi le cud Ohta rate wy ca “es
. ye ‘41 oi if i - if jie? ty ee ia ve adie atih. th ie, ken Hi <a
a uy rae ke LG, Tee fe Ade Enns = eying 1 a) ae
: ; a 7 le - vat aI
’ gl Poo S/n is Test Ping jai gi eo 7 estat T70 aie _
a syle NG Aw iin uy i
7. re § ;
7 on i ” 4
; ; = 7 =F

er AC e lis

BULLE Uh Os be ks Oks

** porjpoymmseyy *Apoq ay} JO JPY TEPNvS 34} UL SATOD IapUdTS OA} OFUL POF 1OLI9}UB 3} JO WOISTAIP
aq} OF anp sf woTELEA JULJIOdMT ysour ayy, *AproLaysod sasva19Ut sa1sULOAUE 9YY JO JU9}Xo [VUIPNFLDUOT ay} PUL Durplog JO ApXo[dttos ey TL, “eovJANs at{] UO
pappoquie SPS Saqy YOUyAd ‘aporAvlo ayy Aq pazdnisazur you sre Ady, ‘sauog oy0193d puv ‘[erodurs} ‘TeYdI990 |t{}_OF—T[NAS Py} JO aseq oY} OF poyoRyzw o1B
PPLULLUODOAUL OT[BYA29 UL, “IT Vela UL payworpur st Yuya ‘aur Suryvsedas sty} ayVoIpULopoIMSyY Sty} JO dUvoyrusis ay} WO1J JOVIJOp OF pattiaap Sear jT sutol
snpunjoid ay} pue stpepyiadns ay} ary [aso] ay} Jv ‘OULSYIUSIS JVo1S OW Jo IN ‘UOHBLILA VT} THOS SP ora, “Payeorpur aie vyeUTTOSOAUL ay} JO pur

DYISJLMDYIS] SMYIUAYLOIUG JO PIS JJ9[ PY} JO SINSY SI} UL UMOYS J1B S2IDWOAUL JY} JO ULIOF JY} UL SUOTPLLILA JY T,

saiomocut ay} Jo ssuryivut pelyiodns oy,

Bula Wise Busy LOLs: PLATE TT.

Three myomeres isolated from the anterior part of the body, 1. e., the plane just in front of the anterior border
of the dorsal fin. The form of the myomeres and of the septa of the region is revealed. The longitudinal
extent of the single myomere is 9cm., though the length of the longest muscle fibers is scant 7mm, The
skeleton and its dorsal extension in the median septum are well shown, ‘The protractor dorsalis and
its supporting, tissues lying between the anterior dorsal foldand the median septum have been removed.

Magnification ¢.

THE DIRECTIVE INFLUENCE OF THE SENSE OF SMELL
IN THE DOGFISH

&

By G. H. Parker
Professor of Zoology, Harvard University

19371°—vol 33—15——5 61

THE DIRECTIVE INFLUENCE OF THE SENSE OF SMELL IN THE
DOGFISH.

*

By G. H. Parker,
Professor of Zoology, Harvard University.

am

The object of the professional fisherman is to bring fish within his reach. In some
instances this is accomplished by direct attack, as in such primitive operations as har-
pooning and spearing, or by circumvention, as in most forms of seining. But in many
cases some deceptive device is set up whereby the fish is lured to its fate. These decep-
tions are practiced on the fish through some of its various sense organs. Noises are
made by the slapping of oars or other such implements on the water and thus fish are
driven into gill nets or entangled in trammels. Such influences probably affect the
fish through the ear, the sense organs of the lateral line system, or the skin, thus involving
hearing and touch. Lights are often carried on boats at night and small fish are thus
attracted in sufficient numbers to be easily caught by hand nets, a procedure dependent
upon the organs of sight. But fishing in the ordinary acceptation of the term implies
the use of some deceptive attraction in the form of bait, which, as originally used,
depends on the feeding habits of the fish and has to do with its senses of smell and of
taste.

The sense of taste has been almost universally attributed to fish, and most fisher-
men, naturalists, and even anatomists have assumed that fishes possessed a sense of
smell. Of recent years some comparative physiologists have denied this sense to
fishes. In their opinion the nasal organ of the fish acts more as an organ of taste
than an organ of smell. This conclusion was thrown in question by the observations
of Aronsohn (1884) and of Baglioni (1909), and was refuted by the experiments of
Parker (1910, 1911), Sheldon (1911), and Copeland (1912). By these studies it was
shown that fishes scent their food in essentially the same manner that air-inhabiting
vertebrates do and with the corresponding sense organ, the nasal organ.

The following investigation has been carried out with the view of ‘scertaining
the exact method by which a fish finds its food or may be caught by a bait. The work
has been done on the smooth dogfish of the Woods Hole region, Mustelus canis (Mitchell).
The gustatory, chemical, and olfactory senses of this fish have already been studied by
Sheldon (1909, rg11), and it was therefore an unusually favorable subject for the investi-
gation of those activities which depend upon the sense organs named.

The experiments were carried out for the most part in a small pen, 24 feet long by
6 feet wide, in one of the pools of the Woods Hole station. The pen was bounded partly

® 63

64 SENSE OF SMELL IN DOGFISH.

by the walls of the pool and partly by chicken wire attached to poles driven into the
bed. It was freely open to the changes of tide. Fish from an adjacent pool were
transferred to it from time to time for experiment and observation.

When three or four dogfish were liberated in the pen, they swam leisurely up and
down its length in a mid depth of water, occasionally resting on the bottom. If, during
their swimming, some crushed crabs wrapped in cheesecloth were thrown into the pen,
they quickly changed their method of locomotion. When a fish in the course of its
ordinary swimming approached to within a few feet of the packet of crab meat, it usually
made a sudden movement to one side with its head, swam at once to the bottom, and
in quick circuitous turns, often in the form of a figure eight, swept the bed of the pen.
In a short time it narrowed its search to the immediate vicinity of the packet and when
at last its mouth was brought close to the bait, this was seized, shaken, and carried off.
Occasionally the packet was not found and the fish then resumed its ordinary method
of locomotion until again by accident it came close to the packet, when it would repeat
the movements already described. The crab meat used in these experiments was
always wrapped in cheesecloth to exclude the possibility of the recognition of the meat
by sight. The dogfish, as is well known (Sheldon, 1911), has extremely poor vision,
at least in daylight, and seldom responds to an object until the latter is within a foot
of it.

If the nostrils of a set of fishes to be experimented upon are filled with cotton wool
before the animals are liberated in the pen, no attention whatever is paid to the packets
of crab meat, though the characteristic reactions return shortly after the removal of the
cotton. These experiments, first tried by Sheldon (1911), show that the dogfish is
excited by food through the stimulation of its nasal organs and that it subsequently
locates the food by the same means; in other words, the dogfish scents its food as air-
inhabiting vertebrates do.

The question that I set to decide was that of the precise method by which the dog-
fish reached its food. After the initial excitement, does the fish move aimlessly about
till, by accident, it runs upon the packet of crab meat, or is it directed toward this packet
by some special form of stimulation? In working on this problem it proved important
to observe two conditions. First, it was found that dogfish of medium size were more
favorable for these experiments than larger or smaller ones. In general, fishes about 2
feet in length were found to be most satisfactory, and these were used in most of the
experiments. Secondly, the dogfish reacted well only when in companies. A single
dogfish in the pen seemed to suffer an inhibition of its activities excepting those directed
toward returning it to the school. It swam about with incessant efforts at escape and
paid very little if any attention to baits or food. It was therefore found best to work
with several fish at once, and usually a set was chosen in which the individuals differed
enough in size and markings to make them severally distinguishable. Often in a given
set the larger ones were prepared in such a way that they showed no response to the
packet of crab meat, and the smallest one of all was made the subject of the special experi-
ment. This one individual was provided with a favorable environment for its own reac-
tions without being subjected to disturbances from others.

If the substances emanating from the packet of crab meat exert a directive influence
on the dogfish through the sense of smell, evidence of this might be discovered by inter-

SENSE OF SMELL IN DOGFISH. 65

fering with one or other of the nasal organs. This might be accomplished by cutting
the olfactory tracts of one side or by temporarily occluding one of the nasal apertures.
Fish in which the olfactory tracts have been cut rarely live more than a few days after
the operation. They are apparently very susceptible to infection from cuts made in
the vicinity of the brain. I therefore abandoned this method of procedure and adopted
that of filling the nostrils with cotton wool. But even this method is not without its
defects. If a single nostril cavity is plugged tightly with cotton the fish will often fail
to respond to the presence of food precisely as it does when both are filled. On removing
the cotton from such cavities they are generally found to be inflamed, if not suppurated,
showing that their surfaces are decidedly delicate. None of these complications arise
when the nasal cavity is only lightly filled with cotton and yet this method seems to be
effective in checking the currents of water through the nose. I therefore adopted a
light plug of cotton wool as a means of excluding the action of a given nasal organ.

To ascertain the state of the normal dogfish, records were taken of the direction
of the head movements, to the right and to the left, in response to the presence of bait,
and of the time consumed between the moment when the packet of crab meat was
first scented by the dogfish and when it was finally seized and carried off. These records
served as a basis for comparison with the reactions of fish especially prepared for tests.
The records of the normal fish are given in table r.

TABLE 1.—RECORDS FROM FivE NORMAL DOGFISH.

| Movements of the head. Z
Time in

Number of fish. | [na |e sccouds to)
| To left. To right. [Secure bait.

Darah rein (stescrotevaveratalaisisistoriciatetatatates toto tetavererere’eteteictorcrerlate ctaleteveratalatois eizie’clotevaherscaiaie Oe raa/overeroe ets Sos sete 14 16 186
Igo OGG Bobo So do DOS ORY E470 OD SOO TA A Src e ero Sem rir Re OL Iria cae Cones ear Pa 4 erent Saal 27 28 108
Gp doi OO Ss CO ANSROUOE RRL D OO Nnb CORO An WHO CSO AMEE DOULOTO SOS OD DOOD EERE oRED ONG BG se sudemr eras | 21 18 67

PA ETH LOS wae cte ris serateisiernisre leer etietate meters tre ieciekestefeheste chet eteie Bertha sese Tee ee icles o Maee cere 20.2 20. 8 | 131-4

From table 1 it will be seen that the dogfish tested found the bait in a little over
an average of two minutes and that they accomplished this operation by making about
as many right-handed as left-handed movements. Many of their movements were
combined and resulted in more or less continuous and characteristic courses in the
form of a figure eight.

In a second series of tests dogfish in which the left nostrils had been lightly filled
with cotton were subjected to the same conditions as were the normal fish whose reac-
tions are recorded in the preceding table. The records of five of the fish with the left
nostrils occluded are given in table 2.

66 SENSE OF SMELL IN DOGFISH.

TABLE 2.—RECORDS FROM Five DocGrFisH witH Lert Nostri, OCCLUDED.

Movements of the head.! ‘Time in

seconds
Number of fish. Raectice
To left. To right. bait.
3 26 132
6 32 116
° 24 IIo
2 16 121
4 32 14r
3 26 124

As table 2 shows, the five dogfish with their left nostrils occluded consumed on
the average about the same length of time to find the bait that the normal dogfish
did, but they accomplished this by a predominance of right-handed movements rather
than by an almost equal number of right and left handed turns. Their movements
were essentially circular and seldom, if ever, in the form of a figure eight.

A third series of dogfish were prepared by filling their right nostrils lightly with
cotton wool. As might have been expected from the results already recorded these
fish found the bait in about the same time as those with the left nostrils occluded, but
by means of left-handed movements chiefly. The records of this set are given in
table 3.

TABLE 3.—RECORDS FROM Five DocrisH witH RIGHT NosTRIL OCCLUDED.

Movements of the head. oe
Time in

Number of fish. seconds to
Toleft. | To right. | Secure bait.

123

92
rir
116
131

cl
x
WuUrHR

24-4 | 2.6 114-6

As tables 2 and 3 show, dogfish with one occluded nostril each tend to seek their
food by moving over a more or less circular path and in such a way that the open nostril
is toward the center of the circle.

Having ascertained that in seeking food or a bait normal dogfish turn about as
much to the right as to the left, and that those with an occluded nostril turn predomi-
nantly toward the side of the open nostril, I attempted as a check series to repeat all
the experiments thus far described on one set of fish. The plan of this series was to test
five dogfish, first as normal individuals, then with their left nostrils occluded, next
with their right nostrils occluded, and finally with both nostrils free. The fish were
allowed to rest one day after each trial with occluded nostrils, so that the whole series
of experiments covered a period of about three days. Unfortunately, during this period
two of the fish made their escape and the series was completed with only three fish.
The records of these three fish throughout the whole set of tests are given in table 4.

SENSE OF SMELL IN DOGFISH. 67

TABLE 4.—RECORDS FROM THREE DOGFISH UNDER SUCCESSIVE CONDITIONS, NoRMAL, LEFT NostRi
OccLUDED, AND BotH NostTRILs OPEN.

Movements of the
head.
State of fish. TE ova 8 aeate
To left. | To right.
|
|
I 2r 26
IM OTINIAL ePyetn,aroic die é w(eelets nie iat ale Sa ale bse cia: Malctstcle ates eeye tia aha raa mie eta clslere setalcbolcisie nc alereideeaaalewreciats 2 18 16
3 3r 28
LU ETAL © nb revors sholtictercinicierarareivineleteistatcin nists raveinic eiemiseicrereisie love vivivie (aie ie sia\eveeinie nla Sinicmvoieselolelsvieiciaieeivere| sreicrnicie cles e 23-3 23-3
| I 3 24
TLeftsnostril occluded 2522 c.0 - cis tclctamce S Sen ea a wnlcted a eee cfdath osc, cielefaca ala elasaiaio siavele(aleiaie eoie’éperescitiorg bretote 2 4 31
\ 3 3 go
NCTA R Cate ray cele taralalettayeisielaraia eiclere raps aciemtela stairs oisieteinisietolels es Tota'ela(e/s/cisin/a]avelere's e'atcre/etelessjais eis /a'a:feisrs isa Jette etetatet= 3-3 28.3
| I 26 3
Right nostrilliocchided \s-4- 6 aco aes ciate cee ler wet du sicelatcctcks ioleidieide's docie oblcwinwe Baise oleate wreoloetet 2 28 °
3 19 4
LNG ERTS da oe agaces naa genos at ocnocaaeHsdscooouo seo cadcosapsegd dea snoDseeddosoane Happoeded Sosenaccos 24-3 2-3
=——
I 19 18
Both nostrils opetien. sacmecnee came sels caciicceine a macleeinsisleicroe siecle ow iafc siete clom umisck einai 2 2I 17
3 16 27
MAN V OLA ON). ai cicin ln; coleialo/orcveleln olaraiaveteicrctereToreromtaetetereferere einiaie ara, uic ielelelepe tote wislelai cleo svepe wiorlo cisjereieieie eversinereis Jeveeeee ees 18.7 24
|

It is quite evident from an inspection of table 4 that the conditions recorded in
the preceding tables are not due to individual peculiarities on the part of the fish used;
that a given fish, which in finding its food normally turns as much to the right as to the
left, can be forced to assume either a predominantly right-handed or left-handed course
by occluding the appropriate nostril, and that after this treatment the same fish can
recover its normal methods of search by the liberation of both nostrils. It must also
be evident from the whole series of records that the dogfish does not run upon its food
by accident, but finds it in response to an influence more less directive in character.

The consistent and striking circular courses that these fishes can be forced to assume
have, in my opinion, more than a superficial resemblance to the so-called circus move-
ments of the invertebrates. These movements are dependent on the differences of
intensity of stimulation on the two sides of the body and this explanation holds, I
believe, for the circular movements of the dogfish. When a dogfish first enters water
permeated with odorous material from its food, it invariably makes a quick turn with
its head which, if the conditions of the water have been disturbed by currents, is always
toward the bait. This movement is followed by other movements of a like kind whereby
the fish eventually reaches the bait. When the normal conditions of the fish are dis-
turbed by the complete occlusion of one nostril, the fish swims as though it were in
water that was highly charged with odorous particles on the side of its body correspond-
ing to the open nostril and devoid of these particles on the opposite side. The fish
therefore turns toward the side of the open nostril, but since, under the artificial con-
ditions of the experiments, this turn does not equalize the stimulus, the motion is con-
tinued and a circular form of locomotion results. Thus, in my opinion, the more or less
circular movements induced in a dogfish with an occluded nostril by an odorous bait
are to be explained upon the same basis as the circus movements of such invertebrates
as crustaceans, insects, etc.

68 SENSE OF SMELL IN DOGFISH.

The movements of a dogfish differ from the movements of these lower animals,
however, in that they are not pure circus movements. A dogfish with a fully occluded
left nostril not only turns to the right but sometimes to the left, and a normal dogfish
so often exhibits movements in the form of a figure eight that it is quite clear that the
whole figure is a single response rather than two separate acts due to alternate excessive
stimulation first on one side and then on the other. Thus, though the responses of the
dogfish to the odors of food may be based predominantly on the principle of symmetrical
stimulation, it is also clear that odorous material calls forth from this animal what are
essentially random movements.

In consequence, the finding of food or of a bait by a dogfish may, thereon be
described as brought about by a combination of movements, partly random and partly
directed, which have resulted from stimulations due to the varying concentrations of
odorous materials in the surrounding sea water. The dogfish, like other elasmobranchs,
has a structure especially favorable for this form of stimulation in that its nostrils are
wide apart, a condition which is immensely exaggerated in the hammerhead shark,
whose nostrils as well as eyes are lodged at the extreme ends of the transversely extended
processes of its head. These conditions make it clear why chumming is so effective with
mackerel and other fishes. When a fisherman spreads bait to attract such fishes from
a distance the response is undoubtedly directive, especially on the part of sharks,
which have been seen to come up to food against the tide from as great a distance as a
quarter of a mile. Such fish must keep within the stream of odorous substance in the
water by responding to the stimulation of their nasal organs in much the same way as
the dogfish was found to do in seeking a bait.

BIBLIOGRAPHY.
Aronsoun, E.

1884. Beitrage zur Physiologie des Geruchs. Archiv fiir Anatomie und Physiologie, physiologische
Abteilung, jg. 1884, p. 163-167.
BaGLIonl, M.
1909. Zur Physiologie des Geruchsinnes und des Tastsinnes der Seetiere. Zentralblatt fiir Physi-
ologie, bd. 22, p. 719-723.
CopELAND, M.
1912. The olfactory reactions of the puffer or swellfish, Spheroides maculatus (Bloch and Schneider).
Journal Experimental Zoology, vol. x11, no. 3, p. 363-368.
Parker, G. H.
1g10. The olfactory reactions in fishes. Journal Experimental Zoology, vol. vim, no. 4, p. 535-542.
tot. The olfactory reactions of the common killifish, Fundulus heteroclitus (Linn.). Ibid, vol. x,
110+ :E; PseT—5-
Parker, G. H., and SHELDON, R. E.
1913. The sense of smell in fishes. Bulletin U. S. Bureau of Fisheries, vol. XXxXn, 1912, p. 33-46.
SHELDON, R. E.
1909. The reactions of the dogfish to chemical stimuli. Journal Comparative Neurology and Psy-
chology, vol. XIX, p. 273-311.
tg11. The sense of smell in selachians. Journal Experimental Zoology, vol. x, no. 1, p. 51-62.

THE STORAGE OF FAT IN THE MUSCULAR TISSUE OF THE
KING SALMON AND ITS RESORPTION DURING THE
FAST OF THE SPAWNING MIGRATION

a

By Charles W. Greene, Ph. D.

Department of Physiology and Pharmacology, Laboratory
of Physiology, University of Missouri

:
Ady vs 3 i At

7 7

a

CONTENTS.

ad

Page
ToikaTeVelTARO».ononagcoso cdedoooddwaccnacdonegonon Be ekopneD ancoroUdsoneDoornnou toad conurage 73
Materialeand methodsiotsinvestigatiotienessrm vere triste ee el vetopelcieiayaleints iota) = sfaciatetalciciens elereiererorere 74
Types of salmon muscular tissue as regards the storage of fat. ... 2.2.2.0... oe eee eee eee 78

Normal loading of fats in the muscles of the king salmon at the time the spawning migration
ESIs5 soa asoopmaaveoe cu de eno CObeb ha pod scA a paGnns coon DUDE ooDEE Sep esneuoopoSUESHEeDoES 80
lero teee) HER AA ARE Oneonta oo bo Onn CoOp ds GEE Ee ORE CC DE e TUE orn UneacE aa orrEe 84
Variation of the amount of fat in the salmon during the spawning migration. ................. 86
Distribution of the fats of the salmon muscle at tide water... 2.2.2... eee 87

Analytical determinations of the percentage of fats in salmon from the mouth of the Colum-
(ELIRU ATG n le nie Sinn EL COOLDE Edd 6 BERGE DARE eaoe On SSanOne UTOGeS Gap a aOU CUUeECROneDoEuS or
EMCO. 6 5 ddoosocsvcovacaooospU pHoabebo UN ROE onnoD co atanDooU LE EDaouDoboODeOOnEBOS 93
Distribution of the fats at an early intermediate stage of the spawning migration............ 98
IB OWE. 5 so hob ooo e.5 pOOOHHBaD Soman eOodE Ho ouosb Enos en SEDOC UBD GOOCeODOdaaobDUdoS 102
Distribution of the fats at a late intermediate stage of the spawning migration.............. 108
Bmw. 5 onadncoooDdEe uso UoOEdouBUrOdodcoNDanS 6oodEocKe cb ucdoecoUseaepeEUUsUEOCS 109
Fat in the tissues of salmon from the spawning grounds... .. 2.22... 0 2c. eee eee eee eee III
Analytical determinations of the percentage of fats in the spawning salmon................ 114
SMOSH oo doadbosheasusnpverabEuboopebnoooouDeopeemaoDNodUGDUOW CURD OO GOIDDODOnOGE 115
Significance of the observed changes of the amount of fat.............. 6.65.02 eee e eee sees 118
‘Transportation of fats in the fasting salmon ............2.. 6. see ec scence cece rete cncscn sc eeees 119
[sbkigell ee con oeocucaucEc cu badenoucanede dad Co ddom eo omAmndobd4 ConuDooOod DOU USO cuOUarL 119
Mechanism of fat transference in the salmon body... ................. 00sec eee eee e eens 124
sBransferenceofratymit hevp ism Kati 1SCle ererateratelerelstaiete rial ete altiers ele eteteke rete ie iataretetctel k=) ereieialal=t= 125
Miransterence oftateanm che) darkamtisclejpeyaeen rier rietaciereterecertaeieteeietsr=ieieiciael ote eeteloerssteienele 127
Relative abundance of stored fat and its disposal in the organs... 2... i.e eee eee eee eee eee ee 129
Sribroyn WEES. comic oousgodoodonucabobdesaahduDoDoUbde FA anee sO IegdcUconsdeccagnacoDnEDnO 130
Saree oF te WECOs. So gecoauaaanooumscons sdmooeHD HD OD DU OHOO As opoDDCUgnaDUGOUDOONNDUM 131
INGE: 6 oo sca obosnaobses Hoban Hone ooGcdnoo lpps cose oeReDcosmaadosocuooDapunnASeDlEnooUaS 136
Explanation ofsplatesisc rare isrersts|<ieleversyatero er siereraleretoieletareKe letsials ele!) lene ol ele fol f~(e/nfoishalol=iwichalainVelerololers)siai=)at= 137

fal

hs obk cridpaie

hence neon Va
‘

12h ie sullyet

wilt or

nil

ii {9-4 ful) Ae
pes!
suortectiea@ wells Teak
‘ -. TWH S,
Je ite
ul Pag tell
ot
es Wii Ww
. t fe
[us ; 1) Sayre
' + tee
‘ a ie i
LecdspFan itt
rsh soi Wares rth
+ vi et ort
=

THE STORAGE OF FAT IN THE MUSCULAR TISSUE OF THE
KING SALMON AND ITS RESORPTION DURING THE FAST OF
THE SPAWNING MIGRATION.

*

By CHARLES W. GREENE, Ph. D.,
Department of Physiology and Pharmacology, Laboratory of Physiology, University of Missouri.

ad
INTRODUCTION.

To the present no study has been made of the distribution of fats in the muscu-
lature of the king salmon, Oncorhynchus tschawytscha. We have the classic studies of
Miescher* on the quantitative chemical variation in the fats of the Rhine salmon,
Salmo salar; also the studies of Noél Paton and his coworkers on the same species from
the Scottish rivers. Miescher found, as did Noél Paton after him, that there was a
great decrease in the quantity of fats as the fish ascended the rivers and as the spawn-
ing season approached.

Miescher also observed fat in the muscle fibers in considerable quantity. This he
considered to be a stage of fatty degeneration preliminary to the use of the fats for the
building up of the immense store of food materials present in the salmon ova at the time
of spawning. Miescher states that the fat granules increase in numbers in midsummer,
and are absent after the salmon have spawned. He also states that the fats are practi-
cally absent in the fibers of the muscles of the smaller fins and of the head.

Mahalanobis,’ as a part of Noél Paton’s series of studies, made a special histo-
logical examination of the fat in the muscles under the subtitle ‘‘ Microscopic observations
on muscle fat in the salmon.” There are six pages of text and seven microphotographic
figures presented in his paper. Less than two pages are used to present the findings
as to the distribution of fats in the muscles. He contrasts the variations in the fat
of ‘‘a fish fresh from the sea”’ with fish taken up the rivers at later dates. The greater
part of the text is consumed in a discussion tending to disprove the ‘‘fatty degenera-
tion” theory which Miescher has assumed, and Miescher is undoubtedly in error
and Mahalanobis right in this particular matter. But Mahalanobis has failed in his
special comparisons between two extreme types of fish from the simple fact that he based
the comparisons on dissimilar muscles. (See p. 121 of this paper.) The studies on the
Atlantic salmon have, therefore, left us without adequate description of the normal
histological distribution of the muscle fats. As for the variations in the microscopic

@Miescher, Friedrich: Statistische und biologische Beitrige zur Kenntniss vom Leben des Rheinlachses im Siisswasser.
Schweizerischer Fischerei-Ausstellung in Berlin, s. 154, 1880. Also reprinted in Die histochemischen und physiologischen
Arbeiten von Friedrich Miescher, s. 116, Leipzig, 1897.

bIn: Paton, D. Noél, Report of the Fishery Board for Scotland, 1898, p. 143.

73

74 BULLETIN OF THE BUREAU OF FISHERIES.

picture as the salmon proceed up the rivers, one may say, with all due respect to the
previous work, that such picture has not yet been given for any species.

The king salmon of our Pacific waters is an entirely distinct genus from that of the
Atlantic salmon. It is decidedly unique in its biological characteristics, as shown in a
number of fundamental ways. These characteristics have been enumerated at various
times,“ but three facts of peculiar importance may again be stated here as most vital
to this investigation:

First, Oncorhynchus tschawytscha is an anadromous fish.

Second, it fasts completely during its entire journey from the sea through the tidal
waters and up the rivers to the spawning grounds.

Third, Oncorhynchus always dies within a short time after spawning.’ It can not,
therefore, return to the sea for a second period of feeding.

These facts, taken in connection with my numerous quantitative chemical deter-
minations revealing variations in the percentages of the fats within wide extremes,
form the scientific background of interest in this investigation. There are enormous
economic interests involved on the Pacific coast in the catching of the king salmon and
the conserving of the flesh as food. There is a well-known great variation in the grade
of the commercial products derived from this fish. These facts add to the scientific
interest in the problem, and emphasize the necessity of a study of the histological distri-
bution and utilization of the fats in the king salmon, the largest and finest of all the
members of the family of Salmonide.

At the time the king salmon leaves the ocean for the fresh water it has reached the
crest of the life cycle as regards the amount of fat deposited inits tissue. In the Colum-
bia River those salmon caught at the lowest point at the mouth of the river are the fattest
of all. The salmon do not feed, i. e., they do not absorb new food materials, during their
return passage in fresh water. There is no other adequate storage of food material than
this fat. One may, therefore, safely assume with Miescher that the fat is the chief
source not only of the fats that go to build up the ovaries but also of the energy liberated
in muscular contractions during the migration journey. One may also assume as a
working hypothesis that the fat will decrease in amount until the spawning is accom-
plished and the death ensues. I have followed the variation in chemical composition in
the king salmon muscles, including the amount of fat at the different stages of the spawn-
ing migration, work that is represented by unpublished results obtained from chemical
analytical determinations, and have also followed the microscopic distribution and varia-
tions of the fat in fat-stained muscular tissues. It is the results of this latter work to
which attention is called in this paper.

MATERIAL AND METHODS OF INVESTIGATION.

The facts presented in this paper concerning the salmon fats have been determined
by the microscope.° The chief reliance has been placed on the special methods for

a Greene, C. W.: Physiological studies of the chinook salmon, Bulletin of the U.S. Bureau of Fisheries, vol. XxIv, 1904,
P. 429-456.

+ Evermann, B. W.: A preliminary report upon salmon investigations in Idaho in 1894, Bulletin U. S. Fish Commission,
vol. XV, 1895 (1896), p. 253. Also, A report upon salmon investigations in the headwaters of the Columbia River in the State
of Idaho in 1895, together with notes upon the fishes observed in that State in 1894 and 1895, ibid., vol. XVI, 1896, p. 184.

cI wish to express my deep obligation to Mr. Thomas J. Heldt, instructor in anatomy, University of Missouri, and to Mr.
George T. Kline, biological artist, University of Missouri, for their valuable assistance in the taking of samples and in the
making of the numerous field observations on fresh materials, on the excellence of which much of the detail of this paper rests.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 75

staining fats for microscopic determination. The particular technique used throughout
the series is the Herxheimer method of staining with scarlet red as modified by Bell.
This direct histochemical method has been supplemented by histological preparations
fixed and sectioned in paraffin and treated with various differential stains.

Selection of salmon types.—The problem of fat variation in the salmon during the
migration is twofold. First, there is variation in any one region of the body at different
stages in the migration journey, and second, there is variation in the amount of fat present
in different parts of the body at any one time. In order to determine the variations of
the first class one must, of course, select typical localities representative of different
stages in the journey. For this comparison I have used salmon from the Columbia
River Basin, selecting four stations, the first one being at the mouth of the river,
working from the town of Ilwaco; the second below the cascades, working at Warrendale;
the third at Celilo Rapids and Celilo Falls, working from Seuferts; and the fourth at
the spawning grounds on the Clackamas River at Cazadero.

There is a great variation in type among the individual fish present at any one sta-
tion at any given time. It therefore becomes a matter requiring some considerable
skill to select consistent types throughout a comparative series. Only after several
years of experience can one select these types with some little assurance that he will be
following similar salmon through the variations that occur at the different stations or
stages in the migration journey. The attempt was made to secure fish just as early in
the migration as possible, i. e., the stage at which the feeding ceases and the migration
begins. In several years of collecting at Hwaco at the mouth of the Columbia River,
I have found that this very desirable stage can not be had within the fishing limits of
the locality; in fact, the cessation of feeding must begin some considerable time before
the fish reach the fishing zone. In order to secure the facts as to the normal fat loading,
it was necessary to make a study of the feeding salmon at Monterey, where salmon that
are known to reenter the Columbia Basin are caught during the feeding stage.

Salmon from the four stations on the Columbia mentioned above were collected
during the months of August and September, 1911, 30 salmon in all being studied
in detail. The stations were visited in the order in which they are named above.
Invaluable aid and cooperation were constantly received from the local fishermen and
from the packers at the various stations.* There is an undoubted seasonal variation
in the condition of salmon on the Columbia River, but by a rapid collection of material
from station to station the effects. of the seasonal variations on the series ought to be
reduced toaminimum. ‘The finest salmon at Ilwaco, for example, as regards the amount
of fat, are reported by the cannery men to come earlier in the season, though the fact
has not yet been determined by scientific methods of measurement.

As a routine practice, salmon from each station were chosen of only two classes,
the choice or best type from the packers’ point of view, and the poorest type, as inter-
preted by the same standards. Only the choice sizes, ranging in length from 80 to 100
centimeters were selected in this series.

@ TY am particularly indebted to the P. J. McGowan Co. for aid in securing salmon types, and for numerous material favors in
the pursuit of the field work at Ilwaco and at Warrendale. The Warren Packing Co. and the Columbia River Packers’ Associ-
ation granted the use of their receiving house on the Ilwaco Dock for a laboratory at that station. Mr. Frank A. Seufert, of The
Dalles, granted an unlimited supply of salmon from his Celilo fish wheels and the Tumwater seining grounds at Celilo Falls.
‘The Cazadero observations were made at the station of the U.S. Bureau of Fisheries at that point on the Clackamas River, where
the hatchery force rendered liberal and invaluable assistance.

76 BULLETIN OF THE BUREAU OF FISHERIES.

Selection of tissue-—In order to follow the variations of fat in different regions of
the body one must, of course, select typical tissues as regards their relations to fat.
For this purpose I have chosen primarily the great lateral muscle, because it represents
by far the greatest mass of the fat-storing tissue of the salmon. This muscle is divided
into two types; the deep or pink muscle and the superficial or dark muscle. Each extends
from the head and pectoral girdle to the base of the caudal fin. There is considerable
variation in the amount of fat in different parts of these muscles. Two type regions
of the lateral muscle have, therefore, been selected for study. The first is the mid
region of the side of the body, in the transverse section which cuts the body just at the
front margin of the dorsal fin. This with the muscle anterior to the section is the fattest
portion of the great lateral muscle mass. Samples from this region are called trunk
muscle, dark or pink, respectively, as the case may be. The second region of the great
lateral muscle chosen is that portion at the base of the tail opposite the fifth and sixth
caudal vertebre, counting from the posterior end. This region is called the caudal
muscle, dark or pink, according to its type. The caudal muscle always has a relatively
less amount of fat than the trunk muscle. It is assumed that there is a more or less
gradual variation from the caudal region to the mid lateral or trunk region as regards
the percentage amount of fat.

While the primary comparisons are with reference to the two regions of the lateral
muscle mentioned, still a considerable number of samples were taken from other muscles;
namely, the anal fin muscles, the ‘‘belly’’ muscles, the intercostals, and the cheek
muscles.

Fixation of tissues.—In view of Bell’s? experience, it was thought better to make
the examination of the fat in the salmon muscle in as fresh condition as possible, and a
full Bardeen freezing microtome equipment was therefore taken into the collecting field.
Samples of the tissues from the type localities were thus cut, stained, mounted, and
studied with the minimum of delay. It was quickly found to be quite impossible,
however, to carry through all the tissues in the fresh condition. There was not time
enough, with a field force of three, to do the necessary work before disintegration began.
It was therefore found necessary to use a fixative, i. e., 10 per cent formalin.

As a matter of routine practice, tissues that were to be held in the prolonged and
tedious processes of working up the material were always fixed in 10 per cent formalin.
This fixation was not found to be detrimental to the preservation of the fat, but on the
other hand, favorable. ‘The frozen sections cut after two hours or more in formalin
were firmer, retained their shape better, and therefore were not so much torn and
distorted in the process of handling necessary to staining. In short, the formalin-
fixed sections enabled one to arrive at a better conception of the relationships
of the fat than is to be had from the tissue cut directly from physiological saline.
After a series of observations made along these lines it was decided to pass all the tissues
through a formalin fixation. It is not claimed that a long immersion of tissues in
formalin is wholly without effect on the fats, yet for salmon muscle we are convinced
that the change produced is very slight. In certain samples sections of fresh tissues
have been compared with sections made after four months’ fixation in formalin. The
difference in quantity of fat shown is not easily determined, but there is some variation

@ Bell, E. T.: The staining of fats in epithelium and muscle fibers. Anatomical Record, vol. Iv, p. 199. 1910.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. Tig.

in the picture—enough to prevent one from using any but freshly fixed tissues for close
comparative estimates.

In taking samples, in all instances where possible the muscles were quickly cut
from the salmon selected while the tissue was yet alive. The muscles were cut in thin
pieces with a razor and immersed in a large excess of 10 per cent formalin. Other
tissues, such as the liver, ceca, stomach, ete., in which fat studies were to be made
were also included. Portions of the material not studied on the collecting grounds
were preserved for future use, some of which still retains its fat in approximately the
normal quantity and relations (after six months).

Portions of the muscles and of certain other tissues were fixed in different standard
preservatives for future study and comparison. But as these studies are not presented
in this paper the detail of fixation will be omitted.

Fat stain used and its preparation.—Sudan III and scarlet red are the fat stains
on which great reliance is placed in the histological identification of the fats. Since
the salmon tissues are so filled with fat it was decided to use the scarlet red as giving
the sharper differentiation. Bell has made a series of special studies on the micro-
chemical staining of fats. It is to him I am indebted for the modifications of the Herx-
heimer method which were followed.

The scarlet red was prepared in saturated solution in alkaline alcohol: Scarlet red,
2 grams; sodium hydroxide, 2 grams; 70 per cent alcohol, 100 cubic centimeters. This
was ripened in a 4-ounce wide-mouth vaseline bottle in a water bath at about 75° C,
for from 20 to 30 minutes. The ripening took place in a closed bottle. But it was found
essential not to heat too strongly lest the reactions lead to the production of a die that
would stain general protoplasm. This stain is supersaturated while warm and some
erystalization will occur on cooling. Bell’s recommendation that the stain be filtered
just before using was followed, also the precautions recommended by him against evap-
oration and sedimentation while staining. This stain retains its properties very well for
a very much longer time than one is led to expect from the directions published by Bell.

Technique of sectioning and staining fresh or formalin-fixed material for fats.—
Either fresh tissue or tissue fixed in 10 per cent formalin was frozen with carbon dioxide
on a Bardeen freezing microtome. The fixed material can be cut as thin as 25-30 4
with comparative ease, but the temperature of the block must be just right. The sec-
tions were received from the microtome knife directly into 70 per cent alcohol and
stained as quickly as possible.

The sections were handled always with a spatula of proper size. They were stained
directly from 70 per cent alcohol. After 5 to 10 minutes in the scarlet red stain the sec-
tions were passed as quickly as possible through 70 and 30 per cent alcohol into water.
The rinsing was sometimes through the alkaline alcohol of the grade in which the stain
was dissolved. This is a safer plan to prevent precipitation of the excess of stain from
adhering to the surface of the sections.

When sections were to be counterstained a short washing in acid water, 0.2 per
cent hydrochloric acid, preceded the hematoxylin counterstain. The acid treatment
was finally adopted as a routine procedure for all sections.

The stained sections were mounted directly from the wash water into pure glycerin.
The glycerine was of course slightly diluted by the adherent water. The glycerin
mounts were sealed with paraffin, or better with paraffin-beeswax cement.

19371°—vol 33—15——6

78 BULLETIN OF THE BUREAU OF FISHERIES.

The splendid keeping qualities of many of the scarlet-red preparations made during
the investigation is attributed to a thorough neutralization of the alkali in the bath
of acid water.

TYPES OF SALMON MUSCULAR TISSUE AS REGARDS THE STORAGE OF FAT.

The musculature of the salmon® is relatively simple. The main mass consists of
the great lateral muscle of Cuvier. A number of smaller muscles are associated with
the various fins and with the structures in the head region. But the storage of fat takes
place chiefly in the great lateral muscles.

Great trunk muscles.—The great lateral muscles include the major masses of muscu-
lar tissue along the sides of the body from the head and the pectoral girdle to the base
of the tail. These masses represent from 60 to 70 per cent of the total weight of the
mature fish when in prime condition.

The great lateral muscle of the salmon has lately been described.? In Amiurus® it
is described as divided into five longitudinal portions more or less distinct. In the salmon
there is a connective tissue septum, the lateral line septum, running the length of the
body just under the lateral line. It separates the lateral muscle into a dorsal and a
ventral portion, each of approximately the same size. Aside from this there is no further
subdivision along the lines designated by MeMurrich.

However, the great lateral muscle is divided into two distinct muscles on the basis
of the well-marked anatomical and histological differentiations that have taken place.
There is a superficial thinner portion which is anatomically distinct and easily identi-
fiable because it is darker in color; and a deeper, wider, and thicker mass which is pink
in color. These two divisions are well separated by a thin lamina of fibrous connective
tissue bearing an excess of adipose tissue in most of its extent. The lateral line septum
separates both the superficial and the deep muscles into dorsal and ventral divisions.

The fin muscles, and to a less extent the head muscles, are the ones in the most
constant but slight activity in the daily life of the fish, and, strangely enough, whether
for this reason or not, these muscles are not largely loaded with reserve fat.

Musculus lateralis superficialis, or dark lateral muscle.—The superficial dark muscle
forms a type of tissue with respect to its loading of fat that has not previous to the
preliminary notices of the present writer? been described, in so far as I can discover,
though it was noted by Miescher.¢ He undoubtedly refers to the superficial lateral
muscle when he says: ‘‘A thin muscle plate lying along the side of the body just be-
neath the skin degenerates most strikingly (Hautmuskel).’’ Miescher makes the refer-
ence quoted ¢ in the brief discussion of the microscopic picture of the fat in the so-called
fatty degeneration.

This muscle is characterized by the following points: (1) Its compact arrangement
of fibers; (2) the smallness and uniformity of size of the fibers; (3) the relatively small

@ It does not seem wise to regard the supracarinales and infracarinales as divisions of the lateral muscle in the ordinary sense
of the designation.

> Greene, C. W. and Carl H.: The skeletal musculature of the king salmon. Bulletin U. S. Bureau of Fisheries, vol.
XXXIM, 1913 (1914), p. 21-60, pl. I-11.

¢ McMurrich, J. P.: The myology of Amiurus catus. Proceedings of Canadian Institute, n. s., vol. m, 1884, p. 328.

dGreene, Charles W.; A new type of fat-storing muscle in the salmon, Oncorhynchus tschawytscha. The American
Journal of Anatomy, vol. xm, p. 175, 1912. Also an undescribed longitudinal differentiation of the great lateral muscle of
the king salmon. Anatomical Record, vol. vu, p. 99, 1913.

¢ Miescher, op. cit., p. 186 (145).

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 79

amount of interstitial connective tissue; and (4) most important of all, by its enormous
loading of fat.

Musculus lateralis profundus, the lateral pink muscle.—The profundus, or the deep divi-
sion of the great lateral muscle, is the pink salmon muscle as it is ordinarily spoken of.
It has a totally different appearance from the superficial dark muscle. The fibers of
the pink muscle vary enormously in size, from 40 up to 250 » in diameter in the adult.
The average size of the fibers varies somewhat in different regions of the pink muscle
even of an adult fish. In the young salmon, ro to 15 cm. long, there is a greater range
of variation than in the adult, as is shown by the outline figure 16, plate x. This
is due to the fact that the fibers are undergoing longitudinal cleavage which is very
unequal. This cleavage leads to a large number of very slender fibers. It is the method
of reproduction of new fibers which leads to the irregular outlines noted in the cross
sections of all the fibers of the profundus of the salmon, both young and adult.

The amount of supporting connective tissue in the pink muscle is relatively great.
Beside the blood vessels, a large amount of adipose tissue is present. It is the adipose
tissue of the pink muscle, which is heavily loaded with fat, that carries the greater
part of the fat of the salmon commercial products.

Myocommata of the great lateral muscles —The myocommata which separate the
muscle myomeres are always crowded with fat in the normal adult salmon. These con-
nective tissue partitions are composed of white fibrous connective tissue into which the
short longitudinally placed muscle fibers are attached. ‘The tissue is largely filled with
adipose cells. Its fat cells are large and relatively uniform in size when filled. They
form a considerable mass within the myocommata most thickly studded near its center.
There are also large numbers of fat cells on the surface and crowded beneath the ends
of the muscle fibers. This fat forms no inconsiderable portion of the storage fat present
in the adult salmon.

Supracarinales.—There are longitudinal-paired muscles along the middle of the
back of the salmon. These extend from the head to the dorsal fin, the supracarinales,
and from the dorsal fin to the adipose fin and to the dorsal lobe of the caudal fin. These
paired muscles are cylindrical in shape and about the size of a lead pencil in the thickest
part, but spreading out somewhat anteriorly. The muscles are of interest in this con-
nection chiefly because they are imbedded in a relatively thick and prominent mass
of adipose tissue. In a prime fish this adipose tissue is crowded with fat.

Infracarinales—A similar collection of adipose tissue is even more striking along
the mid-line of the belly of the salmon. The fat of the belly surrounds the protractor
ischii anteriorly and the retractors posteriorly. The mass is from 1 to 2 cm. thick and
twice as wide in a prime 80 cm. fish. The cylindrical muscles inclosed will be about
0.8 to 1.5 cm. in diameter, and the rest of the area an almost solid mass of fat cells.
In a fish low in fat the fat is taken up, leaving a white fibrous connective tissue mass.
These two areas of adipose tissue form a considerable storehouse of salmon fat.

Muscles of the fins —The storage of fats in the fin muscles has been studied only
sufficiently to demonstrate the type. Adequate comparisons have not been made to
make the observations complete, except in establishing the normal type. The fin muscles
of the pectoral, ventral, dorsal, anal, and caudal fins (the deep caudal muscles) are much
alike in general fat loading. Sections of the erectors and of the depressors of the anal
fin have been most extensively examined. The two muscles are much alike in general

80 BULLETIN OF THE BUREAU OF FISHERIES.

appearance, are a light reddish-brown color—neither pink nor as dark as the super-
ficial lateral muscle. The constituent fibers are loosely attached to the interhemal
septa, and the pair of muscles between adjacent interheemal spines are incased each in
a stout connective tissue sheath. ‘This arrangement is also characteristic of the cor-
responding muscles of the dorsal fin.

Muscles of the head.—The masseter muscle is the largest of the muscles of the head.
Under the name of ‘‘cheek muscle,” the masseter is highly prized as a delicacy by the
fishery folk. It is of good size, has its fibers compactly arranged, and is not colored like
the trunk muscle of the salmon. ‘The cheek muscle of feeding salmon has not been
examined, but the Ilwaco and Cazadero types, representing the mouth of the Columbia
River and the spawning grounds, respectively, are presented in the proper places.
Other muscles of the head region have not been examined for fat.

NORMAL LOADING OF FATS IN THE MUSCLES OF THE KING SALMON AT THE TIME THE
SPAWNING MIGRATION BEGINS.

It is exceedingly difficult to secure salmon just at the moment when fasting begins.
In the first place, it is not easy to determine just when a salmon ceases feeding; that
is, whether a given salmon in hand is one that is still feeding or one that has just ceased
feeding. A second and more important difficulty is that of catching salmon at this
critical stage in the life cycle. There are only limited regions where the king salmon
are captured from the feeding grounds. There is no such place near the mouth of the
Columbia River. The lowest point at the mouth of the Columbia where fish are caught
is between the Canby Lighthouse on the north bank and the end of the Government
jetty on the south shore. Salmon from this locality have already ceased feeding, prob-
ably some little time earlier.

Monterey Bay and its immediate vicinity is a popular ground for king salmon
fishing. When the salmon schools are in this vicinity they are actively feeding and are
readily caught with the trawl. During the spring and early summer months they are
taken in large numbers. At this time a considerable business is done in the salmon
fisheries at the city of Monterey. There is good evidence that the fish caught at Mon-
terey Bay are from schools which ultimately enter both the Sacramento River basin
and the Columbia River basin. ‘Thisis indicated by the fact that on both rivers speci-
mens are occasionally taken which have in their mouths fish hooks of the type used
at Monterey.®

Monterey Bay is about 100 miles south of Golden Gate, the entrance to the Sacra-
mento Basin. It is about 800 miles south of the mouth of the Columbia River. A
well-developed salmon from Monterey would serve very well as a normal type for the
Sacramento Basin, provided one could assure himself that the fish were on its way to
and would enter the Golden Gate and the Sacramento. The amount of change that
could take place due to additional feeding between Monterey Bay and the Golden Gate
would be negligible, assuming a reasonably direct journey. Hi additional fat were stored
it would be too slight to change the average materially. But facts tending to verify
these assumptions can not readily be obtained.

@ Mr. George Warren, of the Warren Packing Co. of Portland, Oreg., showed me a number of such salmon hooks taken from
salmon that have come into their packing establishments on the Columbia River.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 81

It is out of the question to assume that a Monterey fish destined to migrate as far
north as the mouth of the Columbia River is in the condition that will exist at the time
of arrival. Such a salmon is in the growing stage. It will certainly increase in size in
so long a feeding journey and probably will also somewhat increase its fat content.

The Monterey fish are the only ocean-feeding fish available as examples of mature
specimens typical of the Sacramento and to a certain extent of the Columbia Basins.
It follows, therefore, that one must use these specimens as best he can for the purpose.
The matter resolves itself, somewhat, into a question of the ability of the investigator
to select and judge the type that most nearly approaches the mature one.” ‘There is a
wide extreme of maturity represented among the Monterey fish. The small fish give
all evidences of being relatively young and growing specimens. The larger fish are proven
to be the older ones by the work of Gilbert, who finds a close correlation between size
and age. His determinations indicate that these larger fish have been feeding in the
ocean for four, five, or even more years, according to the size of the specimen chosen.
It seems reasonable to assume that such fish will not undergo any very great change
in the average fat content during the intervening months between the time of the Mon-
terey feeding and the beginning of the spawning migration. The larger Monterey fish
may be taken as the best available examples typical of the disposal of the tissue fat in
the late stages of the feeding cycle. On this ground observations and protocols are
presented on Monterey specimens.

In the chapter on the types of salmon muscular tissue as regards the storage of fat
the muscle characteristics are given in sufficient length to enable one to use them in
presenting the picture of the fats and fat variations. It remains now to give the
detailed picture of the normal fat content of salmon muscle at the time when the feed-
ing ceases. Under this category will be presented the following muscle types:

Normal fat content of the trunk pink muscle-—The pink muscle, which represents
the greater proportion of the total mass of muscle of the salmon, contains an enormous
total load of fat at the time the salmon cease feeding. Estimating on the basis of various
chemical studies made in other connections, I would say that this fat loading varies
between 15 and 25 per cent. This great variation represents the normal variation in
fat content.

The fats of the pink muscle are distributed in the connective tissue between the
muscle fibers—i. e., they are intermuscular. The pink muscle carries a relatively large
amount of connective tissue which supports the muscle fibers and the blood vessels,
and this connective tissue has a high percentage of adipose tissue. In it are found enor-
mous numbers of fat droplets, which vary within a wide range of size. The smallest
droplets are often not more than r or 2 4 in diameter, but there are numerous fat globules
of this region that are as muchas 100 win diameter. No figure is presented of this normal
material, but figure 8, plate v1, drawn from an Ilwaco specimen (no. 118), represents
very well the average appearance of the intermuscular fat of the normal tissue.

@ The alternative is to figure back from the first available stage in the spawning migration. For the Columbia River this
latter method I believe gives a truer picture of the normal condition. Attention will be called to this fact in the discussion of
Ilwaco types.

b Prof. Chas. H. Gilbert, who is making extensive studies on the salmon migration and the salmon age, observes that there
is, within certain limits, a close correspondence between size and age. It followsthat the larger fish have a longer ocean-feeding
period, a fact for which we have heretofore had no conclusive proof. Also it is evident that salmon mature sexually at greatly
varying ages. (Gilbert, C. H.: Age at maturity of the Pacific coast salmon of the genus Oncorhynchus, Bulletin of the
U.S. Bureau of Fisheries, Vol. xxxm, 1912, p. 1.)

82 BULLETIN OF THE BUREAU OF FISHERIES.

There is little or no intramuscular fat in the normal pink muscle. The loading of
the fat is intermuscular, in contrast with that in the dark muscle, where it is intramuscu-
lar. In the normal feeding, growing salmon there is no intramuscular fat, or at most
only a trace of fat in the pink muscle. This condition exists up to the time when the
salmon cease to feed. This statement is based on the examination of tissues of the
smaller salmon in the rivers and also on the examination of different sizes, including
the largest adults coming into the market at Monterey, Cal. In the latter there may
be an occasional trace of liposomes within the smallest fibers. Monterey fish that will
enter the Sacramento River basin can not be assumed to be wholly typical salmon that
have ceased feeding. Yet I think it safe to consider these as sufficiently mature adults
to serve for comparative purposes.

In the quite young salmon, from 7 to 16 cm. long, there is no fat in the pink muscle,
either between the fibers or in the fibers along the lateral portion of the body. In the
ventral or ‘“‘belly’’ muscle there is some intermuscular fat at this stage of development.
Salmon of this size are still feeding in fresh water. Of the sizes that one obtains at
Monterey, which of course are feeding in salt water, fat is beginning to be deposited in
the connective tissue between the fibers. This fat is relatively low in amount in the
smaller fish. There is great variation in its amount in different individual fishes at
Monterey, and while the number of fishes studied is very limited one can say that these
indicate that the fat increases in quantity with the size of the fish.

An exception to the above description is found in a narrow zone of pink fibers lying
on the surface of the pink muscle. This zone is immediately covered by the deeper
layer of the dark muscle fibers. In these pink fibers there is always a slight amount of
intracellular fat. This is a special case, the significance of which will be discussed in
the chapter on the mechanism of fat transference in the salmon body. (Page 127.)

Normal fat content of the trunk dark muscle.—The trunk dark muscle is described on
page 78 as characterized by an enormous loading of fat.

The storage fat is both inter- and intramuscular. It is present between the fibers
in a relatively small number of medium-sized drops. These drops vary in size in the
adult salmon from 5 to 20 » in diameter, and are sometimes larger. This fat is most
abundant in the immediate neighborhood of blood vessels. In longitudinal preparations
it is seen not to be uniformly distributed along the length of the fibers.

The peculiar characteristic of the superficial lateral or dark muscle is its storage of
enormous quantities of intramuscular fat. The fat is distributed within the fiber in two
general relations. First, in the region between the sarcolemma and the substance of the
muscle fiber proper, especially in the young fish (fig.7, pl.v). It often happens that there is
almost a complete ring of fat droplets surrounding the fiber and pushing the sarcolemma
out and away from the fiber wall. Ina paraffin preparation there will bea series of vacu-
oles under the sarcolemma, where the fat is extracted. Sometimes these fat drops have
grown so large that they have fused or run together into larger masses of fat, but usually
the droplets are smaller and remain separated. In the maximum loading these droplets
are from 4 to 6 » in diameter. This sarcolemmal fat is not uniformly present in all
regions of the muscle, and in regions where it is absent the sarcolemma is in close approxi-
mation to the outer wall of the muscle fiber as usual.

Second, the intramuscular fat is present in large quantities buried within the substance
of the sarcoplasm. Especially favorable points for deposit are the angles formed by

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 83

Cohnheim’s areas. In these locations very large drops, comparatively speaking, are
present. They are usually quite uniformly distributed over the surface of the fiber as seen
in across section. The intracellular droplets vary in diameter from 3 to6 y. Beside the
larger droplets there are always numerous smaller ones of varying sizes down to a fraction
of a micron. Medium to small droplets may be present in close relation to the larger,
all more or less evenly distributed among the larger droplets. The smallest droplets
are of liposomic size and are deposited in shorter or longer chains between the fibrille or
groups of fibrillea. There is evidence that these liposomes? are arranged with reference to
the striations of the fibrille, and it is suggested that such relation is of significance in
reference to the function of the fats in the muscle.

Occasionally I have found an enormous fat drop filling up the whole central
portion of the dark muscle fiber, the protoplasm of the fiber forming a band-like ring
around the drop (fig. 14, pl. 1x). Even in these cases the protoplasmic ring is closely
studded with smaller fat droplets, in one case as many as 19 droplets 2 to 4 » in diameter
being crowded within the circumference of this protoplasmic ring.

The superficial lateral muscle begins receiving its excess of fat early in the develop-
ment of the fish, at least as early as the fingerling stage. In this respect the muscle is in
marked contrast to the deep lateral muscle, in which there is little or no deposit of inter-
muscular fat until a considerably later stage in the development of the salmon and no
intramuscular fat until maturity.

The main points which characterize the fat disposal in the normal adult trunk
muscles may be summarized as follows:

Summary of jat disposal in the normal muscular tissues.—1. The fat is most heavily
stored in the superficial lateral muscle, where it is present in enormous quantity both
between the muscle fibers and within the fibers. This tissue is heavily loaded with fat
at a very early stage, at least by the 7-centimeter stage, and is always found loaded in
the feeding fish.

2. The great pink muscle contains little or no fat between the fibers in the fingerling
stage, but it has a small amount of such fat in the small Monterey Bay fish. The amount
of this intercellular fat increases to its maximum at the time when feeding ceases. The
intermuscular fat observed at Ilwaco is relatively high. While it is probably less than
at the time of cessation of feeding, it is certainly more than at Monterey Bay.

3. There is no intracellular fat in the pink muscle during the feeding stage, or, at
most, a trace of liposomes in the smallest pink fibers. The liposomic fat makes its
appearance after the fast begins. An exception is found in the superficial zone next the
dark muscle.

4. The fat in the fin and the head muscles is relatively insignificant in amount.
It is both inter- and intracellular in its relations to the muscle fibers. ;

a Various terms have been used to designate the microscopic fat droplets or fat-like droplets. They were first described by
Kélliker as interstitial granules. This was before their fatty nature was sufficiently wellknown. In fact, Kolliker thought they
were not true fat droplets. ‘The term liposome was introduced by Albrecht to describe those interstitial granules of muscle which
are demonstrated by the scarlet red stain. Bell has used the term “‘interstitial granules,’’ but he considers the granules that take
the scarlet red stain as used in his paper as fat bodies to which the term “‘liposomes’’ is applicable. (For historical discussion of
the subject see Bell, Internationale Monatsschrift fiir Anatomie und Physiologie, bd. xxvm, p. 297; also Anatomical Record, vol.
4, P- 199.) Theterm liposome is used in the present report to indicate the microscopic fatty bodies staining with scarlet red and
of small size, usually under 3 y», that take the characteristic scarlet red stain. Itisnot intended to carry any meaning suggestive
of the chemical character as regards the specific kind of fat, though it is the opinion of the writer that neutral fats are the ones
dealt with in the salmon tissue described in this paper.

84 BULLETIN OF THE BUREAU OF FISHERIES.

5. There is a considerable store of adipose fat in the myocommata, in the adipose
tissue around the small longitudinal muscles in the mid-dorsal and mid-ventral lines,
and in the connective tissue of the skin. A slight amount of fat in the viscera should be

mentioned.
PROTOCOLS.

Male salmon (no. 97), length 25.7 cm., taken at Baird, Cal., July 18, ror.

This young salmon was caught with hook and line with salmon-egg bait from the deep pool opposite
the fish hatchery at Baird, Cal., July 18, 1911. It was a relatively large summer fish derived from the
last fall’s hatch, as shown by the scale marks kindly identified for me by Prof. Charles H. Gilbert. It
was kept alive in a fish can for seven days, after which it was killed and examined for fat. On examina-
tion it was found that the testes were well developed, almost mature, and white in appearance. Speci-
mens of the alimentary tract and of the musculature were fixed in formalin for fat staining. Also
samples were preserved for paraffin sectioning.

Microscopic examination of the trunk muscle for stainable fat.—Samples of the lateral muscle preserved
in ro per cent formalin were prepared after five months. Freezing microtome sections were stained for
fat with alkaline alcoholic scarlet red and counterstained with hematoxylin. The fat was present
in the trunk dark muscle in large amounts and had not been extracted in any appreciable amount by
the long immersion in formalin. As the glycerin mounts were beginning to clear there was a stage of
very sharp and distinct differentiation. The fat droplets in the body of the muscle were surrounded
or at least partially surrounded with rings of fibrille.¢ The clear, brilliant scarlet red of the fat drop-
lets contrasted sharply with the palisade-like bands or rings of fibrilla. The sarcolemma at this stage
of clearing made a clearly marked line inclosing fat droplets between it and the fibrillar areas. These
latter fat drops are distinctly outside the areas of fibrilla, yet some of them press slightly into the
interfibrillar spaces. There is not much of this displacement of fibrillz for the reason that the superfi-
cial area of the salmon muscle fiber is bounded by a continuous band of fibrille.

The fibrille are strap-shaped, i. e., their outlines in cross section are rod-shaped. The fibrillz are
set with their flat sides approximating each other and their narrow edges, therefore, bordering on the
surface of the fiber in the case of the superficial area. It is this that gives the palisade-like arrangement
in the superficial coat of the muscle. The continuity of the superficial band is occasionally slightly
interrupted, since the rows of fibrilla as seen in cross section here and there turn in toward the central
portion of the fiber. Where such turns come there is a slightly greater quantity of sarcoplasm present.

Those dark muscle fibers nearest the skin seem more loaded with fat, although the whole layer is
rather uniform in its loading. The striking thing about the material from this fish is the amount of fat
which is under the sarcolemma. In general, the fat droplets in this region are fairly uniform in size and
are spherical. But often a mass of fat seems compressed and spreads somewhat around the surface of
the fiber. Numerous instances are seen in which such masses of fat extend around one-sixth to one-
third the circumference of the fiber. If one gets a view of such a fiber isolated from the mass this type
of fat droplet or group of droplets stands out like a great blister on the side of the fiber. These droplets
are evidently compressed by the pressure of the sarcolemma. They no doubt exist within that sheath
under a certain amount of tension.

The fat droplets within the substance of the fiber vary extremely in size and shape; the average
of the larger drops is about 4.5 to 6 p.

Through a typical section four striking variations from the general picture appear. In each of
these an enormous fat drop has formed in the center of the muscle fiber. One of these fat drops measures
18 4 in diameter, while the fiber containing it measures 33 in diameter. The thinnest part of the mus-
cular ring is 4. and the thickest 8». Evidently in this instance an enormous fat drop has formed in
the center of the fiber and crowded out the muscular substance into a superficial ring of protoplasm.
In this case the ring of protoplasm is filled as full of fat drops as plums in a pudding. There are thirteen
such droplets from 2 to 4 in thickness. There is not so much fat as usual between the sarcolemma and
the muscle substance. Four such fat drops are to be counted in one locality. In another region of the
section a fat cavity in the fiber measures 24 4 in thickness. The ring of protoplasm around it is not so
thick as in the preceding instance, and the fat drop itself has been pushed to one side, though it is still
adherent to the section. Smaller drops of fat are present in the ring of protoplasm. In yet another

aGreene, C. W.: A new type of fat-storing muscle in the salmon, Oncorhynchus tschawytscha. American Journal of
Anatomy, vol. 13, 1912, fig. 1, pl. 1.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 85

region are two adjacent fibers, each containing an extra large fat drop in the center of the substance of
the fiber. One of these drops measures 15 , in diameter, the other 20 nin diameter. Here, also, the sur-
rounding muscle substance is loaded down with the usual type of small fat droplets. One can not assume
that the large drops of fat arise at the expense of formation of the smaller. Rather is it indicated that
these drops are the result of a most active fat storing in this fish at the time it was collected.

The trunk pink muscle is free of fat in the main body of the muscle. There was neither fat between
the fibers nor within the fibers. This statement does not hold for a thin layer of pink muscle lying
just under the dark muscle. In this intermediate zone the pink muscles show a certain amount of
intracellular fat. These fat droplets are never as great in size as in the dark muscle, but are largest in
those fibers lying near the dark muscle layer. Passing from fiber to fiber in the direction away from the
dark muscle, the amount of intramuscular fat rapidly decreases. ‘This zone is, on an average, only five
or six fibers thick. It underlies the whole extent of the thicker portion of the dark muscle.

Microscopic examination of paraffin sections.—These transverse sections were especially fine as giving
a negative picture of the fat in the musculature. The sections show a thin membrane or sheath around
the dark fibers, the sarcolemma. ‘The interest attaches to the location and relations of this sarcolemma
with reference to the substance of the fiber. The sarcolemma is in contact with the sarcoplasm for a
portion of its extent round the fibers, but is distinctly separated from it in most of its circumference.
The picture is as if the membrane were pushed out and away from the fiber. The space between the
sarcolemma and the proper substance of the muscle is subdivided by very delicate strands extending
across the intervening space and continuous with the interfibrillar substance. The form of the spaces,
their size, and arrangement, all strongly support the interpretation that these spaces are filled with
fat in the fresh condition. They are the cavities left when the fat drops are dissolved out, the fat that
in the frozen section is so much more difficult to determine as regards its exact relation to the sarcolemmal
sheath (fig. 7, pl. v).

The central portion of the muscle fiber presents numerous clear areas around which the fibrille
are arranged in irregular circles. Where such a group of fibrilla is unbroken, they usually stand with
their broad dimension radial to the center of the clear area. However, there is no particular
uniformity about the matter. This arrangement is best shown in figure 7, which is a camera-lucida
outline under an oil-immersion lens. The larger angles formed in these whirls of fibrille are more
or less filled with irregularly arranged and smaller fibrille. Between the fibers and forming a slight
border along the rows of fibrille is the sarcoplasm. In most instances this sarcoplasm is sufficient
in quantity to form a very thin sheet surrounding the clear areas already mentioned. The sarco-
plasm can usually be distinguished as an extremely thin sheet around the most superficial fibrille.
It is connected by delicate strands here and there with the sarcolemma.

The pink trunk muscle of these sections exhibits the great variation in size of fibers noted in the
frozen sections. The ends of the fibrilla are very distinct and clear. They are not broad and strap-
shaped, as in the dark muscle, but are more thread-like and smaller. In the deeper portion of the
pink muscle there is no evidence of interfibrillar spaces. In the intermediate zone of pink fibers,
located just under the dark muscle, the fibers are more or less marked by clear spaces. These areas
are relatively large and more numerous in the pink fibers lying nearest the dark and decrease in
number and size in those fibers further away. Some of the pink fibers show irregular groups of small
spaces just under the sarcolemma. In the smaller pink fibers the spaces are more numerous in the
center of the fiber. The arrangement of the transparent spaces within the fibers and between the
fibrillar portion of the muscle and the sarcolemma corresponds with the distribution of the fat drop-
lets in the fibers of the intermediate zone, as shown by the scarlet-red staining.

Salmon (no. 75 and no. 76) collected at Monterey July 24, 1911, length too cm. (estimated).

Microscopic observation of the trunk pink muscle transverse section, oil-immersion lens.—The material
was studied after three days’ fixation in formalin. The striking picture is that of the intermuscular fat,
which is present in large quantity. The fat drops vary in size from 3 » up to 45 » in diameter, the
smaller drops being very numerous. The fat is far greater in amount than in the young specimen
(no. 97) from Baird, on the McCloud River.

This section is well fixed by its three days’ immersion in formalin. It shows a splendid picture
of the fibrillar structure. The muscle fibers are without intracellular fat, or, at best, have only a
trace. The large and most of the medium-sized fibers are perfectly free of fat. There are a few of the

86 BULLETIN OF THE BUREAU OF FISHERIES.

smaller fibers and occasionally a medium-sized one which show a trace of fat around the superficial
ring of protoplasm. Such fibers are surrounded with fat droplets massed on the surface of the fibers
in the connective tissue. Some droplets are undoubtedly under the sarcolemma. It is this fat which
gives the show of color at the superficial coat of fibrilla. In the smallest fibers of the section some
scattered liposomes of minute size are present between the fibrille of the surface of the fiber.

A longitudinal section of pink-trunk muscle (slide H81) shows numerous fat droplets of com-
paratively small size adherent to the surface of the fibers. The sarcoplasm shows the striations in
splendid contrast, but no liposomes were to be found within the fiber.

The intermediate zone of pink fibers.—The line of separation between the pink and the dark trunk
muscle is marked by a connective tissue septum. Occasionally a microscopic group of small fibers
may be found on the dark-muscle side of the septum (sec. H8z). These intermediate pink-muscle
fibers have in their protoplasm a few liposomes, which are limited to the small fibers. There is not
so broad a zone of intermediate fibers as was noted in the young muscle—for example, protocol no. 97.
Well out in the field of pink fibers of section H82 there is an abundance of intercellular fat, but no
evidence of intracellular fat.

Notwithstanding these exceptions, the general picture is that of muscle free from intracellular fat.

Microscopic examination of the trunk dark muscle of fish no. 75 (section H70).—Observation with
one-twelfth oil immersion. The section shows an abundance of fat both between the fibers and within
the fibers. The fat between the fibers is in droplets from 6 to ro in diameter. The muscle fibers
themselves are only from 25 to 50 # in diameter and somewhat irregular in outline. The fat droplets
are rather uniformly distributed among these fibers, though not so great in amount as in the same
type of muscle from Ilwaco.

The intramuscular fat is present in large amount and very uniformly distributed through the
protoplasm of the fibers. The droplets, strictly within the fiber, vary around 4 » in diameter.

It is difficult to determine whether the fat droplets around the superficial zone are under the
sarcolemma, and therefore intracellular, or lie outside this membrane. Certainly in a number of
cases the former is the fact. In comparison with the dark muscle of the younger fish it is noted that
the intracellular droplets of the Monterey muscle are more uniformly distributed through the pro-
toplasm and have a more uniform size.

The intermuscular fat of the dark muscle of fish no. 76 (sec. H73) is in relatively large drops, 304
on an average in diameter; but there is only a small proportion of the intercellular fat present in
the finer droplets. The intracellular fat is rather uniformly distributed through the sarcoplasm; but
the droplets are smaller than in fish no. 75, 2 y, or slightly larger, in diameter.

VARIATION OF THE AMOUNT OF FAT IN THE SALMON DURING THE SPAWNING
MIGRATION.

It is to be expected that the amount of fat present in different portions of the
musculature of the salmon will vary sharply at different times during the migration.
Whether this variation will be directly proportional to the time since the migration
began remains to be discovered. The attempt in this paper is to present the normal
distribution of fats at the end of the feeding period—i. e., the beginning of the migration
phase—and to follow the variations through four typical regions of the Columbia River
Basin. The regions chosen represent (1) the tidewater stage of the migration, (2) an
early intermediate stage in the migration, (3) a later intermediate stage, and (4) the
condition at the spawning ground and at the time of death.

As representing an early stage I have chosen a station at Ilwaco on the Washington
side of the mouth of the Columbia River. At this point P. J. McGowan & Sons have
a canning establishment, the lowest on the river. It is in the midst of the Bakers Bay
field of traps and is the most accessible point to the lowest channel fishing done on the
Columbia River.

For the second stage Warrendale, Oreg., about 6 miles below the cascades in the
canyon of the Columbia, was chosen. The upriver cannery of P. J. McGowan & Sons

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 87

is located here. The region is accessible to fisheries which depend upon the catch of
salmon below the cascades of the Columbia. The samples of this series were chosen
from the fisheries at the seining grounds on the Washington side about 11% miles below
Warrendale.

The third stage was chosen at the Frank A. Seufert’s fishery at The Dalles, on the
Columbia. The fish wheels and seining grounds along the course of the Columbia
below the Celilo Falls furnish splendid opportunity for salmon which have run the
lower mountain course of the Columbia River through the cascades and through the
lower portion of the rapids of The Dalles.

The spawning-ground stage was that on the Clackamas River, Cazadero, Oreg. A
United States fishery is located here. This is the most accessible, in fact, the only point
where spawning salmon can be had during the time of the year in which the field work
was done.

DISTRIBUTION OF THE FATS OF THE SALMON MUSCLE AT TIDEWATER.

At the mouth of the Columbia River the salmon have already ceased feeding and
the muscles have begun to show the first stages of change in the amount and distribu-
tion of the fats. This change is readily detected in the pink muscle, though not so in
the dark muscle. In the dark muscle the amount of fat is so great that one has no
adequate microscopic comparisons for showing the variations. But it is easy to con-
vince one on general comparisons that the storage of fat is even as great in amount
as when the salmon first cease to feed, as they do at some considerable time before this
locality is reached in the migration journey.

Trunk pink musclé.—In the trunk pink muscle the most striking change consists in
the fact of the appearance of intramuscular fat not noted previous to this stage. This
seems to be one of the first histological evidences of the cessation of feeding. At this
time the central core of the pink muscle fibers, and especially of the smaller fibers, is
dotted through with extremely small fat droplets. These fat droplets are rarely as
much as 2 » in diameter, usually not more than 1», and from this size down to droplets
so small as to be scarcely visible by the 1/12 oil immersion. All evidence that I have
points to the fact that this microscopic salmon fat reacts uniformly to the Herxheimer
stain whether the droplets be large or small. The pink muscle fat at this stage is quite
evenly distributed through the cross section of a fiber except in the outer circle of
fibrilla. In this circle there is no intramuscular fat. This gives the fibers the appear-
ance of having a clear surface border as distinguished from the inner portion of the
fiber, which is of course slightly pink from the presence of stained fat. At this stage I
can distinguish a few small and scattered fat droplets between the sarcolemma and the
muscle substance. The intramuscular liposomes are largest in the smallest pink fibers,
usually from two to three times greater in diameter on the average than in the very
large fibers.

The trunk pink fibers show the details of liposome arrangement best in teased
preparations. The liposomes are in short chains consisting of a few individual droplets
in each. At this stage the liposomes in the middle of the chain are largest and they
decrease quite uniformly from the middle toward each end. These chains are loaded
in the interfibrillar spaces. They are present only in certain, not all, spaces between
groups of fibrille. The number of such spaces occupied by the chains of liposomes,

88 BULLETIN OF THE BUREAU OF FISHERIES.

therefore the relative number of liposomes, varies in preparations from different indi-
vidual salmon. The amount of fat in the pink muscle fibers is measured therefore by
two microscopic factors; first, the number of chains in a given mass of fiber; second,
the size of the individual liposomes in the chains.

The pink muscle fibers vary within a wide range of size of fiber, from 25 to 250 p
in diameter. This variation is illustrated in the figure 17, plate x. In the larger fibers
of the Ilwaco fish (notably no. 111 and no. 118, the latter of which is figured in figs. 8
and 9g, pl. v1) the chains of liposomes are quite evenly distributed throughout the
mass of the fiber. However, they are characterized by the relatively small number
of liposomes in the chains and the comparatively small size of the liposomes. In the
large fibers of no. 118 the largest liposomes in the centers of the chains are about 0.5
in diameter and the smallest ones which form the ends of the chains are just identifiable
with the oil immersion. In the smaller fibers of this same fish the liposomic chains are
somewhat larger, the largest liposomes in the chains about double the diameter of the
largest in the large fibers. The liposomes in the small fibers are more thickly dis-
tributed around the central core of the fiber. This variation is not noted in the rela-
tively large mass of the fibers whose diameters run over 200 1.

Fish no. 117 seems an exception to the group from the Ilwaco station. It is cer-
tainly very far below the average of the other specimens as regards the amount of fat
revealed by the microscope. Reference to the protocol will show that this fish came
from a trap some little distance up Bakers Bay. The whole appearance of the salmon,
both its gross appearance and the microscopic appearance, suggests the type of fish
characterized by a certain degree of retrogression. The weight is much below the
standard for the length, as much below the average as certain farther advanced salmon
taken from stations higher up the river. These comparisons lead to the deduction
that salmon no. 117 has been in fresh water some time. Although it has not gone up
the river, the probability is that it has undergone as much migratory change in fats as
specimens that have gone farther up the river. The chemical quantitative determina-
tion of the fats abundantly confirms the above deductions. (See page 92.)

In this salmon the amount of intermuscular fat in the trunk pink muscle is very much
reduced. The number of fat drops is less and the size smaller. The intramuscular fat
is present in all of the trunk pink, but the number of liposomic chains and the size of
the liposomes themselves is reduced. In the very largest fibers there is almost no
liposomie fat. Another point associated with the amount of fat is the decrease in the
intermuscular spaces, so that the fibers themselves seem more compact in arrangement.

Caudal pink muscle-—The pink muscle from the caudal peduncle in each Ilwaco
specimen examined has a strikingly smaller quantity of intermuscular fat than the
muscle from the middle of the body of the same animal. It would seem from the
Ilwaco fish that the intermuscular fat is never laid down in the caudal region in as
great quantity as in the lateral or trunk region of the body. The fat drops are rela-
tively smaller and, in general, fewer in number than from the fatter region of the body.
The intramuscular fat of the caudal pink muscle in the specimens from the Ilwaco
station is less than in the trunk pink muscle. Those conditions which at the beginning
of the migration lead to an infiltration of fat into the muscle cells do not result in as
great a deposit in the caudal pink muscle as in the trunk fibers.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 89

In the fattest fish the caudal pink muscle is characterized by the smaller size of
the liposomes; also by a smaller number of liposomes in the center of the fiber. This
gives the impression of a somewhat greater quantity of fats around the superficial
layer of the fiber.

In those Ilwaco fish which have less fat than the average, for example, no. TI7,
the amount of fat in the caudal pink muscle is very strikingly less than in the middle
of the body. In this particular fish the fat between the fibers is very noticeably less
in quantity; in fact, it is practically absent except in those areas which have a relatively
large amount of connective tissue. The caudal intramuscular fat of no. 117 has almost
disappeared, or at least is present in extremely small amount.

The caudal pink muscle also shows that fish no. 117 has already passed well into
the retrogressive stage which comes with the migration fast, an indication noted in
connection with the discussion of the trunk pink muscle. This is apparent from the
character and arrangement, particularly the arrangement, of the fat in the connective
tissue, as well as in the spaces between the muscle fibers. It is certainly true that the
amount of fat present sharply increases as one proceeds through successive segments
from the caudal peduncle to the mid-lateral region of the body. Light on the signifi-
cance of the above observations is had by considering the condition of the fat in the
dark muscle of the two regions.

Trunk dark muscle.—The dark muscle forms a distinct type of muscle, as previously
announced. In this case the fat has been loaded into the muscle in enormous quan-
tities, both intermuscular and intramuscular. At the Ilwaco station the amount of
fat in the dark muscle is enormous, as illustrated by fishes no. 111, 113, 115, 116, and
118, in all of which the fat deposited in the dark muscle has reached its maximum.

The intermuscular fat is relatively much less than in the pink muscle. This is due
among other things to a structural factor. The muscle fibers are very compactly
arranged, forming a much denser mass than is formed by the pink fibers. ‘The inter-
stitial connective tissue is correspondingly reduced in mass, hence there is not so much
fat carried. On the other hand, the muscle substance has received so great a deposit
of intramuscular fat that one must regard this muscle as a definite fat depot. Atten-
tion has already been called to the fact that deposit in this muscle begins in embryonic
life. It increases in amount up to the time of the cessation of feeding and, we assume,
has not appreciably changed when the good-conditioned fish reach Ilwaco. The trunk
dark muscle contains so much fat in the muscle substance that one can not make
adequate comparisons showing slight variations.

When this tissue is examined in teased preparations, so that a side view is had of
an individual fiber, it is found that the fat droplets are so large and so numerous that
the fibers are difficult to distinguish as individuals. This is shown in figure 1, plate
Ill, where a transparency is figured of a fiber from fish no. 115. Often in the exam-
ination of these teased fibers one notes elongated fat drops or rods. These have formed
in the interfibrillar spaces owing to the fact that the droplets have increased so much
in size that adjacent ones have run together, thus fusing into the mass noted.

Caudal dark muscle.—The superficial or dark muscle from the caudal region in all
these Ilwaco specimens has a very considerably less amount of fat than the correspond-
ing muscle from the lateral region. Even fish no. 113, which is as fat as any in the
series, presents a sharp contrast as regards the comparison of the amount of fat in the

go BULLETIN OF THE BUREAU OF FISHERIES.

caudal dark and in the trunk dark muscle. The size of the intermuscular fat droplets
has sharply decreased, though the number of droplets is as great or even greater than
in the lateral region.

The sharpest contrast lies in the intramuscular fat. In the caudal region this fat
is very markedly less, especially in the size of the larger droplets. Even in the
fatter fish the larger droplets seem to be congregated around the superficial border of the
fibers. The superficial fat droplets are under the sarcolemma, though this fact is often
very difficult to determine. There is a large supply of the finer fat droplets and lipo-
somes scattered through the protoplasm of the caudal muscle substance. The contrast
between this and the fatter regions is not so much a matter of the number of the lipo-
somes as in the size and arrangement, especially of the larger droplets. The largest
droplets in the caudal muscle will not average more than one-half as great in diameter
as in the trunk muscle.

In a salmon like no. 117, which is poor in the general amount of fat of the body,
the caudal dark muscle presents the sharpest contrast in comparison with the standard
of this station. Under the low magnification, sections of the caudal dark show that in
the regions bordering along the blood vessels there are areas which by contrast with
other portions of the section are relatively free of fat. These areas are faded. This is
a condition undoubtedly indicative of the removal of stored fat. The retrogressive
process has already gone so far that one can distinguish the regions in which the active
process of fat resorption is going on with most vigor. This is the first clear-cut picture
showing the process of fat resorption. The appearance of the section is exactly the reverse
of that shown for the dark muscle in the growing stage, also of that in certain pathological
processes wherein fat is being very rapidly laid down.? In discussing later stages it is
argued that these contrasts are due to the irregularity of resorption of the fat from the
tissue. In other words, the fat is being taken up from the tissues and transported to
other parts of the body, to be utilized by the body in the construction of new tissues
(egg yolk, etc.) or in the production of energy. This movement is best facilitated in the
neighborhood of the small blood vessels, and is expressed microscopically by these con-
trasts in fat content. These facts are in further confirmation of the deduction that fish
no. 117 has been for some time on the migration phase of its life cycle.

Fat in the fin muscles.—A few examinations were made of the small muscles of the
fins at the Ilwaco station. The samples selected were the pairs of erector and depressor
muscles located in a single interspace between two interhemal spines. These muscles
are made up of fibers rather loosely bound together. There is a small amotint of inter-
fibrous connective tissue with a tolerably thick sheath around each muscle slip. The
fibers themselves are of a type somewhat like the cheek muscle of the head.

The intermuscular fat is present in droplets of good size, but not in very large num-
bers. In the connective tissue sheaths around the muscles the amount of fat corre-
sponds more nearly to that of the myocommata of the lateral muscle. In general, the
amount of intermuscular fat is considerably lower than that of the pink lateral muscle
of the same salmon.

@ My colleague, Dr. W. J. Calvert, tells me that in his unpublished work on the plague he often noted a striking deposit
of fat in the parenchyma along the immediate border of the blood vessels of the liver. This deposit in the early stages of the
disease extends out only a short distance into the parenchyma of the liver. The course of the smaller blood vessels is easily fol-
lowed througn the parenchymatou: tissue by the bordering deposit of fat. This is, of course, the reverse picture of that described
above. In the salmon the fat is in process of removal; in the plague liver the fat is in process of rapid deposit, but in each case
the histological picture is that of the early, therefore differential, stage in the process.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. gI

The intramuscular fat is relatively small in amount. Many of the fibers show
liposomes of extremely fine size, often so small that one can trace them with difficulty.
There are no rings of fat droplets under the sarcolemma such as characterize muscle that
is beginning to show fat exhaustion. There are certain groups of fibers in these sections
which show a relatively larger amount of intracellular fat. In such fibers the liposomes
will average as much as 2 » in diameter. The liposomes are quite evenly distributed
through the cross section of the fiber and are occasionally quite numerous under the
sarcolemma. This latter type of fiber is suggestive of the dark muscle type. Not
enough comparative work has been done in studying these muscles to determine whether
or not the dark muscle fibers are present in portions of these muscles. There is some
indication that the superficial muscle of the anal fin, the inclinator analis, contains fibers
of the dark type, whereas the erector and depressor muscles are more nearly of the pink
type. If the inclinator contains fibers of the dark type it would suggest that that
muscle is more nearly homologous with the supertficalis lateralis, a homology that needs
further investigation.

Fat of the adductor mandibule or cheek muscle.—The fibers of the muscle are more
compactly arranged and different in appearance from the other portions of the salmon
musculature. They are, however, most like the great lateral pink muscle. At Ilwaco
the intermuscular fat is distributed in scattered but relatively large fat droplets, 60 to
7o in diameter. There is also a comparatively large number of small droplets not over
20 «in diameter.

The intracellular fat is always present. The large fibers in the muscle carry a few
scattered chains of extremely minute liposomes. On the other hand, the smallest fibers
have liposomes about 0.6 y in diameter.

Considering the muscle as a whole at the Ilwaco station the fat distribution is most
nearly like that of the great lateral pink muscle, though both the intermuscular and
intramuscular fat is very much less in quantity. This muscle, like the fin muscles, carries
a relatively small amount of intramuscular fat. This fat is more than adequate for the
uses of the muscle, but the striking fact shown by the sections is that there is never an
excessive accumulation of the fat.

ANALYTICAL DETERMINATIONS OF THE PERCENTAGE OF FATS IN SALMON FROM THE
MOUTH OF THE COLUMBIA RIVER.

When this study was projected it was planned to take a full set of samples of the
muscles studied and make fat determinations by accurate chemical methods. Such a
full set of determinations would have been very valuable in itself but of inestimable
value as corroborative evidence in connection with the microscopic comparisions. It
turned out to be impossible to carry through the full program of the work and the sacri-
fice fell on the chemical series. Chemical samples were taken, however, whenever it
could be done, though the analyses were reserved to be made not in the field but in the
home laboratories. The few samples secured were not analysed until after the micro-
scopic work was completed and the results sent off for publication.

The fat determinations secured on samples from Ilwaco are inserted at this point.
Considering the fact that the eight Ilwaco fishes were chosen to represent the entire
range of types present in the lower Columbia at the time of the expedition, this showing
of fat percentages is most significant. The salmon were taken, first, from the main

G2 BULLETIN OF THE BUREAU OF FISHERIES.

channel as far out toward the end of the jetty as the fishermen go (111, 112); second,
from the main channel south of Sand Island (113, 114, 115, 116); third, from the north
channel leading out of Bakers Bay at a point near Fort Canby (118); and fourth, from
Bakers Bay at the Whitcomb trap (117).

The two chief types of muscle described, the pink muscle and the dark muscle,
were the only ones selected for analyses. The samples were taken in the mid-lateral
region just in the plane that cuts the front of the dorsal fin, the same region from which
histological samples came. ;

The greater amount of fat in the lateral dark muscle as compared with the pink
was revealed by the microscope. But this fact is even more strikingly shown by the
quantitative percentages given in the table below. A glance shows that the percentage
of fat in the dark muscle is roughly twice as great as in the pink. There is no law to be
deduced about it from so few samples. The fattest salmon have relatively the highest
quantity in the pink muscle. The intermediate salmon from this station have a greater
reduction in the fat of the pink than in the dark.

Particular attention is directed to the two females, no. 112 and 117. The
former is from the channel of the Columbia from the farthest point out toward sea.
The latter is from Bakers Bay, quite out of the main channel of the river. Undoubtedly
the great difference is due to the fact that no. 112 was just coming in from the sea.
No. 117 had undoubtedly lost most of its fat and is quite comparable with the salmon
in better condition from the spawning grounds of the Clackamas River at Cazadero.

TaBLe I.—ANALYTICAL DETERMINATIONS OF FAT IN THE TISSUES OF CERTAIN SALMON OF THE
IQII SERIES.

Muscle fats in per cent
¥ of wet weight.
Date. re | Caught in Colmabia River, Ilwaco,
fish. Pink Dark |
muscle. muscle.
IQII.
Aug. 3 1d 15-680 28.018 | Outer channel opposite the end of
the jetty.
Aug. 3 1129 19-655 30. 813 0.
Aug. 4 1132 T4566!) |Meat | Mid-channel, off the upper end of
Sand Island.
Aug. 4 1149 20-179 29-080 Do.
Aug. 7 10. 062 27-420 | Do.
Aug. 8 1169 5-395 23-248 Do.
Aug. 10 1179 2.727 14-324 Bakers Bay, Whitcomb trap.
Aug. 11 118d IOs 507) | |ldsaicines cette North channel (from Bakers Bay),
| McGowan’'s trap near Fort |
| Canby.
} 1 |

Significance of the fat in Ilwaco salmon with reference to the normal quantity of fat at
the beginning of the migration.—In discussing the normal salmon type, the type at the
beginning.of the migration deduced from the study of feeding salmon secured at Monterey
Bay, it was suggested that one might arrive at a better conception of the normal type by
figuring back from the first migration station. But Ilwaco fish show a number of signs of
physiological change presumably due to the cessation of feeding. Among these changes
the most striking are to be had by the examination of the alimentary tract, where, in
the stomach, and especially in the intestine and ceca, one finds extensive evidences of

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 93

degeneration. These changes are almost wholly in the direction of retrogression of
structure.

In the muscular tissue there are two factors mentioned above which are interpreted
as changes that have come on in the amount and distribution of fat in the muscle since
the beginning of the migration fast. These are, first, the abundance of intracellular
fat laid down in the pink-muscle fibers; second, the evidence in the dark-muscle fibers
of removal of fat. Interpreting these two phenomena broadly one may assume that
there has occurred already in the best conditioned channel fish a using up of a certain
amount of fat. Considered from the standpoint of percentages this amount has not
reduced the total storage enough to be readily measured by the microscope except in
the second case. In the poorer fish it is quite obvious that fat is disappearing, undoubt-
edly due to the prolonged fast. This is especially shown in the Bakers Bay type illus-
trated by no. 117. This interpretation is supported by the quantitative chemical
determinations of fat.

There is absolutely no microscopic evidence which can be legitimately interpreted
as meaning a fatty production from the disintegration of protoplasm. If, therefore,
one could follow back the physiological condition of an Ilwaco salmon to that point
in its history where it first ceased to feed, and would examine its tissues for the loading
of fat, he would find that this time represented the maximum amount of fat present in
the animal. In other words, this stage of the beginning of the fast, i. e., end of the
feeding period, represents the climax of fat storage during the salmon’s history. The
microscopic picture obtained by a study of the Iwaco specimens is therefore applicable
to this normal stage, provided, first, that the intramuscular fat of the pink muscle be
omitted, and, second, that all the areas of dark muscle which appear to be losing fat
be considered as uniformly filled with fat. Figure 8, plate v1, representing the fat in a
cross section of pink muscle of salmon no. 118 from Iwaco, would, if the intramuscular
fat were eliminated and the intermuscular fat increased in quantity, represent very well
my conception of the quantity of fat in this tissue when the fast begins. So also in
the dark muscle, figure 3, plate 1, would serve as a type for the dark muscle at the
beginning of the fast. These figures fail in the fact that they have too little inter-
muscular fat to represent the normal, but the percentage difference is one which can not
easily be estimated by the microscope. Salmon no. 114 has nearly twice the amount
of fat in the pink muscle shown by no. 118. This fat is wholly intermuscular and would
show in the microscope in the form of larger drops rather than in a greater number.
Judging wholly by the microscopic comparisons, one would never judge that the difference
is as great as that revealed by the chemical determinations of the fat percentages.

PROTOCOLS.

Male salmon (no. 111) length 950 mm., weight 13,776 grams, taken between the jetty and the black buoy at
the mouth of the Columbia River, August 3, IgIt.

This was a clean, bright, silvery salmon of the short, deep type. It isa perfect looking specimen
oftheseatype. It was caught with a gill net by Mr. Cliff Sweeney. Thissalmon had all the appearance
of a first-class, very fat specimen. Its flesh looked oily and there was a thick layer of cutaneous fat.

Microscopic examination of trunk pink muscle, teased (slide J6).—These isolated fibers of pink muscle
are simply crowded with liposomic fat. The fat is arranged in longitudinal rows or chains of liposomes
between the fibrilla which bear relation to the striations. The liposomes are from o.2 y or less to 2p

19371°—vol 33—15——7

4 BULLETIN OF THE BUREAU OF FISHERIES.

in diameter, rarely larger, as observed in a certain large fiber under examination. This fiber is 200 uw
in diameter. The liposomic chains are comparatively uniform in their disposition throughout the
mass of the fiber. The liposomes themselves are not uniform in diameter in the rows. Adjacent droplets
may alternate between small and large sizes, though in some of the rows the droplets are fused, thus
making an oval droplet extending across the intervening striation membrane. In some of the smaller
muscle fibers the fusions are much more extensive, extending over four or five striations. In the smaller
fibers the fat droplets are relatively larger, averaging between 1.5 and 2 4in diameter. Over the surface
of the fibers and under the sarcolemma there is a sprinkling of small fat droplets from 2 to 5 in diameter.
These are irregularly placed.

The caudal pink muscle was not prepared in this fish.

Trunk dark muscle (longitudinal section, ]1).—The preparation is so filled with fat that the structure
isobscured. The intermuscular fat is in the largest drops observed for the trunk dark muscle, the average
diameter being about 30 p. These drops are often compressed into oval outlines by the pressure of
the fibers. The muscle fibers themselves are only about 4o » in diameter.

There are large quantities of intramuscular fat, the droplets being simply crowded throughout the
whole structure. The larger intramuscular droplets are from 15 to 20,in diameter. These large droplets
often appear in rows along the course of the fiber, giving the appearance of splitting the fiber. However,
they only press the fibrilla apart. There are relatively few of the smallest liposomes present, though
the bundles of fibrillz all show a certain number of these small liposomes. The quantity of fat in this
preparation is the greatest for any dark muscle noted in the whole season’s work. The droplets are
larger, relatively more numerous, and have so distorted the relations of the fibrillz as to break up the
regularity of the structure.

Another section (J33), fixed 18 hours in formalin, gives a much better view of the outlines of the
fibers. The fibers are crowded thickly with relatively large intracellular droplets. They are so numer-
ous as to form almost a continuous layer of drops. A section (J35) stained in sudan shows the same
crowding of fat as those stained with scarlet red. The contrasts are less sharp.

The myocommata of the trunk dark muscle are filled with adipose cells which are crowded with
fat. Many of the cells have ruptured and the fat has run together, but the fat drops of those cells still
intact measure from 50 to 70 # in diameter.

Caudal dark muscle (transverse section, J20).—The dark muscle of the caudal peduncle is very fat,
but not so fat as in the trunk muscle. The drops are relatively smaller. Those between the fibers
are from 6to15 in diameter. The fat within the fibers varies extremely in different parts of the section.
I notice one region in which the fibers are almost free of large drops of fat; only smaller liposomes are
present. In the near neighborhood of this group the fat is gathered around the surface of the fibers,
apparently just under the sarcolemma, where the drops vary from 5» down. The centers of these fibers
have liposomes averaging only about 1 » in diameter. In that portion of the section which is fattest
the central portion of the fibers has larger droplets, not averaging more than 3 1, however. From this
section it seems that the caudal dark muscle must have a greatly reduced amount of fat in comparison with
the trunk region. Slide J21 shows relatively more fat than slide J2o. The fibers are cut somewhat
obliquely, and this brings out the fact that the drops are oval in shape, as in the lateral line region.

Intercostal muscle (longitudinal section, J7).—This section has a large amount of intermuscular fat.
There are numerous large drops averaging 30 ». The connective tissue of the whole section is jotted
fullof very fine fat droplets, from1toroy. There isa trace only of intramuscular fat, nothing comparable
to that in the teased trunk muscle. This fat is in extremely fine liposomes, averaging only a fraction
of a micron in diameter. It seems quite uniformly distributed throughout the substance of the fibers.

Muscles of the anal fin (transverse section, ]23 and 3r).—This section was across the group of muscles
between the interhemal spines and should therefore be of the erector and depressor muscles. There is
a small group of fibers on the outer margin of the section different from the main body, which probably
belongs to the inclinator muscle of the fin.

There is a very small quantity of intermuscular fat. The drops are scattered but relatively large.
The main portion of the muscle has only traces of liposomic fat in extremely fine granules. There are
no rings of fat droplets under the sarcolemma of the type which characterizes fat-exhausted muscle.

The group of fibers on the outer margin of the section has a uniform distribution of intracellular fat
in comparatively large liposomes. These liposomes average 2 4 in diameter. They are quite evenly
distributed throughout the substance of the fiber and under the sarcolemma, where they are somewhat

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 95

more numerous. ‘The liposomes appear as rows running between the fibrillaee which the oil immersion
lens shows have an arrangement with reference to the cross striations. In one fiber six striation
segments have a length of 8 », an average of 1.3 « per striation. The liposomes that are spaced with
reference to these particular striations are from 0.4 to 0.6 » in diameter. The arrangement of fat in
this group of fibers is very like the arrangement in the dark muscle of the lateralis superficialis.

Masseter or cheek muscle (transverse section, J9).—The fibers of the cheek muscle vary considerably
in diameter, running from 30 up to r10 4. This transverse section has a medium amount of intermus-
cular fat distributed in large droplets from 60 to 70 4 in diameter. There are also numerous fat drops
from 15 to 20 » in diameter.

The muscle fibers themselves are much split up by ice crystals in sectioning, yet it is clear that the
fibers contain intracellular fat. This fat is greatest in amount in the smallest fibers, where the droplets
are about o.6 in diameter. The larger fibers also contain intramuscular fat, but the droplets are smaller.

Male salmon (no. 115), length 940 mm., weight 12,225 grams, caught in the Columbia River channel just
above Sand Island, August 7, IgIt.

This salmon was a clean fish, free of sea lice; testes slightly developed and dark venous red in
appearance; stomach relaxed, 5 cm. in diameter with mucous content. Dark muscle teased imme-
diately in physiological saline and figured.

Microscopic examination of trunk pink muscle teased in physiological saline.—These muscle fibers show
extremely fine liposomes within the fiber. In larger fibers the liposomes are difficult to distinguish,
but are readily seen in the small ones. The amount of intramuscular fat is not so great as in the trunk
pink muscle of no. 113. Section Jog, stained with hematoxylin only, differentiates the interfibrillar
sarcoplasm in such a way as greatly to emphasize the outlines of transparent liposomes. The liposomes
themselves appear highly refractive and have not taken more than a trace of the stain (1/12 oil immersion).

Microscopic examination of trunk dark muscle teased in physiological saline (]46).—A drawing of a
fiber from this slide is presented (fig. 1) showing the fat throughout the dark muscle. Practically all the
fibers in this slide are loaded down with fat. There is an enormous quantity of fat present, more than one
can adequately represent by any graphic method. The fat drops within the fiber are relatively large
and are so numerous that they push out the sarcolemma, making its outlines irregular. The drops are
somewhat oval in shape, measuring 8 by 13, 5 by 6, 7 by 16, and smaller. The diameters of some of
the fibers are 35, 36, 40, and 45 4. ‘The fat droplets are in rows. They are relatively large in almost
all parts of the field. This is due to the fact that the liposomes have grown in size until adjacent ones
have fused, a condition throughout the fiber.

With the fusion of droplets the resultant is an oval mass with the long axis with the interfibrillar
space. As the fat mass has grown the fibrils have been forced out of their normal relations. Where the
drops lie outside the sarcoplasm and under the sarcolemma this membrane is seen to be pushed out in
numerous irregular protuberances. This section is unusually clear and transparent, probably because
it was not subjected to formalin. .

Female salmon (no. 116), length 975 mm., weight 14,530 grams, caught in the channel of the Columbia River
opposite the lower end of Sand Island, August 8, grt.

A bright silvery fish, no sea lice, stomach small and contracted with thick walls; intestine one-half
as large as in no. 115; ovaries relatively large, weighing 965 grams.

Microscopic examination of the trunk dark muscle (slides ]77-9r).—The trunk dark muscle of this fish
has less fat than either no. 111 or no. 115. The larger fat drops are between the fibers. They measure
12 to 14 #, but the average is not much over 7 #1.

The intracellular fat is in smaller droplets, from a fraction of a micron to 3 and 4 » in diameter.
There is a massing of the fat droplets around the surface of the fibers.

A series of teased preparations were treated in various ways to test the method The fresh muscle
teased in physiological saline is more transparent than the other preparations and the fat gives the
appearance of a greater quantity, largely because it is more clearly distinguished. Fibers teased in
formalin were very opaque. Those teased in alcohol were somewhat intermediate in character between
the saline and the formalin preparations.

96 BULLETIN OF THE BUREAU OF FISHERIES.

Female salmon (no. 117), length 940 mm., weight 8,245 grams, taken from the Whitcomb trap located in the
bend of Bakers Bay.

This salmon was more slender than no. 111, wasa clean fish, but did not appear in as prime condition.
The ovaries weighed 510 grams. ‘The flesh appeared less oily and was very pale in color, especially in
the caudal peduncle.

Microscopic examination of the trunk pink muscle (transverse section, K31).—This section is taken
ventral to the lateral-line septum. There is very little intermuscular fat. Even along the myocommata
there are only a few droplets and these are of small size, from 1 to ro 2 in diameter. The very largest
drop seen was only 15 in diameter.

The substance of the muscle fiber contains a supply of liposomes arranged in chains throughout
the mass of the fiber. These liposomes run from 0.3 to 0.5 » indiameter. ‘The section is cut obliquely,
making it difficult to interpret the point, yet it is obvious that the liposomes are in greater numbers
along the superficial region of the fibers while the deeper portion of the fibers is relatively poor in lipo-
somes. ‘This gives the section asa whole a mosaic-like appearance. The point is not absolutely certain,
yet I am convinced that this increased quantity of liposomic fat is under the sarcolemma and between
the more superficial layer of fibrils. c

A section taken from the dorsal division of the deep lateral muscle has a less amount of fat than that
from the ventral. The fibers of this section are very compactly arranged and the outlines are corre-
spondingly sharp and angular, similar to that shown in later stages. (See salmon nos. 125 and 126.) The
diameters of the fibers themselves vary between 30 and 150 u, rather smaller than the average pink
muscle fibers.

There is a very small quantity of very fine liposomes within the substance of the fibers. In many
fibers no liposomes are to be distinguished. In the smallest fibers shown in the field, those 30 p in
diameter, the liposomes are present in the middle of the fiber. Ia the medium-sized fibers the lipo-
somes are largely around the superficial area of the fiber and are from o.2 to 0.3 #: in diameter, rarely
larger (1/12 oil immersion). Inthe middle of these medium-sized fibers and in most of the body of the
large fibers, liposomes are much fewer and exceedingly small, scarcely discernible.

The smallest fibers often have rings of very small fat droplets, the droplets running from 2 to 3 4
in diameter. These droplets are just under the sarcolemma. In areas where they are more numer-
ous there are occasional fat drops 15 # in diameter located between the muscle fibers.

Microscopic examination of the caudal pink muscle (transverse sections, K1g and 20).—These sections
of caudal muscle show a markedly less amount of fat than from the dark muscle. The type of arrange-
ment is that of the trunk muscle except that there is less intermuscular fat.

The intramuscular fat is also conspicuously less in quantity. The liposomes are practically absent
from the larger fibers and are very minute and few in numbers in the smaller fibers.

Microscopic examination of the trunk dark muscle (transverse section, K7).—The fat droplets are rela-
tively small in this preparation, notably smaller than in salmon no. 115. ‘The intracellular fat is evenly
scattered through the substance of the dark fibers. The liposomes vary from 2 to 5 » in diameter (tissue
cut fresh). In the regions which have the smallest amount of fat the number of droplets of the larger
size which are so prominent in fish no. rrz are greatly reduced in size, running from 1 to 5 # in diam-
eter. In certain areas of the section lying near the connective tissue partitions there are small groups
of dark fibers which apparently have their fat reduced greatly below the average. Such fibers will
contain irregularly placed liposomes from 1 «down to a just visible size, 0.2 to 0.3, “, while adjacent
fibers will have a more prominent loading of fat in which the droplets average from 3 to 5 # in diameter.
Also the number of droplets in the latter fibers is greater than in the former. This picture suggests
the thought that the fat of the dark muscle is being removed along the course of the larger blood vessels.

Microscopic examination of the caudal dark muscle.—Insufficient study was made of the caudal dark
muscle, but the rather poor sections bring out one point, namely, that the fat is reduced much below
the average and that the fat droplets are massed around the surface of the fiber.

Male salmon (no. 118), length 940 mm., weight 12,470 grams, taken in McGowan & Co.’s trap at the
mouth of Bakers Bay near the Fort Canby Dock.

This fish was a deep smooth specimen, skin bright and clean, no sea lice, head shaped like the
female, medium depth, a splendid specimen apparently comparable to no. 111. The testes were quite
small and immature.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 97

Microscopic examination of the trunk pink muscle (sections K35, 45, K55-58).—The intermuscular
fat is crowded in every angle formed by groups of fibers. The drops vary in size from small ones to
as high as 100 # in diameter. The muscle fibers themselves vary greatly in size, from 50 to 300 p in
diameter. In cross section the muscle fibers are oval to round in outline, the round contour of the
individual fibers being in sharp contrast to the polygonal shaped outlines of fibers of salmon no. 117.

The two sections on slide K4s5 were made free-hand and thick in order to show the relations of
the intermuscular fat. The section includes a tendinous myocomma. Where the muscle fibers are
very close together a single row of large fat drops extends down the length of the fiber from the myo-
comma. In two or three regions the intermuscular space is filled up with two or even more rows
of such fat drops. These fat drops are from 50 to 60 » in diameter. They are somewhat compressed,
having their long axis in the longitudinal axis of the fibers. The myocomma itself is crowded with fat.

The intramuscular fat is present in all the fibers. It consists of extremely fine liposomes, being
most minute in the large muscle fibers and greatest in amount in the small fibers. They are uniformly
distributed throughout all of the body of the muscle fiber, with the exception of the narrow ring of
band-shaped fibrils which forms the surface layer.

The chains of liposomes are rather evenly distributed throughout the substance of the large fibers,
but consist of very small liposomes from those just identifiable up to 0.3 4. In the medium-sized fibers
the liposomes are somewhat larger and in the smaller fibers considerably larger than in the ones just
described. In the latter the liposomes reach the diameter of 1.5 », though the average is less than 14.
In one fiber 56 # in diameter the liposomes were in unusually long chains and large in size, similar to the
arrangement in dark muscle at a late stage in the resorption. Several liposomes in this muscle were
measured which were 2 # in diameter, but the average was from 1 to 1.2 4. Figure 8, plate v1, shows the
distribution of fat in the truak pink of salmon no. 118.

Pink muscle from the belly shows an even greater amount of intermuscular fat; also minute
liposomes in chains throughout the substance of the fibers.

Microscopic examination of the trunk dark muscle (K4r and 42, transverse sections).—There is a large
amount of intramuscular fat in the lateral dark muscle of fish no. 118. The size of the fat droplets in
this region is from 9 to 12 #in diameter. Ina certain interseptal region the fat drops are large, running
as much as 60 2 in diameter. This fat belongs to the adipose tissue proper. A noticeable difference in
the staining character is present between it and the fat in general; i.e., the large fats are lessred. The
muscle fibers of this section are so compact that it is often difficult to determine whether a given fat drop
is within the sarcolemma or without. It is judged that a rather large proportion of the fat which is massed
around the surface of the fiber is under the sarcolemma. The section throughout its whole extent shows
an enormous quantity of fat massed along the lines which separate the fibers.

The teased dark muscle (slide K52) shows numbers of relatively large fat droplets along the sides
of the fiber wall and adherent to the protoplasm. These droplets are smaller on the average than those
of the cross section which were judged to be intermuscular.

Within the dark muscle fibers of this teased material the fat is present in masses—no other word
seems adequately to express the condition. There are numerous fibers, in fact nearly all of them, in
which many chains of liposomes are displaced by long masses or rods of fat. Undoubtedly, these rods
of fat have been produced by the fusion of liposomes in the loading of the fiber with a higher percentage
of fat than is found when liposomes are typically present, as, for example, in salmon no. 132. In the
present section there are four such rods in one microscopic field. In another similar field there are six.
In a fiber 40 # in diameter these rods continue unbroken for as much as 1264. They are located in the
areas between the bundles of fibrilla, where one finds in the ordinary loading either chains of liposomes
or, at most, short oblong droplets.

There are fibers in this teased material that have a somewhat less quantity of fat than that described
in the last paragraph. In one such typical case the smallest liposomes observed measure 1.5 to 2.5 p.
In close proximity to this last chain of liposomes there is a chain of fused liposomes, i. e., a rod, which is
continuous for 1204. This rod has, however, several partial constrictions which undoubtedly represent
points where in the earlier stage of fat deposit the rod is discontinuous.

In the transverse section (1/12 oil immersion) the fat is crowded into the fiber in a way comparable
only tono. 111. The whole surface of the field is taken up with fat droplets almost as thick as they can
stand. There are relatively few liposomes that measure less than 1.8 » in diameter and the size varies

uptosy. There are chains of these smaller liposomes throughout the protoplasm, even in fibers obviously
distorted by the long rods of fat.

.

95 BULLETIN OF THE BUREAU OF FISHERIES.

DISTRIBUTION OF THE FATS AT AN EARLY INTERMEDIATE STAGE OF THE SPAWNING
MIGRATION.

The first station above Ilwaco where salmon were collected was at Warrendale,
Oreg. This station is about 6 miles below Cascade Locks and is in the midst of an exten-
sive fishing field. Salmon taken here have not yet passed the swifter runs of the river,
but have already made a run of about 135 miles above the mouth of the river. The:
station was located at the cannery of P. J. McGowan & Sons, and I am particularly
indebted to the superintendent, Mr. Charles A. McGowan, for many special favors.
This company has a seining ground on the sand bar on the Washington side about 1%
miles below Warrendale. Our specimens were chiefly taken from this point, as the
fish captured there were fine conditioned channel fish.

Fish nos. 120, 121, 122, 125, and 126 were taken at this station during the month
of August. In August one secures salmon which clearly show stages of the removal
of fat from storage localities. There is at this time of the year considerable variation
in the grade of fish at this point. The fatter salmon, for example, no. 120, have their
tissues well loaded with a reserve of fat. The poorer salmon, no. 126, show marked
stages indicative of retrogression as regards the loading of fat.

Trunk pink muscle.—There is wide variation in the microscopic appearance of the
fats in the trunk pink muscle of the fishes at this station. The fattest observed was
no. 120 and the poorest no. 126.

The intermuscular fat is disposed in the muscle according to the same general plan
as in salmon from Ilwaco. However, there is a very great diminution in the amount
of this fat. This is indicated by the decrease in the size of the larger droplets, and toa
less degree in a decrease in the number of droplets. A striking fact in comparison is
that in these Warrendale fish the fat is very much less uniformly distributed among
the fibers than in either the Ilwaco or in the normal tissue. The comparison between
two stations is difficult to make. One can not microscopically measure the number
of fat droplets and compute their diameters and thus the volume of material from the two
stations. Rather he is limited to impressions made by placing the slides side by side.
It is largely on this type of evidence that the above comparisons are made. However,
as regards the intermuscular fat drops, a comparison of the diameters of the largest
drops is illuminating. At Warrendale these largest drops seldom measure over 50 4
in diameter (see the protocol for salmon no. 121), whereas at Ilwaco they often measure
100 # and more in diameter. In observations made at the time the material was col-
lected and sectioned on the grounds at Warrendale it was judged that the amount of
intermuscular fat in fish no. 120 was about one-half to two-thirds as great as in fish
no. 118 from Ilwaco. Each of these fishes is among the best represented at its station.
In those fishes which were relatively poor in fat, as no. 126, the amount of intermuscular
fat is very greatly reduced. The amount of this reduction is best shown by comparing
figures 8 and 10 of plates vrand vu. Judging from the comparison of a large number
of preparations, I would say that the intermuscular fat of this poorest salmon could not
be above 25 to 30 per cent of that of the normal type.

The intramuscular fat is abundant in all of the fibers of the pink muscle from the ;
trunk region. The smaller fibers as usual are more heavily loaded with fat than are the
larger. This is shown chiefly by the larger size of the liposomes in these small fibers.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 99

In the fatter specimen the largest fibers are relatively thickly filled with numerous
chains of liposomes. These chains are more numerous than in fish no. 115 and no. 118
from Ilwaco, and the size of the liposomes in the chains is, if anything, comparatively
greater. When the poorest fish are examined, it is found that the large fibers are strik-
ingly low in fat, for example, no. 126, fig. 10, pl. vi. In fact, it is difficult to distinguish
liposomes in the largest trunk pink fibers of this fish. In numerous instances observed
there were tiny groups of very small liposomes ranged near the surface of the cell, chiefly
under the sarcolemma. If liposomes were present in the body of the large fibers at all,
they were too small to be distinguished with the 1/12 oil immersion. In many sections
of this poor fish the intermediate-sized fibers had their liposomes chiefly at the surface,
whereas the central portion of the fiber was comparatively free of liposomes. Disappear-
ance of fat is not accompanied by any signs of degeneration at this stage. The struc-
tural detail is clearer and very sharp and distinct, as shown in figure 13, plate vu.

In teased preparations where one has a view of a fiber for some considerable length
it appears that the Warrendale pink muscle is relatively rich in liposomic fat. In the
best salmon there is even a greater amount of intracellular fat in the pink muscle than
at the Ilwaco station. The chains of liposomes are more continuous and the size of the
individual liposomes relatively greater. In the small fibers particularly this comparison
holds. In fact, it often happens that in the smallest fibers the liposomes have reached
a size at which adjacent ones coalesce, a phenomenon the significance of which is dis-
cussed in another connection.

While the above comparison is true and striking it is also true that at this station
the range of variation in the amount of liposomic fat in the pink muscle is far greater
than at Ilwaco. The fattest muscles have a greater amount of intracellular fat, the
poorer muscles have a much smaller amount of intracellular fat.

Caudal pink muscle-—Vhe caudal pink muscle of salmon from the Warrendale
station shows the sharpest contrast as regards the amount and arrangement of the fat.
In salmon no. 120 the intermuscular fat is all gone except along the connective
tissue septa where it is present in scattered but medium-sized drops. In the poorer
salmon the amount of intermuscular fat in the caudal pink is practically nil. Here and
there in the thicker septa between bundles of fibers one will find an individual droplet
or a group of three or four droplets not more than 4 or 5 # in diameter.

The intramuscular fat of the caudal pink muscle is very slight indeed even in the
fattest fish. The smallest fibers are fairly well supplied with liposomes which run in
chains comparatively evenly distributed throughout the sarcoplasm. In these instances,
however, there are distinct groups of liposomes under the sarcolemma, but at the surface
of the sarcoplasm. There is a distinct difference in size between the surface liposomes
and the deep ones. The former range from 1 to 1.5 # in diameter, while the latter are
only from o.2 too.4 4 in diameter in fish no. 120. In salmon no. 126 the liposomes are
still present in the small fibers, having much the same arrangement as that just described
and averaging about 0.4 in diameter.

In the intermediate and in the larger sized fibers the amount of intracellular fat is
very small. In the larger fibers only an occasional group of very tiny liposomes at the
surface of the fiber can be seen. In the intermediate fibers there are now and then
fibers which have a comparatively even sprinkling of tiniest liposomes throughout the
mass of the protoplasm with somewhat larger liposomes at the surface of the fibers. On

100 BULLETIN OF THE BUREAU OF FISHERIES.

an average for the station, however, one must say that the presence of liposomes is very
greatly reduced, both in size and number for all the intermediate fibers, while for the
larger fibers it is present only in traces.

Trunk dark muscle.—In the dark muscle of salmon from the Warrendale station
there is even wider variation as regards the loading of fat than in the pink muscle. In
fish nos. 120 and 121 the amount of fat in the trunk dark muscle is very great, while in
no. 125 it is low. In the fatter salmon the loading of fat is almost as great as in the
specimens from Ilwaco, with the exception of Iwaco specimen no. 111 which was an
extraordinarily fat fish. On the other hand, in the poorer specimens the amount of dark
muscle fat is only a small percentage of that at the Ilwaco station.

The intermuscular fat is comparatively plentiful, is located in the connective tissue
septa and in the myocommata. However, the fat droplets average much smaller in
size than in the Ilwaco specimens. Oftentimes the number of these fat droplets, espe-
cially of the smaller ones, seems relatively greater. In Warrendale fish the individual
fibers are usually somewhat more definitely separated and this fact makes it easier to
determine the relation of the intermuscular fat. In fish no. 125 the amount of this
intermuscular fat is very low, but occasionally individual drops are as large in this fish
as in those that have more fat. The amount of intermuscular fat varies in different
regions of one and the same muscle. This variation undoubtedly is associated with a
process of fat erosion which was first observed in certain waco specimens. In Warren-
dale fish the erosion process has gone much further and is more readily followed. In
areas in which the fat has been most fully eliminated the intermuscular fat is reduced
to tiny droplets.

The intramuscular fat of the dark trunk muscle is abundant in all of the fibers of the
fatter fish. In no. 120 the cross sections and the teased preparations show that the
fibers are especially richly supplied with liposomes in their sarcoplasm. The liposomes
are of large size and in relatively long chains. There is considerable fusion of adjacent
liposomes. Especially in fish no. 121 the liposomes are so large that one might better
describe them as droplets. The diameters run from 1 to as much as 4. Certain of the
fibers in this fish and also in fish no. 122 show fusion of liposomes into long rods of fat.
These slender rods usually appear more or less constricted at points corresponding to the
striations of fibrille.

The most striking thing about the fat in the trunk dark muscle ox fish from the War-
rendale station is its great irregularity in different microscopie areas. This has been
spoken of in connection with the very fat fish no. 120, but it isan appearance that marks
every fish examined. If the specimen is one of low grade, as in no. 125, then these irregu-
larities are most prominent. Certain groups of dark muscle fibers will appear richly
loaded with fat while other areas will be almost free, certainly will not contain more than
from 30 to 50 per cent as much as in the fatter areas. In these clear areas the reduction
in fat is due to two factors: First, the great reduction in the average size of the liposomes;
and second, the great decrease in the number of liposomes. In numerous areas where
muscle fibers are in close contact with small blood vessels the fat is very low in amount.
This condition is described in the protocol of fish no. 125. The characteristic picture
presented where a group of fibers lies along the blood vessel is as follows: First, that por-
tion of the fiber next the blood vessel will have no intermuscular fat; second, the intra-
muscular fat will be either absent or greatly reduced in the corresponding area; third,

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. IOI

the portion of the fiber opposite the blood vessel will have a relatively high content of fat;
fourth, as a rule the intermuscular fat in contact with the opposite outer border of the
fiber will still be present.

The teased fibers of this dark muscle show instructive variations. Different lengths
of one and the same fiber show wide variation in the loading of fat. This is expressed
especially through the variation in size of the liposomes. But if the fat is very light
in amount there will also be a variation in the number of liposomes. Careful focusing
will always bring out the fact that the richer portions of the fiber will have a relatively
large amount of fat under the sarcolemma. In many instances the liposomes in the
chains will have fused, forming slender fat rods showing constrictions at the point of
fusion. ‘The poorer areas in the fiber will show a small amount of fat under the sarco-
lemma, smaller liposomes in the chains and little or no fusion. Where the fat is almost
completely eroded the number of liposomes will be obviously reduced. In this case the
reduction takes place more completely near the center of the fiber as compared with its
superficial area.

In numerous instances at the other stations, as well as at Warrendale, I have
noticed that while the fat is being eroded there will appear variations in number and
arrangement of fat droplets under the sarcolemma. In a teased fiber a rather definite
pattern will often be noted in this subsarcolemmal fat, a pattern which coincides with
the blood vessels, the pattern being marked by rows of very small droplets along what
would correspond to the border of the capillaries. Also small rings of droplets will
appear at various points, sometimes in groups. ‘These rings of droplets are arranged
around a clear center. They are interpreted as part of the process of erosion of large
intermuscular fat drops. As lipolysis goes on, the fat that is dissolved away from the
large fat drop will often be redisposed in small droplets within the sarcolemma around
the area which is being compressed by the large drop.

Caudal dark muscle—The caudal dark muscle of the Warrendale fish varies through
even a wider range of fat content than the corresponding muscle from the trunk region.
There is always considerable fat in the myocommata, but the amount of intermus-
cular and intramuscular fat varies exceedingly.

In the fatter fish the intermuscular fat is reduced in the number of droplets present,
but particularly in their size. In no. 125 there is practically no fat in the caudal dark
muscle.

In this salmon certain of the dark fibers are absolutely clear of fat within the
fibers, and the fattest fibers observed contained only a sprinkling of liposomes around
the superficial areas with a trace in the center of the fiber. The whole muscle is as
nearly fat free as any dark muscle examined. It is noticed here also, as in the trunk
muscle, that the fibers freest of fat lie in the neighborhood of small blood vessels.

In a few scattered fibers in the caudal dark muscle of no. 125 an appearance is
noted for the first time that is suggestive of a disintegrative process. We have not been
able to convince ounselves that these fibers are actively breaking down, but they cer-
tainly do show appearances suggestive of the initial stages of water absorption charac-
teristic, for example, of cloudy swelling. The fibers stain lightly in a way which
characterizes an early stage of muscle degeneration. These fibers also contain small
transparent, highly refractive and lightly stained granules. The stain does not have
the usual appearance of fat stain—that is, the color is not the brick red of the scarlet

102 BULLETIN OF THE BUREAU OF FISHERIES.

red dye. Rather it is a more brilliant and dark appearing neutral red. The amount
of stain taken is only slight. These granules do not contain any pigment, as was
noted in degenerating cheek muscle of fish no. 140, to be described later.

PROTOCOLS.
Male, salmon (no. 120), length 937 mm., weight 11,480 grams, Warrendale, August 16, 1911.

This fish was taken from the McGowan seining grounds, 114 miles below Warrendale. It was a fine
fish, in splendid condition; the nose slightly hooked, no large teeth, the testes two-thirds developed,
color normal, but a trace darker than fish at the mouth of the river; back darker but not rusty; fins
perfect.

The muscles were pink and oily. The fish was received fresh from the seining grounds, and the fin
muscles were still alive when samples were taken.

Microscopic examination of the trunk pink muscle (K57, 88, and go).—The intermuscular fat is about
one-half to two-thirds as great as in no. 118 from Ilwaco. Its disposal between the fibers is similar to
the fish taken from the mouth of the Columbia. ‘There is less intermuscular fat from the middle of
the dorsal portion of the great lateral muscle.

The intramuscular fat is abundantly present in all of the fibers. The smaller fibers are more deeply
stained, showing the greatest amount of fat. The small fibers of the teased preparation are filled with
chains of liposomes, the individual liposomes being larger than in the large fibers. In the fibers of large
size the chains of liposomes are not quite so numerous, and the liposomes themselves are relatively
small. Two fibers, side by side, one large and the other small, are in sharp contrast.

Microscopic examination of the caudal pink muscle (transverse section, Kgr).—In this section the inter-
muscular fat is all gone except along the connective tissue septa, where it is present in scattered but
medium large drops (1/12 oil immersion). The substance of these fibers is well fixed in formalin, and
the fibrillar outlines show clearly. In the large fibers of the section there is no fat stain in the body of the
fibers. Occasionally at the very surface there are tiny groups of liposomes. In the smallest fibers there
are in the body of the fibers between the fibrilla numerous extremely small liposomes. ‘There are dis-
tinct masses of liposomes on the surface of the sarcoplasm and under the sarcolemma. The liposomes
within the fiber are from 0.2 to 0.4 in diameter, those on the surface from 1 to 1.5 “in diameter. The
intermediate-sized muscle fibers have a few scattered groups of liposomes immediately under the
sarcolemma, but none in the body of the fiber. These observations are confirmed on fragments of
fibers in which the fibrilla are turned in a horizontal position.

Microscopic examination of the trunk dark muscle (sections K72-76).—The muscle fibers in this
material, both in the transverse sections and in the teased preparations, are especially richly supplied
with fat. The fat is crowded, both between the fibers and throughout the sarcoplasm of the fibers.
The intermuscular fat droplets are numerous, of medium size, but not so numerous nor so large as in
fish no. 111 from the Ilwaco station.

Certain areas in the transverse section have an appreciably smaller quantity of fat. These areas
are associated with connective tissue septa carrying blood vessels, and are similar to those noted in
salmon no. 117 and no. 118, from the mouth of the Columbia. This appearance is undoubtedly
due to the beginning of fat erosion from this type of muscle, and is greater in this section than in the
two Ilwaco fish referred to. The erosion areas have a much less quantity of fat than in fish no. 118.
The fat droplets are not so numerous and are smaller.

On the whole, the amount of fat is somewhat less than in fish no. 118, though the comparison is
difficult tomake. In the transverse section of one fiber roo in diameter, 46 droplets were counted.
They were from 3 to 6 in diameter. In the spaces around the particular fiber and in the same focal
field were 12 droplets oval in shape, 20 » long, but from 4 to 6 y thick.

The intramuscular fat is remarkably uniform in its distribution through the muscle fiber, the larger
droplets averaging from 4 to 6» in diameter. The disposal of the fat is similar in character to that noted ~
in previous fish and is shown in figure 3, plate m1.

Microscopic examination of the caudal dark muscle.—The muscles in this section have very much
less fat than the trunk dark fibers. The intermuscular fat is smaller, 6 to 10 » in diameter, but the
droplets are numerous.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 103

The largest intramuscular liposomes average 3 . in diameter, but there are many smaller liposomes.
There are rings of small droplets around the border of the muscle, these averaging 4 .in diameter. This
superficial fat sharply marks the boundaries of transparent cross sections forming a definite mosaic under
the low magnification. It is almost wholly intramuscular fat lying under the sarcolemma.

Male salmon (no. 121), length 950 mm., weight not given.

A first-class fish from the McGowan seining grounds, 114 miles below Warrendale on the Columbia
River. The testes two-fifths developed.

Microscopic examination of the trunk pink muscle (transverse section, Lt).—The amount of intermuscular
fat is intermediate between salmon no. 115 and no. 117 from Ilwaco. The fat droplets between the
muscles are many of them relatively large but not sonumerous, and do not average so large as in no. 118.
The largest drops are from 45 to 55 « in diameter. The fibers themselves are somewhat more compact
in arrangement, but the outlines of the fibers in cross section are less smooth and circular than in no. 115,
but not so angular, and the fibers do not seem so much compressed as in fish no. 117.

The surprising fact is the great amount of intramuscular fat. This fat is most thickly deposited
through the small fibers, where the liposomes have a size from 1 to 2 4 in diameter. These liposomes
are quite uniformly distributed through the substance of the small fibers. An occasional fiber will
have its liposomes more thickly set around the superficial border. In certain of the smallest fibers,
an example 75 in diameter, also in other regions of the section, there is fat in relatively small droplets
just outside the surface of the sarcoplasm and under the sarcolemma.

Liposomes are present in the largest fibers also, but are exceedingly small and not so plentiful in
the body of the fiber. In these very large fibers many liposomes are found between the fibrille near
the surface of the fiber. They appear as if the liposomes were formed just under the sarcolemma and
between the fibrillz of the most superficial or band-shaped layers. In the inner borders of the band-
shaped fibrils there is a second zone where the liposomes are present in relatively greater numbers.
The liposomes are not larger but more numerous in this zone.

Microscopic examination of the caudal pink muscle (Lo, 10, and rr).—Sections were preserved for 18
hours in formalin. The fibrilla show well indeed. The surface layer of band-shaped fibrille are in
contrast to the smaller fibrille of the body of the muscle. The fibers are compact in arrangement, but
retain a certain amount of round contour which characterizes muscular tissue in prime condition. The
following points characterize the tissue: (a) There is very little, almost no intermuscular fat in the
section. Here and there a small droplet is found in the angles between the fibers. The largest one
observed is only 18 » in diameter. (b) The outlines of the fibers of the caudal pink muscle are defi-
nitely marked by very small fat droplets, measuring from 1 to 2.5 in diameter, many of them even
smaller. The point is difficult to determine, but the droplets seem to be within the sarcolemma.
(c) The caudal pink muscle fibers are relatively low in liposomes. ‘The smallest fibers contain only a
few liposomes. The fibers measuring from 50 to 100 » in diameter have easily identified liposomes,
but the larger fibers are free of liposomes in all but the extreme superficial part of the fiber (1/12 oil
immersion.) The liposomes in the smaller fibers are chiefly around the outer third of the muscle.
In the central portion of the fiber there is not more than one-fourth as much stainable fat as in this
superficial rim.

Microscopic examination of the trunk dark muscle (transverse section, L5 and 6).—The lateral dark
muscle shows an amount of fat greater than in no. 117, but not so great as in no. 111. The fibers are
compactly arranged everywhere in the dark, and under the low magnification their outlines are marked
by the excess of fat in that zone, the fat droplets averaging from 4 to 6 p.

In some of the angles between fibers and in certain regions where the connective tissue is greater there
is unquestioned intermuscular fat. The size of these drops runs from ro to 15 ” in diameter. Along
one exceptionally thick septum this intermuscular fat is absent. The section has the appearance which
indicates the process of resorption of fat (under the 1/12 oil immersion).

The sarcoplasm of these fibers is full of small fat droplets almost as large as those in no. 115. The
droplets are too large to be called liposomes, though every gradation exists between liposomes 1 y in
diameter up to these larger droplets which average 3.6 to 4 4 in diameter. There are certain fields of
the section which contain relatively less fat in the sarcoplasm, the fat droplets being almost gone and
the liposomes relatively smaller but thickly distributed. In these fields there is also relatively less fat
around the border of the fiber, i. e., less fat under the sarcolemma.

104 BULLETIN OF THE BUREAU OF FISHERIES.

Microscopic examination of the caudal dark muscle (L14 and 15).—This transverse section of the dark
muscle from the tail is quite well supplied with fat. I do not see fat droplets that are unquestionably
between the fibers, but groups of droplets that lie just under the sarcolemma were noticed, the largest
of which were 7 » in diameter.

The intramuscular fat is greatly less than that in the dark muscle of the portion of the body where
the largest liposomes measure 3.6» in diameter, but this larger size israre. The largest of the liposomes
run about 2 in diameter. The smaller are from this size down to 0.3 and less in diameter. Judging
from the intensity of the stain, one would say that the caudal dark does not contain more than one-
fourth, possibly one-third, as much fat as the trunk dark.

There are areas, especially along certain septa, which have a strikingly less quantity of fat. This
appearance is associated with the more vascular areas.

Female salmon (no. 122), length 890 mm., weight 8,980 grams, taken at Warrendale, August 17, Tgrr.

This was a good conditioned fish, taken at McGowan’s seining grounds, 114 miles below Warren-
dale. The weight of the ovaries was 680 grams, stomach quite small, appearing one-half degenerated.

Microscopic examination of the trunk pink muscle (transverse section, L38).—This section shows a rela-
tively large quantity of fat between the fibers, not so much, however, as in no. 115 and no. 118 (1/12 oil
immersion). The larger drops measure about 20 » in diameter. The distribution is similar to that in
the fish just mentioned. In this section the smallest fibers, 40 in diameter, are thickly set with lipo-
somes distributed rather uniformly through the fiber. These liposomes vary in size from 0.6 to 1.3
in diameter. In the medium fibers, 75 to 100 » in diameter, the number of liposomes diminishes in the
center of the fiber; also there is a marked diminution of the size of those present. In one fiber, 100 Le
in diameter, the liposomes in the middle of the cross section measure from o.1 to 0.4 » in diameter.

In the larger fibers of the section, those above 100 » in diameter, there is a marked diminution of
liposomes. This diminution is most apparent in the main body of the fiber, i. e., exclusive of the super-
ficial area. This contrast in amount of liposomes between the deep and superficial part of the fiber
is sharp, giving the fiber the appearance in cross section of having a superficial ring of fat. In certain
of the larger fibers, the central liposomes are absent, or, at any rate, one can not distinguish them with
the oil immersion. In these same fibers liposomes around the superficial border will vary greatly in
size, measuring from scarcely identifiable liposomes up to as much as 1.4 in diameter.

An examination of the longitudinal sections (L33), brings out the fact that there is a relatively
high content of fat near the surface of the fiber, both external to the fiber and just under the surface.
The external fat is in the connective tissue, the endomysium, therefore, intermuscular.

Microscopic examination of the caudal pink muscle (section L51).—The fibers are almost free of fat.
There is no intermuscular fat hanging to them, but the connective tissue, myocommata, of the caudal
pink muscle (slide 54) is crowded with adipose fat.

Scattered over the surface of the fiber, all apparently under the sarcolemma, is a good deal of fat
in droplets, from 2 to 3 4 in diameter, not uniform in size. The small and intermediate sized fibers have
tolerably evenly distributed chains of smallest liposomes. In the larger fibers the central portion is
relatively free of chains of liposomes, which, in many instances less than o.2 » in diameter, are so small
they are difficult to identify. Certain of these teased fibers show the striations clearly. In one such
example there are 13 striations in 36 p (slide 54, 15 striation to 36). The diameter of the fiber showing
these striations is 63 ». The chains of finest liposomes in the largest fibers are not always perfect. The
irregularity is due to the dropping out of individual liposomes. In some chains there are more liposomes
than fibrille. This is due to the presence of two liposomes in the space opposite certain striations.
There is not always perfect correspondence in the number of striations and liposomes in the chains in
these pink fibers.

This slide gives examples of the cone-like ends of the fibers. These ends are supplied with lipo-
somes just as in the body of the fiber. All the teased material of the caudal pink fibers shows an increased
quantity of fat at the surface of the fiber. This fat is in tiny droplets varying from the smaller liposomes
0.2 . in diameter up to 2.5 and even 3 4. Many of the droplets are in regular rows, but not so regular
as those deep in the muscle fiber. Undoubtedly this fat is just under the sarcolemma, a fact confirmed
by the appearance when the optical section cuts the middle of the fiber.

Microscopic examination of the trunk dark muscle (sections L1S-25).—These transverse sections show
a relatively large quantity of intermuscular fat. The fibers are more widely separated than is usual

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 105

for dark muscle, and it is correspondingly easy to determine whether the fat is free or under the sar-
colemma. The longitudinal section (L25) gives a fine confirmation of the cross section.

The intermuscular fat is present in large quantity, the larger drops averaging from 12 to 15 p in
diameter. There are many smaller droplets interspersed among the larger.

The fat under the sarcolemma shows well both in the cross sections and in teased preparations.
It is seen best when the focal plane cuts the surface of the fiber. The drops are irregularly placed over
the surface, being held in position by the delicate sarcolemma. When the focal plane cuts the center
of the fiber one can see that the sarcolemma incloses the fat drops.

The intramuscular fat in the body of the fiber is uniform in its distribution, as viewed in cross
section. The droplets average from 2 to 2.5 # in diameter, many of them smaller, but some larger.
Rarely does one see a droplet greater than 3 » in diameter.

In the longitudinal sections the chains of liposomes are more numerous and longer than in any
tissue examined. In some fibers these chains extend across a whole microscopic field. Several fibers
show chains in which the individual liposomes have fused. In such case the fat is in long slender
rods, showing constrictions corresponding to the fibrille (slide L46). In comparison with salmon
no. 111 and no. 115 from Ilwaco, this fish has as many, even more, intramuscular fat droplets, but
the droplets are relatively smaller. The larger droplets, which in the Ilwaco salmon measure as much
as 6 » in diameter, are absent here.

Microscopic examination of the caudal dark muscle, teased (L58).—These teased caudal dark fibers
show considerable variation in the amount of loading of the fat in the different fibers present (1/12 oil
immersion). One fiber contains a relatively large amount of fat on the surface of the fiber, interpreted
as under the sarcolemma. The larger fibers show but little fat in this subsarcolemmal region. , If
present at all, it is relatively small, running 4 to 5 p.

The fat within the substance of the fibers is greatly reduced in amount in comparison with the
fibers from the middle of the body. The liposomic chains are not so numerous and the average size of
the liposomes not so great. In these caudal fibers one often finds a chain of liposomes which has become
fused, asin the trunk dark. Inthe muscle fibers least filled with fat the number of chains of liposomes
is very much less and the size of the liposomes does not average over 1.2 1.

A large salmon (no. 125), length and weight not recorded, taken at Warrendale, Oreg.

Microscopic examination of the trunk pink muscle (transverse section, 1/12 oil immersion).—The amount
of intermuscular fat is relatively small, the drops are often large, as much as 50 » in diameter, but they
are few in number—not more than 1 to every 3 or 4 fibers.

The amount of fat within the fibers is obviously greater in the smaller fibers than in the larger.
In the small fibers, 50 » in diameter and less, the liposomes are fairly uniformly distributed in size,
ranging from 0.6 to 2.5 4 in diameter, but averaging about 1 in diameter. The intermediate fibers
have the fat collected around a superficial zone about 8 to 10 » beneath the surface of the fiber. Some
of the fat droplets in this region are relatively large, 4 “ in diameter, though these are comparatively
rare. In the center of the fibers the liposomes are smaller, the larger ones averaging 1 p. In a fiber
Ioo » in diameter the liposomes are quite uniformly distributed over the surface of the cross section,
but run only o.2 to 0.6 # in diameter. The liposomes in the largest fibers are of practically the same
diameter, but not so numerous. In the largest fibers examined and in those most free of liposomes
there is a noticeably greater number of liposomes near the surface of the fiber. In certain of the fibers
these liposomes are just under the sarcolemma, between the fibrille of the surface layer and in the
zone at the inner border of the superficial or palisade fibrille.

Microscopic examination of the caudal pink muscle.—The intermuscular fat is limited to the myo-
commata and to the thick connective tissue septa.

The intramuscular fat is present as liposomes in the small and intermediate fibers and just under
the sarcolemma of most of the large fibers. Liposomes are distinguished with difficulty in the central
body of the large fibers. .

Microscopic examination of the trunk dark muscle—The intermuscular fat is not so great in amount
as in no. 122, though the fat drops run up to 202 in diameter. There are areas over the section which
have practically all the intermuscular fat as well as much of the intramuscular fat removed.

106 BULLETIN OF THE BUREAU OF FISHERIES.

The fibers are thickly studded with chains of liposomes 0.6 to 2 4 in diameter. Only occasionally
are adjacent liposomes fused as in no. 121 and no. 122. The longitudinal section shows a greater
quantity of fat along the borders than in the bodies of the fibers.

The fat under the sarcolemma is in drops measuring from 4 to 5 » in diameter, occasionally 6 p.
The number of these fat droplets around any given fiber varies greatly, due to the fact that the fat is
being removed in the neighborhood. Choosing an area containing the most fat, the intramuscular fat
is in liposomes from o.4 # in diameter up to as much as 4.3 4. The number of the largest droplets is
relatively small, but when present they are evenly distributed through the fiber. There is great
variation in the size of the liposomes in different portions of the length of one and the same teased
fiber.

In the areas referred to above the intramuscular fat is reduced in amount. The size of liposomes
is affected more than the number of chains. There are several fields through which small blood vessels
go in which that portion of the muscle in contact with the blood vessel is strikingly free of fat. In the
neighborhood of a blood vessel where the fat is most removed the droplets lying under the sarcolemma
are reduced to a few liposomes lying on the side farther from the vascular area. The largest of these
liposomes measure 1.6 in diameter. Through the body of the same fiber in the half opposite the
blood vessels the liposomes are fewer in number and relatively smaller in diameter (0.4 to 1 ») than
in portions of the fiber not in contact with blood vessels. In the third of the muscle lying next to the
vascular area the liposomes are still present, but small—too small to measure accurately. There are
numerous areas in this section (L74) showing contrast as regards the degree of removal of fat.

Microscopic examination of the caudal dark muscle.—There is quite a little fat in the myocommata.
Through the body of the dark muscle, however, there is no intermuscular fat.

Under the oil immersion certain fibers of this section are absolutely clear of fat within the fiber.
In other fibers there are traces of liposomes too small to measure. These traces are confirmed by fibers
which have been tumed to a horizontal position in the handling. In still other fibers there are scattered
and irregularly placed groups of liposomes at points near the surface, but none deep down in the sub-
stance of the sarcoplasm. The fattest fibers observed contained a fairly uniform sprinkling of liposomes
around the superficial area of the fiber and a somewhat smaller quantity in the middle of the fiber. The
whole preparation presents as nearly a fat-free section as has been observed of caudal dark muscle.

A number of fibers contain numerous spherical bodies measuring approximately 2 in diameter
and having a dark-red color (1/12 oil immersion). These bodie are irregularly placed through the
substance of the fiber, as are the brown pigment granules of degeneration. The stained bodies are
spherical, and one might take them for fat bodies. However, if they are fat bodies then the color of the
stain is distinctly different from the type. This color is a brilliant dark neutral red as against the usual
lighter brick red characteristic of this stain. It is possible that we are dealing here with the reaction
of some special fat which stains differentially, according to the observations of Bell.

Female salmon (no. 126), length 780 mm., weight not taken, Warrendale, Columbia River, August 24, IQII.

This salmon was fresh from the McGowan seining ground and was chosen as representative of the
group of fish which show an advanced stage of migration change at this station.

Microscopic examination of the trunk pink muscle (transverse sections M1—3, L78).—The intermuscular
fat isin relatively small amount. The myocommata contain a small amount of fat arranged as a narrow
band of droplets on either side of the tendon. The larger fat drops measure 30 to 4o , in diameter, seldom
more. The intermuscular septa still contain a small amount of intermuscular fat. In the larger of
these sheets of connective tissue a few fat droplets measure as much as 15 # in diameter, most of them
less. These are in the areas where in fish no. 118 the droplets were as much as 1001 in diameter. Most
of the fat drops are small and in relatively small group of 9 or 10 droplets in a group.

The intramuscular fat is very low in thisspecimen. It is present in the small and intermediate
fibers, but difficult to distinguish in the larger fibers. The smallest fibers have their intracellular fat
tolerably thickly sprinkled over the microscopic field. Most of these fibers show a somewhat greater
amount of liposomes near the surface. The medium fibers show great variation. Certain ones are
almost clear of liposomes, while others have a liberal sprinkling. In muscles of this size there is a con-
densation of fine liposomes under the sarcolemma. The same arrangement is true for the smallest fibers.
These small intracellular liposomes are 0.4 to o.6 » in diameter.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 107

In the largest fibers in the section one can scarcely find any liposomes in the body of the fiber.
Around the border and immediately under the sarcolemma, especially in regions which have inter-
cellular fat in the neighborhood, there are groups of liposomes.

These groups are shown in the figure 1o, plate vir, and are characteristic. In one fiber 144 pin
diameter very delicate liposomes are rather thickly dispersed in the superficial layer and more scatter-
ing in the central portion of the fiber. This fiber is smaller than the average of the large size and has
relatively more fat.

The teased pink fibers are filled with liposomes. These are in long chains, which are quite uniform
in appearance. In the small fibers the liposomes are from 1 to 1.5 » and occasionally 2 » in diameter.
In a number of instances observed adjacent liposomes have fused into oblong droplets. In certain
isolated fibrille the liposomes are adherent but irregular in position. In a fiberabout 150” in diameter
I find that the central portion has only scattered chains of liposomes, while near the surface the chains
are more numerous. In either case the liposomes are of irregular size in the chains and not of regular
arrangement, as in the type from down river. The striations in this material are very narrow. I have
not examined carefully enough to determine the exact relation between the liposomes and the stria-
tions. The number of liposomes in a given area varies in different portions of the fiber. There are
irregular patches of fat droplets of liposomic size at the surface of the fiber. From this appearance
and that noted in the cross section one comes to the conclusion that these patches of liposomes are the
ones shown under the sarcolemma in the transverse section. The arrangement of these patches of fat
under the sarcolemma is partly dependent on some pressure factor. At any rate, there isa fairly definite
map shown by them. In some instances this may correspond with the capillary net. Also there are
numerous areas, oftentimes two or three in the same field, in which the fat is arranged in a definite
ring around a clear area. This suggests a relation to some relatively large intercellular fat drop. There
is no other structure present with which such a definite arrangement of fat drops coincide, and the num-
ber of instances observed is too great to be a mere matter of chance. This pattern-like arrangement of
fat under the sarcolemma was often noticed in preparing fresh material in the field. In one of three
fibers of a group examined in this connection there were a number of fused liposomes in the chains.
These fusions have taken place in the chains of liposomes near the surface, but are not of the subsar-
colemmal group.

Microscopic examination of the caudal pink muscle (section Mrr).—The amount of intermuscular
fat is insignificant in this section, almost exclusively limited to the myocommata. The intramuscular
fat is also very small. In the very smallest fibers there is still present quite an appreciable amount of
fat in small liposomes. These liposomes vary in size, about an average of 0.4. They are not so
numerous as in the same size of muscle fiber in the trunk region. There are small fibers which show
groups of liposomes under the sarcolemma. In the medium-sized fibers such liposomes as are present
are limited to the superficial layer of fibrilla and to the space under the sarcolemma. In the large
fibers the only trace of fat is under the sarcolemma, and that is present only in isolated regions where
the liposomes are of scarcely visible size.

Microscopic examination of the trunk dark muscle (sections M7, 16, 25).—Section M7 shows a com-
paratively slight amount of intermuscular fat. That is chiefly along the thicker connective tissue
strands. Among some of these strands which are more vascular the low magnification shows areas in
which the bordering muscle fibers are almost free of fat. The appearance suggests that the fat is in
process of removal. Under the 1/12 oil immersion it is noted that in the compact areas of the muscle
there are scattered droplets of intermuscular fat. The drops are comparatively small in size, 3-5 to 4p,
but occasionally as much as 12 » in diameter.

The intracellular fat is present in medium amount. It is distributed less uniformly over the sur-
face of the section of the fibers. It is noticeably less in amount in the center of the section of many
of the larger fibers. The diminution in the amount of the fat in the middle of the fiber is primarily
due to a great reduction in the size rather than in the number of the liposomes. In the center of a
given fiber under observation the liposomes are from 0.4 to 1“ in diameter, while at the surface of the
same fiber they are 1.6 to 2 in diameter. In this fiber there are fat drops under the sarcolemma which
measure 2.5 to 3 « in diameter.

The examination of a fiber bordering on one of the lightly stained areas mentioned above shows no
large fat droplets, and the liposomes are reduced to an average size of 0.3 » in diameter. There is a
group of liposomes about 1.2 4 in diameter under the sarcolemma of this fiber at the point farthest

,

108 BULLETIN OF THE BUREAU OF FISHERIES.

from the blood vessel. Bordering on the opposite side of the same area there are two fibers which show
a very marked reduction in the number of liposomes next the vascular area, although the liposomes
are not altogether absent. The parts of the fibers opposite the area contain larger liposomes, 1.6 to 2 » in
diameter.

There is a very great variation in the amount of fat in different parts of this section (M25, teased),
if one is to judge by the microscopic size of the liposomes. Oil immersion examination of the dark
fibers shows numerous chains of liposomes very similar in arrangement and appearance to slide M2o
of trunk pink. The liposomes are more numerous than in the trunk pink, but the average size for the
center of the fiber is about the same. Many of these fibers show fat droplets on the surface under the
sarcolemma. Certain of the fibers show that the surface of the fiber has relatively large liposomes. Ina
certain case near the superficial focus are liposomes 2 to 2.4 in diameter and near the center of the fiber
numerous smaller liposomes not over o.6 ». in diameter with an occasional larger one 1.2 #7. The loading
of liposomes varies along the length of the fiber. This might easily happen in a tissue where the fat was
being eroded, since the arrangement of blood vessels can not be uniform with reference to the surface of
the whole fiber (fig. 4, pl. rv).

Microscopic examination of the caudal dark muscle, slides M 16 and 17.—The intercellular fat has
disappeared, or is limited to a tiny droplet here and there in the connective tissue (1/12 oil immersion
examination). There are no larger drops or groups of droplets as in the trunk pink muscle. The myo-
commata still have some fat drops.

The intracellular fat is present in the dark caudal muscle, but the liposomes are extremely small in
size. There are no fibers with the larger liposomes characteristic of the normal dark muscle. The
smaller liposomes average only 0.4 to 0.8 4. In asmall.area which contains more fat, the liposomes are
larger, from 0.4 to 2 4 in diameter. These. liposomes are in a group toward one side of the fiber in an
area about 20,square. The center of the fiber has the smaller liposomes, and there is also a very marked
irregularity in the number in different parts of the field.

In the above fiber and in four others in the immediate neighborhood there are small fat droplets
under the sarcolemma, measuring 2 to 3 , but in each case these droplets are on the side opposite the
adjacent blood vessels. The liposomes throughout the central portion of the fibers in fields in which the
fat is evidently sharply removed are reduced to scarcely distinguishable size, but are comparatively
numerous. On the surface of such fibers the liposomes are about 0.6 to 0.8 4 in diameter and also nu-
merous. Certain portions of the section show the fibers turned horizontally. Liposome chains can be
distinguished in these fibers. In one such fiber the liposomes of the chains are about 0.6 to 0.8 » in
diameter. There are no fused liposomes in this case. This caudal muscle does not have more than
one-half to three-fifths the fat of the trunk muscle.

DISTRIBUTION OF THE FATS AT A LATE INTERMEDIATE STAGE OF THE SPAWNING MIGRATION.

A study was made of the amount of fat present in the tissues of salmon from the
Columbia River at the Celilo Rapids. These fish have passed through a longer stretch
of fresh water and through the relatively swift currents of the canyon of the Cascades.
The famous fishery of Mr. Frank A. Seufert extends along the full extent of The Dalles
of the Columbia.? The numerous fish wheels adapted for the different stages of the water
make it an ideal collecting ground for scientific material. At the time of the visit to the
fishery in August, 1911, active fishing was in progress on the lower Dalles, at Celilo
Falls, and at the Tumwater seining grounds. The samples that were studied came from
a point known as the Cement Wheel, also from the seining grounds at Tumwater imme-
diately below the Celilo Falls. The Cement Wheel is about 300 yards above the mouth
of the Government canal.

The Cement Wheel salmon will have battled only a short distance of the swifter
portion of the rapids of The Dalles. Two salmon were taken from this point, a male and

@ Mr. Seufert has always taken an active interest in the scientific questions concerning the propagation of the salmon and in
work tending to develop and protect the industry. He aided the present work by putting at our disposal every facility for
securing material in the best of condition. 4

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 109

female. The protocol for no. 127 is the male. By reference to the protocol it will be
seen that the trunk pink muscle contains a medium amount of intermuscular fat. The
fat droplets have a distribution that is normal, but the relative size of the droplets is
small so that the amount of fat represented is evidently greatly decreased as compared
with the standards from Iwaco.

The intramuscular fat of the trunk pink muscle was present in much lower amount
than in the fatter specimens from Warrendale but also in greater amount than in the
poorer specimens. The characteristic of these Celilo fish in this regard is in the
relatively low fat in the larger pink muscle fibers. Only scattered traces of intramuscular
fat were found in any of these fibers, and those traces usually in the neighborhood of
intermuscular fat.

Keeping in mind the condition of the dark muscle from the salmon from Warrendale
it was to be expected that the corresponding tissue in the Celilo salmon would have a
considerable loading of fat. This was found to be the case. The specimens from the
Cement Wheel agreed in that the dark muscle fat was definitely diminished in quantity
in comparison with average fish from Warrendale, and sharply diminished in comparison
with Ilwaco specimens. The samples chosen were the average of the type running at
that time, August 22 and 23. The grade of fish at this season is very much lower than
at an earlier date, but this is a factor which characterizes the entire series studied and
does not interfere with the comparison.

The caudal pink muscle was practically free of fat in each of the specimens studied.
The caudal dark muscle had both intermuscular and intramuscular fat, but in each
instance about 40 to 50 per cent as much as in the trunk muscle of the same specimen.
The female of the Cement Wheel specimens showed less fat in both the trunk and caudal
dark muscle than was present in the muscles of the male, notwithstanding the fact that
the male gave other evidences of a greater retrogression in general than did the female.

Little or no variation could be shown between the salmon taken at the Tumwater
seining grounds and those taken at the Cement Wheel. In each instance there was a
fair showing of intermuscular fat in the pink muscle, and an amount of fat within the
fibers of both the pink and dark muscle which characterize an approximately average
grade running below the Cascades.

At the time of this visit no fish were running which were judged to be of as high a
grade as no. 120 and no. 122, described in the protocols from Warrendale.

PROTOCOL.

Male salmon (no. 127), length Soo mm., taken at Seufert’s Cement Wheel, The Dalles of the Columbia,
August 22, IQII.

The fish was silvery in color, with dark dorsal surface; shape that of a half-exhausted specimen;
flesh oily in appearance and to the touch; visceral mass small and degenerated.

Microscopic examination of the trunk pink muscle, I.—The intermuscular fat is medium in quantity.
There are a few large droplets, the largest 40 » in diameter, but many smaller droplets, especially of
the size from 3 to 6 win diameter. The amount of fat is much less than in the specimens like no. 118
from Ilwaco, estimated at 4o per cent. The distribution of the intermuscular fat varies in different
parts of the preparation. In certain parts the amount is not more than 25 per cent that described above.

The intracellular fat varies extremely in the different fibers. In the small fibers, size 40 to 70 # in
diameter, the fat droplets are fairly numerous but of larger size and greater number around the surface
of the fiber. These fibers (1/12 oil immersion) show that the fat droplets aggregated around the surface
are within the sarcolemma and superficial. In the center of the fibers the liposomes are much smaller,

19371°—vol 33—15——_8

IIO BULLETIN OF THE BUREAU OF FISHERIES.

about 0.3 too.5 4. There are droplets under the sarcolemma in practically all these pink fibers. They
measure from o.8 up to 2 # in diameter. The average amount of intramuscular fat in the small fibers
is not more than 4o to 60 per cent of that found at Warrendale. Many of the intermediate-sized fibers
show only traces of fat in the center of the section. Around the circumference there is a somewhat
greater quantity of fat, especially just at the surface. The largest fibers in the section have irregular
outlines, look compressed and are very clear of fat, at least these fibers do not contain fat that stains
in the usual way. I notice an occasional small group of liposomes at some point on the surface of the
section, though these groups are few, often not present at all.

An appearance that is difficult to interpret is due to the presence of very small highly refractive
granules in the protoplasm of the large fibers. These granules do not take stain, at least, if they are
stained at all, it is very different in appearance from the normal, and they do not appear to be uniformly
present in all the large fibers.

Teased muscle (M38 and 39) shows a comparatively small amount of fat. The smallest fibers and
some of the intermediate fibers have chains of liposomes. The chains in the small fibers are not so
numerous as one usually finds. The liposomes themselves are very small, and the picture is one of low
content of fat. In the intermediate fibers the chains of liposomes are more numerous near the surface of
the fiber. Just under the sarcolemma there are groups of small fat droplets rather irregular in arrange-
ment. In the larger fibers of these slides it is difficult to distinguish the chains except at the very
surface.

Microscopic examination of the caudal pink muscle.—The fibers of the caudal region are closely packed
together and are very free of fat (section M46). The connective tissue septa have strands of fat droplets,
most of them small but some medium in size. These measure from 2 to 15 p, chiefly the former size.

The caudal pink muscle fibers are practically free of intracellular fat. Certain ones show traces of
fat at points on the surface, but these are only traces and are limited to areas bordering on the fat-bearing
septa. Extensive areas with no septa between the fibers are free of fat.

The fibers are so compact that their outlines in cross section are irregular polygons.

The caudal pink teased muscle shows practically no fat in the fibers, traces only appear. Certain
of the fibers have a slight bluish tinge and through their substance are opaque granules which are diffi-
cult to identify. These granules are in the interfibrillar spaces.

Microscopic examination of the trunk dark muscle.—The trunk dark muscle still retains a large amount
of fat (slide Mgr under the oil immersion). Drops between the fibers measure on an average from 6 to
9 #t, occasionally as much as 15 pt.

Fat is distributed throughout the substance of the fiber. The droplets vary greatly in size. The
largest ones run from 2 to 2.6 » in diameter. These are more numerous around the superficial portion
of the fiber in most of the material, though groups of fibers are found in which these large liposomes are
quite evenly distributed through the substance. In numerous fibers the central portion is relatively
free of liposomes and the fibers look lighter in color under the microscope. In the light areas, however,
there are liposomes present, though they are very small for dark fibers, 0.6 to 0.8 #, and they are not as
numerous as in the superficial border. Different portions of the section vary greatly in the amount
offat. The muscles freest of fat are those which lie along the septa which carry blood vessels. Under
a low magnification these areas are sharply limited.

Teased fibers (section M44) give a good view of the amount of fat along the course of individual fibers.
There is much variation in different lengths of one and the same fiber. In the fatter areas numerous
groups of fat droplets lie over the surface of the teased fibers. These groups have a configuration such
as was noted in similar muscle from specimens from Warrendale. Undoubtedly the arrangement of fat
bears a definite relation to the arrangement of blood vessels. The chains of liposomes are continuous
in some areas for long distances. The individual liposomes will measure 2 in diameter in the larger
chains, but vary through a range of much smaller sizes according to the relative amount of fat present.
The larger chains are obviously near the surface of the fiber. Occasionally a chain is noted in which
the majority of the liposomes have fused, forming a fat rod such as has previously been described.
These rods, as observed, lie near the surface of the fiber.

Microscopic examination of the caudal dark muscle——The dark tail muscle has enough fat to give
it a relatively deep stain (M56, oil immersion), but this fat is much less in quantity than in the trunk
muscle of this salmon. The fat is condensed around the superficial areas of the fiber. Apparently
there is some intermuscular fat in droplets 3 to 6 » in diameter.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. TEL

The intramuscular fat is arranged practically the same as in the trunk muscle, except that the
size of the droplets is smaller throughout. Droplets 2 » in diameter are relatively rare and the liposomes
in the chains in the middle portion of the fibers run about 0.5 to 0.8 » in diameter, about the size of
the liposomes in the relatively fat-free fibers of the trunk.

The teased fibers give an explanation of the diffuse appearance of the stain in this section
(Ms58-sg). It is due to the extremely small but numerous liposomes. Liposomes are also present in
groups just under the sarcolemma in many places. In the largest chains observed in fibers that seem
relatively better supplied with fat the liposomes measure from 1 to 1.2 u in diameter.

FAT IN THE TISSUES OF SALMON FROM THE SPAWNING GROUNDS.

Salmon at the spawning grounds in the Clackamas River, at Cazadero, Oreg.,
have been fasting from three to four months. During the entire time there is no food
source for the energy which the salmon have expended other than in the tissues and in
the stores of fat. One would certainly expect the great fat storehouse to be sharply
drawn on if not exhausted by the time this stage of the life cycle is reached.

We have taken fish from the spawning grounds so emaciated and so weakened
that they were scarcely able to maintain equilibrium in the river currents. One knows
that some of these fish would die in a day or so even if they were kept in the water in
the most carefully protected condition.* Fat is still present in the muscular tissues
of such salmon. Its percentage is low, yet the microscope reveals an unexpected
quantity.

There is considerable variation among the different individuals present at Cazadero.
All of the early September spawners are undoubtedly of the so-called spring run of salmon.
They have been in the river at least since May or June (estimated). Vet there are
specimens that have very much more fat than the dying salmon mentioned above.
These variations are shown best in a comparative study of the various types of muscle.

Trunk pink muscle.—The preparations of pink muscle show a sharp contrast with
the loading of fat observed in the normal fish and in fish from the mouth of the Columbia
River. Instead of the large intermuscular fat globules there are only small droplets
present and these are few in number. They are located in the angles in the larger
connective tissue masses which mark the points where several fibers are grouped. If
two fibers are compactly pressed to each other there are no fat droplets between them.
Where several fibers are separated by masses of connective tissue there may still be
small groups of fat droplets. In such cases the fat droplets are small in size and few in
number. In one specimen, no. 131, the number of intermuscular fat droplets averages
about one droplet to ten fibers, and the largest droplet present measures only 6 in
diameter. This quantity of intermuscular fat is insignificant.

The presence of a significant amount of intramuscular fat in these most exhausted
specimens from the spawning grounds is in striking contrast to the disappearance of
the intermuscular fat. The intramuscular fat is in minute liposomes less than 0.5
in diameter. The largest liposomes are in the smallest fibers. They are scattered
throughout the substance of the fibers except in the surface band of fibrille, where the
absence of interfibrillar fat gives the appearance of a clear band of fibrils around the
superficial border of the muscle fiber as seen in transverse section. Around the sur-
face of the pink fibers and under the sarcolemma are groups of fat droplets. Where

@ Males were selected from those retained in the spawning pens. In one instance salmon for study were selected from those
most advanced in retrogression yet with no obvious fatal lesions. Of those not used three had died by the next morning. This
is positive evidence that the salmon selected were at the dying stage.

112 BULLETIN OF THE BUREAU OF FISHERIES.

two fibers touch each other a double row of droplets slightly separated can be easily
distinguished. The separating line, of course, is the section through the sarcolemma.
The largest fat droplets in this series under the sarcolemma measure as much as 2 4
in diameter. Often the droplets are slightly compressed, evidently by the pressure of
the sarcolemma, since the radial diameter is a little less than the diameter tangent to
the fiber.

Taking the pink fibers as a whole it seems that:in the better specimens the liposomic
fat is present in greater number of droplets, also in larger droplets, than in the poorest
specimens from Warrendale (compare fig. 10 and 11). Certainly this fat is greater
than in no. 125 and no. 126. In fact, the comparison is close to those fish which have
the highest quantity of intracellular fat at stations lower down the river. In the poorest
fish taken from the spawning grounds the fat is almost completely eliminated both from
the intermuscular and intramuscular regions. ‘This is true for fish no. 140, which has
the lowest amount of fat observed in the lateral pink muscle.

The pink muscle fibers themselves are not plump and round in the fish from the
spawning ground. On the other hand, they form irregular polygons in cross section.
Even the smaller fibers have lost their cylindrical shape. The fibers are more compact
and the whole appearance suggests a diminution in volume (fig. 19, pl. x1).

In teased preparations there is one rather striking deviation from the typical arrange-
ment of liposomes, namely, the deep-lying liposomes are no longer in such regular
spindle-shaped rows as are found in the normal. The chains have the appearance of
broken rows, in which the smaller liposomes are absent, thus giving the chains an irregu-
larity that is rather constantly observed in the fish of this station. Those chains that
are most definite and least interrupted are clearly located near the surface of the fiber.
In the teased material from no. 139 there is a marked difference in the appearance of
the liposomes present in the small fibers as compared with the larger. In the larger
fibers the chains are less numerous and the droplets in the chains smaller. In this
fish the small fibers, 40 to 50 » in diameter, have very evenly distributed liposomes,
the diameters varying from 0.3 to 0.7 . In certain fibers of this section there are
irregularly placed highly refractive bodies which (1/12 oil immersion) are only lightly
stained. These granules are probably associated with an early stage of degeneration.

These teased fibers also show irregular patches of liposomes over the surface of the
fiber and under the sarcolemma. ‘These fat droplets are a trifle larger than those within
the sarcoplasm. In salmon no. 140 we still have a small amount of fat under the sarco-
lemma. In the larger fibers the intracellular fat is present only in traces, the liposomes
being not over 0.2 in diameter and in very short, irregular, and scattered chains. In
the smaller fibers of the material the number of liposomes is still relatively slight, but
the size of the individual liposomes is somewhat larger. Where a liposomic chain is
present it is noted that the arrangement of individual liposomes is very irregular, giving
the chain the appearance of being broken.

Caudal pink muscle—The pink fibers of the caudal region have as nearly no fat
as in any specimen examined. ‘The intermuscular fat is completely eliminated, while
only a trace of intracellular fat is to be found. ‘The teased preparations show that this
trace is made up of definite but tiny liposomes which are only sufficient in quantity to
give a faint stain to the section. Now and then in a fiber near the surface one can note

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 113

a few small liposomes in irregular groups. At best this quantity and distribution can
be called only a trace.

Trunk dark muscle.—In the dark muscle there is no intermuscular fat, or if any is
present at all it isin tiny droplets in those localities which contain the greatest amount of
connective tissue. In a cross section and in a low-power field one might see two or
three such areas.

The intramuscular fat in the dark muscle has markedly changed in its appearance
and arrangement. The relatively large droplets characteristic of the normal tissue,
often measuring as much as 6 » in diameter, have practically disappeared in this stage.
In place of the large droplets the dark muscle now contains a much greater number of
relatively very small sized droplets of the type described as liposomes. Also the fat
droplets under the sarcolemma are now reduced to liposomic size, not larger than 2 p in
diameter and averaging about 1. In across section these tiny fat drops around the sur-
face of the fiber and within the sarcolemma form definite rings which mark the outlines
of the fibers. Examination under the 1/12 oil immersion shows that many of the drop-
lets entering into the composition of this superficial layer of fat are wedged in between
fibrille. Within the central substance of the muscle the liposomes are now interfibrillar
in arrangement. They are small in size but numerous and comparatively evenly dis-
tributed through the sarcoplasm. This description applies to the trunk dark muscle
of the fatter spawning salmon. The poor salmon do not have so many liposomes in the
deep sarcoplasm and different individual fibers vary greatly in their fat content.

In the teased dark muscle fibers from the trunk it is noted that the most fat lies
just under the sarcolemma, but that it is very irregularly placed. The droplets are small
and seldom exceed 2 » in diameter. There are a few chains of liposomes in the body of
the fiber, but these chains are widely separated and consist of extremely small liposomes.
The largest liposomes are about 0.4 » in diameter in fish no. 139. In no. r4o the sur-
face of the fibers has irregular fat droplets often running as much as 3 » in diameter.
But distributed through the substance of the fiber there are only traces of fat except at
the very superficial sheet of sarcoplasm where the liposomes are measurable.

Caudal dark muscle—In the caudal dark muscle there is still some considerable
quantity of intercellular fat, especially in salmon no. 138, although this fat is less than in
the lateral dark muscle of the same specimen.

Within the sarcoplasm of the fibers there are numerous areas in which there are
only traces of fat. Even at the surface of the fiber there is often only a trace of fat.
Under the sarcolemma the fat is in isolated groups of liposomes measuring only a
fraction of a micron.

In fish no. 139 there is more fat in the caudal dark muscle, especially under the
sarcolemma, where the droplets measure from 1 to 3 #. Different portions of the sec-
tion show great variation in the amount of intracellular fat. These areas are similar to
that noted in Warrendale fishes no. 125 and no. 126. In fibers bordering on these
vascular areas the fat droplets are removed from under the sarcolemma, and are absent
except for traces in the body of the fiber.

In comparing the amount and distribution of fat in the Cazadero specimens it is
obvious that the total percentage amount of fat is profoundly reduced. On the other
hand, it is apparent that this reduction has taken place chiefly in the intercellular fat.
The extreme case of exhaustion shows practically no fat either between or within the

114 BULLETIN OF THE BUREAU OF FISHERIES.

fibers. Yet the majority of the salmon taken at Cazadero show on the average as much
intracellular fat as is shown in those salmon taken from the Columbia River at a much
earlier stage in the migration. Certainly they show as much fat in the fibers as all but
the very fattest of the earlier specimens. It is this showing which presents such a strik-
ing factor in the comparison between the pink muscle of different salmon at the various
stages. The pink muscle maintains a surprisingly large amount of intracellular fat
throughout the whole series of stations, even when the fat is practically eliminated from
the great storage depots.

In the case of the dark muscle, which at the early stages is surcharged with fat,
there is an obvious gradual diminution from the mouth of the river to the spawning
ground. On the other hand, there is no complete elimination of fat below that stage of
smallest liposomes which characterizes the pink muscle as a type. The fat may be
eroded from the dark muscle; that is to say, the large drops will gradually decrease in
size but will never be completely eliminated. There is some factor operating which
maintains a supply of liposomes in the active muscle of the major portion of the body.
It is true this supply is not kept up in the caudal muscle, but this undoubtedly is due
to the great and continuous activity of that musculature.

Cheek muscle.—The amount of fat in the cheek muscle has been described for fish
from Ilwaco, but this particular type of muscle has not been studied in all the interme-
diate stations. However, in one specimen, no. 140 from Cazadero, this muscle has been
carefully reexamined. The fibers of the muscle of this fish are even more compact than
noted at Ilwaco. There is only a small amount of interstitial connective tissue and
this carries a few scattered but small fat droplets.

The intracellular fat is present only in traces. In a large proportion of the fibers
no fat can be distinguished; yet in a few of the smallest fibers merest traces are dis-
cernible. E

The striking characteristic of the cheek muscle of no. 140 is found in the evidences
of degeneration. Certain of the fibers take a definite stain not due to the fat, but due
to characteristic degenerative changes in the fibers. These fibers stain a light rose
pink. Under the oil immersion the fibers that take this special stain show signs of
disintegration or atrophy. The bodies of the fibers have greatly shrunken. Their
outlines show that they are compressed as if between adjacent fibers. The fibrillar
structure has likewise disappeared. Slight vacuoles are present. The most diagnostic
feature of the change consists in the pigment granules of muscular atrophy. These
pigment granules are irregularly placed and vary greatly in size. Measured with the
1/12 oil immersion they vary between 0.1 and 0.2 4 in diameter. The degenerative
changes noted are typical of simple atrophy. The changes in this particular muscle
are the most advanced that have been noted. There is some slight indication of atrophy
in the trunk musculature even at an earlier stage of the journey, but nowhere else have
I found definite degenerative pigments, unless the highly refractive bodies noted in
no. 125 and no. 139 be such.

The further details of these degenerative changes are being studied and will be
presented later in a special report.

ANALYTICAL DETERMINATIONS OF THE PERCENTAGE OF FATS IN THE SPAWNING SALMON.

The three Cazadero salmon of this series from which samples were taken for fat
percentage determinations reveal a larger per cent of fat than one would, a priori, expect.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 115

On the whole, they confirm the microscopic findings. It is unfortunate that no fat
percentage determinations were made on no. 140, the only salmon that presented in
the masseter muscle definite and unmistakable evidence of muscle degenerations. This
series should be compared with the Ilwaco series in table 1, page 92.

TaBLe II.—ANaLyTica, DETERMINATIONS OF Fats IN THE TISSUES OF CERTAIN SALMON FROM
CLACKAMAS RIVER, CAZADERO, OREG., TAKEN SEPTEMBER, IQII.

|
Muscle fats in percent-
age.
No. and sex of fish. | Remarks.
Pink Dark

| muscle. muscle.

pee —— —_
1349. | 2.870 6.719 Spawned.
ayes 6.139 9- 662 Spawning.
138d... 3-332 7-962 Spawning, one-half to three-fifths spawned.

PROTOCOLS.

Spawning female salmon (no. 132), length 960 mm., weight 6,840 grams (after artificial spawning).

External appearance first class, body slender in form. An appreciable oedema of the inner surface
of the body cavity walls. Visceral mass exceptionally small.

Microscopic examination of the trunk pink muscle.—In a small portion of section N47 there is a very
small quantity of intermuscular fat, but in general neither the connective tissue septa nor the thicker
strands separating the fibers have more than tiny liposomes. ‘There is a sharp contrast between this fish
and no. 127 from Celilo in this regard. These fat droplets in the connective tissue measure a maximum
of 5 to6m. The average size is very much smaller, between 1 and 2. The trunk pink fibers of this
salmon are somewhat more plump and round in outline than in no. 131. All the larger fibers, however,
are compressed and irregular in outline, suggesting the same type of change noticed in other relatively
fat free tissues.

There is intramuscular fat in the smaller fibers and in the medium in the form of liposomes. These
liposomes are distributed rather uniformly through the substance of the smaller fibers. In the medium-
sized fibers there are not so many liposomes and they are much fewer in the center of the fibers. Around
the circumference of the fibers and under the sarcolemma there is more fat, especially under the sar-
colemma. These small and intermediate fibers have sometimes almost complete rings of fat droplets
under the sarcolemma. ‘The largest fibers have a greatly reduced quantity of fat. The fat is in very
much smaller liposomes, often scarcely visible. Around the surface and under the sarcolemma there
are groups of fat droplets, but not so plentiful as in the medium fibers.

This material on the whole is characterized by the small amount of intermuscular fat and the rela-
tively great amount of intramuscular fat. In some portions of the sections the intramuscular fat is as
great asin fish no. 118 from Ilwaco, far greater than in no. 126 from Warrendale or no. 127 from Seuferts
Cement Wheel. The uniformity of distribution of the fat is not so great as in the Ilwaco specimens.

Microscopic examination of the caudal pink muscle.—There is no intermuscular fat in this section
(N59). In the myocommata shown there is a trifle of fat just at the surface.

The muscle fibers have some intramuscular fat, but the relative amount in the different types of
fibers is difficult to determine, on account of the excess of precipitate present. Some of this so-called
precipitate is in characteristic round granules like unstained fat, but some of it is in the characteristic
scarlet red color, interpreted as a less successful staining manipulation.

Microscopic examination of the trunk dark muscle (slides N53-57).—These sections all show a rela-
tively small amount of fat in the nmryocommata.

The muscle tissue as a whole shows much fat in the critical region at the surface of and between
the fibers. Under the oil immersion it is apparent that a large amount of this fat is between the fibers
in droplets from 2.5 to 3 “in diameter. The muscle fibers are so compact in arrangement that it is diffi-
cult to identify the exact limits of the fiber. Certain unquestioned regions show this fat between the

116 BULLETIN OF THE BUREAU OF FISHERIES.

fibers. In other regions one can as definitely say that there is fat beneath the sarcolemma but outside
the substance of the fiber. At the border of the section a number of fibers have been slightly split
apart and some turned in a horizontal position. These fibers confirm the above.

As arule, through the section there is only a small amount of intramuscular fat in chains of liposomes
through the bodies of the fibers. In the horizontal fibers the size of the liposomes is shown to be from
0.4 to 1.5 p in diameter. The majority of the liposomes are of the larger sizes.

The striking characteristic of this tissue is the great differentiation as between the amount of fat
in the body of the fiber and at the surface. There is apparently more fat in the dark muscle of this
fish than in fish no. 128 and no. 129 from Celilo, undoubtedly more than in fish no. 131 of Cazadero,
but the bodies of the fibers contain relatively less.

Microscopic examination of the caudal dark muscle.—The amount of intermuscular fat is greatly
reduced over that of the trunk muscle, the droplets are smaller, and they are not sonumerous as in that
region. They are, however, sufficient to give a mosaic-like appearance to the section.

The intramuscular fat is extremely small except such as lies just under the sarcolemma. Certain
of the sections show the ends of the fibrillz clear and sharp. The pattern is the same as shown in figure
7, plate v, except that there are no fat spaces present. This certainly indicates that the muscle has
not degenerated, yet nearly all the fat characteristic of the normal muscle is absent and there are only
a very few liposomes in the fibers near the surface.

Spawning male salmon (no. 138), length 840 mm., weight 7,730 grams, from the spawning pens of the
United States Fisheries Station, Cazadero, Oreg., on the Clackamas River, September 4, IgII.

This male salmon was one-half to three-fifths spawned. Color brassy, with black spots. Soft dorsal
decayed and one fungus spot on dorsal fin, A fish in good condition but in late stage of exhaustion.
It would probably have died in the course of 24 to 36 hours.

Microscopic examination of the trunk pink muscle.—The pink fibers of salmon no. 138 are not plump
and round as in fish no. 118. On the other hand, their outlines form irregular polygons. Even the
smaller fibers have this shape. The larger fibers bear histological evidence of great decrease in size
(samples one day in formalin). A set of cross-sectional measurements show the following: 40 by 8o,
60 by 70, 80 by 140, and roo by 220 p, outlines all very irregular. The cross sections show both the
striations and the fibrille very nicely.

The sections are free of intermuscular fat except for a few of the finest droplets and liposomes.
In a large section there is one such group of fat droplets in a large mass of connective tissue. There
are a few thin strands of connective tissue between the fibers, and these carry occasional fat droplets
not over 3 to 44in diameter. The larger size is rare, though fat drops in the same locality in no. 118
from Ilwaco measure 100 7 and more. ‘There are a good many tiny liposomes in this connective tissue,
though the mass is extremely small.

In contrast to the dearth of fat between the fibers there is an unexpectedly large quantity of intra-
muscular fat within. As usual this fat is in liposomes. The size and number are both greater in the
smallest fibers, yet the largest fibers have a pretty even sprinkling of liposomes of extremely minute
size. I can not find a single fiber but that has some fat within its protoplasm. In the smaller fibers
there is a great amount of fat around the surface in the region just under the sarcolemma. Taking the
small fibers as a whole, it seems that the liposomic fat is present in a greater number of droplets and
in slightly larger droplets than in the poorer specimens from Warrendale. There is decidedly more
intramuscular fat than in fish no. 126 from Warrendale.

The teased fibers give a beautiful confirmation of the notes made on the cross sections. They show
one variation from the typical arrangement of liposomes, namely, that the deep-lying liposomes are not
in such regular rows as in the normal. The liposomes have a very irregular appearance, as if the smaller
liposomes had disappeared. Those liposomes that are left are unusually large and uniform in size for
pink fibers. The rows that are most definite and uninterrupted are located near the surface of the fiber.
The larger liposomes are from 2 to 2.5 » in diameter.

Microscopic examination of the caudal pink muscle.—The caudal pink muscle sections are free of all
but traces of fat. The intermuscular fat is limited to a few droplets in the myocommata. Only a few
irregular chains of very small liposomes are present in the fibers. These are more distinct in the smaller
fibers. The striations are distinct and clear, but the fat is faint and scarcely distinguishable.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. Dy7

Microscopic examination of the trunk dark muscle.—The dark trunk fibers are very compactly arranged,
so that it is difficult to distinguish between the intermuscular fat and subsarcolemmal fat. But it is
apparent that there is a small amount of intermuscular fat. This is confirmed by the connective tissue
of teased muscle. The drops are of liposomic size, rarely more than 2 to 2.5 ” in diameter.

The intramuscular fat is present in the usual localities. Under the sarcolemma the droplets are
medium in size, 1.7, 2.8, 3, and 5.4 “in typical droplets. The sarcoplasm of the fibers is rather evenly
studded with unusually small liposomes for this type of muscle. ‘The liposomes are even smaller than
in the trunk pink muscle, are condensed about the surface of the fibers and are larger there, 0.4 to 1.2 p.
In the body of the fiber there are fewer liposomes and the size is from 0.2 to 0.8 yt.

The teased dark fibers confirm the above notes as to the fat disposal. The fibers have exception-
ally clear striations which average 2.6 “in length. ‘The fat is irregular and the liposomes of the chains
can not be followed in relation to the striations. The fat under the sarcolemma is in patches in these
fibers. The whole appearance of the dark fibers closely approaches that of the pink fibers of this salmon.

Microscopic examination of the caudal dark muscle.—In the caudal dark muscle there is still enough
fat between the fibers and under the sarcolemma to give a distinct mosaic-like marking of the outlines
of the fibers. In some portions of the sections this surface fat is gone.

The intramuscular fat is still much less than in the trunk dark muscle. There are fibers that
have the merest trace of liposomes in the muscle substance. Even in the fattest fibers only minute
liposomes are present, the smallest amounts observed in this type of muscle, except in no. 140. ‘There
is the sharpest contrast between this tissue and the Ilwaco type.

Spawning male salmon (no. 140), length 980 mm., weight 8,070 grams.

This fish was selected as a type of spawning fish in good physiological condition but at the last stages
before death. It was taken at the United States Fisheries station at Cazadero, Oreg., on the Clackamas
River, September 6, rgrt.

Microscopic examination of the trunk pink muscle.—There are traces only of intermuscular fat, which
is in fine liposome-like droplets in the larger connective tissue strands and in the myocommata.

The small fibers still contain a sprinkling of liposomes, enough to give them a decidedly pink appear-
ance under the low power. The 1/12 oil immersion shows that these liposomes are gathered chiefly
around the superficial border of the fiber and that they are in groups in the neighborhood of intermuscular
fat. The largest liposome observed in a small fiber measured 1.4 » in diameter, but such are few in
number. The average size of liposomes for these small fibers is only a fraction of a micron.

In a large fiber under the oil immersion there are a few areas of fat in irregular-sized droplets under
the sarcolemma, traces in contrast with Ilwaco salmon. ‘Through the body of the fiber there are chains
of finest liposomes o.1 to 0.2 # in diameter, but the chains are short and irregular. In two other typical
fibers in the field, one medium and the other rather small, the liposomes are present in about the same
number of chains as in the large fiber, but are slightly larger in size. In all these fibers the particular
characteristic feature is the irregularity in the chains of liposomes. The individual chains have not
the usual arrangement of larger liposomes in the middle of the chain and the size tapering down to small
ones at the end of the chain. They are irregular in size throughout the chain.

Microscopic examination of the caudal pink muscle.—There is practically no fat in this caudal muscle.
The trace that is present (1/12 oil) is only enough to give a faint stain to the superficial border of occa-
sional fibers. There are scattered and irregular groups of a few small liposomes just at the surface of the
fibers. Several fibers that have been turned horizontally show no chains of liposomes.

Teased fibers show no liposomes in the body, but occasional traces of liposomes just at the surface
of the fibers. These traces are definitely between the sarcolemma and the sarcoplasm, an arrangement
most often noted in pink fibers poor in fat.

Microscopic examination of the trunk dark muscle-—The outlines of the trunk dark fibers are marked
by rather heavy rings of fat droplets. These markings are least prominent in the neighborhood of
vascular areas. It is not always possible to distinguish between intermuscular and subsarcolemmal fat.
Both are present. Most of the fat observed is judged to be under the sarcolemma. ‘he fat drops
between the fibers range in size from 3 p to 8.

The fat under the sarcolemma seems to be in rather small but numerous droplets. This fat runs
from 2 to 4 4 in diameter. Out in the body of the muscle fibers there is a variable arrangement of

118 BULLETIN OF THE BUREAU OF FISHERIES.

liposomes. In some fibers they are uniformly distributed through the substance of the fibers; in others
there is apparently no fat in the middle of the fiber. In general, liposomes are present in the superficial
areas even in those fibers which have the least fat. The whole appearance suggests a nutritive balance
in which it is just a question whether all the fat will be used or whether there will be an excess sufficient
for deposit.

The teased muscle shows great variation in the quantity of fat in the individual fibers (1/r2 oil).
At the surface of the fiber is a good deal of fat in small droplets, the subsarcolemmal fat. The liposomes
in the chains are small like those in normal pink muscle. The number of chains is also low. In the
fibers carrying the least fat these chains are all but absent.

Microscopic examination of the cheek muscle.—The muscle fibers of the cheek muscle are very com-
pactly arranged. In comparison with the Ilwaco type the fibers are less rotund in outline. The
histological structure is indistinct. In many fibers the fibrille can not be seen because of a disintegra-
tion which marks the first stage of degeneration. Certain fibers scattered irregularly through the section
show a definite protoplasmic degeneration with vacuoles and pigmentation. The pigment granules are
small in size, 0.1 too.2x#in diameter. They are unevenly distributed through the fibers and apparently
somewhat greater near the surface. There is great variation in the quantity of pigment in different
individual fibers.

The fat in the cheek muscle is limited to a very few small groups of intermuscular fat. No traces
of fat could be distinguished within the fibers themselves.

SIGNIFICANCE OF THE OBSERVED CHANGES OF THE AMOUNT OF FAT.

It is obvious that the salmon fat furnishes the food during the migration fast.
The revelations of the microscope are convincing on this point, even if there were no
collateral supporting evidence.

My unpublished chemical analyses of the tissues have revealed a dearth of carbo-
hydrates in the salmon tissues at all stages of the migration. This fact is of vital signifi-
cance in connection with the fat problem. The lack of carbohydrates and the abundance
of fats support Miescher’s assumption that fats furnish the source of the muscular energy
expended by the salmon during the migration. In connection with a series of salmon-
feeding experiments ¢ it was shown that the salmon liver exercises a distinct lipogenic ®
function during the feeding and growing stage. Noél Paton has found that the amount
of fat in the liver of the frog is increased after fat feeding.“ It seems to me that in
animals like the salmon the lipogenic function of the liver becomes a primary function,
taking a rdle quite comparable to that of the glycogenic function of the organ for many
mammals. Fishes of this class are carnivorous. Their food is of a highly oily character,
as is also that of certain birds, and is continuously so. The food is rich in proteins and
fats and in inorganic constituents, but it is poor in carbohydrates. In the adaptations
to such a diet, if for no other reason, the salmon has reached the point in its phylogenetic
development where fats furnish a direct and primary source of foods for the energy

@ Now in manuscript.

b It was Loevenhart (American Journal of Physiology, vol. 6, p. 331, 1901) who first advocated the idea that we might have
a “lipogenesis’’ in the body comparable in character to the “ glycogenesis’’ of Claude Bernard. He suggests that wherever there
is fat storage there will be lipase, and proves it by investigations on a number of tissues that contain fat, for example, the liver,
mammary gland, pancreas, brain, spleen, heart muscle, blood, adipose tissue, etc. He says: “In the case of fats the areolar tissue
is the great primary store, secondary deposits being found in all the tissues. In some animals even this difference in the storing
of fats and carbohydrates is not tobe noted. In many fish, notably the cod, the liver, at certain seasons of the year, becomes the
great depository for fat. The liver we have found to possess powerful lipolytic activity, and hence, under proper conditions, it
should be capable of storing fat. Moreover, this is in accordance with the experiments of Noel Paton, who showed that the fat
contained in the liver of frogs is increased after a fatty meal. It is believed that both phases of lipogenesis are induced by
lipase, a fat-splitting and fat-forming enzyme.”” From my observations I am convinced that lipogenesis is a definite and specific
function of the liver in certain carnivorous animals whose normal food consists of a high percentage of fat, as is the case in the
king salmon.

¢ Paton, D. Noél: On the relationship of the liver to fats. Journal of Physiology, vol. xrx, 1896, p. 167.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 119

liberating tissues. The tissues, in short, can utilize the energy of fats by direct oxida-
tions. It remains to examine the facts submitted and to discover, if possible, the
mechanism whereby this great store of salmon fat is rendered so labile and so wonder-
fully efficient in the execution of the activities of this last lap of the salmon life cycle.

Of all the facts presented it seems to me the most significant are:

I. The appearance of intramuscular fat in the pink muS@le at the beginning of the
spawning migration, and

Il. The maintaining of a relatively uniform distribution of this fat in the fibers
until the death of the animal.

Just as soon as the salmon ceases to feed, and the products of digestion no longer
reach the active musculature, then, and not until then, there is thrown into the pink
muscle fibers a supply of fat adequate to the energy needs of this most critical period
in the salmon life cycle. The excess of fat is deposited in an extremely finely divided
state and is brought into intimate contact with the fibrille, one can almost say with
the sarcous elements. Certainly in many rows of fat droplets between the fibrille the
individual liposomes are in close approximation with corresponding sarcous elements in
the fibrils. One can not escape the inference that this microscopic emulsion of the fat,
its general distribution throughout the muscle fiber, and the intimate relation with the
elemental fibrillar structure, all point to an immediate utilization of the fats in the pro-
duction of muscular energy.

This hypothesis is further supported by the observations on the cheek muscle and
on the fin muscles, all muscles in much more uniform, though less intense, activity than
the lateral muscle. These muscles carry a light but strikingly persistent load of minutely
divided intramuscular fat during the entire migration. They never have a great excess
of storage fat, either intermuscular, as in the pink muscle, or intramuscular, as in the
dark muscle.

The great fat storehouses are the intermuscular fat of the great lateral pink muscle,
the inter- and intra-muscular fat of the dark muscle, and the fat of the adipose connective
tissues. With the cessation of feeding no further fats, proteins, etc., are brought in as
foods. With the external food supply now shut off the physiological mechanism of the
- salmon must turn to the food materials on hand, to the internal food supplies of the
salmon’s own body. ‘The internal supply is limited to body tissues as such, and to the
fats. It isthefats that are immediately drawn upon. From the fat deposits the fat is
gradually but regularly transported to the active muscles, where it is maintained in a
uniform and favorable distribution, and in amount adequate to supply the energy
expended by the salmon in the migration fight against the currents and rapids of the rivers
on its way to the spawning beds.

TRANSPORTATION OF FATS IN THE FASTING SALMON.
HISTORICAL.

The histological observations on the king salmon have given every confirmation
of Miescher’s original assumption, based on his study of the Rhine salmon, that the
fat of the salmon can be transported from one part of the body to another; i. e., from
tissue to tissue. He laid special emphasis on the utilization of the muscle fats in the
building up of the fats of the ovaries, but he also suggested that fat was the source of

120 BULLETIN OF THE BUREAU OF FISHERIES.

the nourishment of the animal.“ It is irrelevant for the present purpose that Miescher
considered the source of the fat to be a fatty degeneration of the muscle tissue. The fact
remains that he demonstrated the presence of intramuscular fat microscopically and
for this he should receive full credit. He must also be given full credit for the concep-
tion that the salmon fat can be transported for purposes of tissue growth, ovaries, and
for use in energy productiéh, muscles. I can not find that he has offered any explanation
of the detail of the processes involved in the fat transference or that he has discussed
the matter, but this work is so important that the three statements he makes are quoted
in full in their setting, and in his own words. On page 186 he says:

Dass wirklich der Seitenrumpfmuskel die wesentlichste Stoffquelle ist, sowohl fiir die Emahrung
des Thieres, als fiir die Gesclechtsreifung, wird evident bestiitigt durch das Mikroskop. Schon die
Winter- und Friihjahrssalmen zeigen nimlich zwischen den feinen quergestreiften Elementarfaiden
(Fibrillen) der ungleich dicken Muskelfasern, besonders in den diinneren, bald mehr bald weniger
ausgesproqhene Fetttrépfchenreihen, wie man sie als Anzeichen sogenannter Entartung des Muskel-
gewebes kennt. Die Menge dieser Fetttrépfchen nimmt gerade im Hochsommer, wenn der Eierstock
rascher zu wachsen beginnt, betrachtlich zu und kann bis zur Undurchsichtigkeit mancher Fasern
fithren. Am starksten degenerirt eine gesonderte diinne Muskelplatte, die an der Seite des K6rpers
direct unter der Haut liegt (Hautmuskel). Dagegen bleiben sozusagen v6llig intact und fettfrei alle
ibrigen Muskeln, Brustflosse, Bauchflosse, Riicken- und Afterflosse, Kiefer- und Zungenbeinmuskeln,
der obere und untere Langsmuskel und die Schwanzmuskeln im engern Sinne. Nur die Bauchflosse
zeigte an einigen Stellen schwache Anzeichen von Degeneration.

The one comparison as to the intramuscular loading of the fibers with liposomes as
the migration time continues is given on page 213.

So findet man denn bei den Friihsalmen jene schwache, hauptsiachlich die diinneren Muskelfasern
in missigem Grade betreffende Durchsetzung mit Reihen feiner Fettkérmchen, bis dann im Frithsommer
das Wachsthum des Eierstocks in seiner geometrischen Progression zu einem absoluten monatlichen
Stoffverbrauch fiihrt, dessen Anforderungen neben der eigentlichen Selbstzehrung sich gebieterisch
in den Vordergrund driingen und wirksamere Hilfsmittel verlangen.

When the Rhine salmon spawn they begin the return migration to the sea with the
associated recuperative processes. Concerning this stage, Miescher makes the following
final statement on page 215 of his monograph (page 171 of the reprint):

Wie ganz anders das Bild, wenn wir Gelegenheit haben, Thiere zu sehen, die auch nur um ro Tage,
besser um ein paar Wochen das Laichen hinter sich haben (leere Weibchen, zu Ende December oder im
Januar gefangen, aber auch eines aus Herrn Glaser’s Fishkiisten, gewiss nicht mehr als ro-Tage von seinen
Eiern befreit). Die Haut ist wieder bliulich glinzend und klar, die Geschwiire tibernarbt oder in
Heilung, das Fleisch durchscheinend, von Fettkérnchen vdllig oder fast véllig befreit; auch die Herz-
fasern in Reinigung begriffen; im Darm keine Spur von Nahrung. Dagegen enthilt der Eierstock bald
mehr bald weniger Eier, die, in einen serésen oder auch etwas eitrigen Erguss der Follikelhaut einge-
bettet, sichtlich zusammenschrumpfen und aufgesogen werden.

Mahalanobis? in Noél Paton’s report on the life history of the salmon also calls atten-
tion to the storage of fat in the muscular tissues, and to the use of this fat in the repro-
ductive organs and in the ‘production of energy. This author calls attention to two
important things, viz, (1) he observes that the fat is present in largest amount in the
muscles of fish entering the estuaries, and in least amount at spawning stage, and (2)
he refutes Miescher’s degeneration theory as a means of accounting for the presence of

@ Miescher: Statistische und biologische Beitrage zur Kenntniss vom Leben des Rheinlachses im Siisswasser, s. 186, 1886;
also reprinted in Die histochemischen und physiologischen Arbeiten, s. 145, 1897.

> Paton, D. Noél: The life history of the salmon. Article 9, by Mahalanobis, S. C., Microscopical observations or muscle
fat in the salmon. Fishery Board for Scotland Report, 1898, p. 106.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. I2I

the fat in the fibers. Mahalanobis considers the fat deposition as a ‘‘fatty infiltration
due to increased accumulation of fat from diminished utilization in the tissues.’’* In
the same paragraph he says:? “Figure 3 is from a fish fresh from the sea—one that had
been actively feeding, and consequently its blood and lymph were rich in fat, whence,
in all probability, the muscle cells absorbed fat and stored it between the fibrils.”’ And
“as already pointed out, the fat granules in fish leaving the sea are more crowded imme-
diately under the sarcolemma (fig. 3, pl. m1).’’ These quotations include all the remarks
tending to show the author’s views as to fat transferences in the salmon tissue. It is
evident that he has given attention only to the mechanism whereby the fat is originally
laid down in the muscle, and his conception is that the process is one of ‘‘infiltration”’ or
absorption from the blood and lymph.

Mahalanobis is hopelessly ‘confused in his studies by the fact that he has failed to
recognize the dissimilarity of two strikingly different tissues. This is shown in the follow-
ing quotation and by the figures referred to therein: ‘In the fish leaving the sea this
accumulation of fat in the fibers sometimes reaches an enormous amount, and a thick
layer occurs under the sarcolemma. This will be evident from a comparison between
figure 2 and figure 3, the former being froma late fish, no. 69, and the latter from no. 79.”’
The figures given by him exhibit differences both in structure and in fat disposal which
bear no relation either to the seasonal type or to the stage of fasting which Mahalanobis
observed, as the studies presented here on the king salmon conclusively show. The
basis of this matter is more fully discussed in this paper in connection with the descrip-
tion of the tissues in question. Mahalanobis compared the dark muscle fiber of his fish
no. 79 with a pink fiber of no. 69. The former muscle is normally loaded with enor-
mous fat droplets in the fiber, whereas the latter muscle never has fat in larger size than
liposomes. His figure 3 from fish no. 69 is from the intermediate zone of pink fibers.
Had he chosen a deeper group of fibers the dearth of fat would have been undoubt-
edly greater. The figures are illustrative of the two normal extremes, are from wholly
different types of muscle, and are, therefore, not directly comparable. This fact he appar-
ently fails to recognize, though his first quotation from Miescher should have helped in
the identification of the muscle types he used.

On the comparative side of the question involved here the recent brilliant work
of Bell should be presented.?_ Bell has studied the liposomes in the muscle fibers of the
ox, dog, cat, rabbit, rat, and the frog. He has also examined the moth (Phlegethontius),

and the fly (Musca). He presents a good historical statement of the work that has
been done along this line, from the discovery of interstitial granules of muscle fibers by

Henle in 1841, down to the publication of his own work in 1911. Bell calls the muscle
interstitial granules that are of a fatty nature “liposomes,” a term introduced by
Albrecht.

Bell, in discussing the granules in the muscle fibers, says (p. 310), ‘All agree with
Kolliker that the granules lie in the sarcoplasm between the fibrils,” and later:

In the skeletal muscles of vertebrates, when the cross markings are wide and distinct, it can usually
be seen that the granules occupy the J-band. But when the striations are narrow the granules seem to
extend the entire distance between adjacent Krause’s membranes. Large granules nearly always lie

@ Paton, op. cit., p. 110.

b Paton, op. cit., pl. 1.

* Mahalanobis, op. cit., p. 108. His fish no. 79 was ‘“‘a fish fresh from the sea.”

d Bell, E. T.: The interstitial granules of striated muscle and their relation to nutrition. Internationalen Monatschrift fiir
Anatomie und Physiologie, bd. xxvuu, s. 297, 1911.

122 BULLETIN OF THE BUREAU OF FISHERIES.

partly at least in the Q-band. In many fibers the granules are arranged in distinct transverse rows,
being apparently limited by Krause’s membrane.

The nature of the muscle granules and particularly of the liposomes has been
extensively studied by Bell. But as the question of the kind of fat has not been espe-
cially investigated in the salmon muscle the review of the discussion will be omitted at
the present time. The contributions by Bell that are of special and far-reaching value
in relation to the questions involved in this paper are two: First, the influence of star-
vation on the fat content of the muscle tissue; second, the influence of fat feeding on
the number and size of the liposomes of the muscles. Under the subject of lack of food,
Bell says:

In every animal there is a gradual disappearance of the liposomes during inanition. As the animal
loses weight the liposomes gradually become smaller and less refractive; and they also stain with decreas-
ing intensity. The muscle fibers of a well-nourished cat are usually full of coarse deeply stained drop-
lets such as is shown in figure 1, from the frog.

Also:

In the rat there is a very rapid decrease in the number, size, refractive power, and staining intensity
of the liposomes. A well-fed rat may contain a large number of strongly refractive liposomes in its
muscle fibers, many of which may be stained with osmic acid. After a reduction in the body weight of
15 to 20 per cent only a few faintly refractive liposomes are usually left. After a reduction of 25 to 30
per cent, it is often found that no liposomes at all can be demonstrated. Every liposome has disappeared.

The remarkable sensitiveness of the liposomes in rat muscle to the food supply undoubtedly accounts
to a considerable extent for the large variations one finds in animals gathered at random. It will be
shown, however, later that the quality of the food is a factor of almost as much importance as the quantity.
A rat whose body weight has been reduced 25 to 30 per cent may develop a large number of deeply
staining liposomes in its muscle fibers (if fed on a diet largely composed of fat meat) though the body
weight remains far below normal.

There is, as has been shown, a marked difference in the number and character of the liposomes of a
well-nourished normal animal and those of an emaciated animal, but the liposomes of an animal in
ordinary condition may not differ essentially from those of a very fat individual. No particular differ-
ences were noted between the muscle liposomes of steers, in which the subcutaneous fatty layer was
6 cm. thick, and those of steers in which this layer was only 5 mm. thick. It was also noted in rats and
dogs that excessive amounts of connective tissue fat are not coordinated with excessive development of
the liposomes.

It is however clear from the above-described disappearance of the liposomes during inanition that
they consist of some form of reserve food substance. This conclusion is in accord with the view that
they consist of true fats or fat-like substances. The gradual decrease in the refractive power and
staining-intensity of the liposomes indicates that the fats are mixed in the liposome with some substance
other than fat.

Under the topic of ‘‘The effect of special feeding on the liposomes”’ Bell says:

Some interesting results were obtained by feeding summer frogs on special rations. It has been
pointed out above that in the summer months (June, July, and August) the muscle fibers contain very
little fat. In a great many animals, in July and early August at least, no liposomes at all can be
demonstrated in the light fibers, and those in the dark fibers are very small and faint and can only be
stained with Herxheimer’s solution. Some young frogs were found in which no liposomes at all could
be shown. It was found that when frogs in this condition were fed heavily on olive oil or fat meat for a
few days the fibers become loaded with liposomes, giving a picture similar to that found in winter
animals.

Bell also tested the fat content of the muscles of frogs caught in the field showing

that the muscles of leopard frogs before feeding had a ‘‘few faint liposomes in the dark
fibers, none in the light fibers,” but after feeding with fat meat and olive oil all the

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 123

fibers are loaded with liposomes. ‘Those in the dark fibers are large and stain with
considerable intensity.

After a series of experiments on leopard and bullfrogs he says: ‘“‘It is apparent that
if the frog be fed an excessive amount of fat the fat will be rapidly stored up in the
muscle fibers.’”’ Similar experiments were performed on rats, which were kept on a
low ration until the liposomes were removed, then were fed on a ration of fat meat.
Under this diet the rats gained in weight and the muscle ‘‘fibers filled with liposomes.”

By these brilliant experiments Bell has conclusively proved that the liposomes in
the muscles of vertebrates, frogs, and mammals bear a distinct relation to the state of
nutrition. The liposomes decrease in number and quantity under a low state of nutri-
tion and they can be increased in size and number when the animals receive a favorable
food. These experiments are of peculiar importance to the problem of the present
paper, since they prove that the presence of the liposomes in muscle tissue is to a certain
extent an index of the nutritive condition of the animal in question. It does not of
necessity follow that the liposome content of the tissues of an animal in the fasting
condition, as in the case of the salmon, will have the same significance. From my
previous work, however, and from numerous field observations, I had arrived at the
working hypothesis that this was the case in the salmon, a position strengthened by the
conclusions of Prof. Bell, which he kindly communicated to me befote his results were
published.

The salmon muscle fat is a filtration fat, not a fatty degeneration. It may be
stated here that the studies on the king salmon tend to disprove Miescher’s theory that
the intracellular fat of the salmon muscle, of whatever type the muscle, is a fatty degen-
eration, a ‘‘Fettentartung’’;* and support the observations of Mahalanobis that the
process is an “‘infiltration.’’ In short, the observations made on the king salmon have
tended to confirm the view expressed above that the intracellular fat of the king salmon
is an expression of the nutritive state of the muscle. It is a loading of fat by a process
of infiltration, as will be explained more fully, and is not a degeneration of the muscle
substance.

It seems surprising that the test of degeneration versus infiltration should not have
been applied to the material under discussion by Miescher and by Mahalanobis. Any
examination of histological sections ought to have shown that there was no appreciable
and adequate conversion of cell proteins into fat, and this observation would have
settled the matter. Transverse sections of dark muscle taken at a late stage in the
migration journey show great regularity of structure, and this structure is of the normal
type. If the muscle protoplasm had undergone fatty degeneration commensurate with
the amount of fat found in this tissue at the time of its greatest load of fat, it is evident
that there would be little normal protein left, and that this little would show pathological
structure. This pathological condition I have never seen except in the extreme ema-
ciated condition found at the time of death. Even then it was found to be extensive in
only one tissue, the great masseter muscle, and this muscle contained no fat.

If argument were still lacking to establish an alibi for the “fatty degeneration”
process of laying down fat in the salmon muscle, it ought to be supplied by the fact that
the young and actively growing dark muscle fibers of the superficialis lateralis muscle
bear a heavy load of intracellular fat. These fibers take on a rich deposit of intracel-

@ Miescher, op. cit., p. 207.

124 BULLETIN OF THE BUREAU OF FISHERIES.

lular fat, both when the muscles are small and immature and when they are larger, also
at a time when they are undergoing longitudinal cleavage.

I have several stages of relatively young fish, from 7 to 16 cm. long, all of which
show a rich deposit of fat in the fibers of the superficialis lateralis. In the older fish,@
as measured by the standard of size, there is a heavy loading of fat in the dark muscle
with corresponding separation of the bundles of fibrille. There is, however, no disap-
pearance of fibrillaee or other unusual characteristic than the distortion that comes from
the presence of such enormous quantities of fat. These remarks all apply, of course,
to the dark type of muscle. In the pink muscle there is little or no intracellular fat in
the muscle fibers during any phase of the growth cycle. ‘This fat appears only after the
feeding ceases.

It seems obvious that intracellular fat of the muscle can not, in the salmon, be
attributed to ‘‘fatty degeneration” in any true sense as signifying a protein degenera-
tion, or, for that matter, a protein cleavage. The amount of protein present does not
justify such a conclusion.

MECHANISM OF FAT TRANSFERENCE IN THE SALMON BODY.

Fat metabolism in most animals, in the Mammalia for example, always involves
the two intertwined problems of most nutrition experiments, namely, fat intake and fat
mobilization. The former carries with it the detail of fat digestion, absorption, and the
laying down of the fats in the fat-storing tissues. The latter involves the taking up of
the fats from the storage tissues and their utilization in the production of new tissue or
in the liberation of energy, as the case may be, and such transferences in the body as
either method of utilization may entail. In most animals complete separation of these
two groups of processes involves more or less abnormal conditions for the animal. But
for the salmon the long fasting period is a perfectly normal process. We, therefore, can
make observations under the grim assurance that the salmon will not, in fact, can not,
eat. There is no added fat being absorbed during this fasting period, hence we have
present at this period only the uncomplicated mobilization and utilization processes.

The discovery of a fat-splitting enzyme, or lipase, was made by Claude Bernard in
1846, and it was early suggested that the fats of absorption might be resynthesized in
the intestinal epithelium. It was not, however, until 1900, when Kastle and Loeven-
hart ? announced their brilliant discovery of the reversible action of lipase, that we
have had an adequate and thorough comprehension of the mechanism whereby the
animal body can transport fats from tissue to tissue. In light of the reversibility of
lipase action it is easy to see how a fatty infiltration can make its appearance in a tissue
so stable in structure as striated muscle, without assuming a disintegration of its pro-
toplasm, as in the fatty degeneration theories.

In the problem before us I have already discussed at length the comparison with
regard to the actual loading of fat in the two chief types of tissue, the lateral dark muscle
and the lateral pink muscle. These are very different types of muscle, and, while the
problem and the controlling factors are essentially similar, it will greatly simplify the

a] have during a number of years found salmon of various sizes entering the fresh waters, all of them exhibiting a great
variation in maturity of sex organs. These two facts, i. e., size and maturity, are independent of each other.
b Kastle and Loevenhart: American Chemical Journal, vol. XxIv, p. 491, 1900.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 125

presentation of the matter to consider the phenomena of fat mobilization in these tissues
separately. I will take the simpler case first, i. e., the pink muscle tissue of the lateralis
profundis muscle.

TRANSFERENCE OF FAT IN THE PINK MUSCLE.

Let the reader recall the characteristics of this muscle up to the time of the entrance
of the salmon into fresh waters, viz: (1) The intermuscular fat of the pink muscle is
slight in the early growing stage as represented by fish 10 to 15 cm. long which are
migrating toward the sea. The intermuscular fat increases in quantity after they reach
the sea, and reaches its maximum when the fish cease feeding—that is, when they begin
the adaptation process preparatory to entering the estuaries. (2) The intramuscular
fat is absent in the young salmon of fresh water, also in the voraciously feeding sea
forms, up to the time the salmon cease to feed, except for traces of fat in the smaller
fibers just at the end of this period. (3) The intramuscular fat after the beginning of
the spawning migration makes its appearance throughout the substance of the pink
muscle fibers of all sizes. It appears in short chains of very small liposomes that are
quite evenly interspersed among the groups of fibrille of the muscle cells. This intra-
cellular fat is present within the pink muscle fibers throughout the migration and at the
time of death after the spawning.

The special contrast is in the distribution of muscular fat just before and just after
the salmon cease to feed. The important change is in the relatively sudden appearance
of the liposomes among the ‘fibrillee of the pink fibers. For this phenomenon the follow-
ing explanation is offered:

Active feeding salmon are also rapidly growing salmon. While growth is taking
place all excess of fat is laid down in the connective tissue or in the dark muscle and
never in the muscle fibers of the pink muscle. The concentration of the fatty products
never exceeds the oxidations in the fibers of the pink muscle, hence no intracellular
deposit occurs.? The transition from a feeding to a fasting state is associated with
numerous tissue changes in other parts of the body, changes which are accompanied by
equally important functional readjustments. Among the functional changes the one
that most concerns the present argument is the increased production of the fat-splitting
enzyme, lipase. Assume for the moment that the products of the last digestion have
been absorbed into the blood and have already been utilized by the tissues. Assume
also that this state has reached a point where the expenditure of energy must be done
by drawing on the body reserves. Then what can happen?

The salmon tissue glycogen is a negligible quantity. There is no adequate supply
in either muscle or liver, as in the mammalia. Glycogenesis can not, therefore, come to
the support of the body in this crisis.

There is an abundant store of fat in the intermuscular depot, great quantities of it,
and a lipogenesis ® comes to the support of the salmon in a way quite comparable to the
glycogenesis of the mammal as conceived by Claude Bernard. Under these conditions
the activity of the muscular tissue is directly dependent on the fat as a source of energy.
The muscle oxidizes fatty bodies in the salmon, just as it oxidizes carbohydrate bodies in
certain other well-known animals.

@ An exception may be found in the border zone of fibers between the pink and the dark muscle.
b Loevenhart: On the relation of lipase to fat metabolism, lipogenesis. American Journal of Physiology, vol. vi, rgor, p. 331.

19371°—vol 33—15 9

126 BULLETIN OF THE BUREAU OF FISHERIES.

It is to be assumed that the muscle fibers absorb the fatty bodies from the lymph
and blood, presumably as soluble fatty acids and glycerin. The fatty bodies of the
blood and lymph are derived from the stored fat by a process of lipolysis. To that
extent to which the store of intermuscular fat of the pink muscle is eroded by this
process of lipolysis will the percentage concentration of the cleavage products of the
pink muscle lymph be high. From the lymph the fat cleavage products dialyse directly
into the pink fibers and become available for oxidation. In the early stages of the fast
there are numerous tissues besides the muscle containing an excess of stored fat; the
digestive tube, the pancreas, the liver, the skin, etc., as well as the connective tissue and
the muscles. Loevenhart % has stated that the limits of lipogenesis are ‘‘nearly propor-
tional to the amount of enzyme acting”’ and “‘nearly independent of an excess of ethyl
butyrate’ in the experiments of his series. Hence, with increasing production of lipase
in the blood there is an ever-increasing percentage of fatty bodies thrown into solution
in the blood and lymph.

Hand in hand with the increase of fatty bodies in the blood and lymph will go an
increase in fatty products in the substance of the muscle fibers. Muscular oxidations are
not rapid enough to keep down the increasing quantity of fatty bodies, hence they will
diffuse through the muscle protoplasm in considerable excess. The lipase of the blood
and lymph will also diffuse into and be present in the muscle fiber, a fact demonstrated
for other muscle tissues. Under the law of reversible lipase action this excess of fatty
cleavage bodies is bound to be reconverted into and deposited as neutral fat. Thus arise
the chains of microscopic liposomes of the pink muscle at the beginning of the salmon
fast.

The number and size of the chains and of the individual liposomes in the pink
muscle, therefore, is a result of the interaction of a number of factors, chief of which are
the following:

a. The relative abundance of the stored fat in the tissues of whatever source, i. e.,
the gross amount of fat available for lipolytic erosion in all parts of the body.

b. The relative abundance of lipase throughout the body, chiefly of the blood and
lymph, but having origin in lipase producing tissues.

c. The structural and physical factors controlling the rapidity of the absorption
from the blood and lymph into the muscle fibers; i. e., the sarcolemma, sarcoplasm, etc.

d. Especially the rapidity with which the fatty bodies are utilized, oxidized, by the
muscle sarcoplasm.

The constants of lipase action have not yet been determined sufficiently to enable
one to apply to this specific instance definite governing laws. It is hoped that some-
thing may be accomplished along this line as this work progresses. At present, however,
one may say that in a general way e, the size and number of the liposomes in any given
fish’s pink muscle, will vary directly as a, the abundance of stored fat, b, the relative
abundance of lipase, c, the structural and physical factors governing the diffusion of the
lipolysed products, and inversely as d, the rate of oxidation of fats in the muscle fibers,
The relation may be expressed as follows:

axbXxclp_,
d

where k is a complex constant representing the unknown facts and relations referred

to above.

@ Lovenhart, op. cit., p. 350

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 277
TRANSFERENCE OF FAT IN THE DARK MUSCLE.

It is stated earlier in the paper that in the active feeding and growing salmon large
quantities of fat are laid down in the dark muscle fibers. This deposit of fat begins in
the earliest stage observed in the young salmon. It reaches its maximum somewhere
near the time when the salmon begin their migration journey. At the Ilwaco station
the amount of fat deposited in this type of muscle is astoundingly large. (See fig. 1,
pl. ur.) The amount is especially significant when it is remembered that the deposit
has taken place as a storage process in a tissue that is supposed to be most active in the
giving off of mechanical energy.

The variations noted in the dark muscle at the different stages in the migration
journey are variations in the amount and character of the distribution of fat. Extensive
discussion has been presented showing the facts as regards this picture at the different
migration stages. Attention is here called especially to two points, first, the striking
variation in the amount of fat of the dark muscle of different parts of one and the same
animal as given in such fish as no. 125 and no. 126. The second factor is the relatively
large amount of fat present as liposomes in the dark fibers at the death of the salmon.

As regards the first point, it is obvious that the diminished quantity of fat along the
courses of certain of the smaller blood vessels, as shown in fish no. 120, also no. 126,
represents a process of fat erosion. It would seem that the fat in the process of being
removed is taken up first along the course of the blood vessels. Apparently we have to
do here with the simple process of lipolysis. If this be the correct view then it is evident
that the fat products of the dark muscle are handled in a way analogous to the fats in the
pink muscle in so far as the process of solution and utilization goes. Therefore there is
nothing peculiar about this tissue in this particular regard.

In an animal in which the fats have reached a certain stage of consumption and in
which the processes of fat solution are going on rapidly we will have the greatest con-
trasts as between the highly vascular and the less vascular areas. The former favor in
every way the rapid solution and removal of the fats. In a comparatively large section
of dark muscle through, say, the trunk region of such an animal, one will notice a decided
mottled or marbled appearance of the section viewed under comparatively low magnifica-
tion. The less fat areas will be lighter, with less of the scarlet red stain, while the fatter
regions will be relatively deeper red in appearance. Often the light areas form distinct
patterns which conform to the smaller veins and arteries.

In the salmon at this stage the contrasts as between the trunk muscle and the caudal
muscle are always sharp. The dark muscle, like the pink muscle, will contain relatively
less fat in the caudal region than in the trunk region. The microscope will show that
this caudal muscle fat is in smaller droplets which are fewer in number than in the trunk
region. In the less vascular areas of the caudal region there will often appear fibers that
have only traces, sometimes no fat. These factors are attributed to the more rapid
using of fat for the production of energy in the caudal muscles as the more active tissues.

The utilization of the fats of the dark muscle does not present as acute a problem as
regards the numerous smaller liposomes which we found in the case of the pink muscle.
The salmon begins the fast with the dark muscle fully loaded with intracellular fat.
Therefore, the first change that will occur in this muscle will be a process of using up the
fat on hand in the cell. When at any time or for any reason this intracellular dark
muscle fat is wholly consumed, then the dark muscle will be in the same category as the

128 BULLETIN OF THE BUREAU OF FISHERIES.

pink muscle in so far as its source of material for the production of energy is concerned.
Regions of dark muscle which have reached this stage are found with the arrangement of
liposomes that is described as typical for the pink muscle. On the other hand, the dark
muscle will have the chains of relatively small liposomes rather uniformly distributed
throughout the muscle mass. ‘These liposomes at this stage are relatively small, as for
example in the fishes described from Cazadero. Rarely will fibers be found with no
liposomes. It seems to me that should a certain area of dark muscle fibers through
excessive activity consume all of its liposomes, then fat would be thrown into those fibers
by the process of lipolysis and fat transference in exactly the same way that it is thrown
into the pink fibers. This detail is fully described in connection with the discussion of
the pink muscle.

As regards the second factor mentioned above, namely, the high percentage of fat
still present in the dark muscle at the time of the death of the salmon, it seems to me
the matter is more complicated. The operation of no ordinary factor would maintain a
higher percentage of fat in the dark muscle at a time when the fats were almost consumed.
One is led to suspect that there is some special factor operative in the dark muscle. In
all probability this factor is the same in the late stage in the life cycle as that operating
in the earlier stage in the salmon development which results in the loading of fats into
this type of muscle. I have observed no special facts which of themselves explain this
situation. There are, however, certain accessory facts which permit of an explanation
which will be.offered as a tentative hypothesis. Of these facts the most important is the
fact of the loading of the dark muscle during the embryonic stage of its development.

Undoubtedly such deposits of fat as occur in no. 97, fig. 7, represent a perfectly normal
process which is to be interpreted as a function of this muscle. Histologically the dark
muscle differs slightly in its structure from the pink muscle. The dark fibers contain
more sarcoplasm and somewhat larger and fewer fibrille. At an early embryonic stage
this difference between the dark and pink muscle is rather more striking than it is later.
This suggests that the dark muscle is a less highly differentiated type of muscle than
the pink. One may assume, therefore, that it retains more primitive characteristics.
In the sections which cut the borderland between dark and pink muscle a few of the
pink fibers of the intermediate zone are found to be filled with liposomes. This loading
of liposomes is greatest in the fibers nearest the surface of the great muscle mass and is
totally absent in the deeper portions of the muscle. The fibers in question are of the
pink fiber type. Their loading of fat must therefore be due to some special factor.
These three facts, namely, (1) the excessive loading of fat in the growing dark muscle,
(2) the more generalized type of dark muscle, and (3) the tendency of the neighboring
zone of pink muscle to load intramuscular fat, all suggest that the dark muscle has still
strongly developed one of its general functions. This function is the production of
lipase. It may be anticipating a bit in the following discussion, but it is evident that
the presence of a relatively high concentration of lipase results in the seizing of the fats
during the growing stage and their concentration in the lipase-producing tissues. This
view is borne out by the deposit of large amounts of fat in the pancreas as well as in the
dark muscle. <A relatively high concentration of lipase in the dark muscle would lead
to a greater concentration of lipase in the tissue lymphs in those tissues which surround
the dark muscle, namely, the connective tissue of the skin and the superficial layer of

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 129

the trunk pink muscle. The deposit of liposomes in the intermediate zone of pink
muscle fibers is occasioned by this greater concentration of lipase.

If it can be granted that the dark muscle retains this assumed power of lipase
production throughout its whole life, then it will follow that at the spawning grounds
we will still have in this particular tissue more than the average amount of lipase in the
muscle fluids. If so, there will be a tendency to hold fats and fatty acid in solution
in this tissue, and, other things being equal, the tendency to maintain a somewhat
higher content of fat in the form of liposomes.

The liposomes will form in those tissue spaces in which there is greater stagnation
of the tissue fluids. In so far as the dark muscle is concerned, this point is immediately
under the sarcolemma. Should the cleavage products of fatty acid and glycerin
diffuse from the sarcoplasm through the sarcolemma, then it would be picked up ulti-
mately by the blood stream and washed away. If, on the other hand, fats are coming
into the pink muscle, then the diffusion will pass first through the sarcolemma and then
the sarcoplasm, which is using fat in its oxidations. Therefore, the central portion of
the muscle fiber will contain fewer liposomes and smaller ones while the superficial
portion and the region immediately under the sarcolemma will contain relatively larger
liposomes. This picture corresponds to the facts whether or not the theory offered in
explanation be true. The factors discussed in previous pages which determine the number
and size of the liposomes are therefore the same for dark muscle as for the pink, except
for one, namely, the factor b, given on page 126, the relative abundance of lipase in the
dark muscle. This factor b is greater than in the pink muscle, since the dark muscle
itself is presumed to be producing lipase.

RELATIVE ABUNDANCE OF STORED FAT AND ITS DISPOSAL IN THE ORGANS.

My chemical analyses have shown that there is always a great variation in the per-
centage of fat taken from any standard region of different individual salmon from any
particular station. No school of salmon contains individuals of uniform characteristics
as regards either size or condition, for example, the specific loading of fat. Considering
only the region under discussion here, namely, the mid-lateral portion of the body, I
find that there is great variation in the total amount of fat present. In the histological
examination of the variation of fats this will show itself in the number and especially in
the relative size of the fat droplets. The loading of fat, as indicated by these facts, is
most constant in the salmon obtained at the mouth of the Columbia River. It is to be
assumed that if one could obtain salmon just in the region where and at the time when
they cease feeding, then the loading of fat would be most nearly constant. Even then
a great variation might be expected, since in salmon of the same size it can not be
assumed that there has been uniformity of opportunity in obtaining food during the
long period of development. Therefore, there are great variations present in the total
loading of fat in the lateral pink muscle, even at the Ilwaco station, variations that are
the expression of a multiplicity of factors.

As salmon pass up the rivers in their migration and since the stored fat forms the
total source of the energy-giving material, it follows that there will be a diminution in
the total fat directly proportional to the length of time and the relative expenditure of
energy since the beginning of the fast.

Instances of variations in fat between the fibers of the lateral pink muscle at the
first station, Ilwaco, are to be obtained by reference to the protocols of fishes nos. 111,

130 BULLETIN OF THE BUREAU OF FISHERIES.

113, 115, and 118, representing very fat fish taken during early August, r911. The
commercial fishermen report that the very fattest specimens are obtained earlier in
the season. Fish no. 114 and especially no. 117 represent the opposite extreme,
i. e., the poorest fish for the August season. Undoubtedly no. 117 was at a stage in
which there was considerable diminution of fat below its maximum. ‘This diminution
may have been due to any one of a number of factors, but one is led to suspect that the
fish has been long without food, even at this low station on the river. The marking
experiments of 1908 @ give evidence that some at least of the August fish remain long
in the waters of the lower river.

It is self-evident that where a fish has a low percentage of stored fat to begin with
there will be a less abundant erosion by the action of the lipases; therefore, other things
being constant, one can not expect in such an animal as high a percentage of concentra-
tion of the fatty products in the blood and tissue fluids as in fish that contain a greater
fat content. This factor a, the variation of the storage fat, has its influence on the
factor e, the number and size of the liposomes, as is indicated by the variations noted

in fishes nos. 120 and 125.
SALMON LIPASES.

The rapidity with which the stored fat of the salmon tissues is eroded when the
fasting begins will, of course, depend chiefly upon the second factor mentioned in the
section on the transference of fat in the pink muscle, namely, the amount or concen-
tration of the lipases.

The phenomena of fat mobilizations in the body belong strictly within the group
of enzyme actions, for which the general law is so admirably stated by Wells:

All metabolism, then, may be considered as a continuous attempt at establishment of equi-
librium by enzymes, perpetuated by prevention of attainment of actual equilibrium through destruc-
tion of some of the participating substances by oxidation or other chemical processes, or by removal
from the cell or entrance into it of materials which overbalance one side of the equation.

The presence of lipase in various animal tissues has been demonstrated often
enough. Hanriot © demonstrates the presence of lipase in the following fluids and
organs: Blood, lymph, urine; liver, pancreas, testes, spleen, thyroid. The lipase was
in greatest amount in the blood, liver, and pancreas. Kastle and Loevenhart 4 showed
the presence of lipase in intestinal epithelium, which on account of its position as an
absorbing tissue is of more than passing interest. The evidence as regards the presence
of lipase in the blood plasma and in the lymph is somewhat contradictory, yet such
lipase is not only demonstrated, as given above, but its quantity has been shown to
vary under certain pathological conditions. Pathologists have followed the variations
in the quantity of lipase in necrosis* with fat formation. The pancreatic cells are
well-known lipase producers. It is to be expected, therefore, that there will be a
variation in the quantity of lipase that will reach the body fluids from this tissue. In
confirmation of this point lipase has been found in the urine of a clinical case of fatty
necrosis associated with inflammation of the pancreas./

a Greene, C. W.: The migration of salmonin the Columbia River. Bulletin U.S. Bureau of Fisheries, vol. XX1X, 1909, p. 131.
b Wells, H. G.: Chemical pathology, p. 68. Philadelphia, 1907.

¢ Hanriot: Comptes rendus de la Société de biologie, 1896, p. 925.

d Kastle and Loevenhart: American Chemical Journal, vol. Xx1v, p. 491, 1900.

e Flexner: Journal Experimental Medicine, vol. 11, p. 194, 1904.

f Opie, Eugene L.: A case of hemorrhagic pancreatitis. The occurrence of a fat-splitting ferment in the urine. Johns
Hopkins Hospital Bulletin, vol. 13, p. 117, 1902.

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 131

I have examined the pancreas of the king salmon for lipase, testing extracts of the
fresh normal glands of salmon caught in active stages of digestion. The experiments
were preliminary only, yet the tests were positive and the reactions vigorous. The salmon
pancreas secretes an active lipase. In my fat-absorption experiments there was always
a vigorous loading of the epithelial cells of the intestinal mucosa, especially in younger
fish. There was a greater mass loading than I have ever seen in mammalian intestinal
epithelium. This, by our current theories of fat absorption, is as strong circumstancial
evidence of the presence of lipase in the salmon mucous epithelium as one could well
expect to discover. Aside from these tests no studies have been made on the lipases
of any of the Salmonide. It is, of course, highly desirable that such studies should
be made. ‘The amount and the variations of the lipase content of the salmon blood, and
especially of the muscles, should be determined by a series of quantitative tests. These
are the tissues of peculiar interest to the problem in hand. There are other tissues
directly concerned in the fat metabolisms of the salmon—the liver, the divisions of
the alimentary tract, the pancreas, etc.

However, on a priori ground, there is every reason for assuming that the salmon
is well supplied with lipase in its fat metabolizing tissues, particularly in the alimentary
mucous epithelium (the gastric epithelium is also fat absorbing), the muscles, the liver,
etc. The blood and the lymph can not escape a lipase content in an animal in which
so many of the tissues are concerned in the lipolytic processes.

In previous discussions the conclusion has been reached that there is a marked
increase in lipolysis at the particular time the salmon cease feeding and begin the migra-
tion journey. This carries the assumption that there is at this time an increase in the
amount, i. e., percentage, of lipase in the blood and the tissues, including the muscular
tissues. Whence come these lipases?

SOURCE OF THE LIPASES.

The presence of a greater percentage of lipase when the cessation of feeding occurs
may be explained on two physiological grounds, both of which are probably active;
first, there may be an absolute increase in the lipase produced in a given time; and,
second, it is possible that with the cessation of feeding the lipase that is usually con-
sumed in the intestine and pyloric ceca in the processes of digestion and absorption
is now thrown more fully into the blood stream. To this extent it raises the percentage
content of lipase in the blood.

The pancreas.—The salmon pancreas is one proven source of lipase. The pancreas
is morphologically of the type described by Legouis? as the diffuse pancreas. The
gland filaments are quite separate from each other. They form an open meshwork
running over the pyloric ceca, the blood vessels and mesentery of the stomach and
intestine, the inner loop of the stomach, and the mesentery of the spleen.

These pancreatic filaments are richly supplied with blood vessels, and these vessels
anastomose with the blood vessels of the ceca, intestine, etc., of the region. Pancreatic
ducts have been described by Legouis as converging to a common duct that enters the
intestine either with, or in the neighborhood of, the bile duct.

@ Legouis, P.: Recherches sur les tubes de Weber et sur le pancreas des poissans osseau. Annales des Sciences Naturelles
Zool., sth ser., t. 17, 1873. See also t. 18, 2d article.

132 BULLETIN OF THE BUREAU OF FISHERIES.

My histological observations, which will be given in detail in a later paper, show that
the pancreatic cells are not all compactly arranged in acini, as is usually the case, but
that they are more or less scattered. Every cross section of the pancreas reveals the fact
that a large amount of adipose tissue is present in association with the pancreatic tissue.
The cross section of a pancreatic lobe is somewhat triangular in shape. The central
portion of the triangle in an adult organ always contains one or two blood vessels of
relatively large size, together with a considerable amount of fatty tissue. The pan-
creatic cells are arranged around the surface of the gland and in the interstices among
the fat cells and blood vessels. There are many acini where definite arrangement of
pancreatic cells around a central lumen can be shown. This lumen rarely forms a space
as large as one sees in the corresponding region of a mammalian pancreas. In these
groups, in favorable preparations, that portion of the pancreatic cell bordering on the
lumen is highly granular and the granules stain in a way characteristic of zymogen
granules. The pancreatic cells around the surface of the gland and in many portions of
the deeper region have a rather scattered arrangement in which it is extremely difficult
to show definite relation to ductules. Such relation is questionable in a large proportion
of the gland. The diminutive size of the pancreatic ducts, the presence of the large
amount of fat in the gland, the rich vascular arrangements of the gland, and the diffuse
arrangement of the gland cells of the salmon pancreas have led me to the conclusion that
the secretion of this gland is largely internal, i. e., that the pancreatic secretion is largely
thrown into the lymph and blood stream of the organ. In this statement it is not meant
to minimize the importance of the gland in the production of the pancreatic juice.
Rather, the intention is, to emphasize the importance of the internal lipase secreting
function.

When digestion stops and the animal begins its fast, the pancreatic gland undoubt-
edly continues to function. This is shown especially by the histological appearance at
different stations of the migration journey. There is obviously no digestive function
to be accomplished by the secretion during the fast. Therefore, during this phase of the
life cycle the internal secretive function becomes the main, in fact, the only one. The
vascularity of the pancreas and of the pyloric caeca does not change in proportion to
the amount of retrogressive change shown in the ceca and in the rest of the alimentary
tract. Putting these facts and deductions together they may be summarized thus:

1. The pancreatic gland is abundantly active during the migration fast.

2. The activity consists in the production of an internal secretion which is not
essentially different in character from the normal secretion produced at an earlier stage
in the life cycle. It is therefore rich in lipases.

3. The pancreatic lipase produced during the fasting period is chiefly, if not wholly,
discharged into the tissue spaces, from whence it reaches the blood stream.

4. The circulation carries the lipase to the tissues of the body, which includes both
the fat-storing tissues and the active fat-using muscles now under especial consideration.

Lipase from the granule cell layer—There is a second tissue filled with zymogen
granules which I believe to be an active source of lipase, namely, the granule cell layer
of the alimentary tract. I have already called attention to the presence of a layer of
granule cells in the mucous coat of the alimentary tract. This layer is especially richly
developed in the pyloric ceca. In every section of the retrogressing caecum, one is
strongly impressed with the continued large size of this granule cell layer. Even in

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 133

fish from the Ilwaco station, specimens which represent the first stage in the changes fol-
lowing the beginning of the fast, this layeris notably large. The granule cell layer is much
thicker and more prominent than in the normal Monterey salmon tissue. I have
made computations that indicate an actual increase of the mass of this granular layer.%
The greater apparent increase in mass may be explained on the ground that the nor-
mal volume of the granule cell layer is retained while the volume of the various other
structures of the alimentary tract is sharply diminishing, a question that is now under
investigation. The structure of the cells in this layer does not materially change while
the other tissues around it are sharply degenerating. The cells remain loaded with
granules which have a strong affinity for the basic dyes. The loading of granules
remains characteristic of the cells even in the most degenerated salmon examined.
Here, again, one must interpret the granules as zymogen granules. Gulland ” has given
the name eosinophile leucocytes to cells of this region, an identification to which I can
not subscribe. By implication he would ascribe to the granule cells a very different
function from that which appears to me to be the true one. The granules, which do stain
sharply with eosin as Gulland observes, are too large to be compared with the granules
of eosinophiles. The granule cells themselves are larger, differently arranged, and have
a very different type of nucleus from the typical eosinophile. Also one never finds
granule cells of the type characteristic of the cells of this layer in the salmon blood
vessels of either the ceca or of the intestine. The matter is more fully discussed in
another paper, but the points are mentioned here in order to meet objections to the
view which is offered for the function of these cells; namely, the granule cell layer of
the pyloric ceca and of the alimentary tract of the salmon is an internal secreting organ
which has for its probable function the production of lipase.

In the normal organ, when active digestion is going on, the lipase produced within
the granule cells is thrown out into the tissue spaces. Some of this lipase quite probably
reaches the epithelial layer and supplements the lipase produced by the epithelial cells.
Such lipase will, of course, facilitate the absorption and the resynthesis of the fats in
the epithelial cells, according to the laws of fat absorption. A considerable portion of
this layer lies outside the stratum compactum. It seems to me improbable that the
internal secretion of those cells will diffuse through the stratum compactum with the
same facility that it will pass out into the tissue spaces of the adjacent muscle coats.
The granule cell layer itself is almost free of blood vessels. The muscle coats, on the
other hand, are richly supplied with blood vessels. It would follow that any secretion
diffusing into the muscle coat would quickly be taken up by the blood, and would be
carried throughout the body. During the fasting period this lipase-producing function
of the granule cell layer is no longer of use in the absorption of fats but is of great sup-
plementary aid to the pancreas in maintaining an adequate supply of lipase in the
blood. This offers an explanation that would account for the persistence of the organ.

Increase in lipase from change in isotonicity of the tissues.—In the recent brilliant work
of the Rockefeller Institute in isolating and growing tissues of pure culture, it has been
shown that physiological activity, as measured by growth, is influenced by the degree
of concentration of the nutritive fluids. Tissues growing in physiological solutions of

a These determinations were made by computing the cross-sectional area of the granule cell layer of ceca from Monterey and
from Ilwaco salmon. ‘The Ilwaco specimens exceed the Monterey by an average of 11 percent. But the probable error is high.

b Paton, Noel: Life history of the salmon. Article 3. The minute structure of the digestive tract of the salmon, and the
changes which occur in it in fresh water, p. 13, 1898.

134 BULLETIN OF THE BUREAU OF FISHERIES.

slightly lower tonicity than that to which the tissues are adapted, undergo a more
active growth than those in fluids of relatively high tonicity. These factors have been
worked out on several tissues, but very strikingly on the rate of healing of wounds in the
skin of frogs which had their body fluids decreased in tonicity by injections of water.*

I have no doubt that these principles are operative in the salmon during the migration.
The salmon tissues in this stage of the life cycle are most of them long past the period
where growth is the prominent physiological activity. There is, however, one marked
exception, namely, the reproductive organs. These organs are quite immature at the
time the migration begins. They suddenly take on an active stage of growth which
proceeds readily till complete development at the time of spawning. Other tissues
show their variations in physiological activity in ways other than growth, the muscular
tissues in increased production of energy, the glandular tissues by variation in secre-
tory activity, ete.

In work on the Sacramento River? I showed that there is a marked decrease in
the depression of the freezing point of salmon serum as the fish migrate from the sea to
fresh water. The depression of the freezing point is a direct measure of the tonicity of
the blood. As this diminution begins with the process of migration from salt water to
fresh water it is evident that it will have the same type of influence on the tissue
activities of the salmon as that shown in the experiments on tissue growth. This is
undoubtedly a factor in the stimulation of the reproductive organs to the sudden
increased growth which takes place in the salmon at this time. It is a safe inference
that this change in isotonicity of the blood and tissue fluids if exerted on other
tissues of the bodies will cause variation in their physiological activity. Applying this
principle to the pancreas and to other tissues where lipase is produced we have a factor
accounting for an increase in metabolites, of which lipase is one, at the critical time of
the beginning of the migration.

Lipase from tissue degenerations.—Another source of lipase may arise through tissue
degenerations. Extensive degenerations are taking place in the alimentary tract.
At Ilwaco the mucous membrane of the intestine and pyloric ceca has already reached
a considerable degree of disintegration. On the current theories explaining degenerative
change one may assume that these cells have already passed through a stage of physio-
logical hypertrophy. Kastle and Loevenhart have shown that these cells are active
lipase producers in the normal condition. Therefore, one may assume that the inflam-
matory condition preliminary to the actual necrosis is associated with an increase in
lipase production. The disintegration of the cells and especially the chromatolysis is
the final step in the process. These pathological changes accelerate the physiological
production of lipase in the mucous epithelium of especially the pyloric caeca just at the
time when such an increase in lipase is of most value to the salmon, namely, at the
beginning of the fast. Lipase from this source would almost cease at a quite early stage.
However, the epithelium is rarely wholly disintegrated.

Lipase from the liver and other tissues.—There is evidence that other tissues of the
salmon are more important in producing lipase. Of these one may mention the liver,
whose function in this regard will be presented in another paper.

a Ruth, E. S.: The influence of distilled water on the healing of skin wounds in the frog. Journal of Experimental Medicine,

vol. 13, P. 422, 1911.
> Greene, C. W.: Physiological studies of the Chinook salmon. Bulletin U.S. Bureau of Fisheries, vol. XXIV, 1904, PP. 446, 449+

STORAGE OF FAT IN MUSCULAR TISSUE OF KING SALMON. 135

Also the group of dark muscle fibers represented in the musculus superficialis lateralis
has already been mentioned as lipase producing tissue. But when all the less important
sources which have been discussed are left out of account there still remains an adequate
lipase producing mechanism in the pancreas and in the granule cell layer of the alimentary
tract to account for the presence of sufficient lipase in the blood and tissues to meet
the need of the fat transference that we have under discussion. Histological evidence
has been given to show that there is no diminution of the activity of the pancreas at
the inauguration of the fasting period. If the pancreatic lipase production even remain
constant then the amount of lipase which this gland will produce as an internal secretion
will tend to raise the total lipase of the blood and tissues. The lipase that is consumed
in the process of digestion during the feeding period will now be left to be thrown into
the circulation. It follows that there will be an increase in the percentage of lipase in
the blood, therefore, according to Loevenhart, an increased solution of the fats with
which this lipase comes in contact. These fats are the stored fats. An increased solu-
tion of the stored fats will raise the fatty acid and glycerin content of the tissue fluid
and the blood. The inevitable result will be an increased supply of these fat cleavage
products to the active muscular tissues. This supply will diffuse through the muscle
spaces, the sarcolemma, and throughout the sarcoplasm of the muscle fiber in an ever
increasing quantity. Since the relative amount of activity of the muscles can not be
assumed to change, i. e., is comparatively constant, it follows that the percentage
amount of fat will increase within the active muscle fibers.

It is shown on page 81 that the pink muscle fibers contain no intramuscular fat
during the feeding stage, or at most, only a trace of such fat at maturity. This is only
another way of saying that the consumption of fatty substances in the muscle fibers of
the feeding salmon is in balance with the fatty bodies penetrating the fibers. There is
never a sufficient excess of fatty acid and glycerin within the fibers to produce resynthe-
sis and deposition of the fat in visible form. But with the increasing percentage of
these substances penetrating the fiber after the fast begins there will be a synthesis of
neutral fats and these will be deposited and can be identified. The liposomes present
in the lateral pink muscle of the salmon taken at Ilwaco represent such deposits that
have taken place since the beginning of the migration. The amount of neutral fat
present in the pink muscle fibers is a measure of the excess of fatty acids and glycerin
brought into the fibers over those oxidized in the muscular activity. If oxidation
diminishes, then fats will be deposited and the excess is expressed in the number and size
of the liposomes (fig. 8, pl.vr).

The character of the liposomes, that is, their number, size, and arrangement in the
pink muscle depends also on one other very different group of factors. This is the struc-
ture of the muscle (fig. 13, pl. vimm). That the fat is laid down in chains of liposomes of
the minute sizes that have been described must depend largely upon the structural
arrangement of the fibrillae and of the interfibrillar sarcoplasm. It is not desired, how-
ever, to discuss this factor beyond merely calling attention to it.

136 BULLETIN OF THE BUREAU OF FISHERIES.

/ th
RESUME.

The points made in this investigation that call for special mention may be cate-
gorically stated as follows:

1. Fat is the prominent and immediate source of the energy of the salmon
expended during the spawning migration.

>. The salmon fat is stored in the body during the stage of feeding and growth,
and reaches a maximum at the time the feeding stage ends, i. e., at the beginning of
the migration fast. This fat can not in any proper sense be looked upon as a fatty
degeneration.

3. The fat storage tissues are primarily the muscles and intermuscular connective
tissues. Storage tissues of minor importance are the cutaneous and other adipose
tissues, the liver, the alimentary tract, and the skeleton.

4. There are two distinct and characteristically different types of muscle—the
superficial lateral or dark and the deep lateral or pink muscle. ‘The latter represents
the major portion of the great lateral muscle mass.

5. The pink muscle is characterized (a) by the enormous load of fat between the
fibers, intermuscular fat, and in the myocommata at the time of maturity; (b) by the
great variation in the size of its fibers.

6. The pink muscle fibers have no intramuscular fat, or at most only traces of fat,
during the feeding stage.

>. Immediately at the beginning of the spawning migration the pink fibers are
loaded with numerous chains of very small liposomes. This loading of liposomes
increases during the early stage in the journey, and then decreases somewhat up to the
spawning time. The fat never wholly disappears even in dying salmon.

8. In the active caudal pink muscle the liposomes are much less constant and are
often completely absent as advanced stages of exhaustion appear.

9. The pink muscle fibers are plump and cylindrical at the time the migration
begins. But at the spawning time the larger fibers have the appearance of being
shrunken by decrease in mass. They become polygonal in cross-sectional outline.
The sides of the polygon are often concave to the exterior, as if compressed by the
adjacent smaller fibers.

to. The dark muscle is characterized (a) by the enormous loading of intramuscular
fat at all stages of the life cycle, but especially at the time the spawning migration
begins; (b) by the relatively small and uniform size of the fibers.

tr. The stored fat of the dark muscle is gradually eroded during the migration
until the fat reaches a quantity and distribution comparable to but still greater than
that in the pink fibers. The fat is never completely eroded and is present in considera-
ble quantity at the death of the salmon after spawning.

12. The smaller muscles of the fins and of the head of the salmon take little part
in the fat storing. The food supply of these muscles, however, is the same, namely,
the fats.

13. Distinct degenerative changes were found in the adductor mandibule muscle
of a spawned male at the dying stage. This degeneration is a simple atrophy with
pigmentation.

EXPLANATION OF PLATES.

The drawings presented were all made from camera lucida outlines. Fat is repre-
sented in the characteristic red color obtained by the scarlet red method of staining fat.
All the drawings and outlines were made for me by Mr. George T. Kline, biological artist

of the University of Missouri.
' PLATE III.

Fic. 1. The transparency of a segment of dark muscle fiber of salmon no. 115 from the mouth of
the Columbia River, Ilwaco, Wash. The most superficial liposomes and fat droplets are represented
somewhat darker, while the paler colored droplets are deeper in the fiber. Magnification, Leitz ocular
2, objective 7.

Fic. 2. A small segment of dark muscle fiber from a young salmon, from the Columbia River,
Warrendale, Oregon. Magnification, Leitz ocular 3, objective 7.

Fic. 3. Section of trunk dark muscle of salmon no. 120, adult in prime condition from the Colum-
bia River at Warrendale, Oreg. The amount of fat present is almost as great as in the Ilwaco fish no.
r1r and no. 115. Magnification, Leitz ocular 2, objective 7.

PLATE IV.

Fic. 4. Trunk dark muscle of salmon no. 126 from the Columbia River at Warrendale, Oreg. This
salmon is representative of a late stage in the fat removal from the tissues. Certain fibers near the
large blood vessel to the right are free of all but the smallest liposomes. Other fibers are still well sup-
plied with fat. Magnification, Leitz ocular 2, objective 7.

Fic. 5. Dark muscle from salmon no. 138, a spawning salmon from the Clackamas River, Cazadero,
Oreg. This figure represents the latest stages in fat removal from the trunk dark muscle. Magnification.
Leitz ocular 2, objective 7.

PLATE V.

Fic. 6. Transverse section of dark muscle from an exhausted, naturally spawned salmon no. 108,
McCloud River, Baird, Cal. This figure represents the extreme exhaustion of fat from the dark muscle.
The salmon was an enormous male which was taken just at the time of natural death. The fat is in
finest liposomes condensed at the surface of the fiber but absent between the fibers. The representa-
tion of the size of the liposomes is somewhat strong. Magnification, Leitz ocular 2, objective 7.

Fic. 7. Dark muscle from young fish no. 97. The preparation is a paraffin section stained with
Mallory’s analine blue connective tissue stain. The figure presents well the excessive number of clear
spaces which represent vacuoles produced by extracting the fat in the imbedding process. One fiber
has recently divided longitudinally into two. This fiber shows no fat along the new portion of sarco-
lemma. Magnification, Leitz ocular 3, objective 1/12. (From American Journal of Anatomy, vol. 13,
1912, p. 175.)

PLATE VI.

Fic. 8. Trunk pink muscle of salmon no. 118 from the mouth of the Columbia River, Ilwaco, Oreg.
Attention is called to the great variation in the size of the fibers, to their characteristic outlines, the
great amount of fat between the fibers, and to the general distribution and extreme fineness of the
liposomes in the fibers, which have come out rather too strong in the reproduction, many of them
being actually just perceptible. This figure without the liposomes in the muscle fibers would represent
the normal condition of the salmon pink muscle at the beginning of the migration fast. Magnification,
Leitz ocular 2, objective 4.

Fic. 9. Segments of two trunk pink fibers with adherent intermuscular fat drops from salmon no.
118. Magnification, Leitz ocular 2, objective 4.

PLATE VII.

Fic. ro. Trunk pink muscle of salmon no. 126 from the Columbia River at Warrendale, Oreg. This
salmon is the one presented asa typical poor condition fish. The intermuscular fat is reduced to groups
of droplets in the stronger connective tissue septa. The intramuscular fat is extremely low, limited to
the smallest and medium sized fibers. These fibers retain their normal histological structure as
shown in figure 13. Magnification, Leitz ocular 2, objective 4.

137

138 BULLETIN OF THE BUREAU OF FISHERIES.

Fic. 11. Trunk pink muscle from salmon no. 132, a spawning female from the Clackamas River,
Cazadero, Oreg. The intermuscular fat is practically eliminated, yet all the fibers except the largest
show a considerable sprinkling of liposomes. In the small fibers these droplets are quite uniformly dis-
tributed, in the medium fibers concentrated around the surface, and in the largest fibers present only
in traces atthe surface. The outline of the largest fiber to the upper right-hand side of the figure indicates
that it is approaching a degeneration stage, though the microscopic fibrillar structure is still normal in
appearance in this particular fiber. Magnification, Leitz ocular 2, objective 4.

PLATE VIII.

Fic. 12.—Cheek muscle of salmon no. 140, a spawned male from the Clackamas River, Cazadero, -
Oreg. Fat is present in a few groups of small droplets in the connective tissue septa. There is no intra-
muscular fat. One fiber in the center of this group is in an advanced stage of atrophy with pigmentation,
shown in the granules of this fiber (not to be confused with similar appearance of liposomes in other
figures). The three fibers to the right of this pigmented one show the first stages of degeneration rep-
resented by a swelling and blending of the fibrille. This detail of structure is not shown in the figure.
Magnification, Leitz ocular 3, objective 4.

Fic. 13. A highly magnified portion of a trunk pink fiber from salmon no, 126, Columbia River,
Warrendale, Oreg. This small segment of a medium-sized pink fiber shows the normal fibrillar
arrangement. The amount of fat present is indicated in figure 10, plate vi. Traces of fat were pres-
ent in this particular segment just under the sarcolemma and between the outer series of fibrille.
This figure is offered in evidence as showing that the elimination of fat from the pink muscle is not
accompanied by any immediate breaking down or degeneration of the finer structure of the tissue.
Magnification, Leitz, ocular 3, objective 1/12 oil immersion. Camera lucida outlines.

PLATE IX.

Fic. 14. The trunk dark muscle of young salmon no. 97 from the McCloud River, Baird, Cal. The
muscle fibers are drawn in outline to show the compact arrangement and relative size of the fibers as
compared with the adult. One particular fiber in this figure showed an exceptionally large fat drop in
the middle of the fiber. Magnification, Leitz ocular 4, objective 3.

Fic. 15. Dark trunk muscle of salmon no. 126 from the Columbia River at Warrendale, Oreg.
Drawing to show the outlines of the fibers of the adult fish after the fat is largely removed. This figure
should be compared with the preceding. Magnification, Leitz ocular 4, objective 3.

PLATE X.

Fic. 16. Outline of the trunk pink fibers of the young fish no. 97 from the McCloud River, Baird,
Cal. The figure shows outlines of the fibers at a stage in which active growth is taking place. The large
number of relatively small fibers have recently split off the larger in the process of fiber multiplication.
Magnification, Leitz ocular 4, objective 3.

Fic. 17. Trunk pink muscle fibers from adult salmon no. 118 from the mouth of the Columbia River
at Ilwaco, Wash. ‘The outlines of the fibers show the relative symmetry of the adult prime condition
muscle. The separation of the fibers is due to the loading of fat in the interstitial connective tissue.
Should be compared with figure 8, plate vr. Magnification, Leitz ocular 4, objective 3.

PLATE XI.

Fic. 18. Trunk pink muscle from salmon no, 122 from the Columbia River, Warrendale, Oreg.
This outline figure shows the more compact arrangement of the fibers of pink muscle that has lost most
of its intermuscular fat. The fibers themselves are normal in outline. In this fish the pink fibers in
general seem somewhat smaller in size than the average for adult fish of mature size. This point should
be kept in mind in comparing the absolute size of the fibers shown in this figure and the preceding.
Magnification, Leitz ocular 4, objective 3.

Fic. 19. Trunk pink muscle from fish no. 140, a spawning male from the Clackamas River, Cazadero,
Oreg. The outlines of the fibers shown in this fish are typical of the stage just before natural death. The
larger fibers do not show any unquestioned structural signs of degeneration, though they have the
decrease in plumpness. Magnification, Leitz ocular 4, objective 3.

IME

PLATE

1913.

So Bik;

U.

BULL.

es

se @
G

R Ya+0e

PLATE IV

Flor 3:

S. B.

Bui Or

Burin Sa banks

} LOL:

PLATE V.

‘Z Win iy
= mpl TN

a
sam
FER

ee \

Fit J

we wes ees
Nii I
x a MUS
ws a =
a

=> LY,
; es
ins

Zip ey B
= elie

“Aires?
U

ei eS

:s

vl

A

Saif

TES
SRD >

PEATE Val.

PLATE VII.

Wyatt. Wie Se Ie Ie see

PLATE VIII.

1913.

Ione, Wins 2sinw de.

BIG. 12:

I

j}

satis
fot
YJ EY)
. aut

\\

: sets
cases
roaws
maces

SO
a

2
>

\“
=
7

FIG. 13.

7 7a 7

%&

Periay
spegiar
isf

PEATE TX,

Towson Be ie, Lorss

Buri Us os berks tos: IN
ya ee oe i : a ey
is 29 of )? - ae
DOO ny,

O
NOES QU: VRC
Bal bp 4e

; O oo eo
INE

Seciopeye.
a & al ; 4 oN
OO

D) GZ if
B® Se a A
Fic. 16.

PrAre Ile

CORRELATIONS OF WEIGHT, LENGTH, AND OTHER BODY
MEASUREMENTS IN THE WEAKFISH, CYNOSCION REGALIS

&*

By William J. Crozier and Selig Hecht
College of the City of New Y ork

iéteue

‘ r\ 9)! Les t lat
Pe pr foe

CORRELATIONS OF WEIGHT, LENGTH, AND OTHER BODY
MEASUREMENTS IN THE WEAKFISH, CYNOSCION REGALIS.

ad

By WILLIAM J. CROZIER and SELIG HECHT,
College of the City of New York.

ot
INTRODUCTION.

During July and August, 1912, an opportunity was afforded at the United States
Fisheries Laboratory, Beaufort, N. C., to make some studies on the correlation of exter-
nal characters in the squeteague; the results are contained in the present paper.*

The weakfish, or squeteague (Cynoscion regalis) is common in Beaufort Harbor, and
during July and August was taken almost every day from the pound net operated by the
laboratory, in quantities up to 300. By far the greatest number of these fish were about
31 cm. long. The specimens used were therefore, to a certain degree, selected according
to length, with a view to having a series covering as large a range as possible. Inasmuch
as the squeteague is known to spawn in late spring, physiological disturbances due to
spawning are negligible. All the fish examined (over 400) were either “spent” or unripe;
so we are sure that none of the weights recorded are influenced by the ripening of the
gonads.

The material was brought from the pound in a live car and immediately removed to
the laboratory. Measurements were made as rapidly as possible, the time for the com-
plete measurement of a single fish rarely occupying more than five minutes. The possi-
bility of shrinkage and of loss of weight through evaporation was carefully considered.
To check this a number of fish were weighed and measured at 11 a. m., placed in a bucket,
and covered (with a towel), our usual procedure, and four hours later no difference in
measurements could be detected.

CORRELATION OF WEIGHT AND LENGTH.

For the determination of the relation of weight to length, 390 fish were examined. Of
these 274 were females, 111 were males, and 5 were too immature for sex identification.
By length is meant total length, from tip of mandible of the closed mouth to the extreme
end of the caudal fin. This was taken by placing the fish on a board, its body perpendicu-
lar to and the tips of its tail just touching a raised end piece. The length was read by
means of a centimeter scale along the line from the mandible to the base of this end piece.
“Weight”? means weight after the surface water and mucous have been removed with a
towel, and is corrected for the weight of the stomach contents. The weighing was done
on a platform balance sensitive to 0.1 gm.

@ We wish to thank Dr. J. F. Abbott, of St. Louis, for his advice in the biometrical treatment of the data; we are also indebted
to Dr. A. J. Goldfarb, of New York, for his suggestions during the course of the work.

b Paton, D. Noél (Report of the Investigations on the life history of the salmon in fresh water, Fisheries Board for Scot-
land, p. 1, 1898), for example, notes that in the European salmon, during April and May, the ovaries are r.2 per cent and the testes
0.15 per cent of the total weight of the fish, whereas in November, near the spawning period, they represent 23.3 and 3.3 per cent
of the total weight, respectively.

19371°—vol 33—15——_10 141

142 BULLETIN OF THE BUREAU OF FISHERIES.

The results are shown graphically in figure 1, where length is abscissa, and weight
ordinate. A large number of the points represent duplicates, triplicates, and even
quadruplicates, and in many cases include both sexes. For example, the point (24.5,

WEIGHT—GMS.

LENGTH—CMS.

Fic. 1.—Showing relation of weight to length in 390 fish.

135) represents 2 females and 1 male; the point (41.5, 680) represents 2 males and 1
female; the point (27.0, 170) represents 2 males and 1 female; and the point (29.0, 220)
represents 2 males and 2 females. From the distribution of the points about the smoothed

—

BODY MEASUREMENTS IN_THE WEAKFISH, CYNOSCION REGALIS. 143

curve, it is clear that sex does not influence the relation between weight and length.?
This does not mean that there is no difference in the weight and length of Cynoscion
regalis of different sexes for the same age; it means that for a given length or weight of
fish sex does not affect the correlation.

The regularity of the curve shown in figure 1 enables its mathematical equation to
be calculated with considerable accuracy. Comparing the abscissas and ordinates of
any two points on the curve, we find that the weight varies as the third power of the
length. The equation therefore will be of the form y=a x*, in which y represents weight,
x length, and a is a constant. the value of which depends on the units used. When length

WEIGHT—GMS.

139 5go 430 S90 739 900 ia 1Zoo 1350 '390 1650 18099 1950 a0 2250 2400 2$50
1% 300 450 6900 i a im = 1350 1500 1650 1600 1330 Zico 2150 2400 2550 ajoo

err it ie
Bee CELE
BEEeZ 2a eaes ||
i | lestesty | | | |

a
ae
ied
Ea
a
Ee
fs

LENGTH—CMS.

|
|
|
|
|
|
|
|
|
|
|

Ieee
|| ef ee
PSS S[e 1s] 8 [eis] 8 EIEN

IRAAEAVAA

Fic. 2.—Correlation table showing closeness of relation between weight and length of fish represented in figure r.

is measured in centimeters, and weight in grams, a has the value 0.008771 + 0.000117,
and the equation becomes, weight = (0.008771 +0.000117) (length) *.

From this it is apparent that the weight in grams of a specimen of Cynoscion regalis
may be obtained by multiplying the cube of its length, in centimeters, by approximately
0.009.

Paton, Fulton, and other investigators® refer to the approximation with which the
weight of fishes vary as the cube of their lengths, but they present no comparable evi-
dence from which such conclusions can be derived with any degree of accuracy.

@ Kellicott, William E. (The growth of the brain and viscera in the smooth dogfish (Mustelus canis), American Journal
of Anatomy, vol. 8, p. 319, 1908), has shown that in the smooth dogfish, the sexes can not be distinguished with respect to either
absolute or relative weights of internal parts, except the gonads.

b Fulton, T. Wemyss: On the rate of growth of fishes, 24th Annual Report of the Fishery Board for Scotland,r90s, Pt. III,
and in other reports. Paton, D. Noel: loc. cit., p. 6. Williamson, Charles H.: On the herrings of the Clyde and other districts,
27th Annual Report of the Fishery Board for Scotland, 1908, Pt. III.

144 BULLETIN OF THE BUREAU OF FISHERIES.

As a measure of the closeness of the relation between weight and length, we have
determined @ the coefficient of correlation 7, which is the index of relation between two
variables, such that the amount of variation in one is a measure of the amount of varia-
tion in the other. Using length as the type and weight as the array, the correlation table
(fig. 2) was constructed. In each square is given the number of specimens which fall
within the weight group and length group indicated. From this arrangement of the
data the coefficient of correlation is found to be r=0.952, with a probable error of +
0.0032. Remembering that unity represents a theoretically perfect correlation, it is
apparent that this coefficient indicates an extremely high correlated variability.

RELATION OF BODY MEASUREMENTS TO TOTAL LENGTH.

With a view to discovering the relation between the dimensions of the external parts
of the fish and its total length, a series of measurements was taken on 123 of the 390
specimens used in the work discussed above. Of these, 80 were females and 43 males.

Fic. 3.—Showing parts of fish measured for comparisons plotted in figure 4.

Referring to the diagram, figure 3, the measurements, in addition to total length and
weight, were:

1. Standard length, from tip of snout to end of last caudal vertebra.

2. Head length, AB, from tip of snout to end of opercular bone, i. e., excluding
the opercular flap.

3. Body length, BD, from the end of the opercular bone to a point on the lateral
line immediately below the posterior limit of the base of the soft dorsal fin.

4. Tail length, DE.

5. Body width, taken at the point C on the line AE, immediately below the origin
of the spinous dorsal.

6. Depth, GF, from the origin of the anal fin, G, to F, on a line perpendicular to
the long axis of the fish.

For the depth measurement, 73 specimens were examined; of these 49 were females
and 24 males. ?

@ Davenport, Charles B.: Statistical methods, with special reference to biological variation, ch. 4, New York, 1904.
b As shown above, and also by the plots in figure 4, sex is a negligible factor in a discussion of this data.

BODY MEASUREMENTS IN THE WEAKFISH, CYNOSCION REGALIS. 145

The lengths were measured by means of a centimeter scale placed on the fish;
width and depth were taken with the aid of spring calipers, using the same scale. For
the measurement of width and depth it was necessary to secure points that would not
be influenced by the amount of food material in the stomach. The abdomen of the sque-
teague is extremely elastic, and its volume varies considerably with the stomach con-
tents. Significant measurements in this region of maximum depth are likewise impossible
after the removal of the contents of the stomach. The places selected fulfilled the
requirements suggested and were found to be sufficiently near the maxima for our
purposes.

The curves shown in figure 4 were derived from the data obtained. For every
specimen total length was plotted as abscissa and the other measurements detailed
above as ordinates.’ From the resulting straight lines it is at once apparent that there
is a simple relation between the dimensions of the external parts of the fish and its total
length. Tt is clear that with increasing length there is a constant, directly proportional
increase in all the body measurements taken.

From the slopes of the lines the rates of growth of the corresponding parts relative
to the growth of the total length may be calculated. Using the units shown on the plot,
the “tangent” of any line is determined by dividing the vertical distance between two
points on this line by the horizontal distance. These tangents are as follows:

Standardplengthstrer races tature ocr. oo cerns tony tro nin area wisietea capes nese aaa o. 840
BO Ghiyseateveps senate rcparsesy sts eusicentataay stata Daaslstaharwta-a cs) gpatevslea yeaa © susleveouste guartie a ee 530
Laney see eee cei states ike asics tind Arasmaisiave toa Lee hs Unie rae keh celass 273
ET eae cote et tatePe eee abe eect ty chat ieie: si esrn Shores SRV n crave ac ansin ous taeht a: eens deeies 215
IDjajdel hers a Sg a oe ott a On aati ere OR Re Sse i ete ae teen arene ese 135
IW dither sass nave oy scrsrersictoniarsrany sexe avecensunrers tretig Mise IES gee aaele Fee ey. ELS

From this it is obvious that, of the body parts, the body has by far the greatest rate
of growth, while the width has the least. It is also clear that the head and tail have
approximately the same rates of growth, and that the depth and width also grow at
about the same rate. It is, of course, to be understood that when the “rate of growth”
is mentioned, we do not mean “rate’’ with regard to time, but relative growth per unit
increase in total length. Thus, for every 10 cm. increase in total length the standard
length will increase 8.40 cm., the body 5.30 cm., the tail 2.73 cm., the head 2.15 cm.,
the depth 1.3 cm., and the width 1.15 cm.

RELATION OF BODY MEASUREMENTS TO WEIGHT.

From the regularities shown in the previous section we may conclude that there
exists a relation between any body measurement and weight similar to that which
exists between total length and weight. Yet another relation, however, may be demon-
strated. Since depth and width are each equal to a constant multiplied by the total
length, we may substitute in the formula for the derivation of weight,’ depth, and
width divided by their respective “tangents,” and thus secure a formula for the weight
in terms of length, width, and depth. This formula is W=k./. w. d. By direct calcu-
lation from figures 4 and 1, k=0.5513+0.0088, and the equation becomes weight =
(0.5513 0.0088) (length) (width) (depth).

@ Here also many of the points represent duplicates and triplicates. > See p. 143.

146 BULLETIN OF THE BUREAU OF FISHERIES.

BRERA.
ee
4. /

pe |
ooo

7

LENGTH—CMS.

Fic. 4.—Showing relative size and proportional increase, or rate of growth, of various parts of fish as compared with length.

BODY MEASUREMENTS IN THE WEAKFISH, CYNOSCION REGALIS. 147
SUMMARY.

I. 1n squeteague of both sexes, between the length of 15 and 70 cm., the correla-
tion of weight and length is extremely close, as expressed by the coefficient of correlation,
¥=0.952.

2. Weight may be accurately expressed by the equation: Weight in gm. = (0.00877)
X (length in cm.)%.

3. Standard length, head length, body length, tail length, width, and depth, are
directly proportional to total length. (See statements of tangent measurements, p. 145,
and fig. 4.)

4. From the curves in figure 4 the growth of these parts relative to total length is
readily calculated. :

5. Weight, as a function of total length, width, and depth, is expressed by the
equation: Weight = (0.5513) (length) (width) (depth).

THE FAT-ABSORBING FUNCTION OF THE ALIMENTARY TRACT
OF THE KING SALMON

Pd
By Charles W. Greene, Ph. D.

Department of Physiology and Pharmacology, Laboratory of Physiology
University of Missourr

149

CONTENTS.

Salacaoin of Gretel sere eo ooaandeh eooedsabonedsenee one ceapoooe ous cou mDbOo UoupegooeT
INfoaneibgaasboner Geibsotolel, .onadnasa doom anew nD soe horn Sura nea cHo en Ba OGOA SATO nAOABocnEoSS
Matte discal miorseeets ciyarsrce oe ety ace ete eit. rst Oe shave toe Sle tielers lala) sahere, Sichetaccrche eueyttesele.s . @)cyessia\a\ayslel ateravers
Generalirelationsjofithe) organs Of abSOnpHOml eo csc.cts acyotersicy tere) sist ele}sveiele “ieinleleisie aisiale ee sie wieiels
Absorptioniondats pygthe: pylOr CieGeCaye ate pecilaiale slapscelara ale stetetels.) spalchetsrefayis @la/intete)siele/s eteleveso eiaierTe
Hvidencesirompsalmon Leedinpenormallyn mame wey cee itso ie aisisictscicte eek wire sis) ree elie as
Histolopicaltappearanceiof tat ansthevepithelialicells yo. -\perneria is oeiein ieiisiion sivas cio
Bradencespumrartiicallys fe disal mow serra ine ate ate cneeas oi ister s12)sie\-7ayel aie fetoso}sislein eereve)elel<televsvla el siel=
Hatankthertunicaprophia Otmtat-ted salmon aeuyemtryseiaistaleteielaiale eieleloieiela s-ereteieictleleietele = leletele

[etre forse) I INf\SEaN YAGI iste, Guelel Od yno gee eennodeacos BeoouoAGppDoaocoNdoDosoane o
Bat-absorbing: power of the salmomaAntestime. oo. 6 oe) se eiereisiclel= or=icle s/oie enee=te' =| + e!aja{oielsiels) sie'eisfs «ielsiaiere
Absorptiontolerat byathersal mortstorm ai. irae eje\« sfe\steiele\ajavate s cyers/aters ofelaysvale¥elole) lm e1@/ evel alelaye\slelels!o\e/e) Je
Absorptonlatterta tree din gepacrte cielo: <epsctaleiele eiieteteiersicreteie steteierer «fete ale eYegene) ls¥ele/eleevasoreicle:sTeje\es-\sioh=
IAbsorptionmim th eycardtaci stomach sya yactatats slatalelaielslayale/sjoleedeleleteie)lersieke e eiesei) fer teic eel ese e) sinus (eins
Absorption im thevpyloricistommachy, 02 \.i.j2-jasjse« «jciete iets els teretnisi= este efel=(-feiniciale)~ = ole)e2 1=\e\nleielon eis
Absorption fat in the tunica propria of the stomach ............... 0. eee ee eee eee eee eee

sh eoreticali consideration smerree mye erieers ite center tee ersicheiapsesttseke eyeyetele cic oe ovane eve) oiciele (eicFeremeastoiane eieleray-

THE FAT-ABSORBING FUNCTION OF THE ALIMENTARY
TRACT OF THE KING SALMON.

Bd

By CHARLES W. GREENE, Ph. D.,
Department of Physiology and Pharmacology, Laboratory of Physiology, University of Missouri.

am
REVIEW OF THE LITERATURE.

The absorption of fats by the alimentary tract of man and animals has been a
subject for discussion and investigation for many decades. Our present views con-
cerning the topic have been arrived at almost exclusively by the study of the higher
mammals. Few observations along this line, especially of an experimental nature,
have been made on fishes. The setting of the problem which has led to the investiga-
tions here presented is found in our current views of fat absorption. These views have
been concisely and admirably summarized by Wells.*

In the intestines fat is split into a mixture of fat, fatty acid, and gylcerin; but as the fatty acid
and glycerin are diffusible, while the fat is not, they are separated from the fat by absorption into the
wall of the intestine. Hence an equilibrium is not reached in the intestine, so the splitting continues
until practically all the fat has been decomposed and the products absorbed. When this mixture of
fatty acid and glycerin first enters the epithelial cells lining the intestines there is no equilibrium, for
there is no fat absorbed with them assuch. Therefore the lipase, which Kastle and Loevenhart showed
was present in these cells, sets about to establish equilibrium by combining them. As a result we
have in the cell a mixture of fat, fatty acid, and glycerin, which will attain equilibrium only when
new additions of the two last substances cease to enter the cell. Now another factor also appears, for
on the other side of the cell is the tissue fluid, containing relatively little fatty acid and glycerin. Into
this the diffusible contents of the cell will tend to pass to establish an osmotic equilibrium, which is
quite independent of the chemical equilibrium. This abstraction of part of the cell contents tends to
again overthrow chemical equilibrium, there now being an excess of fat in the cell. Of course, the
lipase will, under this condition, reverse its action and split the fat it has just built into fatty acid and
glycerin. It is evident that these processes are all going on together, and that, as the composition of
the contents of the intestines and of the blood vessels varies, the direction of the enzyme action will
also vary. In the blood serum, and also in the lymphatic fluid, there is more lipase, which will unite
part of the fatty acid and glycerin, and by removing them from the fluid about the ceils favor osmotic
diffusion from the intestinal epithelium, thus facilitating absorption.

Quite similar must be the process that takes place in the tissue cells throughout the body. In
the blood-serum bathing the cells is a mixture of fat and its constituents, probably nearly in equilibrium,
since lipase accompanies them. If the diffusible substances enter a cell containing lipase, e. g., a liver
cell, the process of building and splitting will be quite the same as in the intestinal epithelium. The
only difference is that here the fatty acid may be removed from the cell by being utilized by oxidation
or some other chemical transformation.

@ Wells, H. G.: Chemical Pathology, p. 67, Philadelphia, 1907.

154 BULLETIN OF THE BUREAU OF FISHERIES.

This point of view has been arrived at through certain classic researches which will
pe reviewed briefly and which for the present purpose may begin with the work of
EH. H. Weber.

In 1847 Weber® demonstrated that during fat absorption the superficial epithelium
of the duodenal villi was filled with fat granules. He says that the cylindrical epithe-
lial cells “swell up and contain chyle granules” and that “the cylindrical cells are no
longer cylindrical or prismatic, but are round, and many become whitish-opaque
while others are filled with transparent oily fluid.’’ This he explains as due to the
power of the cells to absorb the food materials. Weber showed that some of the paren-
chymal cells also became opaque, containing oil-like fluids. This seems to have been
the first microscopic observation of the absorption of fats by the intestinal epithelium.

KOlliker ® in 1857 made similar observations, showing that fat was absorbed in the
stomach. He also used the histological method. Kolliker found highly refractive fat
granules in the fresh tissue. On page 175 is the statement, “I have never failed to
find fat in the stomach epithelium in the dog, cat, or mouse from the second day after
birth on. The mass of fat was indeed very variable. The cells may contain only slight
masses of fine granules or they may be gorged with fat in which not only the finer but
also larger fat drops are present.’’ It is significant that he found the fat only in the
cylindrical cells, i. e., not in the pavement epithelium of the mouse’s stomach. This
definite work of Kélliker on the ability of the gastric mucosa to absorb fat has appar-
ently been overlooked by physiologists until the last few years.

The question of the method by which fat is absorbed has been inseparably associated
with the observations of the actual physical processes of absorption. Our current view,
that fat must be dissociated before absorption, received its first experimental support,
so far as the stomach is concerned, by the work of Marcet° in 1858. This author
delivered a course of lectures in London, discussing in the last of the series the evidence
of the digestion of fat in the stomach. The keynote to his work is explained in the
following quotation: “The experiments were undertaken upon dogs, and repeated four
times with the same result. The animals were made to take a meal, consisting of cooked
meat and sheep’s fat, and were killed from one to five hours later; the contents of the
stomach being at once submitted to examination yielded in every case fatty acids.”

This chemical work of Marcet gave an admirable support to the histological observa-
tions of Kélliker, showing that the gastric mucosa is capable of producing fat-splitting
ferments, or as we now call them, lipolytic enzymes. In the light of our present
knowledge, these two pieces of work, that of Kdélliker and of Marcet, should have estab-
lished the fact of the fat digestion and absorption in the stomach. But no further
reference along this line seems to be available from the literature until the work of Cash 4
in 1880, who, without reference to the previous work of Marcet, again investigated the
fat-splitting properties of the gastric juice and of the extracts of the gastric mucosa.

@ Weber, E. H.: Ueber den Mechanismus der Einsaugung des Speisesaftes beim Menschen and bei einigen Thieren. Miiller.
Archiv fiir Anatomie und Physiologie, bd. 6, 1847, p. 400-402.

> Kolliker, A. von: Einige Bemerkungen ueber die Resorption des Fettes im Darme, ueber das Vorkommen einer physio-
logischen Fettleber bei jungen Saugethieren und ueber die Function der Milz. Verhandlungen der physikalisch-Medicinischen
Gesellsch. Bd. 7, 1857, p. 174-193-

¢Marcet: A course of lectures on the chemistry, physiology, and pathology of human excrements. Lecturev. The Medical
Times and Gazette, vol. 17, 1858, p. 209.

dCash, Th.: Ueber den Antheil des Magens und Pankreas an der Verdauung des Fettes. Archiv fiir Anatomie und Phy-
siologie (Phys. Abth.) 1880, p. 323-333-

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. 155

In 1890 Krehl ® made a restudy of the question of fat absorption from the intestinal
tract. His drawings showing different stages in the microscopic loading of the epithelial
cells with fat granules have become classic in the literature. The most significant fact
on which Krehl lays emphasis is ‘‘the fat is not taken up from the intestine in globular
form, but is absorbed in solution, and is resynthesized’’ giving rise to the droplets
observed in the pyloric epithelial cells which the author presents in his figures. The
conclusion that fat is absorbed in the dissolved state was later advocated by Pfltiger
(1900), after which it received general acceptance.

In 1901 Schilling ® again observed fat in the gastric epithelium of the calf, in this«case
after a meal of milk. Schilling noted that the epithelial cells were thickly studded with
microscopic fat droplets and that fat deposits appeared in the connective tissue of the
tunica propria and parenchyma. He also noted fat in the lymphatic glands during
absorption. He apparently did not investigate the presence of fat in the lymphatic
radicles from the stomach.

In 1908 Van Herwerden ° published the results of extensive and valuable studies on
the gastric digestion in fishes. This subject he investigated under two heads, the second
of which, namely, ‘‘ Enzymes in the gastric mucosa,” concerns us here. Van Herwerden
made his observation chiefly on sharks, but also on bony fishes. These fishes he fed with
olive oil or egg-yolk emulsion, the food being introduced into the stomach by way of the
mouth. Having previously determined that fasting animals were relatively free of fat
granules, he states that upon killing animals after a certain number of hours following
feeding, ‘‘one finds fat drops in great numbers in the superficial epithelium.’ He states
further that ‘‘the fishes contained fat granules everywhere in the submucosa between
the musculature and especially in the lymph vessels which accompany the blood
vessels. In hungering fishes I have never found this to be the case.”

Van Herwerden also tested the activity of glycerin extracts of the gastric mucosa.
He found in Scylliwm a decided increase in the formation of fatty acids; also, in teleosts
his tables show the presence of an active lipolytic enzyme. Extracts previously boiled
gave always negative results, as did also extracts from the muscle walls of the alimentary
canal.

These interesting observations of Van Herwerden seem to be the first that have been
made along this line upon the fishes. This splendid article had escaped my search in
the literature until after the publication of the preliminary report of the present work.

Three previous communications have been made with reference to the present work;
the first relating briefly the observations on fat absorption from the pyloric cceca of
the king salmon,? and the last two, one a preliminary and the other a brief statement
of the facts of fat absorption from the stomach of the king salmon.’

@XKrehl, Ludolf: Ein Beitrag zur Fettresorption. Archiv fiir Anatomie und Physiologie (Anat. Abth.) 1890, p. 97.

Schilling, F.: Die Fettresorption im Magen. Fortschritte der Medicin, bd. 19, r901, p. 613.

¢ Van Herwerden, M.: Zur Magenverdauung der Fische. Zeitschrift fiir Physiologische Chemie, bd. 56, 1908, p. 453-494.

4 Greene, Charles W.: The absorption of fats by the alimentary tract with special reference to the function of the pyloric cceca
in the king salmon, Oncorhynchus tschawytscha. Read before the St. Louis meeting of the American Fisheries Society, 1912.
Transactions American Fisheries Society, 1912, p. 261. 5

¢Greene, Charles W.: The absorption of fat by the salmon stomach. Preliminary notice. Proceedings American Physiolog-
ical Society, American Journal of Physiology, vol. 29, 1912, no. 4, p. XXXVI.

J Greene, Charles W.: Absorption of fat by the salmon stomach. American Journal of Physiology, vol. 30, p. 278, 1912.

156 BULLETIN OF THE BUREAU OF FISHERIES.

During the progress of this work a preliminary notice and final paper have been
published by Greene and Skaer® reinvestigating the fat absorption from the stomach
in mammals; also a paper by Weiss? briefly presenting the fact of fat absorption by
the gastric mucosa in the snake and in mammals.

EXPERIMENTS DEMONSTRATING THE ABSORPTION OF FATS.
METHOD.

The method of determining the character and degree of fat absorption from different
portions of the alimentary tract of the king salmon has been that of microscopic exam-
ination. ‘Tissues were examined fresh and after formalin fixation followed by the
newer fat stains, Sudan III, scarlet red, etc. The chief reliance for staining the fat in
the cells has been on the alkaline scarlet red. These methods of observation have been
confirmed by more careful tissue fixation in Flemming’s osmic acid mixture and by the
corrosive bichromate method of Bensley. Flemming’s solution not only fixes the tissues
but gives the characteristic osmic acid staining of the fats. The Bensley fixation, when
followed up by paraffin sections and differential staining, gives a negative picture, since
the fats are dissolved out by the clearing fluids, leaving only fat vacuoles.

The detail of procedure for staining with scarlet red is as follows: The perfectly
fresh material, living tissue if possible, was dropped into a 1o per cent formalin for two
hours or more. Precautions were taken to insure penetration and proper fixation. The
material fixed in formalin was then frozen in a freezing microtome and cut as thin as
possible. The frozen sections were cut directly into 70 per cent alcohol, and stained in
alcoholic solutions of scarlet red. The stain was made by heating an excess of scarlet
red in 70 per cent alcohol containing 2 per cent sodium hydroxide to a temperature
of about 80° C. This procedure, which is recommended by Bell,° gives a stain which
on cooling leaves a saturated solution of greater staining powers than the ordinary alco-
iolic scarlet red. ‘The stain was always filtered into shallow dishes just before using.
Shallow oval bottom salt cellars were used, and these immediately covered to prevent
evaporation. ‘The sections were lifted from the 70 per cent alcohol, the excess of fluid
quickly removed, and then they were dropped into the stain. Staining is compara-
tively rapid and requires only from 5 to 15 minutes for a successful impregnation.

Sections were taken from the stain, the excess of adherent stain being removed by a
momentary immersion in 70 per cent alcohol, and then were immediately plunged into a
large dish of water. When the particular tissues were delicate, an intermediate grade of
35 per cent alcohol was used. In this case the sections must be in contact with the
alcohol only long enough to remove the adherent stain, otherwise the stain in the tissue
itself will be drawn. As a matter of routine practice it was found desirable to add to
the wash water bath a couple of drops of hydrochloric acid. The faint acidity was found
favorable to the more rapid removal of the traces of alkali. This step contributes
decidedly to the keeping powers and clearness of the sections after they are mounted

aGreene, Charles W., and Skaer, William F.: Absorption of fat by the mammalian stomach, Proceedings American Physi-
ological Society, American Journal of Physiology, vol. 29, no. 4, 1912, P. XXXVI. Evidences of fat absorption by the mucosa of
the mammalian stomach, American Journal of Physiology, vol. 32, 1913, Pp. 358.

b Weiss, Otto: Die Resorption des Fettes im Magen. Pfliiger’s Archiv fiir die gesamte Physiologie, bd. 144, 1912, P. 549-543-

c Bell, E. T.: The staining of fats in epithelium and muscle fibers. Anatomical Record, vol. 4, 1910, p. 199-212.

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. 157

in glycerin. Pure glycerin was used to make the mounts. Sealing with a mixture of
paraffin and beeswax around the cover glass was the final step in the mounting and
preservation of the sections.

The more permanent sections of the tissues fixed as described were made by the
paraffin method, in which no special features in technique were introduced.

The previous fixation in formalin was found to be decidedly advantageous in the
preparation of frozen sections. The brief time of immersion in the formalin does not
introduce a change in the character and distribution of the fats. On the other hand the
tissues are coagulated, hence firmer, and can be carried through the technique with a
much more satisfactory result. When the frozen sections were made directly from
fresh living tissues, then at the moment the frozen section was immersed in the alcohol
preliminary to the scarlet red staining, considerable contraction and sometimes tearing
took place. It wasfound that the distortion of the sections by this step was detrimental
to the securing of normal pictures of the structure and relations of the contained fat.

SELECTION OF SPECIMENS.

Two types of fish were used for the determination of the points detailed in this
report. First, salmon of various sizes and presumably of different ages collected from
the markets in the city of Monterey. The fish selected were those delivered directly
from the fishing boats, which had made their catch by trawling on the ocean fishing
banks in the vicinity. These fish came into the market with living tissues, a fact that
could easily be determined. The alimentary tracts were taken from the salmon at the
slaughtering tables of the fish-packing establishments of the Booth Packing Co.* If
the tissues were proved to be alive in material chosen then histological samples were
selected and placed in fixative immediately, so that the question of prefixation changes
does not enter into consideration.

The second class of material is that derived from young salmon collected from two
stations. The first collecting ground was that of the Brookdale hatchery maintained
in the town of that name on the San Lorenzo River in the Santa Cruz Mountains. Young
salmon were also obtained from the McCloud River in the Shasta Mountains in northern
California. Both these groups of young salmon had never been in salt water. The
ages of the young salmon varied from one to two years, the latter being those obtained
at the ponds from Brookdale.

NORMAL-FEEDING SALMON.

The class of adult salmon mentioned above, which were secured at Monterey, were
in an active aggressive stage of ocean feeding. These salmon come into the markets
often with the stomach and intestinal tract gorged with food. The natural food is of a
varied class, but at Monterey consists mainly of three kinds: First, the squid; second, the
local species of herring; and, third, marine Crustacea, chiefly a rather large amphipod. ?
The king salmon is a voracious feeder and his ability to capture a great variety of food

@ For the courtesy extended by this company I am indebted to Mr. Frank E. Booth, the president.

b In July, 1911, quite a number of salmon were noted with large numbers of these Crustacea in their stomachs. One salmon
stomach in particular contained 4 or 5 (estimated) ounces of such food. It would have been interesting to have counted the
actual number of crustaceans present, but the content of the stomach was partly lost before the thought occurred to make such
an enumeration.

19371°—vol 33—15——11

158 BULLETIN OF THE BUREAU OF FISHERIES.

material besides the forms mentioned above is shown by the various species of fishes
occasionally noted in the food at Monterey. These natural foods are all relatively
oily, the point which particularly concerns us here. As digestion proceeds and the
protein framework is dissolved away these oils are liberated in the alimentary canal
and form no inconsiderable portion of the food of the king salmon. When one remem-
bers the characteristics of the salmon flesh, charged with oil as it is, and evidently storing
great quantities of oil, the interest which attaches to the question of the source of the
oil in the food and the method of digesting and absorbing oils is obvious.

Asa matter of fact it wasin the course of a study of the character and microscopic
distribution of the fats in the salmon tissue that I instituted observations on the ali-
mentary tract of the king salmon which made it obvious that large quantities of oils
were absorbed from the foods in these normal feeding salmon.

FAT-FED SALMON.

The inability to control the relation between the time of taking food and the chance
of securing the fish and making observations of the stage of absorption in the normal
feeding salmon renders it extremely difficult to settle the question of the characteristics
of fat absorption in such. As a matter of fact, my observations made it very clear that
much absorption of fat was taking place in salmon feeding under natural conditions, yet
it was found next to impossible to determine the nature and details of the process from the
specimens available. For this reason the idea of feeding salmon in the aquaria was
conceived and its immediate execution was made possible through the courtesy of the
directors and superintendent of the Brookdale hatchery. Young salmon were transported
in live cans from Brookdale to the Hopkins’ Seaside Laboratory at Pacific Grove, Cal.
Two sizes of salmon were available, one group of yearlings from 6 to 7 centimeters long,
and a group of small 2-year-olds from 14 to 16 centimeters long.

These young salmon were fed olive oil by rectal injection. This was found to be an
extremely reliable and easy way of introducing the oil into the alimentary tract in such
a way as to give one confidence in the accuracy of the results. A medicine dropper was
drawn out in the flame to a slight cone of proper size. A desirable quantity of oil was
then taken into the dropper, the tip inserted into the anal aperture and gentle pressure
maintained until the oil was emptied into the alimentary tract. It is comparatively
easy to hold the young salmon by a firm grip of a lobe of the caudal fin rays, the fish
resting in the palm of the hand in such a way that the head and gills remain under water
to prevent asphyxiation. Under these conditions the fish does not struggle as much as
might be expected. The slight contractions of the muscles of the anal sphincter occurring
when the pipette is first introduced soon relax, but one has always to maintain a gentle
pressure on the pipette for a moment before oil begins to flow into the tract. The alimen-
tary canal of the salmon is a simple S-shaped tube, as has been described and figured in a
previous paper.* When the oil is injected into the posterior end of the canal in sufficient
quantity it flows into the different limbs of the intestine and into the stomach, and from
the stomach will be discharged from the esophagus into the mouth if an excess of oil is
used. In my later experiments this fact was adopted as an index of when the proper
quantity of oil was administered.

aGreene, Charles W.: The anatomy and histology of the alimentary tract of the king salmon, Oncorhynchus tschawytscha.
Bulletin, Bureau of Fisheries, vol. xxxm, 1912, P. 73-100, DI. XXV-XXVIII.

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. 159

A series of artificial feeding experiments was executed at the Hopkins’ Seaside
Laboratory, followed by a more extensive series at the Federal salmon hatchery at
Baird, on the McCloud River in northern California. In this later series the question
of absorption in relation to the time following the administration of oil was especially
investigated. Furthermore, in the Baird series it was possible to maintain the young
fish without food an adequate time to insure the complete elimination of the fat from
the alimentary canal which might previously have been derived from natural foods.

The Monterey series consisted of two salmon of the 2-year-old group with confirma-
tions on two salmon of the small 1-year-old group, with different time allotments for
absorption, ranging from 20 to 70 hours. Careful examinations were made extending
over the stomach, intestine, and various pyloric cceca of each of these series.

GENERAL RELATIONS OF THE ORGANS OF ABSORPTION.

The critical regions for the study of the absorption of fat in the salmon are three,
namely, the stomach with its two divisions, the cardiac and the pyloric ends; the intestine
with its two great divisions, the pyloric and post pyloric; and the numerous pyloric
cceca which have their origin from the pyloric intestine. :

These great divisions are of necessity to be described separately. Logically, one
might take them in the order, stomach, intestine, coeca; but because of the way in which
the evidence was accumulated and other questions attached to the subject it is more
convenient to discuss the details in the reverse order, i. e., absorption in the pyloric
coeca, in the intestine, and in the stomach.

ABSORPTION OF FATS BY THE PYLORIC CCECA.

The gross anatomy and the normal histological structure of the alimentary tract
of the king salmon have both been presented in a previous paper.” Figure 1 of that
paper is an illustration showing the general relations of the cceca to the pyloric end of the
intestine from which they arise in such profuse numbers. Those cceca which originate from
the beginning of the intestine, that is, in the neighborhood of the pyloric valve, are much
longer than those that arise from the posterior end of the series. These cceca often reach
a length of from 10 to 15 centimeters and even more in the adult feeding salmon. They
have a normal diameter of 5 to 8 millimeters. In the sea salmon taken at a time when
food is abundant and digestion has been going on actively for some time the cceca are
always gorged with material and distended to their full length and diameter.

The content of the pyloric coeca under these conditions is peculiar in appearance.
One never finds solid particles of food. Instead, there is only, as Gulland and others
have mentioned, a creamy, yellowish, puslike mass which has a viscid adhesive con-
sistency. This content is never very fluid, i. e., of limpid character. The exact color
of the contents varies with the class of food material which the salmon is digesting at
the time. If the food is made up of Crustacea then the content of the cceca has a darker
color, often of a deep orange red. It is apparent that the viscidity of the mass is due
to the secretion of mucous by the epithelial lining of the cceca themselves.

In the younger salmon the pyloric coeca have the same relative size, but of course
are smaller in proportion to the gross size of the fish. In no instance have I observed

a Greene, Charles W., op. cit.

160 BULLETIN OF THE BUREAU OF FISHERIES.

any extensive mucous content of these young coeca. In the specimens that were fed
fat there was an occasional increase in the transparency, which was interpreted as due
to the presence of oil. In the intestine of such fish the excess of oil was easily and often
shown.

EVIDENCES FROM SALMON FEEDING NORMALLY.

Fat droplets were always observed in the epithelial lining cells of the pyloric coeca
of the Monterey salmon. However, the fat was not present in all cells. Certain por-
tions of the epithelium were filled with fat droplets, while other portions were relatively
free. In almost every animal observed, and in different regions of the same animal,
certain extended portions of the epithelium were observed to contain no fat droplets,
while in the neighboring regions, often in the same section or perhaps in the next mucous
fold, fat would be present. These facts could not readily be explained by the assump-
tion that fat was loaded into these cells by way of storage, being brought in from other
portions of the body. On the other hand, such observations strongly suggest a process
of fat absorption. Previous observations on fat absorption in fishes are apparently
very limited; at any rate the search in the literature has thus far revealed to the writer
only the observations of Van Herwerden* “On Gastric Digestion in Fishes.” This
splendid paper deals largely with digestion and the digestive enzymes. But it definitely
demonstrates fat absorption in Scylliwm. It follows that the chief guide in the inter-
pretation of the present results is that to be found in the comparative literature on fat
absorption in other animals, a portion of which has been referred to and reviewed in a
previous chapter.

The mucous epithelium of the salmon cceca is very extensive, considered in proportion
to the size of the tubes. The measurements of the superficial extent of the mucous coat
show that it is from 6 to 8 times the extent of the external surface of the ccecum itself.
These folds are very complex in arrangement, though the epithelial coat itself is of uni-
form and simple type, a matter that is discussed in the paper presenting the normal
structure of these organs. It is this complex folding, and therefore the relative varia-
tion in the contact of the epithelium to portions of the contained food mass, that explains
the fact of unequal loading of fat in the epithelial cells. Hence there is no doubt that
the fat observed was absorption fat.

HISTOLOGICAL APPEARANCE OF FAT IN THE EPITHELIAL CELLS.

A coecum containing fatty food material in an advanced stage of digestion and ab-
sorption will almost always present epithelial cells in all the stages of fat loading. The
appearance of the cells loaded with fat is characteristic and changes progressively as
absorption proceeds. In a general way, though some allowance must be made for the
comparison, the histological character of the cells would suggest three stages.

Fat absorption, stage 1.—The earliest stage of absorption is that of the passage of fat
into and through the superficial border of the epithelial cells. The methods of staining,
whether they be direct staining of the tissues with scarlet red or fixation and staining of
the fat by the osmic acid mixtures, show a large number of very fine granules in the
most superficial layer of the protoplasm of the cell. These fat granules are extremely

@ Van Herwerden, M., op. cit.

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. 161

small, the largest being less than 1 # in diameter. In some instances they appear in
such minute size that they are only just distinguishable under the oil immersion. As
absorption proceeds the fat granules make their appearance deeper and deeper in the
cell, loading up the zone between the free surface and the nucleus. Here the fat droplets
are relatively large, ofttimes being 4.5 to6in diameter. In the intermediate area and
between the superficial zone and the extra-nuclear zone are all sizes of fat droplets from
the extremely minute ones just described to the large ones in the extra-nuclear zone.
This picture is shown very clearly in figures 6 and 12.

Fat absorption, stage 2—The second characteristic cellular appearance, which is des-
ignated as stage 2, consists in the filling of the inner or basal end of the cell with fat
droplets. Not only that portion external to the nucleus will be loaded with fat, but
the portion between the nucleus and base of the cell will also contain an excess of fat
droplets.

The sizes of the droplets in the end of the cell are similar to those just external to
the nucleus, but the number of droplets is rarely so great. When the cell is fully loaded
it generally happens that fat will be found in the connective tissue of the tunica propria
beneath. If fat absorption is continuous at this stage, as one might legitimately assume
from the histological appearance, it is obvious that as the fat is entering the outer zone
it will at the same time be discharging from the inner zone and passing into the chan-
nels which distribute fat through the body. Knoll? has recently reported experiments
on fat absorption in the mammalia in which this condition is shown to hold.

Fat absorption, stage 3—When absorption from the lumen of the caecum ceases, the
outer margin of the cell begins to clear of fat. This disappearance of fat apparently
slowly and gradually extends over the whole area of the cell external to the nucleus.
In favorable material in this stage epithelial areas will be found in which the outer or
extra-nuclear zone of the mucous epitheiium is almost, sometimes entirely, free of fat
droplets. Still, fat droplets will be present in considerable quantity in the inner or basal
zone. Asa rule the basal portion of the cell will contain relatively large droplets in this
stage and the connective tissue supporting the cell will be similarly loaded with fat
droplets. However, some groups of cells are found in which the fat droplets in the
basal portion of the cell are extremely minute, as shown in figure 13. In this particular
figure the basal areas are heavily loaded with fat in extremely fine subdivision. The
adjacent connective tissue of the tunica propria contains a similar distribution and size
of fat droplets.

These three stages of course are only phases of an orderly and progressive process
in which the fat enters the outer zone of the cell, is disposed within the substance of the
cell in droplets, and is ultimately distributed from the cell to the basal pole, the opposite
from which it entered. The variations in the size of the droplets in different zones of
the epithelial cells, especially the extremely small droplets in the outer portion of the
cells and the fine droplets in the bases of the cells at the time the discharging is almost
complete, are very interesting when considered in relation to the theories of fat absorp-
tion. But the discussion of these theoretical points will be taken up again in a later
section of the paper.

@ Knoll, A.: Chemische und mikrosopische Untersuchungen iiber den Fetttransport durch die Darmwand bei der Resorp-
tion. Pfliiger’s Archiv, bd. 136, 1910, p. 208-247.

162 BULLETIN OF THE BUREAU OF FISHERIES.
EVIDENCES IN ARTIFICIALLY FED SALMON.

A brief report on these experiments has been presented.* The first series of experi-
ments carried on by the method of fat feeding described on the preceding page contains
two young salmon, one 14 and the other 16 centimeters long.

Figures 6 and 7 present the histological picture of the amount of fat in the superficial
epithelium of frozen sections of the cceca from salmon 45. Absorption proved to be
extremely rapid and vigorous in this young salmon, not only in the cceca, but in the
intestine, as will appear later. The epithelial cells, especially of those mucous folds
which extended out into the lumen, were simply gorged with fat. The fat droplets were
extremely large and filled not only the superficial portion of the cells but the basal
portion as well. If the adjacent membranes of a deep fold were in contact with each
other, thus preventing a free contact with the fat of the ccecal content, such places
would show a relatively small amount of absorption fat in the cells. On the free loops
of the mucous folds this situation did not exist, hence these portions were gorged with
fat in all the sections examined. This fact is shown especially well in the high magnifi-
cation of figure 11. In some portions of the tissue in the neighborhood of the areas
drawn in this figure the fat was present in so great a quantity as to burst the cell mem-
brane. It was believed at the time of the preparation that the fat absorption con-
tinued until the quantity within the cells produced a pressure greater than the cell
surface could stand, hence the break, though one can not exclude the possibility of
mechanical pressure during manipulation. Drops often reach a diameter of from 8 to
IO #4, Or even more in the young, which is greater than the normal diameter of the
epithelial cell, even at its largest end.

In figure 6, showing the fat in salmon No. 45 stained with osmic acid, a number of
cells are shown in which the fat droplets of the outer portion of the cell are large enough
to take up the entire diameter of the cell. In different regions of this particular histo-
logical preparation other than shown in the figure there are numerous confirmations of
the above statement.

Both the positive staining and fixation of fat by osmic acid and the arrangement
of fat vacuoles in corrosive fixed and paraffin sectioned material give confirmation of
the direct observations of the fresh material stained with scarlet red. The series of studies
show that fat absorption takes place abundantly in the pyloric ceca. Whatever else these
organs accomplish, it is perfectly clear that the absorption of fat is one of their chief
functions.

FAT IN THE TUNICA PROPRIA OF FAT-FED SALMON.

One of the most interesting confirmatory lines of observation which is largely
cleared up by the fat-feeding experiments is the fact of the presence of fat in the tunica
propria. In fasting salmon, especially in those used for control in the Baird series,
practically no fat is present in the tunica propria. One must be guarded in such state-
ments because this tissue holds on to its fat with great persistence. Fat will persist
in the tunica propria when one can demonstrate absolutely no fat in the epithelial cells.
But when absorption begins, as judged by the amount of fat in the epithelial cells, then

@ Greene, Charles W.: The absorption of fats by the alimentary tract, with special reference to the function of the pyloric
coeca in the kingsalmon, Oncorhynchus tschawylscha. ‘Transactions of the American Fisheries Society, rorr, p. 261.

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. 163

fat begins to appear in the tunica propria in an increased quantity. After 18 hours or
more (see figs. 6, 7, and 8), the connective tissue layer supporting the epithelial cells
becomes extremely full of fat droplets. The fat appears first in the vicinity of the
bases of the epithelial cells, then is distributed through the substance of the tunica
propria. The stratum compactum always forms a definite and striking boundary to
the fat containing tissue of the tunica propria. This is shown especially well in the
figures, particularly figure 6.

The tunica propria acts as a sort of reservoir for the fat immediately following
periods of active absorption. The matter has not been sufficiently studied yet, but it
seems obvious that the tissue building up this stratum of the ccecal wall seizes and holds
fat with unexpected persistence. Fat will be found here in a relatively considerable
number of droplets at a time when the epithelial cells are completely discharged.

The stratum compactum, as described in the discussion of the normal structure,
forms a continuous sheath around the tunica propria. It is a continuous membrane
with no discernible openings other than at the points where blood vessels enter. Any
fat passing through the stratum compactum would have to pass through in solution or
else be carried within in the lumen of the blood vessel. In either case no definite fat
globules as such get by this membrane from the tunica propria.

PROTOCOLS.
BROOKDALE SALMON, FIELD SERIES No. 45, LENGTH 14 CENTIMETERS, TAKEN JULY 6, 1911.

This young salmon was a 2-year-old reared by the Brookdale hatchery, California. It was taken
from an aquarium and transported to the Hopkins Seaside Laboratory, Pacific Grove, Cal. This
salmon was fed fat. In this instance it received first a fat emulsion consisting of 20 per cent olive oil
in coagulated milk injected into the stomach through the mouth. This feeding did not seem to be
very successful and was followed later by an injection of olive oil into the rectum. This method of
feeding proved to be very successful, convenient, and satisfactory. The salmon was killed after allow-
ing 18 hours for absorption of the olive oil (22 hours, counting time from the first attempt to feed by
way of the mouth). Frozen sections were made of the fresh tissues, and certain portions of the tissues
were fixed by different histological methods for later examination.

Fat in the pyloric ceca.—The epithelial cells are simply gorged with fat droplets. Especially are
the outer ends of the cells so filled that the cell boundaries are obscured. The diameter of the droplets
varies widely. The larger drops distort the cells. The basal ends of the cells contain a much smaller
amount of fat.

The tunica propria is also well filled with fat, but not so great an amount as in fish 46. The drops
are more uniform in size. This fat extends deep into the folds of the stratum compactum, but is never
present in its substance.

Fat in the intestine.—The epithelial cells are as much crowded with fat as in the coeca, as shown
in figures 2 and 4. The deeper folds of the intestinal mucosa are not always filled with fat, at
any rate the amount of fat is not nearly so great as in the outer folds.

There is less fat in the intestinal tunica propria than in the cceca of the same animal.

BROOKDALE SALMON, FIELD SERIES No. 46, LENGTH 16 CENTIMETERS, TAKEN JULY 6, IgII.

This young salmon was a mate to no. 45 and was transported at the same time. It was fed fat
by rectal injection only, and was killed after 42 hours of absorption. Frozen sections were prepared
and tissues were also fixed for permanent histological mounts.

Fat in the pyloric ceca.—The amount of fat in the epithelium of the pyloric cceca varied in different
preparations. Those cells on the tips of the folds were crowded with fat, while those in the grooves
between folds were relatively free. Figure 8 and figure 11 are from this specimen.

The tunica propria, as shown in figure 8, was more crowded with fat than in no. 45.

164 BULLETIN OF THE BUREAU OF FISHERIES.

Fat in the stomach.—Sections of the gastric division of the stomach showed the outer ends of
the epithelial cells medium full of fine fat droplets, shown in figure 1. These droplets are in the
extreme outer ends of the cells just within the striated border. They are strikingly smaller than the
droplets present in the intestinal and coecal epithelium of the same specimen. A sprinkling of fat
droplets is present in the inner limbs of the gastric epithelial cells.

McCLoup RIvER SALMON, Fre.p SERIES No. 88, FEMALE, LENGTH 84 MILLIMETERS, TAKEN JULY 23,
IQIt.

This young salmon was seined from the McCloud River and was 1 year old as verified by scales.
It was fed fat by the method of rectal injection and killed after 20 hours.

Fat in the pyloric ceca.—Frozen sections were made of the pyloric coeca and these stained with
scarlet red and counterstained with hematoxylin. Absorption fat was present in moderate quantity,
see figure 9. The greater portion of the fat is limited to the outer ends of the cells, but a few droplets
were present in the inner ends of the cells and a small amount in the tunica propria.

Fat in the stomach.—This specimen showed an unusual amount of absorption in the gastric epithe-
lium. Particularly was the pyloric epithelium loaded with fat. (See fig. 1.) Many of the deep
folds of the pyloric epithelium were practically free of fat, but those cells dipping deepest into the
cavity of the stomach were unexpectedly filled.

Two sections of cardiac stomach were fat-stained only. The slender cylindrical epithelial cells of
the mucous ridges bordering on the lumen of the stomach and those cells extending down into the crypts
of this somewhat contracted stomach all show numbers of droplets. The fat is greatest in amount in
the cells of the free folds. The fat is finely divided in appearance; that is, in minute droplets. It is
greatest in quantity in the outer thirds of the cells. There is a transparent superficial border of the
epithelial coat in which the fat is in the form of finest liposomes, requiring the oil immersion lens for
resolution. (See fig. 1 of osmic acid staining of no. 46, Greene, American Journal of Physiology,
vol. 30, p. 280.) There is also fat in the inner limbs of the cells down to their bases, and this is more
or less continuous with small amounts of fat in the tunica propria.

The gland cells of the secreting portion of the stomach in this section are granular in appearance
and slightly pink with scarlet red. In several regions very small fat droplets, the size of which varies
around o.5 1, are found in the basal portion of many of the gland cells. In connection with the large
majority of the gland tubes in this slide there are cell areas over the surface of the tubes which seem
quite thickly studded with finest fat droplets. This section is not counterstained, so it is difficult
to determine to exactly what tissue these cells belong. In some instances they undoubtedly belong
to the connective tissue of the tunica propria surrounding the gland tubes.

Several of the submucous areas just within the circular muscle and through which blood vessels
run are finely punctate (oil immersion lens) with liposomes. Vascular areas in the longitudinal muscle
coat are also stippled with liposomes, the droplets being located in the endothelial cells and in the
walls of the blood vessels.

McCLoup RIVER SALMon, FIELD SERIES No. 91, MALE, LENGTH 83 MILLIMETERS, TAKEN JULY 25, 1911.

A young salmon seined from the McCloud River, fed olive oil by the method of rectal injection
and killed after 70 hours.

Fat in the pyloric ceca.—Transverse sections of the pyloric cceca were made by the freezing method
and stained for fat. Figure ro shows a characteristic section from this fish. The amount of fat in the
outer portion of the epithelial tissue is unusually great, though a very small amount had penetrated
the inner limbs of the epithelial cells, and practically none to the tunica propria. The length of time
that had been allowed for absorption would justify the expectation that the tunica propria would be
loaded as shown in figure 8, from no. 46. Such was not the case. Possibly fat was late in entering
the particular group of cceca examined.

The amount of fat in this material, as in the cceca of all of the fat-fed salmon, is unquestionably
from fat absorption. The control materials, salmon no. 82 to 86, presented no fat in the epithelium of
the coeca and only traces in the tunica propria.

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. 165

FAT-ABSORBING POWER OF THE SALMON INTESTINE.

In the paper on the normal structure of the alimentary canal it has been shown that
the salmon intestine has a histological structure relatively simple. It possesses the
same epithelial lining coat, the tunica propria, the stratum compactum, and muscular
membranes which are found in its diverticula the coeca. The mucous membrane itself
is shown to be somewhat more complexly folded than in the cceca, a complexity that
increases with the size of the fish. No differentiations are found in the different por-
tions of the epithelial coat of the mucosa. Even in the deepest grooves or pits of
epithelium the cells have the same general form and structural characteristics as in the
most superficial folds.

The intestinal epithelium of the salmon is also a fat-absorbing tissue. Fat is taken
from the lumen of the intestine with the greatest avidity by these cells. The judgment
in this case is based on the histological showing made by the epithelial folds after fat-
absorbing experiments. Unfortunately, no observations were made on the normal-
feeding salmon and no opportunity has arisen to repair the deficiency. The facts
presented here are wholly those derived from the studies of the young salmon which
had been fed fat by the methods described above.

Figure 2, plate xi, from young salmon no. 45 presents a general view of the rela-
tions of the fat under a low magnification. Figure 3, plate x1u, is a highly magnified
drawing showing the fat of one of the loops of one of the mucous folds of the section shown
in figure 2. These figures show the epithelial cells gorged with fat for their whole extent
external to the mucosa. In some instances the fat drops are large and have a diameter
equal to or even greater than that of the normal cells. Often this fat appears in chains
of drops extending from the surface of the cell to the nucleus. In other instances the
droplets are somewhat smaller, but nevertheless crowd the cell body to the margin
These statements are made on the basis of observations in both the pyloric and the post-
pyloric loops of the intestine and on sections of the pyloric portion of the intestine. The
section from which figure 6 is drawn, representing the fat stained with osmic acid in a
fold of the ccecal epithelium, also contains a section through the pyloric intestine. The
intestinal epithelium, too, is crowded with fat. In these epithelial cells the beaded
arrangement of fat droplets is especially prominent. Where the section is accurately
longitudinal through the epithelial cell, the rows of droplets of the larger size are shown
filling up the whole body of the cell and to extend from the free surface to the region of
the nucleus. It is comparatively seldom that fat is present in the basal portion of the
cells, i. e., within the nuclear zone, in any such massive quantity as is so often found in
the external limb of the cell. Here the fat is scattered along in fewer droplets, generally
of fairly large size. In figure 5, plate xu1, the amount of fat in the inner nuclear zone
is comparatively small.

It is to be emphasized that no fat droplets are present in the nuclei themselves.
We have not observed any nuclear fat either in the epithelial or in the connective tissue
nuclei supporting the epithelium of any portion of the alimentary canal. We are inclined
to think that the nucleus does not for some reason ever receive a deposit of fat.

The intestinal epithelium discharges its fat into the connective tissue of the tunica
propria just as observed in the pyloric ceeca. In the material from which figure 3
is drawn this fact is very patent. Here the tunica propria contains a comparatively

166 BULLETIN OF THE BUREAU OF FISHERIES.

heavy loading of fat. Considering the whole of figure 2, the showing of fat in the
loop chosen for figure 3, plate xin, is if anything too low for the tunica propria. In the
intestine also the fat that makes its appearance in the tunica propria is not distributed
over the whole of that structure down to the stratum compactum. The stratum com-
pactum forms a very definite and limiting boundary to the fat-containing tissue.
However, it is believed that this fat present in the tunica propria is not a true storage
fat. No characteristic areolar fat cells are present such as are found in such num-
bers in the pancreas and in certain other definitely adipose tissues of the salmon.
The tunica propria fat of the intestine is in comparatively small drops, rather evenly
distributed over the structure, and bears all the histological evidences characteristic of
the fat in the epithelial cells which is so obviously transient in its character. The fat
in the tunica propria of the intestine is also retained with greater persistence, or at least
for a longer time following periods of fat absorption, than is the fat of the epithelial
cells. This characteristic has already been mentioned in discussing the coeca.

Further studies ought to be made before advancing the point, yet one must mention
here that no obvious lymph channels through which the fat is being removed have been
observed. That is, no structures comparable to the mammalian lymphatic radicles of
the mammalian intestine have been observed during these studies. This is not to be
interpreted as an assumption that there are none, because the observations are insuffi-
cient in number to establish a point of this character. The fact must also be mentioned
that no evidence of accumulation of stainable fat in the cavities of the blood vessels has
been secured. In fact, fat droplets do not appear in any of the coagulated plasma nor
in any of the free blood cells in so far as yet observed either in the intestinal blood vessels
or those of other parts of the body.

Minute liposomes have been found in the endothelial linings of blood vessels and in
the blood vessel walls. Such findings are shown in figures 3 and 6, plate x1. The
quantity of fat disposed in such places is small, but it was found to be present in fishes
in which fat absorption was at its maximum, a fact that suggests but does not prove
a relation to fat absorption.

ABSORPTION OF FAT BY THE SALMON STOMACH.

The fact of the absorption of fat by the epithelial lining of the stomach was first
observed on the young salmon which had been experimentally fed with fat. Obser-
vations were not made on the adult feeding salmon in a way to determine whether or
not gastric fat absorption occurred. The absorption of fat in the young was observed
in both series, i. e., the specimens from Brookdale, Cal., and from the McCloud River
at Baird, Cal. The young salmon in the McCloud River are feeding, but evidently on
a source of food which is not particularly rich in fats. At any rate the specimens
seined directly from the river and examined without further feeding showed only small
amounts of fat in the epithelium of the stomach. In the series of four young fish no
fat could be identified in one, a trace of fat only in one, and two contained obvious
and easily identified fat droplets. These specimens were taken as typical of the average
of those secured from the McCloud River, and were therefore considered as normals.
The specimens that received fat as food by the method previously presented were
examined in comparison with the normal series just given.

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. 167

ABSORPTION AFTER FAT FEEDING.

The amount of fat taken up by the mucous lining membrane of the stomach is not
anywhere near so great in amount as that shown by the mucosa of other portions of
the alimentary tract in one and the same animal. However, this fat is in amount
quite sufficient to form a very striking picture.

The microscopic evidence of fat absorption is largely limited to the superficial
epithelium. At any rate, this tissue is most distinctly loaded with fat droplets, and the
loading apparently occurs earlier than in deeper portions of the gastric mucosa. An
examination of the epithelium of the stomach showed fat droplets present in practically
every portion of that organ. The absorption takes place not only in the cardiac divi-
sion, but also in the pyloric stomach.

The earliest indication of fat absorption is found in the appearance of fat droplets
in the more superficial epithelium and in the distal ends of the cells. As time is allowed
for the digestive and absorptive processes these cells become more fully loaded—in fact
gorged—with fat, first in the outer limbs, then later the droplets appear nearer the
bases of the cells. The glandular tissue of the gastric mucosa also shows the presence
of fat droplets in the later stages of fat absorption. Apparently not only the super-
ficial epithelium and the crypts even down to the neck cells, but also the glandular
cells themselves are capable of taking up fat in quantities sufficient to produce the
numerous, droplets which the microscopic examination reveals. Since the structure
of the gastric mucosa is characteristic and strikingly different in the two divisions of
the salmon stomach, these regions will be discussed separately.

ABSORPTION IN THE CARDIAC STOMACH.

In the series of fish fed at Baird one had little or no fat in the stomach coat, while
three showed the presence of fat in decided quantities. In those fish in which fat was
present in the stomach it was in relatively large amounts as compared with the normals,
That is to say, the amount of fat in the epithelium of the stomach in the fat-fed fish
was larger in amount than in the fish coming directly from the river.

The amount of fat in process of absorption by the epithelium was greatest in fat-
fed fishes nos. 88 and 91 of the McCloud River series. The fat was present in super-
ficial epithelial cells of both the cardiae and the pyloric divisions of the stomach, but
the amount in the cardiac division was very obviously less than in the pyloric division.

The fat in the cardiac stomach is distributed chiefly in the cylindrical cells of the
superficial epithelium. It is in greatest quantity in those cells bordering freely on the
cavity of the stomach. Ina typical section through the cardiac region all that portion of
the epithelium outside the nuclear zone and within the extreme outer clear zone will be
studded with minute fat droplets. The fat droplets here vary much in size, but seldom
reach more than 3 # in diameter. The most of the fat is in such small divisions as to
require an oil immersion lens to distinguish the individual droplets. (See fig. 4, pl.
xu.) In the outer clear zone or border I found fat in only one fish, and in this instance
the droplets were extremely minute, i. e., liposomes. #

@ The size of the fat droplets in the stomach shows every gradation from the larger size of 2 to 3 « diameter down to a size that
is discernable only with the highest magnification. In the salmon stomach, indeed in the salmon tissues in general, I am quite
unable to distinguish any constant differences in appearance among these fats. There is no line to be drawn either as regards
color, size, or contour. It is true that the color shade and the size vary greatly, but not in any way that does not admit of expla-
nation without assuming any characteristic difference in the composition of the fat bodies stained. Under these conditions I
use the term liposomes without reference to the kind of fat, only to designate the extremely small size of the droplets.

168 BULLETIN OF THE BUREAU OF FISHERIES.

The basal parts of the cells have fat droplets, but rather smaller in size and not
so numerous as in the outer limbs of the cells.

In certain regions the fat is present in the cylindrical cells of the lining walls down
in the deeper folds of the crypts, but in other regions it is entirely absent. I have seen
the fat in these cells down as deep into the crypts as the region into which the deepest
gastric gland tubes open. In every case there is a very noticeable difference in the
amount of fat present in the deep-lying epithelium and the more superficial—always in
favor of the greater quantity in the superficial.

The tubes of the gastric glands open into the sides and bottoms of the crypts. There
is a quick transition from the superficial epithelium to the gland cell type at the point
where the mouth of the gland opens. It is not often that a section passes longitudinally
through the mouth of a gland. This is largely due to the fact that the glands are some-
what convoluted in shape, rarely straight and tubular as in the gastric glands of most
mammals. A number of gland tubes usually open into each crypt. Some of these
are very short, and are only a few cells in length, while others extend quite down to the
basement membrane. Occasionally a single tube may be as straight and direct as in
the mammalia, but the majority are irregular. The transition in the epithelium from
the superficial to the glandular type is sudden and sharp. In medium magnification
the superficial epithelium looks darker because of the intense stain (i. e., haamatoxylin),
while the gland cells are more clear and granular.

The gastric glands proper, the differentiated cells of the secreting tubules, seem
never to carry fat in other than the finest division. The gland tubes often show a
distinct reddish shade of color when stained with scarlet red. In the gastric glands
of at least one fish definite droplets were present quite large enough to be conclusively
identified as fat of the usual kind and appearance. ‘These droplets appeared to lie
near the bases of the cells, and, taken with the numerous finest liposomes present,
formed a delicate net-like mosaic. The liposomes were present in greatest numbers
in this specimen, no. 88. It seems to me that in this instance the liposomes bear a
definite relation to the increased amount of fat present in the cylindrical cells and are
to be regarded as absorption fat.

In preparations of the gastric mucosa of young salmon no. 46, fixed in Flemming’s
solution, the osmic acid has stained the fat droplets a brownish black, which brings
them in sharp contrast with the surrounding tissues. Figure 1 of a previous brief
publication concerning these facts 7 shows the superficial epithelial cells of the gastric
stomach containing the absorption fat. This black stain in the ends of the cells forms
a dense black mass, but it is granular in character. At any rate, where the black masses
are broken up granules are seen when examined under the oil immersion. Cross sections
of the necks of the crypts present rings of black granular masses around the lumen.
These masses are the blackened ends of the cells. Where the section cuts the crypt
through the opening of the gastric gland it is noted that the black masses become pro-
gressively smaller in the deeper portion of the crypt and are absent from the surface of
the secreting gland cells. The cell bodies of the superficial epithelium are stained
the dark brown of the osmic fixative. Sections across the cell just beneath the blackened
ends present numerous clear areas. These areas are spherical and very small, though

@ Greene, op. cit.

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. 169

they vary in size. The cells have all the appearance of cells in sections cut through
the pyloric cceca of this fish where large quantities of fat are known to have been present,
but is of course now dissolved out by the oils used in imbedding the material in paraffin.
In the cells of the superficial epithelium of the stomach the clear areas are smaller and
do not form so large a proportion of the body of the cell as in the epithelium of the
pyloric cceca.

Also, through the superficial epithelium one finds black round globules of rela-
tively small size in the middle of the body of the cell. These black dots correspond
to the areas above the nucleus which scarlet red shows to contain fat. (Fig. 4.) In
the base of the cells, especially in the cells of the outer folds of the epithelium, the same
black granules are present. Undoubtedly all these black granules are due to fat stained
with osmic acid. The neck cells of the crypts do not contain the black granules in the
main body. The black staining in salmon no. 45 is limited to the ends of the cells.

ABSORPTION IN THE PYLORIC STOMACH.

The difference in structure of the pyloric stomach mucosa has already been described
in a previous paper presenting the normal structure of the alimentary tract.¢ This divi-
sion of the stomach is a much more active region for fat absorption than is the cardiac
division. The main portion of the pyloric division shows more numerous fat droplets
in the cylindrical cells than is shown by the cardiac epithelium in an experimentally
fed animal.

In the region near the pyloric valve the fat fills the more superficial epithelial cells to
amaximum. The amount more nearly approaches that in the intestinal mucosa, although
the fat droplets never reach the relatively large size of those of the cells of the latter
region. A reference to figures 1, plate x11, and 4, plate xiu, will reveal the comparative
amounts of fat in different gastric epithelial regions. The crypts of this portion of the
pyloric stomach are more open and the fat is more often found in the lining cells of
their walls, even down to the bottoms of the crypts. Yet, in the most heavily fat-
loaded preparations there are always some crypts that show no fat while others may
be quite red with the stained droplets.

In the pyloric stomach, where the epithelial cells are morphologically intermediate
in character between the gastric type and that of the intestine, one can not but make
the inference that the absorptive power is also intermediate in degree. Yet the epi-
thelial cells of the free surface of the mucous fold are distinctly gastric in character, as
has already been described. The cells of the free surfaces are most loaded with fat,
as figure 1, plate x11,shows. Another point of comparison is found in the fact that the
fat droplets in the epithelium of the extreme caudal end of the pyloric stomach are
very much smaller in size than the droplets of the cells of the intestinal epithelium
just on the other or intestinal side of the pyloric valve. Sections of the pyloric stomach
have been prepared in which the first whorl of ccoeca lying close around the wall of this
part of the stomach were also cut. The epithelial coats of the cceca were in every
instance filled with very large fat droplets. The fat droplets were four or five times
larger in diameter than the droplets in the neighboring pyloric gastric epithelium.

@ Greene, op. cit.

170 BULLETIN OF THE BUREAU OF FISHERIES.
ABSORPTION FAT IN THE TUNICA PROPRIA OF THE STOMACH.

The tunica propria of the stomach is very complex in its convolutions because
of the fact that its net supports the irregularly shaped gastric glands. Varying quan-
tities of fat droplets are found in the tunica propria of the young salmon during the time
of absorption of fat. Often it happens that the connective tissue immediately beneath
the superficial epithelium is perfectly free of fat droplets, in fact, is always relatively
free of fat droplets. But during the active stage of absorption in the fat-fed specimens
occasional minute fat droplets are to be found, as shown in figure 1, plate xu. Inthelater
stages the fat seems to accumulate in the tunica propria and is removed only after long
periods of time. Incertain of the younger specimens observed the fat was still present
at a time at which the epithelial cells were approximately free of fat droplets. In these
late absorption stages the tunica propria fat is chiefly limited to that portion which lies
just within the stratum compactum. The droplets are small in size and greater in
number between the bases of the deep gastric glands and the inner border of the stratum
compactum.

It would seem that the connective tissue of the tunica propria, like that in the
intestine and pyloric coeca, holds on to its fat with great persistence. Stating the fact
in other words, the lipolytic process whereby the fat is removed from this region to
other parts of the body must proceed very, very slowly. It has seemed to the writer
that this connective tissue region serves as a temporary storage of absorption fat, also
that the process of dissociation and removal from the region is markedly influenced by
the presence of the stratum compactum. In the paper on the normal structure of the
alimentary tract emphasis was placed on the observation that the stratum compactum
is a continuous membrane. Only at points where it is punctured by blood vessels
entering into the deeper structures within the stratum is it punctured by other tissues.
This mechanical structural feature would throw upon the organs concerned the physio-
logical necessity of disposing of the fat by two possible channels. The first of these is
the vascular channel. In order that the fat may be taken up by the capillaries within
the tunica it must first be dissociated and diffused into the vascular channels. The
second possibility is that the fat may pass through the substance of the stratum com-
pactum. Here dissociation must also take place and be followed by diffusion through
the relatively thick and dense substance of the stratum. Since blood vessels of the
stomach do not form capillary nets in the statum granulosum immediately external to
the stratum compactum it follows that the fat diffusion must be carried through this
coat, i. e., the stratum granulosum, into the submucosa and muscular coats before it
could be taken up by the circulatory system and washed away into the general regions
of the body. In both instances the fat distributing process is comparatively slow,
hence one may expect the removal of the absorbed fat from the tunica propria to be
sharply retarded. These points of view coincide with the facts of observation as
measured against the time which has elapsed from the moment of feeding and absorption
to the time of the preparation of the tissue for examination.

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. I7I

THEORETICAL CONSIDERATIONS.

The observations detailed in the preceding pages made on adult normal feeding
salmon and on younger specimens under artificial and experimental feeding of fats
show beyond doubt that fat is absorbed by all portions of the alimentary tract. The
food of the salmon, which is representative of the carnivorous fishes, is made up of
living organisms. These are wholly marine forms during adult life and are represented
by the crustacean, molluscan, and piscatorial forms. All these classes of animals
possess a high percentage of fat in their tissues, particularly the fishes, which form so
large a portion of the salmon foods. Fats, therefore, form a large percentage of the
normal food substance for the king salmon. The importance of this food material needs
no further emphasis. The question at issue in this paper, therefore, is that of the ability
of the salmon to digest and absorb the fatty elements so rich in quantity in its foods.

It is of vital significance that the fats are digested and absorbed in all the great
divisions of the alimentary canal. It is true that fat digestion as such has not been
followed in this series of experiments, but much collateral evidence has been obtained,
and certain experiments not reported have shown something of the digestive process.
Of all the observations the most important would seem to be the establishment of the
fat-absorbing function of the pyloric cceca on the one hand and, on the other, the fact
that fat is absorbed in the stomach.

As regards the pyloric cceca, the function of these organs has previously been
deduced rather than proven by scientific experiment. Cuvier, at the beginning of the
nineteenth century, considered the coeca as pancreas. Ata still earlier date the general
theoretical view was advanced that the cceca had to do with absorption. In more
recent times statements have been advanced that the coeca are concerned with diges-
tion and absorption. Of course, in any division of the alimentary tract it is a safe
assumption that the function has to do either with digestion or absorption of some
one or more of the food principles.

So far as I can find, no one has, previous to my experiments, attempted to demon-
strate the relation of the pyloric cceca to fat digestion and fat absorption. The preced-
ing observations establish beyond further doubt that the pyloric cceca are primarily
fat absorbers. Incidental observations indicate that fat digestion may and does take
place in these organs as shown further in my first publication of facts from this inves-
tigation.%

The second important observation, that of the fat-absorbing power of the stomach,
is also of great physiological significance. As was indicated in the introductory dis-
cussion of the literature, the fact that fats are digested and absorbed in the stomach
has been established previously by work on mammals. Strange to say, this work has
been largely overlooked or for one reason or another questioned, so that the full accep-
tance of fat digestion and absorption by the stomach has not even yet been granted.
Van Herwerden first showed fat absorption by the stomach in fishes. Following the
publication of my preliminary report,? Weiss® published a brief report on experiments
showing the absorption of fats by the stomach of the snake. Emphasis was laid on
the fact that the fat absorption takes place more readily in the young than in the adults.

a Greene, op. cit. bIdem, op. cit. ¢ Weiss, op. cit.

72 BULLETIN OF THE BUREAU OF FISHERIES.

In fact, Weiss states that in the young cat the stomach has the power to absorb fat,
but this power is lost after a few months. Experiments carried out in this laboratory *
indicate that the ordinary laboratory mammals—the rat, the cat, and the dog—possess
the power to absorb fats not only in the young but in the adult.

It would seem, therefore, that the process of fat absorption in the stomach does
take place with somewhat greater ease and facility in the young than in adults, but
we are convinced that it is a function of the stomach which is retained throughout life
and not lost at an early stage of development, as claimed by Weiss.

The process by which fats are taken up by the mucosa of the alimentary canal is
quite naturally brought in question. The histological method used here does not follow
the digestive processes. But there are certain facts under constant observation which
indicate the nature of the absorptive process. The introductory quotation from Wells
sets forth in terse and concise terms our current views of the mechanism by which fats
are absorbed. Not only that, these views apply to the mechanism of fat transference in
the body in general. Our general conception is that lipases are produced in the body
and that through a process of dissociation the fats are split into easily diffusible forms.
This dissociation takes place in digestion. In the resulting diffusible form the fats can
readily enter the superficial border of the epithelial coat. The laws of lipolysis, as
formulated by Kastle and Loewenhart,° readily account for the resynthesis of fats when
once the fat cleavage products are present and in sufficient concentration within the
walls of the cells. That this is the process in the salmon is indicated by two proven
facts—first, the fact that fat droplets are never found exactly in the striated borders of
the superficial epithelial cells of any portion of the alimentary tract of the salmon; the
second fact is that these cells in the height of absorption are loaded with fat droplets of
such size and numbers as to gorge the bodies of the cells. In fact, numerous histological
pictures indicate that the cell boundaries are under internal tension or pressure. Figure
11, plate xv, as also a number of the other figures presented here, gives one a conception
of the physical condition of the cell when loaded with fat. This condition can be
explained by two links in the chain of evidence assumed by our present theories of
fat mobilization. The first of these is the fact that during rapid digestion of fats, say
in the cavity of a pyloric coecum, the fatty acids and the glycerin will diffuse through
the free wall of the columnar epithelial cells at a very rapid rate. Synthesis within
the cell will convert these fat cleavage products into the relatively inert fat molecules
which accumulate in ever increasing quantities. This removal of the cleavage products
maintains an osmotic condition favorable for further and continued diffusion into the
cell, thus producing a distinct pressure in an already mechanically overdistended tissue.

Emphasis can be laid on this process as an explanation of the enormous loading
of the fats, as shown especially in figures 4, 5, and 8; also in lesser degree in a majority
of the figures presented. In many instances the fat droplets within the cells are so large
as to occupy the full diameter of the cell, and so numerous as to load the entire outer
end of the cell from surface border to nucleus. When an epithelial cell is thus loaded
with fat the fat is of mechanical necessity arranged in the regular beaded rows that give
the diagrammatic appearance which is often presented by the figures.

a Greene and Skaer, op. cit.
b Kastle and Loewenhart: Concerning lipase, the fat-splitting enzyme, and the reversibility of itsaction. American Chemical
Journal, vol. XXIV, 1900, p. 491.

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. 173
SUMMARY.

In summarizing the results presented in the preceding pages the salient facts may
be mentioned as follows:

1. Fats are absorbed through the columnar epithelium of all portions of the
alimentary tract of the king salmon.

2. The primary function of the numerous pyloric cceca is that of fat absorption.
Probably the larger portion of the fats of the food of the salmon are absorbed by way
of these organs.

3. The intestine is a region of active fat absorption. The power of the intestinal
epithelium to take up fat is similar to that of the pyloric coca.

4. The salmon stomach is also a fat-absorbing organ. Fat is absorbed by both the
cardiac and the pyloric types of columnar epithelium.

5. The microscopic indications are that the fats pass through the outer portions
of the columnar epithelial cells in a dissociated form and that resynthesis takes place
within the cell, thus accounting for the numercus large fat droplets present in the cells
during active fat absorption.

DESCRIPTION OF FIGURES.

The following list of figures was drawn for me by Mr. George T. Kline, biological
artist of the University of Missouri. It is difficult, especially in low magnification
figures, to represent the exact amount of fat in the plain of a cross section. But the
relative amount is represented and by the aid of a camera lucida. In the figures of
high magnification the exact size and location of every droplet has been followed with
the greatest care. Figures 5 and 6 represent preparations in which the fat was stained
by osmic acid. All other figures are from sections prepared from frozen section stained
with scarlet red.

PLATE XII.

Fic. 1.—Showing fat absorption by the epithelium of the pyloric portion of the salmon stomach.
This fish was a young specimen from the McCloud River, Baird, Cal. It was fed olive oil 20 hours
before preparation. The superficial epithelium is crowded with fat. Other portions of the same section
show even a greater loading, extending down to the cells of the bottoms of the crypts. Traces of fat
liposomes are noticed in the lymph vessels in the folds. Fat-fed salmon no. 88. Camera lucida outlines.
Magnification, Leitz ocular 1, objective 7.

Fic. 2.—Transverse section showing fat absorption in the posterior loop of the intestine in a fat-fed
salmon from Brookdale, Cal. This young specimen had been fed 18 hours previous to killing. Fat is
crowded into the cylindrical epithelial cells, and has passed in considerable quantity into the spaces of
the tunica propria. The folding of the intestinal mucous epithelium is relatively simple in young
salmon of this age. Brookdale salmon no. 45. Camera lucida outlines. Magnification, Leitz ocular 3,
objective 3 with the lower lens removed.

PLATE XIII.

Fic. 3.—Showing fat absorption in a transverse section through the intestine of fat-fed salmon no. 45.
This figure represents with larger magnification one of the folds shown in figure 2. The general outlines
of the figure are drawn with camera lucida. The fat of the epithelial cells was laid in primarily from
this section, but in part from a comparative study of other sections. The fat-bearing portion of the
epithelium between the two goblet cells to the left was torn in the section, and this portion is all recon-

19371°—vol 33 —15——12

174 BULLETIN OF THE BUREAU OF FISHERIES.

structed from the study of similar folds. ‘The effort was made to present an accurate picture of the
relative amount and distribution of the fat in the cells. The rather regular beaded arrangement of fat is
shown in sections of the same fish, fixed in Flemming, in which the fat droplets are stained black in
figures 5 and 6. It is also shown in material fixed in corrosive sublimate, where the fat has been dis-
solved out, leaving fat vacuoles. The inner ends of the epithelial cells contain only slight quantities
of fat. No fat is ever found in the outer borders of the mucous cells.

The tunica propria is filled with a medium load of fat, the fat being caught in the spaces of the tissue
and in the connective tissue cells. This fat is all laid in with the camera lucida. No fat is present in
the stratum granulosum, either in the cells or in the supporting connective tissue. Traces of fat are
present in the connective tissue surrounding the blood vessels, and also in the vessel endothelial cells.
A few fat droplets are also present in the cells of the muscular coats, especially in the muscularis longi-
tudinalis. Magnification, Leitz ocular 1, objective 7.

Fic. 4.—A high magnification of a section through the superficial fold of cylindrical epithelium of
the cardiac portion of the stomach showing fat absorption in an early stage of the process. The fat is
largely limited to the outer or most superficial zone of the cylindrical cells, but small amounts are present
in the basal zone. ‘This salmon had been fed olive oil by the method of rectal injection, the oil passing
through the intestine and forward into the stomach. Brookdale young salmon, no. 46. Camera lucida
outlines. Magnification, Leitz ocular 1, objective 7.

Fic. 5.—A section of a group of epithelial cells of the pyloric intestine fixed in Flemming’s solu-
tion to show fat absorption. Young salmon no. 45 from Brookdale, Cal., which had been fed fat artifi-
cially. The amount of fat is not so great as present in the section of the caudal length of the intestine
shown in figure 4. The same beaded arrangement of fat droplets is shown, but more smaller droplets
are present in the ends of the cells—a fact showing either an earlier stage or a slower rate of absorption.
Camera lucida drawing. Magnification, Leitz ocular 1, objective 7.

Fic. 6.—Showing fat in the transverse section of a fold of the pyloric ccecum of young salmon no.
45, the same fish asin figure 2. The salmon was previously fed olive oil by rectal injection and the tissue
fixed in Flemming’s solution. This section presents a typical picture of fat absorption in the pyloric
cceca when the process is at its height. It is splendidly fixed, sharply stained, and is reproduced under
camera lucida with the greatest possible care. Note the fine division of the fat droplets shown in the
outer margin of the cylindrical epithelial cells, also the relatively small amount of fat of the zone within
the nucleus. The tunica propria contains an excessive quantity of fat, the boundary limit of which
is sharply marked by the broad band of the stratum compactum. In this specimen an occasional minute
liposome is present in the connective tissue of the stratum granulosum as well as in the muscular coats,
a fact that is very seldom shown. Camera lucida outlines. Magnification, Leitz ocular 1, objective 7.

PLATE XIV.

Fic. 7.—Showing fat absorption in the pyloric coecum 18 hours after fat feeding. Young salmon no.
45. In this specimen fat is crowded in the superficial epithelium, also in the tunica propria. The
details of histological structure are largely omitted in order the better to emphasize the great amount of
fat present. The droplets in the tunica propria are especially numerous in this particular fish. The
stratum compactum forms a sharp outer limit to the fat-bearing zone of the tunica propria. Fat-fed
young salmon no. 45. Magnification, Leitz ocular 2, objective 3. 45.

Fic. 8.—Showing fat absorption in the pyloric coecum 42 hours after fat feeding. Young salmon no.
46, the same fish from which figure 1 was taken. The fat is largely removed from the superficial
epithelium, except in the tips of the folds, but is supercrowded in the tunica propria. Camera lucida
outlines. Magnification, Leitz ocular 2, objective 3.

Fic. 9.—Showing fat absorption in the pyloric coecum of fat-fed young salmon no. 88, from the
McCloud River, Baird, Cal. ‘The structural detail is shown in only one-half the figure. Fat is rather
evenly distributed throughout all portions of the cylindrical epithelium and is present in medium
amount in the tunica propria. There are a few small droplets in the outer muscular coat. Time of
absorption, 20 hours. Magnification, Leitz ocular 1, objective 4.

Fic. 10.—Showing fat absorption in the pyloric coecum of a young McCloud River salmon no. gr, after
70 hours of absorption. This specimen shows the epithelial cells unusually crowded with fat. The fat
has not yet reached the tunica propria, although the time for possible absorption is longer than in no. 88
of the same experimental series. Camera lucida outlines. Magnification, Leitz ocular 2, objective 4.

FAT-ABSORBING FUNCTION OF ALIMENTARY TRACT OF KING SALMON. 175

PLATE XV.

Fic. 11.—Showing fat absorption in the superficial epithelium of the free margin of a mucous fold of
the pyloric ccecum of young salmon no. 46. This figure represents the maximal loading of fat. Many
of the cells are so gorged with fat that their surface outlines are projecting as though under a high internal
osmotic pressure. Camera lucida outlines. Magnification, Leitz ocular 2, objective 1/12.

Fic. 12.—Showing fat absorption in a portion of two adjacent folds of pyloric cylindrical epithelium
from a normally feeding adult salmon from the fishing banks of Monterey Bay. The clear marginal zone
is well shown in this figure, also the characteristic finely divided liposomic fat immediately beneath it.
This zone shades off into one of larger droplets lying just external to the nuclear layer. Note the com-
paratively small amount of fat in the inner zone of the epithelium and in the thin layer of the tunica
propria. The fat droplets are most carefully laid in from camera lucida outlines. Salmon number 22.
Magnification, Leitz ocular 1, objective 1/12.

Fic. 13.—Showing fat absorption in a normally feeding adult salmon, no. 28, Monterey Bay, Cal.
This figure represents a later stage of absorption than the proceeding. It shows a loading of the inner
ends of the cells with finely divided liposomes and a similar charge of fat in the adjacent tunica propria.
The fat has passed the outer zone. This stage of fat absorption was rather rarely observed. Camera
lucida outlines. Magnification, Leitz ocular 1, objective 1/12.

us = ware >
ofa © bit deme
mt is A of wis “Orage? 2
" SP. 6 ie te Oy
Tye iti { i ee ll 7
i et ite Tae >= * th.

i : i : : a os

BUELL. Ue os Becks, 1913; PLATE XII.

5

Bonet, Whe ksh Jes Jey Chey. PATE eli.

oY ee
np \ oe Om as
Roti

bp Et

>

Be0-/- . y yi a5
* > Z ~ 4 aeZ “;

Cae

XIV.

PLATE

Bak, LOLs.

Ss.

UWE

BULL.

10.

Fic.

9.

Fic.

Bury. U.S. B. F., 1973. PLATE XV.

NOTES ON THE HABITS, MORPHOLOGY OF THE REPRODUCTIVE
ORGANS, AND EMBRYOLOGY OF THE VIVIPAROUS
FISH GAMBUSIA AFFINIS

&

By Albert Kuntz, Ph. D.
St. Louis University School of Medicine

177

CONTENTS.

&
Page.
Mritrod ction teyseaeeyreivecsie ais csoicisisieseis ciere ecbelshets vhs Wieveinyeye Wishes bise-cisie ised sisje eyeldlalsve’eaiere ssieiere Biomere eed 181
IRE PIOGUCELV SOLE ANS arey. ype ota! cies cyaisreiais hetavsp wes chene Hew s\ stot cuslayedasann fouel eV ayals i.e eveyecaig aie ouside alerolnus lshanaraver hese 182
err al CBee amr teres caroo eee RTS: Soke gee ter see ei ace fetes fe Gye ane aas tel ges oN ita, oe eosband eralensiensteo es onaD Nays 182
Owanyercers pictee oerirer shsvaeers hse 182
MEAL eect tetetedetetey eget yes Pssst tala fer he ote. ceat etagenen Pete to evas ate taictetae-cereeaset vets aie Gris ai apegsvate eotis etevansfestusiavevetey ars 185
Modifiedtanalsfint iis’; cx acess setae ye eel les seieitin emir ere elaceatbeie iy se cote lela asteeiaiicie che ners 185
Mechanismicontrollingzanial fis sey se rere erereaeie cerrerste e eretey nie co evenerelsfele seieicio wm Sievers sycievoreye ere 185
RI CSEASH peter tonstercre setae SIMrey eet Nap eacrete re es ote Sere re tn, Mat ants redone y Cee pbigng nes er atelnts wre eects fo 186
NES TTA PDEA VO LORY parsers cy TOTTI oS ote shsrove TEI Slaye OD) a lei enenanciavard S/eiSfoneiere ie Wana ey Sue) shelve Lie ch eetele osnaye s Goreyeres 187
(Ova terra Fete re cre casey ees cays teva yesssones easy ei cyeistelnycyarsc vase $15 areuohsva, apakeusy ate ate a endholeichayagaiata a leyelm siacslbadevaisiangieve’e 90s 187
iBlastoderm 2... --cheneace cts s + sek bees ars Se Coe DORR IOS a id, SOE IOC ERT cern 187
Mitterenitiation Oks tMerexr ML YO wwe terres cies eAerer Tee chet eter heel sr ets rs revarere he teevevsye s1S avert lone 187
ater development teers cern rei ateetryae ng rs as ine elaren note cope atau lerebepfensarsensts olseahe aicists 188
CVn a /easoon doce anaseaoondsosoo ea cou Ndenndesooeds ocboeEC dee oraEdaoomcbmpoolacde 189

NOTES ON THE HABITS, MORPHOLOGY OF THE REPRODUCTIVE
ORGANS, AND EMBRYOLOGY OF THE VIVIPAROUS FISH
GAMBUSIA AFFINIS.

&

By ALBERT KUNTZ, PH. D.,
St. Louis University School of Medicine.

*

INTRODUCTION.

Gambusia affinis (Baird and Girard), according to Smith (1907), “is found along the
coast from Delaware to Mexico and reaches inland as far as Illinois. In North Caro-
lina it is excessively abundant in the lowlands, in swamps, ditches, creeks, and also
in the open waters of the rivers.’’? It is known primarily as a fresh-water species,
but occurs also in brackish water. Early in July and again on August 1, 1912, a con-
siderable number of these fish were taken along the swampy borders of Mullet Pond,
Shackleford Banks. The specific gravity of the water in which they were taken on
August 1 was 1.0081. This reading, however, does not represent the normal specific
gravity of the water along these swampy borders, as considerable rain had fallen during
the preceding 12 hours. At the time the above reading was taken, water from the
central part of Mullet Pond showed a specific gravity of 1.0106. On July 24, 1912, a
single specimen was caught in the seine in the terrapin pens at the Beaufort Laboratory.
This was a large female, bearing mature, unfertilized eggs. The water in these pens
is salt, only a very little fresh water entering through small pipes from an artesian well.

Twenty-four of the fish taken in Mullet Pond on August 1 were transferred to sea
water in a small aquarium, where they remained for a period of 19 days. At the end
of this time 1 was found dead. The remaining 23 were apparently in a normal con-
dition; they had, however, lost much of their pigment and their tissues had become
slightly transparent.

During the entire month of July, 1912, these fish were present in abundance in a
brooklet emptying into Beaufort Harbor just east of Beaufort. The water in this
brooklet is supplied largely by springs. It was reddish brown with organic matter
and contained considerable débris. Most of the fish used in this study were taken in
this brooklet.

The generic name Gambusia is derived from the name ‘‘Gambusina,’’ commonly
used in Cuba, which means ‘‘small’’ or ‘‘of no importance.’’ While of no commercial
value, this species has an important economic worth. It feeds largely on insects and
insect larve. Wherever it inhabits waters in which mosquitoes breed the mosquito
larve constitute its principal food. The introduction of these fish into the natural

@ Smith, H. M.: Fishes of North Carolina. North Carolina Geological and Economic Survey, vol. 1, p. 153.
181

182 BULLETIN OF THE BUREAU OF FISHERIES. .

waters as well as into artificial ponds, aquatic gardens, etc., in mosquito-infested regions
may, therefore, play an important réle in the extermination of these pests. Experi-
mental work of this kind already undertaken in New Jersey suggests that the plan of
combating mosquitoes by the introduction of Gambusia and other fishes with similar
habits is entirely feasible.

As was pointed out by Seal % (1908), Gambusia and the related genus Heterandria
possess certain habits and characters which render them superior to all other fishes
as mosquito destroyers. As suggested by their common name, top minnows, they feed
at the surface. Being of small size they readily find their way into shallow waters
which are inaccessible to larger fishes. Gambusia affinis is often found in large numbers
in water less than an inch in depth. Furthermore, it habitually searches for food among
the vegetation and débris along the borders of pond or stream. In Mullet Pond it is
rarely found in the open water, but is present in abundance among the marsh grasses
along the swampy borders, where it not only finds food but is also protected from larger
fishes. :

The small size of this species, its viviparous habits, and its hardy nature ought to
render its introduction and maintenance in new waters comparatively easy. It thrives
under a wide range of conditions. Furthermore, the young, being brought forth in an
advanced stage of development, are not subjected to many of the dangers which beset
the young of oviparous fishes.

The breeding season continues during the spring and summer, several broods being
produced during the season. Seal ° (1911), observing these fish in captivity, has demon-
strated that two or more generations may be born in a summer.

The adult females vary greatly in size, ranging from 3 to 6.5 centimeters in length.
The males are relatively fewer in number and smaller than the females. The adult
males range from 1.8 to 3 centimeters in length. Nearly all of the adult females taken
by the writer during July, 1912, carried either mature ova or embryos.

The present investigation was carried on at the United States Fisheries Laboratory
at Beaufort, N. C., during the summer of 1912.

REPRODUCTIVE ORGANS.

FEMALE.

Ovary.—The ovary is located in the abdominal cavity just beneath the air bladder
and dorsal to the posterior portion of the intestine. It opens directly into the urogenital
sinus, which communicates with the exterior through the urogenital aperture just
posterior to the anal opening. It is a paired tubular organ, but, unlike the ovary of
many teleosts, it is not bifurcated and has no distinct median wall. The left side of the
ovary is always shorter than the right. (Pl. xvi, fig.7.) This disparity in the length of the
two sides of the ovary is due to the position of the stomach, which is located in the left
side of the abdominal cavity. When distended with mature ova or embryos, the ovary
fills the greater part of the abdominal cavity beneath the air bladder and causes con-
siderable distension of the abdominal walls. At the left the ovary in this distended

@ Seal, William P.: Fishes in their relation to the mosquito problem. Bulletin Bureau Fisheries, vol. xxvm, 1908, p. 831-838.
> Seal, William P.: Breeding habits of the viviparous fishes Gambusia holbrookii and Heterandria formosa. Proceedings of
the Biological Society of Washington, vol. xxrv, p. 91-96.

HABITS, MORPHOLOGY, AND EMBRYOLOGY OF GAMBUSIA AFFINIS. 183

condition presses forward against the stomach, while at the right it extends anteriorly
alongside the latter organ. (PI.XxvI, fig. 8.)

Unlike the ovary of many oviparous teleosts, the ovary of Gambusza is not lobulated
and contains relatively few ova. In the same ovary may be found ova in various stages
of development, ranging from almost microscopic dimensions to a diameter of 1.8 milli-
meters attained at maturity. A considerable number of ova apparently reach maturity
at the same time. These being fertilized give rise to a brood of young. After the birth
of this brood, another lot of ova reach maturity, and, being fertilized, give rise to a second
brood. Thus, perhaps, all the ova required to produce the several broods which are
born during a spring and summer may be present in the ovary at the beginning of the
season.

The larger females usually give rise to a larger brood of young than do the smaller
ones. The average number of embryos contained in the ovaries of females 5 to 6 centi-
meters in length, based on a limited number of counts, was found to be 33. The maxi-
mum number removed by the writer from a single female was 76. The number of
embryos contained in the ovaries of the smaller females ranges from 2 or 3 to about 20.

In females of this species taken in the Potomac River early in June, 1912, Smith %
found the average number of embryos contained in the ovary to be 100. This average
is considerably greater than the maximum number observed by the present writer.
This difference is probably due to the fact that the broods observed by Smith were the
first broods produced during the season, while those observed by the present writer
were the second or later broods. As suggested by Smith, the first brood of the season
is probably considerably larger than the later broods.

The ova of Gambusia have no investment of their own save a delicate vitelline
membrane. Each ovum is inclosed in a separate cellular follicle which is attached to a
central rachis (Ryder) by a slender stalk. Running longitudinally in the central rachis
are a pair of vascular trunks from which smaller blood vessels arise and pass out along
the stalks of the follicles. These smaller blood vessels break up into capillaries which
radiate in all directions over the follicular walls. These follicles were described by
Ryder (1885) as follows: ‘‘The ovarian follicles of Gambusia containing mature ova or
foetuses are built up internally of flat or squamous polygonal cells of pavement epithe-
lium, and externally of a network of multipolar fibrous connective tissue cells and
minute capillary blood vessels with cellular walls, which radiate in all directions over
the follicle. From the point at which the main arterial vessel enters it, this vessel,
together with its accompanying vein and investment of fibrous tissue, constitutes the
stalk by which the follicle and its contained naked ovum is suspended to the main
arterial trunk and vein.” ?

In an earlier paper, Ryder (1882) has furthermore described a minute aperture
in the follicular membrane near the stalk of the follicle which he has designated ‘‘the
follicle pore.” Through this pore, he believes, the spermatozoa enters the follicle.

I was able satisfactorily to observe such a pore in the follicular membrane in only a
few instances. I have no reason, however, to doubt the presence of a follicular pore

@ Smith, H. M.: The prolificness of Gambusia. Science, vol. XXXVI, 0. S., M0. 920, 1912, P. 224.

> Ryder, John A.: On the development of viviparous osseous fishes. Proceedings of the U.S. National Museum, vol. vm,
DP. 147-

¢ Ryder, John A.: A contribution to the embryography of osseous fishes, with special reference to the development of the
cod (Gadus morrhua). Report of the U. S. Fish Commission, 1882, p. 461.

184 BULLETIN OF THE BUREAU OF FISHERIES.

in all the ovarian follicles. Without assuming the presence of an aperture in the fol-
licular walls, it would be difficult to understand how the spermatozoa could come in con-
tact with the ova.

I can not agree with Ryder (1785), however, that ‘‘the ovary itself seems to have
no exterior investment, so that the follicles lie directly within the abdominal cavity,
the young fishes upon the completion of their development rupture them and escape
into the latter, and from thence through the abdominal pore into the outer world.” @
The ovary, as stated above, is a tubular organ which opens directly into the urogenital
sinus. When distended with advanced embryos, the exterior walls of the ovary are
very tenuous. The young fishes do not, however, break out into the abdominal cavity,
but pass out of the ovary directly through its opening into the urogenital sinus, thence
to the exterior. There is no aperture leading directly from the abdominal cavity to the
exterior. Furthermore, examination of the ovary of a female immediately after she
has given birth to a brood of young shows the walls of the ovary intact. No ruptured
ovarian follicles communicate with the cceelom. When the exterior walls of the ovary
are dissected off, the ruptured ovarian follicles are found in place ina somewhat shrunken ~«
condition.

It may not be amiss at this point to call attention to an error which appears in the
recent paper by Seal (1911) referred to above. ‘‘ The ova of a full-sized Gambusia are,”
he says, ‘“‘when fully developed, about an eighth of an inch in diameter, transparent
and nonadhesive. Each one is held, apparently, by a thread of membrane to a central
nucleus, the character of which could only be determined by microscopic observation.
The young fish can be seen fully formed, their eyes moving as they turn around in the
egg.” ?

That the author quoted above has mistaken the ovarian follicle for the egg is obvious.
The embryo is developed at the surface of the egg, which has no investment of its own
save the vitelline membrane. When the yolk has been absorbed by the embryo there
remains no trace of the egg. The young fish is then inclosed in the ovarian follicle which
is suspended to the central rachis by the structure referred to in the above quotation
as ‘fa thread of membrane holding the egg to a central nucleus.”

In the paper quoted above (p. 93), Seal describes the extrusion of the young as
follows: ‘‘They are expelled one at a time and the ejection of each fish is so rapid that
they appear as though shot out with some force. This, however, might be due to the
bursting of the follicle and the uncoiling of the fish as it is released from restraint.
* + = The follicles are undoubtedly ruptured at the moment of extrusion, whether
inside or out I have never succeeded in observing, but it appears the more probable
that it is inside.”’

In view of the fact that the ruptured ovarian follicles are found in place in the
ovary after the young fishes are extruded, it is obvious that the rupturing of the follicle
occurs not only within the body of the parent but within the ovary. The young fish
is, doubtless, uncoiled as soon as it leaves the follicle. This uncoiling could, therefore,
add little to the force with which the young fish is extruded. The rapid escape of the

@ Ryder, John A.: On the development of viviparous osseous fishes. Proceedings of the U. S. National Museum, vol. vm,
p. 148. -

b Seal, William P.: Breeding habits of the viviparous fishes Gambusia holbrookii and Heterandria formosa. Proceedings of
the Biological Society of Washington, vol. xxIVv, p. 93, 1911.

HABITS, MORPHOLOGY, AND EMBRYOLOGY OF GAMBUSIA AFFINIS. 185

young fish, if, as is usually the case, it comes out head foremost, may be readily explained
by the tapering form of its body and by its own swimming movements. That some force
is necessary, however, for the extrusion of the young is evidenced by the perceptible
contractions of the muscles of the abdominal walls of the parent just before the young

is extruded.
MALE.

Modified anal fin.—The male members of the species may be readily recognized by
the modified anal fin, which functions as an intromittent organ. The third, fourth,
and fifth rays of this fin are enlarged, greatly elongated, and variously curved. All of
the rays are composed of segments. The diameter of each segment is slightly greater
at the ends than in the middle. Thus each ray shows a series of slight circular ridges.
These ridges are most prominent on the third ray, which is the largest of the elongated
rays and has a slight backward curve near its proximal end. The distal portion of this
ray bears a row of short, pointed spines on its anterior aspect, while posteriorly it is
fringed, a short distance from the tip, by a dentate ridge apparently in the fin membrane. _

The proximal portion of the fourth ray has a gentle forward slope until it comes
into close proximity with the third. From this point the former ray extends distally
parallel with the latter. The fourth ray is slightly longer than the third. Its distal
portion is divided, the two divisions diverging for a short distance and again coming
in contact with each other at the tip. The anterior division bears a few very small
spines anteriorly. The posterior division bears a considerable number of short, slender
spines posteriorly a short distance from the tip. The proximal ones of these spines are
arranged in two groups of three spines each. The fifth ray makes a short, sigmoid
flexure at its proximal end and then extends distally parallel with the fourth. Near its
distal end it makes another slight sigmoid flexure and terminates in a small hammer-
shaped enlargement which interlocks with a slightly recurved hook on the posterior
division of the distal portion of the fourth ray. The third, fourth, and fifth rays of the
anal fin are bound together by the fin membrane. The fifth ray may be brought for-
ward at one side of the fourth until it comes into close or immediate proximity with the
third. In this manner a groove or tube is formed through which the milt is transmitted
into the genital aperture of the female. The first two and the last five rays of the anal
fin are somewhat modified but not elongated.

The rays of the modified anal fin are illustrated in figure 2, plate xvi. Figure 3,
plate xvi, shows the distal portion of the three elongated rays drawn in detail under
higher magnification.

Mechanism controlling anal fin.—The modified anal fin is controlled by a powerful
muscle which is inserted on the proximal ends of the rays of the anal fin and has its
origin on the modified hamal spines of the first three caudal vertebra and a similar
process projecting ventrally from the fourth to the last abdominal vertebra. This
muscle stands in an almost vertical position and is so large that it causes a perceptible
bulging of the body walls just above the vent. (Pl. xv1, fig. 1.)

The process projecting ventrally from the fourth to the last abdominal vertebra
has a slight forward slope. The modified hemal spines of the first three caudal ver-
tebre project forward into the abdominal cavity in an almost horizontal position. The
first hamal spine is nearly straight, having a slight downward curve near its distal end.

186 BULLETIN OF THE BUREAU OF FISHERIES.

The second makes a slight forward curve near its origin. From this point a somewhat
flattened forked process extends posteriorly, one prong of the fork passing on either
side of the third hemal spine. These two prongs terminate in footlike enlargements in
the muscles of the anterior caudal segments. The third hemal spine makes a somewhat
stronger forward curve near its origin than the second. From its proximal portion a
short, flattened, keel-shaped process extends posteriorly. The distal ends of these three
hemal spines are connected by a narrow band of cartilage. (Pl. xvi, fig. 2, HS.)

The interhemals are correspondingly larger in the male than in the female and
are embedded in the large muscle controlling the modified anal fin. The one articulat-
ing with the third ray of the modified anal fin is greatly enlarged and articulates loosely
with the two anterior processes on which the muscle has its origin. (Pl. xv1, fig. 2, 1H.)

The mechanism controlling the modified anal fin projects anteriorly into the abdom-
inal cavity to such an extent that the space allotted to the air bladder becomes somewhat
restricted. Consequently, the latter organ is relatively shorter and occupies a more
oblique position in the male than in the female.

Ryder @ (1885) has given us a brief description of the modified anal fin of Gambusia
and the mechanism by which it is controlled, which is in many respects erroneous. A
comparison of Ryder’s description with the description given above will not be attempted
in this paper. The former description, published more than a quarter of a century ago,
was obviously not the result of an exhaustive study.

Testis.—The testis, like the ovary, is a paired tubular organ and is not distinctly
divided. (Pl. xvi, fig. 4.) It is located in the abdominal cavity dorsal to the posterior
portion of the intestine and just anterior to the large muscle controlling the anal fin.
The testis does not extend as far anteriorly as does the ovary, but, like the latter organ,
the left side of the testis is shorter than the night.

The spermatozoa are contained in spermatophores, which are rounded or spherical
bodies, 0.1 to 0.2 millimeters in diameter. (PI. xvi, fig. 5.) The walls of the spermato-
phores are exceedingly delicate. If the spermatophores are ruptured under the micro-
scope, the spermatozoa may be seen to escape freely even though they are still immature
and inactive. The spermatozoa are comparatively large. Each one is composed of a
comparatively large, elongated, slightly curved and bluntly pointed head, a middle
piece which is nearly as long but more slender than the head, and a long flagellate tail.
(Pl. xvi, fig. 6.)

In most of the spermatophores observed, the spermatozoa were inactive and
apparently curved around a small, bubble-like body, thus forming a more or less com-
plete ring. When the spermatophores were broken many of the spermatozoa were
released from this curved position and freed from the small, bubblelike body. The
tails, however, still retained a slight curve. The heads of the spermatozoa may be
readily observed in the spermatophores under moderately high magnification. They
are closely aggregated but show no regular arrangement. While no spermatophores
were observed in the genital organs of the female, it is highly probable that the sper-
matozoa are transmitted from the male to the female in these bodies.

@ Ryder, John A.: On the development of viviparous osseous fishes. Proceedings of the U. S. National Museum, vol. vim,
PD. 143, 144.

HABITS, MORPHOLOGY, AND EMBRYOLOGY OF GAMBUSIA AFFINIS. 187

EMBRYOLOGY.

OVUM.

The mature ovum is a spherical body having a diameter of about 1.8 millimeters.
It has a gold-yellow color and, being heavily laden with yolk, is quite opaque. It is
invested by no distinct egg membrane such as invests the eggs of most of the oviparous
fishes, but is covered only by a thin, vitelline membrane. Beneath the vitelline mem-
brane the entire surface is more or less completely covered by oil globules of unequal
size and distribution. (Pl. xv, fig. 1.)

BLASTODERM.

The ova are fertilized within the ovarian follicles. Unless the time of fertilization
can be controlled, it becomes difficult to secure the earliest stages of development. The
earliest stages which were secured after fertilization showed a small blastoderm in the
many-cell stage. This blastoderm appears as a small, almost circular cap of cells which
is slightly elevated above the surface of the yolk. (Pl. xvu, fig. 2, B.) The distribu-
tion of the oil globules is not disturbed during the process of cleavage and numerous
globules may be observed through the blastoderm.

As the blastoderm increases in size, the cleavage cavity becomes plainly visible.
The germ ring is never well defined, but appears as a slight thickening of the periphery
of the blastoderm. The cleavage cavity, as observed through the overlying blastoderm,
soon assumes a somewhat triangular outline. The blastoderm becomes slightly
elongated along the axis, which becomes the future axis of the embryo. At the side of
the cleavage cavity on which the thickened area at the periphery of the blastoderm is
broadest, the blastoderm becomes thicker and more opaque. This area is symmetrically
divided by the long axis of the blastoderm and, inasmuch as it gives rise to the embryonic
shield, may be recognized as the posterior pole of the blastoderm. This area increases
in size and distinctness until the embryonic shield is well outlined. (PI. xvun, fig. 4.)

DIFFERENTIATION OF THE EMBRYO.

From the posterior pole of the embryonic shield a narrow thickened area grows
anteriorly. This thickened area alone represents the embryonic area, while the thinner
lateral areas represent the extra-embryonic area of the embryonic shield. (Pl. xvn, fig. 4.)
The embryonic area continues to grow anteriorly over the cleavage cavity and becomes
gradually enlarged at the anterior end. In this manner the head of the future embryo
becomes outlined.

While the embryonic area is becoming differentiated the blastoderm spreads rapidly
over the yolk until the latter is completely covered. The progress of the growth of the
blastoderm over the yolk could not be observed satisfactorily, partly because the germ
ring is not well defined and not easily observed on this very opaque egg and partly
because not all the desired stages of development could be secured. A careful study of
the stages available, however, seems to indicate that the differentiation of the embryo
of Gambusia takes place in a manner which is quite typical for teleosts.

188 BULLETIN OF THE BUREAU OF FISHERIES.
LATER DEVELOPMENT.

After the formation of the embryonic area the embryo soon becomes well outlined.
Plate xvii, figure 5, illustrates a stage at which the tail bud has already grown out and
the anlage of the neural axis is apparent throughout the entire length of the embryo.
The optic vesicles are well formed and from 3 to 4 somites are already apparent.

Plate xvut, figure 6, represents an embryo in which the divisions of the brain are
becoming distinctly outlined. The auditory vesicles and from 12 to 14. somites are
already present. At this stage the heart is becoming differentiated as a simple curved
tube. The heart soon begins to pulsate, and a circulation is set up over the surface
of the yolk. This circulation is at first slow and irregular but soon becomes very
vigorous.

The growing embryo lies in a groove in the surface of the yolk and is inclosed only
by the ovarian follicle. As development advances the ovarian follicle increases in size
and becomes increasingly vascular. The space between the egg and the follicle becomes
filled with a transparent fluid. Thus the embryo lives in a fluid medium. Although
the ovarian follicle becomes highly vascular, a placental or pseudoplacental relation-
ship such as exists in the selachians or even in some of the viviparous teleosts is not
suggested. The embryo develops no structures which would seem to be adapted to
absorb nourishment from a fluid medium. Furthermore, no feces of any kind are ever
observed in the follicle. The abundant yolk supply in the egg is, doubtless, sufficient
to supply all the food material required by the embryo.

It is probable, as was suggested by Ryder (1885), that ‘‘the very intricate meshwork
of fine vessels which covers the follicle supplies the developing foetus with fresh oxygen,
and also serves to carry off the carbon dioxide in much the same way as the placenta or
afterbirth performs a similar duty for the young mammal developing in the uterus of
its parent.”* The analogy between the intra-follicular respiration of the developing
embryo of Gambusia and the intra-uterine respiration of the young mammal must, how-
ever, not be carried too far. The embryo of Gambusia develops gills which apparently
become functional very early. An examination of the gills of an advanced embryo
removed from the ovarian follicle, as Ryder has already observed in the paper quoted
above, shows that the gill filaments are already pinnate and that the pinne contains
loops of blood vessels. This condition of the gill filaments, as is well known, is not
attained by the larve of many oviparous fishes for a considerable interval after hatching.
Furthermore, rythmical breathing movements may be observed as the embryo lies
coiled in the ovarian follicle. It is probable, therefore, that the intra-follicular respira-
tion of the embryo of Gambusia, at least during the later stages of intra-ovarian life, is
more nearly analogous with the respiration of adult fishes than with the intra-uterine
respiration of the young mammal, the fluid in the follicle, by which the embryo is con-
stantly bathed, being aerated by the follicular circulation. -

As the embryo grows, the tail extends posteriorly partly encircling the egg. Soon,
however, it bends indifferently to the right or to the left. (Pl. xvut, fig.8.) This bending
brings the tip of the tail into proximity with the head. Consequently, as the caudal fin
is developed it overlaps the face of the embryo, sometimes partly or completely covering
one or both of the eyes. (PI. XIX, fig. 9.)

@ Ryder, John A.: On the development of viviparous osseous fishes. Proceedings U. S. National Museum, vol. vm, p. 147.

HABITS, MORPHOLOGY, AND EMBRYOLOGY OF GAMBUSIA AFFINIS. 189

Pigmentation begins comparatively early. Scattered pigment spots first appear on
the dorsal surface, being more closely aggregated on the posterior region of the head and
along the dorsal mid-line of the trunk. These pigment spots become more numerous
and more closely aggregated until at birth pigmentation is almost complete.

Embryos which still retain a yolk sac of considerable size when removed from the
ovary show nearly all the characteristic markings of the adult. At birth the yolk sac
is completely absorbed. The newborn fish answers fairly well, except with respect to
dimensions, to the diagnostic description of the species. Its color is light olive, darker
dorsally than ventrally. The number of scales in the lateral and transverse series, respec-
tively, correspond to the number of scales in these series, respectively, in the adult. The
number of rays in the dorsal, anal, and caudal fins also correspond to the number of rays
in these fins, respectively,in the adult. The fine dark line along the side is already present.
The two or three transverse rows of dark spots on the dorsal, the dark margin on the anal,
and the three or four irregular rows of dark spots on the caudal fin, characteristic of
adult females, are already becoming differentiated. The dark purplish blotch on the
side above the vent (absent in males) is not yet apparent. The modified anal fin of the
male was not observed in newborn fishes.

The newborn fishes are 9 to 10 millimeters in length and are very vigorous. Having
been protected from many of the dangers which beset the larve of oviparous fishes, they
are now well prepared to enter upon an independent existence.

Embryos still carrying a yolk sac of considerable size, being removed from the
parent, were able to swim freely in water, where they continued to live, the yolk sac
being gradually absorbed. Such embryos were kept in small aquaria with occasional
changes of fresh water for a period of ro days. At the end of this time the yolk sac was
completely absorbed and the young fishes were apparently in a healthy condition.

SUMMARY.

1. Gambusia affints is known primarily as a fresh-water species, but occurs also in
brackish water. Under experimental conditions, fishes transferred from brackish water
to sea water were kept alive and apparently in a normal condition for a period of 10 days.

2. The ovary of Gambusia is a paired tubular organ without a distinct median wall,
which opens directly into the urogenital sinus. Each ovum is contained in a separate
cellular follicle in which fertilization takes place and the embryo is developed. At the
completion of development the ovarian follicles which are attached to the central rachis
by a slender stalk are ruptured and the young fishes are extruded directly through the
urogenital aperture.

3. The modified anal fin of the male which functions as an intromittent organ is con-
trolled by a powerful muscle which is inserted on the proximal end of the anal fin rays
and has its origin on a bony process projecting ventrally from the fourth to the last
abdominal vertebra and the modified hamal spines of the first three caudal vertebre.
The third, fourth, and fifth rays of the anal fin are enlarged, greatly elongated, and
variously curved, bearing short spines on the distal portions. The interheemal which
articulates with the third ray is enlarged and articulates with the two anterior
processes on which the muscle controlling the anal fin has its origin.

19371°—vol 33—15——13

190 BULLETIN OF THE BUREAU OF FISHERIES.

4. The testis, like the ovary, is a paired tubular organ. The spermatozoa are con-
tained in the spermatophores and are probably transmitted from the male to the female
in these bodies.

5. The formation of the blastoderm and the differentiation of the embryo takes
place in a manner which is quite typical for teleosts.

6. As development advances, the ovarian follicle becomes highly vascular, increases
in size, and is filled with a transparent fluid in which the embryo is constantly bathed.
This fluid is doubtless aerated by the follicular circulation. ‘The gills of the developing
embryo apparently become functional comparatively early. During the later stages of
intra-ovarian life, rhythmical breathing movements of the embryo may be observed.

7. The young are born in an advanced stage of development and show nearly all of
the diagnostic characters of the species. They undergo no marked metamorphic
changes after birth,

Toni; Wh Sh Io Ia, eMey PLATE XVI.

Se

Fic. 1.—Dissection of male Gambusia, showing testis and mechanism controlling anal fin, x 7.2; AF modified anal fin;
HS, modfied hemal spines; /, intestine; 17, muscle controlling anal fin; 7, testis; VP, ventral process of abdominal
vertebra.

Fic. 2.—Skeletal parts of mechanism controlling modified anal fin; AFR, anal fin rays; HS, hemal spines; JH, inter-
hemals; VP, ventral process of abdominal vertebra.

Fic. 3.—Distal portion of modified anal fin greatly enlarged.

Fic. 4.—Testis, x 9.

Fic. 5.—Spermataphore, x 38s.

Fic. 6.—Spermatazoon, X 3,000.

Fic. 7.—Ovary, X 4.
Fic. 8.—Dissection of female Gambusia, showing ovary distended with advanced embryos, x 3.6.

BULL, 5. E., rons: PLATE SOVIL.

Fic. 1.—Mature ovum, x 45.

Fic. 2.—Ovum with early blastoderm, x 455 B, blastoderm.
—Ovum with later blastoderm, x C, cleavage cavity.
—Ovum with blastoderm, showing embryonic shield, » 45; ES, embryonic shield; EA, embryonic area; GR, germ.
ring.

Fic

Fic. 5.—Embryo with 3-4 somites, x 50.

Fic, 6.—Embryo with 12-14 somites, X 50.

F —Embryo with about r2 somites, side view, x 50; H, heart.
Fic.

.—Embryo with pigmentation started, inclosed in ovarian follicle,

PLATE XVIII.

45.

BULL o. Bal, Oi3: PLATE XIX,

aC ENT
Ce, 7
eee use &s

as «: & < Re

4

Fic. 9.—Advanced embryo inclosed in ovarian follicle, x 30

Fic, 10.—Embryo with yolk sac nearly absorbed, removed from ovarian follicle, x 18.
Fic. 11.—New-born fish, dorsal view, xX 15.

Fic. 12,—New-born fish, side view, x 15

—- o>

SPOROZOON PARASITES OF CERTAIN FISHES IN THE VICINITY
OF WOODS HOLE, MASSACHUSETTS

a

By © W. Hahn

IgI

CONTENES:

Occurrence of disease
Methods! of: stud y-oepo4.ci: sfeishe beicre oe Stopate o folor a, alot deus © che be fc Rhatotete al aj araoralehe, eocfeereratacceieral al steer ere rere
Experiments to determine character of infection
Pathological condition ofjthestissnes ts.) fa7a8 afte. c Gm cnet chtori is teeta tecite cacti siete octet
Bacteria associated) with! atrophied Pisses’ esses. wiciecisie + vince uns aelessie cisisisitieiclete oe cleriaieiereisieiestlet ie
Sporozoa associated with atrophied tissues

Stages of Myxobolus musculi

ChHloromiyscuimn ftin d tlie sa apoeyacans, bya yaye wg crs toye © ap egegose osc avat tg ove, sve wvavehccavaler Shere eiayerereeee eye een user
Protozoa related'to:those here: described ALG Fy, teint hs etveishesletenetatche niele tu lsteteteleleravancystelene nel eretetee terse
Prevalence of myxosporidian infection in fish
General conclusions. ..............
Bibliography ss 2 \ie,s.ccaics ves oes bts yaieks leis sche Sls sie ele fol pscaeleel el eue Coie eiche eins eeelee Ries eieiekslaleter ere
Explanation Of plates’ j sciicvatins ot citace.s asieveweyeree svosake eavove ler elaverain ero erepalterel the manera terest iciateperteneer ees

192

SPOROZOON PARASITES OF CERTAIN FISHES IN THE VICINITY
OF WOODS HOLE, MASSACHUSETTS.

Bead
By C. W. HAHN.
&

While studying the Sporozoa in different species of fish at Woods Hole, Mass., in
1909, the myxospore of one was observed in diseased killifish, Fundulus heteroclitus
and Fundulus majalis. Additional material was obtained and some special experiments
were carried out during the seasons of 1910, 1911, 1912, and 1913, the United States
Bureau of Fisheries providing the facilities for this and other similar studies at its
Woods Hole biological laboratory. %

OCCURRENCE OF DISEASE.

When a number of Fundulus of either of the common species (heteroclitus or
majalis) are confined in aquaria for a few days during the warm season, one or more
thickened white or pink areas appear upon the integument of some of the fishes. The
scales of these patches.are more or less loosened. They increase in size and number, and
the number of afflicted fishes also increases. The fins when involved become bloody and
the fin-rays are exposed. Elsewhere the integument disintegrates and the flesh is laid
bare. Considerable excavations into the body muscle are not uncommon. ‘The largest
cavity of this kind observed was in the head region, measuring about 10 to 12 mm. in
diameter and 2 to 3 mm. in depth. Such excavations expose large areas of the skull.
When other parts are attacked, loss of blood or penetration of vital parts causes death
before the lesion becomes conspicuous externally. The integument is thickened around
the sores where the scales are loose. Its color is pink or white. The scales fall out at
the edge of the sores. The caudal fin may be completely removed, also the flesh and
integument of the tail, thus exposing the vertebre, before the fish succumbs to the
disease. Fish frequently give evidence of weakness and depression even before the
flesh has been exposed. There is nothing peculiar about the locomotion except a dimin-
ished activity. In certain cases where there is conspicuous inflammation of the integ-
ument, especially under the head, the fish may be observed to dart downward, and,
with a slight rotation or twist of the body, to scrape the ventral or lateral portion of
the head upon the bottom of the aquarium. The fish slowly lose strength, the smaller
ones first, and the larger ones not until they are greatly mutilated. Apparently all
afflicted fish die unless special care is given to cleanliness, water, and food.

The proportion of fish that are diseased when caught has not been ascertained.
The ratio of those that develop integumentary sores in the first day or two to those
that are healthy depends to a great measure upon the injuries received in handling the

@ Valuable assistance from Dr. Edward Linton and Mr. Vinal E. Edwards is gratefully acknowledged.
193

194 BULLETIN OF THE BUREAU OF FISHERIES.

catch. Sometimes 50 per cent of the /undulus that have been roughly handled, as
when stripped for eggs, will become diseased in 24 hours. Of these, half may be dead
within 12 hours. If a few crabs happen to be confined in the same aquarium with a
large number of Fundulus, they inflict injuries upon practically all the fish and all are
soon diseased. Uninjured Fundulus develop the disease infrequently. (See p. 196.)
Roughly speaking, 3 to 4 per cent of the Fundulus that are brought into the laboratory
at this season (July and August) and confined in small aquaria having but a liter or
two of water for each fish, will be found diseased within two days. Within another
day or two some of these fish die and a large number die in the course of a week. Dis-
eased Fundulus are therefore almost constantly available.

METHODS OF STUDY.

Both fresh and preserved tissues were examined microscopically, the method of
handling the tissues being as follows: The scales having been removed with forceps,
the edge of a slide is drawn over a diseased area with a little pressure, and the mucus
and cellular material thus obtained is spread evenly over the surface of another slide;
or, a portion of integument or muscle which has been removed with a scalpel is ground
between two thick slides by giving to the upper slide a circular motion. It is necessary
to use considerable pressure, and at times cut tough fragments with the sharp edge of
the upper slide. Under these conditions, both slides may be preserved for observation
and still others made from the ground-up material. Some of the smear preparations
made in this manner were examined while fresh and others were fixed and stained.
Altogether about 85 fish were examined microscopically. Fresh smears which were
sometimes supplied with bile and serum were sealed with vaseline, and could then be
examined from time to time, during a period of 24 hours.

For sectioning, tissues were fixed in a saturated solution of corrosive sublimate in
35 per cent alcohol with 0.2 per cent acetic acid and 6 per cent formaldehyde; also in
the ether-formalin-alcohol mixture given below. Some of the smears were fixed in the
same sublimate mixture; others in a solution of corrosive sublimate in 2 parts abso-
lute alcohol and 1 of ether; still others in a mixture of absolute alcohol (60 per cent),
ether (35 per cent), and strong formaldehyde (5 per cent). The mercury preparations
are stained with a modification of Mayer’s hematein. (See Hahn, Archiv ftir Protis-
tenkunde, bd. xvu, no. 3, p. 316, footnote.) Usually the alcohol and formalin prepa-
rations are stained in methylene blue or Giemsa’s stain. The methylene blue was
extracted in a saturate alcoholic (7o per cent) solution of both eosin and orange G.
The Giemsa was washed in water, allowed to dry, and decolorized in carbol-xylol,
without the use of alcohol. Some of the more recent preparations fixed by either of
the above fluids have been more successfully stained by first treating with hematein
for several hours, then decolorizing in 70 per cent alcohol with 1 per cent HCl, returning
through the alcohols to water, and staining in methylene blue or toluidin blue. After
dehydration they were left in a contrast stain (eosin and orange G) for a few minutes
and rapidly run into 95 per cent alcohol, carbol-xylol, and two changes or xylol. Both
smear preparations and sections are mounted in Canada balsam without cover glasses.

Searching is most satisfactorily carried on with an ocular of 1 inch and an objective
of one-fifth inch focal distance. A one-twelfth inch oil immersion objective combined
with the same ocular for ordinary observation is supplemented, when occasion requires,
with a no. 2 compensating ocular. The one-fifth inch objective is not too high to be

SPOROZOON PARASITES OF FISHES. 195

used without a cover glass and reveals most of the details necessary to recognize the
presence of Protozoa or other unusual histological conditions. A mechanical stage is
indispensable.

Three organisms are involved in most of the Fundulus ulcers, rarely a fourth. A
thick, short bacillus is the most abundant. A long, slender bacillus is less common.
The Sporozoa are represented by a species of Myxobolus, and in one case a species of
Chloromyxum. From the evidence in the following account it will be learned that the
primary attack upon healthy tissues, in a certain proportion of the diseased fish, is prob-
ably made by the long bacillus. At least a few and probably many of the diseased fish
are primarily attacked by Myxosporidia. The short bacillus is more or less incapable
of rapid growth in living cells of any kind. While it is not within our province to make
an exhaustive study of the fungus diseases, it has been necessary to ascertain to what
extent they participate in bringing about these pathological conditions.

EXPERIMENTS TO DETERMINE CHARACTER OF INFECTION.

The following experiments were carried out in order to gain some accurate infor-
mation as to the conditions whereby healthy fish are infected and the possibilities of
their recovery. At the time it was not possible to discriminate between fish that were
infected by a fungus and those that were infected by a sporozo6én. It will be apparent
that the experiments are not vitally affected by the kind of parasite present.

Forty fish were divided equally and placed in two 5-gallon aquaria. These fish
had been seined in the usual manner and brought to the laboratory on board the steamer
Phalarope in large milk cans. The trip from the collecting grounds (Menemsha Bight)
usually requires about one and one-half hours. The cans accommodate from 200 to
300 fish each. A hose supplies them with fresh water. The 4o fish used in this case
were examined carefully and found to be free from all visible integumentary disturbance.

First stage—Aquarium no. 1 was carefully cleaned and sterilized. Aquarium no. 2
had contained diseased fish, and 2 diseased fish were allowed to remain with the 20
fish used in the experiment. Contaminated fish from other sources were always kept
in this jar. Both groups were fed about every 48 hours. After a period of 11 days
none of the fish in the clean jar showed any signs of disease. From this fact we con-
cluded that they were free from the disease and suitable for experimentation of a dif-
ferent kind. After the same period (11 days) the contaminated jar had one fish with a
conspicuous sore. It died a day later.

Second stage-—On the eleventh day one of the fish in each of the two jars was
operated upon. A scale or two was removed and the integument pierced with a scalpel
just back of and dorsal to the opercle. More diseased fish were introduced into aquarium
no. 2. Five days later the fish in aquarium no. 1 which had been operated upon died.
The integument, at the point where the incision had been made, had developed a typical
sore. At this time the fish with the pierced integument in no. 2, being a large fish, had
not developed a sore of noticeable extent.

Third stage.—On the sixteenth day of the experiment, all fish having recovered in
both no. 1 and no. 2, scales were removed and the integument of all the fish was pierced
in the same manner as was done with the two above mentioned. ‘Two days later almost
all of those in jar no. 2 had developed marked diseased patches at the very spot where
the integument had been pierced. No noticeable change had taken place in the fish of
the clean jar. Four days later one fish in jar no. 1 died from the effects of the rapidly

196 BULLETIN OF THE BUREAU OF FISHERIES.

advancing disease. Subsequent examinations of the tissues showed that the probable
cause of this disease is a myxosporidian belonging to the genus Chloromyxum, being
unique in this respect. (See p. 205.) Four dead fish taken from jarno. 2 at this time
included two that had been introduced for the purpose of spreading the disease. After
seven days the fish in jar no. 1 were all recovering. The incised integument had closed
and appeared a little white. Of those in jar no. 2, two were dead, three were seriously
diseased and died within 24 hours, and the others had conspicuous sores. The remaining
14 fish from this time began to show signs of recovery, probably because they were not
subjected to contamination and they were fed more regularly. Twelve fish remained
in jar no. 1 and had completely recovered before the experiment was discontinued.

In the above experiment the treatment given to the two jars was as far as possible
the same. Some fish escaped from both jars by jumping out.

The first stage of this experiment, which corresponds with the first 11 days, was not
conclusive. One fish, having contracted a fatal disease from a contaminated environ-
ment, demonstrates the possibility that fish with apparently healthy integument may
acquire the ulcers. The second stage of the test, covering six days, was still less con-
clusive. But the third stage, covering seven days, showed beyond doubt that the infec-
tion enters a lesion of the integument, that contamination favors its entrance, that
some of these diseases may be contracted in tolerably pure water, and that lesions
which are not contaminated heal completely.

Another experiment of this character was then started, making use of some of all
the lots of fish that had been under observation. All were in good condition. Eight
fish of fair size were carefully removed from this stock and, by means of a small steril-
ized scalpel, an incision was made back of the head and a pocket then made under the
integument so as to disturb the tissues as little as possible. Into this pocket was inserted
a bit of the diseased flesh from sores of four fish taken from different aquaria. As a
control, eight more fish of the same size were similarly cut, but nothing was introduced
into the pockets. Of the contaminated fish, four died from the disease in two days,
the balance in four days. In this case the disease spread over the whole upper part of
the body and assumed the characteristic appearance usually encountered. Only one of
the controls died. The others healed and recovered completely. From time to time the
diseased fish which were introduced into the contaminated jar and those used for the
inoculation experiments were examined microscopically. All were infected with bacteria.

This last experiment, covering a period of four days, confirms the results of the
previous experiments as to the infectious nature of the disease as well as the inability
of the fish to throw off strong cultures of the causal agents. We also learn that when
the fish is well nourished and in a wholesome environment, it has considerable natural
immunity and recovers readily from the affliction.

In order to prevent the customary mortality from this kind of affliction, care
should be taken not to injure the fish while collecting; no crabs or other carnivorous
enemies should be confined in the same tanks with the Fundulus, and after establishing
them in an aquarium without crowding, they should be fed on alternate days. The
aquarium should be kept free from dead and diseased fish. With proper circulation of
water, this treatment will no doubt reduce the mortality to a negligible quantity and
preserve the fish for several months.

SPOROZOON PARASITES OF FISHES. 197
PATHOLOGICAL CONDITION OF THE TISSUES.

Those typical sores in which Sporozoa can not be positively demonstrated, and of
which a part may be due to bacteria, present the following histological conditions. They
are probably primarily exogenous ulcers in which there is at times abundant granular
degeneration derived both from lymphocytes and hemocytes. Sometimes at the nidus
of the necrotic area there are small cysts or abcesses containing small lymphocytes.
Usually the vascular tissues abound and erythrocytes preponderate. There is a decided
tendency at times for the epidermis to form a cicatrix. Again it gives evidence of
sloughing off. But so far as the muscle tissue is concerned universal necrosis is common.

The involved epidermis contains numerous nonstaining globules or masses of
variable size (fig. 36, pl. xxI), as to the exact nature of which we are yet in doubt. They
are also to be found in the connective tissue of the dermis and in certain partly atrophied
muscle fibers when adjacent to degenerate tissue. They seem to be more numerous in
the epidermal cells wherein there are obvious signs of disintegration (pp. 198, 201, 203).
Inasmuch as there is a nonstaining zodgloea or secretion about some of the bacilli that are
commonly found in these parts, which frequently prevents them from staining (see p.
200), it is possible that these bodies are of the same nature and contain one or more of the
bacilli. No doubt many are fat globules, but some are certainly not. Some of these
bodies in sections of muscle containing myxoplasms possess a well-defined nucleus.
(Fig. 12, pl. xx.)

In smears of integument, it is occasionally possible to find fragments of considerable
size having the epidermal cells more or less filled with the short bacillus referred to above.
It is not difficult to prove, by the observation of fresh material or by comparison of tis-
sues of different stages of degeneration, that the short bacillus is seldom found in normal
living cells. It is therefore not probable that the primary attack upon the epidermis is
caused by this particular organism. The long slender bacillus is less commonly en-
countered in the dermis and epidermis. There is but little evidence in support of the
view that it is the initial cause of epidermal decadence.

The muscle fibers beneath these infected areas present an interesting condition. To
the naked eye there seem to be numerous white threads running parallel with the muscle
cells. This is especially true of well-advanced ulcers. When seen under the microscope,
such flesh has but few normal fibers with fibrille and cross striz. Most of them have the
sarcolemma and interfibrillar connective tissue still sufficiently intact to retain the general
external structure of the separate fibers, but the myoplasm is in various stages of de-
generation. We conclude, therefore, that the parasite is intracellular and does not pass
readily from one fiber to another. The muscle fibers sometimes undergo degeneration
more or less uniformly throughout their length. In some cases it is more rapid in the
immediate vicinity of the parasites. This we know from sections where the fibrilla show
in places adjacent to degenerate myoplasm in which Sporozoa are numerous. One side
or the middle may be far more degenerate than the rest of the fiber. The parasites have
probably passed through these regions. The first indication of change is the loss of
fibrillation. It is rather difficult to find a parasitized fiber showing normal fibrillation
(fig. 13, pl. xx). The pale bands of muscle fibers next become granular (figs. 1 and 2,
pl. xx) and at length the sarcous elements break up into large pieces. Eventually there
is total granular atrophy of the fiber within the sarcolemma. In certain cases, usually

198 BULLETIN OF THE BUREAU OF FISHERIES.

when the atrophy is hyalin, there are considerable clefts in the sarcoplasm. (Fig. 4,
pl. xx). These spaces may come to be more or less closely packed with erythrocytes or
leucocytes, or both, so that when the cytoplasm of the blood cells has degenerated a
third and common condition is encountered. The nuclei in various stages of degenera-
tion become densely packed and enlarged. They assume amceboid shapes, large alveoli
appear in them, and eventually they fall a prey to the short bacillus (fig. 5 and 6, pl. xx)
elsewhere encountered.

The conditions thus presented are such as to suggest an amoeboid parasite which
has demolished a muscle fiber and simultaneously broken up into innumerable bacillus-
shaped spores by schizogony. (See fig. 10, pl. xx.) The connective tissue nuclei of the
flesh and integument and the nuclei of the gill epithelium give rise to the same degen-
eration phenomena. Such nuclei may be about equally hypertrophied and massed in
such a manner as to completely disguise their true nature. Both muscle and vascular
nuclei may occur in abnormal numbers under the sarcolemma of fibers which are in
almost any state of atrophy but without clefts. (Fig. 5, pl. xx.)

In both fresh and stained muscle the evolution of a curious artifact was observed.
It appears as a dense hyalin body in the sarcolymph, between fibrille. (Fig. 3, pl. Xx.)
Assuming an amceboid form it resembles a rapidly growing organism. (Fig. 2 and 7,
pl. xx.) But the regular distribution (fig. 2 and 3) and numerous variations toward a
crystalline rosette structure are conclusive evidence of their lifeless nature.

Whatever the active cause of the degeneration of muscle fiber, be it bacteria or
Protozoa, the atrophy advances far into one or more muscle fibers without causing any
damage to the adjacent fibers. In cross sections of such tissues there may be a small
group of normal fibers cut in section amongst numerous others that are wholly degenerate.
Capillaries, arteries, veins, and sheets of connective tissue, entirely normal in appear-
ance, may also penetrate these necrotic masses. This is no doubt due to the restraining
influence of the sarcolemma upon either the parasite or toxin. As we have already
noted, the sarcolemma retains its normal relations in completely atrophied fibers.

Restricting our statements to tissues known to be infected by Sporozoa, there are
but two kinds where their action has been observed, namely, muscle, and the connective
tissue of the gill. The pathological condition of the muscle tissue, in such cases, is
not distinguishable, as far as we know, from that which results from the action of
bacteria; but if the pathological changes are to be considered as characteristic of a
parasite when it is known to be the cause of the atrophy, a careful study of those cases
where bacteria are a negligible factor is important. The myxospores, which are the
most easily recognized stages of the Myxosporidia, are common only in smear prepara-
tions and only those which include more or less diseased muscle fibers. These same
smear preparations also contain cells identical in appearance to myxoplasms, pansporo-
blasts, and sporoblasts, which happen to be the only representatives of the Sporozoa that
we have encountered in sectioned material, thus suggesting their myxosporidian character.

Several fragments of tissue, the integument of which was slightly diseased, were
sectioned. They give no evidence of myxospores, but the muscle fibers present prac-
tically the same degenerative changes to be seen elsewhere. The dermis contains
numerous minute unstained lens-shaped structures similar to those described on
page 197. These extend into the ends of the adjacent muscle fibers, becoming less
numerous in the deeper parts. Such fibers show obvious signs of atrophy. Elsewhere

SPOROZOON PARASITES OF FISHES. 199

there are numerous deep fibers containing many large cells, which vary in size and have
conspicuous nuclei. (Fig. 18, pl. xxi.) These are confined by the sarcolemma to a
very few fibers and extend for a long distance through them. A small cavity only is
excavated about each cell. They are usually isolated, though two or more may occupy
the same cavity. The sarcoplasm in such cases is much atrophied, being uniformly
granular or homogeneous. A sharp line of demarcation exists between the infected
and uninfected parts of the muscle fiber, the former being degenerate and the latter
striated and normal. Situated amongst the fibers containing the Protozoa are others
lacking them but atrophied in a typical manner, the sarcoplasm being broken into
irregular fragments. There are several other foreign and unnatural structures in the
sections just referred to, about which the details are given on page 203. Muscle fibers
packed with blood tissues and degenerate nuclei have not been found in any of the
sectioned tissues which contain unmistakable cases of Myxosporidia; but no special
significance has been attributed to this fact.

Smears of gill filaments stained with Giemsa stain present the following conditions:
Both normal and degenerate tissues are encountered. In some places the cartilage
supporting structures have been attacked and are partly disintegrated. The general
external form of the supporting tissue, including the surrounding connective tissue and
epithelium, are, as a rule, partly maintained; but elsewhere the degeneration is com-
plete. Epithelium and connective tissue cells disappear completely, leaving the elastic
fibers and blood elements. Here, as elsewhere, the nuclei of the latter are most persistent,
especially those of the erythrocytes. A large portion of the expressed fluids is composed
of an acidophile substance containing odd-shaped portions of the fused nuclei. The
spaces between the chromatin threads of the latter having become much dilated, fuse
and form large masses of network. These are mechanically separated on crushing the
tissue. Such masses of nucleic acid or degenerate chromatin have unbroken connections
with the normal blood in the arteries or veins of the less disturbed tissue. Where the
blood emerges from partly degenerated blood vessels, they are filled with atrophied
erythrocyte nuclei. It seems probable that very large masses of homogeneous eosinophil
material, which are constantly associated with the degenerate gill tissue, are derived
from hemoglobin, lecithin, etc., of the stroma.

Myxospores abound in these degenerate gill tissues, especially in the purulent
residues of degeneration where nothing else remains recognizable. They also occur
deep in the connective tissue near the cartilage and amongst the capillaries. The
spores, developing spores, sporoblasts, and pansporoblasts, in all stages, are clearly
defined, apparently unaffected by the conditions where tissue cells have become wholly
atrophied. ‘This fact, together with the great abundance of myxospores and developing
myxospores, both occurring in considerable clusters, prove beyond question that the
primary cause of necrosis in this case is the Myxobolus. No bacteria or other possible
agents have been encountered.

BACTERIA ASSOCIATED WITH ATROPHIED TISSUES.

The small bacillus above referred to (p. 195) varies greatly in size. The smallest
(fig. 8, pl. xx) measure less than 0.74 in thickness and 1.54 in length. The large ones
(fig. 9, pl. Xx) average 1.5 in thickness and 7 in length. The former are homogeneous
when stained. The latter frequently appear to have very conspicuous granules just

200 BULLETIN OF THE BUREAU OF FISHERIES.

inside the cell wall. These are probably artifacts. The older bacilli (fig. 9, pl. xx) taper
at one end. They were at first taken for protozoan spores. These bacilli occur by
thousands in and near degenerate epidermis and muscle tissue. It is not unusual to find
them grouped in the form of the cell which they have completely destroyed. They are
then of nearly uniform size (fig. 10, pl. Xx); but between individuals of separate groups,
there is often a great difference in size. They stain, as a rule, with methylene blue,
gentian violet, toluidin blue, and Giemsa stain. Inside the host cells (fig. 6, pl. xx) and
when first set free from them they stain, if at all, with great difficulty. This may no
doubt be due to a zodgleeic condition. In smears, the stretching of this secretion causes
the bacilli to be drawn into long parallel rows. The secretion then resembles elastic
connective tissue fibers and the bacteria replace the connective tissue nuclei. At times
the zodgloea is not noticeable. (Fig. 8 and 10.)

To what extent toxins emanating from the short bacillus are the cause of the death
and disintegration of the host tissues we can judge from the following facts: As already
stated, this bacillus is not to be found throughout large areas of atrophied muscle and
integument. If the toxin emanating by diffusion from a localized organism brought
about the decadence of a tissue, one would expect the evidences of such decadence
to indicate a uniform advance of said toxin in the same direction through a given tissue;
but, as we have seen, the atrophy of muscle fibers is limited to a certain few in a large
number of normal cells, or there may even be a few normal fibers extending through and
far into a necrotic region. The same relations prevail more or less in the epidermis.
If the short bacillus is to be regarded as a saprophyte, then some more virulent primary
organism must be present. In the diseased gills the abundance of M. musculi and the
extent of injury in its immediate presence point to the sporozo6n as the primary agent.
There are a few places in the gill tissue where the short bacillus is abundant, but, as
would be expected of a saprophyte, in very degenerate tissue only. Such seems to be
its relation to all the tissues.

There are also tissues in which nothing but the long bacillus can be recognized as
the agent of primary degeneration. While never abundant, it may be observed more
frequently than the short bacillus in fresh smears of infected tissue. After about 24 hours
the latter appear in clusters in the muscle fibers occupying excavations of regular ovoid
contour. The long type occurs less frequently in tissues that are completely atrophied
than in those which just begin to show signs of decadence. Fresh muscle, in the latter
condition, may have the long bacilli more or less abundantly distributed under the
sarcolemma, but never in compact groups, a condition which is characteristic of the
short form. In sections, the long type has been encountered, one or two at a time,
in muscle fibers at or near the region of advancing degeneration, and occupying irregular
transverse clefts in the scarcoplasm (similar to those in fig. 4, pl. xx). But these cavities
seem to be much too large to be considered the excavations of so few of these minute
organisms. ‘Their toxins may precede them and the transverse cleavage of the muscle
fiber may be due to subsequent mechanical forces. On the other hand, the bacillus is
quite as likely to creep into the crevices in the sarcoplasm as are the blood tissues
(p. 198). Its presence is therefore not necessarily evidence that it is the cause of the
crevices. In one stained smear, some of the muscle fibers of which are completely
hypertrophied, the long bacillus is very abundant, especially in the connective tissue.
There is no evidence of the admixture of fluid from purulent tissue such as is frequently

SPOROZOON PARASITES OF FISHES. 201

common when the short bacillus occurs abundantly; nor are there any of the short
bacilli. The normal striated fibers possess few if any of the germs and they seem to be
numerous in proportion as the sarcoplasm is degenerate. These are not the conditions
we would expect of a virulent parasite unless its primary attack is through the agency
of a toxin. There is a second factor to be considered, however, inasmuch as numerous
myxoplasms and autospores of M. musculi occur in some of the less decomposed por-
tions of the same tissue. With the evidence at hand bearing upon the virulence of the
two bacilli, the most natural conclusion is that the short bacillus is a saprophyte, that
the long bacillus is either a facultative parasite upon the post tissues, which has been
reduced in vigor by the Sporozoa already established therein, or perhaps a true parasite,
in which case there are frequent double infections, the long bacillus and Myxosporidia
together preparing the way for the saprophytic short bacillus.

The long and short bacilli are easily distinguishable by their size, shape, and habits.
The long bacillus is 0.7 in diameter and usually at least 2.54 long, but it may be 22
long, without any noticeable increase in diameter. (Fig. 11, pl. xx.) They have tapered
ends, especially those which have but recently divided. Sometimes the long type
divides, forming short rods, but they are then in chains. ‘They never occur in clusters
asin figure 10, platexx. The short type is never coiled, never so long, and always thicker
than the long bacillus. They are both encountered in smears which include the fluids
of completely broken-down tissues, but the short form is always abundant in such
fluids, while the former is rare. One is frequently clustered and in regular pockets, the
other isolated or scattered and, if in cavities at all, they are irregular crevices.

SPOROZOA ASSOCIATED WITH ATROPHIED TISSUES.

From the evidence in the foregoing pages and borne out by that which is to follow,
it is certain that a sporozoén causes the primary degeneration of muscle, gill, and pos-
sibly integumentary tissues, resulting in pathological conditions which are quite as
characteristic as when the bacillus is the primary parasite. In one tissue which was
sectioned (fig. 18, pl. xx1) the degeneration of the muscle fibers is identical to that where
bacteria alone have been observed (p. 200). The atrophied fibers, which contain numerous
scattered Sporozoa (p. 198, 203), occur in groups of two or three here and there through-
out the fragment of flesh. Frequently, in both sections and smears, degenerate muscle
fibers occur in which there are cells similar to the above but with neither nucleus nor
cytoplasm stained; also large amoeboid masses of granular cytoplasm without any visible
nucleus (fig. 13, pl. xx). Usually such foreign cells occur in tissues when either myxo-
spores or multiplicative stages are more or less abundant.

In one or the other of the above stages the sporozoén has been positively identified
with the disease in 18 of the 85 fish which have been examined. On the other hand,
many degenerating fibers have been encountered both in smears and sections in which
neither Protozoa nor bacteria could be found. In such cases there is about equal
lack of evidence that either of the above are the causal agents of such disintegration.
While it is probable that the majority of the sores are caused by the inoculation of a
wound by a germ, there is less evidence of a primary attack upon the tissues by the
bacteria, except through a widespread toxin, than by Sporozoa. In this connection there
is probably a significant difference in the external appearance of diseased tissues which
are primarily due to the sporozoén attack and those which are caused by bacteria.

202 BULLETIN OF THE BUREAU OF FISHERIES.

Certain fish in which the diseased parts were conspicuously congested (ventral part of
the head, around the anus, and about the eyes) were almost invariably found to con-
tain a large number of myxospores. When we consider the unknown stages of the
Sporozoa which, according to the cyclic habit of these organisms, advance from stage
to stage in a given culture at nearly the same rate, there is reason to attribute to them
more destruction than our observations warrant. Our present lack of knowledge is no
doubt due in part to the inadequate stains that have been employed and in part to the
confusion of tissue cells with certain stages of the myxosporidian cycle. (See also p. 205.)

STAGES OF MYXOBOLUS MUSCULI.

Mention is made in the literature of but one other case of mxyosporidian disease
of the integument and flesh which is closely allied to that of the Fundulus, namely M.
lintont of Cyprinodon variegatus (Linton, 1889-1891). With this one exception, similar
diseases in other American and European salt-water minnows, as far as we can learn,
have not been described. The M. lintoni of the Cyprinodon was at first supposed to be
identical to the M. musculi of Fundulus. But very recently a tumor of the variegated
minnow was encountered. (See p. 206.) Both the spore and the tumor are markedly
different from the common condition of Fundulus.

The myxoplasm of M. muscul: produces a great many pansporoblasts, each with a
single spore. ‘There is a large vacuole in most of the spores which is the characteristic
iodinophilous vacuole of the genus Myxobolus, to which the parasite undoubtedly belongs.

Of the life history we have the spore, pansporoblast, possibly the myxoplasm,
schizont, and multiplicative or autospore. In but 3 of the 18 fish which harbor Sporozoa
have we stages (figs. 20, 21, 26, 27, pl. xx) that can be unmistakably connected with the
spore. By association in the same tissue or by the appearance and staining reaction
we have probably identified the myxoplasms and autospores.

According to Auerbach’s (1910) description of M. bergense, the spore terminates
the life cycle in a given host and starts a new cycle in a new host. We can but assume
that the trophoplasm of /. musculi likewise arises in some way from a primary myxo-
spore. The trophoplasm (fig. 12, pl. xx) is difficult to stain, and therefore its sporozoén
properties are not always certain. (See also Chloromyxum properties, p. 205.) Spherical
or oval spaces in the diseased myoplasm and in the epidermal cells (possibly identical,
fig. 36, pl. Xx1,and p. 197) are very abundant. These are probably multiplicative tropho-
plasms, unless we have confused them with fat or other nonstaining substances. Some-
times these bodies have nuclei (fig. 12) which, though usually faint, may stain deeply.
It is not impossible that some of these small trophoplasms may be those of the Chloro-
myxum. When large, the trophoplasms have a granular structure (fig. 13, pl. xx) and are
doubtless preparing to undergo schizogony. We have encountered but five or six
such schizonts. In one series of sections they are associated in diseased muscle fibers
with cysts containing many spores. (Fig. 14, pl. xx.) The amoeboid form of the mature
schizont is characteristic and distinguishes it from the smaller forms. The schizont
in figure 13 is 334 wide by 74 in length. Some of the cysts are about this size, but
figure 14, which is 19” wide and 244 long, is a section through the small end of a cyst
of only moderate size. The cysts are found both within and between the muscle fibers.
They contain several hundred spores, the nucleus of which, like that of the trophoplasm,
has at times little affinity for the stains we have employed. The spores sometimes

SPOROZOON PARASITES OF FISHES. 203

appear to be spherical in form (fig. 14, pl. xx) and vary somewhat in size. They have
a small faintly-staining nucleus and hyalin nonstaining cytoplasm. Isolated spores
and masses of spores recently discharged from the cysts also occur in the smear prepa-
rations associated with the intrafibrillar masses of material that appear to be equivalent
to schizonts. These spores also occur in small numbers in the diseased gill where
myxospores and sporoblasts are to be found in very great numbers.

The occurrence of a multiplicative process of reproduction amongst the Myxo-
sporidia in the manner here described is not uncommon. We have authentic cases
in gall parasites of the flounder, and they have been described in M. pjetjjert (Keys-
selitz, 1908) and in Henneguya gigantea (Nemeczek, 1911). While there is no question
but that there are multiplicative spores, our evidence that the spore here described
is such is, as with the trophoplasm, far from conclusive. Judging from the meager
evidence at our disposal, there is about equal reason for considering it a young sporo-
blast or a young trophoplast. It is more harmonious to regard these amceboid spores
as the progenitors of both multiplicative and propagative trophoplasts and the oval
spores, which are described below (p. 204) as sporoblasts, more especially since they
apparently arise by free cell formation and in smaller numbers.

The propagative and sporoblast stages have been encountered frequently in both
sections and smear preparations. One series of sections of diseased integument and
muscle (referred to on p. 198), which were cut approximately at nght angles to the body
surface, contains numerous large cells (fig. 18, pl. xx1) with small well-stained (with
hematein) nuclei. Some of the muscle fibers are cut obliquely. They lie in the midst of
healthy tissue and under integument which is apparently healthy. These myxoplasms
are of oblong or spherical form with more or less even surface. The cytoplasm is
tolerably homogeneous and does not retain the stains. The karyoplasm is also unstained,
but the chromatin is somewhat conspicuous. ‘These cells occur abundantly throughout
60 or more sections, in 6 to 8 adjacent fibers, also in other distant fibers. Upwards
of a hundred perfectly normal fibers around them have not a single foreign cell. Such
cells are always intracellular. The sarcoplasm is considerably modified. The fibrillar
structure is lost and the appearance is almost homogeneous.

A cell similar in every respect to the myxoplasm just referred to, occurs abundantly
in smears of muscle tissues. It stains less readily than do leucocytes and has smaller
nuclei with less conspicuous chromatin. Myxospores and pansporoblasts have been
found in their midst, in fact are to be found on slides where this type of cell occurs and
not elsewhere. In this connection it is interesting, and perhaps additional evidence of
relationship, that the same sore from which the sections containing these myxoplasms
(fig. 18, pl. xx) were made also supplied a smear preparation containing numerous myxo-
spores and pansporoblasts represented in figure 26, plate xx1. The sporoblast resembles
the myxoplasm of smear preparations in shape, clear, nonstaining cytoplasm, size, and
feebly staining (with methylene blue) nucleus. It is for the above reasons that this
type is assumed to belong in the propagative cycle.

There is a wide range of conditions to be seen in the nuclei of these myxoplasms,
as well as some variations in size. Some densely-staining, cigar-shaped bodies (fig. 17,
pl. xxr) almost devoid of protoplasm are embedded in the sarcoplasma, and others are
closely applied to the myxoplasms (fig. 18, pl. Xx1, near right-hand upper corner). The

204 BULLETIN OF THE BUREAU OF FISHERIES.

conditions suggest conjugation, but the stages are too few to indicate a succession of
events. One myxoplasm contains two oblong spores. Elsewhere, replacing a degen-
erated muscle fiber, are numerous small cysts (12 in diameter) with eccentric nuclei,
which contain from four to ten or a dozen clearly defined oblong spores (fig. 16, pl. xx).
These spores are found abundantly in other fish. (Fig. 19, pl. xxi.) They appear to
arise by free cell formation. They are characterized by a transparency and a failure to
stain that recall both the trophic stages and the sporoblasts. The nucleus, however,
does stain faintly. It is quite large when the spores are set free. The latter measure
4 by 2.54 and sometimes assume a spherical or amceboid form. Between this condi-
tion and the mature sporoblast we lack recognizable connecting stages. They are
not far removed, however, from the latter, which are spherical cells with very large
nuclei. (Fig. 35, pl. xx.) These occur in the gill above mentioned and have there been
definitely connected with the myxospore. The pansporoblast has been encountered,
along with spores and sporoblasts, in fresh smears of muscle. These are apparently
identically homologous to those described for M. pjeifjeri in the gills of the barbel
(Keysselitz, 1908). If so, the sporogenesis there related would appropriately apply to
M. musculi. Many stages in the genesis of the spore are represented in one of our
smear preparations. These have propagative stages (Keysselitz, 1908) as follows:
First the sporoblast with large nucleus (fig. 35, pl. xx1) and two-parted pansporoblast
(sporocyst) (fig. 22, 23, pl. xx1), which, according to Keysselitz, arises after a process of
autogamous conjugation. The sporocyst apparently sets free the sporocytes before
sporogenesis has proceeded far (fig. 21, pl. xx1). Giemsa stain does not reveal all the
nuclei concerned in sporogenesis. Valve cells are formed (fig. 24, pl. xxr) before the
polar capsules appear as large spherical bodies (fig. 25, pl. xx1). Later the myxospore
becomes elongated and tapered (fig. 20, 26, pl. xx1)._ Two preparations have multitudes
of immature spores. They are all free from the sporocyst protoplasm and have thick
valves. It is therefore rather perplexing to explain figures 20 and 26, Perhaps the
spore is about to be discharged in figure 20. Considerable variation in this respect
occurs amongst some of the gall Myxosporidia.

There are myxospores in 12 of the 85 fish examined. In but 3 of these do they
occur in great numbers. With two exceptions (in diseased gills), the myxospores are
not assembled in a manner that would suggest their origin from cysts or masses of
pansporoblasts, as is common in other species of Myxosporidia. The two cases referred
to may not be interpreted as evidence of this condition, but rather that the pansporo-
blasts, where very numerous, have been packed close together. There are at least a
thousand well-stained spores in the preserved tissues. Not one occurs in the 10 tissues
of which sections have been made. But those same tissues which contain spores have
supplied all the propagative myxoplasms.

The myxospores are very small (fig. 28, 29, pl. xx1). They average 14.3 in length
and 6.7” in width. In one fixed individual the plane at right angles to that passing
through the polar capsules is presented. It measures 6.7% in thickness, from which
we conclude that they are approximately circular in section. But another fresh spore
was flattened in a plane perpendicular to that of the polar capsules and sutures to about
two-thirds its width (fig. 30, pl. xx1). The polar capsules of myxospores average 6.54 in
length and 2 in thickness. When extruded the filament is three to four times the length
of the spore (fig. 29, pl. xx1). Coiled within the polar capsule, the filament makes from

SPOROZOON PARASITES OF FISHES. 205

10 to 14 turns (fig. 26, 28, pl. xx1). In young spores the valves are quite thick and may
be seen at the edges as a pale border to the spore, but in mature spores they are thin
and almost invisible. Young spores are shorter (12”) and wider (7.5) than the
mature spores and the polar capsules are not so long (6). They lengthen out as they
approach maturity. When young, the nuclei stain with great difficulty, if at all. The
sporoplasm occupies all the space at the large end of the spore. A large vacuole is nearly
always visible in the sporoplasm. There are also dense areas and from 1 to 1o nuclei.
(Fig. 20, 28, pl. xx1.) The nuclei are unstained in figure 29, plate xx1. There are some-
times seven greenish-blue nuclei (fig. 28, pl. xxr) and three rather irregular dark-blue
bodies between the polar capsules. It is not possible to be sure that this (10) is the
maximum number as some of these are ill defined. A number of spores have their
nuclei attached near the large end of each polar capsule, thus identifying them as the
“polar capsule” nuclei. Probably the remaining four belong to the sporoplasm. It is
not possible to recognize the ‘‘wall nuclei’ at this stage, and the “resting nuclei”’ of
the pansporoblast are doubtless lost.

CHLOROMYXUM FUNDULI.

The Chloromyxa which have been observed in the muscle of other fish (p. 208) are not
identical to that found in Fundulus. C. quadratum, which resembles the latter, has
myxospores measuring 6 in diameter by 5 along the polar axis, while the Chloro-
myxum of the Fundulus measures 7.54 in diameter and 6 along the polar axis and
differs in shape. The spore of C. quadratum, when seen in line with the polar axis, has
the sides deeply concave, and in the other plane it is more pointed. The polar capsules
are also much shorter. They also differ in the relation of the spore to the pansporoblast
and in the pathological effects. (For description of spore of C. funduli see p. 208.) No
reference to a myxospore of this character has been found by the writer. The name
C. funduli has therefore been applied to this species.

The myxospores of C. junduli have been encountered in but one fish. If the myxo-
plasm occurs in other preparations, it has not been possible to identify it, although many
suspected myxoplasms exist. It is not very probable that they are at all uncommon.
They do not take up a particle of such stains as we have employed. The single slide
containing this species is a smear preparation made from the diseased flesh of a fish
which died in jar no. 1 of the experiment reported on page 196. It is stained with
Giemsa stain.

The muscle of this fish is in an advanced stage of decomposition. When the fresh
slides were examined no myxospores were noticed, being difficult to see without a stain,
but the sporoblasts were observed without recognizing their importance.

Bacteria are present on the slide but lacking in the muscle fibers. The decadence
of the muscle must in this case be ascribed to the Chloromyxum, which is abundant in
the hypertrophied muscle. The muscle is full of cavities containing unstained myxo-
plasms and sporoblasts which are identical in appearance to those of many other prepara-
tions of diseased Fundulus. While this case introduces the possibility that many of the
Fundulus cancers may be caused by Chloromyxum funduli, it gives very substantial sup-
port to the agency of Sporozoa as the cause of these diseases. Since Chloromyxum and
Myxobolus are not uncommon in muscle tissue, double infections are to be expected.
But having failed to encounter myxospores of the Chloromyxum in over 100 stained

19371°—vol 33—15——14 ,

206 BULLETIN OF THE BUREAU OF FISHERIES.

preparations that have been examined, we are inclined to consider the Myxobolus more
abundant and therefore the more common causal agent.

The myxospore of C. junduli is about 7.54 in diameter, with a polar axis somewhat
shorter (6). At right angles to the polar axis, it is circular. There are four polar
capsules, which taper to the apex of the spore, curving so as to conform to the constric-
tion of the spore, which provides it with a blunt pointed apex (fig. 31, 34, pl. xx1). There
are four conspicuous nuclei (black), one near the base of each polar capsule. The sporo-
plasm is stained a pale blue by the Giemsa. The polar capsules do not stain (fig. 31, 34,
pl.xx1). There are occasional myxospores of considerable size to be seen inside the sporo-
blasts when the latter do not take up a particle of stain. Such clear hyalin amceboid
pansporoblasts are numerous throughout the sarcoplasm. They vary in size from a
diameter of about 2 to four or five times the diameter of the spore.

PROTOZOA RELATED TO THOSE HERE DESCRIBED.

Numerous Myxosporidia parasitic upon either integument, gill epithelium, con-
nective tissue, or muscle of fish have been described by other authors. -About most of
them we have very meager information. M. lintona (Linton, 1889; Gurley, 1893) of
Cyprinodon variegatus (short minnow), as already stated (p. 202), more closely resembles
the parasites of Fundulus than any other species of which we know. ‘The difference
at first seemed to be slight and to be easily accounted for by a difference in the age
of the spores. But when a case of the Cyprinodon tumor was finally obtained and
examined, the indentity of the parasites in the two hosts, as well as the nature of
the lesions, was found to be different. The “irregular fungoid elevations,’ described
and figured by Linton and observed again by the writer, are of the nature of cysts con-
taining spores, located in the integument, whereas the elevated scales in Fundulus are
due to an infection of the epidermis by bacilli and a subdermal atrophy of the muscle.
No tumor or spore-filled cyst has ever been encountered. The Cyprinodon tumor which
we examined developed in a comparatively short time, probably less than a week,
though the period can not be accurately stated. It caused the death of the fish the day
following that on which it was first noticed. After the death of the host, the tumor was
8 mm. wide by’ 10 mm. long, and caused a conspicuous elevation from the back of the
fish anterior to the dorsal fin, about 2 to2’% mm. thick. It was of a yellowish-pink color
when seen through the slightly pigmented integument. The scales were practically
undisturbed and the integument was completely intact, in this respect differing remark-
ably from the Fundulus sores. Beneath the tumor, the flesh contained intrafibrillar
myxoplasms and sporoblasts with occasional spores, while the tumor itself was almost
wholly a mass of myxospores, the latter numbering millions. We have already described
(p. 197) a totally different condition in the Fundulus, resulting from the M. musculi.

There is such a difference in the appearance (Linton, 1889, fig. 3) of the spores that
they are readily distinguished. One can not be certain, however, that such differences
are not due to the comparison of different stages in the development of spores of the same
species. We have shown that the spores of M. musculi grow longer as they mature and
the spore wall becomes thinner (p. 205). This fact would explain in part the discrepancy
in the dimensions of the spores from the Fundulus and Cyprinodon. But, since the
spore of M. /intoni measures 13.9 in length, 11 in width, and 8 in thickness (at right

SPOROZOON PARASITES OF FISHES. 207

angles to the planes of the two polar capsules), and the mature spore of M. musculi
measures 14.3 in length by 6.74 thick, and from 4 to 6.7#in width (see p. 204), in indi-
viduals of apparently the same stage of development, it still seems that a sufficient
discrepancy in size exists to supplement the marked differences in the pathological con-
ditions. It may yet prove that the latter are due to the influence of different hosts, inas-
much as we have one case of a Fundulus with a typical M. musculi lesion, but having
spores indistinguishable from those of M. /intoni in either size or appearance.

M. lintoni presents another contrast to the conditions in /undulus. In the former,
calcareous bodies were observed amongst the spores by Linton (1889) and the writer,
whereas nothing of the kind has ever been encountered in the hundreds of Fundulus
tissues which we have examined.

Although the name M. /intoni was for a time retained for the /undulus parasite, the
present state of our knowledge will not permit of this assumption. The species “musculi”’
has been adopted because of the interesting and characteristic attack which the tropho-
plasm makes upon muscle fibers.

The spore of Myxobolus oviformis (Thelohan) resembles M. musculi very much in
appearance, but is less tapered and shorter (Thelohan, 1894).

The following, for one reason or another, are also of interest in their bearing upon
M. musculi. A “Myxosporidian”’ of unknown genus and species was found by Linton
(1899) in the connective tissue of the entire body of Notropis megalops Rafinesque
(albeolus Jordan), the shiner. ‘The epidermis is marked by dark purplish blotches. The
scales are absent in most cases. A “Myxosporidian”’ of unknown genus and species was
observed by Lieberktihn (1854) in the connective tissue of Gastereosteus aculeatus (stickle-
back). The skin is said to have contained cysts. The conditions seem to be unlike those
in Fundulus. Cyprinus leuciscus (Miller, 1841) has been observed with tumors in the
integument caused by a species of Myxobolus. M. oblongatus Gurley produces cysts
under the scaleless skin of the head region in Catostomus tuberculatus Le Sueur (Gurley,
1891, 1893, p. 234). M. transovalis Gurley (1893) of Phoxinus (Clinostomus) junduloides
Girard, occurs under the scales and external to the epidermis. ‘It forms a thin dis-
coidal mass situated in the center of the concave undersurface of the scale.’’ That it
is not identical with M. musculi is certain from the dimensions of the myxospore (length
6u, breadth 8), the diameter of which, at right angles to the polar axis, is greater than
through the polar axis. We have very scanty information concerning the M. strongylurus
(Gurley, 1893, p. 247), which is found encysted in the skin of the head of Synodontis
schal; of M. momurus (Gurley, 1891, p. 416), known from cysts in the subcutaneous inter-
muscular tissue of A phredoderus sayanus Gilliams; of Henneguya niisslini Schuberg und
Schréder (Leger, 1906), which is found in the connective tissue of the dorsal fin of the
trout; of M. gigas Auerbach (1907), which thickens the integument at the ventral angle
of the gill in Abramis brama Linnzus (bream); and of a Myxobolus of unknown species
described by Borne in 1886 (Gurley, 1893, p. 244), which causes great tumors over the
surface of Leuciscus rutilis.

In Coregonus fera there occurs a common disease of the integument caused by a
species (M. zschokket, Gurley, 1893; Zschokkei, 1884), the myxoplasm of which is not
known. The cysts lie in the subcutaneous connective tissue and between the muscles.
It causes irregular thick patches on the skin, from which the scales drop.

208 BULLETIN OF THE BUREAU OF FISHERIES.
PREVALENCE OF MYXOSPORIDIAN INFECTION IN FISH.

The infection of muscle tissue by Myxosporidia is quite common in fish. A parasite
belonging to the genus Chloromyxum occurs in the flesh of the young herring and young
alewife (Linton, 1891). Both the pansporoblast and spores of a Chloromyxum have been
found abundantly by the writer inside the fibers, and the spores also assembled else-
where in large cysts. A fuller account of this species will be published later. The
muscle cells of Callionymus lyra are also subject to an intracellular parasite (Glugea
destruens Thelohan, 1891; Henneguy et Thelohan, 1892; Gurley, 1893), the myxoplasm
of which has not been observed. It causes the muscle fibers to undergo degeneration.
Chloromyxum quadratum (Thelohan, 1894) also occurs in the muscles of this fish. It is
also reported in the flesh of Corts julis, Syngnathus acus, Trachurus trachurus (Minchin,
1903) and Nerophis equoris. In Cottus scorpio the muscle tissue is attacked by Pletsto-
phora typicalis (Thelohan, 1890, and 1891; Gurley, 1893). Both pansporoblast and spores
have been found, but they are intercellular in position. The muscle fibers are displaced
but do not degenerate. Leptotheca perlata (Gurley) occurs in the muscles of Acerina
cernua Linneus. Of these species there are none that closely resemble M. muscult.
Numerous cases of Myxoboli are known to inhabit gill tissues. Auerbach (1911) lists
22 species of Myxobolus which have been described in the gills of fish. But we have
encountered nothing that might be considered identical to M. muscult.

The disease of Fundulus is remarkably like that which has so frequently caused
epidemics amongst the barbel (Barbus barbus Linnzeus) of European rivers. The latter
is caused by M. pfeiffert Thelohan (Raillet, 1890; Ludwig, 1888; Thelohan, 1894). It
produces both tumors and ulcers and occurs encysted and free in muscle, liver, kidney,
spleen, and connective tissue. The tumor when formed does not at all times break
through, either into the body cavity or to the outside. It is not an integumentary
parasite at the beginning as those of Fundulus seem to be. The tumor commonly occurs
amongst the connective tissue and the muscles of the body wall. The parasite may be
encysted in a thin restraining membrane produced by the host. Numerous individuals
of about the same age tend to gather in groups and become isolated in tube-like cysts.
The muscle fiber is invaded and undergoes a ‘‘vitreous alteration’? (Thelohan, 1893)
leaving ‘“‘ yellow granulations as degeneration products’ (Keysselitz, 1908). Thelohan’s
figure 5, plate vi (Thelohan, 1894), representing a muscle fiber containing myxoplasms
in transverse crevices recalls, very vividly the appearances we have encountered in the
degenerate muscle of Fundulus (fig. 4, pl. xx). The tumors may soften and become a
“stinking abscess containing spores” (Ludwig). MW. pfetffert passes through distinct
cycles of development which is no doubt the case in M. musculi. In April it is in a vege-
tative stage in which the multiplicative reproduction prevails; later propagative repro-
duction is encountered and myxospores are developed. The rate of advance of the dis-
ease depends upon the temperature (Keysselitz).

Both Keysselitz and Thelohan describe bacteria in tissues of diseased barbel.
Keysselitz says bacteria contribute liberally to the formation of the tumors. These
bacilli are found only in the tissues infected by Myxosporidia. They prevent the
growth of connective tissue and bring about degeneration (gangrene) of the tissue.
These bacilli are ‘‘as long as the spore”’ (Pfeiffer, 1890) (64, Thelohan) and stain easily
with methylene blue and gentian violet. (This is also true of the bacilli of Fundulus

SPOROZOON PARASITES OF FISHES. 209

diseases.) Pfeiffer mentions threads attached to these bacilli. A coccus is also occa-
sionally found. The presence of bacteria is therefore not necessarily an indication that
they are primary as causal agents of disease since M. pfeifjeri is known to be the cause
of the barbel disease.

GENERAL CONCLUSIONS.

I. The sores of Fundulus are usually caused primarily by lesions. These may
occasionally be due to parasites such as leeches, distomes, and copepods, but usually
to rough handling and carnivorous enemies.

II. At least four kinds of germs invade these lesions and bring about hypertrophy
of the tissue elements and decomposition, namely, two species of bacteria and two species
of Myxosporidia.

III. There is doubt as to the virulence of the bacteria. One species at least is
saprophytic. There is no doubt as to the virulence of the Myxosporidia when present.

IV. Cleanliness, careful feeding, and aeration bring about recovery in practically
allinjured fish. It can not be claimed that fish which are known to have Myxosporidia
are curable.

V. The trophoplasm of both species of Myxosporidia attacks the muscle fibers,
that of the M. musculi also attacks the gill connective tissue.

VI. Blood elements, especially nuclei, give rise to abundant artifacts which are
closely associated with the parasite involved.

VII. Sporogenesis of the Myxobolus occur infrequently in the muscle and gill tissues.

VIII. Multiplicative spores are probably formed in M. musculi in addition to pri-
mary sporocytes.

IX. The myxoplasm of both C. junduli and M. musculi are stained with difficulty
and are therefore not easily found.

BIBLIOGRAPHY.
AUERBACH, MAX.
1907. Weitere Mitteilungen iiber Myxobolus eglefini Auerbach. Zoélogischer Anzeiger, bd.
XXXI, p. 386-391. Leipsig.
toro. Biologische und morphologische Bemerkungen iiber Myxosporidien. Zodlogischer Anzeiger,
bd. xxxv, p. 57-64. Leipsig.
tor. Unsere heutigen Kenntnisse iiber die geographische Verbreitung der Myxosporidien.
Zodlogischer Jahrbucher, Abtheilung fiir Systematik, bd. 30, p. 471-494.
BorRNE, MAX VON DEM.
1886. Handbuch der Fischzucht und Fischerei, p. 211, fig. 215.
Gurey, R. R.
18g. On the classification of the Myxosporidia, a group of protozoan parasites infesting fishes.
Bulletin United States Fish Commission, vol. x1, 1891, p. 407-420. Review in Central-
blatt fiir Bakteriologie und Parasitenkunde, bd. 15, p. 86-88.
1893. The Myxosporidia or psorosperms of fishes, and the epidemics produced by them. Report
U. S. Commission of Fish and Fisheries, 1892, p. 65-304, pl. 1-47.
HENNEGUY, F., ET THELOHAN, P.
1892. Sur un Sporozoaire parasite des muscles de l’ecrevisse. Comptes rendus hebdomadaire
Société de Bioiogie Paris, t. 4, p. 748-749.
KEyYSSELITz, G.
1908. Die Entwicklung von Myxobolus pfeifferi Thelohan, Theil 1 und 2, Archiv fiir Protisten-
kunde, bd. 11, p. 252, 276-308. Jena.
Lecer, L.
1906. Sur une nouvelle myxosporidian de la tauche commune et de la truite indigene. Comptes
rendus de 1’Academie des Sciences, Paris, t. 142, p. 655.
LIEBERKUHN, N.
1854. Uber die Psorospermien. Miiller’s Archiv, p. 9-10, 22, 24, 354, taf. 2, fig. 28; taf. 14, fig. 9-12.
Linton, Epwin.
1889-1891. On certain wart-like excrescences occurring on the short minnow, Cyprinodon variegatus,
due to psorosperms. Bulletin United States Fish Commission, vol. rx, 1889, p. 99-102,
pl. xxxv; ibid. p. 359-361, pl. cxx, fig. 1-3. Review in Centralblatt fiir Bakteriologie
und Parasitenkunde, 1892, bd. 11, p. 475.
Lupwic, H.
1888. Uber die Myxosporidien krankheit der Barben in der Mosel. Jahrbuch des rheinischen
Fischerei Vereins, Bonn, p. 27-36.
Mincuin, E. A.
1903. A Treatise on Zodlogy. Edited by E. Ray Lankester, pt. 1, fase. 2, p. 150-361.
MULLER, J.
1841. Ueber Psorospermien. Miiller’s Archiv fiir Anatomie und Physiologie, p. 477-496, pl. 16.
Abstract in Microscopical Journal, London, 1841-2, p. 123-124.
PFEIFFER, L.
1890. Die Protozoen als Krankheitserreger, Jena, 1 ed.
PLEHN, M.
r905. Uber die Drehkrankheit der Salmoniden (Lentospora cerebralis) Archiv fiir Protistenkunde,
bd. 5, p. 145-166, taf. 1, fig. 7. Jena.
Ramet, M. A.
1890. La maladie des Barbeaux de la Marne. Bulletin Société Centrale d’Aquiculture, Paris,

t. 2, p. 117-120.
210

SPOROZOON PARASITES OF FISHES. 211

THELOHAN, P.

1890. Contribution a 1’étude des Myxosporidies. Annales de Micrographie specialement consacrees
a la bacteriologie, aux protophytes et aux protozoaires. P. 1, t. 2, p. 193-213. Paris.

1890. Recherches sur le developpement des spores chez les Myxosporidies. Comptes rendus heb-
domadaire de la Société de Biologie. Paris, t. 2, p. 602-604. Abstract in Journal Royal
Microscopical Society, 1890, pt. 2, p. 194-195.

1891. Sur deux sporozoaires nouveaux parasites des muscles des poissons. Ibid., t. 62, p. 168-172.

1893. Alterations du tissu musclaire dues a la présence de Myxosporidies et de microbes chez le
barbeau. Ibid., t. 5, p. 267-270. Abstract in Centralblatt fiir Bakteriologie und Para-
sitenkunde, bd. 14, p. 532.

1894. Recherches sur les Myxosporidies. Bulletin Scientifique de la France et de la Belgique,
t. 26, p. 100-394, pl. 7-9. Paris.

WIERZEJSKI, A.

1898. Uber Myxosporidien des Karpfens. Anzieger der Akademie der Wissenchaft in Krakau,
Marz. Résumé in Bulletin International de 1’Academie des Sciences de Cracovie, Comptes
rendus des seances, 1898, p. 129-145. Cracovie.

ZSCHOKKE, F.
1884. Psorosperms de Coregonus fera, Archives de Biologie, t. 5, p. 234-235, pl. 10.

EXPLANATION OF PLATES.

With the exception of figures 20, 27, 30, and 35, the drawings were made with the
aid of a camera lucida. For figures 3, 5 to 11, 15 to 17, 19, 21 to 26, 28, 29, 31 to 34, and
36 a no. 12 Bausch & Lomb compensating ocular and one-twelfth inch oil immersion
objective were used. For figures 1, 2, 12, 14, and 18 a Bausch & Lomb 1-inch occular
was employed with the same objective. The 1-inch occular and a one-fifth inch objec-
tive were used in figures 4 and 13. All figures have been reduced to two-thirds the size
of the camera images. The tube length was 160 mm. and the camera arm 90 mm.

The figures are numbered approximately in the order of development. Figures 2,
3, 5, 6, 7, 8, 9, and ro are made from the same slide, and figures 13, 14, 16, 17, 18, 26,

and 28 are from the same fish.
PLATE XX.

Fic. 1. A bit of infected muscle from a smear of a sore on the side of a small Fundulus heterclitus in
the first stage of disintegration. Fixed in corrosive sublimate and acetic acid and stained with Mayer’s
hematein. The pale bands of the fiber are beginning to become granular at one end. Fibrin threads
have been spread over it in making the smear preparation. ( 860.)

Fic. 2. A bit of degenerating muscle fiber. Numerous artifacts and a degenerate erythrocyte
nucleus occur in the sarcoplasm. The granular strie are degenerated sarcolymph. Note the sarco- .
plasm is also becoming granular. (> 860.)

Fic. 3. From a smear of a bit of degenerating muscle in a sore on the side of Fundulus majalis.
The integument more or less disintegrated, scales entirely absent. Fixed in absolute alcohol, ether,
and formaldehyde. Stained in methylene blue, orange G, and eosin. Sarcous elements have lost
their sharp rectangular form and are becoming granular. A characteristic muscle artifact is distributed
between the sarcostyles and some are just beginning to become amoeboid in form. ( 2000.)

Fic. 4. A characteristic appearance of a degenerating muscle fiber which may or may not be a later
stage than those represented in figures 2 and 3. Neither bacteria nor Myxosporidia are necessarily
present in these spaces. Both have been encountered there. (> 4o0.)

Fic. 5. A fragment of degenerating muscle upon and into which erthrocytes and leucocytes have
entered. The cytoplasm of the latter is disintegrated and the nuclei are in an advanced stage of
degeneration. ( 2000.)

Fic. 6. Atypical mass of degenerate nuclei containing unstained bodies which are probably zodgloea
containing the short bacillus. There are cords of this material in which the bacilli are faintly visible.
Such white areas are not merely transparent spaces but thick masses with stainable protoplasm above
or below. (2000.)

Fic. 7. Artifacts from decomposing muscle fibers. In fresh muscle these are common after Io to
12 hours, appearing first between the sarcostyles. Older stages assume a more compact form. (See
figures 3 and 2.) The stain is a homogeneous pale blue. Maximum length 8.94. (2000.)

Fic. 8. The short bacillus. An isolated group near which are located cells containing white oval-
shaped bodies like those in figure 6. Note the variation in size and shape. That one near the “ x”
sign measures 1.5 by 7.4; that near the “+’’ sign measures 1.84 by 1.12. (22000.) (See also fig. Io.)

Fic. 9. Short bacillus older than figure 8. Nearly the maximum size. Note the taper toward
one end and the stainable granules. The latter are probably artifacts. Left-hand upper one measures
5.2 by 1.44. (X2000.)

Fic. to. A cluster of long bacilli which have caused the complete breakdown of a tissue cell and
rest in situ. (X 2000.)

Fic. 11. Several of the long type of bacilli which are located just under the sarcolemma of a muscle
fiber that shows the first signs of degeneration. The small individual in the middle below has dimen-
sions as follows: Length, 4.8; thickness, 0.7. (2 2000.)

212

SPOROZOON PARASITES OF FISHES. Aig

Fic. 12. A section cut diagonally through a muscle fiber. This fiber is adjacent to the dermis.
On the inner side the sarcoplasm is hypertrophied, on the outer side it retains the fibrillation. The
oval bodies are interpreted as trophoplasms of the M. musculi. The large one has several spherical
bodies which take a deep hematein stain, presumably nuclei. ( 800.)

Fic. 13. A muscle fiber in which there are the first evidences of disintegration. It contains two
or more large trophoplasts or schizonts. The appearance of the cytoplasm is like that of other
stages, pale and unstained, there being no sign of the nucleus. There is evidence of a complex system
of pseudopodial extensions of the cytoplasm which is characteristic of the Myxosporidia. Large indi-
vidual 84.74 by 192.54. (X400.)

Fic. 14. Multiplicative spores of M. musculi, presumably derived from a large trophoplasm such
as figure 13. There is no cyst wall. In adjacent sections are fragments of the schizont nuclei mingled
with the spores. The spores stain feebly with eosin and orange G. The nuclei are not stained deeply.
19.3 in diameter. (> 860.)

Fic. 15. A myxoplasm of M. musculi in muscle from a smear preparation fixed with absolute alcohol
and ether and stained with methylene blue. One side overlies a nucleus of the muscle fiber. The
pale bands of the muscle fiber may be seen. The muscle stained deeply and the parasite pale. The
protoplasm is finely granular and there is only a suggestion of a cytoplasmic network. The nucleus is
vaguely stained. 13.44 by 18.64. (X2000.)

Fic. 16. Formation of sporoblasts of M. musculi. This cyst is one of a mass numbering several
hundred which occupy a position where a muscle fiber has been completely destroyed. The to spores
stain very feebly. They lie in slight cavities of the protoplasm. Diameter of cyst 124; length of

spore 4s. (2000.)
PLATE XXI.

Fic. 17. A possible microgamete of MM. musculi from amongst the numerous myxoplasms of muscle
fibers adjacent to that shown in figure 18. The motile shape of several such structures, the small
amount of cytoplasm, and close approximation to some of the large myxoplasms are noteworthy. (See
right-hand upper region of fig. 18.) 6.5 by 2.24. (><2000.)

Fic. 18. A section of a muscle fiber of Fundulus heteroclitus cut crosswise at a slight angle. The
scales in the region of this infection had dropped off, and the area was almost white, being slightly
discolored by blood. The tissue was fixed in corrosive sublimate and acetic acid and stained first in
Mayer’s hematein, then in methylene blue, later in eosin and orange G. One of the structures in the
sarcoplasm, that to the left in the middle, is the nucleus of a muscle fiber. The others are stages in
the propagative cycle of M. musculi, primary and secondary sporoblasts. The large one in the middle,
at the top, is 12.6 in length and 5.9” in width. (X86o.)

Fic. 19. Three young sporoblasts cf M. musculi from the smaller type of cysts represented in
figure 16, plate xx. Note the increase in the size of the nuclei. They are typically free from cyto-
plasmic stain. (See C. funduli, fig. 31.) Lower individual 4” by 2.54. (2000.)

Fic. 20. A fresh sporoblast of M. musculi containing a spore which is almost mature. From a deep
cavity in the flesh back of the head. Interesting in connection with figure 26. (Free-hand drawing,
not to scale.)

Fic. 21. Sporocyte of M. musculi expelled from pansporoblast. It forms the first stage in the
series represented by figures 23, 24, and 25. The nucleus is small and faintly stained, as is the rest of
the cytoplasm. It has no external envelope. Diameter 11.94. (><2000.)

Fic. 22. A pansporoblast of M. musculi (sporocyst) with two daughter cells, the nuclei of which
are undergoing autogamous conjugation. ( 2000.)

Fic. 23. A pansporoblast of M. musculi after the autogamous conjugation and subsequent division
of the nuclei.

Fic. 24. A sporocyst of M. musculi which has been set free from the pansporoblast. Apparently
the sporoplasm remains attached to one myxospore (fig. 20), and the other is almost devoid of external
protoplasm. The two wall cells are clearly visible, but without nuclei. The capsule nuclei are prob-
ably formed but do not stain. One of the r2 nuclei happens to be in a suitable condition to take the stain.
11.94 by 13.4". (X2000.)

Fic. 25. A myxospore of M. musculi with a remnant of protoplasm. Two polar capsules are
beginning to form. ( 2000.)

214 BULLETIN OF THE BUREAU OF FISHERIES.

Fic. 26. A sporocyst of MW. muscul’ from a smear of diseased integument of the mouth and head
in front of the eyes. Elsewhere the sporocysts have less cytoplasm, It is the only one encountered
in this condition, The failure of the nuclei to take the stain is characteristic. The myxospore is
immature, being less slender than older myxospores. The details of the polar capsules are very trans-
parent and stain dark blue, while the spore wall is a very pale blue. The vacuole and sporoplasm are
prominent, but the nuclei of the spore can not be clearly discerned. Sporocyst, 17.8n by 23.84; spore,
r4y.Su by 7.4u; polar capsule, 7.42 by 2.24. There are 13-14 spirals in the filament. Fixation: Absolute
alcohol, ether, corrosive sublimate, acetic acid. Stain: Mayer’s hematein, methylene blue, orange G.,
eosin. (X2000.)

Fig. 27. Asporoblast of M1. musculi from a fresh smear of degenerated muscle taken from a deep
cavity (the same as fig. 20). Easily distinguished from tissue cells by the three nuclei. Protoplasm
contains much coarsely granular matter. (Drawn free-hand, not to scale.)

Ic. 28. Myxospore from the same slide as figure 26. The mature spore, when compared with that
in the pansporoblast, is longer and more pointed at the polar end. The vacuole is probably an iodi-
nophilous structure. The coiled filaments make ri to 12 turns. The polar capsule wall is visible, but
the spore wall can not be clearly seen, The valves and sutures are also indistinguishable. While there
are as many as 12 blue and green bodies presertt, one can not be sure that all of them are nuclei. Seven
oreight bodies are moderately conspicuous. Two lie in the wallof the polar capsules and are doubtless
the capsule nuclei. 14.8 by 6.2”. (X2000.)

Fic. 29. A myxospore of M. muscu/i from large sores on each side of the tail of a Fundulus heteroclitus,
caudal fin entirely gone. Fixed in absolute alcohol and ether, stained with methylene blue. Six
unstained nuclei in the sporoplasm and one large vacuole. Filament discharged. Spore, 7.42 by 16.44.
Polar capsule, 2.24 by 7.4. (X2000.)

Fic. 30. Diagram of the cross section of a fresh myxospore of M. musculi as if seen from the end.
The specimen was lying so as to present the edge of the valves to view, It is obviously flattened. The
polar capsules also appeared to be, but one can not be certain about this. The sutures are straight and
symmetrical. Fixation: Alcohol, ether, formalin; Giemsa stain. (This drawing not made to scale.)

[Figures 31 to 34 are all from the same smear preparation of diseased muscle from a dead fish, being
one of those taken from jar no, x (see pp. 195, 196).]

Fic. 31. Pansporoblast of Chloromyxum funduli embedded in a degenerated muscle fiber. The
contained myxospore has taken up the stain, but the protoplasm of the pansporoblast is absolutely
devoid of visible structure. Note the even contour of the characteristic lobose pseudopodia. 15.24 by
rau. (X2000.)

Fic. 32. One of a group consisting of free young myxospores of C. funduli. Like the mature
myxospores, they stain readily, but their nuclei are not differentiated. They are, as a rule, not quite
so irregular, but the pseudopodia are always small and angular. Note the contrast between these and
the pansporoblasts. 3.7 by 4.54. (X2000.)

Fic. 33. Myxospore of C. funduli. The outline is approximately circular. The sporoplasm is
homogeneous but dense around the four polar capsules, doubtless because of the greater thickness at
this point. The four nuclei are always associated with the polar capsules, hence are doubtless capsule
nuclei. Diameter, S.oz. (X2000.)

Fic. 34. Myxospore of C. fundulf seen from the side. Note the sporoplasm is not much denser
about the polar capsules. The sporoplasm tapers to a blunt apex. In many itis more pointed. The
polar capsules have long, curved, tapering necks with the large ends far apart. The capsule nuclei
alone stain. 8.2” by 6.74. (X2000.)

Fic. 35. A fresh sporoblast of M. musculi from the same slide as figures 20 and 27. The cyto-
plasm is rich in granules. The nucleus is very large and has a conspicuous karyosome. (Not drawn
to scale.)

Fic. 36. An isolated epidermal cell derived from a mass near the margin of an advanced ulcer,
most of which have numerous unstained bodies like those in the muscle fibers (fig. 12, pl. xx). From
sections. The epidermal cell is not typical in appearance, but the unstained bodies are, and are iden-
tical to those in the adjacent slightly atrophied epidermis. (X 2000.)

Ey AI keke,

BUET Uo Del TO13.

BulLE Uno. Bake Tors: PLATE XXI.

AN ECOLOGICAL RECONNOISSANCE OF THE FISHES OF
DOUGLAS LAKE, CHEBOYGAN COUNTY,
MICHIGAN, IN MIDSUMMER

*

By Jacob Reighard

Professor of Zoology, University of Michigan

215

oT
‘
.

MUD DUM AO OD ae

. PATH eke LE Bavaro: ln ie

i
| }
Pry Shree fi ‘
Wiis * Cribs: i re) ify
i
i

POLO Ia
2A LON
th

CONTENTS.

Barrentstony=shoalua pita teruaersyeerttsterssictetebeiotels cher sicels
stheweretationbalbitataseeitmeritsiriei-teie cil acrii rs

‘Dheideep-waterphabitatiyrrrycyrvarsy= cies ccayesctes steyeyeyeheiasersvevets

pUEVISHES HMR saat sores canre rete mie es hails utissneiG cees favagal SccPA Sys fa Sarai Suave
akepherrin por CiSCOs pape ois cleicusisleriaiew elec eae eeiereeie oie’
Gomimonistick errr acta ctracicitervarn ae oie enricfemi

Hornedidace or creek Chu)... 45 sea jeer cavvoua‘eis soiceieaiererstaceoaie
Calynty ar OW pete ersci ays sic che ker estos canna) Shas storesS viavaje le @iaveishare
‘Spot-talledimunmnoweeery-e semen saeieie c eyecersiicrier
Wommorntshin eres cose ey stasis. arsiawrcereraneteserore even steele terete
Contmomibullheadhrrraniciesstexeriitey tedtcrdsieteccrctere erate cher =
IMAG TIT O Whey sess es co si clara a eyeial ansietohe nsiahcie shane be Sheresefeters 318
Commongpikevor pickere le party ei) seit eater riers ae
PPro ta tayp OF CH sey. scp ctee evs aso rcsyer ea levevencrave, sieheravners) Sraeravaenes deen
PROCICDASS o2y te cercrais efscanciaiani sist sieves esyoyshas sieues Saas) sqanevansMelerets eiere ee.
TS) Gb C-S10 Re rape Sacnidn oo Uli POOR An Ona U cca etc neo Oeraac ae
PUTT PKAMISEE Cee ngageetegevey ears cherevevaretsnersvesecote eis sycfevayereneisieborche gists
Sialllemouthedsblacksibassiy-mieyiesieorateeiels)-rtiae enter rete aie
Warge-motthedsblacks bass: way cc rcisceie ic seas lee ee sletenale eeteyers
BY OLIOWADELCIIE sre reseteencye cusisies a enists erate <G seek tes aid
Iol se aaKS apie reas Cee OO GM Se CORREA COn nematrCR eae ares
OL yeartetecrarreysteytoletets eo cher. sia ph cevsin sie sdceainiees) ora os
theostomaviOwse Fe) snc. euersicreccrei« sieirecin sclera mites mierey-aers
IM erts tha att 0 vaya fatevecovarurseevouareustencieyeratecsiepertie eer erate drarataretete

ishicommimnities of Douglas akessasees ser cicero eee
diheicommunity/of young fishes. ve oe cae ce eee eee
eiheystotiyeshoalkcomm unihyi reise eiselaciseris eect ais
pubenvepetation community. mymsceee er sierra esis
ithevdeep-watericommunity ane eee eect eects

Relation of Douglas Lake species to those of other waters... 1.2.1.2... ee eee cece eee eee eee

GVMNEIR Rs op Sada ee ha eton ne ARO COREE SO aE OT ane aa ote

Pie:

ie "és - > Ot eae a ae mr, yh ee ie a
a — 7 as te 7 i as ew vi ioe :
i os > Cs a 1 - bite wtih z4iget
a) - - o, ae os “ah st ~~ a a a
~~. a ; ; fe? . ‘ 1 7 “” ie
: of - - -@ > a ein i, give Rx ad
a 7 - ™*
7 as 7 ay) we “i ‘ry NM
ire ce wae a , Se ma a ——
> _ _ i n ’ - } ‘a 9 bora ta ign iia f
o oe 7 iz. mu iia wean = ni bert a
- . ' eee co oN alee Say ’
7 a! - : -| oe ees
me 7 : a yp [es soe
od - ee esi Sen ”
; =i* oe . : _ ote a ore ta quit
bl in a ee ee hapa ys
7: ae ; as 4 1S
: fl 7 J p = - 1 7 mary | pan Zt
as oe - ne 7 : a ' 7 H ii Cj
sr, | 4 — har
iv - = a 7 7 6 = 7 iF _ al are} Be on Wy
i 7 _ 7 a> | Fe 7 ; 4 te opin ei ipa
7 H : a ; i - (a ie “oe whee
: a = ' hy 4 a” oo ie pul
> ary oo x it 7
a, i ' Pa (ari
, - 7 : - 1 es ie Bea
™ >} rh ; ay: rah T jug
ST ee: —- . oe =i) tm
rm. | eee i re -s he sau “1 Sapte? &
a — : ‘ - : = his ‘dpiete fade al “i VWirakaig : 5
> rn ar 7 - ? too eee mgm ae ‘ .
— z = ¥ f :
. Phe —— - 1 PAR ag tote
ae - . _ re a Nesta angers ie leaee
: an) a _ an i! a geen! Yara a wis _ U eyes, i » hil ‘
i . - 5 7 7
7 p a ™ > a ; , ‘?e & oe ——
’ ro — - :
. : : yy
=F P -
7 ,
7 - : > - ve
7, g = D ” — + “i a
fs : ~
7 ; 7
7 * 7 ; yy 7 a : ;
7 a » r 7
j a; ;
- : i ‘
_ ¥ . a
i Vs : rn
=. : ; papas

AN ECOLOGICAL RECONNOISSANCE OF THE FISHES OF
DOUGLAS LAKE, CHEBOYGAN COUNTY, MICHIGAN,

IN MIDSUMMER.’
&

By JACOB REIGHARD
Professor of Zoology, University of Michigan.

Bd
INTRODUCTION.

During the summers of 1909, 1911, and 1912 I was in charge of the biological station
of the University of Michigan, situated on Douglas Lake. ‘The little time that remained
to me after my routine duties was given to collection of data bearing on the ecology
of the fishes of the lake. It was my hope after a considerable number of years to reach
general principles by the analysis of data thus collected. It is unlikely that I shall
continue the work. It seems, then, worth while to put on record such facts as I have.
They are few, and the inferences that may be drawn from them are tentative; yet
they may furnish a starting point for some one else or suggest a method. ‘The records
of 1912 were made under my direction by an assistant, Mr. M. E. Houck.

Douglas Lake—Turtle Lake on many older maps—(fig. 1), lies at about latitude
46° 30’ N., in the ‘Southern Peninsula of Michigan, at an altitude of 712 feet above
sea level. Its northern shore is some 15 miles in a direct line from the Straits of
Mackinac. Its greatest length from east to west is 314 miles, its greatest width 21%
miles. The lake has somewhat the form of a fish, the flukes of whose tail form North
and South Fishtail Bays at its eastern end. ‘The total area of the lake, exclusive of
Fairy Island, is about 5.1 square miles; its shore line, including that of Fairy Island,
measures 14 miles. The shores are nearly everywhere a mixed sand of granitic origin.
The water deepens gradually over a terrace or shoal until it is 3 to 6 feet deep. The
terrace varies in width from a few yards to a hundred or more. The bottom then
drops rapidly, in most places into deeper water, forming the “slope” or margin, which
is as steep as loose sand can lie. The slope is that part of the bottom on which vege-
tation ordinarily grows. It extends to the lakeward limit of vegetation, usually at
a depth of not more than 25 feet. The depth of water at the lakeward limit of vege-
tation in Douglas Lake is unknown. The deeper water beyond the slope has a depth
of 82 feet over a limited area near the southern end of South Fishtail Bay, and a depth
of 89 feet at another point. The deeper water does not reach 80 feet over most of the
lake and is not continuous but is interrupted by bars and shoals. Pending the comple-
tion of a hydrographic map, details are not available.

A white disk lowered into the water on August 12, 1913, disappeared at a depth
of 12.5 feet. This indicates that the lake is not rich in plankton, but no plankton

@ Contributions from the Zoological Laboratory of the University of Michigan, no. 143.
219

220 BULLETIN OF THE BUREAU OF FISHERIES.

measurements have been made. The bottom temperature at a depth of 70 feet on
July 10, 1912, was 47° F. (8.3° C.). At a depth of 82 feet in South Fishtail Bay a
temperature of 6° C. has been recorded in July. There is a well-defined thermocline
at a depth of 4o to 45 feet. Its unusual distance from the surface is due to the heavy
winds which cause the surface waters to be intermingled. Above the thermocline
the temperature rises until in August it reaches 20° C. at the surface.

USS VINCENT
Qu LAKE

OOUGLAS LAKE
RESORT

|

,
A Wy Wi
DOUGLAS LAKE RESTOR UTA yy,

Fic. 1.—Map of Douglas Lake, Cheboygan County, Mich. The numbered circles show where collections were made.
1, Maple River; 2, two hundred yards east of Grapevine Point; 3, at the biological station; 4, sand bar; 5s, protected
cove on North Fishtail Bay; 6, deep channel east of Grapevine Point; 7, northeast of Fairy Island; 8, stony shoal on
Grapevine Point; 9, protected bay at Bryant's landing; 10, east shore of South Fishtail Bay; 11, west side of entrance
to North Fishtail Bay; 12, west side of South Fishtail Bay off Grapevine Point.

Above the thermocline in August there is abundance of dissolved oxygen, about
5.5 cc. per liter at the surface and 4.5 cc. at a depth of 33 feet. Below the thermocline
the amount of dissolved oxygen is diminished. It varies in August from about 0.6 cc.
per liter at a depth of 50 feet to nothing at a depth of 63 feet or more (Tucker, 1913).
Below a depth of 45 feet the lake does not afford, in summer, enough oxygen to make

it a suitable habitat for fish.
Little is known of the distribution of the vegetation in the lake. It is briefly
discussed under fish habitats, but should be made the subject of special study.

~ FISHES OF DOUGLAS LAKE, MICHIGAN, 221

Douglas Lake was at one time continuous with the Great Lakes and with Burts,
Mullet, and Crooked Lakes to the south of it. The latter lakes continue to be broadly
connected With Lake Huron by means of Crooked and Cheboygan Rivers. Douglas
Lake, on the other hand, was long since separated from the other lakes. It has no
direct connection with the Great Lakes, but is connected with Burts Lake by the
Maple River. Thus its separation from the Great Lakes antedates that of Crooked,
Mullet, and Burts Lakes and is more complete.

FISH HABITATS OF THE LAKE.

The following four fish habitats of the lake are provisionally recognized. A possi-
ble fifth habitat is suggested on page 246.

BARREN SAND-SHOAL HABITAT.

Wherever the terrace is without stones (pebbles may occur in the sand) it may be
referred to as a barren sand shoal. The sand is loose and shifting in the more exposed
of these shoals and practically always shows ripple marks. In less exposed places the
sand particles are loosely united by a deposit of marl, probably of algal origin. This
gives a certain firmness to the sand, as though it were mixed with clay and makes it
resistant to wave motion. Such protected sand shoals are often free from ripple marks.
The shore bordering all sand shoals is low, without an ice rampart of stones, and the
first land terrace or bluff is at some distance from it. The shoal may be narrow (only
a few yards or more) or wide (a hundred yards or more). The slope on its seaward edge
is steep, as steep as loose sand can lie, and the water over the seaward edge is commonly
about 4 feet deep. Near shore there may be a sparse growth of bulrushes but there is
no other vegetation.

BARREN STONY-SHOAL HABITAT.

Wherever the shore is bordered by a bluff or terrace which is being eroded the
shoreward margin of the shoal contains stones or small bowlders. Along such a shore
there is commonly formed by the action of the ice a rampart of stones, which borders
the shore like a low stone wall. The stony shoal is apt to be wide and the slope beyond
it less steep than that of the sand shoal. The water over its outer edge is often.6 or 7
feet deep. Its bottom may be of shifting sand or of sand agglutinated with marl. Its
shoreward border may support a growth of bulrushes or may be without them. Re-
garded as a fish habitat, its salient feature is the stones. Where the shore of the lake
shows a series of headlands with intervening valleys, it is being eroded along the head-
lands and built up between them. The headlands are bordered by stony shoals and the
intervening low shore by sand shoals. The two pass into one another without sharp

demarcation.
THE VEGETATION HABITAT

If we neglect the scant growth of bulrushes which may occur on the shoals, the
vegetation of the lake is largely limited to the slope. In places the slope is continued
into considerable areas of nearly level bottom covered by water less than 25 feet deep
and overgrown with vegetation. Such an area, known as the “middle ground,” extends

19371°—vol 33—15——15

222 BULLETIN OF THE BUREAU OF FISHERIES.

eastward and a little north from the northern end of Fairy Island to the mainland.
There are similar areas south and west of Fairy Island. The slope, the middle ground,
and similar areas lie in the open lake and are subjected to severe wave action. The
water is usually in motion and the conditions are not in this respect unlike those found
inastream. Among the fish is at least one characteristic stream form, Notropis cornutus,
the common shiner. ‘This habitat may be called the unprotected vegetation habitat.

The vegetation is not emergent and is characterized by absence of water lilies. On
the slope the plants are found on the less steep portions and there form a discontinuous
fringe or zone. Ona still, bright day one may see that the plant growth of the slope
consists of little groups of Potamogeton natans, millfoil, and perhaps other plants which
are 1 or 2 feet apart and in most places do not make dense masses. There are consid-
erable stretches of the slope that are without vegetation. One of these lies opposite
the laboratory on South Fishtail Bay. There are a few places in which the vegetation is
more dense. On the whole, it occurs in patches or islands and within these it is sparse.

Where the shoals are protected from the wave action vegetation gets a foothold,
muck accumulates, and the conditions approach those of a pond with relatively quiet
waters. This is the case on the east and west sides of North Fishtail Bay, in the bay
directly south of Fairy Island, and at the mouth of Bessie Creek. Water lilies occur in
such situations and the large-mouthed black bass is the characteristic but not abundant
fish. The common sunfish is more abundant here than elsewhere. This habitat of
bays and estuaries may be referred to as the protected vegetation habitat. It contains
most of the species of fish to be found in the lake. It merges into the unprotected
vegetation habitat. For the present it seems best to treat the vegetation habitat as a
unit, although in the future it may be advisable to subdivide it.

_THE DEEP-WATER HABITAT.

Beyond the slope near the bottom is the abysmal or deep-water region, where the
bottom is of a soft, black ooze and where there are no large water plants. It extends
from a depth of 25 feet (probably somewhat less) to 89 feet, the extreme depth of the
lake, and comprises the bottom and the layer of water 1 or 2 meters thick above it.
Above the thermocline this layer of water is agitated by the wind, is relatively warm
and well lighted, and contains in summer an abundance of oxygen. Passing downward
along the bottom through the thermocline we encounter within a vertical distance of 7
feet a drop in temperature of some 10° F. As we descend the temperature near the
bottom continues to drop from about 54° F. at the thermocline (July 10) to between
43° and 50°, depending on the depth reached. The water below the thermocline is not
only cold but relatively quiet, unaffected by wave action, and relatively dark. In mid-
summer it contains little oxygen at any level and none at all at a depth of 63 feet or
more. We have taken no fish below the thermocline in midsummer. They are then to
be found only in those parts of the abysmal region that lie above the thermocline
between the lakeward border of the vegetation zone and a depth of about 45 feet.

THE FISHES.

Our data concerning the fishes are given below under each species. The locality
numbers in the tables refer to the map on page 220. The numbers in the column headed
“Water depth” give the distances below the surface at which the fish were taken. They

FISHES OF DOUGLAS LAKE, MICHIGAN. 223

are usually the depths at which gill and fyke nets were set on the bottom. Lengths of
fishes do not include the caudal fin.

LEUCICHTHYS ARTEDI (Le Sueur), lake herring, or cisco.—This species has not been
taken in nets, but adult specimens are frequently cast up on the beach of South Fishtail
Bay. Three of them measured 5,614, and 7 inches, respectively, the latter a male with
slender white testes 14 inch broad. A male 674 inches long and with large testes was
picked up, still living, over deep water in South Fishtail Bay, September 18, 1911. Our
small-meshed gill net has taken suckers of 7 or 8 inches when set on the bottom in water
of 26 and of 42 feet depth. Itshould have taken lake herring if they had been present
there. In midsummer the same net has taken no fish when set on the bottom in water
deeper than 45 feet, although in September a single sucker was taken at 72 feet. The
absence of oxygen in the bottom water below 45 feet in midsummer makes it impossible for
fish to live there. The lake herring must therefore live in deep water at some distance
above the bottom. Perhaps its habitat will be found in the neighborhood of the ther-
mocline. This species is characteristic of the Great Lakes, where its average length is
12 inches. Our largest specimens are only 7 inches long.

CATOSTOMUS COMMERSONII (Lacépéde), common sucker—The records in table 1
show the suckers taken in 1912.

TaBLE I.—RECORDS OF CATOSTOMUS COMMERSONII TAKEN IN DouGLas LAKE IN IogI2.

|
- : Water Local-
No. | Weight. | Length. Sex. depth Apparatus. Stomach contents. ityin | Date.
sy} Fig. 1.

a
8

; ...| Empty
‘Gill; Not examined.

O OI AnNhW WH
wo
AwWRADEAbPENO
a
Sa pa
ANAnNAnAAnn

»
a

p
aw

WAAKHEADRARANDHDH
Lo]
$

Common suckers have been seined in Maple River and have been seen at the mouth
of Bessie Creek. There can be no doubt, then, that they occur over the whole lake and
are amongst its commoner fishes. They are found in all the habitats. On September
23, 1911, one was taken in a gill net drawn from a depth of 72 feet. In July and August
suckers have not been taken below the depth of 43 feet. They are sometimes seen
feeding on the sand shoals in water a foot or two deep. They may therefore occur at
any depth in the lake, but are not known below the thermocline in midsummer.

Food.—The young of this species are seen on the sand shoals in July and August
in company with the young of the yellow perch and the spot-tailed minnow. Ina school
of 475 of these young fish taken on September 1, 1911, five were suckers between 134 and
2inches long. ‘The alimentary canal, from cesophagus to anus, of an individual 2 inches
long was found to measure 3 inches. Its contents formed a brown mass inclosed in a
mucus pellicle. The whole of it could be easily stripped from the canal. The contents

224 BULLETIN OF THE BUREAU OF FISHERIES.

when forced out of the pellicle, proved to be wholly shells of a species of cladoceran,
apparently Chydorus. A three-sixteenth-inch piece was cut from the middle of the
alimentary canal where it is of average size, and the Crustacea in an estimated fifth of
this counted. From this the entire number in the alimentary canal was estimated at
about 2,400. Only 2 or 3 copepods were found in the sample, or 48 for the whole
alimentary canal; the rest were Cladocera and all of one species. There was no sand,
so that it may be safely said that the young sucker is not a bottom feeder, but lives
wholly on the plankton.

The stomachs of 12 of the suckers included in the table were examined and found
to beempty. In August, 1912, Prof. Frank Smith saw the adults feeding on the mate-
rials encrusting the vegetation at the mouth of Bessie Creek. They were sucking off
whatever adhered to the floating stems and leaves of the plants. They went from
plant to plant and mouthed over each branch from base to tip until the whole plant
had been gone over.

The adult suckers may sometimes be seen at dusk or daybreak feeding on the
bottom over the sand shoals. When approached they ordinarily make off at once for
deep water. On July 3, 1912, I found several feeding on the sand shoals at midday,
and each was surrounded by a group of a dozen or more log perch. The log perch
were at that time laying their eggs in the sand and the suckers were feeding on the eggs.
Each sucker was surrounded by a group of log perch which were trying to get such
scraps as might be left from its feeding. It would be interesting to know whether this
commensal relation between sucker and log perch obtains at other seasons and in
deeper water.

While the suckers are thus engaged it is not difficult to approach them until they are
at one’s feet and to watch closely their method of feeding. The sucker moves slowly
over the bottom. At intervals it stops, raises its tail until, if in very shallow water, the
caudal fin breaks the surface. It buries its snout in the sand, often to the nostril, but
sometimes only half so far. The fish then withdraws its snout from the sand and without
moving from the place, works its jaws for several seconds as though chewing. At the
same time a thin stream of sand is seen to come from its mouth. At intervals there
is a sudden spurt of water and sand from its mouth so violent that it disturbs the
bottom. When the fish has ejected all the sand it moves a short distance with its
pectorals in close contact with the bottom and repeats its feeding movements. Wherever
it has thrust its snout into the bottom there is left a deep pit which is usually a sharp
mold of its snout and the lower part of its head. The pits are connected by broad
sinuous trails made by the pectoral fins of the fish. These suggest the tracks of a huge
snail and show oblique parallel lines where the edges of the pectorals have pressed against
the sand at each stroke. These pits and trails are very characteristic impressions, and
are abundant in shallow water throughout the summer. They are more numerous in
protected places where the bottom is made somewhat coherent by the formation of
marl, and where it possibly contains a larger percentage of nutritive matter. These
“tracks’’ of the sucker enable one to tell each morning where they have been feeding
during the night and in what abundance.

Great numbers of dead suckers are thrown up on the beach in South Fishtail Bay
in July and August. Many of these have the characteristic form of starved fish. The

FISHES OF DOUGLAS LAKE, MICHIGAN. 225

back is thin and sharp instead of round, and the head is disproportionately large com-
pared to the body. This is because the head is made up largely of bone, and emaciation
~ can not so greatly reduce its bulk as it does that of the more fleshy body and tail. ‘The
emaciated fish do not appear to be diseased and are not usually parasitized heavily
enough to account for their emaciation. Death seems to be due to starvation.

Hankinson (1908) collected 41 suckers in Walnut Lake and gives their average
weight as 2.5 pounds. Their average length, including the caudal fin, is, from Hankin-
son’s tables, 16.2 inches. From Forbes’s and Richardson’s (1908) figure of the common
sucker, the length from tip of the snout to the base of the caudal is found to be 0.88
of the total length from tip of snout to tip of caudal fin. Applying this correction
to Hankinson’s average length, we get an average length of his suckers of 14.3 inches,
measured in the usual way from tip of snout to base of caudal. In contrast to this the
14 fish taken in Douglas Lake have an average length of 9.5 inches and an average weight
of 0.48 pound.

From the fact that all stomachs of the common sucker were found empty, from their
habit of feeding on the comparatively innutritious materials of the lake bottom and
on those covering the stems and leaves of plants, from the large number of deaths among
them in midsummer, and from their relatively small average size, it may be inferred
that the fish get insufficient food.

In Walnut Lake Hankinson (1908) found, as the result of the examination of the
alimentary canals of 13 common suckers, caddis-worms and cases, small bivalve mollusks,
amphipods, insects, marl, midge larve, and Daphnia. Of these, the caddis-wonns,
amphipods, and midge larve are commonly associated with vegetation. It is not
unlikely that the relatively slight development of vegetation in Douglas Lake makes
it an unfavorable habitat for suckers.

The breeding grounds of the Douglas Lake suckers are unknown. According to the
writer's unpublished observations, suckers breed in streams where there is swift water
and gravel bottom. These conditions are found in Maple River and in Bessie Creek.
Young suckers less than 2 inches long are found in June on the shoals of South Fishtail
Bay, about 6 miles by the shore from either of these streams. They are doubtless fish
of the season and, if the breeding habits of the suckers of Douglas Lake are the same
as elsewhere, the young must have wandered to the shoals from the breeding grounds
in Maple River and Bessie Creek. It is possible, however, that the essential requirement
for breeding is suitable bottom, not running water. Bottom suitable for suckers is
plentiful in Douglas Lake on the shoals, and the young suckers found there may be still
on the breeding grounds.

In figure 2 the lengths and weights of the suckers included in table 1 have been
plotted and a curve sketched to show their relation. It is clear that there is a definite
relation of such a sort that, after a weight of 4 or 5 ounces has been reached, length
increases less rapidly than weight. Thus between the weights of 4 and 5 ounces the
increase in length is about 0.75 inch, while between 14 and 15 ounces it appears to be
scarcely 0.1 inch. At the 15-inch length the line is nearly horizontal. Our data are not
enough to make it advisable to draw the length-weight curve mathematically or to
determine its formula. (See Hecht, 1913.)

PIMEPHALES NOTATUS (Rafinesque), blunt-nosed minnow.—Ten specimens, 2% to 27%
inches long and evidently adult, were taken in the seine at Bryant’s dock (location 9

226 BULLETIN OF THE BUREAU OF FISHERIES.

on the map) on September 20, 1911. One hundred and ninety-three individuals about
134 inches long were taken on stony shoals on the west side of the entrance to North
Fishtail Bay on September 18 of the same year. With them were three Notropis
cayuga and four N. hudsonius. A few were taken in August in company with large
numbers of N. hudsonius on the sand shoals of South Fishtail Bay. In life they
are distinguishable from N. hudsonius by the following field characters: (2) Darker
color; (6) a peculiar jerky movement in progression. The fish do not move directly
ahead, but by a flick of the tail and of the pectorals the head is jerked to one side and
then to the other or several times to one side and several times to the other, so that the
course is zigzag; (c) the body is semitranslucent, so that the vertebral column and the
viscera may be seen faintly from the back; (d) the scales in front of the dorsal on
the back are crowded so as to appear much smaller than the scales behind them.

SF 10 1S 20 25 30
Ounces
Fic. 2.—Graph showing the relation of length and weight for the 14 common suckers, Catostomus commersonii, included
in table 1. Each space on the horizontal line represents 1 ounce; each space on the vertical line x inch. Curve
drawn free-hand.

We have collected this species in numbers only on or very near stony shoals and
in the neighborhood of protected bays. Stony shoals afford it breeding grounds, for
it lays its eggs beneath flat stones and similar objects on the bottom, and the mucky
bottom of protected bays affords it food, for it is a “‘mud eater.”” It has been taken but
rarely and in small numbers on the sand shoals along the south and west shores of South
Fishtail Bay, although frequent collections have been made there. These wave-swept
shoals afford neither stones nor muck.

SEMOTILUS ATROMACULATUS (Mitchill), horned dace, or creek chub, has been taken
only in the vicinity of Bryant’s dock (locality 9). Here the adult was found in consider-
able numbers in company with Pimephales notatus and N. hudsonius. It is abundant
in Maple River near the lake. Bryant's is a resort of fishermen. It is possible that the
horned dace has been introduced here asa bait fish and has not extended its range to other

FISHES OF DOUGLAS LAKE, MICHIGAN. 227

parts of the lake, but it is more likely that it has made its way thither from Maple
River.

NotTropis CAYUGA Meek, Cayuga minnow.—Three specimens only have been taken
in the lake at locality 11, on stony shoals at the west side of North Fishtail Bay at the
entrance. The slopes bordering these shoals are sparsely grown with plants, and they
are so much protected from wave action that there is a thin crust of algal marl uniting
the superficial sand particles.

NOTROPIS HUDSONIUS (De Witt Clinton), spot-tailed minnow, and the common
shiner are the most abundant of the Douglas Lake minnows and the most widely dis-
tributed. On July 29, 107 specimens of the spot-tailed minnow were taken with the
seine on the sand shoals of South Fishtail Bay in about 2 feet of water. They were of
nearly uniform length and averaged 2.8 inches; 64 were females and 24 males. These
were immature fish and were in schools together with young of the yellow perch and
common sucker. Two hundred and sixty-seven immature individuals were seined in
the same place on September 1, 1911. On the 20th of July, 1912, 9 mature individuals
31 to 37's inches long were seined at Bryant’s dock (locality 9), together with mature
Pimephales notatus and S. atromaculatus. Mature individuals 4 to 6 inches long are found
in many places in the lake where there is abundant vegetation on the slope. Here the
fishermen seek them for bait and take them with the baited minnow hook by casting,
as one casts for trout with the fly. They are taken in company with the common shiner
and in about equal abundance. ‘The fishermen locate these schools by the disturbance
of the water’s surface due to their rising, and often visit several patches of vegetation
before they find them. Hence it appears that the fish may travel together in schools
from one patch of vegetation to another.

The alimentary canal of one of the immature individuals taken on the sand shoal was
found to be filled with Cladocera, apparently of the genus Chydorus, the form that makes
up the bulk of the food of the young perch and suckers taken in the same habitat. The
Cladocera were apparently as numerous as in the young perch, but there were no other
Crustacea such as occur in the perch. The short, slender, close-set gill rakers with the
narrow gill openings make an excellent apparatus for the capture of these small Crus-
tacea. The roof and sides of the mouth and the tongue have many short papille set in
curved longitudinal rows, and these may serve to hold the Crustacea while permitting
water to pass backward. ‘There are no records of the stomach contents of the adults
of this species in Douglas Lake. Elsewhere (Forbes and Richardson, 1908) it is known
to feed on insects, crustaceans, and vegetation.

NOTROPIS CORNUTUS (Mitchill), common shiner, is taken on the hook in the same
manner as N. hudsonius and in company with it in patches of vegetation in nearly all
parts of the lake. It is very abundant. Three taken in South Fishtail Bay in August,
IgII, measured, respectively, 3.5, 3.75, and 4.06 inches in length. These were in fine
condition, the mesentery heavily laden with fat. The contents of the alimentary canal
were as follows for the three specimens:

1. About two-thirds Cladocera, apparently Chydorus; one-third insects, apparently
larval.

2. Remains of insects and a small quantity of Cladocera.

3. Some fragments of broad, green leaves on which were bryozoan tubes; some
Gloitrichia; a large number of detached bryozoan branches, some of them with stato-

228 BULLETIN OF THE BUREAU OF FISHERIES.

blasts; an insect larva; a green gelatinous mass including Cladoceran shells and prob-
ably composed of partly digested alge.

Forbes and Richardson (1908) say of the common shiner: “It is especially a minnow
of creeks and the smaller rivers—our coefficients for which are 3 and 2.45, respectively—
scarcely ever occurring in either lakes or the smaller streams. It shows also a marked
preference for clear waters.”” Hankinson found this species in Walnut Lake, chiefly on
shoals with ‘abundant luxuriant aquatic vegetation and black bottom soil.” It was
common on but one shoal. Its abundance and wide distribution in Douglas Lake are
unusual. It occurs not only in the lake but is the commonest fish taken in the seine in
Maple River. According to the writer’s unpublished observations, the species breeds
only in running water on gravel bottom. Maple River and Bessie Creek afford the con-
ditions of its known breeding grounds. Moreover, the young fish have not been recog-
nized with certainty in the lake, which adds to the probability that it does not breed
there. It is more likely that the adults travel from the lake to the breeding grounds in
Maple River and Bessie Creek and that when partly grown the young go from the breed-
ing grounds as far as the eastern end of the lake, a distance along shore of about 6 miles.
The breeding grounds of this most important bait fish of the lake should be located and
preserved.

AMEIURUS NEBULOSUS (Le Sueur), common bullhead, does not appear to be abundant.
In 1911 four were taken on the hook in the vegetation on the east shore of South Fishtail
Bay. Three of them measured 9% inches in length and the fourth 1034 inches. The
records for 1912 are given below:

TaBLE II.—REcORDS OF AMEIURUS NEBULOSUS TAKEN IN DouGLAs LAKE IN IgI2.

Water Local-
No. | Weight. | Length Sex depth Apparatus. Stomach contents. ry in | Date.
Pa
Ounces. | Inches.
I 13 10.4 | Female........... Bumblebee.2 cascee ose 4| Aug. 6
2 Io Tr leans dOtalccentcssae .| Small fish deter 4| Aug. ro
3 EMPEY tyes ciocmiveciseeeete 7 | Aug. 20

| 3 6 Malet seememaccciene

None were taken in gill nets set at greater depth than 12 feet, so that they are
probably confined to the vegetation of the slope and to similar situations elsewhere. The
largest specimen taken is 6 inches shorter than Hankinson’s (1908) largest (Hankinson’s
measurements include the caudal fin), and 7 inches shorter than the Illinois maximum as
recorded by Forbes and Richardson (1908). Examination of the contents of two stom-
achs shows nothing unusual except the inclusion of a bumblebee.

Young individuals of this species were taken in July at various points along the
shore of North Fishtail Bay and in an adjacent beach pool. In 1909 swarms of young
were seen in the same place together with the male. In August they had reached a
length of an inch and a quarter.

Umpra Limt (Kirtland), mud minnow, has not been taken in the lake itself, but is
abundant in the oxbow ponds that have been cut off from Maple River near the lake.
It should occur in the mucky bays and estuaries of the lake itself.

Esox Lucius (Linneus), common pike, or pickerel, is the largest and one of the most
abundant fish in the lake. The following table gives data concerning 22 individuals of
this species.

FISHES OF DOUGLAS LAKE, MICHIGAN. 229

TABLE III.—ReEcoRDS oF Esox Luctus TAKEN IN DoucLas LAKE.

No. | Weight. | Length Sex Water A t St h tent | eg aa
a 7 gth. 3 depth. pparatus. Stomach contents. its in Date.
| a sl
| |
Ounces. | Inches. Feet.
I 104 30.6 150 |, Gills ee oc qeestete @iA-INCO Perch =..miseieeeleineie 12| July 3,19rr
| 28 15 25-30 |..... Oce se eart sees 2 perch, 414 and 334 inches 3 | July 11,1912
ong.
3 12 13-2 4 Fish, not determined 3 | Aug. 3,1912
4 15 15-4 12 3 | Aug. 8,1912
5 15 14-8 12 5 | Aug. 11,1912
6 16 15-2 12 5 | Aug. 12,1912
7] 21 16.4 I2 5 | Aug. 13,1912
8 30 18 26 4| Aug. 15,1912
9 29-5 16.8 26 6 Do.
Io 26 17 26 6 Do.
II 10 12.8 26 6 Do,
12 10 12.8 26 6 Do.
13 18 15-6 26 6| Aug. 16,1912
14 2 17-6 26 6 0.
15 2. 17-6 13 7 | Aug. 20,1912
16 29-5 17-2 13 7 Do.
17 10 13-6 25 7 Do.
8 10.5 13-6 25 ai Do.
19 25 17-8 25 7 Do.
20 27 18.2 12 5 | Aug. 14,1912
2I 12 13-2 12 5
22 59 19.6 45 6| Aug. 18,1912

This fish has been taken in all parts of the lake and at all depths between 4 and 45
feet. It appears not to go below the thermocline in midsummer, but at other seasons
it is possible that, like the sucker, it. goes into deeper water.

Seven of the 22 stomachs examined contained the remains of fish, while the rest
were empty. There is no evidence that in midsummer the Douglas Lake pike takes
other food than fish. It is clear that the spiny fin rays of such fish as perch do not keep
them from the maw of the pike, for in two cases the stomach contents were perch about
4 inches long.

The pike tabulated range in weight from 10 to 104 ounces. We have taken indi-
viduals whose weight we estimated at 10 to 12 pounds and those of 18 pounds have been
reported by fishermen. Of the 20 whose sex was determined, half were males. The
average weight of males is 22.8 ounces, of females 19.6 ounces, but the number of fish
used is too small to make the figures significant and includes a single male of 59 ounces.

A curve showing the relation of length to weight in these 22 fish is sketched in
figure 3. The data are insufficient to show more than the general fact indicated for the
sucker that there is a definite relation between length and weight of such a sort that,
above 8 ounces, the length increases much less rapidly than the weight. The data
represented by the curve, although meager, would be of considerable value if there were
similar data from other lakes for comparison. It is probable that each species in a lake
shows a length-weight curve peculiar to it. It is also probable that curves for the same
species from different lakes might be characteristically different. The form of the curve
for a single species from one locality might show to what extent the conditions of that
locality are favorable to the species. Unfortunately the literature appears to contain
no records full enough for comparison with those of Douglas Lake. Forbes and Rich-
ardson (1908) mention for the pike an average length of 36 inches and an average weight
of 5 pounds. A curve for the Illinois pike, if it were to pass through the point thus

BULLETIN OF THE BUREAU OF FISHERIES.

230

“slOUNTT JO axTd ay} 10J DAM [woHaTIOdAY ¥ St AUT] MaAOIG aYT, ‘puey-desj UMEIP AAD "71 Y8nomy) UMLIP

JOT aJOJIIIT]} ST IAIMD JY} $IYFIOM Sz JO YS v JOF [eMIOT ay} UY} ssa] SEAL YSU] S}I 3eq} OS ‘anzeAino [euIds B peYy cz posjaqunu [eNprlarpurayy “sayour ¢ any peorpaA ayy
UO vovds yore ‘saomMO z syUasaidad aut] [e}WOzZIIOY 94} UO voids YORA “WI Gv} Ul papnpour ‘snt9n] rosy ‘aytd zz ayy JOj YSIOM puL YIZua] Jo uorzYJa1 ay} Zurmoys yderg—e “oT
apts
on Oe Aeon ze )

TT sare Sosdandacdosteesectontectacee’

esas (a ek OS a eR
DiGi i ee H HH
HE cH rH caeeeeeare
ie Beecg Ce
HEEEEEEEEEE EEE er
rH HEEEEECCeeee ana
re PC Elelalmile baa!
I Ea leet ISS te
Bees prne oe er cesT ier leis
| Ay? 7 | ea ar Te (ee
ES a St
CO eae gi
PCC aie aor Fa
appre EEE EE
aa ee at
CI iia Sleiak| BEER EEE EEE
a H EE aa SeR OSE Ames
r Ee ISICIc Is Wise
Obes Fs ST a OS

FISHES OF DOUGLAS LAKE, MICHIGAN. 231

located for a single average fish, would be higher and of different form from that for the
pike of Douglas Lake. Such a hypothetical curve is sketched in broken lines in figure 3
to show how curves for single species might be characteristic of localities. A comparison
of our curve for the pike with that for the sucker shown in figure 2 shows that they
differ. The sucker increases less rapidly in weight with increasing length than does the
pike. A sucker of 12 inches weighs about 14 ounces; a pike of 12 inches, if our curve
is correct, about ro ounces. This is for Douglas Lake. With data enough for many
species from many localities, one might be able to say, from a study of such curves, for
what species of fish the conditions of each locality were most favorable. By defining or
describing these conditions one might then possibly use them as a guide in the practical
operations of fish culture.

It is interesting to note that pike 22 of our list had a large hump on the back due to
curvature of the spine. If this was the result of an injury it had been inflicted so long
before that no external scars remained. The deformity may even have been congenital.
In spite of it the fish had thriven. The conditions of existence were not severe enough
to eliminate it. Its position is shown at 22 on figure 3.

PERCOPSIS GUTTATUS Aggassiz, trout perch, is known only from the numerous
specimens thrown up on the beach of South Fishtail Bay. It has not been taken in
nets. In one specimen 2% inches long the intestine contained the chitinous parts of
an insect larva. This indicates that its habitat is the vegetation zone.

July 17, 1912, following a storm, many adults were picked up on the beach. Among
these were females that gave up eggs freely on slight pressure. On the following day
a search was made of the shoals in the hope of locating the breeding fish, but without
result.

AMBLOPLITES RUPESTRIS (Rafinesque), rock bass.—The data collected concerning
this fish are brought together in table Iv.

TABLE IV.—RECORDS OF AMBLOPLITES RUPESTRIS TAKEN IN DouGLas LAKE IN IgI2.

; Water Local-
No. | Weight. | Length. Sex. Apparatus. Stomach contents. ity in | Date.
| Hopab. Fig. r.
|
Ounces. | Inches.

x 8 7-6

2 7 7-6

3 3% 5-7

4 9 8

5 64 7:2

6 I 4

7 2 5-2

8 I 4 E

9 1% 4-2].
Io I 3- i i
Ir 2 4.8
I2 12 8.8

13 13 9-6].

14 9 8
15 3 4-8 ---||Cambarus virilis; dragon
16 Zz 4:8 | fly larva.
17 2 5

Although the records show it from but two localities, it is taken wherever there is
vegetation in the lake. We have taken it at no greater depth than 51% feet, a depth
which is usually reached a little beyond the edge of the terrace. It may go deeper. I
believe it is sometimes taken at greater depth on the hook, but not beyond the vegetation.

232 BULLETIN OF THE BUREAU OF FISHERIES.

The table shows six specimens with food in the alimentary canal, and we have
records of four others. Five of the 10 fish whose stomach contents were studied con-
tained fish, in one case a sunfish, in the other cases the contents were not determinable.
Three contained the remains of insects, in one case a dragon fly larva. Two contained
each a crayfish; in one this was determined as Cambarus virilis, a species which is
common in the deeper water of the lake and which occurs also under logs and the like
in shallow water. In respect to food, the rock bass of Douglas Lake agree with those
of Walnut Lake and of Illinois waters.

Systematic descriptions give the length for the species as 8 to 10 inches, which
agrees very well with the 9 to 13 inches of our adult specimens. The weight-length

ee ee See
BREREREE DSA
/ 2 3 @ °° 6°78 -G-A0 VW se A? HB 71h
Ounces

Fic. 4.—Graph showing the relation of length and weight in the 17 specimens of rock bass, A mbloplites rupestris, included
in table rv. Each space on the horizontal line represents one-half ounce; each space on the vertical line one-half
inch. Curve drawn free-hand.

curve is given in figure 4, but there are no similar data from other localities for com-
parison. ‘There is nothing in our data to indicate that Douglas Lake is an unfavorable
habitat for this species.

A rock bass 8% inches long noted near the laboratory dock paid no attention to a
baited hook until touched by it, but was then hooked. It proved to be blind, emaciated,
with characteristically large head and without mesenterial fat. It was not too large for
a good sized pike to swallow, and it is remarkable that it should have escaped death so
long. ‘The struggle for existence among the inhabitants of the lake had not been severe
enough to eliminate either the blind rock bass or the crippled pike already referred to.
Both were finally eliminated by man.

LEPOMIS PALLIDUS (Mitchill), b/wegil/ —Records for 8 specimens are given in table v.

FISHES OF DOUGLAS LAKE, MICHIGAN. 233

TABLE V.—RECORDsS OF LEPOMIS PALLIDUS TAKEN IN DoucLAs LAKE.

Local-
No.| Weight. | Length. Sex. Ss Apparatus. Stomach contents. ey in Date.
1g. 1.
cs
Ounces.
I II 2| July 26,1912
a 6 2 Do.
3 I 4] Aug. 5,1912
4 4| Aug. 6,1912
5 5 | Aug. 5, 1011
6]. 5 | Sept. 5,191
7). .| Plants, mostly........... 5 Do.
8 Plants, insects ccesccsee 5 Do.

This species was taken only in shallow water and in vegetation. Although our
records are from but three localities, the fish is taken on the hook wherever there is vege-
tation.

In size our specimens accord with the 5 to 8 inches of the systematic descriptions.
The food of the 4 specimens examined is unusual and is therefore given in detail.

No. 5 had in its stomach 6 terminal buds and leaf whorls of Elodea, 2 terminal buds
and leaf whorls of a Chara, numerous detached leaves of Elodea, and some brown, half-
decayed fragments of vegetation, most of them long, like leaves of one of the Potamoge-
tons. The intestine was crammed with partly digested plant fragments, on some of
which were bryozoan statoblasts, and with numerous heads, wings, and legs of insects,
apparently dipters, all imagoes. No. 6 contained in the stomach a mass of vegetation
but no animal food. The vegetation was apparently water milfoil. The intestine
held the same material in addition to one or two small insects. No. 7 had the stomach
filled with unrecognizable plant fragments, on which were statoblasts. The intestine
contained plant débris, together with two hydrachnids and two or three ostracods. In
no. 8 the stomach was empty, but the intestine was filled with plant débris with numerous
fragments of insects. Some of these appeared to be dipterous imagoes.

The fact that the plant fragments form so large a part of the food and that they con-
sist in so many cases of succulent terminal buds and leaf whorls indicates that plant
tissues form a normal part of the food and are not merely taken adventitiously with
other food. Hankinson (1908) found only animal food in the stomachs of specimens
examined by him, while Forbes and Richardson (1908) record the occurrence of 24 percent
of plant food in some of their specimens, a percentage much less than in our specimens.

EUPOMOTIS GIBBOSUS (Linneus), pumpkinseed, is one of the commoner fishes in the
vegetation of the lake. Our data concerning it are given in table vr.

TABLE VI.—RECORDS OF EUPOMOTIS GIBBOSUS TAKEN IN DouGLAS LAKE.

. Water Local-
No. | Weight. | Length. Sex. depth Apparatus. Stomach contents. ity in Date.
: Fig. 1
SenoDonade = Small shells... ..5......:. 2| July 26,1912
Srselois CO Soxeieyeleieine cto Meter GOW ei teenijetarttie teretiat 2| July 29,1912
coond St ...| Not examined........... 4] Aug. 5,1912
Hae Sista Saal Seep ty eee eens sone 6 | Aug. 12,1912
me Minute shells, insects ... .| 6 | Aug. 16,1912
6. .| Insect larve . . 5 | Sept. 5,1911
5: .| Chara, snails 5 Do.
5: 5 Do.
5: 5 Do.
5: ; . 5 Do.
ahceciniees 3p oy ae we ; t:.|(Crushedisnails!)..2 22... -- 5 Do.
13) || jseciecnnile cee” llegaarpndsecodqanonbod nese topood OnSuooSboccenelaeEat Insect larva, ostracod |........ Do.
shells, sand.

234 BULLETIN OF THE BUREAU OF FISHERIES.

Like the bluegill, the pumpkinseed is found only in shallow water and among vege-
tation, but may occur wherever these are found in the lake.

The contents of the alimentary canal were examined in 9g individuals. In 6
small snails and their crushed shells were found. In 4 of the 6 the alimentary canal
contained no other material than the snails; while in the fifth it contained in addition
to the fragments of 3 small snails a quantity of Chara with orange fruits. The Chara
may have been adventitious. One stomach contained insect larve exclusively; one
contained insect larve in addition to snails; and a third insect larve, snails, and other
material. Snails appear to be the most important element of the food and next to these
insect larvae, but exact percentages are not available. The snails found in no. 8, 11,
and 12 were determined by Mr. H. B. Baker to belong to the species Ammnicola limosa
and Planorbis bicarinata portagensis. ‘The former were adults from one-sixth to one-
eighth inch long, while the latter were young individuals. Forbes and Richardson
(1908) found that snails made up nearly half the food of 9 specimens examined by
them, insects a fifth, and Crustacea a fifth. Hankinson (1908) examined 32 stomachs
and found May-fly larve to be the favorite food, although Crustacea, snails, leeches,
and other insects were included. The evidence on the whole indicates that snails are
the most important element in the food. Fishes appear not to be taken.

Hankinson gives the length of 16 specimens and their average is 5.8 inches, includ-
ing the caudal fin. The length without caudal fin as determined from figures forms 84
per cent of the total length, which makes the average of Hankinson’s specimens 5.1
inches without caudal fin. Our Douglas Lake specimens average 5.2 inches. Judged
by this standard, the conditions are about as favorable in the one lake as in the other.

The pumpkinseed appears to be more resistant to foul water than the bluegill.
When numbers of each were placed together in a pail of water, all the bluegills were
found dead after a time, while the pumpkinseeds were still active.

MICROPTERUS DOLOMIEU Lacépéde, small-mouthed black bass.—The Douglas Lake
data on this fish are given in table vir.

TasieE VII.—ReEcorpDs OF MICROPTERUS DOLOMIEU TAKEN IN DouGLaAs LAKE.

Local-
No. | Weight. | Length. Sex. bees Apparatus, Stomach contents. a in Date.
5 igor.
— | Se ee
Ounces. | Inches. |
I SF elarm wistets aleve : Sri ERIE a, 20k late aera: Sia latino erecta totais este ercisicer | 2| July 26,1912
2 T55 igen 5 d Empty iis hiccsstenacecen 2| July 29,1912
3 Ie5 5-2 : 5 lajdtaja'efaie eisia | siete GS sAigeqadceodoupcaadd | 2| July 21,1912
4 6 8 er Sige d] ahaa che AO)ais atest istaft erste eee doreeoeen <a 2| July 31,1912
Ss 40-5 14 dome: oes dOnecsann 6 | Aug. 15,1912
6 2 fish, 1 crayfis 6 | Aug. 19,1912
” em pt Yoaers-ctalcte eal 7 | Aug. 20,1912
Bi | ccosea wee le a ae TA eae Casea dane viedo eee] OXORTSE. PAM glee et |. ee acl cami pe (els Ge eae Senet 1o| Aug. 8, 1911
9 |. Ma ViLeopard frogs. esc necisecce 10 Do.
10 ie a) ouaall crayfish oc) ec 10 Do.
11 .| Small crayfish, shiner, ro | Aug. 15,1911
12 and frog.
3% inch crayfish......... to | Aug. 21,1911
TSU lass apetelstoreiell| reteieetarae Good-sized crayfish... ....| 1o | Aug. 25,1911

The small-mouthed black bass is found over a large part of the lake, but anglers
seek it along the lakeward margin of the patches of aquatic vegetation, where it is most
abundant. Our data show that in midsummer it ranges to a depth of 45 feet, that is, to

FISHES OF DOUGLAS LAKE, MICHIGAN, 235

the thermocline. In this it agrees with the pike, the perch, and the sucker. It may go
deeper at other seasons of the year, as the sucker appears to do.

Six of the specimens in our table contained food, and we have records of two others.
Of these eight, six had eaten crayfish with or without other food; four had eaten only
crayfish; one had eaten crayfish in addition to a fish; one, crayfish in addition to fish
and a leopard frog. The seventh had eaten a leopard frog only. The eighth contained
numerous specimens of the large cladoceran, Leptodora hyalina. ‘These last could not be
examined with the lens, but the naked eye left practically no doubt as to their identity.
With the exception of no. 6, the fish whose stomachs contained either frogs or fish (no.
g and 11) had been taken on hooks baited with these but not with crayfish. If we
exclude the frogs and fish in no. 9 and 11, there remain seven small-mouthed bass, six of
which had eaten crayfish, accompanied in one case by fish, while one had eaten only
Leptodora. ‘The crayfish were not identified, but that found in the bass taken at 45 feet
was probably Cambarus virilis, which has been several times brought up in the gill net
from that depth, and is the only species known to occur there. There can be no
doubt that in midsummer the crayfish is the chief constituent of the food of the small-
mouthed bass in Douglas Lake. Forbes and Richardson (1908) call attention to the fact
that little is known of the food of this species. They examined the stomachs of three
individuals and found their contents to consist ‘‘wholly of fishes and crayfishes,
approximately a third of the first and two-thirds of the second.”

The size and the weight of the individuals taken in Douglas Lake indicate that
conditions there are favorable to this species. The clear water and sand bottom are
well-known characteristics of its preferred habitat; the pebble-strewn shallows, especially
those about the north end of Fairy Island, afford it ideal breeding grounds (Lydell,
1903; Reighard, 1906).

The young of this species, together with those of the large-mouthed black bass, are
often seen on the sand shoals in late summer in pursuit of the mixed schools of young
suckers, perch, and spot-tailed minnows, which are common there. ‘These they drive
toward shore into shallow water. The bass, which are then 2 or 3 inches long, are two
or three times as deep bodied as their intended victims, so that they are unable to
follow them into very shallow water.

MICROPTERUS SALMOIDES (Lacépéde), large-mouthed black bass, is not common in
Douglas Lake. It occurs in North Fishtail Bay, where the bottom is mucky in places
and the vegetation abundant, and is less common in other localities. We have records
of two taken in South Fishtail Bay. One measured 12 inches and weighed 30 ounces;
the other measured 14.4 inches and weighed 40 ounces. One contained a perch and the
other two fish and a crayfish.

Fishermen tell of catching a peculiar bass on the ‘‘middle ground”’ to the east of
the north end of Fairy Island. The writer has never seen it. It is probably the large-
mouthed bass.

The conditions in Douglas Lake can not be regarded as favorable to this species.
There is little mucky bottom in the shallow water, and few places afford the thick growth
of aquatic plants that the small-mouthed bass prefer for its breeding ground (Lydell,
1903; Reighard, 1906).

PERCA FLAVESCENS (Mitchell), yellow perch, is one of the most abundant fishes in
the lake. Our data concerning it are given in table vu.

236 BULLETIN OF THE BUREAU OF FISHERIES.

TABLE VIII.—ReEcoRDS OF PERCA FLAVESCENS TAKEN IN DouGLaAs LAKE.

fs Local-
No. Weight. | Length. Sex. bere! Apparatus. Stomach contents. ity in Date.
‘ Fig. 1.
Ounces. | Inches.
I-71 dos G4.4 | 14 male 57 fe- 2 insects, 1 copepod, in 6 4| Aug. 6,1912
1 fish examined.

72-77 a. 75 a 4.6 Empty.. a nebieneraaee 4| Aug. 7,1912
3 1.2 Small perch.. Biapated crater etait hctate 4| Aug. 8,1912
hs 6 Rae olectol sigh heated ancnco nese 4 Do.

a, a4 .| Of 20, 14 empty, 6 with 4 Do.
8.4 6 | Aug. 16,1912
WORE Brlianmanooscoca |. lero ilKooort @ooneonncllancas 6 | Aug. 17,1912
8 6 Do.
7 6 Do.
7-6 6 Do.
7-6 6 Do.
8.4 6 Do.
8 6 Do.
8 6 | Aug. 18,1912
Bec eedOw en nanereses| © 43 eee GO. cee cnacfasee GO.. oes 6 Do.
7-6 7 | Aug. 20,1912
8.4 7 Do.
8 7 Do.
8 4 Do.
elalaye fais ict aie 6| Aug. 3,1912
5-8 5 | Sept. 5,1911
5-8 5 Do.
5 5 Do.
4-25 |. 5 5 Do.
4:25 1 crayfish, x dragon fly, 5 Do.
nymph.
4:25 .| Small fish... 5 Do.
4.25 Empty.... 5 Do.
FEE Gan bocacupecnataerony accncoso nd obonde rt aaesd basa doris: Fic. 5 Do.
4.6 Piece earthwor: 5 Do.
5215) 1 with fish, 2 with insect to | Sept. 9,1911
5 rempty, 1 with insects Io Do.
7 Fish muscle. . to | Sept. 14, 1911
725 Not recognizabl 10 Do.
7°75 d 10 Do.
7 1o | Sept. 15,1911
7.25 10 | Sept. 13,4911
7675 10 Do.
8 3 | Sept. 16,1911
9.25 do 3 Do.
7 1 2-inch crayfish. : 3 Do.
7 IM Dt yn teen ee bie 3 Do.

@ Average.

Perch range in midsummer from a depth of a few inches to about 45 feet. They
have not been found below the thermocline even in September. They occur in all parts
of the lake.

Perch of about 2 inches long are common on the shoals in misdummer, but have not
been found elsewhere. In our collections made with the hook from the aquatic vegeta-
tion there are no perch more than 6 inches long (no. 237 to 266, table vit). Larger perch
do not appear to occur there, else we should have taken them among the 29 fish from the
vegetation. For the 221 perch taken in the fyke net in shallow water we have recorded
for most of them only the average length in each lot. There is one fish slightly in
excess of 7 inches. On the other hand, in deep water where we have used a gill net of
1-inch square mesh we have taken no perch under 7 inches and many of 8 inches and
more (no. 223 to 235 and no. 267 to 276). As perch of 4 or 5 inches have a depth of
an inch or more, a net of inch mesh should have taken them if they were present.
The average of the fish taken with the hook in vegetation is 4.6 inches (no. 237 to 245),
while the average of the 25 taken in the gill net is 7.7 inches. It appears, then, that in

FISHES OF DOUGLAS LAKE, MICHIGAN 237

midsummer perch of less than 2.26 inches are found on the shoals, those between 4 and 6
inches in the vegetation, and those of 7 inches or more in deeper water.

Nineteen of the perch in our table showed recognizable stomach contents. Of
these, 11 contained insects only, 1 contained insects together with a crayfish, 3 contained
crayfish only, and 4 contained fish only. The relative importance of the three kinds of
food is perhaps indicated by the frequency of the occurrence of each, which is the ratio:
Insects 3, fish 1, crayfish 1.

The largest perch in our record is 8.4 inches. They appear to be exceptionally
slender, for while in systematic descriptions the depth is given as contained 3.3 to 3.8
times in the length, the depth in Douglas Lake specimens is contained about 4.3 times in
the length. Forbes and Richardson (1908) say “‘the species may reach the length of a
foot and a weight of more than 2 pounds, but does not commonly weigh much more than
a pound.” The heaviest Douglas Lake specimen weighed but 7 ounces. There are no
data from other lakes for comparison with those from Douglas Lake, but Douglas Lake
specimens strike one as being slender, short, and under weight.

After storms perch are found dead on the beach in great numbers. Protruding
from the mouth of such a one is often seen the head of another that has been swallowed
tail first. When pulled out the swallowed perch is, in many cases, found to be almost as
large as that from which it was drawn, and is without doubt the cause of its death.
Cannibalism of this sort is common in the writer’s experience among young wall-eyed
pike kept in confinement and insufficiently fed. It indicates starvation. The small
size of the Douglas Lake perch, the high midsummer mortality, and the occurrence of
the sort of cannibalism described, indicate that the conditions in the lake are not the
most favorable for perch.

In August and September schools of young perch are conspicuous on the sand shoals
of South Fishtail Bay and doubtless on the other shoals. One of these was seined, and
of the 475 individuals captured 203 were found to be young perch from 1.85 to 2.25
inches long. The remainder of the school was made up of 267 spot-tailed minnows and
5 suckers of about the same length as the perch. The young perch are readily distin-
guished in the water from the other species in the school by the seven bars on the sides,
which are more pronounced than is usual with larger perch. They have also a well-
defined black, basal spot at the caudal margin of the first dorsal, which has a more pro-
nounced black border than in the adult.

Placed in an aquarium, the young fish may be seen to feed on plankton. When
the fish are placed in formalin immediately after capture, twisted cords of brown fecal
matter are soon found hanging from the vent. These have on the surface a smooth
coherent pellicle consisting apparently of granular mucus. Under the needle the cord
breaks readily into ovoid masses of equal size each of which is seen to consist, within the
pellicle, of many tests of microcrustacea. The most numerous individuals were those
of a Chydorus-like cladoceran, but Simocephalus and Daphnia were also present in num-
bers and there were a few copepods. These fecal cords varied in length from % inch to
54 inch. By counting the number of tests in one of the 40 ovoid masses into which a 5-
inch piece broke, the number in the whole piece was estimated at 800. There were no
remains of insect larva. The 203 young perch had thus consumed about 162,400 Crus-
tacea, whose remains were found in the fecal cord. The stomach and intestine of one

19371°—vol 33—15——16

238 BULLETIN OF THE BUREAU OF FISHERIES.

individual opened was estimated to contain about four times as many Crustacea as the
fecal cord, whose contents therefore represent one-fifth of the food contained in the
alimentary canal at the time of capture. The total number of Crustacea in the ali-
mentary canals of the 203 young perch at the time of capture was therefore in the neigh-
borhood of 812,000. This is not the daily consumption, but represents rather the food
taken within a comparatively short time. If one should collect the fecal cords passed
by a known number of perch under normal conditions in unit time, there might be
obtained a measure of the total daily consumption of microcrustacea per individual.
The rate at which food consumption increases and the relation of this increase to the
rate of growth might also be thus determined. The whole branchial apparatus of the
young perch with its slender, close-set gill rakers forms an excellent instrument for the
capture of microcrustacea.

By an examination of the gonads, the sex of 237 of the perch included in our table
was determined. Thirty-six were males and 201 females, a ratio of about 1 to 6. In
two instances our records (table viii, no. 1 to 71 and 80 to 221) show a considerable
number of perch taken at one time and place. These two lots include 36 males and 177
females, a ratio of about 1 to 5. The remaining captures consist of from one to three fish,
and in each case in which the sex is recorded the fish are females. The record shows
that all the fish under consideration were taken with the gill and fyke nets. It does not
show the relative size of males and females. If, as is possible, the males are smaller
than the females, the smaller males may pass through the meshes of the nets and the
sex ratio found in our collections may be due to the selective action of the nets used.
One must be sure that the apparatus does not act selectively before a positive state-
ment can be made as to the sex ratio.

The young perch found on the shoals secure an abundance of food, while the shallow
water affords them a refuge from their enemies. When they have reached a length
somewhere between 214 and 4 inches they seek larger food among the aquatic vegetation,
which at the same time affords them a certain protection from their enemies. Here
they remain until they are approaching 6 inches in length. They are then able to
leave the aquatic vegetation and wander into deeper water, for their size affords them
some protection from pike and bass. They now descend as far as the thermocline and
make forays into shallow water beyond the vegetation.

PERCINA CAPRODES (Rafinesque), Jog perch—Mr. H. V. Heimburger, a research
worker at the biological station, reported that he had seen several individuals of this
species in water about 3 feet deep at the edge of the vegetation in North Fishtail Bay.
In August we have seined them in South Fishtail Bay on the sand bottom near vegeta-
tion in 3 feet of water, but they have not been taken in any of the other apparatus
used by us.

We have no records of the stomach contents. Forbes and Richardson (1908) say:
“A third of the food of 11 specimens was found by us to consist of crustaceans (mainly
Entomostraca) and the remainder of insects, the latter chiefly Chironomus larve,
larve of day flies and water bugs (Corixa).’’ ‘Their habitat, as judged by the food, is
probably the bottom in or near patches of vegetation.

Between June 28 and July 8, 1912, between 100 and 200 individuals were breeding
on the sand shoals at the south end of South Fishtail Bay in water from a few inches to

FISHES OF DOUGLAS LAKE, MICHIGAN. 239

2 feet deep. The eggs were laid in the sand and were fed upon by the log perch them-
selves as well as by suckers. A preliminary note on the breeding habits has been
published (Reighard, 1913). A detailed account will be published later.

BOLEOSOMA NIGRUM (Rafinesque), johnny darter, is common in Douglas Lake. Its
distribution in the lake and the factors controlling it are the subjects of a paper by Mr.
H. V. Heimburger, now in manuscript (see also Heimburger, 1913), from which the
following statements are extracted:

It occurs commonly in water of about 18 inches depth, on bottom which contains
some muck and in the neighborhood of aquatic plants. The food of the adult consists,
in midsummer, chiefly of larvee of midges, but partly of Entomostraca. The midges
lay their eggs on the floating leaves of aquatic plants and their larve are found in muck
near by. It is well known that the eggs of this species are laid under stones, sticks,
mussel shells, and the like in shallow water and are guarded by the male fish. Mr.
Heimburger concludes that the localities in which the johnny darter occurs in Douglas
Lake present ‘‘several features in common: (1) Mussel shells, sticks or st ones are found
on the bottom; (2) quiet water protected from prevailing winds so that the bottom is
not subject to violent wave action; (3) a thin deposit of muck in the shallows, with
patches of clear sand exposed; deeper muck deposit is found in the deeper water of the
locality, but a deep muck deposit is not found in the shallow water; (4) masses of Pota-
mogeton, etc., are found near the habitats where both adults and young are found.
These factors are seen to be related very definitely to the food or breeding habits of
Boleosoma and may therefore be regarded as factors determining the local distribution
of this species.”’

ETHEOSTOMA I0W4 Jordan and Meek is rare in Douglas Lake. Two specimens
were taken in the minnow seine along with 75 johnny darters at the west side of the
entrance to North Fishtail Bay in September, 1911. One was taken on sand bottom
at the mouth of Bessie Creek. The species is recorded by Hankinson (1908) from
Walnut Lake. The two Michigan records appear to mark the eastern limit of its range,
which is the Mississippi Valley and northward.

CoTTUS ICTALOPS (Rafinesque), muller’s thumb.—This species is rare in Douglas
Lake, and our records indicate that it is confined to the stony shoals. Its eggs are known
to be laid under stones and other objects lying on the bottom and to be guarded by the
male fish. We have found it in the lake only in localities which afford it nesting sites.
These are Grapevine Point and the west side of North Fishtail Bay. In the latter local-
ity it is recorded as abundant under stones. We have taken it also in Maple River.

LOTA MACULOSA (Le Sueur), burbot.—We have taken a young individual about 2%
inches long at the mouth of Carp Creek, where it enters Burts Lake. It was taken with
Umbra limi, near dense masses of aquatic vegetation. Prof. N. H. Stewart is reported
to have taken two very large specimens in a gill net in Douglas Lake in the summer of
1910.

FISH COMMUNITIES OF DOUGLAS LAKE.

In the second part of this paper there is suggested a classification of the fish habitats
of the lake. The fish found within each of these habitats in midsummer might now be
considered without reference to other habitats. But since fish may pass from one
habitat to another with increase of size and change of food, it seems best to consider

240 BULLETIN OF THE BUREAU OF FISHERIES.

communities of fish primarily with reference to their origin and interrelations. By
community is here meant no more than a group of species or individuals whose mem-
bers are regularly found together for a longer or shorter time. No precise use of the
concept in its relation to habitat is attempted. ‘Thus, the stony-shoal habitat har-
bors two communities, one of which, that of the young fishes, is more characteristic
of sandy shoals. To attempt greater accuracy at the present time would be mislead-
ing. Although it seems best to treat communities of fish rather than the fish of a
habitat, it is most convenient to use for most of these communities the names of the
habitats in which they are best developed.

THE COMMUNITY OF YOUNG FISHES,

On the sand shoals are found mixed schools of young perch, spot-tailed minnows,
and suckers, all about 2 inches long in late summer. These schools doubtless vary in
the proportion of their constituents. Our only record gives perch 203, spot-tailed
minnows 237, suckers 5; but the suckers are often seen to make up a larger per-
centage of the whole. On the stony shoals there are added to these schools many
young blunt-nosed minnows, distinguishable from the other fish of the school by the
jerky mode of swimming and by other field characters already described. What follows
refers to the schools on the sand shoals, but would presumably answer as well for those
on the stony shoals. Perch are known to spawn in the spring. In the hatchery ponds
at Mill Creek, Mich., Mr. Dwight Lydell reports to me that they hang their purse-like
masses of eggs on aquatic plants, brush, and the like. Abbott (1878) found them
spawning on gravel amongst vegetation. HH. M. Smith (1907) gives a similar account,
and also finds the egg masses floating. I have found the floating egg masses in Saginaw
Bay, Mich. Their spawning grounds in Douglas Lake are unknown, but from them
the young perch must make their way to the shoals. Suckers (see under that species)
are not known to spawn except in running water on gravel bottom. Presumably
those of Douglas Lake spawn in such situations in Maple River and Bessie Creek, and
the young fish make their way along shore to the shoals. The breeding habits of the
spot-tailed minnow are unknown. From two or perhaps three different breeding
grounds the young fish forming the communities under discussion reach the sand shoals
and finally live together there in schools through the summer. ‘The schools move
about over the sand shoals, engaged now in feeding, now in fleeing from enemies. If
the water is quiet, they approach the shore; as it becomes rougher they pass toward
the lakeward edge of the shoal and are lost to sight. They may then enter the vegeta-
tion, but as this harbors many enemies it is more probable that they avoid it. All
three species feed on the microcrustacea of the plankton, which at this season are
extremely abundant. The total daily consumption of the whole school referred to
above must reach many millions. Since the alimentary canals of the young fish are at
all times crammed, there appears to be enough food for all, so that the fish are not in
competition with each other.

While the schools of young fishes are on the sand shoals they are often pursued by
young black bass about 3 inches long, small-mouthed and large-mouthed. The water
on the shoals is not obstructed by vegetation, so that the stalking bass are plainly
visible to their prey which attempt to escape the rushes of the bass by short rapid flights,

FISHES OF DOUGLAS LAKE, MICHIGAN. 241

that may end in leaps into the air. Often the flight is toward shore into water so
shallow that the deep-bodied bass are unable to follow.

Two influences seem to keep the young fish on the sand shoals. The first is
abundance of plankton Crustacea, which at this time make up the whole food of the
three species. Since these Crustacea no doubt have a uniform horizontal distribution,
as in other lakes, their abundance merely permits the shoals to be used; it does not
restrict the fish to them. The second influence, that which determines the use of the
shoals, is the relative freedom that they afford from attacks of enemies. This freedom
afforded by the mere presence of the fish on the shoals is increased by their habit of
schooling. The coming together of three or four species to form a large school lessens
the loss to each species and the risk to each individual. This follows in part from
the fact that, if the toll taken by enemies remains constant, a large school loses a
smaller percentage of its constituents than a small school. It follows also because in
a large school there are more eyes on the watch for enemies, and therefore more chance
that any individual will be warned by the flight movements of comrades, and thus be
enabled to escape. Both the habit of schooling and that of frequenting the shoals are
measures of protection, most effective when coincident, for the habit of schooling is
most effective in open water, which affords no lurking place for the enemy, and in shal-
low water in which an enemy’s approach is limited to the lakeward side. It would
probably be ineffective or harmful in vegetation. That the occurrence of young fish
on the shoals may be further influenced by temperature is indicated by the observa-
tions of Michael (Hankinson, 1908, p. 202-204).

By early September the young perch have reached a length of somewhat more than
2 inches. The suckers and spot-tailed minnows are of the same size. A year later the
young perch have reached a length of about 4 inches, have changed their food habits,
and are found in the vegetation. They have left the shoals and the community of young
fishes and have now become for a time a part of the vegetation community. They
do not return to the shoals, but their place there is taken by young fish of the year.
The migration from the shoals probably takes place in late fall or early winter, so soon
as low temperature or ice makes them uninhabitable, and when the perch are prob-
ably less than 3 inches long. The spot-tailed minnows also pass from the shoals into
the vegetation and are thenceforth a part of its community. The young suckers, too,
leave the shoals, and our next knowledge of them is when they are about 7 inches long.
They are then ranging far beyond the vegetation into deeper water. Whether they are
confined to the vegetation when between 2 or 3 and 7 inches in length we do not
know. It is clear that the communities of young fishes are temporary.

The community of young fishes is found on the stony shoals as well as on the
sand shoals, but on the stony shoals there is added to it the blunt-nosed minnow. The
adult males of this species burrow beneath stones or other objects on the bottom, and
thus form nests (Hankinson, 1908). The eggs are attached to the lower sldes of the
stones. The young fish are added to the community last described. Nothing is known
of their food in Douglas Lake. It is presumably plankton Crustacea, the same as
that of the other young fishes of the community, and they are doubtless held to the
community by the same factors. The adults have been found only on or near stony
shoals.

242 BULLETIN OF THE BUREAU OF FISHERIES.
THE STONY-SHOAL COMMUNITY.

It has been pointed out that the presence of stones adds to the community of
young fishes found on the stony shoals the blunt-nosed minnow. ‘This form when
breeding may be regarded as a member of the stony-shoal community. This commu-
nity would then consist of three fishes which lay their eggs beneath stones—the blunt-
nosed minnow, the miller’s thumb, and the johnny darter. ‘The miller’s thumb makes
excavations beneath stones and attaches its eggs in a mass to their lower sides (Han-
kinson, 1908). This has not been observed in Douglas Lake, but the fish themselves
have been found beneath stones on the stony shoals and nowhere else. The johnny
darter also lays its eggs on the lower sides of stones. It has rarely been found in Douglas
Lake or elsewhere than on stony shoals or adjacent to them.

With the three fishes which form this community is found a small crayfish, Cambarus
propinquus Girard. It forms‘burrows beneath stones, and these may be recognized by
the little piles of fresh sand at their mouths.

The stony-shoal community, unlike that of young fishes, is composed of both
young fish and adults and is probably permanent. What happens to its members
when the shoals are ice covered we do not know.

The food of the young johnny darter is known to consist of Entomostraca, while
that of the adult is chiefly midge larvee with an admixture of Entomostraca. (Heim-
burger, 1913.) The blunt-nosed minnow takes a varied food, which, according to
Hankinson (1908), consists chiefly of small organisms taken from the bottom, from
water plants, and from the water. It appears that midges in various stages of develop-
ment formed the chief food of this species in April and May. Besides midges, filamen-
tous alge, desmids, entomostracans, and in one case beetles, were found in the stomach.
The food of the miller’s thumb in Illinois was found to contain about 25 per cent of
small fishes, 4o per cent aquatic larve, and the rest mostly Crustacea (Asellus).
(Forbes and Richardson, 1908.) In Douglas Lake the blunt-nosed minnow and the
johnny darter take similar food and find it in the same place, yet the blunt-nosed min-
now has a somewhat larger choice, and the two forms are not altogether in competition.
The miller’s thumb presumably takes its supply of fish from among the blunt-nosed
minnows and johnny darters, while the crayfish afford it Crustacea.

THE VEGETATION COMMUNITY.

The vegetation community includes the following forms:

Ameiurus nebulosus, bullhead.

Notropis hudsonius, spot-tailed minnow.
Notropis cornutus, common shiner.

Percopsis guttatus, trout perch.

Ambloplites rupestris, rock bass.

Lepomis pallidus, bluegill.

Eupomotis gibbosus, common sunfish,
Micropterus salmoides, large-mouthed black bass.
Perca flavescens, yellow perch, 4 to 6 inches long.
Percina caprodes, log perch.

Of these 10 species 8 have been found only in the vegetation, although, as pointed out
elsewhere, some of them enter other habitats at the breeding season or when young. ‘The

FISHES OF DOUGLAS LAKE, MICHIGAN. 243

inclusion of Percopsis guttatus and Percina caprodes in the list is provisional. We have
found insect remains in the stomach of the former, while Forbes and Richardson (1908) find
the latter to feed on entomostracans and insects, and we found it always on the bottom
near vegetation. What we know of the food of both forms, therefore, suggests the
vegetation habitat. The small-mouthed bass, the sucker, and the pike are found in
the vegetation habitat, but chiefly about its edges, while they are characteristic of the
deep-water community in which we have placed them. The whereabouts of the young
of these forms is unknown to us, but there is little doubt that they will be found in the
vegetation. Here also we are likely to find the young of Lota maculosa.

With the possible exception of the log perch and the trout perch, the forms listed
above are confined to the vegetation. We do not ordinarily find them elsewhere.
They invade the shoals only, so far as known, when they afford them shelter similar
to that to be had amongst vegetation. From the laboratory on South Fishtail Bay a
dock consisting of planks supported on spiles extends from the shore to the seaward
edge of the terrace, which is here an unprotected sand shoal. Beneath this numerous
fishes found shelter—small black bass, both large and small mouthed; bluegills, rock
bass, adult spot-tailed minnows, and small perch. To this retreat they had been able
to make their way readily from the adjacent vegetation, and from it they were able to
harry the schools of young fishes on the shoals. Such shelters were no doubt formerly
afforded by the trunks and branches of fallen trees. With the destruction of the forests,
fallen trees have become rare along the shore and are removed by those who use the
shore as a roadway. When they were more abundant, fish must have been more plenti-
ful on the shoals, for they were then able to find there the shelter conditions of the
vegetation.

Short forays into deeper water in search of food are probably made by several of
the fishes of the vegetation community. By the extension of these the 6-inch perch no
doubt attain in time to a membership in the deep-water community. The breeding
season finds certain of the vegetation fishes outside of the vegetation. The log perch
then betakes itself to the sand shoals and is there closely associated with the adult
suckers, which come to the shoals to feed on its eggs. The common shiner and the
small-mouthed bass must also leave the vegetation to breed.

That the larger small-mouthed bass and the pike make raids from deeper water into
the vegetation in pursuit of prey we can not doubt, for they are captured at its borders.
Moreover, in the stomachs of the pike are found 4-inch perch, which are not known to
occur outside the vegetation.

Three factors suggest themselves as determining the make-up of the vegetation
community—food, shelter from enemies, and breeding habits. All 10 forms, of course,
find their food within the vegetation. Three members of the community—the bullhead,
the rock bass, and the large-mouthed bass—reach a considerable size, feed in part on fishes,
and might obtain these outside the vegetation. But, of these, the bullhead is largely
nocturnal or crepuscular and is slow of movement. It tends to lie in wait for its prey
rather than to seek it actively. The conditions within the vegetation zone are favor-
able for this method of getting food. The rock bass and the large-mouthed bass
include a considerable proportion of small fishes in their diet, and these are most
abundant in the vegetation. The remaining forms feed on invertebrates, largely

244 BULLETIN OF THE BUREAU OF FISHERIES.

insects, and find these most plentiful within the vegetation areas. Here the common
sunfish finds snails and the bluegill plant tissues. The two species of Notropis might
feed on plankton outside the vegetation, while the bullhead, rock bass, and large-
mouthed bass might, like the pike and small-mouthed bass, obtain fish without restrict-
ing themselves to the vegetation. Food alone does not appear to limit these forms to
the vegetation.

In addition to food, the seven smaller forms (excluding A meturus, Ambloplites, and
Micropterus) find within the vegetation a refuge from certain enemies, chiefly pike and
small-mouthed bass. Within its mazes rapid flight is much easier for a small form
than rapid pursuit by a larger enemy.

The breeding habits of one of the forms in our list (the spot-tailed minnow) is
unknown. ‘The writer is familiar with the breeding habits of the remaining forms, and
the papers of Hankinson (1908) and Shelford (1911) contain references to some of them.
Two species (the log perch and the common shiner) breed on plant-free sand or gravel.
A third, Percopsis, according to the unpublished observations of Doctors A. H. Wright
and A. A. Allen, of Cornell University, to which reference is made by their permis-
sion, breeds “in swiftly running water and over a stony bottom.” The common
sunfish also may breed on gravel bottom. Four species (bullhead, rock bass, bluegill,
and large-mouthed bass) lay their eggs, by preference, in nests made by exposing the
fibrous roots of water plants. The bottom material removed in making the nests may
be sand, marl, muck, or mud, but the essential of the nests is the rootlets which form its
bottom. The eggs of the sunfish, rock bass, and bluegill adhere to the rootlets. Those
of the bullhead lie free, united in masses. The bullhead, since its eggs are not adhesive,
may make its nest in situations in which rootlets do not occur, but in the writer’s expe-
rience these nests are always in vegetation. The common sunfish also makes a nest, the
bottom of which is sometimes gravel but more often rootlets. The perch, too, commonly
deposits its masses of eggs amongst vegetation, although when adult it wanders far
beyond the vegetation. Thus there are 6 of the 10 forms that require vegetation as
a part of their breeding environment (A meziurus, usually; A mbloplites; Lepomis, usually;
Eupomotis, usually; M. salmoides; Perca). In some of these forms the relation of the
breeding habits to vegetation is not at first apparent. Thus Lepomis pallidus is often
found breeding where there seems to be no vegetation, but in such cases closer inspec-
tion will show that its eggs are laid on the rootlets of bulrushes which are so near the
denser vegetation as to afford the fish a ready retreat into it. The same is often true of
the large-mouthed bass and the common sunfish.

Of the 10 forms included in our vegetation community we may repeat that all
obtain food in the vegetation., Although some of them might obtain it elsewhere it is
doubtful whether they could obtain it as well. At least four of these probably need
the vegetation for protection (the two species of Notropis, Percopsis guttatus,and Percina
caprodes). ‘This need is scarcely less in the bluegill, the sunfish, and the perch, and in
the young of the other forms. The two species of Notropis are not bottom forms.
While they might obtain plankton food outside the vegetation, they would be very
conspicuous there and could scarcely long escape their enemies. Percopsis and Percina
are bottom forms and would more readily escape detection outside the vegetation.
It is possible that they may wander to some distance from it. The relation of these
10 forms to the vegetation is summarized in table rx.

FISHES OF DOUGLAS LAKE, MICHIGAN. 245

The cross in the table indicates that the species avails itself of the factor indicated;
the zero indicates that it does not, or, in the case of the column headed ‘‘ Protection,”’
that it does not need to. Adult Ameiurus, Ambloplites, and Micropterus are larger
than the perch which exist outside the vegetation. They must be held to it by some
other factor than the need of protection. This may be food or some factor not included
in our table. Since some of the species may obtain their food outside the vegetation,
the data in our table indicate that no one of the factors considered necessitates the occurrence
together of the species which make up the vegetation community.

TABLE IX.—SHOWING THE RELATION OF THE SPECIES OF FISHES INCLUDED IN THE VEGETATION
COMMUNITY TO THE THREE Factors OF Foop, PROTECTION, AND BREEDING AFFORDED BY THE
VEGETATION.

Species. | Food. Protec Breeding. Species. Food. Pos | Breeding.
| | | 2 |
Ameiurus nebulosus.....-....) r< ° > Lepomis pallidus............ x | x XxX
Notropis hudsonius. . . eel x | x (2) || Eupomotis gibposus....... x | x< x
Notropis cornutus..... | oe x ° || Micropterus salmoides . are x ° x
Percopsis guttatus.... x x ° || Perca flavescens, 4-6 inches x 4 x
Ambloplites rupestris......... x ° x | Percina caprodes...:....-..... | x DG °

@ Unknown.
THE DEEP-WATER COMMUNITY.

In the deeper water near the bottom, down to about 45 feet, the following forms
have been taken in midsummer:

Catostomus commersonii (common sucker), from 7 to 12 inches long, approximately.
Esox lucius (pike), between 12 and 30 inches long.

Micropterus dolomieu (small-mouthed black bass), between 12 and 16 inches long.
Perca flavescens (yellow perch), 814 inches long.

Lota maculosa (burbot), large.

All these fishes wander far from the vegetation. The small-mouthed black bass
feeds chiefly on crayfish. Presumably the larger pike prey on the other fishes of the
community, but our records do not show that any pike had eaten fishes more than 4%
inches long. These were perch of a size found only among vegetation and were doubt-
less taken by the pike at its borders. It seems probable that the five fishes of this com-
munity are protected from one another in part by their size, for the individuals of each
kind are usually too large to be eaten by the others except by the largest pike or burbot.
The perch remain in the aquatic vegetation until they are about 6 inches long and large
enough to enter the deep-water community.

The fishes of the deep-water community, except possibly the burbot, are much
about the borders of the patches of vegetation and more or less within these patches.
Here the pike (no. 1, 2, 3, 5, 6, 15, 16, table 111) obtains the smaller perch and probably
the other fishes of the plant zone. To a lesser degree, the small-mouthed black bass
may obtain fish from the same source; at any rate it is commonly taken on the lake-
ward margin of the plant zone on hooks baited with shiners or spot-tailed minnows.
The sucker is also known to enter the patches of aquatic plants. The characteristic
feature of the deep-water community is then that its members occur near the bottom in
deeper water outside the patches of aquatic plants, not that they may not also occur

246 BULLETIN OF THE BUREAU OF FISHERIES.

within these patches. In this they differ from the rock bass, the small-mouthed bass, and
the bullhead, which have been taken only in or near vegetation and in shallow water,
although their size would apparently enable them to enter the deep-water community.

In addition to the five species listed, one other, the cisco, or lake herring, must
occur at some level considerably above the bottom. This may be inferred from the
fact that it has never been taken by us in fine-meshed gill nets set on the bottom in
either shallow or deep water. It is probable that it should be regarded as a member of
a nekton community characteristic of a mid-water habitat. No other fishes are known
to be associated with it.

RELATION OF DOUGLAS LAKE SPECIES TO THOSE OF OTHER WATERS.

Douglas Lake has large areas of bare sand or gravel bottom, comparatively clear
water, kept well agitated by the wind, and a relatively sparse growth of vegetation,
It would be of interest to learn: (1) Whether its fishes give preference in other regions
to the conditions that they find in Douglas Lake, and (2) whether their distribution
over the continent is such as to afford these conditions. Forbes and Richardson (1908)
give the only data known to me on the habitat preferences of American fresh-water
fishes. For many species they indicate by coefficients or percentages the kind of water
preferred (whether large rivers, small rivers, creeks, lakes, or ponds), the kind of bot-
tom (mud, rock and sand, mud and sand), and the amount of current (swift to mod-
erate, slow to stagnant, variable). I am unable to interpret these data in such a way
as to make them available for a detailed comparison of the habitat preferences of the
fishes of Douglas Lake with those of Illinois, and therefore restrict myself to noting
two points:

There is no mud bottom in Douglas Lake, none at least in its shallower parts.
The bottom is sand or gravel, with an overlying stratum of muck in the deeper water
and in protected situations in shallow water. None of the species occurring in the
lake are among those given by Forbes and Richardson as preferring mud bottom in
other waters, and but two species (Ameiurus nebulosus, and Umbra limi) are commonly
found on such bottom in other waters. The other Douglas Lake species, in so far as
their preferences are indicated for Illinois, are found with greatest frequency on a
bottom which includes rock or sand or both.

Among the fishes in our list the following are found by Forbes and Richardson
to show a preference for small rivers or creeks: Catostomus commersonit, Semotilus
atromaculatus, Notropis cornutus, Ambloplites rupestris, Micropterus dolomieu, Percina
caprodes, Boleosoma nigrum. Suckers, rock bass, and small-mouthed bass occur often
in lakes, but the horned dace and the common shiner are rare in lakes. Forbes and
Richardson give the water preference of the horned dace as large rivers 1.67, creeks
3.77, lowland lakes 0.11. The species is of local occurrence in Douglas Lake and is
possibly introduced. For the common shiner the Illinois preferences are large rivers
0.11, small rivers 2.45, creeks 3.00, lowland lakes 0.02, upland lakes 0.20. The species
is one of the most widely distributed and abundant of the Douglas Lake fishes. Its
abundance, together with the presence of the other species showing preference for small
rivers and creeks, indicates that in the character of its bottom, the movement of its
waters and the sparseness of its vegetation Douglas Lake affords the small-river-creek
conditions preferred by these species.

FISHES OF DOUGLAS LAKE, MICHIGAN. 247

For the purpose of discussing the general distribution of the fishes of Illinois,
Forbes and Richardson have divided the region over which they occur into 12 districts
and have tabulated the distribution of each species in these districts. The number of
Douglas Lake species found within each of these districts is shown in table x, arranged
in numerical order.

TaBLE X.—SHOWING NuMBER OF DouGLaAs LAKE FISHES FouND IN OTHER REGIONS.

GreatiVakesibasiiawstirray te ekerdtts aat ey eee Pee cep neces ciao oreo sitions IEE Sones Danae 22
WippermMississip pijan deMissouriay alleyseaercrrtle stirrer ir eerste, kee secede onnk 21
Wower Mississ(ppican de© nowy el Levsnene open e seiersels tere ot aici Se eciee oe iets sine cle sien ake 20
OuebeciauduNewal mg lar eerste ve vecssteferel tes clea voce easyer ye eletale sins? ol oisrate apes Boers, ae eieeiok cee eed hs 19
North Atlantic; NewsHngiandito|Chesapeake Bayes wce-cneccrc---asies2 esse sts eteiecececcsceeee. 17
SouthrAtlantic, Chesapeake: Bayztovbloridas sveceqsajat styelaleqaeie eloiei-ati-iss.ct bide cis cle ast ks nee cincw ees 13
15 ote R Oya LEKKI noon vasa oddone 0.00 GS RA SN Od OS SENG OOSDO RTD OTOEIEDE SORE Sees Oe aete eine cena 12
Far north, north of Mississippi drainage, between Rocky Mountains and Lake Superior drainage... 12
Basti Gulfidistrict mtOpMississip pid ralnace Oni wWestsacmeci- ieee eeicca aa leheeiiisiorersisisis slevee ec ciee else cae 10
West Gulf district, westward from Mississippi drainage, including Rio Grande.................... 4
Hlorida ge emiristtlaaerarsas ee ee erst era stere Telok ere ieke sisiercia eek ster eke a eee ae Stare elaine tome nl aise MEANS B
BargrlorthivwestamestaLopRoc cys Mountains sae aerctveciletecpsecrayd oe orae ames sietetite elaine iene el arscieie arera anaceee I

It is clear from the table that the Douglas Lake species are northern and north-
eastern rather than southern or southwestern in their range. A single species, the
pike, crosses the Rocky Mountains into the far northwest, but the species is of common
occurrence in the Northern Hemisphere. Three species, the large-mouthed bass, the
bluegill, and the bullhead, occur in Florida, while four species are found in the west
Gulf and Rio Grande region. All the Douglas Lake species, with the exception of the
cisco, occur in the upper Mississippi and Missouri Valleys. In the lower Mississippi and
Ohio Valleys all are found, with the exception of the cisco and Etheostoma iowe. In
Quebec and New England Etheostoma iowe, Notropis cayuga, and the bluegill are lack-
ing. These three, with the cisco and the miller’s thumb, are lacking in the north
Atlantic district, leaving 17 species. This number is reduced to 13 in the south Atlantic
district, 10 in the east Gulf, and 4 in the west Gulf district.

In general, more Douglas Lake species are to be found in clear, rock and sand-
bottomed, northern waters than in the more turbid southern and southwestern waters.
Forbes and Richardson (1908) publish a list of 34 Illinois species that avoid the turbid
waters of the lower IIlinoisan glaciation. Ten of these are also Douglas Lake species.
They give also a list of 37 species that tolerate the lower Illinoisan glaciation. In this
list are but two Douglas Lake species.

The fishes of Douglas Lake appear, then, to give preference in Illinois, and pre-
sumably elsewhere, to those conditions of water and bottom that are available to them
in Douglas Lake and to be distributed over the continent in districts in which such

conditions are found.
SUMMARY.

1. Four fish habitats are provisionally recognized in Douglas Lake—the barren
sand-shoal, the barren stony-shoal, the vegetation, and the deep-water. Each is defined
(pp. 221, 222).

2. Twenty-two species of fish are listed from Douglas Lake, and detailed data on
their occurrence, weight, length, food, and interrelations are tabulated and discussed

(p. 229).

248 BULLETIN OF THE BUREAU OF FISHERIES.

3. Four fish communities are provisionally recognized in Douglas Lake in mid-
summer—the community of young fishes, the stony-shoal community, the vegetation
community, and the deep-water community (p. 239-246).

4. The community of young fishes is characteristic of the shoals. On the sand
shoals it consists of perch (Perca flavescens), spot-tailed minnows (Notropis hudsonius),
and suckers (Catostomus commersonit), all in schools together and all about 2 inches
long in late August. On the stony shoals, young blunt-nosed minnows (Pimephales
notatus) may be added to these schools (p. 241).

5. All the members of the young-fish community feed exclusively on plankton
Crustacea (p. 241).

6. The occurrence of young fish of several species (a) in large schools and (b) on
open shoals are conditions which, when they occur together, favor escape from preda-
tory enemies. It is held to be the presence of such enemies that keeps the members
of the young-fish community together and on the shoals (p. 241).

7. The community of young fishes is temporary. Before their second season its
members forsake the shoals (p. 241).

8. The stony-shoal community consists of the young and adults of three species
which lay their eggs beneath stones—the blunt-nosed minnow (Pimephales notatus), the
johnny darter (Boleosoma nigrum), and the miller’s thumb (Cottus tctalops). With these
is associated a small crayfish (Cambarus propinquus Girard) (p. 242).

g. The factor which holds the members of the stony-shoal community together
is the presence of stones or other similar objects which afford the conditions necessary
for breeding (p. 242).

10. The stony-shoal community is permanent, except as it may be interfered with
by winter conditions (p. 242).

11. The vegetation community consists of 10 species which, with one exception,
are unknown except in or very near vegetation (p. 242).

12. The occurrence together of the members of the vegetation community is not
attributed to a single factor, but to two or more factors, of which food, protection, and
breeding conditions are specified (p. 243).

13. All members of the vegetation community find their food in the vegetation;
in addition seven of them find there probably necessary protection, and six find in con-
nection with vegetation their usual breeding conditions (p. 243).

14. The deep-water community consists of four or five species—the common sucker
(Catostomus commersonit), the pike (Esox /ucius),the small-mouthed black bass (Microp-
terus salmoides), the burbot (Lota maculosa). All of these occur near the bottom in
deep water outside vegetation, although they may also penetrate vegetation and invade
shallow water (p. 245).

15. The members of the deep-water community obtain their food in the deeper
water and about vegetation. Their size is held to enable them to leave the vegetation,
since by it each species is in some degree protected from enemies (p. 245).

16. The species of fishes found in Douglas Lake give preference in Illinois (Forbes
and Richardson) to those conditions of water and bottom that are available to them
in Douglas Lake (p. 246, 247).

17. The fishes of Douglas Lake are distributed over the continent in those districts
in which the conditions available to them in Douglas Lake occur (p. 247).

FISHES OF DOUGLAS LAKE, MICHIGAN. 249

LITERATURE CITED.
Asszorr, C. C.

1878. Notes on some fishes of the Delaware River. Report of the Commissioner of Fish and

Fisheries for 1875-76, Appendix B, p. 825-845. Washington.
Forses, S. A., and RIcHARDSON, R. E.

1908. The fishes of Illinois. Natural History Survey of Illinois, State Laboratory of Natural
History, vol. m1, CXxxI+357 p., with colored plates, 76 text fig., and atlas of 103 maps.
Danville. 7

Hanxinson, T. L.

1908. A biological survey of Walnut Lake, Michigan. A Report of the Biological Survey of the
State of Michigan, published by the State Board of Geological Survey as a part of the
report for 1907, p. 157-288, pl. x1I-LXxv, 6 text fig. Lansing.

HEcat, S.

1913. The relation of weight to length in the smooth dogfish, Mustelus canis. Anatomical Record,

vol. vu, p. 39-41. Philadelphia.
HEIMBURGER, H. V.

1913. The factors that determine the distribution of Baleosoma nigrum in Douglas Lake, Che-

boygan County, Mich. Fifteenth Report Michigan Academy of Science, p. 120. Lansing.
LypeE.L, D.

1903. The habits and culture of the black bass. Bulletin U. $. Fish Commission, vol. xx11, 1902,

p- 39-44, 1 pl. Washington.
REIGHARD, J.

1906. The breeding habits, development, and propagation of the black bass. Sixteenth Biennial
Report, Michigan State Board of Fish Commissioners, appendix, p. 1-73, 2 pl., 12 text
fig.; also as Bulletin No. 7, Michigan Fish Commission. Lansing.

1913. The breeding habits of the log perch (Percina caprodes). Fifteenth Report Michigan Acad-
emy of Science, p. 104-5. Lansing.

SHELFORD, V. E.

1911. Ecological succession: m1. A reconnaissance of its causes in ponds with particular reference
to fish. Bulletin of the Marine Biological Laboratory at Woods Hole, Mass., vol. xxm,
p- 1-38. Lancaster.

Smits, H. M.

1907. The fishes of North Carolina. North Carolina Geological and Economic Survey, vol. n,

453 p., 21 pl., 188 text fig. Raleigh.
Tucker, D. A.

1913. The oxygen content of the waters of Douglas Lake, Michigan. Fifteenth Report Michigan

Academy of Science, p. 121-128. Lansing.

THE POTAMOGETONS IN RELATION TO POND CULTURE

om

By Emmeline Moore

Contribution from the Department of Limnology, Cornell University

CONTENTS:
oad

Page.

err trom 1 Horie net rite Re Pore ake Tetetec tore erelcle = teers e oiasekcCaielolin hove re valajs/ sist sieie|sjeisvcVa'ejeve aieha! Scfeioiaisis(eleie\e7e7s 255
ELTSCOT 1 Call peeetetetetete te ve teteeet eee etee tee Pace Tai ete tcl oteetey eles aveserofevayeVey<ieforaratayaterecovevalerate/ovalapctate ererale\sveleferclarelnieys 255
PE ClESlOMmROLAMOPeLOM VEST SALE ape yepataterels(eleiatesohciol-tclstereitciels s/stelsfolelsiers\«\sielalase\erelele/efeloioieie sre; sierelofs 260
General survey of life conditions of the species investigated. .............. cece cee e cece eee eees 261
SPY ATTI CVU CANIUS Rpts teers otase atch tare craters clo sevevs/sictalsye s cisiayetepele:sfetelejesele cleveyeve(ere/oie+« ereie a cicishese ofsisteleverereis 261

le, Greapo IN FHES, 9 soo seceatacunagndd opbanosoMencocooncaraTen cdopdo CCU OnOUEbeMooOouTOoOnSn 261

12s ND weroy oH HALTS, cc ooauvine dopEMMOoo Ae DAaes ConAS GOO mEnH 6 opUoneoNdoooe sdoanadaoountaocads 262

12, WAVOVELTES, cooccoudaneonn on GUS aod SGODUNE CS Ono Of Onanod COUR DH POU Oooo DDOGOTEOnODD 263

124 GARD ou ocansoddodosoucdneb douDODDUdoUdo DU deo puOnUdoodondcoUEROdonuGCUduBONDOODSS 264
PERZOSUOLULO LUIS a say cteretes feten sce rae Fat ooh eVo eGov ato ert a Toa ts coh a tne Peale eee Sy cate fal okayavewe fate eye veyetaier sy e1siere oceiehs 264

PMO DtusttOliustae face: wes aicieh cutlelo cep sc cies emits Asien Seah Gono vnooRnonanodobUdnoDOOGdC 265

12), 1llboTErET ca sdooke abs dao bp nb osen sé Be ob hoade oon soos coonestoodueao bon daocunbudoooCKbUNS 265

1D, MECHA S an ao ceasecocansocopsoces none COOn Goce onc oDoocauanocadaUuECU DCO DD UOmOOnODOOE 265

IAPR G ba DAnsiil em yarrtet are teye avehe ve fe clexciiis ichatoneke evasensuerel euciere eieratevey sharelsi innate o1iailorelatolese wiavate/o: a) sieye/eie: iss 61 267
Summary gO mC tiiralifeaciInes are reiere scree aeeyotaicicieersysie lteter el ststetene evaertafehevose eels /elarclaievhafoie,« cl olayeieyexeiers 268
Naturaleandtartificialy propaga tom rer pert ctrtperereytetaletensisl loyal foicinr= afeVelevoie steiatoso/<'el ofeteravere cl sieYoierais/eye1= 268
PLOPASALLOMN VALID CTS sey cy vars te teeter retetete ete fepe el atanetata) atefatniar-Veveretsted=yata7elcfotalsveretels efereiseielelerereversiclelerer= 269
Propapationap yattt eros TOOtStOCKS smeterstayr eras icleisiayor=istelsisis =ielstereyele’aletols!cfere’eraleiereveieje s\eyesaVl=io19 “Vel 272
Propagation by, subterranean stems mMOt tuberous, ee \cjre elec a elo eleleicicle.c sieisiele eis eleisieleiele leis © 273
Propagation byawiti tera OMe ere ereyaevare ons systas oleae «tate a\ay ayers (atsteiatstersye otoie alojal+ielatelclateveseleters ielefelers) <r 273
SPrOPAGa CONG Va DING epee terre terete sche snc aerate haar ctetetc) tel atenciershe lo stevens hetoichaiers ele! sicievopetels leper erst essere 274
Propagationibyasirapments Ole Stems cpa -y-sal= erels ehe)=1 acest sie edatelay viaya elele) hele aketejatareieleselese er oiet= asic 277
Propagation ygSeedsmwmp ener st tet rekecs ate <efeters) stele crear epet cya) ott s\sfoseyotosess tol cleter-lel= eiolerayererefolotevelslopetol <i> 278
Production of seeds and vegetative propagative structures .......... eee e eee eee eee eee eee 279
Economiciaspects:ofseotamag etOns a ysiavaatcietersiaieveystel rete ele ere) =sekateleteteral=(ols}evessvalals ole/="efelelohereleieieleloiei sisi ie 281
Species of Potamogeton and the animals foraging upon them................. 0: cess ener eeeee 284
WomChisiomersrriercta es ne erect holeleter eras sletoreeieeeteleroter sr cvaleraicroyeustetel reveyeravekey leper iayetorsietetsde clatetoonsrs iolee choneis 285
shi ay-5 0250) ths bane ad recone ro Ona ah COG 4.5 en cide Deana UUM R Oca ne Drop ic tenn OOo OOOO a aon bard. 286
EXP LANACLOMMOL PLACES | << oysla/a/esyoie/efetelelevelevercioteisraielsiarsietelaisiolelere)clersyetelsse\elalsiefelsieie sfetereielatsisiafece)sisieisiacel 1: 289

19371°—vol 33—15——17 253

THE POTAMOGETONS IN RELATION TO POND CULTURE.

ad

By EMMELINE MOORE,
Contribution from the Department of Limnology, Cornell University.

ad
INTRODUCTION.

The cultivation of lakes, ponds, and streams follows as a natural consequence the
biological investigation of the aquatic life within them. Herbivores and carnivores
live their life in the water, and if we ponder over their means of sustenance we are
struck by the fact that the natural food supply has rarely been augmented by cultural
methods.

“The larger aquatic plants,’”’ says Pond (1903), ‘form a link in the chain of nutri-
tive relations that stretches from the water and soil to the higher fishes.”’ If such is the
importance of these plants, the great mass of vegetation which comes to maturity each
season is a national asset. Yet the annual yield has never been estimated or given a
place in the Government crop reports.

Aquatic plants have contested for possession of the waters much as the grasses
have contended for supremacy on land, until it may be said that the dominant forage
crop of our lakes, ponds, and streams is to be found among the pondweeds, the Pota-
mogetons. Variety in form, adaptability to environment, and diversity in range have
all contributed their share in giving prominence to this group and in furthering a natu-
ral resource whose propagation and control are vital factors in the economic relations
of the life of inland waters.

The object of this investigation is to present such observations and experiments
on the natural and artificial propagation of the Potamogetons as will render cultural
methods economical and practical.

The work herein recorded was carried on at Cornell University under the direction
of Dr. James G. Needham, to whom I wish to express my grateful thanks for help and
suggestion.

HISTORICAL.

The cultivation of aquatic plants was an ancient occupation, one which concerned
itself with the beautification of pools and fountains. In modern times, too, aquatic
plants have been used in variety and profusion in the ornamentation of artificial or
natural ponds. But the cultivation of aquatics from an economic standpoint is a new
idea, so new, in fact, that data regarding it are just beginning to appear in bulletin
form in the Government compilations of scattered and isolated experiments. In the
bulletins plants of the genus Potamogeton have received the larger measure of notice
because observations on the feeding habits of animals associated with them point to
the important réle of these plants in the economy of nature.

255

256 BULLETIN OF THE BUREAU OF FISHERIES.

Further contribution to the present status of the Potamogetons incorporates of
necessity a considerable body of observation pertaining to the systematic, morpho-
logical, and biological aspects of this group, and renders it highly desirable to set forth
the historical background of each of these three phases of the subject.

JOHN GERARDE, 1633.

A beginning in the classification of the Potamogetons was made by the old herbalists, medical
men, who found it necessary to study plants in detail in order to discriminate the kinds employed for
different purposes. The special virtue in Potamogetons, for example, resided in the leaves, which
were applied to reduce inflammation. In the herbal of John Gerarde the group Potamogeton (Pota-
mogeiton in the old spelling and pondweed or water spike in the common parlance of the time) con-
sisted of four species—a broad-leaved pondweed, a narrow-leaved pondweed, a small pondweed, and
a long sharp-leaved pondweed. There was a figure of the entire plant accompanied by the Latin and
English name. Then followed the “description, place, time, names, nature, and virtues agreeing
with the best received opinions.’’ A “fennel-leaved water milfoile’’ illustrated by a figure easily
recognized as our fennel-leaved pondweed, Potamogeton pectinatus, was given a place among the
Myriophyllums. Such was one of the earliest attempts to classify the group.

CHAMISSO AND SCHLECHTENDAL, 1827.

The first important monograph of the Potamogetons was the work of Chamisso and Schlechtendal,
who, in Linnea, volume 2, 1827, systematized the results of scientific observation during the latter
part of the eighteenth and the beginning of the nineteenth centuries. Under the family name of
Alismacez 21 species were described and illustrated by drawings of fruit and leaf, including among
them many of the common and widely distributed species of to-day. Several other Potamogetons
were listed as uncertain in position and difficult to classify, a condition which holds as true to-day as
then, when Chamisso and Schlechtendal struggled to bring order out of chaos in this puzzling group
and recorded this pertinent observation: ‘Species Potamogetonum habitum mutantes in alias sepe
transire videntur, alienaque speciei habitum mentientes scrutatorum irrident,’’ which translated
is, “Species of Potamogeton changing their habit seem often to pass into others, and feigning the habit
of other species baffle research.”’

REICHENBACH, 1845.

Reichenbach’s monograph of the Potamogetons, in his Icones Flores Germanice et Helvetice,
followed in 1845. More intensive in scope than any preceding work, it marked a distinct advance
both in the method of description and in the matter of illustration. Several reproductions, especially
of flower and fruit, which were drawn with great cleamess and accuracy, have found their way in the
latest authoritative works on the subject. In this monograph the author introduced the figure of the
so-called “‘bur,’’ the vegetative propagative body of P. crispus Linneus, though he apparently did
not recognize its significance in the rapid propagation of this species. It is interesting to note that
the figure is inserted without further description or comment. Moreover, it is erroneously drawn, and
the error has been copied time without end.

IrMISCH, 1851.
Thilo Irmisch, in a published note in Flora, 1851, first recognized the presence of tubers on P.
pectinatus.

AGARDH, 1852.
A year later J. C. Agardh, in Verhandlungen der K. Schwedischen Akademie der Wissenschaften,

recorded several observations on the tubers of this species of Potamogeton.

Cros, D., 1856.
D. Clos was the first to publish an account of the origin of the “bur’’ of crispus, though his obser-
vations are incomplete regarding both their development and their germination.

Irmiscu, 1858.

In aremarkable monograph by Irmisch, Uber einige Arten aus der naturlichen Pflanzenfamilie der
Potamogeton, the history of the development of the tuberous growths on P. pectinatus is recorded and
their morphological and anatomical structure described. The author states that, at the end of the vegeta-

POTAMOGETONS IN RELATION TO POND CULTURE. PANG

tive period in the fall, the shoots of recent formation have a singular appearance, the last two thin
internodes bearing tubers at the end. At first the tuberousend resembles a conical terminal shoot or bud
surrounded by scales. Internodes make their appearance and soon become thickened; eventually the
scales split and disclose a tuber of two swollen internodes. Simultaneously a slender bud forms at the
distal end of the tuber, and axial outgrowths develop from the sides that bespeak the shoots of ordinary
branches. These axial shoots in turn develop swollen internodes which follow two thinner ones as in the
preceding case and produce a series of tubers dichasial in form. The excellent series of drawings by
means of which the author depicts the transition from internode to tuber leaves nothing to be desired
in the morphological interpretation of them. They are clearly two modified internodes.

Tuber-bearing shoots grow out of the upper leaf axils also, and follow the usual development of genera-
tions of internodes with leafy shoots, besides the tuber-bearing ones in two or three series. The anatom-
ical structure of the tuber resembles that of the stem excepting that all tissue not fibro-vascular is filled
with starch. The observations on P. obtusifolius are incomplete, but the presence of winter buds is
noted. For P. natans and P. lucens the morphology of the rootstock, stem, and shoot is completely
determined, and the details are clearly shown in the drawings. The method of branching is fundamen-
tally the same in the two species. In brief, the growing tips of the rootstocks branch dichotomously,
giving an erect axis and a horizontal one. Each generation of the developing rootstock brings forth two
horizontal internodes and a bud which is the incipient erect axis. The terminal bud at the end of the
horizontal axis reproduces this condition as long as the plant lives. In the development of the erect
shoot, the scales, usually three in number, grade into stipular sheath, phyllodes, and foliage leaves. A
two-fifths arrangement of leaves is noted and the shoots follow the same order. The winter condition
of P. lucens consists of rootstocks by means of which the plant propagates itself rapidly in the spring.
The internodes of these rootstocks are shorter and thicker than the ordinary ones and are borne in a suc-
cession of three or more with terminal and axillary buds containing the incipient axes of the horizontal
and erect shoots.

Irmisch made observations also on P. crispus, investigating especially the “burs’’ or propagative
shoots, although this work was anticipated in part by D. Clos in his Mode de Propagation particulier au
Potamogeton crispus L. Irmisch, however, found two forms of the bur of crispus, the slender spicular bur
as well as the stout, horny, denticulate one observed by Clos. The former bur he observed growing in
the axils of detached shoots in late autumn and afterwards breaking away from the axils and settling in
the mud. The origin of the latter form he did not observe, but he found it in the muddy bottoms of
ponds in great abundance. These ‘‘burs’’ or modified twigs, as Irmisch sometimes called them, he con-
sidered important examples of propagative structures.

In connection with these plants, Irmisch first pointed out the ‘‘Scheiden-Schiippchen, squamule
intravaginales,’’ scale-like structures developed at the leaf bases, having as a possible function the pro-
duction of slime or mucilage for the protection of young and slender shoots.

This monograph is of special importance in presenting the morphological data of a few species of
Potamogeton. From time to time further contributions have been made to the subject by other investi-
gators in the field, but this still remains the greatest work of its kind.

As a result of these studies on the tubers of P. pectinatus, the rootstocks of P. lucens, and the burs of
P. crispus, Irmisch came to appreciate the advantage of artificial propagation in this group and remarked
in conclusion, in an observation that is prophetic of present day interest, ‘‘That many of the Potamoge-
tons, as well as other aquatic plants, possess in a singular way that possibility of domestication which
has given us the tame animals from the wild ones.’’

RoBBIns, 1867. .

Thus far the work of the Potamogetons was confined principally to European species. In 1867,
however, the American species were reduced to something like a complete intelligible systematic shape
by Dr. G. W. Robbins, whose descriptions, as far as they came within the range, were incorporated in
Gray’s Manual, edition 5. Later descriptions of the western species were published as they became
known.

MORONG, 1893.

The greatest contribution to the literature on the North American species of Potamogeton is by
Thomas Morong in his Naiadacee of North America, a monograph which includes 37 North American
species, 14 of which are confined to this country. Many of these species were studied through succes-

258 BULLETIN OF THE BUREAU OF FISHERIES.

sive seasons of the year and a considerable body of knowledge pertaining to the development of the
plants was accumulated. It is recorded that 17 of the described species are propagated vegetatively by
one or more of the following structures: Rootstocks, tubers, winter buds, and stems.

SAUVAGEAU, 1894.

The work of Sauvageau is particularly a contribution to the biology of the Potamogetons. While
there are additions in morphology and anatomy extending the observations of Irmisch to other members
of the genus, the most noteworthy investigations pertain to the origin and the development of those
vegetative structures which greatly facilitate the multiplication of species during the vegetative
period.

Sauvageau devotes a special memoir to P. crisbus. He observed both forms of the so-called ‘‘burs’’
of this species, the slender spicular one and the more common denticulate one, noting their origin,
growth, and germination.

Experiments conducted in aquaria show that detached fragments of stems of various forms as
P. lucens, P. densus, P. perfoliatus, and P. crispus develop roots, shoots, and buds, and that such
detached parts of plants constitute a rapid means of propagation. Investigation of the growth habit of
P. natans discloses a condition in marked contrast to the above-mentioned species. No special propaga-
tive bodies exist, but the species perpetuates itself by the continuance of the rhizome anchored in
the mud, a rhizome which maintains itself through the winter rest period with the submersed shoots
in various stages of growth.

Experiments on seed germination indicate a latent period of considerable variability. In P. crispus
germination occurs within a year; in P. natans in from three to four years.

FRYER, 1900.

The first two installments of a fine quarto work, The Potamogetons of the British Isles, by Alfred
Fryer, appeared in 1900. The monograph includes the varying forms and states as well as the recog-
nized species, with accompanying plates, by the artist, Robert Morgan, who has reproduced the plants
in color with singular beauty and accuracy. Unfortunately for science, the author’s death occurred
before this important work was finished.

Fryer had an intimate acquaintance with the Potamogetons and their habits. He grew many speci-
mens in tanks in his garden, watching developments there and in their native haunts at different times
of the year. He grew Potamogetons in order better to classify them, for he recognized the necessity of
having a long series of specimens of the same form. “One set,’’ as he says, “would contain a series of
forms from lucens to heterophyllus without a single gap. This would show the way in which two quite
distinct species pass from one to the other without a missing link.’’ Asa result of these observations a
long and valuable series of communications on the genus, under the title “Notes on Pondweeds,”’
appeared in the Journal of Botany from 1883 to 1899.

BENNETT, 1880-1914.
In the Journal of Botany Mr. Arthur Bennett’s “Notes on Pondweeds’’ have appeared regularly
from 1880 to the present time. He has become the acknowledged authority on the classification of the

genus.

PIETERS, 1902.

In a Contribution to the Biology of the Great Lakes, Mr. A. J. Pieters notes the distribution of
aquatic plants, describes the forms occurring in diverse situations, presents details of structure, and
records various methods of vegetative reproduction. The Potamogetons, he observes, form a conspic-
uous feature of the aquatic vegetation, predominating, as a rule, in aquatic associations or flourishing in
isolated patches. P. heterophyllus, he says, exemplifies the latter condition in that it thrives in a surf-
beaten sandbar, where its runners ramify in all directions among the stones and pebbles, and its roots
penetrate the underlying clay. Details of structure which are figured for P. americanus suggest the
special adaptation of a thin, broad-leaved form, whose leaves are submerged, for withstanding diminu-
tion of light and rapid motion of water. The so-called hibernacula, or winter buds, represent the more
familiar forms of vegetative reproduction observed by the author.

POND, 1903.
Further contributions to the biological literature of aquatic plants have been made by R. H. Pond.
Two papers are presented on this subject. In the first, The Biological Relation of Aquatic Plants to the

POTAMOGETONS IN RELATION TO POND CULTURE. 259

Substratum, the author showed that rooted aquatics depend on the soil substratum for the supply of
nitrates. In conducting the experiments various aquatic plants were used, among which were P. per-
foliatus and P. obtusifolius. It was found that both of these plants are dependent on the soil substratum
for optimum growth, though the cuttings which were employed behaved differently in manner of growth:
P. perfoliatus showed an increase of growth through the development of new rhizomes; P. obtusifolius
manifested it in a continuation of the branches already present. The behavior of P. perfoliatus is in
accord with the observations of Sauvageau in his experiments on the propagation of Potamogetons by
fragments of stems.

The second paper of the author, The Larger Aquatic Vegetation, to appear in Ward’s American
Fresh-water Biology, supplements the work of the first by additional observations, discussions, and
generalizations. From his observations on the substratum of the larger aquatics it appears that they
may be found growing on gravelly, sandy, or loamy soil, the loamy soil supporting the greatest variety
of species. Direct experiments on this point, with the natural conditions reproduced as nearly as
possible, bear out this observation. The author states, moreover, that the character of the soil is so
important a factor that it is possible to predict the nature of the bottom from the species that are found
growing in it. For example, ‘‘Among the islands of western Lake Erie Potamogeton heterophyllus is
common on the reefs and pebbly shores, but it is not noticeable in the coves where a good soil sub-
stratum exists, and so prominent is it in the former places that its presence may be considered char-
acteristic of the flora.’”’

JEPSON, 1905.

In a popular article in the Sunset Magazine for February, 1905, Prof. W. L. Jepson has set forth the
possibilities of the marshes as a feeding ground for ducks. He has taken as a concrete illustration the
Suisun Marshes in California, marshes which abound in the fennel-leaved pondweed, P. pectinatus,
and which afford natural feeding grounds for the various kinds of wild ducks, more particularly the
canvasbacks. The canvasback and the broadbill, both diving ducks which visit these marshes, devour
greedily the tubers that are developed in abundance on the rootstocks and upper portions of the stems
of this Potamogeton in the autumn. It is claimed that these tubers give the fine nutty flavor to the
canvasback at this season of the year. Other ducks, nondiving species, feed on the tender rootstocks
and leafy stems which are brought to the surface in the feeding operations.

ASCHERSON AND GRAEBNER, 1907.

Ascherson and Graebner have published the last important monograph on the group, the Potamo-
getonacee, in Das Pflanzenreich. In this work the whole number of the described species has reached
87. Of these North America has 38, 14 of which are exclusively American. The numerous forms and
varieties that are listed, though some common ones are omitted, illustrate how difficult the problem
of classifying the Potamogetons still remains. In addition to the literature on classification the authors
have assembled much important data on the anatomy and morphology of the group from foreign sources
not generally accessible. Under the caption, Uberwinterungsformen und Vegetative Vermehrung bei
Potamogeton, the following propagative structures are figured: The slender, spicular bur of P. crispus;
the large denticulate bur of P. crispus; the tuberous rhizome of P. lucens; the tubers of P. pectinatus;
and the winter bud of P. obtusifolius. All except the first are reproductions of Irmisch’s celebrated
monograph.

MCATEE, Ig!I.

In a bulletin of the Biological Survey entitled Three Important Wild Duck Foods, Mr. McAtee has
assembled for Government publication important data regarding these foods, in the hope that they may
become more widely known and propagated for the preservation of wild ducks. Analyses of the food
content in the stomachs of the more important species of the game ducks show that the pondweeds, the
Potamogetons, are a favorite plant food. The ducks which apparently show a special fondness for it
are the canvasback, the redhead, the scaup, and others, the first of which takes a very large proportion
of the Potamogeton, the amount being nearly 50 per cent of the food eaten.

The best known duck food among the Potamogetons is P. pectinatus, of which the seeds, the tender
rootstocks, and the tubers are eaten. It is general in distribution, thriving in fresh, brackish, or salt
water. This and other widely distributed species are figured, and suggestions on how, when, and where
to plant them are given.

260 BULLETIN OF THE BUREAU OF FISHERIES.

MICKLE, I912.

A Canadian bulletin, The Possibilities of Northern Ontario as a Breeding Ground for Ducks, by
G. R. Mickle, is an investigation of the shoal waters of that Province with a view to their utilization
for the propagation of wild game. In the preliminary survey the approximate amount of shoal waters
is estimated to be 2,800,000 acres, on which various edible water plants grow. But it is hoped that
the natural supply may be augmented considerably by transplanting such of the larger aquatics as will
contribute especially to the food of wild ducks. Among the valuable water plants suitable for trans-
planting, the author names several species of Potamogeton, P. natans, because of its abundant seed
habit, P. perfoliatus and P. crispus because of their edible leaves.

MICKLE AND THOMPSON, 1913.

A second Canadian bulletin by G. R. Mickle, written in collaboration with R. B. Thompson,
supplements the work already done in this line. A table giving the estimated percentages of the
various constituents of duck food shows that both P. heterophyllus and P. perfoliatus form an im-
portant food constituent in the diet of wild ducks.

It will be seen from the foregoing résumé of the literature on the genus Potamoge-
ton, that an important extension of the subject is in the field of biologic research, an
aspect of the study which regards also the economic significance of the group. It is
apparent, too, that this field of research concerns itself primarily with the propagation
of Potamogeton by such structures as tend readily and effectually to distribute the
group, viz, by burs, tubers, rootstocks, and winter buds. These have been described
generally in Europe and in America, and one may consider their production a natural
phenomenon.

SPECIES OF POTAMOGETON INVESTIGATED.

The species which are included in this investigation have been selected from the more
or less common forms growing in the lakes and ponds at Ithaca, N. Y., and vicinity
(Spencer and North Fairhaven). And these species have been chosen because they
offer variety in habitat and in methods of propagation, and because they serve an
important rdle in the economic relations of aquatic life, affording food, shelter, and
support to many forms of animals which exist among them. The list of species follows:

P. americanus C. and §. | P. zosterifolius Schumacher.
P. amplifolius Tuckerm. P. obtusifolius M. and K.
P. heterophyllus forma terréstris Schlechtd. P. filiformis Pers.

P. perfoliatus L. P. pectinatus L.

P. crispus L. P. Robbinsti Oakes.

These Potamogetons were studied from September, 1912, to June 1914, in their
natural habitats and in aquaria. Entire plants were thus observed throughout the
period of development of those structures which are valuable in the vegetative propa-
gation of the species. From time to time collections were made of entire plants with
their subterranean systems intact. In shallow waters the plants were uprooted by
hand, but in the deeper waters they were obtained by means of a rake or a grapple
thrown over the side or the stern of a rowboat. No collections were made in mid-
winter, i. e., from the latter part of December to the middle of February, when the
frozen condition of the lakes and streams rendered it impracticable. P. crispus is an
exception, since it was collected from spring pools at all times of the year.

These studies have afforded an opportunity to observe the animals that are inti-
mately associated with the Potamogetons. Such have been noted, especially those
forms which depend upon these plants for food, support, or shelter.

POTAMOGETONS IN RELATION TO POND CULTURE. 261

GENERAL SURVEY OF LIFE CONDITIONS OF THE SPECIES INVESTIGATED.
POTAMOGETON AMERICANUS.

This species, which has been grown from seed and cultivated through two succes-
sive seasons, will receive more specific treatment later under the caption “Natural
and artificial propagation.’ In its natural habitat this plant has been observed
growing near the mouth of Fall Creek, a tributary of Lake Cayuga, and in a near-by
cove of the lake, at varying depths of 3 to 4 feet. It has been observed also at Spencer
Lake, at about the same depth but in much swifter water. In the latter situation
the blades of the leaves are conspicuously attenuated. According to Fryer (1900),
who has observed this plant in various localities, it is a plant of upland streams and
rivers rather than of stagnant waters.

By uprooting the entire plant in the growing season, it is found that the stem
springs from a rootstock that is deeply anchored in the mud, where new shoots radiate
horizontally from the established parts of the plant. During the summer these young
rootstocks produce large buds at their tips (fig. 6). After the plant dies down, which
may occur as early as August, the subterranean system remains intact for several
weeks. The new rootstocks, however, carrying the buds at the tips, become eventually
detached through the disorganization of the parent stem and in time die away, leaving
but little beyond the isolated buds to perpetuate the plant the following spring. Such
buds, since they remain in a quiescent state during the winter, may be called winter
buds or hibernacula, a term applied to structures of a similar nature and function.
Mr. A. J. Pieters (1901) doubtless referred to propagative structures of this kind when
he recorded for P. americanus (P. lonchites) ‘‘extensive runners bearing buds at their
ends,” though no figures are given and no further observations are noted.

Fryer (1888) mentions an autumnal state of P. americanus (P. fluitans) in which
the leaves are all narrowly linear or grasslike. These later growths, he says, are
developed in the axils of old leaves during the natural decay of the lower part of the
stem. They are ultimately set free as fascicles of narrow leaves which, after rootlets
are formed at the base of the new growth, sink to the bottom and continue the life of
the species. Such structures, which would be analogous to the winter buds of P. obtu-
sifolius and P. zosterifolius, have not been observed in P. americanus under investiga-~
tion, though they may have been overlooked in the changes of water level during the
autumn.

POTAMOGETON AMPLIFOLIUS.

This is an American species distributed quite generally throughout the continent.
It forms large patches in the open vegetation but thrives also in close association with
P. Robbinsti, Heteranthera dubia, Ceratophyllum demersum, Elodea canadensis, and
other plants of aquatic meadows. As a forage plant it may be regarded as one of the
best, growing continuously from early spring to late fall or early winter and producing
an abundant herbage by reason of its numerous large leaves. The rankest growths
have been found in the more quiet waters of Lake Cayuga and ‘“‘The Pond” at North
Fairhaven, at depths of 5 to 7 feet, in a substratum of mud rich in vegetable mold.
Propagation is rapid. The dense patches of stems, more or less unbranched, arise in
great numbers from an intricately developed subterranean system (fig. 7).. This

262 BULLETIN OF THE BUREAU OF FISHERIES.

extensive ramification of underground stems, richly provisioned with starch, remains
more or less intact during the winter, carrying at alternate nodes undeveloped shoots
which quickly establish new extensions of the plant in the spring. Another means-of
vegetative propagation is found in the detached tips of branches which, after separation
from the decaying parent stem in the fall or spring, sink to the bottom and become new
centers of growth. New shoots also develop at the nodes of decaying stems and, on
separation, sink to the bottom and take root as in the case of detached tips and stems.
Besides these vegetative means of growth this Potamogeton produces an abundance of

seeds.
POTAMOGETON HETEROPHYLLUS.

Various forms of this species occur throughout almost all of North America except
the extreme north. One of the numerous forms, forma terréstris Schlectd., is repre-
sented in this investigation and all data herein recorded pertain to this plant. It isa
so-called land form of Potamogeton and briefly characterized in Gray, seventh edition,
as ‘freely creeping in exsiccated places, producing numerous branches which bear tufts
of oblong or oval coriaceous leaves but no fruit.’’ This plant, which grows in the open
air after being left entirely uncovered by water, has been observed in two places along
the shores of ake Cayuga—one in a railroad pool 2 miles east of Ludlowville and the
other on a sandbar at Myers Point. In each of these places it is interesting to note
that gradations in habit accompany the varying changes in habitat. The railroad
pool is a particularly favorable spot for the growth of this Potamogeton. It is an
artificial pond which has been developed by building a railroad embankment near the
foot of the bluff bordering the lake. In consequence, a long, trough-like depression
exists between the bluff on one side and the railroad embankment on the other, with
water from the lake seeping through and maintaining itself at about the level of the
lake. It is a situation especially favorable to the growth of this plant, because the
annual withdrawal of the water is gradual, following the natural lowering of the lake level
during the summer months. The bottom of the pool is covered with black mud, largely
marl in composition, a foot or more in thickness, over which water may rise to the height
of 12 to 16 inches. During high-water level in the spring, this Potamogeton grows
submerged in the pond with its tuberous rootstalks anchored in the mud (fig. 8). Upon
the withdrawal of the water, following the lowering of the lake level in the summer,
drought conditions prevail, and then the submerged leafy stems give place to the land
forms. Upon the approach of drought conditions the previously submerged leaves die
and from the main rootstock or from those arising from the axils of the lower leaves
(upper in some cases, Bennett, 1880) runners extend horizontally to a depth of 2 to 6
inches below the surface of the mud. From the fertile nodes of these runners erect
axes arise, bearing tufts or rosettes of leaves which cover the ground in great numbers
and compete with mosses and small forms of sedges, carpeting the surface of the mud.
The leaves of the rosette (fig. 12) are unlike the elongated, membranous, submerged ones.
They are more rounded in form and coriaceous in texture, and by the presence of sto-
mates on the upper surface of the lamina, they are enabled to function as ordinary

leaves.

POTAMOGETONS IN RELATION TO POND CULTURE. 263

During the season of 1913 the plants which flourished in a submerged condition
during the month of May gradually changed their habitat upon the withdrawal of the
water during June and became land forms by the first of July. At this time the tuberous
rootstocks which perpetuate the plant vegetatively were well developed, and waited
only the final stages in the curing process to become the perfected vegetative structures
which tide this species over the unfavorable season of growth.

On the sand bar at Myers Point, the other station where this Potamogeton thrives,
the life conditions are not so sharply marked by the complete withdrawal of the water
during the dry season, and the various stages exhibited in the transmutation from aquatic
to land forms were easily observed. In water about 10 inches deep the continuously
submerged plants developed low bushy stems, with a few coriaceous leaves at the top.
In shallower water the plants behaved in the same way, producing bushy, stunted-
looking stems, which finally graded into land form with leaves in tufts or rosettes resting
on the exposed surface of the sand bar. The rootstocks, which were twisted and con-
torted in their effort to become established in the pebbly and gravelly sand bar, were
buried from 2 to 4 inches beneath the surface in the rich, black soil of the bar. All
of thé internodes of these subterranean stems were more or less thickened and often
attained a length of 8 to 14 inches.

No fruiting plants were found, and this observation is in accordance with the
generally accepted opinion that this form of heterophyllus is propagated entirely by
vegetative means. Observations on the artificial propagation of this species are recorded
in a later chapter of this paper.

POTAMOGETON PERFOLIATUS.

The leaves of this plant afford valuable forage material, though the season of
growth is comparatively short, the plants appearing late in the spring and dying quite
early in the autumn. In the environs of Ithaca this species flourishes in quiet waters
either in a substratum of sand at the relatively shallow depths of 2 to 3 feet, or in
“aquatic meadows”’ in a substratum of mud at depths of 3 to 5 feet. The observa-
tions of Pieters (1901), in “The Piants of Lake Clair,’ and of Thompson (1897), in
“The Biological Examination of Lake Michigan” extend the range of depth at which
this species exists to 12 feet. During the growing season the vigorous undergronnd
stems increase rapidly the output of forage material, since a single subterranean system
produces a large number of erect, much branched, leafy stems. The experiments of
Pond (1903) and Sauvageau (1894) and the observations of R. B. Thompson (1913)
afford evidence of other means whereby the rapid extension of this plant takes place.
In accordance with their observations, young branches, which are easily detachable,
float away and rapidly become new centers of growth. In winter the vigorous and
abundant subterranean system decays, leaving only the terminal shoots of two or
three nodes (Fryer, 1900) to continue the plant the following spring. This plant,
therefore, has three important means of vegetative propagation: By readily detached
leafy stems, and by extensions of the subterranean system, both of which operate to
multiply the plants during the growing season; and by the terminal portions of root-
stocks which, remaining in a quiescent state during the winter, establish new plants
in the spring.

264. BULLETIN OF THE BUREAU OF FISHERIES.

POTAMOGETON CRISPUS.

This species, a native of Europe, was recorded in this country by Pursh as early
as 1814 (Arthur Bennett, 1901). Since that time it has become established over an
extensive area because of the remarkable facility for multiplying itself vegetatively.
It is the most abundant Potamogeton in the vicinity of Ithaca, where it flourishes in
various habitats—in deep or shallow water, in sand or mud bottoms, and in stagnant
pools or flowing streams. It is singularly adaptive in each situation.. It has been
collected with P. pectinatus growing at depths of 8 feet, in which habitat the internodes
are extremely elongate; it has been found in pools where the substratum is an accu-
mulation of débris from ash heaps and dumping grounds; and it is not uncommon in
the swifter parts of streams and along the lake shore in sandy situations where the
substratum is thrown into ripples by wave and current action. In the latter situation
it has always possessed short, stocky stems and a general dwarfish appearance,

P. crispus grows the year round and spreads with great rapidity. It is propagated
primarily by “burs,” peculiarly distinctive structures to which there is nothing quite
comparable in our native species. Morphologically they are branches, but in the stage
most frequently seen they are scarcely recognizable as such members of the plant
structure. They have a horny look and a reddish color. ‘The shortened internodes
and thickened persistent leaf bases combine to give the characteristic bur-like appear-
ance (fig. 22).

POTAMOGETON ZOSTERIFOLIUS.

This flat, grass-like species of Potamogeton is not largely foraged upon by aquatic
herbivores, yet it appears in greater or less abundance in most ponds and lakes and
doubtless serves an important réle in the economy of life by furnishing support and
shelter to the countless small forms which have been found upon it.

P. zosterifolius.is among the earliest of the Potamogetons to appear in the spring, as
well as among the first of them to disappear in the autumn. It flourishes in a sub-
stratum of mud in still or running waters, and while it is not adapted to possess the soil
so completely as P. crispus, nevertheless it has effective means of perpetuating itself.
Mr. A. J. Pieters (1901) remarks that this species, which he has observed growing in
abundance in Lake. Erie, may be losing the power to produce seeds. Indeed, during the
past season few plants matured seeds in the several regions where they were observed,
but all developed winter buds in great abundance (fig. 33).

Large quantities of vegetation, that is, the accumulation of the varied and abundant
mass that still exists in the autumn, have been hauled up to the surface for examination,
and it was both surprising and astonishing to see the vast number of winter buds of
this Potamogeton that were entangled among the stems of other plants. It suggests to
an extent how well this species accommodates itself to its surroundings. It never
forms dense patches of growth, but it often occurs with aquatic plants that form them
more or less densely. By virtue of its slender, grasslike habit, it occupies the interstices
of the more rank aquatic flora, and it occupies these spaces as simple individual plants,
not as erect axes of a complete and intricate subterranean system. The plants are
anchored to the substratum by the roots only, which develop from the winter bud, and
because of this loose hold in the soil they are readily pulled up. The large number of

POTAMOGETONS IN RELATION TO POND CULTURE. 265

plants which have been uprooted appeared always to possess a comparatively simple,
erect stem which developed from a winter bud without the ramifications of rootstock
which are characteristic of other species of Potamogeton not grasslike in habit.

POTAMOGETON OBTUSIFOLIUS.

This species is apparently an important aquatic forage plant, for its delicate leaves
show abundant evidence of larval depredations throughout the growing season. It is
somewhat grasslike, yet less stiff and harsh than the preceding species. It is a rare
Potamogeton in the flora and has been observed in one place only, Spencer Lake, where
it is found in a muddy substratum in shallow waters of more or less swiftness. The
plant has a bushy habit of growth, branching widely toward the summit, a habit which
tends to produce dense patches of these plants. At one place in the station it grows in
such dense masses as to choke up the mouth of a small stream entering the lake.

The plants are late in appearing among the other aquatic forms in the spring,
lagging behind P. zostertfolius a month or more. The bushy habit of the plant begins
to show itself early in the summer, when branches arising near the base of the plant
ramify toward the top until the characteristic bushy habit is attained. Fruit is pro-
duced abundantly, but doubtless an equally important structure in the reproduction and
distribution of the species is to be found in the large winter buds. These appear on the
much-branched stems in great numbers and differ in no essential respect from those of
P. zosterifolius except that they are much less stiff. As in the above-mentioned species,
they fall away from the parent plant when mature and sink to the bottom. Like
P. zosterijolius, too, there is characteristic simplicity in the underground system. The
mature plants which have been collected show no tendency to produce ramifications in
the substratum, nor any indication of a perennial habit, but the plants become readily
propagated vegetatively by means of winter buds or hibernacula.

POTAMOGETON FILIFORMIS.

A habit sketch of this plant is shown in figure 36. Morong (1893) states that this
is a rare species in the United States. One collection only was obtained. The specimens
were found early in July near Canoga on Lake Cayuga, where the plant flourished in
shallow water and among calcareous rocks along the shore. The plants were short and
bushy in habit and bore abundant fruit. In all cases the erect axes developed from a
tuberous rootstock which, judging from the numerous erect shoots that grew therefrom,
is the common method of vegetative propagation in this species. The tubers (fig. 37)
occurred in series of 3 to 5 on the rootstock. Although no opportunity was afforded
for studying this plant during the successive seasons, it is deemed worth while to
record the observations of one collection of plants, since this species of Potamogeton
is unique in its habitat and promising in the possibility of seed and tuber production.

POTAMOGETON PECTINATUS.

This species possesses many important characteristics which recommend it to the
culturist of aquatic plants. It is one of the most abundant and widespread of the
Potamogetons. P. pectinatus is regularly found in quiet waters, though it has a variable
habitat in other respects, occurring in a substratum that is sandy or muddy and in waters

266 BULLETIN OF THE BUREAU OF FISHERIES.

that are deep or shallow, fresh, salt, or brackish. It is also extremely variable in growth
habit. Two of its remarkable forms which occur in Lake Cayuga and its environs and
which Dudley (1886) describes as a slender form® and a gigantic form? are included in
the present investigation of this species.

P. pectinatus, the species which is common everywhere, is among the first of the
Potamogetons to sprout in the spring, making its appearance early in April. Of such
plants which appeared early in the season, over a hundred individual specimens were
uprooted to determine the agent of propagation. In all cases these plants developed
from tubers which were buried in the mud or sand. Figure 38 shows the general habit
of growth from these reproductive structures. The new plant quickly establishes itself
by developing simultaneously with shoot formation an extensive subterranean system
of stems, which in turn send up leafy shoots in great numbers. By this ramification of
the underground stems, P. pectiznatus encroaches upon the soil so effectively as to produce
dense patches of growth, to the exclusion, in some cases, of other species of aquatics.
The plants bear fruit more or less abundantly, but, in general, tuber formation doubt-
less equals or surpasses seed production.

Tubers of various size occur, the size being dependent, more or less, on the nature
of the environment. The largest and finest specimens were found at North Fairhaven
in the quiet waters of Sterling Creek, where P. pectinatus forms a part of an equatic
meadow renowned for its luxuriance of vegetation. These large tuber-bearing plants
grow in the rich, mucky substratum at a depth of 6 to 10 feet in association with
Elodea canadensis, Myriophyllum spicatum, Ceratophyllum demersum, Utricularia vulgaris
var. Americana, Nymphea advena and other Potamogetons, such as amplifolius and zosteri-
folius. In this situation the plants are rapidly propagated from the tubers. On June
21 several specimens were collected which illustrate the complete cycle of tuber forma-
tion. Plants retained intact the old tubers, the new shoot—a tall, leafy, erect axis
bearing in some cases a floral spike—and the new rootstocks bearing tubers. On
many plants in this most favorable environment the tubers were greatly in excess of
the matured fruits, and often the only reproductive structures. The plant dies down
early in autumn. In October attempts were made to collect underground stems to
determine, if possible, a perennial habit in this region. Only portions of the rootstock
were secured, but in every instance disorganization had progressed to a considerable
extent. The appearance of the tuber in the spring, when many of the plants were
uprooted and observed with shoots growing from them, indicates a complete and
natural separation from the parent stem, probably in the autumn. It may be inferred,
then, that the tubers are the only vegetative structures that do survive the unfavorable
growing season.

The slender form described by Dudley (1886) was found still occupying the same
nesteny in er dise Lake where it was observed a him many Es ago. The plants

@ too7. Var. @) with slender elongated stems (1 to 1!4meters); nodes remote, as are the w) faite of the spike, whose peduncle
is usually over one-lourth meter long. Leaves few and slender, plants sometimes proliferous. Near the lighthouse, Cayuga L.
Dr. Robbins “found no parallel for this remarkable form,"’ in his own observations. Dudley, William R.: The Cayuga Flora.
Bull. Cornell Univ. (Science), p. 107.

+ 1008. var. (2) a giganticform growingin deep water northwest and northeast of the lighthouse, Cayuga L. Not yet
found in flower or fruit, though examined more or less frequently during ro years past. It is frequently proliferous, especially if
detached. It grows in banks, the plumelike bushy tops reaching the surface of the water. The leaves and sheaths are similar to
P. pectinatus, except in length. Dr. Robbins remarked that he had “nothing that comes near to it in length of leaves—usque ad
10."’ Stipules are usually much shorter than in P. pectinatus. Specimens were obtained in 1874 from 4 to 544 meters long. This
form was also noticed by Mr. H. B. Lord, probably somewhat earlier than 1874. Loc. cit.

POTAMOGETONS IN RELATION TO POND CULTURE. 267

grew in banks in sand and silt bottoms at a depth of 5 to 7 feet. They were quite
unmixed with other aquatics. In July and early August the long heavily fruited
spikes floated in dense masses at the surface and gave to these areas of the water a
characteristic brown look. Proliferations were not found on these plants during the
summer; fruits, however, were more abundant than on any other form of pectinatus.

The gigantic form of pectinatus grows in deep water. Plants 8 feet long are com-
mon, although many average but 5 feet at the end of the growing season. ‘This form
grows in a substratum of sand and silt at depths varying from 6 to 12 feet in a region
of the lake exposed to a more or less constant sweep of the wind. ‘The plants, there-
fore, which grow practically to the surface of the water, are subjected at times to
vigorous wave action. Altogether these environmental conditions favor a growth of
remarkable luxuriance. The plants grow in banks, and so thickly as to preclude the
possibility of encroachment by other forms of vegetation, though in shallow places,
where the growth becomes sparser, a few scattered representatives of P. crispus,
P. perfoliatus, and Heteranthera dubia occur.

This form of pectinatus begins growth early in the spring. In May, 1913, the
plants already approached the surface of the water. On June 21, 1913, a plant bearing
a single floral spike was found, although in several collections made thereafter neither
flower nor fruit was obtained. This appears to be the first record of a floral spike on
this form of pectinatus. From the collections made in November a few tubers were
found on the tips of the foliage sprays of the plants that were uprooted from their
natural moorings, although they were found more commonly on sprays that were floating
in the drift. This latter observation is an agreement with Dudley (1886), who observed
and described this form in Lake Cayuga. No rootstocks were secured, since attempts
to uproot the plants at such depths with a grapple resulted always in breaking the
stem just short of the subterranean system. This appeared to be embedded firmly and
deeply in the substratum, at least more deeply than the length of the grapple teeth,
which measured 4 inches. However, the bases of the erect stems, the parts which
develop just above the rootstocks, possessed remarkable examples of proliferation.
Thickened runners, more or less contorted, arose from leaf axils at the bases of the
erect stems (fig. 50, A), terminated by large, elongate tubers. The bases of the stems
were hard and woody, more especially so in the regions where they became detached
from the underground system. This condition suggests a continuation of the woody
structure in the subterranean parts. It may be inferred perhaps, from the general
habit of the plant and the attendant conditions of growth, that the rootstocks are per-
ennial, and that the basal runners, which bear in abundance large tubers and green
shoots, are the chief propagative structures of this form of pectinatus.

POTAMOGETON ROBBINSII.

Although this Potamogeton is less well known than the other species, it is
destined to be regarded as an important aquatic forage plant, first, because it is very
prolific, and, second, because the foliage is very generally eaten. The habitat of this
species, where it has been under observation, is not unlike that of P. amplijolius, with
which it is often found in association. It has been observed in the quiet waters of lakes
and ponds at depths of 3 to 5 feet in a substratum of rich, black mud. The stems
ascend from a somewhat creeping base and branch profusely in a more or less two-

268 BULLETIN OF THE BUREAU OF FISHERIES.

ranked arrangement, forming large, broad, flat sprays of foliage, which often cover the
bottom in large patches. It is the rarest of all Potamogetons to fruit, at least in the
situations where it was observed, but because of the tendency to branch profusely
propagation is readily effected. The branches, especially those whose internodes remain
short, become thickened and hardened through the storage of starch, and when
detached function as propagative structures. This enlargement and induration may
occur also at various points along the main axis that bears the propagative branches,
so that the final dismemberment of the whole plant provides enormous possibilities
in the multiplication of the species. Dismemberment may occur in the autumn, but
the plant is hardy, and this natural separation of parts may be deferred till spring,
then long rootlets develop at the nodes and establish the plant at once. The plant is
tardy in beginning its growth in the spring, but this tardiness in growth is obviously
advantageous to a plant that propagates mainly by vegetative means in the manner
of this species. Moreover, a very material advantage accrues in that the full and
complete foliage of this Potamogeton appears late in the season when many other
aquatics, including Potamogetons, show signs of decay. ‘Growth occurs during the
winter. It is not great, however, and manifests itself only in a slight elongation of the
branches, producing fresh, green tips of foliage, which are foraged upon by aquatic
herbivores almost as fast as the leaves appear.

SUMMARY OF CULTURAL FEATURES,

The Potamogetons which yield important forage products fall into two groups:
Those which produce abundant herbage in their leaves—P. americanus, amplifolius,
perfoliatus, crispus, and Robbinsti—and those which develop a large supply of starchy
food products in the tubers and tuberous rootstocks—P. pectinatus, filiformis, and
heterophyllus.

The species which grow best in the currents of streams are P. americanus and
obtustfolius; in deep water, P. pectinatus, especially the slender and gigantic forms of
Dudley; in calcareous regions, P. heterophyllus and filiformis; in exsiccated places,
P. heterophyllus.

The species appearing early in the spring are P. americanus, zosterifolius, pectinatus,
heterophyllus, crispus, and ampiltfolius; those growing late in the autumn and continu-
ing throughout the winter are P. crispus, amplifolius, and Robbinsit.

Abundant fruit is produced in P. perfoliatus, obtusifolius, and filiformis, and on
pectinatus in most situations. Vegetative reproduction occurs freely in all species.
The important vegetative structures are: Winter buds or hibernacula in P. obtustfolius
and zosterifolius; modified branches in P. crispus and Robbinsti; tubers in P. pecti-
natus and filiformis; tuberous rootstocks in P. heterophyllus, and subterranean buds in
P. americanus, ampiifolius, and perfoliatus.

NATURAL AND ARTIFICIAL PROPAGATION.

The natural propagation of Potamogetons has been touched upon in a general survey
of life conditions, and it has been seen that these plants propagate freely by means of
various vegetative structures. At this point it is desirable to consider this method of
propagation in greater detail, and to present data which will afford a means of com-
parison between the general seed habit and the tendency to produce vegetative propa-
gative structures.

POTAMOGETONS IN RELATION TO POND CULTURE. 269

PROPAGATION BY TUBERS.

A conspicuous method of vegetative propagation is seen in the development of
plants from tubers in P. pectinatus and P. filiformis. The tuber-forming habit of
pectinatus has been described by Irmisch (1858), who carefully worked out the morpho-
logical details of the tubers in terms of the ordinary stem structure. His figures illus-
trate the development of tubers on detached parts of leafy stems, on the erect axis, and
on the underground stems. It is not clear from which forms of pectinatus these drawings
were made. In general, however, they bear a close resemblance to our most common
representative of pectinatus, though no hint of the variability in this species is given
beyond the fact that some plants were collected in deep water, and that the tubers
were varied in shape, some being more cylindrical than others.

The work of Sauvageau (1894) confirms the observations of Irmisch as regards the
tuber-forming habits of pectinatus, but this investigator also makes no allusion to the
remarkable forms that exist in this species. His drawings, moreover, are, as he states,
modifications of those by Irmisch. Both of these workers in this field recorded the time
of tuber formation to be in the autumn. Jepson (1905) suggests an earlier development
for those on the rootstock and the erect stem. , He says: ‘‘The slender threads which
develop one, two, and even three tubers at the end, are not only borne on the horizontal
rootstocks and on the soil at the bottom of the ponds, but are also produced on the
upright stems, and at the end of the season on the uppermost leafy portion.”

Regarding the presence of tubers on rootstock, stem, and spray, the present investi-
gation is confirmatory. Tubers have often been observed on all these parts of the plants.
Additional figures and observations relate more especially to the season in which they
occur and to their artificial propagation. Collections of plants made on the 15th of May,
1913, and thereafter throughout the growing season, show the presence of tubers in great
numbers on the proliferating shoots of the rootstocks. Many of these tubers are well
grown in May, though others subsequently arise on the extensions of the subterranean
system which develop after this time.

Figure 41 represents the basal part of a small immature plant of P. pectinatus col-
lected in shallow water June 20, 1913. Many plants at this time were more nearly
mature and bore larger tubers, but it seemed desirable for illustration to select a small
plant because in such all parts may be preserved intact during the collection of material,
a task that is attended with considerable difficulty when the plant has attained a large
size and great complexity of parts, especially in the subterranean region, where the
underground stems are exceedingly brittle and tender. This figure (fig. 41) illustrates
the general sequence of growth in what may be termed the typical vegetative life cycle
of the plant. The order of development is as follows: The production of a leafy, erect
shoot (C) from the tuber of the preceding season (A); the growth of the horizontal axis
or rootstock (D); and the production of the stolon-like branch or runner which in turn
bears a tuber or tubers at the end (B).

As the season progresses the tubers become solidly packed with starch in sufficient
amount, apparently, to bring the plants developed from them to a very advanced stage
of growth, at least to render them quite independent of the soil for a considerable length
of time. Figure 45, B illustrates the typical condition in this respect when tubers sus-
pended in aquaria without contact with the substratum produce the future propaga-
tive structure. Thus the continued dependence of the plant upon the stored starch in

19371°—vol 33—15——18

270 BULLETIN OF THE BUREAU OF FISHERIES.

the tuber would seem to be advantageous, especially if growth occurred under untoward
conditions.

The tubers, hardened by the great quantity of starch that is packed into the tissues,
normally pass through the winter in a dormant state. This, however, is quite easily
disturbed, and by supplying continuously ordinary room temperatures the tubers may
send forth shoots as early as October. Figure 45, noted above, illustrates such a response
to growth conditions, the plant having been developed between the dates of October 22
and December 20.

The propagation of tubers in aquaria has shown that when tubers occur in twos, for
example, figure 40, the larger one develops the shoot. The smaller one has never been
seen to sprout unless by chance it became detached. In that case it developed an
individual plant. It has been frequently observed that plants of this species when
propagated in aquaria never attain their full size or vigor when deprived of a soil sub-
stratum, an observation that is in accord with the results of Pond’s (1903) experiments
on rooted aquatic plants.

The remarkable versatility of P. pectinatus as regards the origin of tuber-bearing
runners has been clearly shown by Irmisch (1858). There is, moreover, in each of these
situations, on rootstock, stem, and spray, a considerable variation in size and number of
tubers. For example, an underground stem or rootstock may develop them at the ends
of slender, stolon-like branches which arise from the axils of fertile nodes as shown in
figure 43. These have been found singly or in pairs, large or small, depending upon
the richness of the substratum and the size of the plant. Again, the rootstock itself may
be terminated by tubers which occur singly, in pairs (fig. 42), or in threes (fig. 39).
Plants bearing rootstocks of this character have been collected at various times during
the growing season, and from each collection the specimens have shown comparatively
short underground stems without other tuber-bearing structures. Some rootstocks
have shown no tendency to produce tuber-bearing runners or tubers at the end of the
horizontal axis, but send up a succession of leafy shoots from the fertile nodes. It is
suspected, however, that had such plants been undisturbed tubers might have devel-
oped, especially since at the base of these upright shoots there was always a bud, either
latent or showing a tardy development.

In autumn pectinatus develops tubers on the leafy spray. They are generally
smaller than those which occur on the rootstock, but quite conspicuous because of their
pale, yellow color. They are borne singly or in pairs at the ends of runners that are
bright green and stouter than the stems from which they arise (fig. 44, B, C). These
structures are readily distinguishable about the time the plant begins to show signs of
decay. They may occur on attached or detached parts of the plant. The remarkable
prolificity of these sprays is a characteristic of this species. Repeatedly detached
parts of the leafy spray have been placed in aquaria and tubers have been developed in
abundance until the spray became completely disorganized. It is interesting to note
that when this species grows in the currents of the stream the tendency to form pro-
liferations on the leafy spray is conspicuously lessened, although portions of these plants
when caught in the drift and carried to quiet water readily produce them in the new
environment.

For the most part tubers are more numerous on sprays devoid of fruiting spikes,
although exceptions are frequent. In examples of this kind, figure 44, B, C, shows the

POTAMOGETONS IN RELATION TO POND CULTURE. 271

origin of tube-bearing structures, one arising near the base of the peduncle, the other
solitary from the axil of a leaf. Figure 44, A, is a detail of such a spray showing the
usual character of the runner. Runners arise also on the lower parts of the stem
(Irmisch, 1858). These, like many on the tips of the spray, may develop so late in the
autumn that tubers never mature. What their fate is during the winter can only be
conjectured. It is a fact, however, that when placed in aquaria they continue to grow
slowly and eventually produce small tubers, or remain for a time in a quiescent state
and then send forth leaves and other runners.

In the gigantic form of pectinatus (Dudley, 1886) the tubers are elongate and large
in size. They are born on runners at the bases of the stems just above the substratum
of mud, and are therefore several feet beneath the surface of the water. Figure 61
shows the entire leafy axis with a tuberous runner attached at the base of the stem.
This is the normal position for what appears to be the chief propagative structure of
this form of pectinatus, and the usual condition at the approach of winter. The runner
is seen in detail in figure 50. The tubers are yellowish in color, and when stripped of
scales, which envelop them at this season, appear as in figure 51. The remainder of
the runner is dark green in color, more or less contorted and tuberous, and hardened
throughout by storage of starch (fig. 50, B). Secondary runners bearing tubers (fig. 50,
1, 2) are additional features in what withal is a remarkable propagative structure.
Peculiar tuberous internodes, transition stages, perhaps, in the formation of tubers,
appear frequently and characterize the more hardened and resistant portions exclusive
of the terminal tubers (fig. 52). On germination a leafy shoot and runner are pro-
duced. Figure 55, an illustration of a similar feature repeated in a series, was devel-
oped in an aquarium from the terminal tuber of a small runner. It illustrates how
resourceful in the propagation of this species so small a structure may become.

Young, green, leafy shoots arise from the fertile nodes of the runner (fig. 50, D)
and doubtless function in perfecting the propagative structures of this persistent part
of the plant, for at this time—that is, in the autumn—the leaves of the main axis
begin to disorganize. The young shoots retain their greenness through the winter,
remaining in a quiescent state meanwhile, and produce the main axis of the new plant
the following spring. When these structures are transferred to aquaria, they pass
through a winter-rest period, a period which is tess easily disturbed, however, in this
form of pectinatus than in others of the same species. Extreme plasticity is character-
istic of various portions of the runner. Fertile nodes produce either tubers direct, or
leafy tips, or runners, any .one of which may in turn produce a runner. The tip of a
secondary runner may produce a leafy shoot (fig. 53), and a tuber, instead of elongating
its axis in the natural way, may develop precociously a reserve bud which produces
the leafy stolon (fig. 54).

As in pectinatus generally, the detached sprays of the gigantic form show a greater
tendency to produce tubers than the attached ones. Likewise the runner is the im-
portant structure which bears them. Such tubers may become very numerous. As
many as 15 have been counted on a single plant (fig. 62). Detached portions of the
plant bearing tubers float away in the drift, from whence they may or may not find a
favorable place of growth in the spring. The tuber-bearing runners developed at the
bases of the stems rarely become loosened from the tangle of vegetation at the bottom
and must therefore repopulate the area year after year, encroaching but slowly on the
surrounding region.

iS}

~I
No

BULLETIN OF THE BUREAU OF FISHERIES.

P. filijormis represents a tuber-forming species which produces these propagative
structures apparently in the manner of P. pectinatus. Since material was collected but
once during the summer, no definite data can be recorded regarding the details of tuber
formation beyond the fact that the plants develop from tubers, as the collected materials
show (fig. 36, 37), and that these tubers, whether they occur singly or in a series of
two or more, have a likeness to those of pectinatus, in size resembling the common form
and in shape approaching more nearly the deep-water form. In details of structure the
tubers of filiformis are similar to those of pectinatus. Judging from the general habit
of the plant it seems fair to assume that the tubers have arisen in the same way and that
vegetative propagation would depend largely upon them.

PROPAGATION BY TUBEROUS ROOTSTOCKS.

The vegetative structures of P. hetcrophyllus terréstris are illustrated in figures ro
and 11. Morphologically they are a series of more or less shortened and hardened
internodes richly provisioned with starch. They are borne at the terminal portions of
the underground stems. Well-developed buds, the incipient, erect axes, occur at alter-
nate nodes of the structures, while the intervening nodes remain sterile, as in the case of
undifferentiated rootstocks. Figure 8 represents a typical plant collected early in
May. At this season the plant is still submerged. The tuberous rootstock of the pre-
vious year sends up young, erect shoots from the fertile nodes, and extends the growth
horizontally by an elongation of the terminal bud to form the new rootstock.

The underground stems acquire a distinctly tuberous appearance very early in the
summer. At Myers Point, where the collections were made frequently, the tuberous
character became apparent at the time when drought conditions began to prevail in
the pools; that is, when the water level was reduced to such an extent that the sub-
merged, leafy shoots gave place to the later-formed, erect shoots topped with tufts or
rosettes of aerial leaves which rest upon the mud. Figure 12 represents a plant of this
kind. By comparing the plants shown in figures 12 and 16 the origin of the tuberous
rootstocks is clear. In figure 16 tuberous structures appear at the ends of the new
underground stems, B and C. This tendency to produce the tuberous growth may
appear early when the plant is still submerged, though it may be deferred till drought
conditions prevail, when the new type of leaves forming the rosette above the ground
function to produce the abundant storage of starch which is found in the mature tuberous
rootstocks.

Some underground stems, throughout the growing season, continue to produce
internodes nontuberous in structure (fig. 9), but they are exceptional rather than the
rule. The tip of the rootstock that is destined to become tuberous generally shows
this character very early. ‘The internodes at the end do not elongate in the usual way,
but appear serially in a more or less bead-like form (fig. 1o and 11). Figures 13 and 14
represent the tuberous rootstocks partially developed. Figure 10 shows a fully mature
one. These structures, and many others in similar stages of development, were col-
lected in July and it is interesting to note that while some are only approximately mature
others are fully so thus early in the season. In November all evidences of other plant
parts have disappeared and the tuberous rootstocks only are left isolated in the mud,
where they remain in a quiescent state through the winter. A typical structure, as it

POTAMOGETONS IN RELATION TO POND CULTURE. 273

appears at the beginning of the winter, is seen in figure 11, although variations in the
length and thickness of internodes are not uncommon.

Tuberous rootstocks have been transferred to aquaria, where the growth has cor-
responded exactly with that exhibited in the natural habitat except in one respect, the
development of aerial tufts of leaves. But the explanation of this omission in the
cycle of development is clear, since the plants remained submerged in the aquaria.
The period of desiccation not having been interpolated, it is assumed that the tuber
formation progressed in a natural manner for the species. Figures 16 and 17, drawn
from aquarium specimens, show how in the purely aquatic phase of its existence the
natural habit of growth and reproduction in this Potamogeton is reproduced under
artificial cultivation.

PROPAGATION BY SUBTERRANEAN STEMS NOT TUBEROUS.

Among the species studied, P. perfoliatus, P. amplifolius, and P. americanus are
propagated in this manner. The plants are carried over the winter by means of the
terminal portions of underground stems, which are generally stouter than the ordinary
ones and which bear conspicuous scaly buds. These buds are the incipient shoots from
which the elaborate plant structures of the following season are developed. Sauvageau
(1894) has figured this propagative structure for P. perfoliatus as he found it at the
approach of winter. He states that the entire plant dies in autumn, except a few inter-
nodes which bear the buds for the continuation of growth in the spring. In figure
18 is represented a portion of an underground stem that survived the winter and pro-
duced the first few internodes of growth. The scales on the part that lasted through
the year are distinctive in appearance. They are larger and looser than the ordinary

eones, black in color, and leathery in texture (fig. 18, A.)

In P. amplifolius perennial parts are also found in the underground stem. Figure
7 represents the characteristic features of such a structure at the beginning of the winter.
The young, erect shoots A, A, A, wit partially unfolded leaves at the tips, pass
the winter unchanged and serve to promote rapid growth in the spring. The buds
terminating the horizontal stems remain latent through the winter and on unfolding in
the spring push out in all directions through the substratum. In these ramifications a
subterranean system of interlocking stems and roots is developed that fixes the plant with
exceeding firmness in the soil.

In P. americanus vegetative propagation is accomplished by subterranean scaly
buds which generally grow in pairs at the end of the rootstock (fig. 4and 5). The general
structure of the bud resembles that of P. perfoliatus. It is an incipient shoot, possessing
a succession of very short internodes and young leaves, with scales surrounding the whole
axis. A small portion of the rootstock generally remains attached to the buds and
persists through the winter.

PROPAGATION BY WINTER BUDS.

The winter buds afford the only means of vegetative propagation which have been
observed for P. zostertfolius and P. obtusifolius. These structures develop at the ends
of the shoots. The terminal internodes remain short and, becoming completely cov-
ered by closely overlapping leaves and stipules, form a hard, compact, cone-like bud.

274 BULLETIN OF THE BUREAU OF FISHERIES.

Such buds become conspicuous during the month of August. Later when they are
mature they easily fall away from the parent axis, which thereafter dies down com-
pletely. Being heavier than water, the buds sink to the bottom and by the middle of
October they have either disappeared or have become entangled in the accumulations
of Elodea, Myriophyllum, Ceratophyllum, ete., which still remain intact. In the dis-
organization of this mass of vegetation, a gradual settling of the entangled buds takes
place and they eventually find lodgment with the others in the substratum of mud,
where they remain in a quiescent state till early spring. Such buds may properly be
called hibernacula, since they pass through the unfavorable winter season in a state of
rest.

The general external aspect of the winter buds is seen in figures 33, 63, and 64. In
size and form the two buds are quite similar but the leaves of obtustjolius are less stiff and
harsh. In the internal structure of the bud (fig. 34) the typical branch-like character
is apparent with the young leaves closely crowded toward the tip.

Plants of both species have been reared in aquaria by anchoring the buds in sand
ormud. The latter operation is not necessary, however, since mature buds sink naturally
to the bottom, but it was a precautionary measure against the disturbance of buds under
observation in aquaria. The plants of zosterijfolius thus propagated did not bloom, but
produced winter buds; those of obtusijolius bore flowers and fruit early in August.

During the winter the loose leaves on the outside of the bud decay, but, on the whole,
the entire bud is well preserved. This resistant character is more especially true of
zosterifolius, in which many of the enveloping leaves of the bud persist long after the
new plant has become established. In the spring the first sign of growth is manifested
by a spreading of the inclosing leaves. Then follows the development of roots from
successive nodes (fig. 35) and the elongation of the internodes at the tip of the bud. .
This elongation carries the young leaves forward and upward, and in a short time the
general habit of the plant becomes apparent (fig. 65). The various stages in the growth
of the bud in the spring are, in so far as they have been observed, similar in those two
species of Potamogeton, except that obtusifolius lags behind zostertfolius.

PROPAGATION BY BURS.

P. crispus is the single example of such vegetative propagation. The first
evidence of propagative structures by means of which the growth of this species is
rapidly extended became noticeable early in May. At that time the so-called ‘‘burs”’
(fig. 22) made their appearance. They were enormously abundant, appearing in the
axils of nearly all the leaves. Many of them became fully mature by the middle of
the month; especially those which developed in pools of standing water where the daily
temperature of the water was comparatively high. In the colder waters of spring pools
and of the open lake these propagative structures, like the flowers and fruit, were
retarded in development, maturing about two weeks later. As the summer advanced
the development of the burs decreased until by the middle of July only scattered
individuals were to be found.

Asa rule, the burs occur in the axils of the leaves. They may, however, terminate
the growth of the axis (fig. 30). In this latter position they may occur in pairs (Savau-
geau 1894), often with a flowering spike. They may develop from the rootstock

POTAMOGETONS IN RELATION TO POND CULTURE. 275

directly (fig. 31), though this occurs but seldom. On the maturity of the bur detach-
ment from the parent stem is an easy and natural process. The tissue just below
the pointed base of the bur becomes softened and the burs fall away, either by their
own weight or by accidental contact with other objects. On reaching the bottom,
anchorage in the substratum is facilitated by the peculiar shape of the bur, a sharp-
pointed, spindle-shaped structure that is heavier than water. A rest period occurs
before germination takes place. This rest period is apparently a varied one, depending
on the season when the bur is matured. Those which matured early in the season, in
so far as it could be determined, germinated in the fall, and in October bore shoots
from 6 to ro inches long (fig. 59). Those maturing late passed the winter in the
quiescent state and germinated early the following spring.

The slender, spicular burs (fig. 21) described by Irmisch (1858) and by Sauvageau
(1894) were found more or less commonly in the axils at the base of the erect stem,
and always few in number compared with the stouter form. It is interesting to note
in this connection that these spicular burs appeared more abundantly on the so-called
“state’’ of P. crispus, a plant with flat, not undulate leaves, said to be a young state of
crispus (Fryer, 1900). In one of the spring pools from which collections were made
the spicular buds predominated on what appeared to be matured plants of this flat-
leaved form. The plants were never so vigorous looking as those in the other situations,
and the appearance of the spicular burs upon them may be explained by differences
in habitat. Generally they appear to be poorly conditioned plants, and from observa-
tions on their development it would seem that they are a starved state of crispus rather
than a young state.

The development of the large bur (fig. 22), which Sauvageau (1894) described in
part, has been observed in the field and in aquaria throughout the various stages, from
its beginning as a small branch to its completion as a mature bur. Since the steps in
the formation are essentially the same under natural or artificial conditions, observa-
tions will be presented on the material under control in the laboratory.

Vigorous looking plants were collected in the latter part of March and anchored
in a soil substratum in aquaria with running water. Cuttings also were used, some of
which were anchored in the soil and others allowed to float on the surface of the water.
Three weeks later, short, stunted-looking branches appeared in the axils (fig. 26, A).
They exhibited at once a noticeable thickness of the axis and later the peculiar
denticulate appearance at the base of the leaves (fig. 22, a). When the diameter of
the branch had become considerably augmented and the denticulate margin conspicu-
ous, disorganization of the leaves commenced from the distal end and proceeded toward
the base. Disorganization ceased at the tip of the denticulate base (fig. 22, a, 1). By
this time the basal portion of the leaf was hardened, thickened, and horny like the
axis, and the entire structure presented the characteristic burlike appearance. Figure
60 shows several small-sized denticulate burs in various stages of development.

Essentially the two kinds of burs are similar, differing only in certain minor details.
In the bur shown in figure 22 the leaf bases are large and always denticulate, the buds
in the axils are relatively small, and the internodes are short. In the spicular bur
(fig. 21) the opposite is true. The leaf bases are small and spinous with a smooth margin,
the buds are well developed, and the internodes are comparatively long. A difference

276 BULLETIN OF THE BUREAU OF FISHERIES.

between them is also apparent in the time of occurrence and in position on the stem.
Irmisch (1858) recognized a disparity between them and suggested a difference in origin,
though he was not able to determine this for both forms. The spicular burs he found
originating from the axillary buds of decaying, floating stems in autumn. ‘The den-
ticulate ones he found always mature and detached from the parent stem in muddy
bottoms of pools. Sauvegeau (1894) describes and figures both forms of burs, giving
their origin as well. My observations, however, are not in full agreement with their
representation on the stem as expressed in Sauvageau’s figures. According to his
illustrations, both forms are abundant on the same branch and at the same season of
the year. This has not been found to be the established order in vigorous and healthy-
looking plants. Numerous collections of P. crispus indicate that when the denticulate
burs are abundant—that is, in the early part of the growing season—the spicular burs
are scarce, and if present on the same stem they are sparsely represented at the base of
the axis. In every case the large denticulate bur seems to be the product of strong
and vigorous-looking plants, and the spicular bur a result of poorly conditioned ones.
That the spicular bur is a weakling would appear to be borne out by observations on
their development. When grown in aquaria they have been found on sickly-looking
plants and when germinating burs have been deprived of their vigorously growing
shoots, small shoots bearing spicular burs have replaced them. In this instance a dis-
turbance of the natural trend of growth would be the occasion of their formation.
When the spicular burs germinate they produce shoots bearing leaves not crisped,
but narrow and flat (fig. 25).

The internal structure of the bur is fundamentally like that of the ordinary stem.
No new features appear in the tissues of any part of the bur, but starch grains are
present in such great quantities that the cells become distended with them. In the
fully developed bur (fig. 71) the cells become so greatly expanded that the air cavities
are practically obliterated. It is to these distended cells so compactly stored with
starch that the hardened, indurated character is due.

The accumulation of starch in the bur furnishes an abundant storage supply for
rapid growth, after a rest period of greater or less prolongation, depending upon the
time of formation. Burs formed early in the summer may germinate early in the fall,
or, like those of later development, pass the winter in a quiescent state. Figure 23
shows a stage of germination which is usual in the early spring. It is obvious from the
general appearance of the shoots that burs of this character passed the early part of the
winter in the resting stage. At the same time burs much more advanced in stage of
growth (fig. 32) are frequent, and it is assumed that these are comparable to burs that
germinated in the fall (fig. 59) and grew but little during the winter. In aquaria a varia-
ble rest period is common. Under these conditions burs have been germinated after
periods of six weeks and of three months.

In the germination of a bur there are as many possibilities for the production of
stems as there are axillary buds on it, although usually not all of the buds germinate.
The greater number of burs bear but one shoot eventually, but several may begin
growth and produce short shoots (fig. 23). By experiment it has been found that when
a bur is broken into bits with one bud per node, each bud will produce a shoot. In
the development of a plant from the bur, progress in the growth of a shoot manifests

POTAMOGETONS IN RELATION TO POND CULTURE. P3579

itself first by the establishment of an erect axis, from which very soon a subterranean
system arises in the manner shown in figure 27. By further extensions of these axes
the number of branches is greatly augmented and the capacity for multiplication
greatly increased. E

P. crispus, like most of the Potamogetons, propagates readily by detached stems.
Many of these have been picked up in the drift along the lake shore where under favor-
able circumstances some, doubtless, find lodgment and establish new centers of growth.
Besides, in the spring there have been found leafy axes which, while still remaining
attached to the parent stem, lie prone upon the muddy or sandy substratum and, be-
coming rooted at the nodes, send up a long series of erect stems (fig. 20). In this manner
P. crispus combines the rapid growth from stolons with the normal spread of the subter-
ranean system and forms an effective means of possessing the soil.

The large number of burs which are developed indicate that they are the chief source
of distribution in this species. Some plants doubtless develop from seed, though they
can not represent any great number of the whole since comparatively few seeds mature.
To obtain some data on this point a large number of young plants were pulled up and to
the most of them a bur was attached, an observation which shows that, for the region
at least, this structure was the chief agent of propagation. From the standpoint of
prolificity, P. crispus represents a desirable species for cultivation. It remains to be
shown that this abundant herbage is of importance in the economy of aquatic life. Data
relative to this are recorded under the heading ‘‘ Economic aspects of Potamogetons.”’

PROPAGATION BY FRAGMENTS OF STEMS.

In P. Robbinsi the propagation occurs exclusively by vegetative means, depending
upon a more or less complete dismemberment of the plant. This breaking of the plant
into propagative structures does not take place at random, but occurs at very definite
points throughout the leaf-bearing part of the plant. At intervals along the axes of the
stems, a few internodes develop which are very short, and in them starch is stored
so abundantly that they become hardened and stiff and noticeably thickened in diameter.
At the limits of these indurated regions where the stems appear constricted, the tissues
soften when the structures are mature, and dismemberment becomes a natural operation.
The process of separation is similar to that which is met with in P. crispus and which
causes the detachment of the bud from its parent stem. Besides the main axes of the
plant which break up into many potential units, there are also numerous short, axillary
branches which possess the characteristic feature of the propagative structure. The
internodes are likewise short and stiff and conspicuously augmented by the deposition
of starch. Moreover, they are always provided with a growing terminal bud, a feature
which facilitates rapid propagation. When an axillary shoot becomes 6 or more inches
long it behaves like the main axis of the stem eventually breaking up into several propa-
gative structures. In figure 67 is represented a single branch showing the constricted
appearance which distinguishes a stem bearing more than one propagative structure.

In the spring, often before a general dismemberment of the plant occurs, very long,
white rootlets are developed at the nodes (fig. 57). “These rootlets serve to anchor the
new growth, whether it be an attached part of the plant or a scattered fragment of the
stem. The provision for the initial growth in these fragments of stems lies in the storage

278 BULLETIN OF THE BUREAU OF FISHERIES.

of starch within the tissues. Starch is so abundant that the air cavities are considerably
reduced by the distension of the cells (fig. 71). In portions of the stem where the tissues ©
are not obscured by the deposition of starch, it is seen (figs. 69, 70) that mechanical tissue
is scattered through the stem in greater abundance than is common in the other Pota-
mogetons, serving to support the heavy sprays of foliage and to give the rigidity of stem
which is characteristic of this species.

In P. amplifolius the tip ends of the branches function as propagative structures in a
manner similar to P. Robbinsw (fig. 58). These structures appear in the autumn devel-
oping only at the tips of the branches. The internodes are short and thick and densely
packed with starch. At the end there are a few partially unfolded leaves which con-
tinue to grow slowly or, at least, remain green all winter. These rapidly expand when
the roots develop in the spring and the entire structure forms an effective and rapid

means of propagation.
PROPAGATION BY SEEDS.

While the main purpose of this paper is a consideration of the vegetative means of
propagation, yet it is important by way of comparison to present such data as are
available on the propagation of these plants by seeds. In reviewing the literature on the
seed germination of Potamogetons, it appears that Irmisch (1858) and Sauvageau (1894)
have made the only contributions of importance.” Irmisch figures the germinating seeds
and two small seedlings of P. natans but otherwise gives no data concerning them.
Sauvageau found that P. crispus, perfoliatus, and pectinatus germinate in less than a
year and that P. natans remains dormant three or more years. No figures accompany
his account of their behavior.

In the course of the present investigation additional observations have been made on
P. pectinatus and P. americanus. ‘The seeds of both species were gathered in October
and kept in cold storage through the winter. On January 24 seeds of each kind were
placed in aquaria and kept at ordinary room temperatures. On February 14, the seeds
of pectinatus began to germinate, but this process was very irregular, extending over a
period of three or more weeks. These seedlings lacked vigor and nothing came of them.
On March 15 other seeds of the same species were taken from cold storage and placed
in aquaria as before. In this later planting germination was more uniform, the majority
of seeds sprouting within a few days of each other. Subsequent growth was rapid and
vigorous. It appears from the behavior of the seeds in the two experiments that the later
planting is advantageous. Figures 46 and 47 represent seedlings of the second planting
3 and 5 days old, respectively. Figure 48 represents a seedling of the same species
about ro days old, and figure 49, one about 3 weeks old.

The seeds of P. americanus planted on January 24, showed no signs of life till May 5.
Those of the second planting germinated between June 13 and 15. In this species also
the later planting proved to be more successful. Figures 2 and 3 represent seedlings,
respectively, 5 and 14 days old. When the seedlings were about 3 weeks old they
were transplanted and kept in outdoor aquaria with running water till October. Figure
4 shows one of these seedlings which produced winter buds during the latter part of the
growing season. These winter buds described in a preceding chapter are the vegetative

an a recent publication by Esenbeck the seedlings of P. coloratusaredescribed. (Esenbeck, Emst: Beitrige zur biologie
der gattungen Potamogeton und Scirpus. Flora, bd. 7, June, 1914, p. 151-212, fig. 59.)

POTAMOGETONS IN RELATION TO POND CULTURE. 279

propagative structures characteristic of the species. All of the seedlings produced them.
Figure 5 represents the first shoot in a germinating winter bud. It may be assumed from
the general behavior of the seedlings and the growth from the hibernacula that in this
species vegetative structures only are matured the first year, and that seed formation
is deferred at least until the second year.

At present definite knowledge regarding the young stages of Potamogeton, in
general, is very meager and this is doubtless attributable to the fact that the plants are
small and inconspicuous the first year and fail to develop fruit until one or more vege-
tative reproductions of the plant have taken place.

PRODUCTION OF SEEDS AND VEGETATIVE PROPAGATIVE STRUCTURES.

The abundance of P. crispus and P. pectinatus in the local flora have made it possible
to observe the relative production of seeds and vegetative structures in a considerable
number of these plants. Besides, the formation of the conspicuous vegetative structure
in both species is practically synchronous with seed formation. The observations on
mature plants selected at random form the basis of the following tables:

TABULATIONS OF PROPAGATIVE STRUCTURES IN POTAMOGETON CRISPUS, JUNE 16, 1913.

(A) BUR FORMATION.

Denticulate burs.
Plants with both burs | Spicular
Number of bs
plants: Withibuse and floral spikes. Le eare
poly num- :
er.
Floral
Burs. spikes.
4 I 60 Sc
I I Do on I
2 ne on I
4 2 OC me
I ee 2 2
9 4 os bn
2 aa 4 ae I
- 3 Be 4 3
I 56 on 2
I 5 56 56 2
I Be 5 2 I
Ir 6 oe O10 ae
6 ni 6 I Oe
6 7 do ste I
3 ae 8 I -
2 ae 8 2
I 50 8 3
9 8 «s a0
2 9 ae I
I 9 20 4
2 Ob 9 2 I
12 ee 190 oe
I Sa 10 I
I oe Io 2
5 II ae ate
I oo Ir 2 a
I mle: 12 I I
3 12 0 we 2
I a 12 3
I 15 a4
I ae 15 2 I
I 2r a 2 2
I0o III 135 32 20

280 . BULLETIN OF THE BUREAU OF FISHERIES.

TABULATIONS OF PROPAGATIVE STRUCTURES IN POTAMOGETON CRISPUS, JUNE 16, 1913—Continued.

(B) SEED FORMATION.

Plants bearing sterile spikes. Plants bearing fertile spikes.
| Number of
| = umber 0
Number of aie ‘of Ruanbes ot Number of ee 2 f fertile Number of
plants. plant. spike. plants. plant. spikes per | seeds set.
plant.
52 I 6-7 4 I nd 2-4
10 2 4-7 3 2 I 1-2
16 3 5-8 6 3 I I-5
6 4 5-9 2 4 I 2-8
2 5 5-8 I 6 I 2
3 6 5-7 I 2 2 6
T 7 Sad I 9 bs IT
50 139 Sot | 18 53 19

Table A shows a preponderance of burs over floral spikes; table B, a preponderance
of sterile spikes over fertile ones. A comparison of the tables A and B shows that bur
formation exceeds seed production; that is, the important mode of increase is by vegeta-
tive means. It should be remembered in this connection, however, that the bur which
is the most conspicuous is but one of several vegetative structures contributing to the
rapid extension of this species, and that seed production is, therefore, even less important
relatively than the tables suggest it to be.

TABULATION OF PROPAGATIVE STRUCTURES IN POTAMOGETON PECTINATUS, SEPTEMBER 30, 1913.

Number of fruiting
ikes and tubers on
Niurmnercs ote plant. Number of | Number of | Number of Number of
plants. fruiting tubers immature sabterran
Fertile Seka spikes only. only. stolons. mi eantereniat

spikes. s
I 4 a: 2 (a)
T os 10 x I
I 2 17 se 3 (a)
I 18 I (a)
s 25 2 (a)
I 5 AG (a)
I * 6 - (a)
2 2 5) 4 (a)
by oo 12 He (2)
I 5 =o I
2 ee 2 a 2
2 ae I
3 i 3 2 (a)
I 3 Ir I (a)
I 6 2 I I
I a0 6
7 2
I 3 (a)
I 4
I 1s (a)
H 8 (a)
2 oo x!
I 5
35 24 70 ae 86 31 | 43

@ Imperfect record. > This plant bore one sterile spike.

The fertile spikes of P. pectinatus produce, in general, from 1o to 15 seeds. The
tubers occur singly, in pairs, and in threes. Bearing these possibilities in mind, the -
tabulation of P. pectinatus indicates a close approximation to equality in the produc-
tion of seeds and tubers. The small number of plants from which the data were collected
is an objection which could be justly put forward, yet the results conform in general
with field observations in restricted areas where the common form of pectinatus

POTAMOGETONS IN RELATION TO POND CULTURE. 281

predominates. Propagation by tubers is, as we have seen, the more rapid method and
the one which produces a luxuriant foliage early in the growing season.

In view of the observations and experiments, it is clear that in any project in which
the propagation of Potamogetons is an important feature, success will be measured by
adherence to the general principle that vegetative reproduction is the dominant mode
of increase in the genus Potamogeton.

ECONOMIC ASPECTS OF POTAMOGETONS.

In the study of the various phenomena attending the propagation of Potamogetons
oppertunity was afforded to observe, more or less closely, various aquatic animals which
abounded on these plants. Their presence in such great numbers suggested the pos-
sibility that the Potamogetons might play an important réle in the economy of life
beyond that of mere shelter and support, or other mechanical and indirect relations
which have been ascribed to the larger aquatic plants for many years.

It has been stated by Pond (1905) that—

The larger aquatic plants, as such, are, while living, little used as food by aquatic animals, yet they
greatly increase the surface available for the attachment of microscopic plant forms, which are eaten
by smaller animals, and the latter in turn by the fishes.

In the very recent publication by Shelford (1913), bearing on the life relations of
aquatic animals, but little importance is attached to the larger aquatic plants beyond
the various mechanical and indirect relations that have so long been attributed to
them. He says:

The smaller aquatic animals are commonly either alga-eaters or predatory. The larger aquatic
animals are commonly predatory or scavengers. The rooted vegetation is eaten only to a small extent.
Small floating or swimming plants and animals are the basis of the food supply of larger animals. We
could probably remove all the larger rooted plants and substitute something else of the same form and
texture without greatly affecting the conditions of life in the water; that is, so far as the life habits of
the animals are concemed. * * * Plants in water are of particular use to animals as clinging and
nesting places.

Recent research bids fair to modify these generalizations by Shelford. Such a
relation as Pond describes has frequently been observed in P. pectinatus in the autumn
when myriads of midge (chironomid) cases have been found applied to the leaves (fig. 56).
The leaves are not eaten but they are thickly covered with diatoms and other small
alge which, doubtless, afford foraging materials for the larve. A small caddis fly
(hydroptilid) larva, with characteristic elliptical case, has also been observed in con-
siderable numbers in the same relation with pectinatus, the larve apparently feeding
on the epiphytic algal growth. The larve of both of these insects, after wintering on
the algal-covered leaves, have emerged as adults in the spring. Other midges and caddis
flies, flies (aquatic Diptera), moths (aquatic Lepidoptera), and beetles (Coleoptera)
have been found in great numbers on the various species of Potamogeton. The other
smaller invertebrate animals most frequently seen on these plants are Crustacea, snails,
and worms.

Another interesting relation existing between the Potamogetons and aquatic insect
forms is seen in the striking resemblance between the cases of a caddis fly (Leptoceride)
and the stipules of the leaf of P. americanus (fig. 75). The cases in which both larvze
and pupz dwell are attached along the stems and leaves in so characteristic a manner
as to become almost, if not quite, indistinguishable from the plant parts.

282 BULLETIN OF THE BUREAU OF FISHERIES.

Reighard (1894) has expressed in a table ‘‘a part of the imperfectly known relation-
ships existing between the various groups of plants and the invertebrate animals on the
one hand and the fishes on the other.’’ One of the great gaps in the chain of relations
therein expressed is a lack of definite knowledge concerning the réle of the higher plants.

Some definite research in this direction has been begun. Recent investigations
on the food habits of aquatic insects have shown that the larger aquatic plants do
serve as forage materials. According to Hart (1895), the larvee of Nymphula sp. (Para-
ponyx), an aquatic lepidopterous insect, feed voraciously on Potamogeton natans. Need-
ham (1907) mentions the presence of Nymphea advena in the diet of Chironomus albis-
tria, and Morgan (1912) found that the higher plant tissues formed an important part
of the stomach content of May-fly larvae. In view of these investigations the leaves
and other edible parts of Potamogeton were closely scrutinized for evidences of their
use as food. In my own investigations the first indication that the living tissues of
Potamogeton was being eaten was seen in the young growing tips of P. crispus, which
had been transferred from a pond to an aquarium in the laboratory. The leaves of
several plants were mined by a small larval form which proved to be a chironomid
(midge). The characteristic leaf mine is shown in figure 72. Miss Tilbury (1913),
who was working in the Cornell laboratory on the feeding habits of the midge, taking
advantage of this observation, reared her species, Chironomus cayuge Johannsen,
mainly on P. crispus and entirely on Potamogeton.

On examining the leaves of other Potamogetons it was found that practically all
species were foraged upon to a greater or less extent. Larval depredations were most
common on P. Robbinsii. In this plant the aquatic lepidopterous larva Nymphula sp.
(Paraponyx) is the chief herbivore, and so voracious is its appetite that a large proportion
of the growing tips are constantly being defoliated in the manner shown by figure 68.
Portions of the ieaf are cut out also by the larva, applied together by means of silk,
and used as a protective case or retreat during the larval and pupal stages. Nymphula
sp. is by far the most conspicuous larva feeding upon P. Robbinsiz, yet other important
smaller forms are numerous. The limy incrustation that accumulates very freely on
P. Robbinsii offers apparently especial inducements to certain case-making insects, as
midges and caddis flies. Such larve are exceedingly numerous on this species of plant,
and the limy incrustation is the chief material used in the construction of the cases.

A few of the chironomid larve that were common on P. Robbinsii collected at North
Fairhaven in October were segregated and fed exclusively on this Potamogeton. They
passed successfully through the pupal and adult stages and proved to be the midge,
Chironomous aberrans. ‘The larval and pupal stages have been hitherto unrecognized
in the life history of this species.”

The leaves of P. amplifolius were conspicuously mined by the dipterous larva
Hydrellia sp. (Ephydride). The pupa were collected on the leaves August 6. Several
flies and their parasites were reared from them, emergence occurring between August 16
and 20. The larva makes a wide, irregular mine through the leaf, and in each case under
observation pupates at the end of the mine toward the base of the leaf blade where the
edges naturally roll together and form a protecting furrow (fig. 73). Nymphula sp.
(Paraponyx) is also common on this Potamogeton and many of the young leaves are eaten
by them. Oftentimes the larva cuts out a portion of the leaf for its case with the neat-

® Determinations of dipterous larvee have been made by Prof. O. A. Johannsen; of caddis-fly larve, by Mr. J. T. Lloyd.

POTAMOGETONS IN RELATION TO POND CULTURE. 283

ness and precision of a leaf cutter bee (fig. 74), though usually there is less regularity
of outline. .

On the floating leaves of P. americanus collected early in August were found eggs
of Paraponyx and of chironomid.* Those of Paraponyx covered broad areas of the
under surfaces of the leaves and presented the appearance of minute six-sided cells of
honeycomb, yellowish in color. In a few days the larve hatched and began at once to
feed and to cut portions from the leaves for larval cases. Fryer (1888), in connection
with his studies on P. fluwitans, mentions that the larve of Nymphula (Hydrocampa
potamogata) entirely destroy the floating leaves of this species, and thus indirectly
induce the development of fascicles of leaves, structures which are analogous to the
winter buds of P. obtusifolius. The eggs of the chironomid, which were found on the
leaves of P. americanus, were inclosed in small elongate cases blackish in color, suspended
from the edges and from the underside of the leaf, and from the petiole. These eggs
hatched within a few days, but their entire life history was not observed.

The leaves of P. obtustfolius harbor a large number of chironomids, and apparently
offer a valuable supply of food to many of them. A few of the larve were segregated
in small diskes and supplied with fresh leaves of this Potamogeton. An undescribed
species of Chironomus was reared. Cricotopus trijasciatus and Tanypus flavellus were
the most abundant species on the leaves.

Other plant parts besidesleaves wereeaten. The tubersof P. pectinatus and the burs of
P.crispus were devoured by thelarve of Paraponyx and bythe larve of the Chironomide.

The underground stems of P. pectinatus® are provided with large and numerous
air spaces (fig. 66), and these were found to be an important air-supplying source for
the Donacia larve. The larvee attached to the subterranean stems of this Potamogeton
were collected from the muddy substratum at North Fairhaven August 14, 1913.
Stems on which the larve were not attached showed, quite generally, the characteristic
punctures, or double scars, made by the caudal spines in tapping the air supply.

Jepson (1905) called attention to the value of the tubers of P. pectinatus in the
diet of our wild game birds. He says, ‘‘The diving ducks, such as the canvasback
and broadbill, eagerly seek these tubers, devoting most of their time to this pursuit
until the duck-shooting season opens.’”’ McAtee (1911) and Thompson (1913) in their
researches on the diet of wild game birds have shown that a large percentage of the
food taken is Potamogeton.

The stomach content of 5 canvasbacks has come under my observation recently.
One duck shot in October had been feeding in rich aquatic meadows where Potamo-
getons flourished with Myriophyllum, Elodea, etc. Its stomach was filled exclusively
with tubers of P. pectinatus. Four ducks shot at the close of the season in January
had apparently exercised a choice in the matter of food. Feeding in an abundant
mixed vegetation, they had selected only Potamogeton—P. Fresii and P. pusillus.
The parts of the plants available were the winter buds which at this season have
settled in the mud at the bottom along with the hibernacula of Myriophyllum, Elodea,
and other aquatic plants.

@ During subsequent observations in June, 1914, masses of eggs almost infinite in variety and number have been found attached
to the stems and leaves of the various Potamogetons. It would seem that these plants, diverse as they are in habit and form,
offer especially suitable conditions for the attachment of the eggs of aquatic animals. ‘The eggs of the water mite (Hydracarina)
are exceedingly abundant. The eggs of insects that have been recognized are as follows: Stratiomyiide, Coriscide#, Gyrinide,
Donaciine, Hydrophylide, Pyralide, Corduline (Tetragoneuria), Hydrobatide and Tricoptera.

» Since this observation was recorded Donacia larve have been found on the underground stems of P. americanus.

284 BULLETIN OF THE BUREAU OF FISHERIES.

SPECIES OF POTAMOGETON AND THE ANIMALS FORAGING UPON THEM.

To facilitate reference, a list is given of the species of Potamogeton, together with
the smaller animals which have appeared to be intimately associated with them. Other
forms of animal life were often found upon these plants, but none seemed to be so char-
acteristically on their own ground, so to speak, as the forms listed below. Those
animals are starred (*) which have been observed feeding on the living plant tissue.

List OF POTAMOGETONS AND SMALL ANIMAL ForMS ASSOCIATED WITH THEM.@

Plant. Animal.

PPE ATICTICATINS ca aissnclclepisie selele'seieisteicrre sts Insecta...) Diptera:

Chironomid (undetermined).
Lepidoptera:

Pyralidx—

E * Nymphula sp. (Paraponyx).

Tricoptera:

Leptocerida—

Two species.

Mollusca. . Ancylus.

Pe ATID LOMAS va jscornrerelotere stele elnis\elelsie vienie’e/eisielciersicte Insecta. .| Diptera:
Ephydride—
* Hydrellia sp.
Chironomide—
* Chironomus sp.
* Tanytarsus sp.
Tricoptera:
Undetrmined.
Lepidoptera:
Pyralide—
* Nymphula sp. (Paraponyx).
Mollusca. . Ancylus.

Ps PETIOLIACIS 2 9s :c10:0,i0je/01e16/0°s|s/sjeFe\ejo.e 6[0's%eiele{s ster sie Insecta. .| Diptera:
Chironomida—

* Tanytarsus sp.
PX CTISPUUSia1-wiclelslsieieislelealeisielelclsie eiaveheleiniccleteleleiereistets Insecta. .| Diptera:

* Chironomid (undetermined).
Lepidoptera:

Pyralide—

* Nymphula sp. (Hydrocampa).
IPS ZOSLELLLOLILIS 7. <je1es\afs'e'e/ols\s eieis’e)s\e's(elsisieisieleisiaiscis Insecta. .! Diptera:

Chironomida—

* Tanytarsus sp.
Lepidoptera:
Pyralide—
* Nymphula sp. (Paraponyx).
Tricoptera:
Undetermined.
PP ObttisifOltis 5 creieic c\-leiejsielalsisismcisia'e stoisiefehaleleloinisis Insecta. .| Diptera:
Chironomida—

* Chironomous sp.
Cricotopus trifasciatus.
Tanytarsus flavellus.

PsP Mectin ats saccsesccicctiee dese sseiccerecnecere Insecta. .| Diptera:

Chironomida—
Tanytarsus flavellus
Cricopteris sp.

Tricoptera:

Hydroptilide (in autumn).
Coleoptera:

Donaciine—

Donacia sp.
PAECtINAENS saa. 5 snisicinia seisiciciejslowleslninatalelsian parse Insecta. .| Diptera:

(Gigantic form) Chironomida2—
Chironomous sp.
Tanytarsus sp.
Crustacea...| Amphipoda (very abundant).

Gammarus.
Hyallela.
Eucrangonyx.
Vermes.. Nais (very abundant in autumn).
PY Robbitishis vorcicisc te steers vices eacelssisesica selene Insecta..| Diptera:
Chironomida—

* Chironomus aberrans.
Tanytarsus sp.
Tricoptera—
Undetermined.
Lepidoptera:
* Nymphula sp. (Paraponyx).

@ A Potamogeton which came under casual observation only, P. Epihydrus, may be mentioned as an important addition te
the plants actually foraged upon by insect larve. Several specimens of this species of Potamogeton, collected at Spencer Lake
in August, were quite thickly dotted with caddis-fly larve (Leptoceride), which were feeding upon the fresh green leaves.

POTAMOGETONS IN RELATION TO POND CULTURE. 285

The Mollusca—Planorbis, Limnea, and Physa—were common on all of the Pota-
Teogetons.

These observations on the animal life associated with the Potamogetons afford
an additional contribution to the biological relations of the Chironomide, Pyralide,
Leptoceride, Hydroptilide, and Ephydride, groups in which one or more members
have been observed in their feeding operations. Of these animals it has already
been recorded by Reighard (1894) that the Chironomide are an important fish food.
Scattered reference is made by others of the value of aquatic insect larve in the diet
of fish. The fact that these insects eat the living plant tissue of the Potamogetons
adds greatly to the importance of these plants from an economic standpoint.

CONCLUSION.

In all contributions bearing on the life conditions of the Potamogetons, the promi-
nence of these plants in the shoal waters has been recognized, and where special effort
has been directed toward the study of their life relations, an economic value has been
ascribed to them. The present investigation affords further evidence of the economic
value of these plants, and contributes the results of observation and experiment on the
cultivation of several species. These results warrant the expenditure of additional
thought and effort on what purports to be one of the most important resources of our
lakes, ponds, and streams.

19371°—vol 33—15——19

BIBLIOGRAPHY.

AGARDH, J. G.
1852. (Some observations on Potamogeton pectinatus.) Botanische Verhandlungen der K,
Schwedischen Akademie der Wissenschaften im Jahre 1852. Entry taken from a review
in Flora, 1854, p. 755-757.
ASCHERSON, P., and GRAEBNER, P.
1907. Potamogetonacee, Das Pflanzenreich, 4, 11, p. 184, text fig. 221. Leipzig.
Bape, E.
1909. Das Stiswasser-Aquarium, p. 896. Berlin.
BauMANN, E.
tg11. Die Flora des Untersees. Stuttgart.
BENNETT, ARTHUR.
1880. Notes on pondweeds. Journal of Botany, p. 380.
Britton, N. L., and Brown, A.
1913. Illustrated flora of the northern United States, Canada, and the British possessions. 3 vols.
New York.
CHAMISSO and SCHLECHTEND.
1827. Linnea, vol. 2, p. 157-233.
Cios, D.
1856. Mode de propagation particulier au Potamogeton crispus L. Bulletin de la Société Bo-
tanique de France, t. 3, p. 350-352.
CONSTANTIN, J.
1884. Recherche sur la structure de la tige des plantes aquatiques. Annales des Sciences Natu-
relles, 6° sér., p. 287-331, pl. 14-17.
1885. Observations critiques sur l’epiderme des feuilles des végéteaux aquatiques. Bulletin de
la Société Botanique de France, t. 32, p. 83-92.
1886. Etudes sur les feuilles des plantes aquatiques. Annales des Sciences Naturelles, 7° sér.,
t. 3, p. 94-162; pl. 2-6.
Davis, CHas. A.
1907. The flora of Walnut Lake. In: Report of Biological Survey of Michigan. Published by
State Board of Geological Survey as part of report for 1907, p. 217-231, pl. 60, 3 maps.
DuDLEY, WILLIAM R.
1886. The Cayuga flora. Bulletin of Cornell University (Science), p. 132. 2 maps.
DycueE, L. L.
1910. Ponds, pond fish, and pond fish culture. Part 1, Bulletin no. 1, State Department Fish
and Game, Topeka, Kans., p. 36, fig. 4.
ESENBECK, ERNST.
1914. Beitrige zur Biologie der Gattungen Potamogeton und Scirpus. Flora, bd. 7, p. 151-212,
text fig. 59.
FiscuEr, G.
1907. Die bayerischen Potamogetonen und Zaunichellien. Berichte der Bayerischen botanische
Gesellschaft, bd. 11, p. 20-162.

Fores, 8. A.
1880. The food of fishes. Bulletin Illinois State Laboratory Natural History, vol. 1, no. 3, p.
18-79.

1888. Studies of the food of fresh-water fishes. Ibid., vol. 2, p. 433-474.

1903. Vegetation. (In his: Plankton of the Illinois River.) Ibid., vol. 6, p. 236-252.
FRYER, ALFRED.

1888. Notes on pondweeds. Journal of Botany, vol. 26, p. 273-278.

1900. The Potamogetons of the British Isles. Parts 7, 8, 9, p. 56. pl.

286

6. London.

we

BULLETIN OF THE BUREAU OF FISHERIES. 287

Griick, Huco.
1905-6. Biologische und Morphologische Untersuchungen iiber Wasser und Sumpfgewichse.
2 vols. Jena.
Hankinson, Tuos. L.
1907. A biological survey of Walnut Lake, Michigan. Report of Biological Survey of State of
Michigan. p. 288, pl. 60, 2 maps.
Hart, C. A.
1895. Entomology of the Illinois River and adjacent waters. (In his: Lepidoptera, p. 164-183.)
Art. 6, Bulletin Illinois State Laboratory Natural History, vol. 4, p. 148-284, pl. 15.
Eni, E. J.
1898. Eleocharis melanocarpa, a proliferous plant. Bulletin of the Torrey Botanical Club, vol.
25, 00. 7, Pp. 392-394, pl. r.
1898. Potamogeton Robbinsii. Botanical Gazette, vol. 25, p. 195-196, pl. 15.
Ho irerty, G. M.
1go1. Ovule and embryo of Potamogeton natans. Botanical Gazette, vol. 31, p. 239-346, pl.
II-I1I.
Hut, E. D.
1913. The advance of Potamogeton crispus. Rhodora, vol. 15, no. 177, p. 171-172.
Irmiscu, THILO.
1851. Uber die Inflorescenzen der deutschen Potameen. Flora, no. 6, p. 81-93, pl. 1.
1858. Uber einige Arten aus der Naturlichen Pflanzenfamilie der Potameen, p. 56, pl. 3. Berlin.
Jerson, W. L.
1905. Where ducks dine. Sunset Magazine, February.
KERNER, A., and OLIVER, F. W.
1895. The natural history of plants, their forms, growth, reproduction, and distribution. 2 vol.
New York.
MaccILLIvRay, ALEX. D.
1903. Donaciine. (In his: Aquatic Chrysomelide, a footnote, p. 325.) Bulletin New York
State Museum, no. 68, p. 288-327, pl. 21-31.
McATEE, W. L.
i911. Three important wild-duck foods. Circular no. 81, Bureau Biological Survey, United States
Department of Agriculture, p. 1-10, fig. 19.
MicKLE, G. R.
1912. Possibilities of northern Ontario as a breeding ground for ducks (published by L. K. Cam-
eron), p. 3-8, text fig. 2. Toronto.
1913. The increase of the food supply for ducks in northern Ontario. Ibid., p. 17, text fig. 3,
figs. 1-5b.
MorcGan, ANNA HAVEN.
1913. A contribution to the biology of May flies. Annals of the Entomological Society of America,
vol. 6, no. 3, p. 371-413, text fig. 3, pl. 42-54.
MoronG, THos.
1893. The Naiadacee of North America. Memoirs of the Torrey Botanical Club, vol. 3, no. 2,
1892-93, p- 65, pl. 55.
NEEDHAM, J. G.
1907. Notes on the aquatic insects of Walnut Lake. In: A biological survey of Walnut Lake, Mich-
igan. Report of Biological Survey of State of Michigan, p. 252-271, fig. 19-21, pl. 1, 1 map.
PIETERS, A. J.
1894. The plants of Lake St. Clair. Bulletin of the Michigan Fish Commission, p. 10, 1 map.
Lansing.
1gor. The plants of western Lake Erie with observations on their distribution. Bulletin United
States Fish Commission, vol. 21, p. 57-79, text fig. 9, pl. 8. Washington.
Ponp, R. H.
1903. The biological relations of aquatic plants to the substratum. Report of United States Com-
mission of Fish and Fisheries, p. 483-526, fig. 6. Washington.
(In press). The larger aquatic vegetation. American Fresh Water Biology, p. 32, fig. 20.

288 BULLETIN OF THE BUREAU OF FISHERIES.

RAUNKIAER, C.
1903. Anatomical Potamogeton studies and Potamogeton fluitans. Botanisk Tiddskrift, vol. 25,
NO. 3, P. 253-280, text fig. 9.
REICHENBACH, LUDOVICUS.
1845. Icones Flore Germanice et Helvetice, vol. 7, p. 40, pl. 71.
REIGHARD, J. E.
1894. A biological examination of Lake St. Clair. Bulletin Michigan Fish Commission, no. 4, p.
60, text fig. 1, pl. 2, 1 map. Lansing.
SAUVAGEAU, M. C.
1889. Contribution a 1’étude du systéme mécanique dans la racine des plantes aquatiques les Pota-
mogeton. Journal de Botanique, 3° année, no. 4, p. 61-82, fig. 9.
1890. Observations sur la structure des feuilles des plantes aquatiques. Ibid., 4° année, p. 41, 68,
IN75/ 129300735 LOLs 220, 2 ai
1891. Sur les feuilles de quelques Monocotyledones aquatiques. Annales des Sciences Natu-
relles Botanique, 7° sér., p. 103-296. .
1894. Notes Biologiques sur les Potamogeton. Journal de Botanique, 8° année, p. 1, 21, 45, 98,
112, 140, 166, fig. 31.
SHELFORD, Victor E.
1913. Conditions of existence of aquatic animals, p. 58-72. In his: Animal communities in tem-
perate America. Chicago, University of Chicago press, 1913.
SHERFF, Earu E.
1913. Vegetation of Skokie Marsh. Bulletin of the Illinois State Laboratory, vol. 9, art. 11, p.
575-614, pl. 86-97.
SCHENCK, H.
1886. Die Biologie der Wassergewachse, pl. 2. Bonn. Entry taken from a review in Botanisches
Centralblatt, no. 24, p. 355.
TuHompson, H. D.
1897. Report on Plants. In: The biological examination of Lake Michigan. Appendix 1, Bulletin
Michigan Fish Commission, no. 6, p. 72-75.
THompson, R. B.
1913. Description of edible plants. In: The increase of the food supply for ducks in northern
Ontario. Toronto, p. 8-17, text fig. 2, fig. 1-sb.
TILBURY, Mary RvuTH.
1913. Notes on the feeding and rearing of the midge, Chironomus cayuge johannsen. Journal
New York Entomological Society, vol. 21, p. 305-308.
TITCOMB, JOHN W.
1909. Aquatic plants in pond culture. Bureau of Fisheries, doc. no. 643, 31 p., 32 text fig., 2 pl.
UspEensky, E. E.
1913. Zur Phylogenie und Okologie der Gattung Potamogeton. Moscou. Entry taken from a note
in Flora, bd. 7, p. 187.
WarmINnG, EUGENE.
1909. Oecology of plants. An introduction to the study of plant communities, p. 422. Oxford.
WIEGAND, K. M.
1898. Embryology of Potamogeton. Botanical Gazette, vol. 25, p. 116-117.
1899. Development of microsporangium and microspores in Potamogeton. Ibid., vol. 28, p.
328-359, pl. 24-25.
York, HARLAN.
1905. The hibernacula of Ohio water plants. Ohio Naturalist, vol. 5, no. 4, p. 3, fig. 3. Columbus.

EXPLANATION OF PLATES.

All of figures on Plates XXXIV-XXXIX, with the exception of the photo-micro-
graphs, are photographs of plants floating in water, in aquaria, or in specimen jars.

PLATE XXII.

1. Potamogeton americanus, seed, 114 times natural size.
2. Potamogeton americanus, seedling 5 days old, 1% times natural size.
Fic. 3. Potamogeton americanus, seedling 14 days old, 114 times natural size.
4. Potamogeton americanus, seedling of four months, !4 natural size; A, winter buds.
Fic. 5. Potamogeton americanus, germinating winter bud, natural size; the last two winter buds
belated in development. Aquarium specimen, January 24, 1914.
Fic. 6. Potamogeton americanus, rootstock with winter bud A, natural size. September.

PLATE XXIII.

Fic. 7. Potamogeton amplifolius, rootstock, 44 natural size; A, A, A, young shoots which continue
to grow slowly through winter. November.

Fic. 8. Potamogeton heterophyllus, submerged plant, 14 natural size; A, tuberous rootstock; B, B,
submerged shoots; a, 6, and c, details of leaves. May 4.

Fic. 9. Potamogeton heterophyllus, rootstock not tuberous, natural size. July 7.

Fic. 10. Potamogeton heterophyllus, tuberous rootstock, natural size. July 7.

Fic. 11. Potamogeton heterophyllus, tuberous rootstock, natural size; A, A, incipient shoots

November 17. a ee
LATE XXIV.

Fic. 12. Potamogeton heterophyllus, typical habit of land form, 114 times natural size. Terminal
portion of rootstock tuberous. July 7.

Fic. 13. Potamogeton heterophyllus, terminal portion of rootstock showing tendency to become
tuberous, 114 times natural size. July 7.

Fic. 14. Same, more advanced stage, 114 times natural size. July 7.

Fic. 15. Potamogeton heterophyllus, aquarium specimen, natural size; A, tuberous internode placed
in aquarium in November; B, B, new shoots and rootstock. January 26.

PLATE XXV.

Fic. 16. Potamogeton heterophyllus, aquarium specimen, natural size; A, B, C, D, new shoot and
tuberous rootstocks in various stages of growth. March 14.

Fic. 17. Potamogeton heterophyllus, aquarium specimen, natural size.

Fic. 18. Potamogeton perfoliatus, winter shoot showing elongation of internodes, natural size; A,
leathery scale; a, detail of scale. June.

Fic. 19. Growing tips of same.
PLATE XXVI.

Fic. 20. Potamogeton crispus, recumbent branch, showing development oi new shoots on old stem,
¥ natural size. March.

Fic. 21. Potamogeton crispus, spicular bur, 114 times natural size. Aquarium. July.

Fic. 22. Potamogeton crispus, denticulate bur, 34 natural size; a, detail of leaf with denticulate
base, 2 times natural size; 1, line delimiting starch storage.

Fic. 23. Potamogeton crispus, sprouting bur, 114 times natural size. March.

Fic. 24. Potamogeton crispus, denticulate bur with sprout, natural size. Aquarium. July.

Fic. 25. Potamogeton crispus, spicular bur with sprout, 14 natural size. Aquarium. July.

289

290 BULLETIN OF THE BUREAU OF FISHERIES.

Puate XXVIII.

Fic. 26. Potamogeton crispus, cutting showing bur development, 1% natural size; A, immature burs.
Aquarium. March 22—April 7.

Fic. 27. Potamogeton crispus, sprouting bur, showing development of rootstock and erect shoots,
natural size.

Fic. 28. Potamogeton crispus, shoot with spicular bur at base, natural size.

Fic. 29. Potamogeton crispus, similar structure rooting above tip of bur, natural size.

Fic. 30. Potamogeton crispus, bur formation at tip of branch, 114 times natural size. Aquarium.

PLATE XXVIII.

Fic. 31. Potamogeton crispus, subterranean stem, showing bur in axil of scale, natural size. June.
Fic. 32. Potamogeton crispus, sprouting bur, 34 natural size. March.

Fic. 33. Potamogeton zosterifolius, winter bud, 34 natural size. October.

Fic. 34. Potamogeton zosterifolius, long section of winter bud, 1% times natural size.

Fic. 35. Potamogeton zosterifolius, winter bud sprouting, 1% natural size. April.

PLATE XXIX.

Fic. 36. Potamogeton filiformis, habit sketch, 4 natural size. July.

Fic. 37. Potamogeton filiformis, detail of tuberous rootstock, 114 times natural size.

Fic. 38. Potamogeton pectinatus, young plant developing from tuber A, 14 natural size. May.
Fic. 39. Potamogeton pectinatus, series of tubers, slender form of Dudley, 34 natural size.

Fic. 40. Potamogeton pectinatus, tubers, showing details of early growth, 1/4 times natural size.

PLATE XXX.

Fic. 41. Potamogeton pectinatus, tubers of two successive seasons; A, old; B, young; C, erect stem;
D, subterranean stem; 114 times natural size.

Fic. 42. Potamogeton pectinatus, terminal portion of subterranean stem, 1% times natural size. A
condition which may be present from June to October.

Fic. 43. Potamogeton pectinatus, mature portion of rootstock bearing tuberous runners; A, tuber-
bearing runner; natural size. September.

Fic. 44. Potamogeton pectinatus, spray showing fruiting spike and tuberous runners, A, B, C, %
natural size; a, detail of runner; A, A, young green shoots.

PLATE XXXII.

Fic. 45. Potamogeton pectinatus, plant developed in aquarium, suspended in water, }2 natural size;
A, old tuber which produced plant; B, young tuber.

Fic. 46. Potamogeton pectinatus, sprouting seed, 3 times natural size.

Fic. 47. Same, later stage, 3 times natural size.

Fic. 48. Potamogeton pectinaius, seedling about ro days old, outer testa of seed removed, 114 times
natural size; a, inner hard testa, showing characteristic lid-like portion thrust open, 114 times natural
size; b, seedling with hard testa of seed removed showing foot-like expansion.

Fic. 49. Potamogeton pectinatus, seedling three weeks old, natural size.

PLATE XXXII.

Fic. 50. Potamogeton pectinatus, gigantic form of Dudley, runner B, from base of erect shoot A,
¥ natural size; C, tuber; D, D, D, young green shoots; 1, 2, secondary runners. November.

Fic. 51. Potamogeton pectinatus, gigantic form of Dudley, tuber devoid of scales, 144 times natural
size.

Fic. 52. Potamogeton pectinatus, gigantic form of Dudley, portion of a tuberous runner, natural size.

Fic. 53. Potamogeton pectinatus, gigantic form of Dudley, growing tip of secondary runner, natural
size.

EXPLANATION OF PLATES. 291

PLATE XXXII.

Fic. 54. Potamogeton pectinatus, gigantic form of Dudley, sprouting tuber, 114 times natural size.
Aquarium, February.

Fic. 55. Potamogeton pectinatus, gigantic form of Dudley, tuberous runner, natural size.

Fic. 56. Potamogeton pectinatus, spray, showing cases of chironomids, natural size. September-
November.

Fic. 57. Potamogeton Robbinsii, characteristic vegetative structure, rooting at nodes, 114 times

natural size. May.
PLATE XXXIV.

Fic. 58. Potamogeton amplifolius, rooted tip of branch, a propagative structure. April.
Fic. 59. Potamogeton crispus, germinating burs. October.
Fic. 60. Potamogeton crispus, burs in various stages of development. June.

PLATE XXXV.

Fic. 61. Potamogeton pectinatus, gigantic form of Dudley, erect axis of plant, 5 feet 2 inches tall,
bearing runner at base of stem. November.
Fic. 62. Potamogeion pectinatus, gigantic form of Dudley, portion of leafy spray showing tubers.

November.
PrateE XXXVI.

Fic. 63. Potamogeton obtusifolius, winter buds. October.
Fic. 64. Potamogeton obtusifolius, erect axes bearing winter buds. October.
Fic. 65. Potamogeton zosterifolius, sprouting winter bud. Aquarium specimen. February.

PLATE XXXVII.

Fic. 66. Potamogeton pectinatus, cross section through stem, showing numerous air spaces.
Fic. 67. Potamogeton Robbinsii, branch, showing points where dismemberment occurs, 1, 2, 3.
Fic. 68. Potamogeton Robbinsii, branch defoliated by larve of Nymphula sp. (Paraponyx). Larval

cases, I, 2, 3.
3 PLATE X XXVIII.

Fic. 69. Potamogeton Robbinsii, photo-micrograph of section through old stem, showing arrange-
ment of mechanical tissue. :

Fic. 70. Detail of fig. 69.

Fic. 71. Potamogeton crispus, photo-micrograph of section through stem of starch-filled vegetative
structure; a, cell with starch grains. November.

PLATE XXXIX,

Fic. 72. Potamogeton crispus, leaves, showing characteristic leaf mining of chironomids.

Fic. 73. Potamogeton amplifolius, leaves, showing characteristic mines of Hydrellia sp.; a, b, pupa
cases at end of mines. August.

Fic. 74. Potamogeton amplifolius; a, b, leaves, showing circular pieces cut away by larva of Nym-
phula sp. (Paraponyx); 1, larva in case; c, dead leaf, showing cases of Chironomus sp.

Fic. 75. Potamogeton americanus, spray showing attachment of cases of caddis fly (fam. Leptoce-
ride); a, case of caddis fly; b, stipule of leaf. June 31, 1914.

Pi ho
“oe er Pech Mic <i

we aes |

- ™ ae _ an ma
~_ SS nye 8s ie ai oes Any a ty farted “Soar

7 ie ey oth al

7 ple me jinn? ese nasal pula te Oa ie jo aanate a
—— © "Take Venter is taste, a Pa " arvrvit _ tbe hice: ai lalyt A
7 ws
7 : — anny b 7. a Pad
— WEDS da aw ute ae raat eta ey tae
“ae 7 a ae 7 ae ae aN ny pre ~ fee : Wes
7 - Se ot, «. Le me “ wee Sacer “ie 4 = ;
i, a : ana ve of F vet nie 9 ral teed tid ea a

' = a i MUA We ot) ie Hehe aes oie we =
ns | 7 ; - i uu sy ips “atunglg i cn ee i Liv ost i ib ie

- 7 . : _ : 9 ‘ i
a? 7 Ly om ee ? us cee af. ae SM i i
7 ie aes ve *s ee sees * ae)
Pes im re rh Ny Buns! os * ‘ - ht i ‘ a y rata 1, PRP: apy “ti hag? vad « : md
i pte i f '
™ : ‘T. en Am ‘i tat Ws fa pens ‘aga
: Uno ig Ut tomes. hak ies 2 Nia Mg =) a
7 : Oe 1 ; 7 - 7 ag
7 t 7 . “abr® F u a ape Cr | om pene oe 7 . ee ae wi 9 ae ae
: 7 in ma Pann Ve Packie Lt Oe a al er ye ae
7 % a 7 n ee » aii 4] en ey oe — _-) Pry ee foal » - ae
ee v aEPa by Papeete psi eae ven aa ves Gi Panay, interme hs al on oa
i ys, 3 Sg na a 4
: a 7 vg a a a © oe] 7
: 7 c efi 7 7 or W 7
aa ee Pat Musi tinione - a he @ Ae Al Bol, 43- Wh ae rap a ieee ar
7 _ : , «@ ce - ‘ Wn mens oth i He ai q Ree en auth a Haat eal Filed athe a i rh
} as ra re a vy Sa Moe: di nh Tb 2 ae
Pe ——'@ | Ri en we : a
7 _ - = os! = aly ee i= fi . 5 ae pea F es ls
— pas Sete 2 caays iy ty on wie et one Sih ai oh rr
7 a et ee a ee ar Ete aa
>. _ i a ie 4 mas) ba ti fag :
Pe tA ee uF rT Sind gee are enies fio! : ¥ yp tuauee® Loree ‘f. Nhl Aves
— - aT, a i a he i ay ; i. dite a
: - p i Dee i= ~
' ~ _ roe fen ; - Oi ahr are + a 7 1 a oe
— 7 " ; 1 : “< 78 _ :
| AN ee Sek ie! iN seen? m er a Pah ‘alee eae VY P'. a
‘’ a : - Sy oy nate a ‘ 4 ie é ie Avy i Ce? gail: 1a ih hee eli oe a:
y ee ) a : ; ir ” : ary ) 7 44 ost! Ashes : vm, +9 wr 7.
_ ; 7 7 i at = ar Ty AD - ee an 7 aa ” or ' Wes oh - 7 ar pall ‘ if. eens «
7 : i . a 7 ae wt Ay : ro eae ae, psy wine a] 7 i= 4 % “aS
7 : peels tf in i ts af eae am Niet - af wee ee ae : ta a" ta ome ry - 7 ee
- ie an? Phantge ‘w.6,, hh 5 ae ae
“ t : = _—
~~ 7 t , 7 ' - ~~ ; __ : f, a's : 7a i : or
=e ng) yoo iar i:
oy -* oy, 77 ow ag vil _ “=e " ar : _ --
: TF i) Pee 7 a :
-— .!) : : v
_ - i 0 im | : A man :
3 7 7 » Cay, > Vi >|! 7 7
\ cane
sy — e 7 ca
: - i
a —
7 7 ia i
. r + : :
~~. is
yas ae. 7 ' _
e 7 ,
- ‘oe > wa i ms MM
= a : a) ae a “a aa

| y Yer S A
| Z

b=
Va

4

JSOAME, Wie wp Ja Ica aiepuek PLATE XXV.

y

PEATE Sexe iL.

TBhopesey LW Ss IBS Ikan iKopesys

ar
: = <=)
oe Z
oF Ne uy
—~r a
FS f
= Ss
\ ew
i = il. 4
f= aH =,
Ao 5
2

Gy
tal
| j
eV,

Wes

PLATE XOXSVIT:

BULE Uso. Bar, Tors:

BULL Us osebs Ey, LOLs:

PLATE XXVIII.

=>

—

SSS

———

—

=

—S

—_——

aX

PLATE XXIX.

BULL Wn on Ba, LOrg.

Burm Seb. Hen rons: PLATE XXX.

Kite

BuLUy Wey Bake tons:

Nes

aa

Burt. Us Ss. Bs EH. 1913: PEATE: XOXO Vi

XXXV.

PLATE

Be Heron.

S.

Burt. Ul

62

PAD s xe KOOVEle

Burr Uso be bs. LOLS.

BuLE Us on Bak, LOLs: PLATE XXXVII.

PLATE XX XVIII.

iBjoncies AWM 18, ANG, acon , PLATE XXXIX.

GENERAL INDEX.
&

Page.
affinis, Gambusia, habits, morphology, and embryol-
Retctelelefetatstetetstaicteislatcivicletelaisietsistaletetercioletelelelaisierstalersiafatclats 181-190
Anchovia mitchilli, embryology andlarvaldevelopment 3-19
Bairdiella chrysura, embryology and larval develop-

MIPETAL che folatelaietayainteiataia’s/=/elniatalalatsinlels}ejeistsisieleieisteluie!=iainiaia(a|a‘alaieie 3-19
canis, Mustelus, sense of smell. ...........secseeceeecees 63-68
Chiryatira Hairaiellaccmnrcyeteereisismalcteisaieiicisinicisioeiesterters cits 4
Cynoscion regalis, correlations of weight and body meas-

ITCTTICTELS or eteialatciora sicfeletetataiclolefaeialoln/<fe]elatuia eleinistalelaletalelaieie 141-147
Diseases in fishes:

IAtrophied Cisse seem emicicstnisrcintelsieiecicieisieletesiersictcinetelesisine 201

(Chiftiieke gitel iii GN co sascoogecescocaosaenaoocedsos 205

MEG KODOMUS MSCHL -1n ats efelalelere elalinisieloweiniec cleeln sieise's viele 202

Myxosporidaniinfecttom Jace = vicie’s «esse eicinis <i claie viniais/siz 202,208

See also Sporozo6n parasites.

Dogfish, directive influence of sense of smell............. 63-68
Douglas Lake, Mich., ecological study of fishes in mid-

219-249

s0006 239

222

221

species, relations to those of other waters -+. 246
Embryology, Gambusia affinis................0.000005 181-190
Embryology and larval development, Bairdiella chry-

sura and Anchovia mitchilli..............0.seceeeeeee 3-19

Fishes, Douglas Lake, Mich
diseases. See Sporozoén parasites.

Parasites... .c-.ss---- « 193-214
sense of smell, dogfish....... ++. 63-68
viviparous, Gambusia affinis.. . 181-190
onaG (THC y A anadesoanmocoooad eo mca coo SNDONOCDORDECOOOD 193
Gambusia affinis, habits, morphology, and embryology 181-190
LSE CLOCIIE TIS ke 0201 Cl 12100 5 oteieicin'e)=(a)oloselalsia)einia/stoislaia/e|s/eisia/s/s/e/4 193
LGIIN TENS ai apaagacaseobeaadecootucbcocoso eco cOosouedone 193
King salmon, fat-absorbing function, alimentary tract. 153-175
COLI Mert oie RLV eT tes oteyetelatotetatelaisis’= siateisiataiwiaielatetelotaiatatelsiejsi=rs)a 91
fat stored, resorption during fast of spawning migra-
RL OME eratsieieiatatatatdacstetetsterc nia) alaisicieleye(avalaie's|ala]a]siniaiatetars steioyataia 73-138
fat loading, beginning spawning migration.:.......... 80
Sikeletalantscttlabure sam sniceyeiiem lnc ciieiewisis eieieieman merci 25-59
MAPATEST writ ITS. 5 AeshpoooassonpadogacgnDe Jddodopocsne 193

Page.
Michigan, Douglas Lake, fishes... ..............0e0005 219-249
Mitchillis An CHOVIA. ccsepetalsieiemiatis cise ste ercistatetsts’aicie sielelsia's 13
Morphology, reproductive organs, Gambuasi affinis... 181-190
skeletal musculature, king salmon..................-- 25-59
Mustelusicanis, Sense Of SIEM eerie ciniciels clcis(clsiels'cintsielsieie «le 63-68
Oncorhynchus tschaw iShares cies sleicteatelsicicie cela 25, 73,153
Pacific saltnom'saic aces sleleineeeeciaviedcieiclesiewienieiste cleiseicte 74
Parasites” SHOLOZOOM a \e.aiicisiasiele'e claleie vi siaiela'eleles'eisiniais aisle 193-214
Physiology, alimentary tract, king salmon............ 153-175
skeletal musculature, king salmon..................-- 25-59
Sense ot smell doptist serie vateticie sinicle sinteialelelcialsta/aeietcleiate 63-68
reproductive organs, Gambusia affinis.............. 181-190
Pond culture, Potamogetons...........0.scssecceceses 255-291
Potamogetons, pond culture. ............cecee eee ee eee 255-291
lconomicaspect Olercterscriccienisteesiet slelseraneleeieisteleteetetels 281
propagation, natural and artificial.................65- 268
SPECIES IN VESLIVALEM wecieicicisaseleie wicivie sie eleleisieleteieieieieiiociete 260
species, animals foraging O1.............eeeeeeeeeeeees 284
regalis, Cynoscion, correlations of weight and body
FTICASHTEIICTICS wretetelelele terior foteietetetetay=[elete lato seletai=istateleleteleletate I4I-147
Salmon, king, fat-absorbing function of alimentary
MELA CE ainya xi lelo'=(niais|asejele-s.c.efu(eleiein.cieivieleisisinivveiaivislsiefolete(e\siele 153-175
Colin bia RIV CG i. sieciccelseineenicleetsiniaieicis eratels love wisiaistalaisi= or
fat stored, resorption during fast of spawning migra-
LOIN. < srersjeleists qjatelofaielare afa'a aiaterctale sistas loreietote efeisieiatevw aye istere 73-138
fat loading, beginning spawning migration............ 80
Skeletaliriiscelatre veces clelsiclecislcieieleicielcicisiersieleleisre 25-59
sense of smell, dogfish, directive influence of............. 63-68
Skeletal musculature, king salmon..................2.05 25-59
Sporozo6n parasites, vicinity Woods Hole, Mass...... 193-214
Chloromyxum funduli...............-..+000- 205
Myxobolus musculi. . 202
Myxosporidian infection 208
WPTOLOZOA ais steyolala)ofatetal=tals'atetelsinlatsalerviats so
Squeteague, relations of weight and lengths. «+. 141-147
tschawytscha, Oncorhynchus.............-. + 25) 739153
Viviparous fish, Gambusia affinis...............2.0055 181-190
Weakfish, correlations of weight and body measure-
141-147
193-214

293

O

rent

. ct ott, Pn, ) Cntiyyelae it AUD © Mie thle eg YN (MA Ph
= ee Cee 2 Aes Yee) iiNet sae ee
_ tp sai AH ala, Art Toon fo 18) rae Veerticn Shkbed oes) ‘Yui Nad antalyil
: — | Sey eS, Arie er ete Gh Tres yynden. 1 Gea
me a : ~ Ls 6 ie aL ) i hy
ale SOA y 7 ae _— + - a 7 : a ‘ - BV yin
af Tce. Wisy — (1 Ly WY = « (x re ) “ @é in
7 ne Fi) PUTT ie ae | ., = iy oe] ) a) iph : : a a
% D4 S Ixy 5 él. . Saeed : on a
‘p> at y, 1 Snes a5 ij i Goes
i c, wy a » | ay 7 ra a an - Amati
~ 5-0) a eee oF ti aa - oad) 1) Bart
asl yh - a y - Fae b ai, ar ; _ la OTT A
heath on 7 a . > 7h 10 mts 1 24 oe * ee fre rae
: ES Vice Lay i he a AE) Trsatvay e7
- neu te + Soe it mas ns ar Wate Oat +e eg ets
a ead Te ‘uy 2 ee a mt vice: ~~ 2 @ al!) Vu! (ibis _~ : mao #3
re as is trie ‘fa My rs we ie SF Uh ant =
, iy eine D 4 be etpins ie. . : ' ig _ on
=~ S- a ia : a vi i ail rr a iD
uM ce YO ae wet) is) Wie ge =}
ie ws _ Ai. ae on f, Ot. es oe Oo ad sane fer Or
weg - a - ee m= Fr tog oe 2) a By rrr (Wea ple _
7 is : i
—_ ga” F ot = aa Eo = 7 nm elt stoked ihe Kah ea ae
oo 7 tif. s Wea te 7 ni : On ir eit) <a
3 - L Ses ue *4 Bs - fs | 7 , 7 is 7 = a 4
_ : ee fi ay yes 7S ie? — _—
: ' : = { “Al ty @* ani © a4 a
“ee fen "rs ae mi na Uy a ay eee eT ny no) 1h = {ion
1 _ Duos te ee ee ; ie aft mah
n) Hi nny ap (Sauls 115 itive rt ni it: Nous © leaflet

: is ‘ ly) hi © ay igi ane roa i a
: siawe aa. oe v pando tm) aL, aha alee POM, =; mt poeseiai otra

at ae
ll ou a Cars n er ane 2) ae Poth ALY iy Saat i ete _ Hines
— ay an F —_ y ro 7 ki er viv =egee aa «i°f c@0li ee
be Wey i ae aii By eedcaly range ee syst aly iid vee me waif
5 Ay enya any Peri iy cet ay gen ae ; ors ‘ Sante '
ue ae sa “trae? ihp4 St) an) ‘eg =e SURAT vue yan Peat dbind tite
ae hia es eal he LY - =. ay, fom : 1%) indy Dee
o iy — ent nig fi Tr os == om rs 7 , ai oe elit 9® Seeds aT
—_ ie if Tee MEU ce Se Dai a + . . aie
iit \ J ae :
: ;
4 ‘Py tr -
a : _ @ gas
fi oo 7 7 -
: an i i o
p ' ' ' ay
7 ’ ’
a - - _
f '
1A = -
Z =
‘ :
: -
1 7 a
7 A -
- o-
a
: ! S pelae 7
, : te
7
- 7

ho a ee.