[From The Johns Hopkins University Circulars, No. 111, May, 1894.] 
THE DEVELOPMENT OF THE FINS OF 
TELEOSTS. , 
ROSS GRANVILLLE HARRISON 
[This work, undertaken at the suggestion of Prof. M. Nussbaum, was 
carried on partly in the Anatomical Institute in Bonn, and partly in the 
Biological Laboratory of this University.] 
(Preliminary Communication.) 
Excepting the elasmobranch fishes, we have no complete knowledge of 
the development of the extremities of any group of vertebrates. The 
skeleton alone has received due attention. The muscular system of the 
limbs of the higher vertebrates has been supposed by recent writers who 
have touched upon the subject, to take its origin from masses of cells 
derived from the myotomes. These myotomic cells are in a general way 
to be regarded as homologous with the cells of the muscle-buds (Muskel- 
knospen), out of which the definitive muscles of the elasmobranch fins are 
known to develop. The following observations made upon the salmon 
(Salmo salar) render a modification of this view necessary. 
The Unpaiked Fins. 
The unpaired fins arise as proliferations of the mesenchyme cells, which, 
in the form of a loose mesh work, fill the median fin folds. The caudal fin 
is the first to appear, and is followed by the dorsal, the anal, and the adipose, 
in the order named. 
Shortly after the dorsal thickening has appeared, muscle-buds appear at 
the anterior dorsal angles of the myotomes of that region, and grow rapidly 
into the fin-rudiment, as has been described by Dohrn (Studien IX. Mit- 
theilungen aus der Zobl. Sla. zu Neapel, Bd. VI). These processes converge 
considerably towards one another, so that, while the middle ones project at 
right angles to the long axis of the body of the fish, those at each end of 
the fin cut the axis at an angle of about forty-five degrees. Cross sections 
show that the buds are solid; a few cells are enclosed by an epithelium 
of similar cells. The nuclei closely resemble those of the mesenchyme 
and the cell boundaries are indistinct. The buds are continuous with the 
cells of the lateral layer of the myotomes (cutis-plate). Similar buds grow 
out from the anterior ventral angle of the myotomes in the region of the 
anal fin. The tail fin also receives outgrowths from several of the terminal 
myotomes; the adipose fin never contains muscle-buds or muscle-tissue of 
any kind. As the buds grow farther into the fin-rudiment, their outer ends become enlarged and somewhat flattened against the epidermis, in which a 
considerable bulging is caused. The stalk now disintegrates, and its com- 
ponent cells can no longer be distinguished from the mesenchyme. In the 
meantime the nuclei which lie in the median half of each bud accumulate 
considerable cytoplasm as a first step towards differentiation into muscle 
cells. These masses of embryonic muscle cells now grow centripetally; 
they ultimately become the erector muscles of the fin-rays. The lateral 
half of the bud now loses its identity as a cell-mass, having become undis- 
tinguishable from mesenchyme, but very soon the cells which lie opposite 
the peripheral end of each bud accumulate more cytoplasm, and these 
masses also grow towards the body, remaining close to the epidermis. They 
ultimately become the superficial muscles of the rays, which in the adult 
take origin from the skin. By this time the mesenchyme has developed to 
such an extent that the muscle masses are not at all clearly defined, so that 
it is impossible to draw a sharp dividing line between nuclei which will 
ultimately belong to muscle and those of the connective tissue. 
By this time cartilaginous rods, alternating with the muscle-pairs, have 
appeared in the middle plane of the fin. These become the interspinal 
bones which support the rays of the dorsal fin; in the anal fin they are the 
interhaemals. The chondrification takes place centripetally. I shall call 
them ray-supports. 
Horizontal sections of this stage show clearly the serial arrangement of 
the various structures of the fin. Opposite the cartilaginous rods, the ecto- 
derm is constricted to a marked degree. At these constrictions mesenchyme 
cells have aggregated close to the ectoderm, forming loose strands of cells, 
one on each side of each ray-support. This is the beginning of the defini- 
tive depressor muscles of the fin-rays. 
The dermal rays now begin to develop, in lines which are distally con- 
tinuous with each erector muscle rudiment, the cells of which are not 
separated from those of the corresponding rays by any sharp dividing line. 
Considerably later a small nodule of cartilage is formed at the tip of each 
cartilaginous ray-support, which has in the meantime become considerably 
bent with its convexity forward. Each pair of dermal rays grasps with its 
basal end the corresponding cartilaginous ball, and a strong fibrous tissue 
binds them together. Muscles now become inserted into this mass, in such 
a manner that each ray receives one pair of each of the three muscles 
belonging to each segment of the fin. Anterior to the pivot on which the 
cartilaginous ball rests, the erector is attached; posterior to the pivot, the 
depressor, and the muscle which takes origin from the skin. The depressor 
and the erector arise from the ray-supports. The fin has now practically 
reached its adult condition except that the cartilaginous skeleton has not 
yet ossified. 
The above account holds good only for those segments which do not lie 
at the ends of the fin. In the first two or three segments the course of 
development is considerably modified, although the same definitive arrange- 
ment is reached. 
The number of myotomes which produce buds is variable, but as a rule 
ten or eleven reach the dorsal fin, and eight the anal. Both anterior and posterior to these, buds may be formed, but they do not reach more than 
rudimentary development. The number of these buds is very variable. 
When the cartilage has just begun to appear, and the muscle-masses are on 
their way to segregation from the surrounding mesenchyme, these rudi- 
mentary buds have disappeared, presumably having disintegrated into 
ordinary mesenchyme tissue. Both anterior and posterior to the regularly 
formed muscles, are paired masses of closely packed mesenchyme cells which 
form a distinct layer under the epidermis. These masses are undoubtedly 
derived both from the original mesenchyme of the fin and from the breaking 
down of the rudimentary buds. The posterior mass of each side differen- 
tiates into two muscles, one corresponding to the erectors, the other to the 
skin muscles. These become attached to the posterior ray of the fin. A 
depressor muscle is wanting in the case of this ray. 
Anterior to the first of the regularly formed muscles is a cartilaginous 
ray-support, and anterior to this are the masses or laminae of undifferenti- 
ated cells described above. In the middle plane, anterior to the first car- 
tilage, at the regular distance existing between the other ray-supports, cells 
aggregate, and later chondrify, forming another ray-support. That portion 
of the undifferentiated cell mass lying between this new cartilage and the 
one next succeeding it, segregates from the rest, and from it, muscles cor- 
responding to those of the region of the muscle-buds are developed. This 
process continues until the usual number of rays and muscles found in the 
adult fin is reached. In Salmo salar this is fourteen in the dorsal and ten 
in the anal. The anterior ray-support and its muscles are not so fully 
developed as the others. 
Each fin is innervated by a series of spinal nerves. The nerves of each 
fin are connected by a longitudinal commissure which is a branch of the 
ramus lateralis vagi. lam unable at present to trace the development of 
this interesting plexus. 
In late embryonic and in the adult stages the metamerism of the fin 
corresponds to that of the body of the fish. In earlier stages the fin is more 
concentrated, as exhibited by the strong convergence of the muscle buds 
which enter into it. This varies, however, greatly in different species and 
is a matter to which but little importance is to be attached. 
The Ventral Fin. 
The ventral fin appears considerably later than the median fins. The 
first traces of it to be seen are slight aggregations of cells in the body wall 
just below the ventral edge of the myotomes which lie in the region of 
the dorsal fin. About the same time the epidermis covering these parts 
becomes considerably thickened through multiplication of its cells. Before 
the aggregations of mesenchyme cells become very conspicuous, muscle- 
buds grow out from the anterior ventral angle of each myotome in this 
region. About six enter the fin-rudiment; those at each end of it pro- 
jecting very obliquely to the axis of the body, converge towards the middle 
of the fin. Dohrn (Studien IX, p. 401) draws a sharp distinction between the mode 
of origin of the musculature of the anal fin and that of the ventral. In 
this he is followed by Kaestner {Arch. f. Anat. u. Physiol. Anat. AM., 1892, 
p. 200). Dohrn remarks that while the anal fin derives its muscles from 
muscle buds, the musculature of the ventral originates “ ohne Yermittelung 
von Muskelknospen, direct durch Einwachsen der Musculatur vomUrwirbel 
aus, wie sich leicht an Lachs- und Forellenembryonen nachweisen lasst.” I 
am unable to confirm this statement as my preparations, both surface views 
and sections, show distinctly that the muscle-buds which grow into the 
ventral fins are similar to those which enter the anal, except that the latter 
are considerably larger. 
At this stage nerve fibres from the spinal nerves of corresponding seg- 
ments which give off muscle-buds may be detected in the fin. This is a 
very much earlier stage than the earliest at which nerves could be seen in 
the median fins. 
The mesenchyme proliferates rapidly, while the ectoderm is raised into 
a fold which projects from the ventro-lateral surface of the body, parallel 
to its long axis. The mesenchyme forms a compact mass lying under the 
epidermis; the region next to the somatopleure is filled with less densely 
packed tissue. The muscle-buds project far into the fin, but, unlike those 
in the paired fins, are separated from the epidermis by a layer of mesen- 
chyme. Very soon the buds disintegrate. 
The region, in which the first steps towards differentiation of muscle first 
appear, is the space previously occupied by the muscle-buds. From the 
very first there are no traces of metamerism in this muscle, although it 
is safe to assume that the cells from the buds take part in its formation. 
About the same time, at a corresponding position on the opposite or inner 
side of the fin, a similar differentiation takes place. It is not so likely 
that cells from the muscle-buds take part in this. Between these two 
muscle-layers the cartilaginous skeleton has by this time appeared. The 
development of the skeleton has received such thorough treatment at the 
hands of Wiedersheim and others as to render further mention of it here 
unnecessary. 
The fin rotates so that its line of attachment to the body makes an angle 
of about forty-five degrees with the long axis of the body. The inner 
muscle, of which the beginning was described above, becomes the pro- 
tractor or abductor profundus, and the outer, the retractor or adductor 
profundus. The superficial muscles develope before the twisting of the 
fin takes place. They are formed through differentiation of the mesen- 
chyme cells which lie between the deeper muscles and the epidermis. It 
is extremely improbable that cells from the muscle-buds take any part in 
their make up. The muscles and skeleton grow forward in the body wall 
between the ventral ends of the myotomes, so that eventually only a very 
small portion of each of the muscles lies in the free extremity. In embry- 
onic stages these muscle-masses are continued distally, without sharp 
dividing line, into the mother-cells of the dermal rays. The Pectoral Fin. 
The pectoral fin diverges from the primitive type more than the other 
fins, both in its definitive structure and in its course of development. It 
develops considerably earlier than the others, and lack of histological 
differentiation of the tissues at that time renders its study more difficult. 
The first trace of this fin is to be seen in a thickening of the somato- 
pleure; the thickening of the ectoderm and its fold arise later. This is in 
accordance with Boyer’s observations on Fundulus (Bulletin Mus. Comp. 
Zool, Vol. XXIII, No. 2). The thickened portion of the somatopleure 
is not confined to the “pectoral plate,” but extends to the portions of 
the splanchnopleure, on the same level, and through the nephrostome 
to the Wolffian duct. This thickened portion of the peritoneum is due to 
the cuboidal or columnar character of the epithelum composing it. Ante- 
riorly, laterally, and posteriorly to it, the cells flatten out. There is, just 
anterior to it, a portion of the body wall in which are numerous mesen- 
chyme cells, derived from the head mesoderm. Ziegler {Arch.f. Mikr. Anal., 
Ed. XXX) has regarded this as the rudiment of the fin. Study of the later 
stages shows, however, that this region lies completely anterior to that in 
which the fin develops. 
At a somewhat later stage, the cells belonging to the pectoral plate 
become much more distinctly columnar than the others; and, multiplying 
rapidly, soon become several layers thick, and are much more densely 
packed than those lying anterior to them. A thickening of the epidermis 
now takes place, which, unlike that of the ventral fin, consists in an 
increase in size of the individual cells and not in a multiplication of the 
same. At the crest of the prominence which the proliferation of the meso- 
derm has caused, the ectoderm is thrown into a fold, parallel to the axis of 
the body and extending through three somites. In cross section the struc- 
ture is triangular; the somatopleure, which extends out over the yolk sac, 
is its base, and is nearly horizontal. In profile the crest is semicircular. 
Through rapid proliferation of the mesoderm cells, the prominence becomes 
much more pronounced and soon the height greatly exceeds the breadth. 
The cells which lie near the base are not so densely packed as those lining 
the ectodermal walls. 
In the meantime the myotomes have sent out processes from their ven- 
tral growing edges, but instead of entering into the fin-rudiment, as given 
by Kaestner {Arch. f. Anal. u. Phys. Anat. Ahth. Jahry., 1892, p. 200) for 
Salmo and Boyer for Fundulus, they become greatly elongated, and grow- 
ing forward, give rise to the coraco-hyoid muscle, as has been described by 
van Wijhe (Verh. d. Konink. Akad. van Wdenschappen, Amsterdam, Deel 
22) for Pristiurus, and by van Bemmelen {Anat. Anzieger, Bd. IV) for 
Lacerta. The first myotome is at this stage quite rudimentary; the second 
and third lie entirely anterior to the fin which is on a level with the 
fourth, fifth and sixth. The anterior end of the Wolffian duct is opposite 
the middle or posterior portion of the fifth segment. The first myotome 
has no ventral process. The second, third and fourth send out long strands 
consisting entirely of cells from the cutis plate. These grow ventrally in the somatopleuric wall of the pericardial cavity. After a certain time 
the first one atrophies, the second and third bend forwards, and are followed 
by the fourth (from the fifth myotome), which grows straight forwards and 
slightly ventrally. The foremost one becomes attached to the base of the 
hyoid arch by means of a tendon, the stalks connecting the buds with the 
myotomes atrophy, and the three buds unite to form a muscle which 
takes origin from the membranous shoulder-girdle, and is attached to the 
urohyal. This muscle is still divided into three segments, at least in young 
fish, in which the yolk-sac is entirely absorbed. The last of these buds 
extends for one whole segment under the pectoral fin with the cells of which 
it is in close contact. Sections through this, a little anterior to the point of 
origin of the outgrowth, give such a figure as has been drawn by Kaestner 
(Fig. 32). This outgrowth is less well defined than the others, and in cross 
section it can often scarcely be distinguished from the mesoderm of the fin. 
It is not at all unlikely that individual cells may detach themselves from 
it and remain in the fin, but it is certain that, as a mass it lakes no part in the 
formation of the fin muscles. 
The sixth myotome has a ventral process, which, however, does not grow 
forward as the others do. It ultimately pinches itself off from the whole 
length of the myotome, and becomes an independent longitudinal strand of 
muscle fibres which runs dorsal to the attachment to the fin, but which is 
afterwards probably incorporated into the lateral muscle masses. The 
seventh and succeeding myotomes grow ventrally and are concerned in 
forming the ventral muscles of the fish. 
The changes that the fin has undergone are now considerable. The 
attachment has constricted considerably, at least in comparison with the 
free portion, which has become a fan-like expansion. With the absorp- 
tion of the yolk the fin is brought to the ventro-lateral surface of the body, 
and rotating on its axis, so that the line of attachment, instead of being 
parallel to the axis of the body, now makes an angle of about forty-five 
degrees with it, the anterior extremity is thus brought into a corresponding 
position with that described in the posterior. The internal changes that 
have taken place during this time are the differentiation of the central core 
of cells into cartilage, and of the proximal portions of the superficial mesen- 
chyme layer into muscle. It may be reyarded as certain that the cells which yive 
rise to these muscles originate from the somatopleuric thickening, and, as is the case 
with most of the muscles of the ventral fin, are in no way connected with the myo- 
tomes. At first there are but two muscle-masses, a primitive abductor or 
protractor, lying on the outer side of the cartilaginous skeleton, and an ad- 
ductor or retractor on the inner side. A superficial muscle is developed 
later from a mass of cells lying just within the fin between the deeper 
abductor and the inner epidermic wall. The superficial protractor or ab- 
ductor does not appear till much later and probably arises through delami- 
nation from the primitive muscle, though X am not perfectly convinced 
of this. 
The nerves of this fin are distinguished very early in its development, 
just as in the ventral, i. e. before any differentiation of the tissue has taken place. They arise from the first four spinal roots. The first root corresponds 
to the second myotome, and its ramus ventralis unites soon with the second 
nerve to form the hypoglossal. This gives off a branch to the fin-plexus 
and one to the coraco-hyoid muscle. The arrangement is completed very 
early in the life history of the individual and seems to be quite typical 
for the teleosts. 
The mesodermic structures of the median fins are derived from mesenchyme 
cells derived from the sklerotome, and from muscle-buds, which are out- 
growths of either the dorsal or the ventral edge of the myotomes. To a 
certain extent these fins retain their primitive metamerism, in that each 
muscle-bud may be traced directly into a certain muscle of each segment of 
the fin. Other muscles are derived from cells which are indistinguishable 
from mesenchyme cells, and which are in all probability to some extent de- 
rived from the same. The segmentation of the extreme anterior portion of 
the fins is secondary, although in the adult no difference can be seen between 
the two portions. 
Recapitulation. 
The ventral fins show in the early stages of development traces of a similar 
metamerism. The buds, in this case, soon disintegrate and in the space 
occupied by them a single muscle-mass develops, the adductor or retractor 
profundus. The other three muscles of this fin are developed from cells 
which have arisen from the somatopleure and perhaps also from the sklero- 
tome. This condition in the teleosts seems to be a step between the Elas- 
mobranchs and the Amphibia. In Triton a few isolated cells break off from 
the ventral edge of several myotomes and mingle with the cells of the pos- 
terior extremity which are, however, mostly derived from the somatopleure. 
According to Paterson {Quart. Journ. Mic. Sci., Vol. XXVIII), the myo- 
tomes in the chick take no part in the formation of the muscles of the 
limbs. Kaestner has cast doubt upon this statement, but it is doubtful 
whether his grounds for so doing are sufficient. 
The pectoral fin is derived entirely from somatopleuric cells. The muscle 
buds of this region are greatly modified and take part in the formation of 
the coraco-hyoid muscle. 
I wish to postpone the full discussion of the meaning of this diversity in 
the origin of the muscles, until some observations on other forms are com- 
pleted. It is in all probability to be referred to delay in the differentiation 
of the component parts of the fin until they take up their position within it. 
In other words, instead of so much connective and skeletal tissue, and so 
much muscle being contributed to the fin, it receives cells which still retain 
the potentiality to become any of these, and their position with regard to 
surrounding cells rather than their origin determines their ultimate fate.