410 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
the parent. I however was fortunate enough to obtain a fish, in the uterus of which I discovered a 
considerable quantity of fluid, and on collecting it and submitting it to chemical tests, I found that this 
fluid contained a considerable quantity of animal substance, resembling, to a certain extent, some of the 
compounds that are formed from albumen during the progress of digestion. The fluid was of yellowish 
color, translucent, deposited on standing some small globules which under the microscope strongly 
refracted the light, were not altered by acetic acid, but dissolved in ether; probably fat globules; 
when heated there was no coagulation, although the fluid was not quite so clear; solution of HgCL, 
caused no precipitate ; tannin in solution caused a yellowish precipitate. In adding ether to a por- 
tion of the fluid, there was a free disengagement of gas, a white flocculent precipitate was formed, 
and on allowing the vessel to stand the fluid separated itself into three portions : the upper portion 
consisting of pure ether apparently, then a layer containing white flocculi, which occupied about the 
fourth part of the fluid, and below this the remains of the original fluid, but little altered in appear- 
ance. There can, I think, be little doubt but that it is through the medium of this fluid that the 
foetus obtains its nourishment. The considerable portion of animal matter it contains, and that too 
in a state particularly fitted for absorption and for conversion into tissue, fits it for furnishing the 
foetus with the elements necessary for its growth by absorption through the large surface of capillary 
vessels which are found in the vascular digitations that surround the foetus, and which are constantly 
bathed in the fluid. The difficulty that up to the present time has attended every attempt to trace 
the connection between the parent and foetus in these embiotocoid fishes is owing, in the first place, 
to the extreme delicacy of the vascular digitations of the foetus, which prevents their being observed 
in preserved specimens, and also to the fact that in almost every case the fluid secreted by the uterus 
is entirely expelled by the violent struggles of the fish when removed from the water, so that it was 
almost by a rare accident that I succeeded in obtaining any. I hope, however, during the coming 
season to be able more fully to carry out these researches. (San Francisco, January 21, 1867.) 
Later lie published the following note, v, 371-372: 
Some months since I presented a communication to the Academy pointing out the manner in which 
the foetus of the embiotocoid fishes was nourished whilst it was being developed within the ovisac. 
I there stated that the ingress of water into the ovisac would not take place at all freely, as the organ 
communicated with the surface by a narrow canal surrounded by muscular fibres. This structure of 
the oviduct would evidently oppose an obstacle to the entrance of the semen into the ovisac for the 
purpose of impregnation, unless some means exist by which the ventral surfaces of the fish can be 
maintained in contact during the act of copulation, as the penis consists of a slightly developed 
tubercle which can not penetrate for any distance into the oviduct. From the direction of the orifices 
of the penis and oviduct it is evident that anything like a perfect contact of these organs can only be 
maintained whilst the fishes are in a reversed .position, so that the head of one fish is towards the tail 
of the other. In order that contact may be maintained whilst in this position, we find the anal fin of 
the male fish furnished with certain appendages which enable it to give a firm hold to the ventral fins 
of the female, so that close contact of the ventral surfaces can be maintained. 
These appendages are of two kinds. In Emiiotoca, Damaliclithys, and some other genera we find a 
well-developed mammary elevation situated near the anterior part of the anal fin on both sides, termi- 
nating in front by a teat-like process. In Amphisticlius, Holconotus, and some other genera this mam- 
mary appendage is wanting; but its place is supplied by a bony transverse plate with serrated edges, 
inserted in the fin some distance farther back and parallel to the fin rays. In addition to these plates 
there are also found cartilaginous ridges with roughened borders, placed in front of the plates and 
running parallel with the edge of the fin. I think there can be no doubt but that these fin append- 
ages serve the purpose I have assigned to them, for on placing the fish in the reversed position, with 
the orifice of the oviduct and penis in contact, it will be seen that they enable the ventral fins of the 
female to secure a firm hold on the anal fin of the male, so as to keep the fish in contact during the 
process of copulation. At the season of copulation the anterior surface of the anal fin in the male 
becomes covered with a thick layer of firm epithelium. As this commences at a short distance from the 
ventral attachment of the fin, a well-marked groove is formed at the base of the fin, which affords an 
additional hold for the ventral fin of the female. After the season of copulation is over and the testi- 
cles regain their quiescent state, this epithelium almost disappears. At the same time the mammary 
sack diminishes very much in size, so that when the testicles are reduced to their smallest size hardly 
a trace of the sack remains. One or the other of these forms of appendages have been found on the 
anal fin of the male in all the species of embiotocoid fishes I have examined. 
