Medical Officers of the Army of India.

55

      In some of the Embiotocidœ, according to Blake and Ryder, the vertical
fins of the embryo are greatly enlarged and have a special blood-supply, by
which means oxygen, and probably nourishment also, are absorbed from the
vascular ovarian wall of the mother, these specially vasculized fins being, in short,
a sort of fœtal placenta.

      In other Embiotocidœ, according to Eigenmann, the ovarian wall -secretes
a nutritive fluid by the dissolution of its own epithelium, the fluid being further
enriched by spermatozoa left over from the original impregnation by the male,
this fluid is ingested and assimilated by the embryo, while respiration is carried
on partly by the general surface of the body and fins (through which also some
nutriment may be absorbed) where they come in contact with the ovarian wall,
and partly by the current of ovarian fluid passing through the alimentary canal.
According to Eigenmann the surplus spermatozoa of the male parent here play
an important part, for not only do they form part of the food-supply of the
developing young, but by their activity they prevent the ovarian fluid from
stagnating, by setting up useful currents.

      In Anableps alone, in which as in other Cyprinodonts the yolk sac is vas-
cular, do we find anything at all approaching the yolk-sac placenta of the
Elasmobranch fishes. In Anableps, according to Wymann, the yolk-sack
increases greatly in size after all the yolk has been used up, and its surface
becomes covered with small papillæ or villi whose function is probably to absorb
nutriment furnished by the wall of the maternal ovary.

      On the " Investigator" we have discovered a small section of the family
Ophidiidœ, including the three genera Dipiacanthopoma, Saccogaster, and
Hephthocara, to include species that are truly viviparous, and I believe, from
significant external characters, that all the species of this section will be found
to bring forth their young alive.

      Of the three members of this family noticed by the " Investigator " to be
viviparous, namely Diplacanthopoma rivers-andersoni, Saccogaster maculata,
and Hephthocara simum, only the first named is represented by material suffi-
ciently well preserved for satisfactory investigation, but the arrangements seen
in it are certainly the same in Hephthocara simum.

      I have described the intra-ovarian gestation of Diplacanthopoma rivers-
andersoni, in the Annals and Magazine of Natural History for August 1895.

      Both ovaries of a specimen over 15 inches long were distended and had
exceedingly vascular, though thin and tough, walls.

      On opening an ovary the contents shelled out without the slightest forcing.

      The contents were embryos and unchanged eggs, the embryos forming a
thick surface layer, the unaltered eggs forming a sort of kernel.

P