146 On a new Species of Viviparous Fish. 
The ovaries of Diplacanthopoma Rivers-Andersoni consist 
of a densely packed mass of embryos and ova enclosed in a 
thin but extremely tough capsule. The capsule is abundantly 
supplied with blood by the ramifications of +a large branch of 
the mesenteric artery. 
There is no attachment or adhesion of any kind between 
the ovarian capsule and its contents. 
The embryos form a thick surface layer immediately 
beneath the capsule, enclosing a central mass of largish (a 
little over 1 millim. diameter) ova, which consist entirely of 
yolk-spherules, withont any trace of an embryo or even of a 
germinal area. 
Whether these unchanged ova would have developed sub- 
sequently to the birth of the present superficial layers of 
embryos, or whether they were destined for the intra-ovarian 
nourishment of the present embryos, are questions which it 
is impossible in an isolated case to discuss; but from their 
large size, which precludes any suggestion of immaturity, it 
would seem probable that they were intended for present use 
rather than for a future brood. 
The embryos, which are long and eel-like—6 to 8 millim. 
long—lie matted together, and firmly adhering to one another 
by their tails by means of a coagulated secretion. 
The vertical fins only are represented by a long fold of 
integument, which runs from the occiput, round the tip of the 
tail, to the vent. This fold of integument consists of layer 
upon layer of large-nucleated cells. ‘The remains of the yolk- 
sac are enclosed in the abdomen, causing a bulging of the 
abdominal wall along its whole length, from the throat to the 
vent; but there is no vitelline constriction or pedicle. 
T am inclined to think that the vertical fold of the integu- 
ment, which is really only an extended sheet of embryonic 
cells, is an absorbent (nutritive) surface, somewhat as in the 
embryos of certain fishes of the family Embiotocidee, in which 
the interradial membranes of the vertical fins have been 
shown to play the part of a foetal placenta. 
In the present case, however, there is no vascular con- 
nexion, at any rate on the foetal side; and I am inclined to 
think that the nutrient material is absorbed not so much from 
the thin tough ovarian capsule as from the ovary itself, 
namely from those ova in which no trace of a germinal 
vesicle can be found. 
An embryo taken at random measures 8 millim., namely 
2 millim from the snout to the vent and 6 millim. from the 
vent to the tip of the tail. 
