926 The American Naturalist. [October, 
The embryo at hatching is in a very r 
tail is not yet free from the yolk and is in 
tissues over the base of the yolk lengthen and the embryo begins to 
straighten itself. The tail does not begin to grow out until some time 
later. Several days are undoubtedly consumed after hatching in reach- 
ing the equivalent stage of oviparous fishes at the time of hatching. 
The hastening in hatching is due to the absorption by the embryo 
of food supplied by the ovary. The embryo thus soon fills the cavity 
left between the egg and the membrane at maturation, and the mem- 
brane then bursts. The hatching process is therefore decidedly differ- 
ent from the hatching of oviparous fishes. 
If the newly-hatched oviparous fish is compared with the corres- 
ponding stage in Micrometrus another ancestral trait will be discovered 
in the latter. While the yolk in the latter is minute as compared with 
that of the former, the yolksack is just as large. The yolk fills but a 
very small portion of it ; almost all the yolksack is occupied by the 
enormous pericardial chamber through which the thin tubular heart 
passes upward and forward. 
The development of the whole family of fishes of which Microme- 
trus is a representative (the Ditremidae) is characterized by the hyper- 
trophy of the hind gut, first pointed out by Ryder. This is already 
well developed at hatching, and at the time Kupffer's vesicle appears 
it very probably extends considerable beyond the point where the 
latter is situated. In all stages until the last this hind gut protrudes 
greatly from the ventral profile. 
Yolk absorption is not completed until quite late in the develop- 
ment, that is, not until the embryo has reached a stage homologous 
with the corresponding stage of oviparous fishes at the time of the 
completion of yolk absorption. That the yolk is not sufficient to 
account for a tithe of the growth of the embryo during this time goes 
without saying. The delay in the total absorption of the yolk is 
simply another trait inherited from oviparous ancestors. 
The food absorption is of considerable interest. The earliest ab- 
sorption is undoubtedly similar to the preplacental absorption of food 
by the embryos of placentalian mammals. At the time the embryo 
has developed a tail and circulation, and before the mouth is open, the 
first giU slit is open, and a continuous stream of mucus mixed with 
spermatozoa enters the intestinal tract at this point, and passes out of the 
anus apparently unchanged. 
With the opening of the mouth villi appear in the hind gut. Food 
secreted by the ovary is now taken in through the mouth, and can be 
