422 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
not been observed in any other egg (see under yolk nucleus , p. 423). The proportions 
of yolk and germ as found in other teleosts being reversed, at first glance the two 
structures might be taken for each other, and, iudeed, I observed the yolk in a favor- 
able egg for a considerable time, expecting it to show some signs of segmentation, before 
I became aware of the true state of affairs. The axis of the egg is slightly longer than 
the transverse diameters. The yolk, being the smaller segment of the sphere, is of a 
smaller diameter than the overlying germ. 
Several eggs measured have the following dimensions in jj. 
Total diam- 
eter from 
zona to 
zona. 
Axis. 
Diameter 
of germ. 
Diameter 
of yolk. 
Height of | 
yolk. 
Yolk 
| nucleus. 
300 
276 
258 
230 
98 
231 
213 
89 
249 
179 
53 
36X45 
244 
200 
The proportion of the yolk is seen to vary considerably, as is also the size of the 
germ (200 to 258) — I did not look for extremes — all of which shows that the size of the 
egg of Cymatogaster is in a state of unstable equilibrium. 
Intimately connected, of course, are the small size of the egg and the great reduc- 
tion of deutoplasm. The nearest approach to the size of the egg of Cymatogaster is 
probably that of Clupea , in which the germ also has a comparatively large size. There 
are very few yolk spheres in the germ of Cymatogaster , so that the small amount of yolk 
present if evenly distributed in the germ would scarcely prevent it from undergoing 
complete and regular segmentation. As large a proportion of deutoplasm is probably 
found in some holoblastic eggs. The meroblastic condition of segmentation is here 
a purely ancestral trait. The small size of the yolk is unquestionably the result 
of viviparity and may have been brought about through natural selection. We need 
only hasten the time of maturation of the average teleostean egg by two or three months* 
to produce an egg not unlike that of Cymatogaster , for, as I have shown elsewhere, the 
growth of the teleostean egg — i. e., yolk formation — takes place chiefly during the two 
or three months before maturation. Should any eggs of an oviparous fish show any 
tendency to so early maturation the young would invariably perish during its larval 
stages, owing to the lack of food. Not so in a viviparous species. As will be seen later, 
in Cymatogaster intracellular digestion of the ovarian fluid takes place before the embryo 
is freed from the zona, which occurs very early. Eggs of such viviparous species with 
a tendency to early maturation, having all food necessary for their development in 
the surrounding liquid, would not necessarily perish, but would in fact have-an advan- 
tage over eggs maturing later, which they could deprive of nourishment, if they did uot 
feed on them. It would need but the hastening of the processes of maturation, as 
stated above, to produce an egg similar to that of Cymatogaster. 
The teleostean ovary has a period of physiological rest and a period of great physi- 
ological activity, the latter in oviparous forms occupying the months just before 
oviposition. A hastening of the process of maturation and reduction of the number 
of eggs would therefore deprive viviparous forms of the type of the Embiotocidce of 
their periods of greatest activity; but they are not deprived of this activity because 
the young, where only a few are born, remain for a long period in the ovary and are 
