420 
BULLETIN OP THE UNITED STATES FISH COMMISSION. 
Copulation takes place in Cymatogaster during June or early July , although the eggs 
are not fertilized till the following December! This fact strongly bears out my opinion 
expressed in the American Naturalist (1890, p. 925), that the small size of the eggs of 
Cymatogaster is due to a hastening in the process of maturation, and the consequent 
nonformation of yolk. (See chapter on the egg and segmentation, p. 421.) The normal 
or former period for maturation of the oviparous ancestors of Cymatogaster very 
probably coincided with the present time of copulation, or nearly with the present 
time of extrusion of the young; the eggs, however, which should be fertilized by these 
spermatozoa, have just been set free in the shape of living young. They, therefore, 
remain in the ovary in a dormant condition till the next series of eggs become mature, 
which they fertilize. It is difficult to imagine why the maturation of the male has 
not been similarly hastened with the hastening of the maturation period of the female, 
for the process must have been a gradual one. We must imagine a time when the 
result of the ovarian gestation was much less perfect than now, and the larvae were 
freed in a much less mature stage than at present. Under such circumstances it 
would be quite possible for a number of spermatozoa to live through the period of 
gestation, and protract their stay in the ovary till other eggs were mature, and it is 
possible that the presence of spermatozoa may incite the eggs to the earliest 
maturation possible. If the eggs of a given season should be fertilized -before the 
maturity of the males the spermatozoa introduced during the subsequent maturity 
of the males would necessarily remain in the ovary till the next series of ova were 
mature or else be destroyed. The hastening in the maturation of the ova of an 
oviparous fish would not be followed by the disastrous consequences to the offspring 
that would result in the too early maturation of an oviparous egg, and in the conse- 
quent reduced amount of food material, since the food necessary in the eggs of the 
oviparous fishes for their larval existence is supplied by the ovary. 
It is in some such manner that we must explain the peculiar conditions existing 
in Cymatogaster. 
Since the statement that copulation takes place about five months before the eggs 
are ready to be fertilized is somewhat paradoxical, it is perhaps best to state upon 
what facts I base this conclusion.* 
During the period of segmentation few males are found with the females, which 
are then in shallow water, near wharves, etc. The testes of the males at this season 
are small and evidently during their period of physiological rest. The latter part 
of May I found a male with enormous testes, which I supposed to be iu a patho- 
logical condition. Shortly afterward, May 31, 1891, I obtained in San Francisco a 
large number of males, all of which had enlarged testes, but none of which were 
mature. The males were now common in shallow water. The next lot of males was 
obtained nearly two months afterward, July 29, 1891, at San Diego. In these males 
the testes were nearly reduced to their resting condition again. On examining a 
* This was written before Jordan’s and Gage’s papers on Diemyctehis appeared, from which it is 
evident that the spermatozoa can live a considerable time in the female Diemyctelus. A still more 
striking observation has been made by Bumpus in invertebrates. In the young lobster the transfer 
of spermatozoa may occur a year or even two years before fertilization takes place. I have not seen 
his paper and do not know whether Bumpus has demonstrated that the spermatozoa live from the time 
of the transference to the fecundation of the first lot of eggs or whether a new transference takes place 
