VIVIPAKOUS FISHES OF THE PACIFIC COAST. 
413 
examined, fresh. In this way many stages of growth of the egg could be observed 
in situ. The ovaries containing embryos or larvae are slightly transluceut and, after 
some practice, can be distinguished from those not yet mature.* 
Methods of studying living larvae. — The same process of procuring the larvae is 
used as for obtaining the eggs. The very young are, however, much more difficult 
to see than the eggs, and it is frequently necessary to spread out the tissues and ex- 
amine them microscopically before the larvae can be found. It is to be borne in mind 
that in this species the maximum number of eggs maturing in a season is only 22, and 
since the eggs are minute it would frequently be difficult to determine whether an 
ovary was immature or contained young embryos if it were not for the slight difference 
in the ovary itself as mentioned before. In these earliest stages the larvae are very 
sensitive to the condition of the mother. It has frequently happened that the larvae 
have become distorted if the mother had been confined in a pail of water for half an 
hour and before she showed any signs of exhaustion. They are most sensitive when 
the circulation is just established, but are frequently found to be nearly dead after they 
have attained 20 mm. in length, if the mother has been confined in standing water for 
several hours. For many stages this fact is very useful, since the young can be thus 
stupefied and studied more readily. For the latest intraovarian stages this method 
again produces distortions and coagulations in the vessels of the fins before the young 
is sufficiently stupefied. Partly asphyxiated larvae, 10 to 20 mm. long, can frequently be 
revived by simply placing them in some of the ovarian fluid on a glass slip, where they 
are exposed to the air. The method of removing a blood clot from the blood vessels 
ean thus be observed very readily. 
Connection of the developing egg and larva with the ovarian structures . — Part of 
the maturation processes of the egg are undergone in the follicle, but in all probability 
the egg is freed from the follicle before segmentation. Unsegmented eggs have been 
found free between the ovarian lamellae. The larvae always fall out of the lamellae if 
the ovary is immersed in any preservative fluid, and in the older stages the young can 
be seen to change head for tail in the ovary, so it may be safely said that the eggs 
and young are, during no stage of their development, connected with any portion of 
the ovary. Sections of some stages of ovaries containing eggs and young tell the 
same story. 
Position of larvae in the ovary . — Girard stated that in some species the embryos were 
regularly arranged in the ovary and that in Hysterocarpus traski (1859, p. 1G) in which 
the young were nearly ready to be born u all of them had their heads in the same direc- 
tion as that of the mother, a circumstance for the first time noticed.” That the larvae 
are not definitely arranged in the species examined by me has already been mentioned 
and is emphasized by the fact that in later stages they can change their position. 
* It will be found to be more difficult to study tbe different stages of the developiug egg of these 
fishes than pelagic eggs, or even mammalian eggs, because the age of the contained eggs can not be 
known, as in the case of mammals, and the different stages,- if they are not incidentally procured, must, 
therefore, be sought by chance, a process which often necessitates the examination of many individ- 
uals and consumes much time. To add to the difficulty, the fishes, after heavy rains or a slight fall 
of temperature, seek deeper water and cannot be procured. Stages thus lost had to be sought in the 
next series (smaller specimens) to mature— a process which did not always prove successful. To 
balance this, there is an almost unlimited supply of specimens when the conditions are favorable. 
