386 Prof. Agassiz on Viviparous Fishes from California. 
differ from all the other viviparous species known to us. There 
is another feature about them of considerable interest, that while 
the two adult differ markedly in coloration, the young have the 
same dress, light yellowish olive with deeper and brighter trans- 
verse bands, something like the young trouts and salmons in their 
Parr dress. ‘The transversely banded species may therefore be 
considered as inferior to the other, since it preserves through life 
the system of coloration of the embryo. — 
It will be a matter of deep interest to trace the early stages of 
growth of these fishes, to examine the structure of the ovary. and 
the eggs before fecundation takes place, etc., etc. The state of 
preservation of the specimens in my hands, precluded every such 
investigation. 
hough I know thus far only one single genus of this type, I 
do not think it right to combine the generic characters with those 
of the family, as is generally done in such cases, as I would also 
object to the practice of omitting any specific characteristics where 
only one species is known of a genus. This showsan entire mis- 
apprehension of the relative value and subordination of the char- 
acters of animals. I would therefore characterize as follows the 
genus 
E'mbiotoca, Agass. 
_ Body much compressed and elevated. Head small, with scales 
only on the cheeks and opercular pieces. Teeth in both jaws, 
jaws, and arranged like pavement. Dorsal fin with nine or more 
spinous rays. ‘T’he first three rays of the anal fin, spinous, and 
much shorter than the following articulated rays, which are 
always finer and more numerous than the corresponding rays of 
the dorsal fin. The lateral line is continuous to the base of the 
caudal fin. Whether the peculiar mode of reproduction is a 
family or a generic character, remains to be ascertained by further 
Investigations. It is however probable that with some slight 
modifications it will be found the same in all the members of the 
family. 
Some differences between the two species observed, might ren- 
der it doubtful whether they ought to be considered as belonging 
to as many distinct genera or not. But we know that in genera 
differing greatly from others, ®he range of the specific differences 
is also wider than in genera with many species; so until I am 
taught differently by new discoveries, I would refer them both to 
the same genus. Such doubts could scarcely be enter- 
tained respecting families with many genera, where a standard to 
estimate genuine generic differences is easily obtained. 
