30 
BLIND FISHES IN GENERAL. 
placed in a fold of the skin of the head, and probably of little use, 
as these young remain buried in the sand ; but as they attain ma- 
turity, and, with it, the parasitic habits of the adult, their eyes are 
developed to a fair size, thus reversing the general rule in the class. 
In most other fishes the eyes are developed to a full and even 
remarkable extent as to size and perfection of sight in water. 
In Anableps, or the so called four eyed fish of the fresh waters of 
Central and South America, which belongs to a closely allied fam- 
ily with our blind fish, the Cyprinodontidm, the eyes are not only 
fully developed, but are divided into an upper and lower portion 
in such a way, by an opaque horizontal line, as to give the effect 
of two pupils, by which the fish probably sees as well when follow- 
ing its prey on the surface with its eyes out of water, as when 
under water. But it is in the interesting family of cat fishes {/Silu- 
ridce) that we find the most singular arrangement of eyes in per- 
fect adaptation to the diversified modes of life of the numerous 
species. In this family the eyes assume nearly every possible mod- 
ification from partial and even total blindness to perfectly develop- 
ed eyes, and these organs are placed in almost every conceivable 
position in .a fish’s head ; from the ordinary large eyes on the side, 
to small ones on top of the head, enabling the fish to see only what 
is above ; to the oval eyes on the side, in some just back of the 
mouth, situated in such a way that the fish can only see what is in 
close proximity to its jaws or even below them. Many genera of 
this family found in South America,^ Africa f and Asia,| have the 
eyes so small and buried under the skin or protected by folds or 
cartilage, as evidently to be of no more use than simply to distin- 
guish light from darkness. 
Among the most interesting forms of this family, in this respect, 
is the genus described by Prof. Cope under the name of Gronias 
nigrilabris. This fish is very closely allied to our common bull 
thinp: akin to ‘‘alternate generation,” or of larvfe capable of reproduction? Without 
having any facts to support such an assumption, except that, on general principles, the 
young of Myxine would probably be very much like Branchiostoma, and that its young 
is not known, while Branchiostoma has only been found in waters where some species 
of Myxinoid exists, I think that before the position of the lancelet is firmly established 
Ave must know the embryology of the Myxinoids; for should the lancelet prove not to 
be the young of the Myxinoids, it must necessarily form a distinct class of animals, 
perhaps as near to the mollusks as to the vertebrates. 
* Pvmelodiis cyclopiiini of Humboldt, Helogenes, Agoniosus and other genera. 
■\ Eiitropius congensis. 
X Ailia, Shilbichthys, Bagroides and other genera. 
