94 PISCES. 



of the branchiae, and even on bones behind these arches, at- 

 tached like them to the hyoides, called pharyngeal bones. 



The varieties of these combinations, as well as those of the 

 form of the teeth placed at each point, are innumerable. 



Besides the apparatus of the branchial arches, the hyoid 

 bone is furnished on each side with rays which support the 

 branchial membrane. A sort of lid composed of three bony 

 pieces, the operculum, the suboperculum, and the inter- 

 operculum, unites with this membrane in closing the great 

 opening of the gills ; it is articulated with the tympanal bone, 

 and plays on one called the preoperculum. In many of the 

 Chondropterygii this apparatus is wanting, 



The stomach and intestines differ in size, figure, thickness 

 and circumvolutions, as greatly as in the other classes. The 

 pancreas, except in the Chondropterygii, is replaced either 

 by c£ecums of a peculiar tissue situated round the pylorus, or 

 by the tissue itself applied to the beginning of the intestine. 



The kidneys are situated along the sides of the spine, but 

 the bladder is above the rectum, and opens behind the anus 

 and behind the orifice of generation; exactly the inverse of 

 what we find in the Mammalia. 



The testes are two enormous glands commonly termed 

 milts; and the ovaries, two sacs about the same form and size, 

 in whose internal folds are deposited the eggs. Some fishes 

 copulate and are viviparous ; the young fry are hatched in the 

 ovary and issue through a very short canal. The Selachians 

 alone, besides the ovary, have long oviducts which frequently 

 open into a true matrix, and they produce either living ones 

 or eggs enveloped with a horny substance. In most Fishes, 

 however, copulation does not take place, the female depositing 

 her ova, and the male impregnating them after extrusion. 



Of all the classes of animals, that of fishes is the most diffi- 

 cult to sub-divide into orders from fixed and sensible charac- 

 ters. After many attempts, I have decided upon adopting the 

 following arrangement, which, though it militates in some in- 

 stances against precision, does not separate natural families. 



Fishes form two distinct series, that of Fishes, properly 



