FISHES. 83 



of the upper jaw, whose place is supplied by those of the palate : their 

 whole structure also exhibits evident analogies which we will describe : it 

 is divided into three orders. 



The Cyclostomi, whose jaws are soldered in an immoveable ring, with 

 branchiaa containing numerous openings. 



The Selachii, which have the branchiee of the Cyclostomi, but not 

 tlieir jaws. 



The Sturiones, whose branchiffi are opened by the usual fissure fur- 

 nished with an operculum. 



The other series, or that of the Ordinary Fishes, presents a primary 

 division in those where the maxillary bone and the palatine arch are fixed 

 to the cranium: they constitute an order which 1 call that of the Plec- 

 TOGNATHi, and are divided into two families: the Gymnodontes and the 



SCLERODERMI. 



I next find fishes with perfect jaws, but whose branchiee, instead of 

 being pectiniforra, as is the case in all the others, resemble a series of 

 small tufts; they also constitute an order which I call Lophobranciiii, 

 that comprises but a single family. 



There then remains an immense number of fishes, to which no other 

 characters can be applied than those of the external organs of motion. 

 After an extensive research, I have found that the least objectionable of 

 these characters is the one employed by Ray and Artedi, drawn from the 

 nature of the first rays of the dorsal and anal fins. Thus the ordinary 

 fishes are divided into Malacopterygii, in which all the rays are soft, 

 with the occasional exception of the first of the dorsal or of the pectorals ; 

 and into Acanthopterygii, in which the first portion of the dorsal, or of 

 the first dorsal where there are two, is always supported by spinous rays, 

 and in which some of the same are always found in the anal fin, and at 

 least one in each of the ventral ones. 



The first may be conveniently divided by a reference to the position of 

 their ventral fins, which are sometimes situated behind the abdomen, 

 sometimes suspended to the apparatus of the shoulder, or are totally 

 wanting. 



We thus arrive at the three orders of the Malacopterygii abdomi- 

 nales, the Subbrachiati and the Apodes, each of which comprises 

 certain natural families to be described. The first is particularly nu- 

 merous. 



It is impossible, however, to apply this mode of division to the Acan- 

 thopterygii ; and their subdivision in any other way than by that of 

 natural families is a problem that I have hitherto vainly endeavoured to 

 solve. Fortunately, many of these families are possessed of characters 

 nearly as exact as those which could be given to true orders. 



