150 



He commenced by remarking that among the genera referred by 

 Cuvier to this family there were several, such as Pacilia, Lebias, &c., 

 vhich possessed maxillary teeth and a large number of branchio- 

 stegous rays. These genera, he conceived, ought to be excluded 

 from the Cyprinida ; and the family be considered as limited to fishes 

 ■with mouths destitute of teeth, and having few branchiostegous 

 rays. 



To the family thus reduced the nearest affinities appeared to him 

 to be the genera Atherina and Mugil. In internal organization 

 the Cijprinida agree nearly Avith those genera ; and this considera- 

 tion, M. Agassiz conceives, is of much higher importance in the natū- 

 rai arrangement than the extemal character founded on thepresence 

 or absence of spinous rays in the dorsal and other fins. The affinity 

 of the CyprinidcB to the Siluridm he regards as extremely doubtful : 

 and although from the bearded Carps to the bearded Siluri there 

 appears to be a natūrai transition by means of the bearded Loaches, 

 it is important to distinguish that in these latter, as well as in the 

 Carps and other Cyprinidce, the beards, as they are called, are merely 

 processes of the skin; 'vvhile in the Siluri, the cirri of the angles 

 of the mouth are actually prolongations of the maxillary bones, 

 becoming gradually cartilaginous and tapering into thread-like ex- 

 tremities. 



In the subdivision of the Cyprinidce, M. Agassiz regards the form 

 of the fins, and especiaUy of the dorsal and anai, as furnishing indi- 

 cations of the highest value ; and the form of the pharj'ngeal teeth 

 as aflFording the characters next in importance. He first distinguishes 

 the groiip coraprising the genera Anahleps, Cubitis and Botia, the 

 latter established by Mr. Gray for the reception oi those Loaches in 

 Avhich the suborbital bone is armed ■with a moveable spine. He then 

 distinguishes another group comprising four genera : 1 . Cyprinus, 

 in -vyliich the pharyngeal teeth are large, and, \vhen worn, resemble 

 the molars of some Rodent Mammalia, such as the Hare ; 2. Barbus, 

 in which there are three rows of lengthened conical hooked teeth on 

 each side of the j9Aaryw,r ; 3. G'oizo, in which the pharyngeal teeth 

 have the šame form as those of the Barhels, but are more slender, 

 and constitute only two rows ; and 4. Tinca, the pharyngeal teeth of 

 which are club-shaped, rounded at the end, and placed in a single 

 row. In the genus Leuciscus, which M. Agassiz limits to Leve. Al- 

 bumus and three allied species, the mouth is cleft obliquely, and the 

 teeth, consisting of elongated cones, are disposed in four rows. 

 From these the Cyprinus Našus is to be generically distinguished as 

 possessing six rows of pharyngeal teeth : its mouth is transverse and 

 inferior, with the edges cutting. A third genus, containing many 

 species, also reąuires to be distinguished, as having only two rows 

 of teeth, one of which is hooked : in these the opening of the 

 mouth is rounded. There remains the genus Abramis, distinguished 

 by its long anai fin, in which the teeth are hevilled ofF and have 



