292 CLASSIFICATION OF FISHES. 



family or sub-family of Clupeince, we shall just advert 

 to the analogies which seem to result from the disposition 

 we have made of the various groups,, by instituting a 

 comparison between the 



Analogies of the Clupeince and the Salmonid^e. 



Primary Types of the Analogies. Primary Types of the 



Herrings. to Salmons. 



rBody much compressed; mouth) -. or c. r . 

 Clupea. ) small, obliquely vertical ; body i ^Sf' 



C sharp and serrated. ) lasalmo. 



C Mouth larger, more horizontal ; ) 

 Elops. \ belly (typically) smooth; hody > Salmo. 



C not much compressed. j 



C Mouth excessively large; teeth) 

 Osteoglossum. s strong, numerous; head com- V Xiphostoma. . 



C pressed. 3 



„. . .v ra \ C Mouth completely vertical; no? CJ „ ± 

 Odontognatkus (?). J anal fins. ^Stemoptyx. 



Chirocentrus. Dorsal fin close to the caudal. Sudis (?). 



The reader will remember that we have considered 

 Lacepede's genus Serrasalmo as one of the typical re- 

 presentatives of the American salmons ; and it is interest- 

 ing to see how much they accord with that of Clupea : 

 both have the body greatly compressed, and much 

 broader than any of their congeners ; both have the mouth 

 small and obliquely vertical — at least such is the direction 

 of the lower portion of the jaws in Serrasalmo; and 

 both have the very unusual character of the belly being 

 sharp and serrated. In Elops and Salmo, on the other 

 hand_, the form of the body is more lengthened; the 

 belly is neither sharp nor serrated ; the mouth is cleft 

 almost horizontally ; and the very aspect of the two 

 groups show an intimate resemblance. This, indeed, 

 cannot be said of the external form of Osteoglossum and 

 Xiphostoma; and yet there are points which intimate a 

 similarity of structure, and, consequently, of habits : each 

 has the largest mouth in its own circle ; the teeth in both 

 are slender, unequal, and numerous : but the truth is, that 

 each contains very few species ; so that, as there are no 

 very aberrant examples, there are no intermediate links of 

 connection. We before remarked, that Cuvier intimates 

 a resemblance between Osteoglossum and Sudis; and we 



