CLUPEIN-E. HYODON. 291 



further innovations on the existing dispositions of these 

 groups. We have a strong conviction, however, that 

 Hyodon "will hereafter be brought in among the aberrant 

 types of the SalmonincB, with which, in every thing but 

 its single dorsal fin, it bears, both externally and inter- 

 nally, the strongest resemblance, — a resemblance which 

 extends to its habits, food, and the fresh waters to which 

 both are more especially appropriated. If we ventured 

 a conjecture upon its true station, we should place it 

 between Osteoglossum and Xiphostoma. The absence 

 or presence of an adipose fin is considered the sole dis- 

 tinction between the Cyprince and the SalmonincB, and 

 between the latter and the Clupeince — and it is, doubt- 

 less, one of the most typical characters that we yet know 

 of; but at the confines of each of these groups we must 

 look for considerable variation in this respect, because 

 every naturalist is fully aware that when nature is about 

 to quit one type of form, she modifies her structures in 

 such a way that many of the strongest characters she 

 had been employing are lost, and are exchanged for 

 others which only exist in their full perfection in the 

 next group to which she is advancing. Besides this, 

 we have several instances of two closely allied genera, in 

 which one has two dorsal fins, and the other only one. 

 The most striking of these that at present occurs to our 

 mind is in the case of Loricaria and Hypostoma among 

 the Siluridce, — two types which Cuvier places only as 

 sub-genera, although the latter has an adipose fin, 

 while the former has none. We have before expressed 

 our suspicion that some of the sub-genera of Brazilian 

 salmon, as Prochilodus Agass., and even Anodus, may 

 eventually be found to enter among the carps ; and for 

 the above reasons Hyodon would not be absolutely ex- 

 cluded from forming an aberrant group among the sal- 

 mons, merely because it has not an adipose fin. How- 

 ever this may be, we feel perfectly satisfied on the 

 situation we assign to Chirocentrus, whose whole struc- 

 ture is intermediate between the herrings and the pikes. 

 (246.) Having now closed our survey of the entire 

 u 2 



