Mr. 0. Salvin on the Genus Cinclus, lO'J 



VI. — On the Genus Cinclus. By Osbert Salvin, M.A., 

 F.L.S., F.Z.S., &c. 



(Plate II.) 



When proper attention is given to all the slightly varied forms 

 under which many species present themselves, the difficulty of 

 classifying such forms increases with the attention bestowed 

 upon them, — that is, if the relationship which each bears to 

 another be reduced to its proper value and yet brought into 

 systematic arrangement. The definition of species founded on 

 the presumed inability of hybrids to produce ofiFspring is one that 

 is not, and never has been, applied in practice : the fact is as- 

 sumed from differences of form and colour, and the supposed 

 absence of individuals intermediate in character between two 

 distinct forms. What we really have to consider is, not that two 

 allied species cannot interbreed, but that they do not — sometimes, 

 perhaps, because they are never brought in contact in nature, at 

 others because they will not, even when occupying the same 

 country. That closely allied species do not interbreed is inferred 

 from observation of facts of greater or less value, which, in the 

 aggregate, tend to show that they keep themselves distinct. 

 Differences of apparent character and their supposed constancy 

 form the actual basis upon which species are differentiated ; and 

 it seems unreasonable, as the fact of constancy becomes more 

 certain from the examination of a number of individuals, to 

 reject a difference, however slight, that may exist between two 

 supposed species and to refuse them specific rank. 



The Common, the Ring-necked, and Japanese Pheasants are 

 known to produce perfectly fertile crosses, and are good in- 

 stances of easily recognizable forms interbreeding when arti- 

 ficially brought into contact. A scheme of nomenclature must 

 assign such well-marked forms a place in any system of nature, 

 and that by giving each a name. Our method, then, if it is to 

 take cognizance of every constant form, must do so by naming 

 each such form ; and the only practical guide we have to deter- 

 mine our judgment in defining species is that of the stability of 

 recognizable characters ; and it is to the fact of the existence of 

 such characters, rather than to their amount, that we have to 



