CHARACTERS OF THE RASORIAL TYPE. Q63 



facts upon which these results are founded^ be pointed 

 out, it will then be time enough to talk of " want of 

 demonstration." We would even wish, that every thing 

 that has been hitherto said upon the natural system were 

 considered as an hypothesis, which remained to be 

 proved in those volumes where the details of each class 

 of animals are to be entered upon. 



(322.) There is one question which the reader may 

 possibly wish answered, arising out of the assertion 

 just made, that the dog is a rasorial type ; and it would 

 be this : — The dog belongs to the carnivorous order of 

 FercB ; how, then, can it be of a different type ? does 

 not this imply a contradiction ? By no means. The 

 FercE, as a whole, is a sub-typical group (305.) ; but 

 as it is a natural group, it contains, within itself, a ra- 

 sorial division. Hence, although this animal belongs to 

 the FercB, it is one of the rasorial examples of that 

 order ; and, accordingly, we find it the most intelligent 

 and docile of all its congeners. If the order Fera had 

 no rasorial types, then it would be an imperfect group; 

 and so, from not being circular, there might be reason 

 to suspect it was not a natural one. But, as aU its 

 divisions assimilate to the other types here defined, it is, 

 therefore, presumed to be both natural and perfect. So 

 in like manner, among birds, Melanerpes Sw. is the 

 aquatic (or fissirostral) type of the sub-family of wood- 

 peckers {PicianeB Sw.) ; although this group, of which 

 it forms a part, is, as a whole, a scansorial type. 



(323.) In substantiating the proposition with which 

 we commenced this chapter, we have endeavoured to 

 consult, in our definitions and examples, brevity and 

 perspicuity. It is obvious, that where the modifications 

 of a particular form pervade the whole animal creation, 

 the difficulty of definition becomes vastly increased. We 

 have, therefore, drawn our explanations and illustrations 

 chiefly from the best known vertebrated groups, and 

 from such only as have been submitted to analysis. A 

 far greater number might have been adduced from the 

 annidose kingdom, reposing on the same grounds ; but 



