ON ]S T OSOLOGY. 449 



no means vouch. I have many difficulties still with regard to 

 them, and I think I could still point out considerable improve- 

 ments ; but my labour is to study genera in the first place, and, 

 when I can arrive at them more certainly, species. If I can dis- 

 tinguish these with tolerable accuracy, I am not concerned with 

 my distribution into orders and classes being more or less exact. 

 It is from our accuracy with respect to the species that all the 

 rest of the divisions can be properly established ; and in no- 

 sology indeed we are very much relieved from a more particular 

 attention to classes and orders. 



Wherever the species of things are very numerous, as in 

 the case of plants and insects, it is absolutely necessary to 

 attempt classes and orders. But there are many depart- 

 ments, even of natural history, where this is not so neces- 

 sary, because the species are not so numerous ; or if the species 

 are, because the genera at least are not, so that we need not go 

 higher. The first is the case with respect to quadrupeds ; and 

 the second, perhaps the first, with respect to Nosology. Nobody 

 has inveighed more vehemently against the use of method than 

 M. de Buffon. He says the whole known species of quadrupeds 

 on the face of the earth are not above two hundred : now, we 

 can easily remember the separate characters of two hundred 

 things or species without uniting them into genera, far less 

 throwing them into orders and forming classes. I say the same 

 will apply in the case of Nosology, where the species are not very 

 numerous, as appears pretty clearly from the abridgment of 

 what have been called genera in my synopsis ; and without lay- 

 ing down characters of either classes or orders, it is easy to re- 

 member the separate characters of a hundred and fifty genera 

 which I have there marked ; (and this is what I meant by the 

 passage in my prolegomena, to which I have already referred.) 

 But I do not mean to push this too far ; for there are advantages 

 which, I say, may, and I hope will arise from the study of even 

 classes and orders ; and M. de Buffon is much in the wrong in 

 rejecting such an attempt with regard to his quadrupeds. The 

 attempt to establish genera and species is always the study of 

 the affinities between different things ; and though we distinguish 

 them by their differences only, we arrive at a great deal of know- 

 ledge also from their affinities ; and we would never have marked 



