51 



CHAPTER VI. 



MAMMALS CLASSIFICATION AND MUTUAL AFFINITIES. 



Starting -with the principles which I have laid down in the previous chapters, and by which I 

 mean to be guided throughout the remainder of this work, the affinity of species becomes of vital 

 importance to our whole inquiry ; and not merely the simple question, whether this or that species 

 be allied or not, but the degree of affinity, becomes almost as necessary to be known. While it 

 is indisputable that two different species starting on the career of change at the same time, and 

 from the same terminus, can never bo expected to make equal progress ia their journey, as one 

 will certainly outstrip the other, stiU it is equally clear that when they do not start simultaneously, 

 those which have started first ought to have the advantage on the whole. Thus we can hardly 

 escape from the conclusion that when we find two animals both apparently derived from the same 

 stock, but one more removed from its typical character than the other, that one dates its connexion 

 with it from the more distant period of time. 



The reader will see how important such indications may be when questions arise as to the 

 relative antiquity of the separation of diffi?rent lands, or their alternate separation and reunion. 

 No work can deal satisfactorily with geographical distribution which does not take a large account 

 of questions of affinity. 



A few words upon the chief difficulties which meet us in our attempts to classify the animal 

 kingdom, and more particularly, mammals, will therefore be a fitting preamble to th-s details 

 on which we are about to enter. 



Assuming that species are derived, the one from the other, the most perfect system of 

 classification would of course be simply a genealogical tree showing the descent of each. 



The materials for making such a tree are, however, beyond our reach. The records that 

 have been kept in the pages of geological strata, are imperfect and Luterrupted, and we do 

 not even know that we can always read the language ; and of by far the greater portion no 

 record has been preserved at all. 



All that we can do, therefore, is by the study of the anatomy and physiology of those living 

 species to which we have access, to endeavour to ascertain their affinities, and to make up a fictitious 

 tree, in the best way we can from the materials we possess. 



. The genealogical form of classification has, however, this disadvantage, that wo can at no 

 rate, and by no possible contrivance, squeeze it iuto a linear arrangement. Each species requires 

 a separate tree for itself. Scarcely one of all the thousands of species inhabiting the globe can 

 come into the same arrangement with another, for if two appeared in the same genealogy, one 

 of them must either be the parent or descendant of the other, and the cases in which there is the least 

 reason for supposing that this has been the case with living species, are few in the extreme. But 

 although this bo the case with individual species, it is not necessarily so with larger groups ; 



