652 LAND MAMMALS IN THE WESTERN HEMISPHERE 



fagriochoerids from the foreodonts and the fEntelonychia 

 from the ftoxodonts. From time to time attempts have been 

 made to miite two or more of these groups, but in each case 

 better material and fuller knowledge have demonstrated the 

 unnatural character of such association and the separate 

 origin of the peculiar structure. 



Admitting the reality and frequency of these modes of de- 

 velopment, a far more difficult problem is to determine the ex- 

 tent to which such independent acquisition of similar structures 

 has actually been carried, and it is at this point that the widest 

 divergences of opinion are to be found. As yet, our knowledge 

 is far too imperfect to permit the making of positive statements, 

 but there is no evidence which would justify the conclusion 

 that the same genus, family or order of mammals ever arose 

 independently from radically different ancestors. We have 

 no reason to believe that identical groups of mammals were 

 ever separately developed in land areas which through long 

 periods of time had no means of intercommunication. If such 

 a thing ever happened, it must have been the rarest of excep- 

 tions. On the other hand, parallelism, by which related forms 

 pass through similar stages of development, would seem to 

 have been so exceedingly common, as fairly to deserve being 

 called a normal method of evolution. As more and better 

 material has been gathered, it has grown increasingly clear 

 that almost every large group of generic, family or higher 

 rank, whose history is known in any adequate measure, con- 

 sists of several distinct, though related phyla, which pursued 

 more or less closely parallel courses of modification, though 

 diverging from one another sufficiently to make the distinction 

 of them comparatively easy. The parallelism was thus not 

 exact, however perfect it may have been in particular structures, 

 and the longer the phyla persisted, the more distinctly did 

 they diverge. 



A typical problem, which involves these principles, is 

 afforded by the very curious and interesting group of South 



