SUMMARY 521 



1. Proportions of extinct to living species or genera afford no satis- 

 factory comparison of age in two extinct faunas. The proportion of liv- 

 ing genera of mammals recorded in any Tertiary formation depends a 

 great deal on the families and orders that happen to he best represented, 

 some being more conservative than others ; on the region where it is found 

 and the facies that it represents, most of the modern types appearing 

 earlier in the Holarctic region and the open plains and uplands rather 

 than forest and swamp environment and the more peripheral regions of 

 the terrestrial world, where the older and more primitive stages are apt 

 to persist ; on the completeness of the specimens found and the conserva- 

 tism of the describer, since fragmentary specimens of an extinct genus or 

 species will often be indistinguishable from an existing one nearly re- 

 lated, and the standards of specific and generic distinction are far from 

 uniform. 



The Pleistocene fauna of South America, for instance, contains an 

 extraordinary proportion of extinct species, genera, and even families, in 

 comparison with that of North America and Europe. On this account it 

 was very generally regarded as older, referred to the Pliocene. The Ter- 

 tiary faunas of western Europe described by older writers from frag- 

 mentary material have been more largely referred to existing genera and 

 species than a modern revision of the fauna with more complete material 

 would admit. In consequence these faunas appear to be of more modern 

 type than they really are. 



2. The general equivalence in evolutionary stage of two faunas is like- 

 wise an unsafe guide. It may be fairly dependable, provided the faunas 

 are equidistant from their principal center of dispersal. Otherwise it is 

 very likely to be misleading. The modern fauna of central Africa has a 

 great deal of resemblance to late Tertiary fauna of the Mediterranean 

 basin. The Pleistocene fauna of Mexico has much in common with the 

 Pliocene fauna of the United States. On the other hand, the successive 

 Tertiary faunas of the United States and Europe, on opposite sides of the 

 great Asiatic land-mass, correspond on the whole pretty closely.' 



3. Satisfactory conclusions as to the correlation of vertebrate faunas 

 must depend on an analysis of each fauna and study of the origin and 

 dispersal of the different groups of which it is composed. But if the more 

 exact correlations are thus to rest on theories of dispersal, it is very neces- 

 sary that they should be checked and tested at every available point, not 

 merely by the dispersal of other groups of mammals, but by. the marine 

 faunas, stratigraphic and physiographic geology, etcetera. Otherwise we 

 may come to such absurd conclusions as the recent solemn pronounce- 

 ment of a great German authority on Proboscidea, that the American 



XXXVIII— Bull. Geol. Soc. Am., Vol. 27, 1915 



