RELATIONS TO OTHER SCIENCES 553 



proportions and developed horns, the fauna might be described 

 without hesitation as Middle or Upper Eocene. Groups of mammals 

 progressively differing from this original race in (1) the larger size 

 of the brain, (2) special adaptation of the teeth to flesh-tearing or 

 vegetable-grinding, (3) greater mobility of the neck, and (4) adapta- 

 tion of the feet either to grasping prey or to running on hard ground , 

 mark successive geological periods. The general succession is always 

 the same whatever may be the local circumstances; and for this 

 reason it is impossible to accept the published conclusions of the 

 brothers Ameghino as to the age of the various mammal-bearing 

 Tertiary deposits of Patagonia. The mammals of South America are 

 certainly anomalous, but the marine fossils intercalated between 

 some of the deposits containing bones in Patagonia prove that the 

 rate of mammalian evolution was much the same there as in other 

 lands. Even Australia, which is in many respects a remnant of the 

 Mesozoic world, can be readily recognized by its mammals as modern 

 Tertiary. The monotremes are certainly a very ancient type, but 

 their large brains, peculiar skulls, and rudimentary or lost teeth 

 show that they belong to a far later period than that at which their 

 lowly tribe flourished. Similarly the kangaroos have highly special- 

 ized teeth and feet which cannot be misinterpreted. 



Relations to Cosmical Physics 



While fossils prove that the succession of life during geological time 

 has been essentially the same everywhere, it is still impossible to 

 determine exactly which faunas were contemporaneous in different 

 parts of the world. A deposit containing Carboniferous fossils, for 

 example, in Australia was not necessarily formed at precisely the 

 same time as a rock yielding similar fossils in the Arctic regions. 

 There may have been migration, and the Carboniferous animals and 

 plants may have quitted the Australian region long before they 

 reached the Arctic Circle, or vice versa. To obviate the use of the 

 word "contemporaneous" in referring to such a case, Huxley long 

 ago proposed the more indefinite adjective "homotaxial," which 

 postulates nothing more than the identity of two rocks in their fossil 

 contents; and there is at present no prospect of dispensing with this 

 provisional term. It is therefore unfortunate, but true, that paleon- 

 tology gives only very uncertain information about the distribution 

 of heat over the surface of the globe in past ages. It is perfectly clear 

 from fossils that climates have changed in nearly all if not all parts 

 of the world; it is not equally evident how these changes of climate 

 in different regions were related to each other. 



Fossils, however, can only be used as tests of climate with special 

 caution. When, by analogy with the existing world of life, a whole 



