I7O EVOLUTION OF TO-DAY. 



no surprise, therefore, if the different genera of the 

 same family should be found to be sometimes widely 

 separated from each other, particularly if fossil evi- 

 dence show that the intermediate localities were 

 formerly also occupied by the family. Larger groups 

 than families are so old that no geographical rela- 

 tions could be predicted. (2) It would necessarily 

 follow that barriers, which limit the migration of a 

 species, must be the limits of the extension of the 

 species. The same species ought not to be found on 

 both sides of such a barrier, unless it can be shown 

 that the barrier in question is of recent origin, or 

 that it really does not entirely prevent migration. 

 (3) The theory would require a close relationship 

 between the present and past inhabitants of any 

 country, since the latter have descended from the 

 former. This would, therefore, make it desirable 

 that a direct relation should be found between the 

 present inhabitants of any locality and its recent fos- 

 sils, after making deductions due to recent migra- 

 tions. Geographical distribution should thus be a 

 science by which the localities of species, genera, etc., 

 could be explained from the consideration of the 

 relationships and powers of migration of the organ- 

 isms, and the changes in the configuration and 

 climate of the land. Anatomical relationship should 

 be parallel with geographical relationship. Unless 

 these laws are found to be verified with more or less 

 exactness, it would be impossible to accept evolu- 

 tion ; if they are verified, evolution finds a certain 

 amount of support. 



