382 THE SONORAN REGION. [CHAP. X. 



species, there is surely no reason why they should not occasionally 

 occur in the case of genera. It would, therefore, seem by no 

 means improbable that the species of the genus Equus which 

 inhabited the eastern and western halves of the northern hemi- 

 sphere during the close of the Tertiary period may have been 

 evolved from the closely-allied but separate ancestral equine 

 stocks. 



The matter does not, however, by any means end here. In 

 an earlier chapter 1 it has been shown that the same species of a 

 genus of fish (Galaxias) occurs in countries so remote from one 

 another as New Zealand and Australia on the one hand, and 

 Patagonia on the other. With the evidence of the cave-animals 

 before us, is it absolutely impossible that these apparently iden- 

 tical fishes can have been evolved independently of each other ? 

 Should this be so, it will engender increased caution in drawing 

 any inferences as to former land-connection from the evidence 

 of single animals. But such instances of independent evolution, 

 if they do occur, must be of extreme rarity, and will in no case 

 interfere with deductions drawn from the presence of a number 

 of allied species or genera of animals in distant countries. 



1 Suprh, p. 125. Recently another species has been described from South 

 Africa. See Steindachner, SB. Ac. Wien, vol. ciii. p. 460 (1894). 



