DISPERSAL AND MIGRATIONS 147 



in the equatorial region, whence they have 

 spread in two grand streams north and south 

 towards the Poles. This evidence is too com- 

 plicated to be discussed in a popular little 

 volume like the present, but the reader may 

 consult the work on the migration of birds 1 

 by the present writer, where his theories of 

 avine dispersal are fully elaborated. To our 

 mind innumerable facts of dispersal suggest 

 that species never emigrate or extend their 

 area in the Northern Hemisphere in a southerly 

 direction, nor do they so extend their range 

 in the Southern Hemisphere in a northerly direc- 

 tion. Farther, there is no evidence to suggest 

 that a species retires from adverse conditions 

 (as the retreat from a glacial epoch), or that 

 such have any influence whatever in expanding 

 its geographical area of occupation. We believe 

 that breeding conditions are the sole determin- 

 ing factor in such range expansion, and that a 

 species extends its area as a natural result of 

 increase. The soundness of these views is 

 confirmed by many facts of geographical dis- 

 tribution and migration. Not only do we find 

 many species absent from certain areas pre- 

 senting few or no barriers to emigration save 



1 The Migration of Birds. Amended Edition, 1897. 



