102 THE RIDDLE OF MIGRATION 



at present a heavily forested area was, only a few 

 thousand years ago, a continuous sheet of ice. 

 With each local shift of environment the factors of 

 selection must have changed for that district. On 

 such changes probably hinge the details of the mi- 

 gratory movements of the Lapland longspur and the 

 same may be said of many another northern 

 migration. 



In this hypothetical reconstruction of the origin 

 of the migratory habits of the longspur, a number 

 of assumptions have been made. They are not of 

 equal value and we may briefly review them. 



The operation of the laws of natural selection 

 have been taken for granted. They represent a 

 biological principle the workings of which can be 

 seen on all sides and at all times. In this assump- 

 tion we are assuredly justified. 



We have assumed that the rate of reproduction 

 of the longspur must have been such as to force it 

 to spread from an original center. This assumption 

 is subject to no serious objection for we know that 

 exactly this happens in the case of many other birds 

 today. 



We have assumed that longspurs have wandered 

 in every direction at the beginning and close of the 

 breeding season. We know that such wandering 

 is characteristic of most — possibly all — species 

 today. 



