100 THE RIDDLE OF MIGRATION 



longspurs. Yet this may be an extreme attitude 

 to adopt, for the return to the north with the re- 

 estabHshment of spring conditions and sex fever may 

 conceivably have been achieved, in part at least, 

 with the aid of memory and a rudimentary intelli- 

 gence. Similarly the presence of the sun, always to 

 the south during the major portion of the day, may 

 have been a sort of general drawing card to that 

 point of the compass during the shortening days of 

 autumn. Such factors might materially have has- 

 tened the process, but the assumption is superfluous. 

 But the present distribution of the Lapland long- 

 spur includes only northern breeders wintering in 

 the southern and central States. The explanation 

 of such a distribution can, of course, never be 

 satisfactorily demonstrated; but with a migratory 

 swing once established, each spring would see a 

 spreading further and further north as long as sum- 

 mer conditions remained favorable in all respects. 

 Natural selection however is constantly and every- 

 where in force and selects as certainly in the sum- 

 mer as in the winter. The selective forces, even 

 today, are in many cases extremely difficult or even 

 impossible to detect. The crow (Corvus brachyhryn- 

 chos) is just now getting consistently more abundant 

 in western Canada in spite of the heaviest persecu- 

 tion that man can devise. Some altered factor in 



