DISPERSAL AND MIGRATIONS 161 



Another, if less generally attributed cause, is the 

 rise or fall of temperature. Other ornithologists 

 believe that it had its origin in the accidental 

 wanderings of birds, or even the aimless roam- 

 ing of young birds in quest of food ; whilst a 

 very ingenious hypothesis attributes it not to 

 want of food, but to want of light, during the 

 several months of polar night in the Arctic 

 Regions, ages ago, when those latitudes enjoyed 

 a much warmer climate in winter than they do 

 now. It will be noted that all these hypotheses 

 on the cause of migration really mean a retreat 

 from adverse conditions of life. Failure of food 

 supply can scarcely be a tenable hypothesis, be- 

 cause, unfortunately, birds begin their migrations 

 in numberless cases before food shows any sign 

 of becoming scarce ; neither can migration arise 

 from accidental roaming, because birds, of all 

 creatures, are closely attached to certain spots, 

 and exhibit no propensity for straying beyond 

 their normal limits, which are determined by 

 other and perfectly well-known causes. After a 

 long and close study of the facts, I am disposed 

 to attribute the spring migration of birds to the 

 range expansion or colonisation of a species 

 gradually increasing the distance from its winter 

 base. As we have already seen, the general trend 



