112 THE RIDDLE OF MIGRATION 



back by the young. Thus crows, for instance, 

 destroying eggs of ducks, grebes, and other birds, 

 force repeated laying and incubation on many indi- 

 viduals of these species and so delay their normal 

 southward passage. The whole migration thus 

 becomes lengthened. Wetmore points out (p. 178) 

 that on September 6, 1920, when he noted golden 

 plovers in Paraguay, the same species was on the 

 same date recorded from Lake Athabasca (Alberta) , 

 6,000 miles farther north. The latter were un- 

 doubtedly young, the former probably adults. 



Breeding birds of Alberta, whose departure can be 

 accurately noted, are on the whole astonishingly 

 precise in their departure dates. Crows are a good 

 example. As a general thing one may be quite 

 certain that the main exodus will commence during 

 the last half of August and will be virtually over 

 by the 10th of September. The birds begin to con- 

 gregate in common roosts in July, the numbers get- 

 ting greater as the later families are added. About 

 the 20th of August the first bands pull out, the roost 

 continuing to be occupied by ever-decreasing num- 

 bers till the last contingent leaves. As the birds 

 forgather at sundown there may be, on any given 

 night, hundreds less than there were the previous 

 night. This happens simultaneously throughout 

 central Alberta. In the last week of March of the 



