93 UPLAND AND MEADOW. 



It is necessary to presume, of course, that Wilson was 

 correct in his statements to make my own conclusions 

 wortiiy of consideration ; but the later authors, Audu- 

 bon and Nuttall, follow so closely in their remarks upon 

 the same birds, it is not probable that any serious blun- 

 ders were committed. At present, the irregularities in 

 migration are most noticeable in autumn ; and one 

 potent fact that the return southward is then under 

 protest, is that the action of frost is very variable, and 

 sheltered nooks, of perhaps an acre or two, may escape 

 for several weeks, while the marshes elsewhere have 

 only wilted weeds to cover them ; and these sheltered 

 nooks are often so densely overcrowded with the birds, 

 and birds, too, of other species, that it is difficult to con- 

 ceive how they keep alive. This occurs, at times, among 

 the strictly upland birds, but much less frequently. 

 They, on the other hand, I suppose, often find food be- 

 coming less and less abundant, and leave before tiiere is 

 any apparent reason. I have known some of our birds, 

 notably thrushes, to quit a neighborhood directly after 

 nesting, or early in July, and not one again to be seen 

 until the following spring, when they returned, as usual, 

 and remained in abundance until there was not only 

 hoar-frost, but thin ice. Temperature and migration 

 are largely coincident, but cannot be considered as cause 

 and effect. Whether the bird is an insect eater or 

 granivorous, it matters not. The food, in either case, 

 has its allotted time of growth, and while this continues 

 the birds are sure of a supply, for premature frosts do 

 not affect either animal or vegetable life so far as to 

 cut off the food supply suddenly, and thus the bird has 



