Xo.40;{] 



CIIAXGES OF COLOR IN BIRDS 



37 



the light for a week or two and meal-worms were added to 

 its diet. This invariably resulted in a full resumption 

 of song. Even in the middle of winter a tanager or a 

 bobolink would make the room ring with its spring notes 

 and with this phenomenon was correlated a slight de- 

 crease in weight. This phase of the experiment could 

 not be repeated indefinitely, however, for the song period 

 seemed limited, just as it is under normal conditions, 

 although the nuptial plumage remained unchanged 

 throughout the winter. As one of my keepers pithily put 

 it, "We have their calendar twisted backward." 



I found that a sudden alteration in temperature— 

 either lower or higher — wrought a radical change in the 

 physical metabolism of the birds. They would stop feed- 

 ing almost altogether, and one tanager lost weight rapidly. 

 A few feathers on the neck fell out, and in the course 

 of some two weeks this bird moulted almost every feather 

 and came strongly into his normal winter plumage of 

 olive green. The metabolism set up by the change in 

 temperature, in its extent and rapidity, seems compar- 

 able only to the growth of a deer's antlers. 



Karly in the following spring individual tanagers and 

 bobolinks were gradually brought under normal condi- 

 tions and activities, with quick result: just as the wild 

 birds in their winter haunts in South America were ar 

 that time shedding their winter garb and assuming the 

 more brilliant hues of summer, so the birds under my 

 observation also moulted into the colors appropriate 

 to the season. The old scarlet and black feathers fell 

 from the tanagers and were replaced by others of the 

 same color; from buff, cream and black, the bobolinks 

 moulted into buff, cream and black! There was no ex- 

 ception; the moult was from nuptial to nuptial; not from 

 nuptial to winter plumage. The dull colors of the winter 

 season had been skipped. 



I think we thus have proof that the sequence of plu- 

 mage in these birds is not in any way predestined 



