118 DWIGHT 



It has been asserted that various species of caged birds have 

 been seen to change color without feather loss. Well, it is 

 perfectly true that some do change color, but in these birds, 

 as can be proved, no actual pigmentary change takes place. 

 Colors concealed by feather edgings or optical effects produced 

 by structure may be intensified by gradual loss of parts of the 

 feathers and as these minute parts will not be found in the cage, 

 those who are ignorant of simple principles of wear will see a 

 "brightening by influx of pigment." The adult Purple Finch 

 (yCarpodaciis piirpnreiis) and adult Indigo Bunting {Passerina 

 cyanca) both brighten very perceptibly by wear alone as I can 

 testify. The former has no prenuptial moult, the latter has one 

 confined to the body plumage. Theorists class both together 

 and lump with them a lot of other species, among which 

 similar changes are observed by them or by their friends and 

 the crudest observations and baldest statements are offered as 

 "proof" of new color. 



From the extensive literature of the subject we learn that ob- 

 servers of caged birds have failed to bar out the possibility of 

 moult, and in species that regularly only renew a small portion 

 of their plumage at the prenuptial moult (particularly the small 

 feathers about the head) I have no doubt that moult has oc- 

 curred, the tiny feathers being whirled out of the cage by a puff 

 of air or rolled up to almost nothing if sprinkled from the bath 

 or drinking cup. It is not often that many are cast off at one 

 time and they are so extremely small that the entire plumage of 

 the head of a bird the size of a Song Sparrow {Melospiza fasciatd) 

 may be held between the thumb and forefinger. Did any of the 

 theorists ever try blowing away such a pinch of feathers even in 

 a small room and see how many he will find ? Some idea of the 

 relative size of feathers may be gained from plate I, on which 

 those of a Bobolink {polichoriyx oryzivorus), a good sized Pas- 

 serine bird are figured. In order to give some idea of their 

 relative number, I have made actual count of all the contour 

 feathers on a spring male. The enumeration of the minute 

 down-feathers, semiplumes and filoplumes I leave to others. 

 The result of my count is as follows : 



