ic8 DwiGHT, Variations of the American Goldfinch. lApr 



character by which adult males may be distinguished from young 

 males is found in the lesser coverts or ' shoulders ' which with the 

 median coverts are bright canary-yellow. In some specimens the 

 lesser coverts are greenish tinged, being dusky basally, and there 

 is much white in the median coverts. Such birds usually show a 

 white spot on the primaries, while the yellower-shouldered birds do 

 not, but whether they represent the second winter and the others 

 the third or later winter plumages, I cannot say. Osteological 

 characters show that none of these are of the first winter, although 

 some resemble very closely bright colored young birds. The same 

 influences of wear produce the same effects in adults as in young 

 birds, bleaching them rapidly during the winter months. There 

 is great variation in winter specimens, whether of the first or later 

 winters, also there are age variations from winter to winter, and 

 the individual resistance to fading varies, but all these variations 

 so overlap that it is hardly possible even with large series to estab- 

 lish which of them has been most potent in any given case. At 

 about the same time as in young birds, or usually a little earlier, 

 the second winter dress begins to be replaced by the second 

 nuptial. 



The postnuptial moult in the female occurs a little later than in 

 the male who is not occupied so long in caring for the brood. 

 The differences between first and second or later winter plumages 

 are relative, and, although the colors appear to average deeper 

 with age, the age of a specimen may only be told with certainty 

 by osteological characters. The lesser wing-coverts are more fre- 

 quently greenish in adult females. 



The postnuptial moult of salicamans begins fully a month or two 

 earlier than that of tristis. Two ragged males of August 12, 

 evidently passing from first nuptial plumage, have acquired three 

 new primaries, a few tertiaries and wing-coverts and some of the 

 body plumage, while a specimen of August 27 (probably a year or 

 more older than the much worn birds just mentioned) is further 

 advanced, having six new primaries, and four pairs of rectrices 

 and much of the new body plumage and coverts. Comparing these 

 birds with September or October specimens of tristis just complet- 

 ing the moult, I find that the browns are equally dark and appar- 

 ently the two forms indistinguishable at this stage. As in tristis. 



