Plumages of the Male Crossbill. 667 



such birds, obtained in June and July, in whicli onl}^ a trace 

 of the juvenile feathers remained. Some of these yellow 

 males are approached in colour by some of the brightest 

 females, but the latter appear never to be so bright ou the 

 crown and mantle. 



The striped juvenile plumage is moulted in the first 

 autumn entirely except for the wings, tail, primary coverts, 

 and the outermost three or four greater coverts ; at each sub- 

 sequent autumn every feather is moulted ; I cannot find any 

 evidence of any spring moult. The changes which take piaffe 

 as the breeding season advances is variable : some red birds 

 become duller or coppery-red and some yellow birds become 

 bronzy yellow, otlier red birds become a rather brighter red, 

 others again seem to show little change. Some colour 

 change, due to abrasion, in parts or the plumage, especially 

 ou the crown, is I think possible, as in some the red of the 

 feathers is confined to the tips, and when these are worn off, 

 the bron25e-yellow of the bases of the featliers would become 

 visible. The brightest red males in first plumage are tinged 

 with red on the scapulars and wing-coverts and even on the 

 bend of the wing (the so-called ^^ rubrifasciata ''). 



The period of moult in the Crossbill has long been a puzzle 

 to me, as being such an early breeder one might suppose 

 that, like a Rook, it would moult earlier than do most passerine 

 birds, but this does not appear to be the case. Naturally, the 

 period of moult is a rather variable one according to the 

 earliness or lateness of the broods, but generally speaking 

 it is much the same as in the other finches. Of the juvenile 

 birds, I find the earliest to have finished its post-juvenile 

 moult is one on June 23, but this and another on the 30th 

 are the earliest which showed any moult at all. July and 

 August appear to be the months in which this plumage is 

 cast, but I find several in these months, and one even as late 

 as September 28, in which no moult had begun. The latest 

 bird in whicli moult was not complete was an odd one killed 

 on Nov. 30. Of birds one or more years old, the earliest 

 in moult was one on July 16, and the earliest completed 



