32 Allen's naturalist's library. 



Grouse is fairly complete, and many of the finely mottled 

 rufous and black autumn flank-feathers are replaced by widely, 

 and often irregularly, barred buff and black feathers, similar to 

 those of the chest. It must be particularly noted that in 7i07ie 

 of the many females examined, in breeding plumage, were 

 the whole of the autumn flank-feathers cast or changed in the 

 summer moult, a large proportion being retained, unchanged 

 in colour, till the next (autumn) moult. The summer flank- 

 feathers are produced in two ways, either by a gradual re- 

 arrangement and change in the pigment of the autumn feathers 

 (PI. III., Figs. 6-8) or by moult (PI. III., Fig. 9). In some birds 

 the whole of the alteration in the plumage of the flanks is pro- 

 duced by change of pattern in the old autumn feathers, in 

 others the change is entirely produced by moult, while some- 

 times both methods are employed by the same individual. 

 In the former case, the first indication of the coming change 

 may be observed in the beginning of November, or even 

 earlier, when many of the flank-feathers show traces of an 

 irregular buff stripe or spot near the terminal half of the 

 shaft (Fig. 7). As the bird only changes about half its flank- 

 feathers, these buff marks are only to be observed on such as 

 are destined to undergo alteration of pattern, which, roughly 

 speaking, means every second or third feather. The buff spot 

 gradually enlarges and spreads along the shaft, then becomes 

 constricted at intervals and broken up into patches which 

 gradually extend laterally towards the margins of the webs, 

 forming wide irregular buff bands (Fig. 8). Meanwhile the 

 interspaces become black, and the rufous of autumn dies out. 



When the summer feathers are supplied by moult, they 

 usually begin to make their appearance about the beginning of 

 March, and even when fully grown, may generally be recog- 

 nised from those produced by change of pattern, by their more 

 regular black and buff barring (PI. III., Fig. 9) The change of 

 pattern without a moult appears to take a long time to become 

 complete, for we find, as already shown, that though autumn 



