452 Mr. J. G. Millais on Change to 



the manner in which these are eflfected, form a subject which 

 few naturalists have studied. The reason of this is probably 

 that none except the field-naturalist, who obtains fresh speci- 

 mens in the flesh and examines them at all seasons, can 

 correctly grasp the case, for, with some exceptions (as in the 

 LimicolcB and Gavi(E),i^e feather-casting of birds in confine- 

 ment is very irregular, owing to improper feeding and want 

 of exercise. 



For over twenty years, during which I have collected in 

 all parts of the British Islands, I have made " the change-of- 

 plumage question '' a special study. In my opinion there is 

 nothing more interesting than the discovery of links in the 

 chain of evidence which goes to show the life-history of any 

 one species whose change of dress is at all elaborate. In 

 some birds, like the Goldeneye, Sheld-duck, Eider, Long- 

 tailed Duck, Gannet, Great Black-backed and Herring Gulls, 

 &c., we see a great number of gradual changes taking place 

 before the adult plumage is reached, lasting, in some cases, 

 five years ; whilst the reasons which govern the seasonal 

 changes of adult Grouse, Waders, and others are by no 

 means easy to determine without a very great deal of close 

 observation and study. 



Putting aside the question of immature plumages, let us 

 now look at the laws that govern those birds which make an 

 annual change from the winter garb to the summer breeding- 

 dress. We find that this great change, with the exception 

 of perching-birds, takes place in one of the three following 

 methods : — 



(I.) By a nearly complete moult. 



(II.) By a partial moult, in which the old winter feathers 

 are replaced by new summer ones, and by many of the old 

 winter feathers themselves changing colour. 



(III.) By a complete recoloration of the feathers in new 

 form throughout the whole plumage, only a few being moulted 

 and replaced by new summer ones. 



Taking these three methods in the order in which they 

 come, the species which will be found to follow method No. I. 

 are practically but few, of which the Long-tailed Duck 



