60 MOSTLY MAMMALS 
It was doubtless assumed that, had the experiment been 
extended over a longer period, the white would have 
gradually extended downwards till the whole hair became 
blanched. But had this been the normal way in which 
the change from a black to a white coat is brought about, 
it is obvious that animals ought frequently to be captured 
in which the coat is in the same condition as that of the 
lemming. So far, however, as I am aware, no such con- 
dition has ever been described. 
Moreover, it is perfectly well known that, apart from 
those which turn white in winter, a large number of 
animals have a winter coat differing markedly in colour, 
as well as in length, from the summer dress. The roebuck, 
for instance, is of a brilliant foxy red in summer, while in 
winter it is grey-fawn with a large patch of pure white on 
the buttocks, And it is quite clear that the change from 
red to grey, and the development of the white rump-patch, 
is due to the shedding of the short summer coat and its 
replacement by the longer winter dress. Obviously, there- 
fore, it is natural to expect that a similar change of coat 
takes place in the case of mammals which turn white in 
winter. 
That the change in spring from a white to a dark dress 
is due to a shedding of the fur seems to be admitted on 
all hands, for it would obviously be quite impossible for 
long hairs to become short, or for white ones to turn 
brown. And even in animals which do not alter their 
colour in any very marked degree according to season, the 
spring change of coat is sufficiently obvious. For the 
winter coat, owing to the long time it is carried and the 
inclemency of the season when it is in use, becomes much 
faded and worn by the time spring comes, and the con- 
trast between it and the fresh and brilliant summer coat 
