WILLOW PTARMIGAN 181 
very numerous again and total disappearance in 1876; 
1882, a few observed; 1883, 1884 and 1885, great 
abundance; 1887, disappeared entirely; 1891, a few 
seen, gradually increasing each year till 1895, when 
there were considerable quantities; 1897, none; 1903 
and 1904, abundant, and a few seen every winter since 
to date, 1909. 
“At one time it was supposed that these years of 
abundance on the coast were due to heavy sleet in the 
interior, covering up all the buds and preventing the 
birds from feeding, and thus forcing them to seek food 
elsewhere. I have noticed that this will affect them 
to some slight extent, but the dates given show too 
much regularity for this to be the true cause. My be- 
lief is that it is due to the food supply. Having ex- 
amined thousands of the crops of these birds I found 
that over ninety per cent. contained the buds of a 
species of willow, popularly known here as pussy wil- 
low, Salix arctica? The balance were buds of the 
birch, poplar and the mountain ash and its berry. I 
also noticed a few seeds that I could not identify. 
“After a year or two of great abundance, all the 
willows are destroyed by the breaking of the tips and 
the buds, and the shrub takes about two years to re- 
cover, which is generally by fresh sprouts from the 
roots. As the food fails the birds have to move else- 
where. It would, therefore, appear from my data 
that it takes about ten years to go around their circle 
of migration. I say circle because their line of flight 
seems to indicate this. They first appear on the 
