247 
a burrow beneath snow the surface of which was frozen 
into ice. In the next place, the enemy must approach 
the grouse through the burrow as a ferret does in the 
case of a rabbit’s burrow; and, in the third place, the 
burrow must be long enough to give the animal a suffi 
cient warning and opportunity for escape. Now, it would 
appear that, in Russia, the grouse do not burrow in the 
snow, until it is at least an inch and a half deep, nor 
towards the end of winter when crusts of ice form on 
the top of the snow. The birds burrow in the snow that 
lies lightly and through which it is easy to fly. Thus the 
first condition of safety is secured. As to whether the 
beast of prey would approach the grouse through the tun 
nel, that does not seem to me to be so absolutely certain. 
We are told that the bird, when it has reached the end 
of the burrow, “ pokes a hole through the top with its 
head for air.” Now, if we may credit the attacking animal 
with an ordinary amount of intelligence, we may suppose 
that it will understand that the grouse lies under this 
air-hole; in which case it is quite possible that it might 
approach it from above. 
In Russia, I am told that the fox will scratch the snow 
away to get at the field-mice beneath; and if the fox 
can smell the mice beneath the snow, it is reasonable to 
suppose that it would still more be able to find the grouse, 
and the sleeping animal, being attacked from above, would 
not be able to fly away. In this case, it is obvious that 
the length of the burrow would make no difference to the 
safety of the animal. But let us assume, for the sake of 
argument, that the attack is made through the burrow. 
We are instructed to picture a state of things in which 
the precise length of the burrow will make the difference 
between the life and death of the individual, between 
the extinction or preservation of the race. In this safe