MUSK-EATS AND SQUIRRELS. 77 



gest themselves. That it is in some way connected with 

 their food-supply in winter is probable, and there may 

 be some connection between the habits and the general 

 condition of the springs and the abundance of surface 

 water. I have thought that when the springs were "up" 

 and the swamps full, in autumn, these houses were more 

 likely to be built, and vice versa; but I have noticed 

 several exceptions, and do not feel assured that I am right 

 in my surmise. At all events, there is no reasonable 

 ground whatever upon which to base the assumption that 

 musk-rats can foretell meteorological changes, beyond such 

 as may occur within twenty-four or thirty-six hours. Far 

 be it from me to deprive any of our mammals of all due 

 credit ! On the contraiy, my whole sympathy is with 

 that view of their origin and of man's that gives the same 

 powers, but in different degrees, to them and myself ; but 

 there is as yet nothing discovered in the life of a musk-rat 

 that even vaguely hints at the possibility of a prophetic 

 power beyond the capabilities of any ordinary weather- 

 witch. 



Let us now consider, in this connection, the habit of 

 many mammals of storing away quantities of food for 

 winter use. As the musk-rat does not hibernate, and is 

 not affected in its movements by the severity of winter 

 weather, it requires no such magazines of supplies, and 

 consequently has none. The winter-houses it erects have 

 a bearing, I doubt not, upon the food-supply, which is 

 probably more accessible from these structures than from 

 their ordinary burrows. On the other hand, many mam- 

 mals have but one class of retreat, summer and winter, 

 and these gather in autumn, for winter use, enormous 

 quantities of nuts and seeds, which are their sole depend- 

 ence. Now, the same question arises here as in the case 

 of the musk-rat : Have these " harvesting " animals any 



