ECOLOGY OF ISLE ROY ALE. 409 



of described muskrats, as given in Elliot's recent Check List ('05, pp. 

 252-255), clearly shows that very little is known of the range of these 

 common animals; and at the same time their somewhat anomalous 

 geographic relations suggest that the interrelations of these species 

 must be imperfectly understood. Similar relations are suggested by 

 the data concerning zibethicus, of which there are five varieties in ad- 

 dition to the typical form. This latter form ranges from Labrador 

 to the Gulf States and northward, east of the Rocky Mountains, to 

 Keewatin. As to its occurrence in earlier geological deposits, musk- 

 rat remains have been found in the Pleistocene deposits of South 

 Carolina, New Jersey and Pennsylvania. 



The post-Glacial changes within the glaciated portion of their 

 range presents an interesting problem. Thus starting with a 

 poorly drained glacial topography, the inwash from the sur- 

 rounding hills, the accumulation of vegetable and animal re- 

 mains, the perfecting of drainage lines due to the down-cutting of 

 outlets, and other effects of running water, would all tend to encroach 

 upon the poorly drained areas and convert them into dry land habitats; 

 while, at the same time, there would tend to be a corresponding in- 

 crease of stream habitats to a certain degree. Such changes as these 

 would begin on the surface first exposed by the retreat of the ice; and 

 since the ice retreated in a northerly direction, the southern margin 

 of this drift would first be exposed to the general processes of meta- 

 morphism (in the sense of Van Hise) in the zone of weathering (so 

 far as the soil was concerned) and to erosion (so far as the topography 

 was concerned) ; and as the retreat continued these processes would 

 extend their range of influence northward, and thus give to the en- 

 vironment a definite dynamic trend. 



FoBtunately, direct observation clearly shows that the processes just 

 outlined in a general way have been active on all the drift surface. 

 In general, the drift first exposed is the most metamorphosed and 

 eroded and the best drained, while those regions which were later 

 exposed are less metamorphosed and eroded, and imperfectly drained. 



This gives ground for the opinion that as the ice retreated to the 

 north there has been a general extinction, from the south northward 

 of the poorly drained habitats whose origin was due to the glacial 

 topography. Perhaps a more definite statement of these effects, from 

 the standpoint of processes, would be that the direction of extinction 

 was a resultant, determined by the direction of the ice retreat and the 

 lines along which the drainage later developed. 



The above remarks on the dynamics and history of the muskrat en- 

 vironment are of special interest on account of their bearing upon two 

 problems as follows: first, the probable post-Glacial migrations of their 

 optimum environment, as it thus seems probable that there has been 

 a post-Glacial northward migration of the most favorable habitat for 

 the muskrat; and second, on account of its influence upon the habits 

 of muskrats. The muskrat is essentially a burrowing animal, and this 

 is perhaps an older habit than house building. It is therefore of interest 

 to know that the muskrats of the southern range are primarily bur- 

 rowers, rather than lodge builders. In the south, below the glacial 

 lake area, they frequent the sea coast, coastal plains, and streams, 



