224 HOWARD E. PULLING 



occupied by Lake Winnipeg and surrounding country and, ex- 

 tending northward, includes the western part of the Churchill 

 river to a point below Southern Indian lake. The Nelson river, 

 rising in Lake Winnipeg, emerges from the old lake bed about a 

 degree south of Split lake and does not reenter it. In this lake 

 bed a heavy deposit of silicious clay has rounded the rugged out- 

 lines of the pre-cambrian rock and produced a rolling country. 



One feature of the topography is the occurrence of muskegs 

 on rather high hills where precipitation accumulates in cups 

 dammed by the ancient rock or by the deposit of compact clay 

 barriers that are but slightly permeable to water. Another note- 

 worthy feature is a sharp discontinuity in the depth of soil, for 

 bare rock may alternate with deep drift and of two neighboring 

 hills, one may be solid rock and the other of glacial origin. How- 

 ever, because of temperature conditions these apparently deep 

 soils are, in reality, often verjr shallow from the standpoint of 

 penetrability. Except in the more southern portions of the 

 region they are perpetually frozen to within a relatively short 

 distance of the surface and mechanically this frozen layer is, of 

 course, about as effective a barrier to root growth as rock would 

 be. 



The point to be emphasized in this paper is that through the 

 combined agencies of past history and present climatic conditions 

 the soils of the far north are, in general, shallow. In about 

 latitude 56°N. and longitude 97°W. for example, frozen soil was 

 encountered in August, 1916, at depths ranging from 2 meters on 

 bare slopes facing south to 3 cm. on flat beaches with a north 

 exposure. Because of these mechanical barriers to root penetra- 

 tion large trees of deep rooted species would not be expected 

 unless this habit became modified, for an inadequate root system, 

 of course, precludes the possibility of an extensive top. 



If one considers the characteristic form of the visible portions 

 of trees, habits so pronounced that the dominant trees may often 

 be determined at a distance, it should not be surprising to find 

 that the subterranean portions — the roots — also have definite 

 habits. These characteristic root habits are well recognized in 



