387] H. E. Pulling 189 



formed from rock flour in the bed of ancient lake Aggasiz. 

 The upper limit of frozen soil encountered during the sum- 

 mer varied from a depth of a few inches, near the water's 

 edge on a shore with a north exposure, to about six feet on 

 a slope well above the water line and with a southeast ex- 

 posure. It is highly probable that one of the most effective 

 agencies conditioning local distribution of plant species is 

 the depth at which frozen soil is encountered. Large trees 

 and other deeply rooted plants could not exist in soils made 

 too shallow by the presence of perpetually frozen soil near 

 the surface. 



The soil of the spruce forest, which is the characteristic 

 type of this region, is covered chiefly by sphagnum, often to 

 a depth of several feet. Large amounts of water are held 

 by the moss so that these forests resemble those of temperate 

 regions at the borders of swamps and marshes. If the forest 

 is situated on a hillside, however, the soil underneath the 

 moss is usually dry and if exposed in windy weather will 

 blow as dust. This may perhaps be explained in the light 

 of knowledge of the conditions above and below the dry 

 layer. This dry stratum rests upon frozen soil, which be- 

 cause of its lower vapor pressure and of other probably less 

 effective properties, should continually absorb water from 

 the adjacent, unfrozen soil. Thus, making-the easily justified 

 assumption that the soil was originally wet, the conditions 

 exist for almost completely drying it, provided it should not 

 regain the water lost. The light snowfall in this region 

 would be unlikely to produce large amounts of water in the 

 spring, especially on these slopes where drainage in the spring 

 is rapid over the frozen surface of the soil, the relatively 

 small amount remaining being conceivably retained by the 

 highly absorbent sphagnum covering. The summer rains, 

 which, although frequent, bring comparatively little water, 

 are apparently no more than sufficient to supply the trans- 

 piration loss of the plants exposed to almost continuous winds 

 and often to bright sunshine for many hours a day. 



Eoots penetrate this dry layer only to a slight extent, 



