388 Andu. Lcndagkr 



when trodden upon. This is often found to he the case in localities 

 with ;i continuous covering of plants, e.g. Cassiope. On the other 

 hand, at the foot of steeper snowdrifts, these hollows are not often 

 known to occur, the reason heing that only at its outermost edge 

 is the snow-covering thin enough to permit the radiation to penetrate; 

 besides, at its foot, a steep snowdrift easily becomes ice, owing to 

 the melting snow, and then forms a compact mass on the ground. 

 At the foot of such a perennial snowdrift I found, late in the year, 

 only a few moss-fragments and a crust of Xostoc with Gloeocapsa 

 mayiia and Phormidium aiihimnale. The presence of the moss- 

 fragments proved that the snow had not always been lying there; 

 but in this place it was not possible to demonstrate which species 

 approach nearest to such an ice-edge, and are the first to take pos- 

 session of the ground. When such an enormous snowdrift melts, 

 the result often is that the ground below it remains a barren slush 

 of clayey mud, until it again becomes frost-bound. 



