328 BOTANICAL GAZETTE [may 



directly to its final destination on the canon floors. It is likely, how- 

 ever, that a greater amount is brought down by the streams from the 

 snow banks at the base of the upper cUffs, upon which it had fallen, 

 perhaps years before. The wet surface of the drifts would hold all 

 such fine material deposited upon it. This deposit is not so slight as 

 might be imagined, for during a single summer the drifts become so 

 coated with dust that the discoloration is conspicuous at a great dis- 

 tance. The fate of the dust thus deposited is as follows: Most of 

 the run-off from a large snow bank in an arid mountain region comes 

 from the melting of the compact basal portion. The water which 

 results from surface melting nearly all evaporates. Thus the layer 

 of dirty snow upon the surface of the drift is slowly but continually 

 sinking. Each winter a fresh layer of snow is deposited, which, during 

 the following summer, accumulates a layer of dust upon its surface. 

 Frequently a section through a snow bank will reveal a series of these 

 dirt bands, roughly marking off what remains of each winter's snow- 

 fall. It is evident that because of the basal melting, these dirt bands 

 will finally sink to the bottom of the drift, and that the matter Avhich 

 the dirty layers contain will be carried down with the stream of water 

 which issues from beneath the snow bank, often with great velocity. 

 This material, converging from numerous sources to a few restricted 

 drainage areas, is by no means to be "neglected, and being chiefly 

 organic in nature, it is a valuable contribution to the making of the 

 first soil. Being exceedingly fine material, it may be carried to a 

 considerable distance, and is deposited in basins, or wherever the 

 velocity of the' stream slackens. The disintegrated rock material, on 

 the other hand, being comparatively coarse, if carried away at all, is 

 deposited very near to its point of origin. 



As soil producers, Hchens have accompHshed practically nothing 

 in the glacial canons. The fact that the rock is a hard granite, and 

 the smoothness of its surface, as well as the comparatively short 

 time that the rock has been exposed, are sufficient to explain this. 



B. The invasion of the canons.— The'unimportance of lichens 

 in preparing for the establishment of the higher plants has just been 

 mentioned. The advance of the latter must therefore be conditioned 

 by the other sources of soil material, rather than by any furnished 

 by lichens. And so we find two groups of invaders, one making use 



