338 The Ohio Journal of Science [Vol. XIX, No. 6, 



seeds which otherwise would have drifted clear across the 

 barren flats without finding any lodgment. The seeds thus 

 caught had sprouted and grown wel 1 , proving that in that 

 particular place, at least, the principal deterrent to re vegetation 

 was the inability of seeds to catch hold. When examined a 

 year later, the plants growing up in the tracks had made a 

 notable growth, as may be seen by comparing pictures taken 

 in the two years. (See pages 336 and 337). 



WIND EROSION A GREAT DETERRENT TO REVEGETATION. 



But the effect of the wind on vegetation is not to be measured 

 merely by its influence as a seed disseminator. Much more 

 important is its effect on the soil itself. It is so violent that it 

 keeps the surface of the ground over large areas always in an 

 unstable shifting condition, so that plants have little oppor- 

 tunity to start. 



The wind is, indeed, one of the most important factors 

 retarding the revegetation of the devastated district. In 

 another place I have shown how important it is in the vicinity 

 of Kodiak.* Near the Volcano the total devastation and the 

 conformation of Katmai Valley give it a clear sweep so as to 

 greatly intensify its effects and augment its importance. 



Here the snowdrifts which accumulate during the winter are 

 buried under a mantle of wind blown sand which is often more 

 than half a meter thick. Our observations on such drifts in 

 1916 showed uniformly that the sand had all accumulated after 

 the snow fell, for it lay as a sharply distinct layer on top of the 

 snow and the two were not interbedded. This indicates that 

 it was all accumulated during a short period in the spring. 

 (See page 339). Such sand-blanketed snow is very slow in 

 melting and in places shows little wastage even as late as the 

 first of August, which of itself is a factor of considerable moment 

 in retarding the renewal of vegetation on the snow covered 

 areas. 



The abrasive power of this shifting sand, as it is carried by 

 the wind, is very considerable. There are large tracts in the 

 upper valley in which the sandblast has cut away the bark 

 and even abraded the wood on the northwest side of the dead 

 trees, leaving them uninjured on the lee side. A forest of such 

 trees is most striking testimony of what the wind can do. 



*The first paper of this series, pages 37-3&. 



