RESULTS FROM AVALANCHE ACTION 413 



ash accumulated to a thickness of 15 to 30 centimeters or more on the 

 steep part of the old cone near the crater. From time to time the coat of 

 new material became water-soaked from the heavy tropical rains and slid 

 down the mountain in more or less of a sheet avalanche. On the collect- 

 ing ground of the steep upper cone planation and grooving were not prom- 

 inent, but on the middle ground of the Morne Saint Martin, where the 

 force of the avalanches spent itself, planation and grooving were pro- 

 nounced. In June, 1902, the striated surface of the old agglomerate, 

 with here and there a heap of unassorted ash upon it, suggested closely the 

 appearance of a regularly glaciated surface with its overburden of till. 

 At the lower end of the slope of the morne there is a flat transverse valley, 

 with a short, abrupt rise below it, before the lower slope of 3 to 5 degrees 

 begins. The avalanches were checked here and diverted into the canyon 

 of the Eiviere Seche, apparently doing no more planation of the kind 

 described. 



Eounded and subangular pebbles and bowlders were seen in abundance 

 in the new ash, but none was noticed showing striations to indicate that it 

 might have been an agent or one of the tools in producing the sand-blast 

 or the avalanche abrasion, though it seems as if diligent search might 

 have brought some to light. 



SAND-BLAST ACTION ON THE SOUFRIfiRE 



Such sand-blast work as that just described as occurring in many places 

 on Mount Pele was likewise observed on the Soufriere of Saint Vincent, 

 but through the nature of the old agglomerate and the fact that the soil 

 was not so extensively removed, the evidences of the abrasion seem now to 

 have been obliterated by atmospheric erosion and the growth of vegetation. 

 In 1902 an interesting feature of the sand-blast erosion was the sharpen- 

 ing and charring of the ends of tree roots and branches that pointed 

 toward the crater. Planing due to the sliding of the new ash over the 

 old agglomerate ridges was not noted in Saint Vincent, but the eroding 

 power of sliding water-soaked ash and ash-saturated water was manifest 

 in another and perhaps more interesting fashion — by the excavation of 

 the U-shaped valleys about to be described. 



U-shaped Valleys 



Several of the radial valleys on the Soufriere of Saint Vincent show the 

 U-shaped cross-section in a rock gorge that is usually thought to be con- 

 fined to valle} r s that have been excavated by the action of glaciers. These 

 Saint Vincent valleys, though small, are typical, the best and most accessi- 



