Jan., 1919] Effects of the Eruption on Vegetation 179 



only 8 to 12 inches, 4 while at Katmai it was 40 feet on the 

 crater rim and 15 feet at a distance of five miles. The flat flood 

 plain of Katmai Valley with its forest of large trees is the place 

 which should have shown the most definite evidence of hot 

 tornadoes. But this was swept by a terrific flood which so tore 

 up the dead forest as to cover up any lesser damage which may 

 have preceded it. Nevertheless it must be noted that the trees are 

 still standing everywhere throughout the Valley, except in areas 

 swept bare by the flood. On the other hand, while violent hot 

 blasts like those that devastated St. Pierre have not been 

 known to retain their power for great distances from the vent, 

 the dead trees under discussion were some of them situated as 

 much as 25 miles away. 



But, while there is no evidence of hot blasts of tornadic vio- 

 lence, it is difficult to imagine such widespread destruction as 

 we are dealing with as due to other causes than withering blasts 

 from the volcano, and there is considerable evidence that these 

 were the agents of destruction. Those trees that survived in 

 areas of otherwise complete destruction were invariably so sit- 

 uated as to be sheltered from winds coming from the direction 

 of the crater. 



In lower Katmai Valley, which bends so that the east side is 

 sheltered from the volcano while the west side is exposed toward 

 it, all trees were killed on the exposed west side and in the 

 middle of the valley, but under the protecting mountainsides on 

 the east bank isolated buds on many trees survived. The 

 destructiveness of the eruption was even more mitigated in 

 Soluka Valley, which, although only half as far from the crater, 

 was sheltered by the ridge of the Barrier Range. (See map.) 

 This difference is made the more striking by the fact that while 

 the ashfall was six feet in Soluka Creek it was less than two 

 feet at the east side of Katmai Valley, and less than a foot on 

 the west side where the destruction to trees was greatest. 



At Russian Anchorage in the left arm of Amalik^ Bay, 23 

 miles from the vent, which the writer visited the year after the 

 eruption, the alders on the mountainsides exposed toward the 

 volcano showed no signs of life while on the opposite sheltered 

 slopes they were green with new leaves. This effect could 

 hardly have been produced except by withering winds from the 



4 Worcester, D. C. National Geographic Mag. 23 : 359. 1912. 



