230 



POPULAR SCIENCE MONTHLY. 



there was no means of judging forward motion. The discharge from 

 a cannon would 'roll, squirt and seethe' without showing any forward 

 motion, if we were looking into the cannon's mouth. There can be 

 no question^ however, that there was an avalanche of solid materials 

 which accompanied the second black cloud seen by Anderson and 

 Flett, and I am tempted to ask, may there not have been a similar 

 non-incandescent avalanche with the first cloud, representing the 

 relatively cool material which cleared the crater's throat? 



Fig. 3. St. Pierre from High Slope of Pei ee : showing Course of Blast as seen from 



THE Crater. 



(Photo for T. A. Jaggar, by E. C. Rost.) 



If this were true we should have with each puff a sequence of 

 events somewhat as follows : 



1. Straight puff upwards. 



2. Avalanche of heavy materials downwards. 



3. Horizontal dust-cloud outwards. 



The sharply differentiated horizontal and vertical blasts had been 

 noted in all the eruptions of these volcanoes. They may be seen in 

 the photograph of the eruption of July 16, where one cloud lowers over 

 the water, the great column is upright, and its crest is bent to the 

 east by the high counter trade-wind. (fig- 1-) 



The three stages above mentioned, in an explosive discharge from 

 a fissure upward, might have been expected on purely a priori grounds. 



