66 PRATT. 



the lake suddenly rose about 2.5 meters. The inud at Banadero was cool and 

 fell like rain. The activity decreased gradually, and by daylight the cloud from 

 the volcano was again white. 



This main eruption awakened many people in Manila, and in Dagupan, 240 

 kilometers north of the volcano, people say they were awakened at about this 

 time by hearing a noise. The effect in Manila was that of a tremendous vibra- 

 tion, accompanied by the rattling of windows and loose doors. The noise seemed 

 to pass over the city in a great wave, so that it was heard farther away, after 

 it had ceased close at hand. No earthquake was recorded in Manila at this time. 



Eruption cloud. — The cloud which rose over Taal Volcano during 

 the recent eruption, seems to have been like those noted over other volca- 

 noes during similar times. It emerged from the crater with explosive 

 violence, roae rapidly in the air to a great height, and finally spread 

 out at the top in a horizontal layer. 



No observations were made as to the height of this cloud. A photo- 

 graph taken the following morning shows a steam cloud, being carried 

 away from the camera, which by comparison with the full height of 

 the volcano appears to be at least 3,500 meters. While the steam cloud 

 was probably higher at the time of the main eruption than at others, it 

 is doubtful if its burden of solid ejecta reached a greater height than the 

 estimate just given. 



Estimates ranging from 10 to 15 kilometers have been submitted, based on 

 photographs of the electrical discharges in the cloud, the calculation involv- 

 ing the focal length of the camera, the size of the image and the distance from 

 the camera to the cloud, which is assumed to be vertically above the volcano. 

 The probable error in these estimates arises from the fact that in all cases the 

 cloud, by its rapid lateral expansion has moved toward the photographer. The 

 lightning streak on the surface of the cloud (the streak must be on the surface 

 of the cloud toward the camera, otherwise it would be obscured and appear as 

 a flash if seen at all) is always closer to the camera than the foregoing assump- 

 tion makes it. Estimates based on measuring the horizontal angle from a 

 known point to the apparent top of the cloud are obviously liable to similar 

 error. If the lightning flash shown in Plate VII of the paper by Charles Martin 

 was directly over the crater, then it was about 4,500 meters long and 100 meters 

 wide. 



A feature of the movement of the explosion cloud not generally 

 observed by the people who watched it, probably because of the dark- 

 ness, but abundantly indicated by study of the devastated area, was its 

 terrific sweep do-\vnward and outward from the crater rim. The princi- 

 pal evidence of this movement is obtained from its effect on the vegeta- 

 tion on the lower slopes of the volcano, and the west shore of Lake 

 Bombon. The heavy growth of cogon grass was flattened absolutely, 

 and the tips point radially away from the crater. The occasional 

 patches of scant forest, except where protected by the natural topog- 

 raphy, were completely destroyed although not burned. Even on the 



