336 VULCANIC EEUPTIONS ON MARTINIQUE AND ST, VINCENT. 



inu) the air. If, as .seems probable, the energy displayed by steam 

 columns of the No. 2 tj^pe is all than can be supplied by the steam pro- 

 duced from rain and jjercolating water in the upper part of the con- 

 duit, it follows that durino- explosions of the No. 3 type both rock and 

 steam are rising from a depth in the volcanic conduits. Presumably, 

 then, during- eruptions of the types No. 3 and No. -i molten rock is being- 

 forced up within the conduit of a volcano and, owing to relief of pres- 

 sure as it rises, the steam dissolved in the molten magma escapes with 

 tremendous violence. There are thus two sources for the steam which 

 furnish the energy displayed in the summit portions of ascending lav^a 

 columns — one from the rain and percolating water, and the other from 

 a deeper but unknown source. But this attempt to follow the volcanic 

 conduits downward in fancy has brought us to the region of specula- 

 tion and it is time to stop, at least for the present. 



To the four types of volcanic-steam columns dcscriljcd above, a lifth 

 might be added to include volcanic explosions like that of Krakatoa. 



Prodiidx. of the eruptions. — The material discharged from Mont 

 Pelee and La Souf riere may be divided into two portions — lirst, steam 

 and gases, and, second, solid rock debris. Up to the present time no 

 observations indicate that molten rock has been extruded; that is, no 

 lava streams have flowed over the surface from the crater of either 

 volcanoes or issued from fissures in their sides. 



As to the discharge of vast volumes of steam, there is no difi'erence 

 of opinion to be formed in the various reports already rendered. 

 Observers who have visited the craters from which the recent erup- 

 tions came, and have even ventured within them, report onh' faint 

 traces of gases. The conditions, however, between the time when a 

 crater is quiescent and when violent explosions occur within it are 

 no doubt different, and as yet but little evidence concerning the gases 

 that mav have been present during the times of most violent activity 

 has been ol)taincd. The most that can be accepted in this connection 

 is a plainly perceptible odor of sulphui'ous acid noticealde in the air, 

 even at a distance of some 8 miles at sea, when the \<)l('anoes were in 

 a compai'ativ(dy mild state of activity and while walking over the 

 debris they sliowered on their respective islands. A nuich fainter 

 odor of sulphuretted hydrogen is reported to have l)een present, as, for 

 example, among the ruins of St. Pierre, but whether due to gas emit- 

 ted from the volcano or arising from organic matter buried beneath 

 the still hot debris is not clear. The presence of carbon dioxide, 

 although asserted or surmised to have been discharged during the 

 greater eruptions, has not been proven. The consideration of :dl the 

 available evidence points strongly to the conclusion that steam was the 

 chief vaporous or gaseous substance^ emitted, but mingled with it were 

 minor (juantities of sulphurous and no doubt other gases. 



In the above connection it should be noted that flames above the 



