Eruptions on Martinique and St Vincent 419 



sions due to surface water gaining access 

 to beds of hot debris and the explosions 

 in the summit portions of true volcanic 

 conduits. In the former instances sur- 

 face water descends into hot rock debris; 

 and, from the fact that water is present 

 in the superficial portion of the earth's 

 crust, it seems equally manifest in the 

 latter instance that highly heated rock 

 rises from deep within the earth and 

 meets the surface waters. In each in- 

 stance steam explosions result. 



/ 'aviation in the Eruptions of the Pri- 

 mary Craters. — The variations presented 

 by the steam columns which ascend 

 from active volcanoes — of which the so- 

 called pine tree of Vesuvius is a well- 

 known example — and which in man}^ 

 instances afford the most spectacular of 

 the awe-inspiring phenomena associated 

 with them, have been described by 

 several observers who have recently 

 visited Martinique and St Vincent, but 

 most graphically by George Kennan. 

 The variations referred to are indicative 

 of what takes place in an active crater 

 and in the upper part of the conduit 

 leading to it, and furnish evidence in 

 reference to the changes there in pro- 

 gress. A classification of the various 

 phases presented by the steam column 

 rising from Mont Pelee has been pre- 

 sented by the gifted traveler just re- 

 ferred to, which is instructive. 



' ' The vapor column ascending from 

 Mont Pelee," writes Kennan, "varies 

 greatly from day to day and sometimes 

 from hour to hour, not only in density, 

 but in color, form and general appear- 

 ance. In its varying aspects it may be 

 described as follows : 



" 1 . The vapor of quiescence : a 

 slowly ascending column of pure white 

 steam, which has neither sharp, clearly 

 defined outlines, nor puff-like convolu- 

 tions, and which suggests steam rising 

 from the hot water of a geyser basin or 

 from the escape-pipe of a big ocean 

 steamer. ' ' 



The explanation of the account of 



such a column from a crater of the type 

 of the one at the summit of Mont Pelee 

 seems to be that the top of the lava 

 column is well below the bottom of the 

 crater, and that the hot rocks are dis- 

 charging steam, owing to the contact 

 with them of water percolating in from 

 the crater walls or falling of rain. The 

 generation of steam is a surface phe- 

 nomena, and due essentially to the same 

 cause as the escape of steam from hot 

 debris ejected by a volcano and accumu- 

 lating in valleys, etc. The notable fea- 

 ture is the absefice of convolutions and 

 more or less individualized fleece-like 

 masses in the ascending column, such 

 as are produced by small steam explo- 

 sions from liquid lava, as is frequently 

 the case at Vesuvius. 



"2. The vapor of moderate activity : 

 a column of greater density and some- 

 what darker color, which rolls and un- 

 folds a little as it rises, and looks like 

 steam mixed with brownish or yellowish 

 smoke from the chimney of a manu- 

 factory." 



This stage may reasonably be sup- 

 posed to indicate conditions similar to 

 those mentioned in the first instance, 

 but more intense. A considerable vol- 

 ume of water gaining access to the deep 

 funnel-shaped crater might not be va- 

 porized before descending to the summit 

 of the column of liquid or but partially 

 congealed column of rock within the 

 conduit, and energetic explosions result. 

 The steam columns indicative of mod- 

 erate activity thus correspond with the 

 columns produced by the maximum ex- 

 plosions of pseudo-craters. The inner 

 slope of the crater of Mont Pelee is pre- 

 cipitous, and, as several observers have 

 reported, portions of its walls overhang- 

 ing. The fall of blocks of rocks from 

 the crater walls would no doubt cause a 

 conspicuous column of dust-laden steam 

 to ascend. 



"3. The vapor of dangerous activ- 

 ity : a sharply defined dark-yellow col- 

 umn of what appears to be liquid mud, 



