43 



The National Geographic Magazine 



ord from which to judge of the nature 

 of certain phases of several of the recent 

 eruptions. In the instance cited, there 

 does not seem to have been a loft}' col- 

 umn of dust-charged steam standing 

 above the summit of the mountain, 

 which deflected the upward blast from 

 the vertical conduit of the volcano, but, 

 owing to the density of the mixture of 

 steam and dust driven out, it overflowed 

 the lip of the crater and rolled down the 

 mountain side. That is, one condition, 

 as previously stated, which may bring 

 about a marked variation in the nature 

 of an eruption, and in fact furnish the 

 chief control of the secondary phe- 

 nomena, is the density with which the 

 steam extruded is charged with solid 

 matter. This condition may obtain 

 control even when the explosive vio- 

 lence is not enough to drive the dust- 

 laden steam to a great height. The 

 degree of comminution would no doubt 

 be another factor influencing the result. 

 The fiuer the solid material was com- 

 minuted, the more fluid-like would be 

 the mixture. 



In an eruption like that described 

 above the topography may exert a de- 

 cided influence. During eruptions of 

 great but of not the maximum intensity 

 the deep notch in the southwest portion 

 of the rim of the active crater of Mont 

 Pelee would give direction to the escap- 

 ing dust-charged steam and determine 

 the course the expanding avalanche 

 would take. In the absence of such a 

 notch, as in the case of La Soufriere, 

 the overflow would be radial. A more 

 intense eruption from Mont Pelee might 

 also be radial, the notch in its rim fail- 

 ing to influence so completely the direc- 

 tion taken by the greater discharge. 

 This is what seems to have occurred 

 during the later eruptions of the vol- 

 cano, when Morne Rouge and other 

 villages were destroyed. 



Mud Flows. — The valleys on the lower 

 slopes of both Mont Pelee and La Sou- 

 friere have in numerous instances been 



filled to a depth of 40 to 60 or more feet 

 with hot dust and stones. The streams 

 have thus been displaced, and are striv- 

 ing to regain their right of way, but as 

 3 r et, owing largely to the washing down 

 of dust, lapilli, etc., from bordering 

 slopes, are making but slow progress 

 with their work. In many instances, 

 in fact, the high-grade rills are bring- 

 ing to the main drainage channels more 

 debris than the master streams can re- 

 move, and the process of valley-filling 

 continues. Water finding its way into 

 these beds of hot debris, as already 

 mentioned, causes steam explosions, 

 sometimes of such energy as to resem- 

 ble a primary eruption from one of the 

 main or true craters. These eruptions 

 at times hurl large quantities of debris 

 into the stream channels which have 

 been partiall}' cleared, thus producing 

 dams and causing small lakes to form. 

 These water bodies rise until the}' over- 

 flow the accumulation of loose material 

 restraining them, when they are rap- 

 idly drained, and floods of water heav- 

 ily charged with debris occur below 

 where the temporary dams were formed. 

 The mud flows originating in these and 

 other similar ways have been frequent 

 on both Martinique and St Vincent, and 

 have in several instances been referred 

 to as lava flows. 



Erosion. — Much that is highly in- 

 structive centers about the manner in 

 which the surface waters are removing 

 the f reshly added material from the sur- 

 faces of Martinique and St Vincent. 

 Instead of being a protection to the sur- 

 face on which it rests, the fresh debris 

 is in many instances of assistance in its 

 more rapid erosion. On steep slopes 

 and even when the surface is nearly 

 level, the rills formed during the numer- 

 ous tropical showers quickl}' cut through 

 the loose surface material and, aided by 

 the angular particles in suspension, cor- 

 rode the soil or rocks beneath. The 

 rains, as it seems, are heavier than 

 usual, owing to two causes: First, the 



