Volcanic Disturbances in West Indies 289 



is evident. Flow lines occur in the 

 glass about the elongated vesicles, and 

 its whole aspect indicates clearly that 

 the material now glassy was once soft 

 and fluent like paste, and while in that 

 molten condition contained gases or 

 vapors which expanded and formed 

 vesicles, whose elongation took place in 

 the direction of motion within the mass. 



Here we have the seat of expansive 

 and explosive energy, which must have 

 played a most important part in the 

 eruption, and we may briefly consider 

 it, but only from a petrographic stand- 

 point. Many of the vesicles completely 

 closed must still contain in some form 

 more or less of the gaseous agent that 

 produced the vesicle, but these are 

 almost wholly lost in the preparation of 

 thin sections for microscopic study, and 

 their ultimate determination must be 

 relegated to the chemist. There can 

 be no doubt, however, that a large part 

 of the gaseous matter given off is steam. 

 The source of this water and other gases 

 contained in the molten material, magma, 

 as it rises to the earth's surface in erup- 

 tion we will not discuss, but what hap- 

 pens to it in the process of crystalliza- 

 tion concerns the specimens we are con- 

 sidering. 



Crystallization is a process of exclu- 

 sion ; so it happened that as the magma of 

 Mont Pelee approached eruption crystals 

 of magnetite and hypersthene developed 

 with a still larger number of plagioclase, 

 and the absorbed gases rejected in this 

 process became concentrated in the un- 

 consolidated portion. When the press- 

 ure was relieved the liberated gases 

 expanded, producing the vesicles and 

 giving rise to pumice of different de- 

 grees of porosity. Specimen i6«, al- 

 though very vesicular, is much less so 

 than 16b, for it quickly sinks in water, 

 while 16b readily floats. 



We can easily imagine the process of 

 expansion toadvance beyond the pumice 

 stage even until the bubbles burst and 

 the whole mass be blown to fragments, 



giving rise to lapilli and dust, of which 

 the collection contains a number of ex- 

 amples. The gases filling the vesicles 

 were the most mobile part of the mass, 

 and like bubbles in water responding 

 to the common impulse, moved in the 

 same general direction, but slowly, and 

 accumulated to form the big bubbles 

 which produced the explosions. The 

 eruptions of Mont Pelee, as in vol- 

 canoes generally, are a series of explo- 

 sions due to the rupture of great bub- 

 bles formed from the accumulation of 

 smaller ones like those of the pumice 

 just described. 



According to Professors Hill and Rus- 

 sell, there were no regular flows of 

 molten lava from Mont Pelee during 

 the recent eruptions. The magma was 

 so full of absorbed gases that it was all 

 blown out before effusion took place. 

 The only flows were the so-called ' ' mud 

 flows," consisting of ejected material, 

 lapilli, sand, and dust, so saturated 

 with water as to flow in landslide 

 fashion on slopes sufficiently steep. 



The fragments of pumice described 

 above probably represent more closely 

 than any other material we have from 

 Mont Pelee the general composition of 

 the erupting magma. A chemical 

 analysis by Mr W. F. Hillebrand is 

 No. 1 in the table, page 291. Having 

 considered the erupting magma, the 

 ejected material will be taken up, be- 

 ginning with that of May 8. 



Sand and Dust from the Eruption of 

 May 8. — The British ship Roddam was 

 in the harbor of St Pierre at the time 

 of the great explosion on Mont Pelee, 

 about 7.45 a. m. , May 8, and specimen 

 21 is dust which fell on the deck of the 

 Roddam at that time. 



The material was washed and sepa- 

 rated by decantation to facilitate study. 

 The largest particle of the specimen has 

 a diameter of about 1 . 5 centimeters ; 

 but, judging from reports, much larger 

 fragments must have fallen on deck at 

 that time. 



