2 go The National Geographic Magazine 



The material consists of small lapilli, 

 sand, and dust ranging from the diam- 

 eter stated down to particles invisible to 

 the naked eye. Its color is dark gray. 

 Plagioclase crystals and fragments are 

 abundant ; hypersthene and grains of 

 magnetite somewhat less so ; but at least 

 half of the mass is dark microlitic, more 

 or less felty, but not vesicular ground- 

 mass, often inclosing or clinging to 

 crystals, and appears identical with the 

 groundmass of the lavas of Mont Pelee 

 antedating the last eruption. 



There is a small amount of clear glass, 

 which may represent the molten mate- 

 rial in which the gas moved to accumu- 

 late for explosion, and this view is ren- 

 dered more probable b} r the occurrence 

 of particles of pumice similar to 1 6a and 

 1 6b ; but it appears certain that the 

 greater portion of the material which 

 fell on the Roddam was derived from the 

 pulverization of solid rock about the vol- 

 canic vent of Mont Pelee, and only a 

 small part from the molten magma 

 which was the seat of the explosion. 



Sand and Dust from the Eruption of 

 May 20. — The steamship Potomac was 

 in the harbor of Fort de France, about 

 20 miles from the place of the explosion 

 on the slopes of Mont Pelee, May 20. 

 Specimens 23 and 24, which are iden- 

 tical, illustrate the character of the sand 

 and dust that fell at Fort de France on 

 that occasion. The color is pale buff 

 gray, and the range in size of particles 

 up to about one millimeter. Having 

 traveled through the air much farther 

 from its source than that collected on 

 the Roddam May 8, it is much finer. 



The particles are too fine for discrim- 

 ination with the naked eye, but under 

 the microscope are found to be crystals 

 of plagioclase, hypersthene, and mag- 

 netite, mixed with a larger proportion 

 of dark microlitic groundmass, just as 

 in the sand from the Roddam of May 8, 

 excepting that the proportion of ground- 

 mass to crystals seems larger at the 

 greater distance, and this is to be ex- 



pected , for the brittle and lighter ground- 

 mass pulverizes and floats in the air 

 more easily than the mineral particles. 

 Some particles of clear glass occur, but 

 they afford not more than a trace of the 

 magma represented b}^ the pumice. 



A comparison of the material ejected 

 May 8 and May 20 discovers under the 

 microscope no certain difference between 

 the two magmas, but chemical analysis, 

 which deals with a larger quantity than 

 a thin' section, is more likely to recog- 

 nize small differences. For this reason 

 anatyses were made by Dr Hillebrand, 

 and are given in the following table 

 (2 and 3, p. 291). 



Cinders from the Streets of St Pieire, 

 Martinique. — One of the specimens, No. 

 22 in the collection, has the above label, 

 and represents the character of the rock 

 rain upon that ill-fated city. It con- 

 tains products of both eruptions — May 

 8 and May 20 — and is a mixture of all 

 sizes of fragments from .001 millimeter 

 up to two centimeters in diameter. The 

 largest pieces, of which there are only 

 a few in a quart of material, are com- 

 posed of smaller fragments cemented 

 by a dark substance which Dr Hille- 

 brand has tested and found organic. 



In composition this specimen is like 

 21 and 23, excepting that it is some- 

 what coarser, and contains organic mat- 

 ter added after the eruption. One small 

 fragment of hornblende was fonnd and 

 most likely came from one of the older 

 lavas, for none was observed in the 

 pumice. 



In the rain of volcanic ejecta at St 

 Pierre the fine and light were carried 

 down with the larger fragments, but it 

 is evident that much of the finer and 

 lighter material would be shot upward 

 by the explosions to great heights above 

 the large, heavy fragments to be spread 

 far and wide b}^ upper currents. Be- 

 yond the Island of Martinique the vol- 

 canic sand and dust from Mont Pelee 

 are not represented in Mr Hill's collec- 

 tion. 



