488 
DRS. T. ANDERSON AND J. S. FLETT ON THE ERUPTIONS OF THE 
chemical composition already published. It was astonishingly uniformly fine-grained ; 
small lapilli were present, spongy, brownish-grey, and porphyritic, and were most 
numerous where the finer stuff had been removed by the rains, but large bombs are 
rare, except on the slopes around the crater, and we saw none near the city of 
St. Pierre. There were a few ejected blocks, so few that most of them already bore 
the marks of the hammers of the French and American geologists, who had been 
there before us. They were brittle, and flaked off on the surface like those already 
described as found on the Soufriere. All that we saw were old and somewhat 
decomposed porphyritic andesites. 
The products of the magma of Pelee were distinctly more acid than those of the 
Soufriere, and more of the dust consi.sted of felspar, which gave it a pale grey colour. 
It is less rich in the ferro-magnesian silicates, especially augite, but contains a fair 
amount of hypersthene and some hornblende. No olivine has been reported. Mixed 
with the broken and entire crystals of these minerals are the remains of a base 
or ground-mass, which is often filled with fine microlites of hypersthene,"'" and 
this material of the later period of consolidation is, on the whole, more abundant 
than in the dust of the Soufriere. The felspars are also more rich in the albite 
molecule, and contain less of anorthite. 
The area of devastation in Martinique is not onl}^ sharply defined, it is small, 
much smaller than that of St. Vincent, and occupies only a segment of the volcanic 
cone. Its base stretches along the coast from Cai'bet to Roche La Perle ; its apex is 
a little to the north of the summit of Pelee; its boundaries are fairly straight, but 
are apt to be deflected by the irregularities of the surface and the obstruction offered 
by the projecting lidges to the outward flow of the black cloud near its margin 
Within this region the desolation varies almost directly with the proximity to the 
crater. It is complete and total on the upper slopes and in the north end of 
St. Pierre, but decreases rapidly as traced southwards through the city and along 
the coast to Car bet. On the west side, PiAcheur is covered with ashes, but it was 
not swept by the awful Idast. The wind constantly floats the fine ash projected 
from the crater with the rising steam clouds in this direction. The trees are powdered 
over, the vegetation blighted, the streets paved with dust, but much of this is due, 
not to the great eruption, but to the subsequent activity of the volcano. 
Only about a third part of the mountain has been ravaged by the blast. The 
north and east sides have hardly suffered, or at any rate the deposit of ash on them 
was so thin that when we were in Martinique it had mostly been washed away, 
and the beautiful verdure which characterises that lovely island flourished to all 
appearance as if no eruption had ever taken place. 
The area affected is rather more than one half that which has been blasted 
in St. Vincent, and the total mass of material ejected hy Pelee, may be, perhaps, 
* A. Lacroix, “ Sur les Roches rejeteus par I’Eniption actiielle de la IMontagne Pelee,” ‘ Comptes 
Remlus,’ vol. cxxv., p. 453, 1902. 
