SOUFRIERE, AND ON A VISIT TO MONTAGNE PELEE, IN 1902. 
487 
windward side, alDout a mile above Lot 14, so far as the condition of vegetation in the 
city was concerned. There were no houses, ol course, in that region in St. Vincent, 
In St. Pierre, as in St. Vincent, the blast was heavy and crept along the ground. 
At any rate, when it reached the lower slopes it flowed along, and was deflected by 
j^rominent ridges in its path. The bluff behind St. Pierre directed it along the coast, 
where it swept the shore to Carhet, killing or scorching all the cocoa-nut palms 
on the low beach, and singeing the face of the steep hank, hut unable to mount 
the heights and devastate the country behind. This when we saw it was green; 
it had, jDerhaps, been covered with a thin sheet of ashes, hut the crops, the trees, 
and even the buildings were little damaged, and nowhere was the line between the 
blasted, desolate, fire-swept area and that which had Ijeen injured, hut not beyond 
the power of rapid recovery, so perfectly sharp and definite. In the length of one 
garden, near the statue of the "Virgin on the cliff top, we could find every transition, 
from utter destruction at one end to very fair preservation at the other. The edge 
of the dust cloud had been as well defined between St. Pierre and Carhet as in any 
part of St. Vincent, f 
The depth of the layer of ashes was quite inconsiderable when compared with that 
on the south side of the Soufriere. On the cane fields at the north end of the town it 
had been originally perhaps a foot, but near the sea-shore, behind the rum distilleries, 
and near the mouth of the Riviere des Peres, perhaps 3 feet or more. In the town 
itself so largely had the deposit been re-arranged by water that it was not safe to 
deduce its original thickness, and everywhere the heavy rains had scoured away the 
loose, ashy sand, and only indirectly could we infer how much had once been there. 
The upper surface was wonderfully smooth, but Professor Lacroix and his colleagues | 
describe ridges “ like sand dunes,” which reseml)le the hog-hacked mounds on the 
Wallibu Dry River. There did not appear to have been any great deposits in the 
valleys, which were mostly open and shallow, not deep, narrow ravines, hut there 
was more in such situations than on the flat, and it was obvious that much of what 
had originally obstructed the stream courses had been cut out l)y running water and 
carried to the sea. In the Riviere Blanche we Ijelieve there is some depth of hot 
sand, which, as described by Professor Heilprin and Professor Lacroix, is slipping 
into the water and giving out gushes of steam, recalling on a smaller scale the 
explosions on the Wallibu. The sand was distinctly lighter in coloui' than that of 
the ISoufriere, as may be inferred from the descriptions of its mineralogical and 
* Mr. Hovey states that “in many places the line of demarcation passed through single trees, leaving- 
one side scorched and brown, while the other side remained as green as if no eruption had occurred.” 
E. 0. Hovey, “ Martinique and St. Vincent: a Preliminary Report,” ‘ Bull. Amer. Museum Nat. Hist.,’ 
vol. xvi., 1902, p. .347. 
t MM. A. Lacroix, Rollet de i/Isle & Giraud, “Sur I’Eruption de la Martinique,” ‘Comptes 
Rendus,’ vol. cxxxv., p. 382, 1902. 
+ MM. A. Lacroix, Rollet de l’Isle Sz Giraud, “Sur I’Eruption de la Martinique,” ‘ Conqrtes 
Rendus,’ vol. cxxxv., jr. 421, 1902. 
