SOUFRIERE, AND ON A VISIT TO MONTAGNE PELEE, IN 1902. 
489 
one-tenth of that which the Sonfriere has furnished. The great avalanches of dust 
which fill the Rabaka and Wallibu valleys enormously surpass in magnitude any 
deposits of the same kind in Martinicpie, and the general sheet of deposit over the 
Sonfriere is not only larger but j^robably also five times as deep as that which can be 
found on Pelee. The dust from the Sonfriere fell in Barbados (1,700,000 tons), and 
all over the sea for 700 miles to the south-east of this. 
The cloud which surged from the crater of Pelee had only one outlet. On north 
and east it was hemmed in by walls of rock ; only to the south-west could it find a 
fissure of escape, and through this it rushed and swept down the slopes on the plain 
beneath and on the devoted city. It spread out somewhat, and the area it covered is 
fan-shaj^ed, hut all its violence was concentrated in that small space, and the havoc 
it wrought was fearful in consequence. Proljably the force of the hot Idast was as 
great in Martinique as in St. Vincent at equal distances from the point of origin. 
But in St. Vincent the total enei'gy expended during the great explosion was vastly 
greater tlian in Martinique, for the avalanche poured over the whole lip of the great 
circular crater, though mostly over the notch on the south-west side, and radiated 
out in all directions from the centre. Still, that the blast in Pelee was so violent, 
while the amount of material ejected Avas so small, is one of the most interesting- 
features of the eruptions, and calls for special discussion when we come to consider 
the mechanism of these discharo’es. 
O 
In St. Vincent no fissures are known to have opened on the hill-sides, no flows of 
hot mud lava have AA'elled out of the mountain, and there is a complete absence of 
* fumaroles or minor steam emissions, exce})t from the boiling lake Avithin tlie crater. 
But in Martinique early in the eruption the ri\"ers began to floAv full of muddy AA'ater, 
/ 
and on May 5th the Etang Sec broke through its barriers and, dischai'ging Into the 
Iliviere Blanche, sent doAvn a torrent of hot mud Avhich buried the Usine Guei'in, 
Fissure formation and hot mud-floAvs did not end here, they are one of the dominant 
features of the eruption of Pelee, and some interesting facts regai'ding them are 
contained in the Preliminary Report of the French Commissioners.'^ Along the course 
of the Riviere Seche, and in the district betAveen that and the Riviere Blanche, 
numerous fumaroles have appeared which indicate the presence of a series of fissures 
in that quarter. They are probably radial in direction, and may be crossed by 
another series (tangential) on which lie certain fumai'oles near the sea-shore. On the 
north-Avest side of Montague Pelee tliere is, at Ajoupa Bouillon, a fumarole Avhich lias 
poured out large quantities of muddy Avatei', and this may lie on a prolongation of 
the radial fissures on the south-east side of the cratei-. They are not cliaracterised 
by gaping cracks, but rise througli the old or iieAv ash deposits, and giA^e evidence of 
their presence by emanations of steam and sulphuretted hydrogen. Some of them 
are at a temperature sufficiently high to melt pieces of lead Inserted into the orifice. 
* MM. A. Lacroix, Rollet de lTsle & Giraud, “ Siu- I’Eraptiou de la Martinique,’' ‘ Comptes 
Rendus,’ vol. cxxxv., p. 381, 1902. 
3 11 
VOL. CO.—A. 
