480 
DES. T. ANDERSON AND J. S. FLEET ON THE ERUPTIONS OF THE 
all tlie depth and picturesqiieness of those of the Soufriere, and where the volcanic 
blasts have withemd up and swept away the vegetation, the same general resemblaiice 
to the rugged, naked canons of the western American desert country is seen in both 
cases."' Vertical clitfs of lava alternate with sloping taluses of ash ; every variation 
in the underground structure is reflected in the contours of the surface, and where , 
the rich Antillean forest clothes the eroded surface, the beauty of the landscapes and 
the variety of colour lend a striking charm to the distant hills and the highly- 
cultivated shores. 
In both mountains the conical volcanic form is well exhibited, though scored 
with deep radial ravines. Pelee is 4428 feet high, while the Somma rim of the 
Simfriere is 4050, but had the latter mountain not lost by some great explosion the 
upper pait of its cone, it would pi'obably have been someAvhat the more lofty of the 
tw(r In each case the diameter of the cone at sea level is almost 8 miles, and the 
restless action of the waves has eaten back the land, and formed ranges of lofty cliffs 
vliich face the ocean. 
At the summit of the Soufriere we ha\ e a concentric structure—a crater within a 
crater. The gi'eat convulsion during which the huge Somma crater t of the mountain 
was produced, antedates authentic histor 3 ^ The lower lip of this great “caldera” must 
have been the sonthei'ii, near which the main crater of the present day stands, and 
this in turn bears on its north-east side another smaller parasitic vent (that of 1812). 
Pelee had at its apex a single small crater surrounded by a serrate range of cliffs some 
200 feet high or less, and in the bowl-shaped depression lay a little lake, the Lac des 
Palmistes, 150 metres in circumference. But on Pelee were nmnerons lateral orifices, 
the parent sources of “ soufrieres ” and hot springs, one of which, at Ajoupa Bouillon, 
has since attained notoriety as the focus of a minor eruption on September 3rd, 1902. 
Another laj" near the goi'ge of the Biviere Seche to the south of the summit ; it emitted 
steam and sulphuretted liydrogen, and around it the I'ocks were decomposed by acids 
and often crusted over with sulphur. This wos surrounded l)y higli cliffs, and was 
very rarelj^ visited, but from near the top of the moimtain a view might sometimes be 
oljtained of its interioi'. It lay apparently on one side of the canon, and trom the 
descriptions (jf it which we have received must have closely resembled many of the 
soufrieres of Dominica. 
This Soufriere has now assumed a new impoi'tance, for, according to some accounts, 
it w'as from it that lose the cloud of suffocating gases and red-hot dust which laid 
the fair city on the shore below in ruin and ashes. When we were at St. Pierre 
great toweling balloon-shaped steam clouds would freqneiitl}^ ascend from the neigh- 
* Tejipest x4ndekson and John S. Flett, “Preliminary Report on the Recent Eruirtion of the 
Soufriere in St. Vincent,” ‘Proc. Roy. Soc.,’ vol. Ixx., p. 11, 1902. 
t Mr. R. C. Hill suggests (‘ National Geographic Magazine,’ vol. xiii., p. 233) that the upper part 
of the mountain was blown away in the eruption of 1718. Mr. Anderson’s description of the crater 
as it was in 1784 disproves this. 
