400 
DKS. T. ANDERSON AND J. S. FLETT ON THE EliUPTlONS OF THE 
Their maximum activity does not coincide closely with the outburst of steam clouds 
from the crater. 
When these fumaroles lie in the course of a stream or in the sea, the water flows 
into the open tube in periods of quiescence; when, on the other hand, the activity 
resumes, it is discharged as an overflow of hot mud, alternating with steam puffs and 
jets of muddy water. To this is due the sudden variatioiis in the volume and the 
temperature of the streams. As in St. Vincent, the banks of hot ash overlooking the 
current are subject to frequent landslide, which give rise to ascending steam clouds 
when they meet the water; and this must not be confounded with the open 
fumaroles from which steam is also emitted. Professor Lacroix and his colleagues 
also remark that the large number of dead fishes cast ashore on some days in the Bay 
of St. Pieri-e may have been killed by the action of similar fumaroles beneath the sea, 
along a prolongation of the same fissures, and that the telegraph cable was 
broken apparently where it crossed this line, and when recovered its insulating 
material was found to have been melted. 
On the north-western side of the mountain also, mud-flows have taken place on a 
large scale. One lias covered the village of Basse Pointe, another flows down to 
Macouba. Even Ijefore the great eruption there were mud currents in some of the 
streams, and thougli tliey niay be in part due to the washing action of rains on the ash 
covered soil, and consequently similar to the muddy rivers of St. Vincent, yet they 
cannot all be explained in this way. Some are due to the discharge of the crater 
lakes through fissures, others no doidjt to the re-emergence of Avater Avhich has been 
engulfed into fumaroles and open cracks, Imt for a fuller discussion and explanation 
of their origin and nature we must Avait till the detailed reports of the French 
Commissioners ai'e to hand. 
The long preliminary stage of the eruption at Pelee, the scarcity of earthquakes, 
the deA^elopment of fissures and of fumaroles, and the extensiA’e 11oaa"s of mud, 
together AAuth the more acid nature and more uniform fineness of the ejecta, and the 
greater strength of the blast in proportion to the small amount of dust deposited, 
are tlie main differences Avhich Ave find in making a comparison of the great 
eruptions of 8t. Vincent and Martinique. In all their principal features the tAVO 
outbursts are parallel, and belong to a clearly-defined and highly-destructiA^e type of 
A'olcanic action. These A'olcanoes are of the explosive class, and in their dust 
avalanches exhibit one of tlie most I'eniarkable efiects of the expansiA-e poAver of the 
superheated steam in an igneous magma. Their fatal Auolence is oAving rather to 
the physical pi'ocesses at Avork and to the form their ejecta assume than to the 
magnitude of the eruptions themseL'es, for Avhile that of the Soufriere Avas really 
a considerable eruption, and produced notable geological consequences, that of 
Pelee Avas comparatively small, and its geological effects are of no great impor¬ 
tance, and most of them Avill disappear Avithin a feAv months of the cessation of the 
explosions. 
