RECENT VOLCANIC ERUPTIONS IN THE WEST INDIES. 
275 
two unbiassed observers, who had seen it and that of May, declared this 
was the larger of the two. 
Our limited time was now coming to an end, but on leaving for 
Dominica two days later we were able, from the deck of a steamer, to 
make some examination of the slopes of the mountain down which we 
had seen the incandescent avalanche descend. The whole district from 
just beyond St. Pierre to near Precheur, a distance of about 4 miles, 
was covered with a deposit of light grey ash of varying thickness, 
perhaps averaging a few inches, but evidently much deeper in the 
valleys of the Riviere Blanche and Riviere Seche, which descend from the 
mountain about 2 miles beyond St. Pierre, and drain the slopes below 
the large fissure out of which we saw the eruption descend. The water 
of these rivers was boiling as it fell into the sea—in fact, it was repro¬ 
ducing on a small scale the phenomena of boiling mud which are 
described above in the cases of the Wallibu and Rabaka rivers in St. 
Vincent, though how far up the mountain these Wallibu effects extend, 
and where they give place to true volcanic discharges, it is difficult to 
describe as yet; we must wait further observations by M. Lacroix and 
his colleagues. 
Returning now to the mechanism of the hot blast and the source of 
the power which propelled it, both my colleague and I are convinced 
of the inadequacy of previous explanations, such as electricity, vortices, 
or explosions in passages pointing laterally and downwards, or explo¬ 
sions confined and directed down by the weight of the air above. Such 
passages into the mountain, which, to be effective, would require to be 
caverns closed above, and not mere open ravines, do not exist in the 
case of the Soufriere, and we are not aware that they have been ob¬ 
served in Mont Pelee ; and as to the weight of the air, this did not 
prevent the explosions in the pipe of the Soufriere from projecting 
sand and ashes right through the whole thickness of the trade-winds 
till they were caught by the anti-trade current above and carried to 
Barbados. Moreover, the black cloud, as we saw it emerge from Mont 
Pelee, seemed to balance itself at the top of the mountain, start slowly 
to descend, and gather speed in its course, and the second incandescent 
discharge tollowed the same rule. We believe that the motive power 
lor the descent w r as gravity, as in the case of any ordinary avalanche. 
The accepted mechanism of a volcanic eruption is that a molten 
magma rises in the volcano chimney. It consists of fusible silicates 
and other more or less refractory minerals, sometimes already partly 
crystallized, and the whole highly charged with water and gases, which 
are kept absorbed in the liquid, partly by the immense pressure to 
which they are subjected. When the mass rises nearer the surface and 
the pressure is diminished, the water and gases expand into vapour and 
blow a certain portion of the more or less solidified materials to powder, 
or, short of this, form pumice stone, which is really solidified froth 
