RECENT VOLCANIC ERUPTIONS IN THE WES^T INDIES. 
267 
its tributaries, which drain into the sea on the windward coast north of 
George Town. The western, which is rather more extensive, is much 
eroded into deep valleys and ravines, often with almost precipitous 
sides, in which flow the rivers and torrents, which are often dry, and 
go by the names of Wallibu, Wallibu Dry river, Bozeau, Morne Bonde, 
and Larikai. The rocks of which this part of the mountain is com¬ 
posed are almost entirely tuff's and agglomerates formed of fragmentary 
material, similar to that ejected from the volcanoes during the recent 
eruption. We saw a few beds of lava, but they were chiefly exposed 
in the deeper parts of ravines, and were obviously of much older date 
than the more superficial portions. This great transverse valley, and 
especially its western portion, the valleys of the Wallibu, received the 
greater part of the products of the eruption, amongst which we saw no 
lava, and do not believe any was erupted. A considerable number of 
ejected blocks* were noticed, but the great bulk of the material consisted 
of fine sand and ashes—in fact, lava blown to pieces by the sudden ex¬ 
pansion of its imprisoned gases. The trade-wind blows steadily from the 
east or north-east, and a certain amount of the finer particles would be 
caught by the wind and deposited in the Wallibu valley. A considerable 
amount was also driven directly upwards so violently as to go through 
the whole thickness of the trade-wind and get caught by an upper 
current in the reverse direction, by which it was carried to the east as 
far as Barbados and the surrounding sea. A certain quantity fell on 
the north slopes of the volcano beyond the Somma ridge and along the 
east coast as far as to beyond George Town, but the greatest deposit 
which we saw, and in comparison with which all the others were trivial, 
was in the Babaka and Wallibu valleys. 
The tropical rains have cut deeply into the soft strata which form 
the cone of the Soufriere and the slopes at its foot, and have produced a 
series of branching valleys with steep or almost precipitous sides, and 
separated by the narrowest of ridges—in fact, often mere knife-edges. 
The whole, before the eruption, was clothed with the most luxuriant 
tropical vegetation. Into this series of valleys was discharged from 
the Soufriere a black cloud, so heavily laden with incandescent dust 
that it might most fitly be described as an incandescent avalanche. 
The mechanism of its production is discussed below, but its immediate 
geographical effect was to fill the Babaka and Wallibu valleys and some 
of their tributaries to a depth in some places as great as 200 feet. The 
motion of the mass was sufficient to prevent any large amount of hot 
sand being deposited on the ridges, while the hollows received the 
greater part, and the whole was smoothed over by the blast, and lay 
in rolling masses like drifted snow. Thus in the course of a single 
*The nature of these will form the subject of a special report to the Royal Society 
by my colleague, Dr. Flett. 
