270 
RECENT VOLCANIC ERUPTIONS IN THE WEST INDIES. 
soon bo renewed, and its contained lake has been discharged, but it is 
already beginning to fill again. Any one who knew it before and 
visited it now would notice a considerable change, but, if he deferred 
his visit for a few years, would probably see no marked difference. 
Another curious little secondary result deserves notice. Water will 
hold more mud in suspension when it is flowing down a steep slope. 
The water in the steeper upper parts of the valleys was charged with 
mud to the utmost, but where it descended on to more gentle slopes, and 
consequently moved more slowly, it could not carry so much, and 
deposited part, especially where it moved slowly at the side of the 
stream. We saw two places where dams had thus been formed across 
the mouths of small lateral valleys, and small lakes or large ponds had 
been produced. As the dams were only soft mud, these may only be 
temporary; but I have seen a permanent lake of several acres formed in 
this way in Iceland, by a bank of shingle brought down by a rapid 
glacial river. Dr. Plett thinks, and I agree that the explanation is 
feasible, that these lateral dams are the remains of the avalanche which 
filled the valley during the eruption, and that the centre part only has 
since been washed away. 
A watercourse formerly existed which supplied all the plantations in 
the Carib country on the east or windward slope of the mountain with 
water taken from the Babaka river high up. The river has now 
changed its course, and no water enters the conduit, which, moreover, 
in places is blocked up with ashes. This, however, can doubtless be 
remedied, but, as far as it goes, is a geographical change. The planta¬ 
tion buildings were not of sufficient size for their destruction to be of 
geographical importance, and the black population will, according to all 
previous experience, return after the cessation of the eruptions. 
On the whole, the permanent geographical changes in St. A incent 
are comparatively small. It remains to discuss the mechanism of the 
eruption ; but this will be better considered along with that of Mont 
Pelee. 
Turning now to the consideration of Martinique, I may remark 
that our instructions were to proceed first to St. k incent and devote 
our chief attention to that island. This we did, but later on we went 
also to Martinique, for the purpose of making such an examination as 
would enable us to compare the phenomena of the two volcanoes. 
There is remarkable similarity between the islands of St. Vincent 
and Martinique. Both are roughly oval in form, with the long axis 
almost north and south. The north-west portion of each is occupied by 
a volcano, the Soufriere and Mont Pelee, more strictly called Montague 
Pelee, which have many points in common. Both volcanoes show a 
single or practically single vent, a remarkable absence of parasitic 
cones, and a scarcity of dykes. In both a transverse valley exists to 
the south of the volcanoes, and the main discharge of ejecta during 
