360 
DRS. T. ANDERSON AND J. S. FLETT ON THE ERUPTIONS OF THE 
island to the sea, so that a comparatively small submergence would convert it into a 
strait.”^ But a better instance of this is the broad open valley on the south side of 
the Soufriere, between that mountain and the Morne Garu. It is a sigrnificant fact 
that there is no good road from the leeward to the windward side over the mountains, 
and that the only passable footpath was an old Carib track which started from 
Chateaubelair and went to Georgetown. It led right up to the edge of the crater of 
the Soufriere before descending on the other side, an ascent of 3000 feet. It stuck to the 
summits of the spurs between the valleys, where the vegetation was least dense, and 
there was least probability of a wash-out due to the sudden rains and floods. At flrst 
glance it would have appeared easier to follow the course of the valley across the 
island, hut this would have passed through dense forests, and the lateral streamlets 
descending the side of the valley to join the rivers would have been difficult to cross 
in time of flood, and would have caused much damage to any road carried over them. 
In such a situation a road would also have been exposed to landslides and falls of 
rock, which would have often rendered it impassable. 
In consequence of the small size of the island, none of the rivers of St. Vincent are 
of any magnitude. Their courses are remarkably short and straight, being mostly 
nearly at right angles to the coast-line. The drainage is of a very simple “ conse¬ 
quent ” type, and intimately connected with the geological structure of the island 
The most striking feature of the stream-valleys is their great depth, especially in 
their upper parts, where they present in any of the characteristics of canons. This is 
best seen in the devastated country, where the surfaces are no\v quite hare and have 
a desert aspect.! The casual streamlets which descend the valley walls after sudden 
rains form niches which furrow the surface, and the alternation of vertical cliffs of 
lava with sloping taluses of weathered ash produces features which strongly recall 
those of the canons of western North America. As the streams are still cuttino- 
rapidly along most of their length, it is rare to find any extensive plains of alluvial 
deposit except near their mouths. The sea is too deep for the formation of deltas, 
ljut in many of the bays there is a small stretch of flat ground with nsuafly a curved 
storm beach facing the sea. Higher up the valleys narrow flood plains may be found, 
but they are few; and sugar cultivation, wliicli is best carried on on level ground, is 
almost entirely confined to the coast. 
The whole of St. Vincent is of volcanic origin. There are no marine sedimentaries, 
and no organic limestones. All the available evidence goes to show that the island is 
the product of prolonged volcanic activity, and that the materials accumulated on a 
land surface, without any intervals of depression during which marine sediments were 
formed. By far the most common type of rock is a coarse volcanic tuff or agglome- 
* J. A. Spencer, “The Windward Islands of the West Indies,” ‘Trans. Canadian Inst.,’ vol. 7, 
p. SGI. 
t Tempest Anderson and John S. Flett, “ Preliminary Report on the recent Eruption of the 
Soufriere in St. Yincent,” ‘ Proc. Rojn Soc.,’ vol. 70, Plate 2, 1902. 
