4.34 
DES. T. AXDEESOX AND J. S. FLETT ON THE EEUPTIONS OF THE 
bends. The loose sand, undermined in this way, would tumble down, and at 
the contact of the hot ash witli the water a great column of steam would shoot 
into the air. Very little noise attended these outbursts; as a rule, only a low, faint 
rumble. In dry weather the river of steaming mud flowed quietly along its channel, 
effecting little erosion, and few explosions were to be seen. But when it came down 
in flood, the fine ash was ploughed out like snow, and enormous outbursts followed 
one another in rapid succession. 
It is not desirable to be caught in a tropical downjoour of rain on the Soufriere, as 
the streams rise so rapidly that they cannot be crossed, the more so as they are filled 
with boiling mud. But on one occasion, when we were descending from the crater, a 
short but heavy shower overtook us. The Rozeau Dry River lies to the west of the 
trail. It had been dry when we were going up ; as we came down it was filled vdth 
a thundering inky torrent, 30 feet broad, roaring along at 20 miles an hour. The 
new ash had already been very thoroughly cleaned out of the uj^per portion of the 
valley, but down below there were still remains of the hot sand deposit, and as the 
torrent spread down the stream, it carved into the banks of ash, which seemed to 
melt away before its attack. As the water touched the hot sand, pillars of black mud 
capped with a great club-shaped cloud of steam towered into the air. It was a 
marvellous and beautiful spectacle, the column of water curving outward at its apex, 
and dropj^ing down like a fountain, while over it 23layed the delicate feathery steam 
cloud, the height of which we estimated at 700 to 800 feet. It might well be com¬ 
pared to a geyser of boiling mud, and the contrast between the black base and the 
pearly apex heightened its beauty. 
The Carib Country and the Valley of the Rahaka Dry River. 
On the windward side of the island, in the CArib Country, and the ravine of the 
Rahaka Dry Ptiver, the 2 )henomena already described in the valley of the AVallibu 
and Wallibu Dry River, on the leeward side, am repeated in all their essential 
features, though with a few not unimportant differences. 'The topographical condi¬ 
tions are here considerably simpler than on the western side, as in the Carib Country 
we have a broad triangular plain, the base of which rests on the sea shore, and 
extends from south of Georgetown to Overland Village, while the apex is placed some 
distance afjove Lot 14 estate (see Plate 2l, fig. 1). It slopes gently to the sea 
coast, and though its surface shows tiaces of old eroded terraces, masked with an 
accumulation of loose surface deposits, it is crossed by no prominent ridges, but forms 
a smooth expanse, which is entirely under cultivation, and I)efore the eruption was 
considered the best agricultural district in the whole island. Across this sloping 
plain several streams flow in shallow valleys, the most important being the Rahaka 
Diy River. 
The conditions already described as prevailing at Georgetown are typical of the 
