322 ERUPTION OF THE SOUFRIEKE IN ST. VINCENT. 



water can be witnessed. The river plows its way deeper and deeper, 

 constantl}^ sweeping- the material into the sea. The valleys, at first 

 almost obliterated, are now reassuming their old appearance. Ter- 

 races on their sides give evidence of former levels at which the streams 

 flowed. There are five or six such terraces on the Wallilni. This 

 river flows in gushes of hot steaming black mud, its intermittent flow 

 being due to small landslides temporarily damming up its channel, 

 only to be swept away as the pressure of the water increases. On the 

 Rabaca Dry River ,here has been less erosion, and only after heavy 

 rains does it reach the sea, as the water from the smaller showers is 

 apparently eva})orated in its passage through the banks of hot ashes. 

 After rains Ijoth rivers can be seen steaming all along the lower parts 

 of their courses. 



When one of these streams comes down in force it undermines its 

 banks by washing out the soft new ashes at their base. Then land- 

 slides take place, and a curious spectacle results. When the hot ash 

 tumliles down into the water an immense cloud of steam rises in the 

 air to heights of hundreds of feet. It expands in great glol)ular 

 masses exactly like the steam explosions from a crater, and as it drifts 

 away before the wind fine dust rains from the cloud. We had the 

 good fortune to witness a magnificent series of these explosions one 

 day as we were descending from the summit of the Soufriere. It was 

 in the valley of the Rozeau Dry River. After every landslip a column 

 of nuiddy water rose to about 200 feet, carrying Avith it pieces of stone. 

 Immense, quantities of steam shot up to 700 or SOO feet in the air. It 

 resembled an enormous geyser of l)lack mud and steam. In the Wal- 

 libu River after every shower these steam explosions may be wit- 

 nessed taking place on a large scale. After a few hours of dry 

 weather they cease, though the river can still be seen to be steaming 

 strongly as it flows along. 



The structural modifications produced ui)on the hill by this ci'iiption 

 have been astonishingly slight. W^e saw no fissures, no parasitic 

 craters or cones, and no lava streams. Even the craters at th(> sununit 

 retain essentially their old configuration. All the evidence points to 

 the supposition that it was from the large or old crater that this 

 eruption for the most part proceeded. But the snialU>r cratiM- has not 

 disappeared, nor has it been filled up. We did not see it. hut we can 

 rely on the evidence of several observers who knew it well before the 

 eruption and have seen it since. The narrow ridge between it and 

 the large crater still stands, though probably somewiiat lower than 

 before, and possibly is slipping down in landslides on l)()tli sides. 



Like all the higher mountains of the Windward Islands, the Soufriere 

 has usually its sununit capped with cloud, especially during th(> rainy 

 season, and this was the case on both the daj^s on which we made the 

 ascent. On the flrst occasion the mist lifted for a few miiuites and 



