7 6 



The National Geographic Magazine 



channels through the fresh layer and 

 exposed the cultivated soil beneath. 

 The steep-sided trenches, with well- 

 marked terraces, to which many tribu- 

 tary rill channels converged, were, in 

 miniature, canyons like those of the 

 Colorado region. Some of these in- 

 structive details may be recognized in 

 the accompanying photographs. 



Among the larger features of the re- 

 cent changes that especially attracted 

 attention was the manner in which the 

 steep-sided valley of Wallibu River, per- 

 haps a quarter of a mile broad, had been 

 deeply filled with fresh, hot debris, and 

 the way the displaced stream was en- 

 deavoring to regain its right of way. 

 The valley had been filled, as estimated 

 by Mr McDonald, to a depth of 50 or 60 

 feet with freshly fallen debris. Through 

 this material, surface water was work- 

 ing its way, and, meeting the still hot 

 stones and dirt, was being changed to 

 steam, which, escaping from thousands 

 of vents, formed white columns that 

 rose at times hundreds of feet into the 

 air. This wonderful display of steam 

 jets and geyser-like eruptions, varied in 

 grandeur with the amount of surface 

 water present. During the intervals 

 between the occasional heavy down- 

 pour of rain, the energy of the escap- 

 ing steam would decrease, and a person 

 could walk in safety over a miniature 

 crater, from which steam had previously 

 been seen to rush out as from the escape 

 valve of a steamship, but with ten or a 

 hundred times its volume. During a 

 heavy shower, however, as happened at 

 the close of my second excursion to the 

 Wallibu region, the volume of steam be- 

 came so great that the entire landscape 

 was obscured, and the upward rolling 

 clouds ascended for thousands of feet. 

 On such occasions the roar of the escap- 

 ing steam could be heard a mile or more. 

 The steam jets, at times, had such energy 

 that black columns, consisting of what 

 may be termed mud, were shot upward 

 like geysers, to a height of fully a hun- 



dred feet, and would play for several 

 minutes. These miniature eruptions 

 have been referred to, in several news- 

 paper accounts of the strange scenes on 

 St Vincent, as volcanic eruptions from 

 newly formed craters ; but this is a mis- 

 take, as they were clearly due to the sur- 

 face waters working their wa^^ through 

 thick beds of hot dust and stones. A 

 similar phenomenon was witnessed by 

 me near St Pierre, and was seen again 

 near Georgetown, and in each instance 

 the cause was the same. 



Wallibu River, as I have stated, was 

 displaced from its former bed b}' the 

 vast quantity of debris precipitated into 

 its channel or washed from the border- 

 ing uplands. At the time of my visit 

 the stream was behaving in a most pe- 

 culiar and interesting manner. Not 

 only was it a stream of hot w y ater from 

 which steam was being given off in large 

 volume, but, owing to the vast quantity 

 of loose material present, was overloaded. 

 The debris checked its flow, and for a 

 time would hold back the water and act 

 as a dam, the stream bed downstream 

 becoming dry, and as the pressure of 

 water increased, the dam would give 

 way, and a large body of steaming water, 

 black with material in suspension, would 

 rush down the previously dry channel, 

 and with a roar plunge into the sea. 

 The stream made these pulsations at in- 

 tervals, on an average, of perhaps twenty 

 seconds, and between each swift rush of 

 black, seething water its channel was 

 vacant. A similar behavior of the 

 stream near St Pierre, and also of those 

 on St Vincent which reach the sea near 

 Georgetown, was observed. Such ex- 

 amples of what niay justly be termed 

 overloaded and pulsating streams are 

 certainly novel to students of the life 

 histories of rivers. 



From the heights above Richmond 

 House the entire western slope of La 

 Soufriere was in full view during our 

 visit, and, like the corresponding side 

 of Mont Pelee, was without life. Not 



