SOUFRIERE, AND ON A VISIT TO MONTAGNE PELEE, IN 1902. 
441 
study of the photographs will give a better idea of the features ot these rills than 
pages of description. 
In many places the sides of the valleys were absolutely bare, and not only had the 
new ash been removed, but with it much of the old soil had also disappeared, 
as might be seen from the manner in which the roots of the broken trees projected 
above the surface, and in the lakes of water previously mentioned as forming in the 
lateral valleys of the Habaka Dry River, a red mud from the eroded soil Avas mixed 
with the black mud furnished by the new Amlcanic ash. Often we could see areas of 
many hundred square yards absolutely divested of the covering of black mud, and 
sometimes the whole side of a valley showed it only in one or two patches (see 
Plate 34, fig. 2). 
Fine sand with little lapilli had formed the greater part of the material which lay 
upon the slopes. Stones of any size were very few, for AAdren they fell they could not 
rest unless they buried themselves in the hot sand, but rolled down into the ravines. 
This was, in fact, one of the dangers of the ascent, as the rains had loosened many 
stones in the old agglomerates, and when once they were set in motion nothing could 
stop them till they landed in the stream hundreds of feet below. The ash, as a rule, 
held together A^ery Avell when it was AA^et, hut every noAv and then the pressure of the 
foot started small landslips, AAdiich tumbled doAvn the slopes. 
In the deep narrow valleys Avhich score this side of the mountain little of the neAV 
ash w’as left (see Plate 30, fig. 2). At the higher elevations a patch here and there, 
on the extremities of the bends or behind a projecting bank, was all that could be seen. 
Further down, flows of the black liquid mud had congealed in the channels, unable to 
traA'ol further. Once a sharp shoAver, brief but heavy, OA^ertook us as Ave were 
descending. Below us was a place, “ the rRer bed,” where the path crossed a tributary 
of the Rabaka, dry except after rains. As we did not Avish to be cut off by a raging 
torrent, we hurried down to get OA^er the crossing. The stream was pouring in cascades 
OA^er the rocks in the upper part of its course, but when A^^e reached the “ river bed ” 
there was no water there. Looking up the stream we saAv a creeping mass of stiff 
black mud slowly winding its way down the channel like a serpent. The water was 
so loaded with sand and mud that it almost ceased to floAv. Had the rain continued, 
of course an inky torrent Avould have rushed doAvn into the lake of mud in the lower 
end of the valley. Further down, where the ravines entered the plains of the Carib 
Country, and on the flat ground near Wallibu, some part of the masses of ash originally 
piled up there still remained (see Plate 31, fig. 2). 
We have mentioned the explosions of hot ash and water in the lower part of the 
Rozeau Dry River. The remains of the deposit were terraced and deeply cut into, 
a mere shadow of their former selves. So great had been the erosion along the Avhole 
course of these steep valleys, that it would be rash to say to what extent they had been 
encumbered Avith hot sand unmediately after the great eruption. In more than one 
place there were still 40 feet of ashes in the bottom of the gorges (see Plate 30, fig. 2). 
A"OL. CC.—A. 3 L 
