282 DR. TEMPEST ANDERSON ON THE ERUPTIONS OF THE SOUFRIERE IN 
Further slides are reported to have taken place at the times of the eruptions in 
September, 1902, and March, 1903, but this is not very certain. The cliffs in this 
part are higher than further south, but present similar structural features, viz. :—tuff 
capped with new ash (Plate 14). A new beach is forming along the base, the materials 
of which appear to be furnished partly from the cliffs, partly by that brought down 
by the rivers. The gorges cut through this plateau are often very narrow, and their 
steep sides show fine sections of the tuffs and ash beds.* 
The Soufriere. 
The Upper Slopes .—The upper slopes of the mountain are chiefly formed of beds 
of tuff like the lower parts, but contain perhaps a larger proportion of ejected blocks 
which naturally fell in greater abundance nearer the crater. They are much cut up 
into deep ravines which are separated by the ridges, the slopes of which are often 
very steep, as described in Part I.t Even in 1902 the new ash, which had never 
been thick on these slopes, was in great part washed away, and it was only on the 
ridges and in some of the valley bottoms that any considerable amount of that ash 
remained, and this process, of course, has gone on ever since. 
The ash on the ridges still remains, and its surface is consolidated into a thin 
but hard crust similar to that on the plateaux. This extends generally only to a 
width of a few feet, and often not more than one or two. On each side of the ridge 
where the ash has been washed away the old soil has been exposed, and it would 
doubtless also have been in a great measure removed if it had not been held together 
by the roots of plants which, as mentioned below, were in many cases not killed. 
Even when dead they no doubt held the soil together to a large extent while the 
new vegetation has been re-establishing itself. 
The old Carib track to the summit ran along one of these ridges (Plate 22, fig. 2), 
and the consolidated ash forms in most parts an excellent footpath along the former 
lines. It is true that in places the ridge has been carried away by landslides, but a * 
comparatively small amount of labour would suffice to restore it to a perfectly useful 
condition. The path for a considerable distance overlooks the upper part of the 
Ptozeau valley, which extends nearly to the summit of the mountain, and this is a 
good example of a high valley in which extensive ash deposits were formed. It was 
here that Mr. T. M. Macdonald saw explosions in the earlier stages of the eruption, 
and these were supposed to proceed from parasitic craters, i.e., side branches of the 
main chimney. The place was pointed out to me by his brother in 1907. It was 
impossible to approach it closely except at unjustifiable risk, but an examination from 
a distance of perhaps 100 yards revealed nothing but a hollow in a bed of ashes, which 
no doubt was merely the locality of a secondary explosion in the hot ash like those 
* Plate 12, fig. 2. In lower Trespe valley, 
t Part I., Plate 31, fig. 2, and Plate 35. 
