300 DR. TEMPEST ANDERSON ON THE ERUPTIONS OF THE SOUFRIERE IN 
to increase the resemblances and minimise the differences. Thus in March, 1902, 
there was in the latter valley a number of fumaroles, some of which were active 
enough to have led to the supposition of their being parasitic craters. Further 
examination has satisfied Professor Lacroix of their superficial nature, i.e., their 
origin in the deposits of hot ash, and their consequent similarity to those in the 
Wallibu. # They are all now cool and extinct except one group, which I had the 
pleasure of visiting in company with M. Guinoiseau, Adjutant in charge of the 
Observatory of Morne des Cadets. They are situated at a height of about 1350 feet 
above the sea, on the low ridge between the Rivieres Blanche and Claire, which here 
How down one broad valley apparently much in the same way as two rivers often 
occupy each one side of a big old valley in St. Vincent. The ridge itself is perhaps 
50 feet higher than the valley on each side, and broad in proportion. It consists of 
fragmentary ejecta of the 1902 eruption, including a good many large blocks. The 
temperature of the fumaroles had been as high as 300 C. last year, but was 
gradually decreasing. M. Guinoiseau found it to he about 230° C. at the time of 
our visit. This group is the only one about the nature of which any doubt now 
exists, but it appears most probable that its origin, like that of the others, is 
superficial. 
The differences between the two valleys are not by any means so great as the 
resemblances, and they appear to be all traceable to two causes, viz., the repeated, or 
rather, the at one time almost constant, passage of the incandescent avalanches, and 
the fact that owing to the configuration of the crater these all descend through the 
V-shaped gap right down the valley on which they spend their whole force. Thus 
the denudation of the deposits is in a somewhat less advanced stage than in the 
Wallibu, and the bedding somewhat more complicated owing to an alternation of 
water-sorted beds with those of fresh ash. Moreover, apparently owing to the very 
direct course of the valley from the V gap to the sea, and its steep inclination, the 
number of large ejected blocks even in the lower part of the valley is much greater 
than in the corresponding part of the Wallibu. This is only what might have been 
expected when 1 recall the stones which we saw descending by leaps and bounds in 
the incandescent avalanche of July, 1902. They were large enough to be visible at 
a distance of several miles and distinctly a brighter red than the rest of the avalanche 
material. 
In this connection also ought to be mentioned another phenomenon, which though 
not altogether absent in St. Vincent is much more conspicuous in Martinique, viz., 
a scoring and grooving of the rocks of the sides of the Blanche valley. The part 
I specially noticed was a cliff nearly 200 feet high in the ridge between the Blanche 
and Seche Rivers and perhaps half a mile nearer the sea than the active fumaroles 
above mentioned. The valley here is somewhat narrower than higher up, and much 
narrower than lower down, and it is just the part where the avalanches might be 
* Lacroix, p. 400. 
