SOUFPJEEE, AND ON A VISIT TO MONTAGNE PELFE, IN 1902. 
477 
especially the absence of the premonitory symptoms which invariably at the Soufriere 
have preceded any resumption of volcanic action, and as the crater had then been 
quiescent for eighteen months (since June, 1812), it is unlikely that any violent 
eruption should have broken out without giving warning beforehand. It is a 
rule, to which we know of no exceptions, that the longer the period of quiescence at 
the Soufriere, the more marked are the disturbances and earthquakes which are 
observed before an eruption. The outburst of May 18th this year gave apparently 
no warning, though of this we cannot be. quite certain, as it took place after dark ; 
before all the others, earthquakes have been numerous, violent, and continued. 
It is said that a steam column shot upwards from the crater and, reaching the 
higher currents of the air, floated away on the anti-trades. It spread out to form a 
cloud which, as the emission was brief, was soon separated from its parent stem and, 
having become detached, was Ijorne away to the horizon. There Is no description of 
any rain of ashes, but a few days afterwards large stones were found lying around 
the old crater, some of them several hundred yards from the rim. As by that time 
a covering of green vegetation had probably formed in some part of the upper sloj^es, 
which had been devastated two years before—-though Mr. Anderson, in 1784, 
describes them as quite barren and stony—and as the oljserver was resident near and 
well acquainted with the volcano, there is not much chance of his having been in 
error on this point. Their distribution led to the iDelief that it was from the old 
crater the steam arose, which is quite likely, as undoubtedly that crater had taken 
the chief part in the eruption of 1812, and had discharged the great masses of sand 
which blocked the Wallibu and Rabaka Valleys, and were accompanied by the loud 
detonations that awakened alarm in Barbados. 
This description of the steam cloud, “ part of which rolled down the side of the 
mountain,” is curiously reminiscent of the behaviour of the great black cloud. But 
it is in no way probable that this phenomenon appeared on this occasion. It has 
never been known, except in the major eruptions,, and had an avalanche of dust rolled 
down the hillside it would have left traces visible wlien tlie mountain was again 
visited three days afterwards, and too obvious to have been overlooked. Moreover, 
we are told that the water was still left in tlie crater, and little change was to be 
perceived in the aspect of the interior. Now the great black cloud never appears till 
after the crater lakes are emptied, and the ascending column of lava has forced its 
way to the surface. 
In all probability this was merely an exacerbation ol solfataric activity, such as has 
been experienced by more than one of the Soufrieres of the Windward Islands, and 
after the emission of a great mass of vapour and sulphurous gases the short-lived 
eruption came to an end. Such an emission would carry with it many stones from 
the bottom of the crater lake, would blast any vegetation growing within the crater, 
produce minor changes in some part of the crater walls, and project into the air a 
column of mud and dense muddy vapours which would fall partly outside the lip and 
