315 
THE ERUPTIONS OF MAY, 1902, AT THE SQUEEILRE IN ST. VINCENT. 
St. Vincent, as since 1812 the Soufriere crater was occupied by a lake of water and 
the escape of steam did not take place on a large scale. In Dominica and St. Lucia 
there are many “ Soufrieres ” discharging sulphurous gases. The crater of Pelee 
contained a “ Soufriere ” of this kind, and Professor Lacroix has given a full account 
of the products of fumarole action in Martinique. 
It is rather a remarkable fact that of all the specimens which we collected of older 
andesitic rocks ejected during the 1902 eruptions only one shows contact alteration. 
The fresher specimens could not be expected to suffer much change by being again 
raised to the temperature at which they consolidated. With the weathered rocks it 
is different, for their secondary minerals would certainly be modified had they been 
enveloped in the new incandescent lava. Hence we may infer that the ejected blocks 
above described were mainly the materials which plugged the orifice of the volcano 
and were cleared out by the rise of the magma. 
The specimen which does show contact alteration is a fine grained, dark coloured 
and banded rock. It consists essentially of dark brown biotite in small scales, 
rounded or nearly idiomorphic when enclosed in the felspars ; hypersthene in minute 
pleochroic grains; a little augite of similar habit, iron oxides and granular, usually 
untwinned felspar. The latter is andesine and labradorite; the whole rock has 
typical hornfels or “ pilaster” structure and is indistinctly spotted. Its composition 
shows that it is igneous, and there can be little doubt that it was originally a fine 
banded andesitic tuff. In one part the specimen shows larger phenocrysts of 
bytownite-labradorite felspar, with aggregates of hypersthene and augite (chondritic 
groups) replacing porphyritic pyroxene. These lie in a fine matrix rich in biotite like 
that above described. This portion of the specimen seems to be a fragment of 
andesite in the tuff. The rock is crossed by little veins filled with recrystallised 
hypersthene. It presents many points of resemblance to the contact-altered andesites 
of the Cheviots and of Lome. 
Quartz Andesites. 
In the ash beds of the 1902 eruptions there are a limited number of andesitic rocks 
which present some remarkable characters which seem to indicate that they have 
undergone alteration of an unusual type. These rocks are milky white or pale grey, 
often with dark patches which represent the ferro-magnesian minerals. Many of 
them are brecciform; in fact, they look like whitened or bleached tuffs. It is 
difficult to prepare microscopic sections of them, as they fall to pieces when being 
mounted. 
Under the microscope they prove to be full of quartz which occurs only in the 
matrix and not as phenocrysts (Plate 26, fig. 5). The original basic felspars of the 
andesites remain, though small grains of quartz may form along their margins. These 
felspars often show intense zoning and have much the same characters as those of 
the andesites and hypersthene basalts. The primary olivine, augite, hypersthene, 
2 s 2 
