314 
DR. JOHN S. FLETT: PETROGRAPHICAL NOTES ON THE PRODUCTS OF 
have arisen in several different ways. On the Soufriere they are rare, and as they 
occur only in scattered blocks among the ashes, there is nothing to indicate in what 
manner they have originated. 
The alterations which the ejected blocks of former lavas have experienced are due 
to weathering, to fumarole action and propylitisation, and to contact action. 
The weathered blocks show changes of a familiar type.- Their olivine is replaced 
hy serpentine and carbonates; their hypersthene by bastite ; the felspars become 
cloudy with kaolin and carbonates. In the groundmass the glass is devitrified and 
stained with limonite and chlorite. 
In some cases the groundmass has been replaced extensively by silica. This may 
be brownish chalcedony showing spherulitic structure, almost the same in appearance 
as a spherulitic glass, but harder than steel when tested with a knife. The fibres 
of the spherulites have positive elongation. More common is quartz in small irregular 
patches replacing the glassy base and enclosing microlites of felspar. This gives these 
rocks a secondary micropoikilitic structure. 
In small cavities in many of these ejected blocks, scales of tridymite are found, and 
Professor Lacroix lias observed cristobalite also in some of them.* It has been 
pointed out by WeinscHenk f that tridymite is a fumarole deposit, and indicates that 
the rocks have been subjected for some time, in the solid condition, to the action of 
steam at a high temperature. In St. Vincent, as in Martinique, it does not occur in 
the new bombs of the first eruptions of 1902. Professor Lacroix]; has made some 
very interesting observations on its development in the ejecta of Pelee. It began to 
appear in the vesicular andesites during the winter 1902-03, that is to say, six or 
eight months after the eruptions had begun. In the materials cast out in 1904, after 
the dome had stood for some time, tridymite was abundant. It appeared in the 
enclosed blocks before it was found in the new lava itself. 
Pocks of propylite type are also found among the ejected blocks. They are grey 
or greenish grey masses in which the felspars are often decomposed along certain 
zones or at their centres, but elsewhere fresh. In these rocks pyrite is common and 
indicates the operation of sulphurous gases from fumaroles. Chlorite and epidote 
replace the augite, the hypersthene often yields a pale coloured fibrous mineral with 
strong double refraction, perhaps iddingsite or a secondary mica. Quartz occurs in 
some of these rocks; carbonates are scarce. The igneous structures may be perfectly 
retained, even when the original minerals have entirely disappeared; the whole mass 
between crossed nicols is nearly isotropic, and apparently consists mainly of kaolin 
and limonite, but the outlines of all the porphyritic crystals are clearly visible in 
ordinary light. 
The ultimate stages of fumarole decomposition are not well illustrated in 
* ‘La Montague Pelee,’ p. 593, 1904. 
t ‘Die gesteinsbiklenden Mineralien,’ Edit. I., p. 76, 1901. 
1 Ibid., p. 519. 
