308 
DR. JOHN S. FLETT: PETROGRAPHICAL NOTES ON THE PRODUCTS OF 
The pleochroism is strong and is of the usual character. The optic axial angle 
(negative) is large (2E over 120 degrees, 2V over 60 degrees). Glass enclosures and 
grains of magnetite are common in the pyroxenes. Olivine occurs in small grains; 
perfect crystals are not found, but only splinters which often show a conchoidal 
fracture. It is colourless or pale brown (perhaps from oxidation in presence of 
hydrochloric acid gas and air), and its presence is easily confirmed by treating the 
powder with strong cold acid when it is covered with a film of gelatinous silica, even 
before the basic felspars are attacked. Hornblende, as reported by Falconer and 
Klein, # is found occasionally. So far as we have seen it is always brownish-green 
and occurs only in irregular fragments. Most of the iron oxide is titaniferous 
magnetite in small rounded grains; pyrites, apatite, and possibly zircon are also 
present. 
Non-crystalline or glassy material occurs in all the dusts, though less abundant 
than the crystals. It is partly in minute fragments, partly in thin films surrounding 
the crystals, and partly in the form of small, rounded vesicular lapilli. The steam 
cavities which abound in it are empty. Between crossed nicols some of these lapilli 
contain small crystals evidently belonging to the second generation, showing that 
crystallisation was preceding in this portion of the rock as it ascended in the crater. 
Much of the glass, however, is perfectly vitreous, and this portion was probably liquid 
at the moment of projection. Occasionally the glass has been drawn out into threads 
while still in a viscous state. The lapilli, however, are not highly pumiceous and the 
splinters do not have those arc-shaped, concave outlines which are found in shattered 
pumiceous glasses ( e.g ., those emitted by Ivrakatoa in 1883).t 
In the first part of this report we have stated our belief that the magma of the 
Soufriere, though in large part crystallised, was not entirely solid at the moment 
when it was shattered into dust and the great black cloud emerged from the crater. 
Professor Lacroix^ has found that the later nuees arclentes or Peleean clouds of 
Montagne Pelee were produced by the disruption of a solidified magma. But the 
facts of the eruption of May 7th in St. Vincent, as given by us in the first part of this 
report, are sufficient to prove that the magma was a semiliquid mass as it ascended 
in the crater. The crater lake was at its usual level at 11 a.m. on Tuesday (May 7th); 
at midday on Wednesday the lake was discharged by overflowing the southern 
lip of the crater. Thus in 24 hours it had risen about 1000 feet, or 40 feet per 
hour. Yet, according to the evidence of all who knew it before and have seen it since, 
the interior of the crater was not greatly modified and many of its old features were 
recognisable after the eruption. Only a plastic, semiliquid mass could have risen so 
rapidly and could have been ejected with so little disturbance and so little permanent 
alteration of the topography of the interior of the crater. The great rock column 
* Loc. tit. 
t Symons, G. J., ‘ The Eruption of Krakatoa and subsequent Phenomena,’ Plate 4 (1888). 
t ‘ Montagne PeBe,’ pp. 206, 208, 383, 
