466 
JOHNS : EXPERIMENTAL PETROLOGY. 
sufficiently to cause any crystals already formed to enter, or 
commence to enter, into solution again. Even if the liberation 
of sensible heat were not enough to raise the temperature of 
the mass, it might be sufficient to maintain it stationary for 
a time, and thus cause structural modifications. Further, at 
any stage the mother liquor might arrive at a composition 
which would freeze as a solid solution or glass. 
As a small contribution to the work, the results of certain 
experiments made recently by the present writer are offered, 
with the warning that the conclusions are but tentative and onh- 
to be adopted with caution. Silica and silicates may be looked 
upon as the chief solvents in igneous rocks. It is true that 
silica also, at some stage or another, appears as a solute and 
separates out in most of the igneous rocks, but that does not 
seriously affect the view that we are dealing with solutions of 
various minerals in fused silicates of a composition that varies 
as the consolidation of the rock mass progresses, and that the 
solubility of these minerals in any particular silicate is a function 
of temperature, except when it is capable of forming a solid 
solution. Silica would therefore appear to offer a promising 
line of attack. 
A quantity of pure quartz sand was taken and exposed at 
a temperature exceeding 1,800° C. to an atmosphere containing 
much finely divided magnetic oxide of iron. The temperature 
was just sufficient to fuse the quartz, thus confirming Boudouard's 
determination that the fusion point of silica is 1,830° C. When 
the mass of quartz sand was withdrawn, it was found that the 
surface, to the depth of eight millimetres, had been fused, and 
that it was coated with magnetic oxide. The coating was a little 
less than one millimetre in thickness, but after cooling it could be 
easily removed with the finger nail. It was very evident that 
magnetic oxide is not soluble in fused silica. Further experi- 
ments were then made, during which silica was exposed at 
varying temperatures for a period extending over several weeks, 
to finely divided magnetic oxide. In no case was there any sign 
of solution, and it seems established that magnetic oxide is not 
soluble in silica between 200° C. and 1,830° C, which were the 
limits during the experiments. 
