344 4, RUTLEY ON STRAIN IN CONNEXION WITH CRYSTALLIZATION 
The phenomenon is superficial and seems only to follow the cracks. 
We may feel tolerably sure that the glass was not cracked in this 
manner before the house was burnt. ‘The cracks are precisely such 
as would be produced in glass or rock crystal by heat, and subsequent 
immersion in a cold solution; and these conditions would, in this 
case, be fulfilled by the fire and the play of water from the fire- 
engines. Whether these little circular crystalline aggregates repre- 
sent crystallization set up actually zm the glass, or result from the 
crystallization of some substance other than that of the glass, intro- 
duced in solution into the fissures, I am not prepared to say. How 
far water has taken a part in the production of spherulitic and perlitic 
structures in vitreous rocks is a question yet to be ascertained; but 
it is well known that perlites contain from 2 to 4 per cent. of water. 
An analysis by Richter of the Spechthausen pitchstone given in 
Roth’s ‘ Gestein’s Analysen’ (Berlin: 1861), shows the presence of 
5‘o per cent. of water, while another analysis by Erdmann of a 
spherulite from the pitchstone of the same locality shows only °3 per 
cent. of water. It appears, then, that the spherules themselves contain 
less water than the surrounding glass. To follow this subject one 
step further, I may mention that in a section of perlite from Buschbad, 
near Meissen, the perlitic fission is rendered visible between crossed 
nicols by strings of doubly refracting spherules which closely follow 
the cracks; while in another section of perlite from Schemnitz the 
perlitic fissures are also luminous between crossed nicols; but the 
character and disposition of the crystalline particles cannot be 
resolved with a 4-inch objective. A section cut from the weathered 
surface of a specimen of pitchstone from Monamore Mill, in Arran, 
is very instructive. A spherulitic structure pervades the whole 
section, but it especially favours very fine cracks which run in 
different directions but in tolerably straight lines through the pre- 
paration, and along these cracks the spherules are very thickly 
clustered. There are a few felspar crystals in this section, and around 
them the spherules have also collected in swarms, each crystal, or 
group of crystals, having a well-defined border of them. The 
ground-mass of the section consists of a brownish glass thickly 
charged with fine dusty matter and small spicules. If we assume 
that access of water along the fissures has had anything to do with 
the formation of the spherules, we have to face the question how 
it is that the felspar crystals are also bordered by them, and how it 
is that isolated spherules occur plentifully in the glassy ground- 
mass. We may imagine the existence of diminutive fissures 
close around the crystals of which all evidence is obliterated 
by the spherulitic crystallization ; but we find isolated spherules in 
obsidians which, in thin section, are as clear as window-glass and 
where there are no traces of even microscopically small cracks. 
Clearly, then, we must regard the spherules in the same light in 
which we regard the individual porphyritic crystals which are 
formed in a glassy magma in a manner which appears to us utterly 
capricious. As with the spherules, so with the more definitely 
developed crystals, there is always a tendency to segregate around 
