350 REPORTS ON THE STATE OF SCIENCE.—1912, 
such slags,** and minute crystallites, which from their form must have 
segregated from the mass after it became rigid, may often be detected.'? 
The early stages of devitrification are easily seen in slags which have 
been granulated by means of air or water.’® 
Natural glassy rocks offer a still better series of examples. Here, 
again, all stages of devitrification may be observed, the spherulitic 
mode of crystallisation being frequent, as in rhyolites. It is some- 
times possible to show that the crystallisation must have taken place 
at a comparatively low temperature. Thus, in the pitchstones of Arran, 
the phenocrysts are of augite or enstatite, but the minute crystallites 
which are so characteristic of that rock are of hornblende, a low- 
temperature mineral.'* The great viscosity of the mass at the time 
of formation of the hornblende is shown by the clear areas, free from 
microliths, surrounding the crystallites. Similar crystallites are found 
in the lava of Mauna Loa,'> whilst the extreme slowness of devitrifica- 
tion is shown by their absence from some recent lavas, even with a 
high percentage of iron, as that of Kilauea. Remelting the dolerite of 
Rowley Regis also yields a clear glass. That the process of devitrifica- 
tion—that is, of passage into the stable condition—proceeds to com- 
pletion if given sufficient time is shown by the fact, noticed by 
Harker, that glassy rocks of Paleozoic age are almost unknown, 
although many rocks of that period show, by the presence of perlitic 
and other structures, that they have previously been glassy. The 
effects of devitrification may be produced artificially by heating pitch- 
stones and other glassy rocks.*® 
The crystallites formed in all the above instances are of different 
composition from the glassy mass, and their formation therefore 
involves diffusion. The devitrification of such a substance as silica, 
however, is merely a process of molecular change, not necessarily 
accompanied by diffusion. Some geologists have attributed still greater 
importance to the growth of crystals in a glassy magma at low tempera- 
tures, and have explained a number of structures frequently observed 
in rocks to the growth of previously formed crystals at the expense of 
a small quantity of glassy ground-mass.*” 
Porcelain is probably partly glassy and partly microcrystalline. 
The diffusion of solid carbon, either in the amorphous form or as 
graphite, into porcelain at 1000°-1500° has been recorded on more than 
one occasion.!® The mechanism of the process has not been studied, 
but the presence of the carbon at a considerable depth below the surface 
has been determined both microscopically and chemically. 
11 Gee, for instance, W. Mathesius, Stahl. u. Hisen, 1908, 28, 1121; M. Theusner, 
Metallurgie, 1908, 5, 657. 
12 H, Vogelsang, ‘ Die Krystalliten ’ (Bonn, 1875). 
13 C, H. Desch, ‘ The Chemistry and Testing of Cement ’ (London, 1911), p. 98. 
14 A, Harker, ‘ The Natural History of Igneous Rocks’ (London, 1909), p. 226. 
16 B.S. Dana, Amer. Jour. Sci., 1889 [iii.], 37, 441. 
16 F, Rutley, Proc. Roy. Soc., 40, 430. 
{17 J. W. Judd, Quart. Jour. Geol. Soc., 1889, 45, 175. 
18 R. §. Marsden, Proc. Roy. Soc. Hdin., 1880, 10, 712; J. Violle, Compt. rend., 
1882, 94, 28. 
