_ Part I. Secor. iv.§ 1.] EXPERIMENTS IN FUSION. 295 
vitreous varieties, from 0 to 4 per cent.’ More recently Mr. Mallet 
has observed that plate glass (taken as representative of acid or 
siliceous rocks) in passing from the liquid condition into solid glass 
contracts 1°59 per cent., 100 parts of the molten liquid measuring 
_ 98:41 when solidified; while iron-slag (having a composition not 
unlike that of many basic igneous rocks) contracts 6-7 per cent., 
100 parts of the molten mass measuring 93°3 when cold. By 
the contraction due to such changes in the internal condition of 
subterranean masses of rock minor oscillations of level of the surface 
may be accounted for, as already stated (p. 284). Thus the vitreous 
solidification of a molten mass of siliceous rock 1000 feet thick might 
cause a subsidence of about 16 feet, while, if the rock were basic, 
the amount of subsidence might be 67 feet. 
Difference between the products of artificial fusion and 
natural lavas.—lIn the experiments of De Saussure, Dolomieu, Hall, 
and subsequent observers, it has been found impossible to obtain 
from a piece of fused rock a crystalline substance exactly resembling 
the original mass. Externally it may appear quite stony, but its 
internal structure, as revealed by the microscope, shows it to be 
essentially a slag or glass, and not a truly crystalline rock. There 
is another fundamental difference between the natural and artificial 
products. When a compound containing substances of different 
fusibilities is artificially melted, and allowed thereafter to cool in 
such a way that the various ingredients may separate from each 
- other, they appear in their order of fusibility, the most refractory 
coming first, and the most fusible being the last to take a solid form. 
But in rocks which have crystallized naturally from a fluid condition, 
it is often to be observed that the component minerals have been far 
from obeying what might have been supposed to be their invariable 
law. Thus, in all parts of the world, granite presents the very 
striking fact that its quartz, which we call an infusible mineral, has 
actually solidified after the more fusible felspar. In the Vesuvian 
lavas the difficultly fusible leucite may be seen to have enclosed 
crystals already formed of the fusible augite. In some ancient 
crystalline rocks the pyroxenic constituents, which offer a less 
resistance to fusion, have assumed a crystalline form before the more 
refractory triclinic felspars. From these facts it is clear that, in the 
fusion of rocks and in their subsequent consolidation, there have been 
conditions under which the normal order of appearance of the 
minerals might be disturbed or reversed. 
Yet another fact may be mentioned to show further the difference 
between the kind of fusion which has frequently obtained in nature 
1 Bull. Soc. Géol. France, 1847, p. 1390. Bischof had determined the contraction of 
granite to be as much as 25 per cent. (Leonhard und Bronn Jahrb. 1841). The correct- 
ness of this determination was disputed by D. Forbes (Geol. Mag. 1870, p. 1), who 
found from his own experiments that the amount of contraction must be much less. 
The vaiues given were still so much in excess of those recently obtained with much care 
by Mallet, that some defect in their determination may be suspected. 
2 Phil. Trans. clxiii. pp. 201, 204; clxv.; Proc, Roy. Soc. xxii. p. 328. 
