492 Prof. S. J. Shand — Saturation in Petrograjjhy. 



transforms into forsterite and glass. It follows from these results 

 that forsterite and free silica can onlj^ coexist in a magma which 

 commenced to freeze at some temperature in excess of 1,557°. There 

 is not a shadow of evidence that any natural magma had such an 

 enormously high, freezing-point (meaning the point at which freezing 

 commenced). Every one of the common rock minerals whose 

 melting-point is known freezes below 1,550°, and the majority of 

 them much below it. The freezing-point of a magma must therefore 

 be much lower still, in accordance with Raoult's Law, apart from all 

 consideration of tlie action of ' mineralizers '. Brun ' and others have 

 shown that the melting-points of fresh lavas lie between 950° and 

 1,200°; the Palisade diabase of New York ^ is readily fluid, in the 

 crucible, at about 1,200° ; Goldschmidt^ has estimated the maximum 

 temperature which prevailed in contact zones in the Christiania 

 district at 1,000^ to 1,200°. The fact is, that instead of disproving 

 my contention, the work of Anderson and Bowen actually proves 

 it by showing that (in a dry melt) forsterite and free silica do not 

 form side by side below 1,557°. Olivine includes fayalite, however, 

 as well as forsterite, and the occurrence of fayalite together with 

 quartz or tridymite in the lithophysse of acid lavas and the druses of 

 granites has been described by Judd, Lacroix, Iddings, Quensel, and 

 others. Iddings* offers the following explanation : — 



"... the concurrence of fayalite and quartz in lithopliysse in highly 

 sihceous rocks shows that silicon may be deterred from forming the higher 

 silicate under certain conditions, which appear to be, either those that permit 

 the water present in rock-magmas to enter into combination with the silica, 

 forming a hydrogen silicate H4 Si O4, which . . . subsequently breaks up into 

 H2 O and Si O.2 which may appear as quartz ; or such that water may act 

 catalytically to promote the immediate separation of quartz before the separation 

 of the ohvine compound from the solution." 



This is a possible explanation of the conjunction of fayalite and 

 quartz ; it would apply especially to the contents of vapour cavities 

 and druses, where magmatic water would be concentrated towards 

 the completion of the freezing of the magma. But a simpler 

 explanation is available which involves no supposition whatever. 

 It is just this, that the compound Fe Si O3 is incapable of independent 

 existence under any conditions realized in nature — it is not known as 

 a mineral — hence the lower silicate, fayalite, is stable even in presence 

 of quartz. I ought, on this account, to correct my former lists of 

 saturated and unsaturated minerals by adding the rare fayalite to 

 the former, retaining only the magnesian olivines and forsterite in the 

 latter. The change is a slight one, since practically all common 

 olivine is magnesian and therefore remains on the unsaturated list. 

 "Washington, in his recent description of the volcanic rocks of 

 Pantelleria,^ describes an ' olivine ' existing in small amounts in the 

 acid lavas, and shows by analysis that this is a nearly pure fayalite, 



^ Recherches sur Vexlialaison volcanique, ch. vi. 



^ E. B. Sosman & H. E. Merwin, Journ. Washington Acad. Sci., 1913. 



^ Die Gesetze der Gesteinsmetamorphose , Christiania, 1912. 



* Igneous Rocks, vol. ii, 1913. 



* Journal of Geology, 1914. 



