436 PROF. T. O. BONNEY ON DEVITRIFICATION [Nov. I9O3, 



differences in the composition of the minerals present as conspicuous 

 crystals and in the groundmass, or by variations in the amouDt of 

 water in the mixture, for that also must produce some effect on the 

 conditions of consolidation. 



Before proceeding farther we shall find it convenient to give a 

 brief summary of the different types of structure. These may be 

 classified (more for convenience than as implying hard-and-fast 

 divisions) as the linear and the granular. The linear may be 

 subdivided into (a) the rectilinear, (b) the curvilinear. 



Linear Structure. 



(a) The rectilinear. — For this a mere mention will suffice: 

 its earliest stage is the formation of microliths like the felspars in 

 an andesitic glass, its latest that of the porphyritic crystals often 

 found in holocrystalline rocks. The augites from near Predazzo, the 

 leucites of Somma, the felspars of the Mairus porphyroid and the 

 Lamorna granite (to quote a few conspicuous instances) probably 

 imply that the magma in which they formed was at one time super- 

 saturated with the constituents of a particular mineral, which were 

 not indeed wholly removed by crystallization, but were, so to say, 

 reduced by it, so that they could be kept in check by the represen- 

 tatives of other minerals — for I apprehend that the formation 

 of an eutectic is equivalent to a temporary deadlock 

 in a struggle for priority. 



(l>) The curvilinear, into which we have virtually been led 

 in the last few words. It is represented in its first stage by tri- 

 chites, but is usually found in groups, as in some spherulites and in 

 the micrographic, or micropegmatitic structure, 1 of which that long 

 known as graphic is only a variety. We shall find, I think, that 

 so far as there is any difference between these two structures, it 

 depends very largely on the nature of the obstruction offered. 



Spherulitic structure in its simplest stage is apparently no more 

 than a radial grouping of molecules, 2 as perhaps in some of the 

 clear, almost structureless spherulites which on crossing the nicols 

 give rather distinct black crosses ; but it is generally associated (as 

 taking place in a ' mineral mixture ') with some amount of separa- 

 tion. Occasionally this is extrusive, as (to a slight extent) in 

 ordinary banded spherulites, and more conspicuously in the holo- 

 crystalline spherulites of orbicular granite or corsite — but commonly 

 the magma separates into two minerals (exclusive of minute iron- 



1 Now commonly designated the ' granopkyric.' Apart from the fact that the 

 word itself, like all but one ending in ' phy re ' or ' phyric,' is nonsense, 

 Vogelsang, its author (as I believe), used it in another and partly-appropriate 

 sense. 



2 This must be the case in the spherulitic structure occasionally found in 

 chalcedony. 



