440 PROF. T. G. BONNEY ON DEVITRIFICATION [NOV. I903,. 



the crystallites in the former, the more irregular are the outlines 

 in the latter. The three structures have much the same origin, but 

 in two of them crystal-building, owing to local circumstances, has 

 proceeded on a rather more definite plan than in the third ; and we 

 observe that the more root-like the micrographic structure the 

 more irregular are the granular outlines. Thus the departure 

 from a rectilinear habit in the components of a de- 

 vitrified rock is a measure, figuratively speaking, of 

 the severity of the struggle between them. 



Secondary Devitrification. 



We pass now to secondary devitrification. Here also the prin- 

 ciples which we have endeavoured to establish may be applied. 

 Although it would perhaps be rash to assert that the occurrence of 

 spherulites invariably signifies an elevation of temperature almost 

 enough to melt the glass, this is certainly most effective in producing 

 them. We may even say that, in secondary devitrification (at ordinary 

 temperatures, so far as we are aware), not only is the ' trachytoidal ' 

 (microlithic) structure unknown, but also, apart from spherulites, 

 any form of rectilinear structure is rather rare ; indeed, the circum- 

 stances uoder which we most commonly find the latter are them- 

 selves suggestive. It is associated with a perlitic structure (itself 

 a very strong presumption in favour of the rock having once been 

 a glass), and in close relation, as Mr. Parkinson and I have inde- 

 pendently observed, with the perlitic cracks. To these the lines 

 separating the quartz and felspar are often rudely perpendicular, so 

 that the interval between two concentric cracks may be occupied by 

 an alternation of these two minerals, while in uncracked portions of 

 the slice they may appear as ordinary curvilinear grains. Naturally- 

 devitrificd rocks, without any perlitic cracks, commonly exhibit a 

 speckly, rather minute, and irregularly-outlined structure, resem- 

 bling that frequent in cherts. This might be expected, because the 

 material of the latter probably crystallizes under constraint ; for, in 

 some cases, it was originally a colloid and subsequently became 

 microcrystalline, in others the crystals in forming had to deal with 

 mechanical obstacles (particles of clay, etc.). We sometimes find 

 this speckly structure restricted to parts of a slice which are 

 rendered less translucent by the presence of minute dust (probably 

 ferruginous), while the clearer are more coarsely granular ; or to 

 parts which are free from perlitic cracks, the ' grain ' becoming 

 coarser when these appear. 



The fact that curvilinear, more or less ragged-edged, granulation 

 is the normal type in the secondary devitrification of a glassy rock 

 points in the same direction, and is confirmed by the fact that 

 annealed steel often shows the same structure. 1 



1 See figures in Mr. Stead's paper on the ' Crystalline Structure of Iron & 

 Steel ' Journ. Iron & Steel Inst. vol. liii (1898), one or two of which might serve 

 for outline-drawings of some felstones. Here the steel had been heated to a 

 temperature rather above 750° C, and the coarseness of the granulation mainly 

 depended on the time for which this had been maintained. 



