Vol. 59.] PEOF. T. G-. BONNEY ON DEVITRIFICATION. 435 



Part IL— By Prof. T. G. Bonney, D.Sc, LL.D., F.R.S., F.G.S. 



The solidification of a mass of given chemical composition is a 

 question of temperature, its crystallization of environment. We 

 may regard a glass as a mixture of molecules, each endowed with 

 polarities enabling it to join in building a crystal belonging to a 

 particular group, but at present associated without the orientation 

 which is essential for- the existence of a crystal. 1 -But in a solution 

 crystallization may involve not only change in orientation, as when 

 opal is converted into chalcedony, but also change of place, an 

 aggregation of certain molecules which previously were mixed up 

 with others. This is illustrated by the 'clarification' of a tachy- 

 lyte when the ultra-microscopic particles of iron-oxide in the brown 

 glass begin to collect together in visible granules, by the formation 

 of feathery groups of crystallites (hornblende) from the surrounding 

 dusty glass in some of the Arran pitchstones, and in all non- 

 crystalline rocks consisting of two or more minerals. This grouping 

 of like with like is due to an attractive force which, under certain 

 circumstances, though to a more limited extent, can also act in a 

 body while it remains solid ; as when, in some stalactites, minutely- 

 crystalline calcite becomes coarsely crystalline, or steel and other 

 metals become crystalline under pressures and vibrations, 2 and glass 

 softened by heat is devitrified. Any local discontinuity, such as the 

 existence of an outer surface or of an included solid, is favourable 

 to crystallization, because that is a process like building, and thus 

 is facilitated by a ready-made foundation. Moreover, as heat is 

 generally lost by radiation from the outer part of a mass, crystal- 

 lization naturally begins here, at the coolest part. When the con- 

 ditions in the neighbourhood are uniform, then, if crystallization 

 starts from a (non-mathematical) point within the mass, it will 

 proceed uniformly in all directions and produce a spherulite ; if from 

 a line, an axiolite ; if from a surface, some dependent form. In 

 the case of a plane (such as the outside of a piece of glass) the crys- 

 tallites might be arranged like a mass of parallel rods, but more 

 usually, since certain spots in it afford slight advantages, they 

 form tufts diverging from centres, which are occasionally so far 

 separated as to produce i hemispherical spherulites.' 



As crystallization usually requires time, 3 a slow fall of tempera- 

 ture is favourable to it, and indirectly to the reduction of a solution 

 supersaturated with one or more minerals to an eutectic, because 

 thus the constituents in excess are separated out. Here we often 

 meet with apparent anomalies, such as the separation of both 

 quartz and felspar from an acid magma, or that of both felspar and 

 augite from a basic one. These may be explained, either by slight 



1 The optical effects of strain in a colloid indicate a temporary orientation ot 

 its molecules. 



2 Journ. Iron & Steel Inst. vol. liv (1898) p. 185 & vol. lviii (1900) p. 60. 



3 Mr. J. E. Stead remarks, Journ. Iron & Steel Inst. vol. liii (1898) pp. 151— 

 52, that ' the crystallization of steel requires a certain amount of time as well 

 as a certain degree of heat.' 



