330 Reports and Proceedings — Geological Society of London. 



appear in tlie crystallizing glass. ' Porous sphernlites ' are once 

 more mentioned, in order to call attention to the ' feather-like ' 

 crystals which often distinguish them, and of which an explanation 

 is given in Part 11. Eeference is made to the conditions which 

 favoured primary devitrification at Obsidian Cliff; and the author, 

 leaving general principles to be discussed in the second part, 

 mentions one or two special types of primary devitrification. These 

 are concerned with the probable formation of eutectic zones, or 

 patches ; either following the crystallization of an overplus of any 

 given material, or as a residuum. After a brief reference to 

 secondary devitrification, this part of the paper concludes with 

 a summary in which the several relations of secondary to primary 

 devitrification structures are given. 



Part II. — By Professor T. Gr. Bonney. 



Crystallization in a colloid mass involves an orientation and 

 commonly a separation of the molecules ; a process illustrated in an 

 early stage by the formation of microliths in a glass, and the 

 devitrification of the latter when it is heated without actual 

 melting, or by a metal becoming crystalline under strains. Certain 

 conditions, such as slow cooling, supersaturation, and the presence 

 of inclusions — anything causing discontinuity — are favourable to 

 crystallization, some special cases of which are discussed in the 

 paper. The structures thus formed in rocks may be classified as 

 (1) the linear and (2) the granular; and the former may be 

 subdivided into (a) the rectilinear, (h) the curvilinear. Spherulitic 

 structure in its simpler form falls under (a), and is at first little 

 more than a radial grouping of molecules, the process becoming, 

 as described, gradually more complicated. Of this, ' graphic ' or 

 'pegmatitic' structure is a final stage, where two minerals are 

 crystallizing out of a solution, and one has slightly the advantage 

 over the other, so that it virtually forms a skeleton crystal. Into 

 this the ordinary radial growth of a spherulite may be seen to pass ; 

 likewise also examples of (a) into those of (b) : the latter being due 

 to the ' leading ' mineral meeting with a rather stronger resistance, 

 as if a crystal were forming in a very tough jelly. An experiment 

 of Messrs. J. I' Anson and E. A. Pankhurst (Min. Mag., vol. v, 1884, 

 p. 34) on the formation of tubes of colloid silica from a fluid 

 alkaline silicate, affords a good illustration of this curvilinear 

 growth. Resistances, as the author has pointed out in an earlier 

 paper, are favourable to actinolitic and branching growths, and the 

 various types of structure mentioned above can be shown to be 

 dependent on them. 



The granular structure is next discussed, and explanations are 

 offered of its varieties. This, on a microscopic scale, is often 

 a result of devitrification, where (so far as is known) there has 

 been no marked rise of temperature ; and the author shows how 

 this is affected by greater or less freedom of molecular motion, 

 discussing also cases in which a crystalline mass, like a spherulite, 

 has undergone a later rearrangement. 



