488 Dr. Johnston-Lavis — Lithophyses in Lipari Obsidian. 



ami may throw light also on the problem of the origin of quartz- 

 bearing basalts. It is noteworthy that in the districts where 

 these various abnormal quartz-bearing types occur, the association of 

 igneous rocks of very different chemical compositions and the order 

 of their succession lead us, on quite independent grounds, to the 

 hypothesis of the stratified magma-basin. 



III. — Note on the Lithophyses in Obsidian of the Eocche 

 KossE, Lipari.^ 



By H. J. Johnston-Lavis, M.D., M.R.C.S., F.G.S. 



IT is a pretty well recognized fact that Obsidian is a hydrous silicate 

 of indefinite composition and your fusion experiments confirm 

 that fact. Now molten obsidian may give off its H.,0 in consequence 

 of two processes — by simple vesiculation, or by the individualization 

 from the glass of a definite anhydrous silicate. More commonly 

 tliese two processes have gone on together, but modified by the 

 vicissitudes of temperature that the molten aquiferous glass is exposed 

 to during its eruption. 



A free surface is alike favourable to the separation of a gas from 

 solution, and to the starting of crystallization. Crystallization may 

 start from a vesicle (as I could show you at this present moment in 

 a specimen of ice) or vesicles will form on the surface of crystals. 



A spherulite, however, usually starts from some solid particle, as 

 a microlith or a grain of dust enclosed in the glass, and proceeds to 

 grow outicards in a radiating manner, but as it grows it liberates 

 some vapour by the conversion of the hydrous glass to an anhydrous 

 felspar ; and likewise the spherulite acts, when the pressure is 

 diminishing, as a catalytic or free surface, and simple vesiculation 

 takes place. 



The crystallization vapour will collect between the outer ends of 

 the fibres, press them apart, and be joined by its ally the vesiculation 

 vapour. The result is that they form a spherical vapour shell around 

 the spherulite. The highly viscous glass (as proved by faulting in it) 

 draws with it the outer ends of the radiating needles forming the 

 greater part of the spherulite, with the exception of the most resist- 

 ing part or parts on one or more sides which carries with it the 

 central remnants of the spherulitic concretion as shewn in your 

 figure 3, The spherulite may have broken up into a number of 

 cones scattered over the interior of the vesicle-walls where they are 

 attached by their bases. At the same time the sides of the cones 

 have shrunk by the crystallization of the remaining glass between 

 the component needles and the escape of the similarly enclosed 

 vapour which would cause contraction in a lateral direction, so that 



' Prof. G. A. T. Cole and Mr. G. "W. Bntler, the authors of a recent paper "On 

 the Lithophyses in the Obsidian of the Eocche Eosse, Lipari" (Q.J.G.S. August, 

 1892), having found that Dr. Jolmston-Lavis had been led to an interpretation of 

 the facts considered, different from theirs, and hitherto unpublished, asked him to 

 add to the discussion of the matter a brief abstract of his views. In accordance 

 ■with his suggestion, his reply is here published. — Edit. Geol. Mag. 



