442 PKOF. G. A. J. COLE AND ME. G. W. BUTLER ON THE 



tube ; evidence of this etching may be seen in the moat-like depres- 

 sions around the bases of many of the cones or ' mushrooms/ 

 Conical groups of fibres become thus deposited, projecting into the 

 cavity in continuity with those already developed in the wails. In 

 the same manner, the divergent groups of fibres are deposited on 

 the ends of the cones and on the ' bridges.' Considering how the 

 whole mass of a lava is permeated by fluids, materials may be 

 brought, even from some distance, into the local sphere of hydro- 

 thermal action. Tridymite finally arises, and even the fayalite 

 described by Messrs. Iddings and Penficld.^ 



Lastly, the possible intrusion of glass by pressure from without, 

 as in the Yellowstone Park ^ and the Tynemouth Dyke,^ must not 

 be lost sight of in seeking the source of the materials in the vesicles. 

 Such intruded glass would rapidly become attacked, if the vapour 

 were still present in the cavity, and would quickly help in the 

 infilling. 



The following experiments show the differences between the 

 materials in and around the lithophyses, and help to strengthen the 

 above suggestions. 



(i) The black obsidian contains sufficient water to allow of its 

 fusion at about 4 of Von Kobell's scale ; it froths up during fusion 

 into a pumice or 'thread -lace scoria.' 



(ii) The outermost porcellanous zone of the lithophyses behaves 

 similarly, becoming transparent just before fusion, probably by 

 reabsorption of its globulitic matter ; when again cool, it remains 

 transparent. 



(iii) A compact grey rock from the Forgia Yecchia, appearing 

 brown in section, and composed of densely-set spherulites, resem- 

 bling the material of the porcelJanous zone at the E,ocche Rosse, also 

 froths up as if it were a simple obsidian. 



(iv) But the compact blue-grey spherulitic growths formed in the 

 centre of the lithophyses of Lipari are only fusible at a temperature 

 as high as that required for orthoclase, though at last they intumesce 

 like the obsidian. 



(v) The fibrous white rays, partly projecting into the cavities, 

 also fuse at about 5 of Von Kobell's scale. 



The fact that empty vesicles of all sizes lie in the pumice and 

 obsidian side by side with almost completely infilled examples shows 

 that the liberation of gas may occur prior to rather than during 

 crystallization. There is doubtless every link between the structures 

 at the Rocche Rosse and the lithophyses with large central cavities 

 described by Von Richthofen * and Iddings, though we may differ 

 as to the mode of origin of the cavities. Von Richthofen believes 

 that hydrated silicates crystallize out and form an aggregate, which, 

 on relief from pressure, becomes dehydrated and consequently 



1 Amer. Journ. Sci. vol. xl. (1890) p. 77. 



2 Iddings, Seventh Ann. Ee, ort U. S. Geol. Survey, p. 283. 



3 Teall, ♦ Amygdaloids of the Tynemouth Dyke,' Geol. Mag. for 1889, p. 481. 

 * ' Studien aus den ungari-ch-siebenbiirgischen Trachytgebirgen,' Jahrb. der 



k.-k. geol. Eeichsanstalt, 18G0, p. 180. 



