164 G. A. J. COLE OTT HOLLOW SPHEETJLITES Al^D THEIK 



There is, then, no essential difference between the composition of 

 the matrix and the solid portions of the nodules, and the latter are 

 therefore far more likely to have been produced from the rock by- 

 devitrification during- cooling than by any subsequent alteration. 



Two determinations bearing out this view are given of the 

 specific gravity of the gToundmass, viz. 2'*! 10 and 2*403, that of the 

 material of the "Lithophyse" being 2-420. The figures expressing 

 "loss on ignition" in the analyses may, slight as the difference is, 

 suggest that decomposition has already begun in the devitrified 

 portions of the rock. 



J. Roth*, commenting on Yon Hauer's analyses, fully agrees 

 with Szabo in regarding the hollow structure as the result of the 

 decomposition of spherulites, and this opinion appears to be generally 

 adopted. Zirkel t has thus referred to " Lithophyses " from the 

 Yellowstone area as " sphEerolites nearly as thick as a walnut, that 

 develop, by decomposition, the concentric layer-structure," all stages 

 occurring " between sphserolites in the natural state and cavities in 

 which there are five or six shells with their isolated borders." Mr. 

 William Semmons has kindly allowed me the use of specimens 

 showing similar features, which he collected from the obsidian 

 cliffs of Beaver Lake. 



E. Weiss X, dealing with a group of rocks from the Thiiringian 

 Porest, insists on the fact that here also the hollow spheroids met 

 with are nothing but larger spherulites, and points out that the 

 solid parts of both show radial structure in thin sections ; but he 

 returns towards the older theory of Yon Richthofen in suggesting 

 that material separating from a glass may be built up around a 

 vesicle. 



We are left, then, with two possible explanations of the hollow 

 nodules found in lavas. Their outer and solid portion may safely be 

 ascribed to spherulitic segregation from the matrix ; but the inner 

 cavity may be either the result of weathering upon a complete 

 spherulite, or a steam-vesicle that has acted, like an included crystal 

 or foreign body, as a centre of devitrification during cooling. 



It does not appear that any real evidence has been brought forward 

 in support of the latter view ; but as a vesicular origin has been 

 assigned to some of the Welsh nodular fel8ites§, and perhaps, by 

 implication, to those of the Llanberis Pass, it may be well to inquire 

 whether the decomposition of spherulites is not in reality sufiicient 

 tD explain the characters of these rocks. 



The determination of the conditions under which exceptionally 

 large spherulites arise must be left to those who are studying the 

 devitrification of artificial glass. Such forms are so frequently 

 associated with small ones that the variation in size of the Silurian 

 nodules is no evidence against their being all of common and 



* Beitrage zur Petrograpbie der plutonisehen Gesteine, 1869, p. 168. Also 

 Allgemeine und chemische Geologie, 1883, toI. ii. p. 9. 



t U.S. Exploration of the 40th Parallel, Microscopical Petrography, p. 212. 

 \ Zeitschrift der deutschen geol. Gesell. 1877, vol. xxix. p. 421. 

 § Quart. Journ. Gleol. Soc. vol. xxxviii. p. 295. 



