166 



G. A. J. COLE ON HOLLOW SPHEEULITES AND THEIK 



on a large scale the imperfect type of spherulitic structure, the famous 

 Hungarian examples bear the same relation to the concentric and 

 well-marked radial types. The " Lithophysen " in a rhyolite brought 

 by Prof. Judd from Lipari have affinities with both groups ; and 

 reflection will show that towards the centre of such a felsitic segre- 

 gation as has been described the material may often tend to become 

 more and more fully crystalline. Thus in smaller cases — an obsidian 

 from Yulcano, for example — the colourless radial groups of crystal- 

 lites form independent well-bounded spherulites in the midst of 

 brown clouds of globulitic dust. 



A superb example of hollow spherulitic structure occurs in an 

 obsidian brought from Iceland by Mr. J. Starkie Gardner. The 

 brown spherules, sometimes single, sometimes intergrown, show 

 radial structure to the naked eye, and, as might perhaps be expected, 

 a sharp crack divides their outer surfaces from the glass. Even in 

 the cases one may meet with in breaking the specimen across, the 

 inner cavities are often very large in proportion to the fibrous shell ; 

 but here, again, we see nothing of the smooth curving surface of a 

 steam-vesicle. The microscope shows the solid portion of the sphe- 

 rules to consist of delicate radially grouped fibres, with some trace of a 

 concentric structure near the outer margin. The inner extremities of 

 these fibres are, as it were, frayed out, with signs of the formation 

 of secondary minerals between them (PL IV. fig. 3). Everything, 

 indeed, again points to the conclusion that the hollow structure is 

 only a result of decomposition. 



iN'ow lava-streams are especially subject to the attacks of acid 

 vapours, of steam, and of water of high temperature, during the later 

 stages of eruption *, and the numerous cracks of the glassy 

 varieties will afford access to all portions of their mass. The parts 

 most likely to be affected are any felspathic crystals that may be 

 included, the crystallites of magnetite that may have separated out, 

 and the spherulites that are most differentiated from the glass. 



Long ago Scrope f pointed out that such " globular concretions " 

 are first acted upon by decomposition, and give rise to some of the 

 forms of " variolite," though he does not definitely refer the hollow 

 spherulites that he noticed in the perlites of the Ponza Isles J to 

 any process of alteration. The more developed their radial structure, 

 the more planes of weakness will the spherulites contain ; and at 

 the very centre, where the individual fibres terminate, the decom- 

 posing agents will probably find good hold. At the same time the 

 outermost layer of the spherule, between which and the glass there 

 is often small cohesion, will become reddened with iron alteration- 

 products and perhaps appreciably dissolved. Where radial and 

 partly vitreous layers alternate, the former perish while the latter 

 remain. This may be illustrated by banded obsidians from Lipari, 

 in which all stages of the process can be made out easily with a 



* Compare Judd, " Volcano of Schemnitz," Quart. Journ. Geol. Soc. vol. 

 rxxii. p. 322. 



t Considerations on Volcanoes, 1825, p. 118. 



I Trans. Geol. Soc. series ii. vol. ii. pp. 202 and 218. 



