302 Grenville A. J. Cole — Rhyolites of the Vosges. 



that the spherules had originally been formed as compact and normal 

 aggregations, but that the rock had undergone a second fusion ; the 

 glass had then penetrated the spherulites along their lines of radial 

 structure, corroding but not destroying them, nor yet disturbing, by 

 shifting of their parts, the parallelism of the earliest-formed crystal- 

 lites within. Each spherule became thus split up into a bundle of 

 diverging and rounded rays ; and Vogelsang remarks in a later 

 passage ^ on the similarity between this structure and that of certain 

 felsospherites in the " kugelporphyr " of the Vosges, in which the 

 place of the glass, however, is taken by a granular siliceous aggre- 

 gate affecting polarised light. 



That Vogelsang's explanation will not satisfy certain cases, while 

 it may probably in others be regarded more as an ingenious than an 

 inevitable deduction, is clear from sections of Hungarian " strahlige 

 sphaerolithe " in the collection of the Normal School of Science and 

 Eoyal School of Mines. In the instance figured (Fig. 1), the fluidal 

 structure of the obsidian passes continuously through the matrix 

 and the spherulitic matter, and is merely disturbed, not broken, in the 

 interspaces between the rays. While this continuity excludes the idea 

 of a second fusion, the local disturbances must have arisen from the 

 opposition offered by the rays themselves. In a glassy lava moving 

 with some rapidity, large spherulites are impossible, the separated 

 globulitic materials being carried out into parallel bands. As the 

 movement becomes slower, however, and as consolidation proceeds, 

 nodular aggregations appear locally along this banding, and, if the 

 crystallisation is finally arrested at this stage, their outer margins 

 are seen to be ill-defined and ragged, set with, in fact, irregular but 

 fairly radial rays.^ It is conceivable that as we frequently meet in 

 rocks with the mere skeletons of crystals planned on a scale too 

 ambitious for the time occupied in the consolidation of the mass, so 

 even skeleton- s-pJierulites may arise, the whole product being similar 

 in structure to the arrested external layers of the instances above 

 described. The diverging rays, set at all angles to the direction of 

 flow, will gradually interfere, after the manner of embedded crystals, 

 with the further uniform progress of the glass ; and the last move- 

 ments in the mass, together with the pressure of upper layers, will 

 serve to distort the lines of crystallites between the rays, or even to 

 break them through and rearrange the particles into local lines of flow. 



In some of the Hungarian cases examined, the interstitial matrix 

 has itself become finally spherulitic, but on a minute and delicate 

 scale, showing a mosaic of dark-cross areas when viewed between 

 crossed Nicols. In the rock of Wuenheim, however, it seems for 

 the most part to have remained glassy until the period of its 

 secondary alteration into granules. The individualised fibrous 

 structure of the spherulitic rays, which are well compared to axiolites 

 by Eosenbusch,^ would in itself assign to them an origin independent 

 of one another in all but the tendency to develope in groups about 



1 Ibid. p. 168 2 See Quart. Joum. Geol. Soc. vol. xli. plate iv. fig. 1. 



3 Mikroskop. Physiogr. 2nd edit. (1886), p. 395. 



