168 G, A. J. COLE ON HOLLOW SPHERIJLITES AISTD THEIR 



Sometimes singly, sometimes in groups, these centres of tlie larger 

 nodules point to the origin of much of the hollow structure ; while 

 in other cases, just as in. the Yellowstone obsidian, the decom- 

 position may have ramified through fairly uniform segregations. 



The matrix, now much altered, is magnificently perlitic, and the 

 removal of any kaolin would thus prove easy ; but the rock is also 

 rudely cleaved, the spherules themselves have occasionally been 

 faulted, and ample time has been allowed since the dying-out of 

 Snowdonian fumaroles and hot springs for the work which, doubt- 

 less, began to be silently carried on to our own day. If, as I hope, 

 one of the more perplexing structures of our ancient lavas may be 

 adduced as additional evidence of their similarity to those of modern 

 days, the object of these notes will have been amply served. It is 

 probable, however, that no petrographical characters, nothing, indeed, 

 short of stratigraphical evidence and comparison from point to point, 

 can absolutely define the relations of an igneous mass. The 

 orbicular segregations of a deep-seated rock, classed as spherulites 

 by Vogelsang, but defined as " belonospherites " rather than 

 "• felsospherites " *, might in time give rise to hollow nodular 

 structure. On the other hand, the complete demolition of the 

 spherulite, an operation going on in the Icelandic specimen of Mr. 

 Starkie Gardner, may in future ages reduce such a rock, whether it 

 occur as an interbedded or intrusive mass, to the level of a 

 scoriaceous obsidian or an amygdaloid. 



In conclusion, my best thanks are due to Prof. Judd for kind help 

 and many illustrative specimens, and also, as before mentioned, to 

 Mr. William Semmons. Most of the rock -sections referred to have 

 been prepared in the Geological Laboratory of the ]N"ormal School of 

 Science and Eoyal School of Mines. 



EXPLANATION OF PLATE lY. 



Fig. 1. Section of obsidian from Iceland, with ill-defined spherulites developing 

 along the fines of banding. X 12. 



2. Section of obsidian from Beaver Lake, Yellowstone Park, for compa- 



rison with the above, showing part of the outer margin of a large 

 hollow spheruHte, with the banded structure of the glass passing 

 into it without interruption. Some of the bauds are composed of 

 minute spherulites. X 7. 



3. Section of part of the inner margin of a large hollow spherulite in 



obsidian from. Iceland, showing a tendency to the formation of 

 secondary minerals betweeu the fibres of which it is composed, x 75. 



4. Broken surface of altered rhyolite from the Wrekin area, showing large 



spherulite with structure characteristic of " Lithophysen," the cavities 

 being lined with crystallized quartz. X 1^. 



5. Section of large hollow spherulite in altered perlite from Esgair-felen, 



Pass of Llanberis. The left-hand portion contains an inner nucleus 

 with well-marked radial structure ; the right-hand portion contains an 

 irregular cavity partly filled with quartz. X 5|-. 



* Die Krystalliten, p. 134. 



