On the Hollow Spherulites of die Yellowstone. 157 



the Yellowstone region. A brief description is given (i) of the 

 small bluish-grey solid spherulites common in the obsidian of 

 Obsidian Cliff, and (ii) of a hollow variety in which radial 

 structure is barely discernible. In the latter, the spherulitic 

 part is represented by a whitish, rather crumbly material con- 

 sisting of felspar, tridymite, and quartz. 



The hollow spherulites proper are divided into two groups — 

 (i) those containing cavities without definite form, and (ii) those in 

 which the cavities are related to the shape and structure of the 

 spherulite. The latter include the well-known lithophysae. The 

 manner in which these occur, and the relation of the cavities- 

 to the enclosing spherulite, are described. Attention is drawn 

 (a) to the porous character of the latter, and (b) to the network of 

 felspathic fibres, studded with crystals of tridymite, which usually 

 distinguish the spherulite near a cavity. 



Hypotheses framed to account for these varying structures would 

 take one of two directions : — (i) Hollow spherulites are the result 

 of some property of the original magma, or (ii) are due to the 

 decomposition of an originally solid spherulite by heated waters. 

 Taking the second alternative first, a description is given of the 

 effect of solfataric action on the rhyolites of the Yellowstone Canon. 

 The conclusion reached is ' that the action of hot waters charged 

 with silica may be to remove portions of the rock, or to permeate it 

 without destroying its characteristic structure; that we obtain,, 

 however, no evidence to show that the spherulites are most easily 

 attacked, but rather the reverse.' Explanation, therefore, is most 

 naturally sought in some property of the original magma, and that 

 propounded by Prof. Iddings appears the nearest in accord with 

 facts. Exception is taken to certain physical processes postulated 

 by Prof. Iddings in a recent memoir, but with his earlier work the 

 present writer is substantially in agreement. 



In the second part of the paper direct comparison is drawn between 

 the structures exhibited by the hollow spherulites from Obsidian 

 Cliff and those of examples from Shropshire, Jersey, and other 

 localities. Attention is called to the presence in the latter of 

 quartzose amygdaloids, crescentic in shape, and having a relation to 

 the edge of the nodule. Sometimes a series of such are found parallel 

 one to the other, not infrequently (at AYrockwardine) becoming more 

 or less completely circular. Projecting into such an amygdaloid, or 

 occupying an end, we find in many instances a network of felspathic 

 fibres comparable with the fibrous structure which characterizes the 

 American examples. 



A description is given of a series of rocks from Boulay Bay, once 

 very vesicular, and containing the remains of crystals — probably 

 felspars, — analogous to the crystals found encrusting the cavities of 

 lithophysoe from Obsidian Cliff. Traces of a mineral which resembles 

 the tridymite from the latter locality are described from AYrock- 

 wardine. 



Taking into consideration the resemblances between the hollow 

 spherulites of the Yellowstone region and those of Great Britain,. 



