IN OBSIDIAN FROM CALIFORNIA. 427 



of doubly refracting fibres or matted microliths. From tbis fact 

 alone it is evident that it would be useless to attempt to determine 

 the presence of tridymite, either by solubility of the pellets in a 

 solution of an alkaline carbonate or by specific gravity. 



5th. Where spaces occur between the small pellets or primitive 

 spherulites, they are filled with crypto-crystalline matter associated 

 with isotropic matter, which may be glass or which may be tri- 

 dymite. 



(jth. On comparing what 1 here regard as the secondary structure 

 developed in this spherulite with the structures produced by the 

 devitrification of artificially formed glass, there appears to be a very 

 strong resemblance in most cases, making allowance, of course, for 

 differences in texture. The ramifications of the crystalline rods here 

 indicate the existence of divergent crystalline bundles, which are 

 probably bcmnded by radial planes of arrest, planes which in this 

 instance evade detection by reason of the delicate nature of the rods 

 composing each bundle, and the consequently confused appearance 

 caused by the overlap of several fasciculi within the thickness of the 

 section ; whereas, in a more coarse phase of devitrification, a single 

 bundle would occupy the entire thickness of the section, so that, in 

 such a case, the arrest-planes would be well defined. 



Now it is evident that if the initial points of divergent groups 

 originate very close to one another along one of the circumferential 

 bauds of a spherulite (which band may be taken to represent a 

 pause in the crystalline development) those rods which diverge 

 mos.t strongly from any radius of the spherulite will be (juickly 

 arrested by the development of similar rods in the adjacent bundles, 

 and such rods will be excessively short, while those which approxi- 

 mate more closely to radii of the spherulite will be comparatively long. 

 Under such circumstances, the crystalline rods belonging to successive 

 zones will exhibit an apparent continuity. This seems to support 

 the view that the divergent crystalline structure in these spherulites 

 differs in no essential respect from the structure commonly observed 

 in manufactured glass which has been artificially devitritied. 



Taking all these points into consideration, I am inclined to think 

 that the conclusions arrived at in this paper are substantially 

 correct ; but, on the other hand, the wide experience of Mr. Iddings 

 in connexion with obsidians and the structures developed in them, 

 renders his opinion worthy of the most careful attention. 



1 trust that I have in no way misinterpreted his views, and I 

 would gladly substitute them for my own if, in this instance, I felt 

 that they offered a more satisfactory explanation of the phenomena 

 here described. 



EXPLANATION OF PLATE XVIT. 



Fig. 1. Border of composite s])Iierulite in black obsidian, from Hot-f^prings, 



near Little Lake, California, X ^5. Ordinary transniithMl light. 



The light upper portion ol" the drawing represents tlie obsidian. The 



bomidary of the spherulite is here ?een to lx> irn>g(ilar, owing to the 



partial recei)tiou of two smaller sphernlites. The latter were ti-aversed 



