14 



PEOCEEDINGS OF THE NATIONAL MUSEUM 



VOL. 6S 



absorbed gases upon the molten frlass from which they were liberated 

 during the crystalization consequent upon cooling.'' 



Below is given the normative composition of the glass, spheru- 

 lites, and lithophysae, calculated from the analyses given by 

 Iddings,-- a comparison of which will show the close relationship 

 between them and immediately suggest the origin of the minerals 

 from the glass. 



Normative composition of ohsidian, spherulitcs, and litliophysac 



These bodies, are in fact, as other investigators have pointed out, 

 but special types of crystallization of obsidian. It only remains to 

 determine the mechanics of a crystallization that will give rise to 

 unstable forms of silica and the other characteristic features of 

 lithophysal structure. 



We have already seen that the lithophysae are closely related to 

 the spherulites. Since they are but one step advanced, they must 

 have their beginnings in spherulitic crystallization. Iddings has 

 shown that the crystallization of the spherulite began in the separa- 

 tion of a microlite from which the further crystallization proceeded 

 in all directions. If its environment was rich in dissolved water, 

 this would be liberated and begin its work of solution immediately. 

 The glass, being easily soluble, will be readily attacked, but a satu- 

 rated solution of glass will be supersaturated with respect to the crys- 

 talline phases and one or more of them must crystallize out. In Fig- 

 ure 5 the relation of the silica minerals to temperature and solubility 

 is given. The actual value of the solubility of these minerals is of 

 course not known, but their relative solubility is known from their 

 stability relations. Glass, being the least stable, is the most soluble, 

 followed in turn b}^ cristobalite, tridymite. and, finally by quartz, the 

 stable phase, which is the least soluble. A solution at the tempera- 

 ture and concentration X (fig. 5), for instance, in contact with glass, 

 would then be supersaturated with respect to all the crystalline 

 phases of silica ; cristobalite. tridymite, and quartz, and one or more 

 of these must separate out in crystalline form. The fact that all 

 three crystalline phases are apparently stable in contact with each 



2= 7th Ann. Rep. U. S. Geo], Sui-v., 1888, p. 291. 



