BASALTIC (iLASS. 161 



These analyses are directly comparable with those of the basalt and 

 basaltic obsidian and with the andesite and andesitic obsidian from Clear 

 Lake, and the character of the differences is manifestly the same. In each 

 case the glass is comparatively very rich in alkalis and silica and contains 

 only a little lime or magnesia. 



Inferences. — lu tlic rocks ft'om Anfuxta and the Rossberg only small blebs 

 or streaks of acid glass are found. At Clear Lake, on the other hand, im- 

 mense quantities of glass, covering large areas, accompany crystallized rocks 

 in such a maimer as to leave no doubt of their direct connection. The 

 natui-e of the cases is the same, but the size of the masses is veiy different, 

 and I am not aware that any instance has ever been studied in which areas 

 of glass which must be measured by the square mile are thus connected 

 with crystallized rocks of a different chemical composition. It is plain from 

 these occurrences that associated masses of very different chemical composi- 

 tion and of great volume sometimes form portions of the same eruptions. 

 They pass over into one another by transitions, but, whether they never 

 have been more thoroughly mingled than they now are or whether, having 

 been intimately mingled, they have separated by eliquation, it is perhaps 

 impossible to decide at present. The conditions show that they were in 

 contact in a fluid state and that the passage from the crystalline to the 

 amorphous rocks is a gradual one. 



It is manifest that, in the case of these comparatively recent and super- 

 ficial rocks, the crystallization has been governed by the chemical compo- 

 sition, for the glassy and crystalline masses, while of different composition, 

 have been subjected to physical conditions which were nearly identical. It 

 cannot be doubted that there are many cases in which the differences in 

 structure of massive rocks are referable to chemical variations which are 

 perhaps numerically small. Even in the lavas it is not an infrequent thing 

 to find rounded masses wliich differ greatly in mineralogical composition 

 from the surrounding mass, and yet these have been subjected to ex- 

 actly the same physical conditions as the material in which they are em- 

 bedded. Even, therefore, if no chemical difference known to be significant 

 could be discovered, it would inevitably follow that such a difference nev- 

 ertheless existed, for variations in texture must be due to variations either 

 MON xiri— 11 



