OTHER MODES OF CLASSIFICATION. 217 



Again, these two views, which are usually held by the same persons, 

 are by them connected with a third view, in regard to the original con- 

 stitution of the earth's crust. On first cooling, the outer layer is sup- 

 posed to have been highly oxidized, highly siliceous, and therefore com- 

 paratively light — in other words, granitic ; beneath this was a less oxi- 

 dized, less acid layer, and so on progressively, the deeper layers becom- 

 ing heavier and heavier, and more and more basic. The first eruptions 

 were from the outer layer, and therefore granitic. Afterward, as the 

 crust grew thicker and thicker, the eruptions were from deeper and 

 deeper layers, and therefore denser and denser, and more and more 

 basic. 



But, in answer to these views, it may be said that, as to age, there 

 can be no doubt that granite, though most commonly associated with 

 the older rocks, is found in strata of all ages up to the middle Tertiary, 

 and fissure eruptions have occurred in all ages up to the latest Tertiary. 

 The granite of Mont Blanc was pushed up at the end of the Eocene 

 (Lyell), and the great fissure-eruptions of the Northwest took place 

 at the end of the Miocene and during the Pliocene.* Also, as to com- 

 position, trachyte and liparite have much the same chemical composi- 

 tion as granite, except that more of the silica is in combination and less 

 of it free in the former than in the latter. Some early diorites and 

 gabbros have much the same chemical (if not mineralogical) composi- 

 tion as basalt. 



Again, others, with much reason, think that all the differences be- 

 tween the three groups in mineralogical character and crystalline 

 structure are due wdiolly to the different depths at which and the slow- 

 ness with which solidification took place. They think, therefore, that 

 if trachyte (Fig. 180, c) could be traced downward deep enough it would 

 pass into porphyry (b), and finally into granite (g), and similarly basalt 

 would pass into dolerite and diorite, and finally into olivin-diabase-t On 

 this view, what we can not do, has been done for us by erosion ; and 

 granite is most commonly associated with older rocks only because 

 these have been most eroded, and therefore their deeper parts, or even 

 the fountain-reservoirs from which eruptions have come, have been 

 exposed. Similarly, a less extreme erosion of the mesozoic rocks has 

 exposed the porphyritic and dioritic dikes through which eruptions 

 came up ; while, of the modern lavas, only the upper or overflowed 

 parts are exposed. This view explains completely all the phenomena 

 of igneous rocks, and the gradations between them, in chemical and 



* American Journal of Science, vol. vii, p. 167, 1874. 



f This gradual change has very recently been distinctly observed in Southeastern 

 Europe by Judd (Geological Magazine, 1876, vol. xxxii, p. 292), and also in Colorado by 

 Peale (Hayden's Report for 1873, p. 261), and also by Hague and Iddings in Nevada 

 (United States Geological Survey Bulletin, No. 17, 1885). 



