12 SHALER ON THE CLASSIFICATION OF LAVAS 



come from depths below the levels to which geological sections ever give us access. Nor 

 do I mean to question the fact that the present superficial character of many of our traps 

 depends upon comparatively recent alterations of their constitution by the agents of 

 decomposition; allowing for all these perplexing accidents, it seems still that the diversity 

 in the lavas of different geological districts, and indeed their diversity in the same district, 

 can only be explained by the hypothesis of their local and comparatively superficial origin. 



The second part of the last proposition rests upon the well known chemical composition 

 of the lavas, and also upon the fact that the highly metamorphosed rocks seem in all 

 cases to show more complete alteration of their very siliceous elements than of the 

 refractory clays and limestones. It is fairly a matter of surprise when we consider how 

 large an element the sandstones form in all our metamorphosed sections, that we so rarely 

 find extensive beds of rocks of this nature in our deposits which bear the marks of 

 high temperature. I am inclined to think that it is reasonable to suppose that the 

 deposits which were originally very siliceous have been forced out by the expansion of 

 their gases as lavas, and now appear in the neighboring dykes, or have been driven to the 

 surface as true volcanic ejections and have since been lost by erosion. 



The depth of the overlying rocks which is required before the rents in the rocks can be 

 filled by dykes is a matter that cannot be determined. It is evident that it must be great, 

 for as in parts of Eastern Tennessee it is clear that the surface has lost a section of at least 

 ten thousand feet in thickness, since the close of the Carboniferous, owing to the peculiar 

 intensity of the denudation there ; yet there is no trace of dykes in the district, though 

 there are many veins. 



3. Dykes formed from the melting of particular parts of a section, may penetrate 

 either downward or upward from the point of origin. 



The opening of fissures in metamorphic rocks brought about by contraction or other 

 strains, will necessarily lead to the frequent downward extension of dj^kes. If the section 

 contains extensive beds of fusible materials on one level, and infusible materials above and 

 below, the result will necessarily be the penetration of the lava into any fissures that 

 may be contiguous with it, whether these fissures extend downwards or upwards from the 

 fusinc bed. This suggestion may be of no great importance, yet I have seen places where 

 it would aid in the understanding of the distribution of certain dyke stones. 



4. The pressure of the superincumbent unmelted beds cooperates with the pressure of 

 the gases contained in the fusible bed, to impel the molten or semi-molten matter into 

 the dyke fissures ; and also to a certain extent the friction thus brought about aids 

 in the fusion of the lavas. 



The pressure of a vertical section having a depth of twenty thousand feet, amounts 

 to about 4,000,000 pounds per square foot. This pressure will doubtless be sufficient to 

 aid the flow of any of the mass-rocks which are somewhat softened by heat, and the con- 

 tained water. As soon as the movement begins, the heat derived from friction will greatly 

 aid in the fusing of the lava. In many of our ancient dykes the superincumbent deposits 

 have doubtless much exceeded twenty thousand feet in thickness, so that the pressure may 

 have much exceeded the 2,000 tons to the square foot, which would be given by five miles 



