ORIGIN OF THE GLASSES AND MINERALS. 79 



thinking it in great part formed during the consolidation and cooling of 

 the glass. It is consonant with this that the feldspars formed during 

 this cooling, especially those in the spherulites, are quite acid, while 

 Hawes found very basic feldspars an abundant constituent of normal trap. 



Boiled with strong hydrochloric acid and treated with fuchsin, there 

 is no trace of decomposition and the optical characters indicate a very 

 acid feldspar. The C0 2 brought into the mass by the waters from the 

 coal-bearing sandstones below may have taken possession of a large 

 portion of the Ca, leaving the Na to go into the newly made feldspar. 



The similarity of these aqueous feldspars to those in a metamorphic 

 schist is remarkable, and it is interesting to find diopside and the blue 

 segerine-augite and hematite formed with it, thus making a very peculiar 

 crystalline schist in a very peculiar position. 



It is again remarkable that delessite and its serpentinous decomposi- 

 tion product is rare in these glasses and the associated traps. This mili- 

 tates against the idea that the penetration of the ground waters into the 

 liquid trap is the cause of its chloritization. 



Resume. 



The normal tuff formed of water-borne trap fragments, cemented when 

 cold by rust, is contrasted with rocks made in the upper or lower parts 

 of submarine lava flows by the intimate commingling of mud with the 

 liquid lava. This takes place by two quite distinct processes. 



The fine mud is carried out over the surface of the advancing sheet 

 and sinks and becomes variously blended with the liquid lava for a depth 

 of 10 or 20 feet. This takes place over an area of the surface which, 

 measured about parallel to the advancing front of the sheet, exceeds 10 

 miles, and measured at right angles to the front, about half a mile. 



Because so large an amount of mud was carried out over the sheet 

 while it remained so liquid that the two could blend, it is assumed that 

 convection currents set up in the water by the heat of the trap, acting as 

 indraughts from the adjoining muddy bottom to the lava sheet, spread 

 the mud more rapidly than ordinary currents would have done. 



As exactly the same mud impregnation occurs at the base of the sheet, 

 even where the sheet rests on coarse gravel (arkose), and as it is confined 

 to the same limits, it is assumed that the superficial blended layer was 

 carried forward and down the front of the moving sheet, and so under- 

 rolled upon the gravels on which the sheet had advanced. 



As the mud is mingled with the surface layer of the trap when cooled 

 nearly to its solidifying point, only slight chemical changes occur. This, 

 and the fact that the formation occurs at the top and bottom of the bed, 

 distinguishes this process from the other. 



