226 WALKS AND TALKS. 



clay; and the silica would, in a similar way, mix with other 

 substances or form by itself beds of pure quartz. 



If these views are correct, it appears that the mineral con- 

 stituents of sea-water date from the primeval time. The table 

 salt so familiar to our senses is a venerable product. The 

 sodium had endured the heat of the molten globe and consoli- 

 dated as a silicate in the first crust. The chlorine had drifted 

 about, a constituent of the heterogeneous atmosphere, until 

 the war of the elements came ; when it was dragged down as 

 an incident of the conflict. Here it found occasion for un- 

 wonted activity ; and all over the world, chlorine atoms busied 

 themselves in soliciting sodium atoms, to gratify a newly dis- 

 covered liking, and create, in unlimited abundance, a sub- 

 stance destined to acquire unlimited usefulness. So our 

 common salt, which knows so well how to grow old unchanged, 

 is employed by us to assist other substances to grow old 

 unchanged. 



We may inspect, now, the oldest rocks accessible to inves- 

 tigation, and see if they are the kinds which should be 

 expected, according to the above reasoning. Yes, in a general 

 way they are. Feldspars are everywhere disseminated in the 

 bottom rocks. There are great beds of crystalline limestone, 

 and of argillites. There are micas, hornblende, and augite, 

 all of which, like the feldspars, are essentially silicates of 

 alumina and other bases. But here, also, are conglomerates, 

 composed of rounded or angular fragments of older rocks. 

 Here, indeed, are vast formations of quartzites j but they are 

 formed of grains, instead of being those purely vitreous masses 

 which would result from the precipitation of free silica. But 

 each separate grain is such vitreous quartz, and it appears MS 

 a fragment of some rock in which the whole mass was purely 

 vitreous. At some time then, earlier than the formation of 

 these granular quartzites there were formed quartzites not com- 

 posed of grains, but continuous like a mass of glass. We 

 have to conclude that the oldest rocks accessible are <>( rein- 

 Dan ts of the primeval precipitates. This is the conclusion 

 pointed out also by the presence of conglomerates. 



