134 GEOLOGY OF THE HIGH PLATEAUS. 



assumptions, with a view to ascertaining, as well as we are able, how 

 nearly they approach the truth. 



1st. It is assumed that the density is some approximately definite func- 

 tion of the percentage of silica. There are indeed considerable variations 

 from exactness in this respect, and we may select two or more species of 

 rock having the same silica contents, but which differ conspicuously in 

 density. Yet nothing is more certain than the fact that as a general rule 

 the assumption is very near the truth. This is so well known that further 

 discussion is probably unnecessary. 



2d. It is assumed that melting temperatures also bear an approximately 

 definite ratio to the silica. Here the variations from exactness are no 

 doubt somewhat greater than in the case of density. Still, we know that 

 on the whole the law strongly prevails, and that the melting temperature 

 diminishes with the acidity of the rock.* The blast-furnace slags present 

 often very close approximations to many of the volcanic rocks, and these 

 approximations are not infrequently so close as to be fairly comparable. 

 In such cases it is familiar to those who are acquainted with the practical 

 working of furnaces that the more basic slags are much more easily fused 

 than the more acid ones. The absolute melting temperatures, however, 

 are not accurately known. 



3d. The assumption that the two curves (density and fusion) will ordi- 

 narily cut each other at the ordinate of 60 per cent, of silica is one which 

 presents greater diificulty. Translating grajihical terms into concrete lan- 

 guage, the meaning of it is this : It assumes that rocks having a normal 

 percentage of about 60 per cent, of silica, and corresponding lithologically 

 to the hornblendic propylites are fused and rendered light enough to 

 erupt at one and the same temperature ; while rocks more basic are fused 

 at a lower temperature, but require a higher one to be sufficiently ex- 

 panded ; and rocks more acid are sufficiently expanded at a lower tem- 

 perature, but i-equire a higher one to fuse them. Is there any independent 

 evidence of the verity of this assumption f The point is a very important 

 one ; indeed, vital. For if the intersection of the two curves be elsewhei'e, 



*Sec observations of Bischof on fusion of igneous rock, D'ArcMac, vol. iii, and results of Devillo 

 and Delesse, Bui. Soc. Geol. Franco, 2d scr. iv. D. Forbes Chein. News, xviii. 



