I.] THEORY OF IGNEOUS ROCKS AND VOLCANOES. 3 



to time ejected from below. They are generally regarded as 

 evidences both of the igneous fusion of the interior of our 

 planet, and of a direct communication between the surface and 

 the fluid nucleus, which is supposed to be the source of the 

 various ejected rocks. 



These intrusive masses, however, offer very great diversities 

 in their composition, from the highly silicious and feldspathic 

 granites, eurites, and trachytes, in which lime, magnesia, and 

 iron are present in very small quantities, and in which potash 

 is the predominant alkali, to the denser basic rocks, dolerite, 

 diorite, trap, and basalt ; in these, lime, magnesia, and iron-oxide 

 are abundant, and soda prevails over the potash. To account 

 for these differences in the composition of the injected rocks, 

 Phillips, and after him Durocher, suppose the interior fluid 

 mass to have separated into a denser stratum of the basic sili- 

 cates, upon which a lighter and more silicious portion floats like 

 oil upon water ; and that these two liquids, occasionally more 

 or less modified by a partial crystallization and eliquation, or 

 by a refusion, give rise to the principal varieties of silicious and 

 basic rocks ; while from the mingling of the two zones of liquid 

 matter intermediate rocks are formed. (Phillips's Manual of 

 Geology, p. 556, and Durocher, Annales des Mines, 1857, Vol. 

 I. p. 217.) 



An analogous view was suggested by Bunsen in his researches 

 on the volcanic rocks of Iceland, and extended by Streng to 

 similar rocks in Hungary and Armenia. These investigators 

 suppose the existence beneath the earth's crust of a trachytic 

 and a pyroxenic magma of constant composition, representing 

 respectively the two great divisions of rocks which we have 

 just distinguished ; and have endeavored to calculate from the 

 amount of silica in any intermediate variety, the proportions in 

 which these two magmas must have been mingled to produce 

 it, and consequently the proportions of alumina, lime, magnesia, 

 iron-oxide and alkalies which such a rock may be expected to 

 contain. But the amounts thus calculated, as may be seen 

 from Dr. Streng's results, do not always correspond with the 

 results of analysis. (Streng, Annales de Chimie et de Physique, 



