188 
IDDINGS. 
would urge it upward more freely and thus cause a separa¬ 
tion. The more important feature of his conception, how¬ 
ever, is the idea that the components of a molten magma do 
not necessarily represent fixed mineral molecules, and that 
the magnesia and silica, and consequently the other oxides, 
are capable of entering various compound molecules, accord¬ 
ing to conditions of temperature. 
A wholly novel theory of the origin of the variability of 
volcanic rocks was conceived by Bunsen and published two 
years after that of Dana. It was one involving the synthesis 
of two distinct magmas, and, as Sartorius von Waltershausen 
has remarked, was purely imaginary and unnatural. It 
emphasized the differences and not the resemblances of 
rocks, and has maintained a strong hold on the minds of 
those who have been most impressed with the distinguish¬ 
ing characteristics of volcanic rocks. 
The idea of the essential nature of the differences among 
igneous rocks took a more rational and permanent shape in 
the theory of Von Waltershausen, who imagined the mate¬ 
rial of the earth to be arranged in zones of increasing specific 
gravity from the surface inward, and that the eruptive rocks 
varied in composition with the geological age of their erup¬ 
tion ; so that the more silicious and specifically lighter ones 
were drawn from nearer the surface than the more basic 
rocks, and were erupted at a time when the crust of the 
earth was not as thick as it became subsequently by the 
gradual refrigeration of the globe. 
The transitions in composition between different kinds of 
rock were due to the gradations of the material of the zones, 
which rested on the supposed primary arrangement or dif¬ 
ferentiation of the original fluid magma of the earth by a 
process of gravity separation. This is essentially different 
from Bunsen’s synthetical theory, though it is frequently 
alluded to as a modification of the latter. It was based on 
a primary differentiation and has no element of synthesis 
in it. It was in accord with a widely accepted view of the 
order of eruption of igneous rocks, namely, that the compo- 
