THE ORIGIN OF IGNEOUS ROCKS. 107 
The application of his views as to the causes of ejection 
and the crystallization of rock at depths in consequence of 
the relief of pressure and of Von Waltershausen’s law is 
that in the earliest periods the eruptive rocks must have 
been mostly silicious. The beginning of the volcanic period 
(the Tertiary) opens with basic rocks and is followed by still 
more basic ones. 
As an exception to this rule he notices trachyte and rhy¬ 
olite, but considers them as “ a notable but only apparent 
anomaly,” more closely related to andesites and only very 
remotely related to basalt. They are due to secondary pro¬ 
cesses, and are subordinate in regard to quantity. 
The order in which the volcanic rocks were erupted is, 
first, propylite, then andesite , trachyte, rhyolite, and basalt . 
He considers volcanic action to be closely related to 
massive eruption, but not so deeply seated, and remarks 
that two neighboring volcanoes may be erupting different 
kinds of lava and be independent of each other. Volcanic 
action is the last stage of massive eruption, but the order of 
succession is the same as with massive eruptions. 
Von Richthofen advocates Von Waltershausen’s theory, 
assumes that Bunsen’s law is of universal application, and 
reiterates the old notion that the earliest eruptive rocks were 
the most highly silicious, and that they have been getting 
more and more basic to the present. His views of the action 
of volcanoes agree with those of Durocher, that they are de¬ 
rived from localized reservoirs that have branched off from 
the main supply. 
In 1872 Scrope, in the second edition of his “ Volcanos,” 
advocates the theory that the separation of minerals of a 
molten lava within the focus of a volcano may be brought 
about by their differences in specific gravity, the feldspars 
rising and the augites sinking. He remarks that Darwin 
was of the same opinion, hnd refers to his book on the Vol¬ 
canic Islands, which has already been cited. Scrope adds 
that “ a somewhat similar effect may be occasioned by the 
mechanical squeezing out of the finer particles, or the more 
