110 
IDDINGS. 
manence of these shells, or the eruption of pyroxenic material 
upward through the silicious shell, involves physical diffi¬ 
culties which appear to be altogether insuperable. 
“ If, however, by any means, temporary local lakes were 
formed, resulting from the fusion of a thin shell of the crust, 
it w r ould seem that the arrangement into two zones—a lighter 
overlying a heavier one—would, from the nature of things, 
gradually assert itself within the limits of the enclosed fused 
region.” 
Assuming that his order of succession for each period of 
eruption is correct, namely, that the pyroxenic modification 
is always the later, the acidic being the first, he concludes 
“ that wherever fusion is developed on a considerable scale, 
by whatever means, the fused material divides itself into two 
parts—the acidic or lighter coming to the surface before the 
basic and heavier. 
“ There are two methods by which the separation within 
the limits of a fused lake might be made: First, while in a 
state of fusion, on well-understood principles, the heavier 
liquid might concentrate at the bottom of the lake, leaving 
a supernatant couche of lighter matter; or, secondly, in the 
act of crystallization, which all present facts tend to prove is 
a subterraneous process, the actually formed crystals might 
separate themselves according to their differences of specific 
gravity; ” an idea which he credits to Scrope. 
King’s theory is briefly as follows : Each one of Richt¬ 
hofen’s orders, with its acidic and pyroxenic members, would 
be the product of a single ephemeral lake. A period of 
erosion would result in the formation of a lake. The cessa¬ 
tion of erosion would define the time limits and the vertical 
expansion of the lake. Refrigeration, continuing from that 
time, would result in the crystallization of the various mineral 
species. There is sufficient difference in the specific gravity 
of the minerals of the acid and basic rocks to account for the 
separation. The magma through which they moved is 
partly the isotropic glass of the groundmass, partly the 
groundmass itself. 
