THE ORIGIN OF IGNEOUS ROCKS. 
Ill 
Since this separation would be an affair of some time and 
the causes which determined eruption might supervene when 
crystallization had begun and before specific gravity separa¬ 
tion had completed its work, it would be natural to expect 
that eruption would frequently occur before the complete 
genesis of species. The latest lake of fusion after gravity 
separation would result in a layer of rhyolite floating upon 
a layer of basalt. 
“ Wherever a molten lake should be formed within the 
acidic shells of the earth, after separation by specific gravity, 
the relative proportions would show a great preponderance 
of the acidic member.” 
Refrigeration may advance so far before an eruption takes 
place that the upper layers of the lake may become solidified 
and only the lower portions be erupted. 
Rocks of average composition may either result from the 
eruption of the fused lake before separation commences, or 
they may be the intermediate zone between the upper and 
lower layers. 
In the secular refrigeration of the globe these temporary 
lakes of fusion would necessarily occur at greater and greater 
successive depths. The deepest of all would be the latest 
(neolite) lake, or that which produces rhyolite and basalt. 
King recognizes the difficulty of explaining the presence 
of mica and hornblende among the lighter minerals of the 
acidic rocks, since their specific gravity should have carried 
them down into the lower portions of the lake. His expla¬ 
nation, however, is quite insufficient. 
From the foregoing it would appear that the five kinds of 
volcanic rocks distinguished by Yon Richthofen, namely, 
propylite, andesite, trachyte, rhyolite, and basalt, were each 
the product of a single ephemeral lake of fusion, except 
that the last two were derived from one source. Each 
would require for its production a special act of erosion, of 
cessation of erosion, and a crystallizing process before erup¬ 
tion. 
