150 
IDDINGS. 
which they were erupted. They express only the variation 
from one chemical extreme to the other, considering the 
group as a whole. The introduction of a third variable, if 
it were possible to fix the order of eruption for each rock 
analyzed, would necessitate a third set of coordinates, in¬ 
volving cubical projection, which would not be feasible for 
so complicated a set of lines. 
It is also to be observed that the lavas which constitute 
the rocks whose order of succession is from intermediate to 
extreme varieties are not bodies with a particular or uniform 
composition in any case where their volume is large. They 
are variable within certain limits, and the limits of a body 
of later eruption may overlap those of the previous body to 
a greater or less extent. Thus the statements about to be 
made regarding the changes in the composition of rocks of 
different eruptions must be understood in a general sense. 
The rocks of intermediate composition, which form the 
earliest eruptions in the Yellowstone Park region, have a 
range of small extent, of perhaps 6 per cent, of silica. In 
several localities the oldest rocks contain 61.50 per cent, of 
silica, and possess the other elements in almost the same 
proportions as the average of 64 analyses of the Yellowstone 
Park rocks, as it has been reckoned by Prof. F. W. Clarke * 
in a paper on “ The Relative Abundance of the Chemical 
Elements,” the silica in the average being 61.89 per cent. 
From this mean the successive eruptions through alternating 
and interlocking series pass into basalts with 47 per cent, of 
silica, and into rhyolites with 75 per cent, of silica. 
In particular localities, as at Crandall basin, the whole of 
the great series may not be represented as eruptions from 
one center, but a partial series may occur in which the same 
relations exist with certain modifications. Thus the earliest 
eruptions from the Crandall center were more basic than the 
general average of all of the rocks of the Yellowstone region. 
They probably ranged from 58 to 53 per cent, of silica. As 
* Clarke (F. W.) Bull. Phil. Soc. of Washington. 8°. Washington, 
1892, yoI. 11, pp. 131-142. 
