380 HENRY S. WASHINGTON 
formation of any mineral in igneous rocks must partake more or less 
of the nature of begging the question, in the absence of properly con- 
trolled physico-chemical experimental data. Any conclusions based 
on our present knowledge of the physico-chemical relations of minerals 
must rest on very insecure bases, and it were better not to attempt to 
draw them. 
A few words may be devoted to the subsidiary point of the relative 
abundance of leucitic rocks. We have seen above (p. 271) that the 
size of the leucitic area is much smaller than those of the non-leucitic 
ones, the relations varying with the content in anorthite and femic 
molecules, but with a maximum of 21.61 per cent. of the whole in 
peralkalic and persalic magmas. It follows from this that the num- 
ber of leucitic rocks actually to be observed should be considerably 
L' 
1davey, ae 
less than the non-leucitic ones, though it does not follow that the 
actual relations must be exactly the theoretical ones. This would 
depend upon the general and the average compositions of all igneous 
rocks and the correspondence of these with the theoretical area OLM. 
As a matter of fact, we find that these do not correspond, the center 
of gravity of OLM falling at the point X (on the line O*N"), and the 
locus of the average rock well to the left of and below this, as shown 
in the adjoining figure. That is, igneous rocks, as they actually 
occur, are higher in silica and lower in potash than our theoretical 
magmas, of which amounts proportional to the theoretical limits of 
each are represented in our diagrams. It follows from this that 
leucitic rocks should actually be much more rarely met with than 
is indicated by the relative sizes of the various theoretical areas. 
That this is actually the case is shown by the following figures. 
