360 HENRY S. WASHINGTON 
fall within the leucitic areas corresponding to their content in molecules 
of anorthite and femic minerals. This is the crucial test of the 
validity of our assumed order of affinity for silica; for it follows from 
the preceding discussion that for any given set of magmas containing 
a definite amount of anorthite and femic molecules in the norm 
(however these may be crystallized in the mode), but with varying 
amounts of silica and potash the loci of normatively leucitic rocks 
(which should also be modally leucitic) derived from them should 
fall within the corresponding leucitic area; and those free from norma- 
tive leucite (which should also be free from this in the mode) should 
fall outside of this, either to the right or to the left. 
We have seen that the introduction of salic lime into the magma 
causes only a slight change in the positions of the leucite-nephelite 
and orthoclase-nephelite boundaries of the leucitic area, though it 
has a much greater effect on the leucite-orthoclase boundary above, 
lowering this very decidedly. The effect of the introduction of the 
constituents of the femic minerals, on the other hand, is much more 
marked and varied in its character. The leucitic area shifts some- 
what to the left if pyroxene enters into the norm,’ while with olivine, 
and with magnetite and other minerals free from silica, the shift is 
still greater and toward the right or less siliceous end. At the same 
time, the upper boundary is lowered to an extent even greater than 
is caused by salic lime. It is to be remembered, however, that, what- 
ever the constituents thus introduced or assumed to be present, the 
angle LNO and the inclination of the leucitic area to the horizontal 
line QAMN do not vary, the lines LN and OWN always being parallel. 
A few simple cases have been plotted on Plate II, but it is obvious 
that, for purposes of adequate comparison of actual rocks with their 
theoretical leucitic areas, some means must be had of determining 
the true position of the leucitic area corresponding to any given rock, 
whatever be its composition and however complex it may be as 
regards its mode or norm. Fortunately this determination is not 
difficult, nor does it involve an inordinate amount of calculation. 
The method consists, in brief, of the calculation of the locus 
(expressed in terms of the percentages of silica and of potash) of one 
t This shift toward higher silica would be still greater if aegirite enters into the 
problem, as it not infrequently does. 
