FORMATION OF LEUCITE IN IGNEOUS ROCKS 273 
thermore, the areas are shifted successively toward the more siliceous 
end of the diagram, owing to the higher percentage of the pyroxene 
in silica than in nephelite, It will also be noted that the influence of 
the introduction of anorthite is exactly similar in character to that 
already noted in the case of the persalanes, though the change is less 
quantitatively, as shown by the lines L4N4 and L5N5, and O4N4 
and O5N5, being closer together than was the case with the homolo- 
gous lines in the persalanes. But it seems needless to go into greater 
details, as the principles will be the same as with the preceding cases. 
We may next assume that the femic molecule is an olivine, exactly 
intermediate between forsterite and fayalite—that is, with MgO= 
FeO. The molecular weight of this is 172, and the silica percentage 
is 35.88. As in the preceding case, we shall consider both dosalic 
and salfemic magmas, and in the former shall assume the magmas to 
be peralkalic and domalkalic, while in the salfemanes only peralkalic 
ones will be considered. Pyroxene is assumed to be absent in all 
these cases. 
The leucitic areas of these various olivinic magmas are shown in 
Plate II, the peralkalic and domalkalic dosalanes respectively by 
green lines made up of very long dashes and of the same with alter- 
nating short ones, while the salfemic magmas are represented by red 
lines made up of very long dashes. Here again the triangular leucitic 
areas are of exactly the same form and parallel to the homologous 
ones in the persalanes and the pyroxenic magmas, and, furthermore, 
the sizes of the olivinic areas are the same as those of the corres- 
ponding pyroxenic ones. Thus, L7O7N7is the same size as L4O4N4, 
and L9O2N° as L°O°N®S, and so on. This, of course, is due to the 
fact that the relative amounts of potash in homologous magmas 
remains the same whether olivine or pyroxene is introduced. But it 
will be seen that the introduction of olivine, rather than pyroxene, 
shifts all these leucitic areas well toward the right, or less siliceous, 
end of the diagram, a consequence of the lower silica percentage of 
the olivine. Furthermore, as in the preceding cases, increase in the 
- amount of olivine lessens the size of the triangle until, for perfemic 
magmas, they would vanish at the point Ol, where SiO, =o0.598 in 
the line QN, which is the locus of our assumed olivine, and which is 
also the intersection of lines passing through L, L7, and L®, and 
