FORMATION OF LEUCITE IN IGNEOUS ROCKS 363 
locus of a rock will fall inside or outside its corresponding leucitic 
area according as it does or does not show leucite in the norm. 
It will be observed that the great majority of these rocks which 
show both normative and modal leucite are distinctly high in potash, 
for the most part containing over .o50 mol. prop. of K,O, and that 
silica and potash both tend to increase with decrease in the amount 
of femic molecules, the black and the green circles lying above and 
to the left, and the red ones to the right and below; also the range 
in silica is greater with increasing potash. A study of the list of 
analyses will show that, with the exception of a leucite-syenite of 
Magnet Cove and the missourite of the Highwood Mountains, all 
these rocks are effusive flows. 
But there are a considerable number of cases which do not con- 
form to our assumed order and which are to be regarded as con- 
stituting exceptions to the rule, and which therefore demand an expla- 
nation. ‘These exceptions are of two kinds—namely, rocks with nor- 
mative leucite, but free from it in the mode, though falling inside 
their leucitic areas; and rocks with leucite developed modally, though 
free from it normatively, and falling outside their leucitic areas. 
The loci of the first kind of these exceptions, indicated by uncolored 
circles, fall for the most part well toward the less siliceous end of the 
diagram, and most of these rocks are low in potash and belong to the 
salfemane class. The loci of the other kind of exceptions, indicated 
by colored squares, fall mostly well toward the more siliceous end, the 
rocks are often quite high in potash, and they are pretty well distrib- 
uted among the persalane, dosalane, and salfemane classes. 
The cases of the first kind are exceptional, and do not conform to 
the assumed order of affinity, because no leucite has crystallized, 
though it should be present according to our theory; while the cases 
of the second kind are exceptional in that albite has been formed 
rather than nephelite, and leucite rather than orthoclase, at least in 
part, in these soda having an apparently greater affinity for silica 
than has potash. 
But before considering these exceptions it will be pertinent to 
the discussion to assume another order of affinity, and to calculate 
the norms and their loci on the basis that soda has a greater affinity 
for silica than potash has, or, expressed mineralogically, that albite 
