276 HENRY S. WASHINGTON 
line O'N’, yet for any given set of magmas, homologous as regards the 
relative amounts of anorthite and given femic minerals, and variable 
only as regards feldspars and lenads and soda and potash, and conse- 
quently with fixed limits (LQV) for all possible magmas, there will 
be no disturbance of the ratios of the normatively leucitic ones and 
those normatively free from leucite. 
From the above discussion and study of the diagrams, and on the 
assumption that potash has a greater affinity for silica than has soda 
or any other of the bases, we may draw the following conclusions as 
to the relations between the chemical composition of the magma and 
the formation of leucite, which are to be tested later by comparison 
with the data furnished by actual rocks. 
1. There should be no leucitic rocks with an excess of silica over 
that necessary to bring all the bases to their highest silicated condition 
—that is, leucite should not occur in rocks whose norms show the 
presence of normative quartz. 
ta). Asa corollary to this we should never expect to find modally 
leucitic rocks with a silica percentage greater than 64.75, that of 
orthoclase; and as femic minerals and anorthite enter into the 
magma, the silica percentage of possible leucitic rocks should never 
exceed that of a mixture of orthoclase with the appropriate amounts 
of anorthite and femic minerals. 
2. On the other hand, we can expect to find leucitic rocks with 
silica percentages running down to zero as a limiting value. 
3. We should expect to find the greatest range in the percentage 
of potash consistent with the presence of leucite, and consequently 
the greatest number of leucitic rocks, with silica percentages about 
equal to those of mixtures of leucite and the appropriate amounts 
of anorthite and femic minerals. From these maxima the range 
in potash and the number of leucitic rocks should diminish slowly 
as silica falls, and should diminish more rapidly as silica rises. 
4. We should not expect to find rocks without normative leucite 
carrying this mineral in the mode, and conversely we should expect 
rocks with normative leucite to carry this mineral modally. 
4a). As a corollary to this it follows that, if the locus of a rock, 
expressed in terms of the percentages of silica and potash, and plotted 
as on Plate I, falls within the leucitic area corresponding to its char- 
