364 HENRY S. WASHINGTON 
will form rather than nephelite, and leucite rather than orthoclase, in 
which respect the cases of the second kind are peculiar. For this 
purpose we need only consider the simplest case, that of a persalic 
and peralkalic magma, without excess of silica and with alumina 
equal to the total alkalies, as formerly. 
It is clear that on this assumption the line representing purely 
feldspathic norms, composed only of orthoclase and albite, will be 
identical with that calculated on the previous assumption. It will 
be the line O'Aton Plate II (p. 270). Also the line representing purely 
lenic norms, composed only of leucite and nephelite, will coincide with 
L*N*. Similarly, if the magmas are perpotassic (that is, quite free 
from soda), the upper limiting line will coincide with the previous 
Graves ILAOMOLS 
The divergence from the previous assumption will be expressed by 
the line representing norms composed only of albite and leucite, 
along which these are in equilibrium, so that magmas with less silica 
for any given percentage of potash will yield norms composed of 
albite, nephelite, and leucite, while those with more silica will yield 
norms composed of orthoclase, albite, and leucite. This is the 
dotted black line LtA* of Plate II. The orthoclase-albite-leucite 
area L*O*tA? is comparatively small, and with a range in silica only 
from that of albite to that of leucite, while the albite-leucite-nephelite 
area L'A'N? is much larger and with a range in silica from that of 
albite to that of nephelite. 
On this assumption, therefore, we should expect to find very many 
more rocks with leucite than rocks with orthoclase, and orthoclase 
rocks should be relatively less abundant on this assumption than 
should leucite rocks be on the other. Also we could expect to find 
leucite rocks with silica percentages up to 68.70, and the range in 
silica of leucitic rocks should increase with decreasing potash. It need 
scarcely be pointed out that the facts of petrography are directly 
contrary to these deductions. Furthermore, not a single leucite 
rock falls within the area L'O'A",? and the position of the line L'N’, 
and the shape and position of the areas L'O'At and L'A'N’, are 
wholly discordant with the positions of the leucitic rocks as actually 
observed and plotted. Similar discordances would be observed were 
' The apparent exception seen in the diagram will be discussed later. 
