252 HAROLD L. ALLING 
the extinction angles of the plagioclase feldspars does not involve 
such a large error in the determination of the composition as it 
does in the case of the potash-soda feldspars. This is, perhaps, 
the reason that petrographers have thought it was impossible to 
determine, microscopically, the chemical composition of the latter 
series. 
Specific Gravities: We have seen that the specific-gravity 
curves for the potash-soda series probably are represented by two 
non-parallel lines; the upper lines representing the density of the 
monoclinic modifications and the lower those of the triclinic forms. 
The specific gravities of the plagioclase feldspars, Ab-An, can 
properly be represented by a straight line inclined upward from 
albite to anorthite. ‘Taking the curves for these binary systems 
it has been possible to construct the diagrams 6 and 7 in Figure 12 
for the ternary system. ‘They are instructive, even though they 
lack confirmation on the potash-lime side. They also illustrate 
the difficulty in the determination of the potash-soda series in that 
only a few lines intersect the potash-soda binary side, while a 
greater number cut the plagioclase side. 
The diagrams indicate an inclined flat surface of the space 
model which they represent. One diagram is for the orthoclase- 
barbierite-anorthite system; the other, the microcline-albite- 
anorthite feldspars. Two others might be drawn: orthoclase- 
albite-anorthite and microcline-barbierite-anorthite. 
, CLASSIFICATION 
It is important that we attempt to sum up the ternary system, 
potash-soda-lime feldspars, by offering a reasonable classification 
of the same. . To accomplish this the writer has had to venture 
upon untrodden ground and therefore realizes his limitations. 
To overcome some of the objections that may be raised against it 
on the ground that it is too comprehensive or too complicated, the 
scheme is offered in two forms: one which may be called the tech- 
nical classification and the other the popular one. They are 
shown in Figure 13. 
In proposing these classifications absolutely new names have 
been avoided so far as possible. For the row of areas beginning 
with potash oligoclase and ending with potash bytownite the 
