Editorial Comment. 1 1 7 
In the next place he separates the content of Hme into 
two categories, viz : that concerned in the formation of the 
feldspars, and that which enters into the ferromagnesian bisili- 
cates. This separation, however, is intimately associated with 
the disposition of the alumina present, which goes first to satu- 
rate the alkalies, the remainder entering into the hornblendes, 
micas, etc. 
Thus, the chemical elements, and the composing minerals 
of all crystalline rocks coming from magmas, are separable 
between the two types, the alkaline-earthly type and the fer- 
romagnesian type. The dominant characters in each are 
given leading roles in his scheme of graphic representation. 
In the first he constructs a triangle of potash, soda and lime, 
giving each element a position, according to its amount, on 
perpendicular ordinates and abscissas, coloring it blue in hir. 
plates. The ferromagnesian triangle is composed in a similar 
manner by arbitrarily assigning the leading elements in the 
ferromagnesian group to positive or negative positions on the 
same axes. This triangle is distinguished by being lined in 
black. The exceptional combinations and occasional excesses 
of any of the elements are adjusted among themselves by cer- 
tain convenient compensations or by special representations. 
The triangles thus constituted, placed in juxtaposition, af¥ord 
a fundamental graphic presentation of the composition of sim- 
ilar or of contrasting rocks. The proportion of silica con- 
tained in a rock does not figure in this classification. 
The alkaline elements of the magma as represented, in- 
clude soda, potash and so much of the lime as is recjuired by 
the feldspars. Any excess of soda, as in the soda-bearing 
bisilicates, is specially expressed. 
The ferromagnesian portion of the magma has a triangle 
which expresses the proportion of magnesia, iron and the lime 
which remains after the formation of the feldspars. If there be 
no lime remaining this triangle is reduced to a vertical line. 
Also, in case of an excess of alumina above that required for 
the feldspars, this is assigned to the ferromagnesian triangle 
and may take the place of a lack of lime in restoring the tri- 
angle. 
Nearly all grouped chemical analyses hitherto published of 
igneous rocks that have been considered allied in origin, such 
