TIE CLASSIFICATION (OF THE IGNEOUS ROCKS Tel 
in molecular ratios, ferrous and ferric iron being entered upon the 
same radius vector, and silica, because so much in excess of 
the others, being evenly divided between the two horizontal 
radius vectors. A broken line joining the intercepts on the 
eight radius vectors forms a polygon, which may be long and 
narrow, or short and thick, convex above or below, or reéntrant 
in any portion, left or right handed, etc., according to the chem- 
ical constitution of the rock. 
2) 
% 
g 
SiQ 
Wie, ip 
When viewed in this diagram, the rock comes to have a 
handwriting by which it may be instantly recognized. When 
drawn to scale, this diagram not only shows the chemical char- 
acter of the rock but all the results of analysis. may be quickly 
read from it numerically.* In it, as in all other successful 
diagrams the molecular ratio is substituted for the percentage of 
each constituent. Some authors now publish these ratios with 
every rockanalysis. It is to be hoped that this will soon become 
a general custom. _ 
THE COMPOSITE ROCK DIAGRAM 
The principal objection to Brégger’s diagram is that it rep- 
resents not a rock species or a rock type but only an individual 
analysis, the rock type covering a considerable range of differ- 
ing analyses. So far as I know, only isolated attempts have 
been made to average rock analyses to secure an adequate con- 
ception of the chemical constitution of the rock type, although 
the method of averaging results is so successfully used in other 
fields of science. 
*On plates I, IV, V, and VI, .or equals 1™™. 
