EDITORIAL 187 
be overestimated. They not only fix in an easily comprehended 
form facts already vaguely apprehended, but often suggest rela- 
tionships not previously suspected. With all machines the 
product turned out depends on the material operated on. And 
while the machine itself may be perfect, the product may be 
open to criticism. 
The diagrams in question tend to give more definite impres- 
sions of the relative quantities of the chemical elements in rocks 
than are obtained from the usual statements of analyses. But, 
if instead of actual rock compositions there is substituted an 
average of various rocks, it is clear that there is danger of 
placing too much value on the apparently definite expression 
conveyed by the composite diagram. Everything depends upon 
what rocks have been grouped together. Defects in grouping 
vitiate the diagram. For this reason it is desirable to distinguish 
between the use of graphical methods of presenting an assem- 
blage of diverse quantities, which is highly commendable, and 
the practice of averaging diverse quantities, which is open to 
serious criticism. 
Joe Pails 
