1907—8.] 
Classification of Igneous Rocks. 
85 
VII. — Classification of Igneous Rocks according to their Chemical 
Composition. By H. Warth, D.Sc., Tubingen, late Dep.-Super- 
intendent, Geological Survey of India. Communicated by Professor 
James Geikie, F.R.S. 
(MS. received June 22, 1907. Read July 15* 1907.) 
In my previous paper (see Geological Magazine, March 1906, p. 131) I 
explained a method of separating igneous rocks into sixteen groups of 
chemically similar individual rocks. A more minute subdivision appears, 
however, desirable for the special purpose of identifying and describing 
particular rocks. I have therefore devised the following further classifica- 
tion with four times as many groups, and I trust that the proximity thus 
attained between the individual rocks of each group will amply meet all 
requirements. 
The method is still founded upon the fact previously pointed out by 
Harker,* that any large number of igneous rocks, selected at random from 
those commonly examined and analysed, yields approximately the same 
average composition. The average composition of 1000 rocks utilised for 
the present purpose (see page 150) is practically identical with that of 500 
rocks used in the former case. The mean composition of the rocks 
remaining practically the same, we may feel confident that the classifica- 
tion which answers for 1000 rocks will also answer for any additional 
specimens, and will in fact be generally applicable. 
We will now first describe the method of classification. It will be 
noticed that in the tables of composition the second place of decimals has 
been omitted, and that not more than seven basic oxides have been recorded. 
The rarer bases such as MnO, BaO, SrO were included respectively under 
FeO and CaO. The following quantities of oxides derived from the 
general average composition are added together, 
(1) Al 2 Og + (Fe 2 0 3 + FeO) + MgO + CaO, 
the result being found =33*7. We now separate all the 1000 rocks into 
two sets, those which yield individually for the same bases a sum smaller 
than 33'7, &nd those which yield a sum larger than 33*7. All the former 
* A. Harker, Geological Magazine, May 1899. 
