36 The American Geologist, January, i905 
unsafe to undertake to do this with the microscope so they 
were himped together under the general head feldspar. The 
different species were separable with Thoulet's solution, so 
that by this means the granite was separated into quartz, 
orthoclase. oligiclase, and iron ores ; while with the micro- 
scope alone but three groups could be made, viz., quartz, 
feldspar, ores. 
(a). METHOD BY CHEMICAL ANALYSIS. 
The chemical composition of the Westerly granite as 
given in professor Kemp's article follows : 
SiO' 73.05 per cent. 
APO' 14.53 per cent. 
Fe-0= 1 
PgQ - 2 . 96 per cent. 
MnO Trace. 
CaO 2 . 06 per cent. 
MgO Trace. 
K^O 5.39 per cent. 
Na^O 1.72 per cent. 
H'O 29 per cent. 
100.00 
Calculation of the standard mineral composition by the 
method of Cross-Iddings-Pirsson-Washington, as set forth 
in their "Quantitative Classification of Igneous Rocks," 
from this analysis, was not possible in all desired detail be- 
cause the two oxides of iron were determined together. Al- 
though the total iron is not high and in this individual in- 
stance would likely make little difference in the results 
whether treated as all Fe'O^ or FeO, the fact that consider- 
able magnetite was present as well as lesser proportions of 
the magnesium-iron-silicates made it seem desirable to have 
a new analysis in which the two oxides of iron would be 
separated. It is also especially important that all the other 
tests be made upon the identical specimen from which the 
sample for chemical analysis is taken. The following analy- 
sis was made by the writer, under the helpful guidance of 
Dr. H. S. Washington, in the chemical laboratory of the 
Department of Geology, Columbia University. The rarer 
elements were not determined as the object of the analysis 
did not require them : 
