SYSTEMATIC PETROGRAPHY 353 
and much more common than had been supposed. From these 
causes the petrography of igneous rocks grew to be the principal 
part of the science instead of being relegated to an ‘‘appendix,” 
as had been done at an earlier day. 
The chemical analysis of igneous rocks was mainly directed 
to a determination of their mineral constituents and to the ques- 
tion as to whether the various oxides were present in simple 
or constant proportions or not. For the latter inquiry the ratio 
of the oxygen contained in the silica to that of the bases, taken 
together or in groups, was calculated and comparisons instituted. 
G. Bischof set up as a means of comparison the so-called “ oxy- 
gen quotient,’ obtained by dividing the total oxygen of the 
bases by that of the silica, as shown in the analysis of any given 
rock. That the faulty analyses of the time could not in any 
case have yielded trustworthy evidence of stoichiometric or 
constant proportions of the constituents is now clear; but even 
with the imperfect analyses the complexities of the problem 
were such as to lead nearly all chemical students of rocks 
to abandon attempts to deduce chemical formule or simple 
ratios for them. 
For many years chemical investigations were necessary in 
determining the approximate mineral composition of many rocks, 
and important discoveries were undoubtedly made; but it is 
pathetic to recall the years spent in hard labor by many of the 
foremost men of their time in endeavoring to work out the 
mineral composition of fine-grained or aphanitic rocks from 
incorrect or inadequate bulk analyses, and with very imperfect 
knowledge of the constitution of some rock-making minerals. 
Abich, Bunsen, von Waltershausen, Bischof, Scheerer, Roth, 
Streng, Delesse, Haughton, and many others devoted much 
time to this research without what appears to us, at this time, to 
have been commensurate results. 
Classification by feldspars—The principal effect of this chem- 
ical work upon petrographic classification appears to have been 
to perpetuate the mistake of giving an undue importance to the 
feldspars as rock constituents. As different kinds of feldspar 
