REPORT OF THE DIRECTOR, 1922 107 
if the chief aim were to divide and subdivide and to discover still 
more intricate or arbitrary bases of further subdivision. Descrip- 
tion has been recognized as a more or less necessary accompaniment, 
but this usually covered chiefly the features useful in the scheme 
of classification. Otherwise it became a more or less rambling 
account of everything that could be seen. 
Thus to be able to describe and classify rocks marked one as an 
accomplished petrographer and to have discovered a slightly differ- 
ent or new mineral proportion or a slightly different chemical pro- 
portion from anything previously described and to be thoughtful 
enough to give it a new name, marked one as an active contributor 
to the science. 
In an effort to reach the last word in this direction of intimate 
discrimination, some of the foremost petrographers of our own day 
have gone so far as wholly to destroy the actual rock by a complete 
chemical analysis before even beginning the task of classification. 
Sometimes by this method mineral constituents were listed in the 
recast of this analysis, and taken into serious account, which never 
occurred at all in the live rock itself, whereas the much more sug- 
gestive constituents and structures that it did have were largely 
disregarded. 
The most serious offender in this respect is the so-called quanti-' 
tative system of rock classification originated by four eminent 
American petrographers. In saying these things there is no wish to 
be understood as attempting to belittle all of this pioneer work. 
Much of it, doubtless, had to be done to lay the foundation of crit- 
ical inspection and discrimination and comparison. The quantita- 
tive principle was a good one to emphasize. The principle itself has 
come to stay; but a classification as mechanical as the quantitative 
system can be, at best, only a side issue in the real petrology that is 
already coming. 
Much as we must value the contributions that have been made by 
the systematic petrographer and the chemical petrographer through 
their nice points of discrimination and their increased appreciation 
of sound chemistry and their rigid requirements of quantitative 
thinking, I am quite convinced that the apparent object in view is 
not a sufficiently high goal for the present stage of development of 
geologic science. This is not the chief or the most promising aim of 
rock study. 
To classify and describe rocks by any system now in use is at best 
only a first step toward a more rational petrology, just as learning 
