3152 WHITMAN CROSS 
von Leonhard advanced structure also to a prominent place. 
Both declared that the system and nomenclature of rocks should 
be founded upon characteristics, and that the various treatments 
rocks must further receive from the geologist would be facilita- 
ted rather than retarded by such an arrangement. Brongniart 
made a suggestion, apparently unconsciously, in subdividing 
Heterogeneous rocks, that geological conditions of origin might 
serve as a practical basis in classification; but he made only lim- 
ited application of that factor, and does not appear to have 
grasped the broader meaning of his own proposition, 
After von Leonhard and Brongniart had placed petrography 
upon a firm basis, no great advance was made in the systematic 
part of the science until the middle of the century. The cause 
of the unsatisfactory elements in the early schemes was recog- 
nized to lie in the ignorance concerning the character of many 
rocks, and a large number of investigators devoted themselves 
to the study of the composition of rocks onthe one hand and to 
their geological relations on the other. While the object of this 
review is to trace the application of principles or of knowledge 
in classification rather than to follow the course of investigation, 
it is necessary to refer briefly to the studies made during the 
second quarter of the century, noting the use of new knowledge 
in system. 
Investigations into the chemical composition of rocks — Among 
the characters of rocks not taken into appropriate account in 
the first systems, by far the most important is chemical compo- 
sition. But from the beginnings under Cordier more and more 
attention was paid to chemical investigation, until other things 
became quite subordinate. First, qualitative test was applied, 
then partial analysis, and finally bulk analysis. Chemical 
research was naturally applied with most valuable results to 
the denser rocks, chiefly of igneous origin, and in this way, with 
ever-extending field investigation, it was found that these sub- 
stances were vastly more complex, more varied in character, 
