SYSTEMATIC PETROGRAPHY 361 
In this work Senft gives detailed tables for the determination 
of rocks, the first of which presents the following general scheme 
of classification : 
A. Inorganic rocks (‘‘Anorganolithe’’). 
I. Crystalline. 
I. Simple. 
2. Composite. 
II. Clastic. : 
a. Indurated. 
1. Pseudoclastic. 
2. Hemiclastic. 
3. Holoclastic. 
6, Unconsolidated. 
I. Gravel, sand, etc. 
2 Soll coExdkrumernss): 
B. Organic rocks. 
Curiously enough Senft’s discussion of principles of classifi- 
cation begins with the consideration of the means for subdivid- 
ing the smallest divisions of this table, which he calls ‘‘ Classes.” 
Senft believed that rocks of these classes should be further 
arranged by their characters. Reviewing these, he points out 
the great difficulties in so classifying them, owing to the fine 
grain, variable composition and texture of many kinds. The 
facts lead him to assert that the main systematic divisions, such 
as orders, groups, etc., cannot be founded on structure ( Gefiige ) 
or other outer habit of the rocks, because the former factor 
would at once separate similar things, and the latter bring differ- 
ent things together. Mineral composition would be the best 
basis for arrangement were it not for the fact that, in the fine- 
grained or aphanitic rocks, chemical tests are necessary to 
determine the mineral components. As it is, the chemical rela- 
tion of rocks—and especially of the crystalline—to certain 
solvents affords the only safe means by which rocks can be 
ciassified. This factor of solubility is actually applied to the 
formation of orders, suborders, and groups under the classes 
provided by the scheme already outlined. Senft is, however, 
not fully consistent, for the first division of the class of com- 
posite crystalline rocks is on mineral composition, into the two 
