8 RICHTHOFEN — THE NATURAL SYSTEM 
not of one only but of various principles, compares and weighs the results obtained by 
each of them, and accepts them as final only when perfectly harmonizing among each 
other. Itis then that it tries to determine what principles are most available for 
establishing the higher orders, and which for the subdivisions. The singular complica- 
tion which is peculiar to the classification of rocks, is, besides the reasons already 
mentioned, due in a great measure to the fact that geology combines the double func- 
tions of a historical and an inductive science, while in petrology we have besides the 
requirements of a descriptive natural science. The natural system of rocks should 
therefore be based, not only upon the entire range of their petrographical characters, 
such as mineral composition, chemical composition, texture, and specific gravity, but 
also upon their mode of origin and geological occurrence. Classification of objects 
and classification of relations are, with them, closely connected, and should be made 
to assist each other. 
The question may be raised, whether a natural system of rocks based upon 
such principles can be established at all, and if it can, whether it would be of any use 
for the advancement of science. To the first question, the answer must be in the 
negative, as far as sedimentary rocks are concerned. They have been formed by a 
complexity of circumstances, and just so complex and infinite in variety are they, in 
respect to chemical and mineral composition and all external characters. To analyze 
in detail their mode of origin, and the sources from which their material has been 
derived, transcends the faculty of human intellect, and it would be a hopeless task to 
attempt to discover any laws regulating the boundless differences of their composition. 
They are thus debarred from natural classification, though its principles may be applied 
imperfectly to the establishment of some general groups. We arrive at similar con- 
clusions in regard to those rocks, the sedimentary origin and subsequent metamorphism 
of which can be proved. Accidental and local circumstances have played as conspic- 
uous a part in their first deposition as was the case in regard to those sedimentary 
rocks to which the term metamorphic has not been usually applied. But as meta- 
morphie processes of a certain nature have ordinarily affected extensive tracts of 
these rocks, and similarly pervaded great thicknesses of them, the local differences of 
their action having been apparently more in degree than in mode, they have occa- 
sioned a certain similarity of effect which partly conceals the original differences in 
the composition of the rocks affected ; and it appears that the differences in the kind 
and intensity of metamorphic action, though recognizable only in their final results, 
will, when better known, afford a convenient principle for a classification which may 
have some similarity with, but not the full requirements of, the natural system. It is 
different with those rocks which on the surface of the globe appear as intrusive or 
eruptive masses. Notwithstanding their infinite variety in character and composition, 
they are connected by definite relations which bring their elementary composition 
even within range of mathematical calculation. Their recurrence in the most widely 
separated countries, with similar external character, identical chemical composition, 
and in analogous relative order of succession, is another distinguishing feature of 
eruptive rocks. For these reasons, as well as in virtue of other peculiar characters 
(46) 
