220 The American Geologist. April, i90(i 
these distinguishing characteristics shall strictly and thor- 
oughly agree, it is not possible, in systematic classification, to 
give equal consideration to all seven of these principles or 
even to any two of them. It has been found quite untrue that 
a given mineralogic combination, for example, has either a 
constant chemical composition, a constant mode of occurrence, 
a constant origin, or any of the rest. It is, therefore, neces- 
sary to decide upon some single characteristic upon which the 
classification shall be actually primarily based, and this prin- 
ciple must be the most important. 
In the earliest morning of our present geologic science, 
when all that was actually knowm of the igneous rocks could 
be put into a page of print, these rocks naturally came to be 
considered as separated chiefly according to their manner of 
occurrence, and the present consideration which the manner 
of occurrence of rocks has as a prime element in their classi- 
fication, seems to be a direct inheritance from these earliest 
times. During the times of Werner and Hutton some vol- 
canic rocks were, of course, recognized as igneous, but granite 
or plutonic rocks in general were considered always either 
chemical precipitates or metamorphosed sediments until Hut- 
ton proved their igneous origin. The conception of the grand 
division of igneous rocks into plutonic and volcanic, with the 
intermediate occurrence as dikes or veins, was, then, fonnu- 
lated in the mind of Hutton before the actual study of the 
igneous rocks had begun and wdien the application of miner- 
alogic and chemical criteria to rock subdivision was practically 
unknown. 
Later, in the swift development of geologic science (Hutton 
proved the igneous origin of granite in 1785), when fossils 
had been proven to be organic remains and had begun to be 
used as the basis of the geologic system, the element of geo- 
logic time became naturally of great importance in the minds 
of men as regards rock classification, and it was natural to at- 
tempt to classify the crystalline rocks also in this manner. At 
so early a period, when it was for a time generally believed that 
the age of sedimentary rocks could be divined from their tex- 
ture and amount or metamorphism, it was natural to suppose 
that the age of igneous rocks might be determined by similar 
criteria. We are not far awav as vet from the shadow of this 
