222 The Ametican Geologist. April, 1900 
entirely upon chemistry and indeed upon a very few principles 
of that science, the classification depending entirely upon the 
relative amounts of lime, soda, and potassium in a rock. As 
an example of other chemical schemes of classification the 
recent one of Lewinson-Lessing* may be mentioned. F. H. 
Hatch J offers a classification which is based (i) on chemistry, 
(2) mode of occurrence. Among the chemical considerations 
the relative proportion of silica is the only point that is con- 
sidered and rocks are divided into five groups on this ground. 
Under the mode of occurrence there are three divisions, cor- 
responding essentially to the plutonic, dike, and effusive rocks. 
Classifications based primarily on chemistry have never 
been seriously considered by the great majority of petrograph- 
ers and the reasons against these systems are so obvious that 
it is hardly necessary to go into detailed arguments against 
ihem. We have already mentioned that rocks of different 
mineralogic combination, such as certain syenytes and certain 
leucite-basalts, have the same chemical composition, and also 
that the chemical composition of a given igneous rock would 
be identical with that of certain metamorphic, sedimentary, or 
even chemically precipitated rocks. 
Fouque and Michel-Levy J, in their tabulated rock classi- 
fication, make the following principles the most important — 
(i) structure, (2) age, (3) mineralogic composition. The chief 
structures are three and are substituted for the chief modes of 
occurrence as adopted by Rosenbusch. They are granitoid, 
porphyric, and trachytoid. In point of age, Michel-Levy 
agrees essentially w-ith Rosenbusch in making his porphyric 
rocks pre-Tertiary, while his trachytoid rocks are necessarily 
Tertiary and post-Tertiary. Judging from the context ac- 
companying Fouque and Michel-Levy's tabulation, however, it 
would seem as if they had been drawn into this exact form of 
tabulation much by the influence of Rosenbusch's table, for 
in this text, we find the chief elements of classification stated 
as (i) mineralogy, (2) structure, and (3) age; and on page 42 
we find the statement that "the giving of different names to 
the Tertiary and pre-Tertiary series is as arbitrary as pos- 
*See review in the American Geologist, vol. 23, No. 6, p. 346. 
tintroduction to the study of Petrology, London, i8gi, p. 78. 
^Structures et classification des Roches Eruptives, Paris, iSSg p. 42. 
