498 General Notes. [ May, 
somewhat peculiar views. The petrographical descriptions are, 
however, largely taken from the work of others, while those 
which are original are not sufficiently detailed; the generaliza- 
tions are often broader than the facts thus far accumulated woul 
seem to warrant; and even the statement of the writer’s most 
original ideas regarding rock nomenclature and classification is 
not in certain points altogether free from ambiguity. 
Chapter first, containing nine sections, deals with the interior 
structure of the earth; the origin and alterations of rocks and of 
their constituent minerals ; the methods of rock-classification 
hitherto followed and their ‘defects, and lastly, the proposal of the 
author’s system of classification. 
Sedimentary and eruptive rocks are held never to grade into 
one another, as sometimes appears to be the case. Each class 
mainly to the alteration of the older ones. The minerals tend 
to constantly pass from less stable compounds to those which are 
more stable for the conditions now existing on the earth. The 
alteration therefore varies with the age, and also, under the same 
_ conditions, inversely as the amount of silica which the rocks 
contain. Foliation or schistose structure is no necessary proof ot 
ei Deets» Us origin of a rock. Inasmuch as the alteration of 
ces on in some cases much more rapidly than in others, 
lithological character can be regarded as no index of age. 
The mineral constituents of an eruptive rock are divided into 
three classes: 1st, those present in the magma before its extru- 
sion (foreign) ; 2d, those formed at the time of the consolidation 
of the magma (indigenous) 3d, alteration products (secondary). 
The first class is regarded as composed entirely of foreign inclu- 
sions, no account .being taken, as it seems, of such minerals as 
leucite, olivine, etc.,which may crystallize out of the molten magma 
long before it is extruded or solidifies! Hornblende appears to be 
regarded as always belonging to either the first or third class 
The present systems of rock-classification, based on chemical 
composition, structure, mineral constituents and geological age, 
are reviewed in turn and pronounced artificial and unsatisfactory. 
Section viii contains the statement of thirteen principles which 
the author thinks should underlie a natural classification of rocks. 
It must be confessed, however, that this attempt is not altogether 
- Satisfactory. It is stated that a// the petrological (field), oe 
cal (microscopical) and chemical characters of a rock must be 
used in ining its species, but in what way is not made 
clear. Mineral composition is sufficient to define varieties but 
_ Rot species. All rocks which may be followed from one form to 
