214 JOURNAL OF THE PLYMOUTH INSTITUTION. 
characters. Minerals are definite combinations of elements. 
Chemical analysis tells us clearly enough what are the elements, 
and what their proportions, which go to form a mineral; and 
chemical analysis indicates what are the elementary bodies, and 
what their proportions, in the mixture of minerals which con- 
stitute a rock. But chemistry, save in rare instances, cannot say 
in what proportions these elements are distributed among the 
component minerals, and therefore what these minerals are. 
The chief rock-forming minerals—taking the genera, and not 
the species—are various oxides of iron, olivine, enstatite, hyper- 
sthene, augite, hornblende, mica, felspar, and quartz. Excluding the 
oxides of iron (all of which contain iron and oxygen), every one of 
these minerals contains silica, and some of them notable quantities 
of alumina, magnesia, lime, potash, soda, or other alkalies. The 
same elements, differently proportioned, thus occurring in whole 
or in part in several associated minerals, no satisfactory informa- 
tion is obtained by mixed analysis. Moreover, there are minerals, 
practically identical in composition, which vary chiefly or wholly 
in their crystalline form. Calcite and arragonite have precisely 
the same components, and differ only in the fact that arragonite 
crystallizes in the rhombic and calcite in the hexagonal system. 
Neither the hand-lens nor chemistry, therefore, affords us direct 
aid in dealing with minutely-crystalline rocks, though by the use 
of specific-gravity solutions with powdered rocks, several minerals 
may be and are isolated for identification. 
We owe the present position of petrological science to the 
microscope. Opaque and transparent are for the most part relative 
terms. The densest and most opaque rock, if cut into slices 
sufficiently thin, is transparent, or at the least translucent, and its 
structure and composition are then capable of examination in the 
fullest detail by various methods of microscopical investigation. 
Not only may the nature of the minutest components of a rock 
be ascertained, but the history of the rock, and the changes 
through which it has passed, may frequently be made out. 
Without attempting further detail, I may add that microscopic 
petrology in its origin is purely English. Mr. H. Witham ground 
thin rock sections for microscopic examination, in the study of 
fossil plants, so far back as 1831. But the real father of this 
branch of science is Mr. H. C. Sorby; and it is not yet a quarter 
of a century since he commenced the microscopic study of purely 
