160 MELLO: MIOROSCOPICAL STRUCTURE OF ROCKS. 
also seen to be strongly dichroic, and by its marked dichroism is 
readily distinguished from hornblende, which is indeed rather 
pleochroic than dichroic. Again the metamorphosis frequently present 
in some of the constituents of a rock is well shown by means of the 
microscope, as for instance the olivine, a common mineral in the 
dolerites, is not seldom found to be undergoing a gradual transforma- 
tion into serpentine, from Avhich it has been surmised that when a 
rock is highly charged with olivine, this metamorphic action may 
sometimes be so extensive as to cause the change of the entire rock 
from dolerite or basalt into serpentine. 
We may gather then that when the structure of a rock has been 
carefully determined, we s'hall hold the key which will enable us to 
open the secrets of its history, the modes of its formation, and the 
origin of the rock itself. 
It is very important that we should learn to discriminate between 
minerals which belonged to tlie rock at its original formation and those 
of secondary character, and it will appear that certain crystals have 
separated from the magma at a period quite different from that of the 
formation of the other crystals, and the only way in which we can 
ascertain to which class certain minerals belong is by studying the 
relations which they bear to one another as seen under the micro- 
scope. Noting the position of the several crystals of the different 
minerals whicli have separated during the same period of consolida- 
tion, we shall observe that crystals of one mineral are often enclosed 
in those of another species, and again, the crystalline form of one 
mineral will be produced at the expense of others, so that w^e shall be 
able to establish the order in which they were crystallized." 
By means of the microscope we have the power to unravel the 
minute details of structure in the finest grained rocks, be they 
basalts or slates, or any analogous forms. We shall at once be 
enabled to distinguish one rock from another, say dolerite from 
diorite, the peculiar behaviour of the hornblende in the latter being 
so ma-ked that it cannot be mistaken for the augite of a dolerite. 
The microscope will also show us how closely alHed certain rocks 
* Prof. Judd, Q. J G-. S., Vol. xlii., pt. 1. 
