350 
SECOND REPORT— 1832 . 
affinities give way to stronger ; carbonates are changed into 
sulphates ; metallic substances are oxidized ; copper is replaced 
by iron, &c. 
I have not spoken of improvement in the methods of analysing 
minerals, of which many made in recent times might easily be 
enumerated; since these processes seem rather to belong to the 
history of chemistry. Nor have I attempted to give the results 
of analyses of particular minerals ; for these, though valuable 
materials of mineralogical knowledge, cannot be introduced into 
a general view like the present one, till they have been con¬ 
nected by some principle of dependence or relation. 
5. Classification. 
I. Distinction of species .—It will probably give to common 
hearers and readers a strong impression of the confusion still 
prevailing in the science of mineralogy, when we state that it 
is still a matter of dispute what are the limiting conditions and 
definition of Species in general; and that very wide and nu¬ 
merous differences of opinion prevail as to the identity and 
diversity of species in particular cases. Indeed it would be 
almost difficult to mention a species which is free from such 
doubts. This uncertainty is however not so fatal to real 
science as might at first sight appear. The formation of 
definitions, and the establishment of unerring distinctions, are 
among the last, and not the first, steps of systematic know¬ 
ledge. 
Haiiy’ s definition of a mineral species, “The same ingredients 
combined in the same form,” acquired a kind of celebrity at the 
time, and it has been adopted by many succeeding mineralogists. 
The definition which seems to be recognised in the crvstallo- 
metrical school of more modern times is, “ The same primary 
form with the same fundamental angles of cleavage, combined 
with an approximate identity of chemical and physical cha¬ 
racters.” But both these definitions were announced as axioms 
when they should have been tried as guesses. It was impos¬ 
sible to know, independently of experience, that the sensible 
differences of minerals corresponded universally to determinate 
differences in their ingredients. It is now certain that they do 
not: for, without calling in the doctrine of isomorphism, we 
know that scarcely two analyses of minerals of the same kind 
give identical results ; so that the Haiiyian definition of species 
is inapplicable without some reformation of its terms, and to make 
it unexceptionable will be found no easy task. In like manner it 
was impossible to know, independently of experience, whether 
minerals which resembled each other so as to have no constant 
