346 
SECOND REPORT- 1832. 
unfortunate that algebraical symbols should be used in a man- 
ner contradictory of the first principles of algebra, as is the case 
in the notation of Berzelius, it is probable that the general ac¬ 
ceptation of a notation of any kind will be mainly influenced by 
the amount and value of the information which it is employed to 
convey ; and in this respect the Swedish system possesses an 
advantage in which it cannot easily or soon be matched by any 
rival system. 
In speaking of the connexion of chemistry and mineralogy, I 
ought to mention the general law announced by M. Kupffer. 
This law professes to give the dependence of the crystalline 
form on the specific gravity and atomic weight of the body, 
and thus, if well established, would be a principle of a very 
high and comprehensive character in our science. I do not 
think, however, that any one who examines M. Kupffer’s Me¬ 
moir critically, will be satisfied with the kind and quantity of 
facts from which this induction is held to be collected. The 
selection of substances belonging to one particular system of 
crystallization, (for instance, the rhombohedral,) for com¬ 
parison with each other, seems to be quite an arbitrary step, 
and is in no way explained by the law so asserted. But not to 
insist on this objection, the mode in which the dimensions of 
the primary form are compared with the other quantities is 
such as would enable the author to prove almost any law with 
equal facility ; for he holds it to be an unimportant matter 
whether he takes what is usually considered the primary 
form, the primary rhombohedron for instance, or any other 
rhomboliedron which can be derived from it. Thus in com¬ 
paring calc spar with rhombohedral iron oxide, he takes the 
number expressing the axis of the primary form of the one 
substance, but in the other substance he multiplies the axis by 
four, thus substituting for the primary rhombohedron that which 
arises from truncating its edges. And it is by using a similar 
license in other cases that he exhibits an approximate verifica¬ 
tion of the formula which he states. There appears to be little 
hope of any valuable result to be obtained by comparison of 
numerical results, except the properties which the numbers ex¬ 
press be clearly the same property in the different cases which 
are compared. If the cleavage rhombohedron of one case is 
analogous to anything, it must be analogous to the cleavage 
rhombohedron in the other. 
Still it is impossible for any one to take an interest in this 
portion of science, without seeing that the connexion of chemi¬ 
cal composition with crystalline form is one of the great pro¬ 
blems to be solved; and it is very natural that those who feci 
this should be tempted to hazard a guess concerning the solu- 
