: 
: 
Classification of simple Boilies. 44\ 
with substances belonging to two different genera. In this case, 
they indicate between these two genera an analogy which it will 
perhaps be difficult to ascertain without their assistance, but 
which is not the less real, and according to which we ought to 
place them in succession after each other, so that the body which 
establishes the link of the chain is at the end of the first, or the 
beginning of the second, in order to be always between two 
bodies which they resemble by common characters. Nothing 
then remains but to determine with which of the two genera it 
ought to be definitively united, by comparing the properties which 
it shares with the one, and those which are common to it with 
the other, in order to decide according to the number and im- 
portance of the analogies which result from these properties. 
The analogies to which these researches wil] lead us, will fix in 
an invariable manner the natural order of the simple bodies, con- 
formably to the general idea which I am about to give of them. 
The last head of this paper will have for its object to examine 
once more the various genera into which all these bodies shall 
have been distributed according to the data laid down in the 
preceding head, in order to assign to each of them a distinctive 
eharacter formed by the union of some remarkable properties, 
chosen in such a way that they cannot be found at once but in 
a body appertaining to the genus which it is wanted to charac- 
terize, and to see at the same time according to what principles 
of nomenclature we could, if necessary, establish for each genus 
a denomination common to all the bodies which form part of it. 
§ I. On the impossibility of reconciling the manner in which 
chemists have hitherto ranged simple bodies, and the distine- 
tions which have been established between them; with a 
classification deduced from the whole of their properties ; 
and on the order which it is proper to adopt, to unite as 
much as possible those which present the most characters in 
common. 
The first source of the artificial classifications hitherto used, 
seems to me to have arisen from the old distinction of the metals 
and non-metallic bodies. I must confess, however, that this di- 
vision leads us to separate but a very small number of bodies 
which we ought to unite in the natural order; and that it is in 
general tolerably conformable to the classification which results 
from the comparison of all the properties of bodies; and that it 
will even be sufficient, in order that it may embrace none of the 
genera which I regard as natural, to separate from the metals 
three substances which are generally united with them, arsenic, 
tellurium, and silicium: but then it becomes very difficult to as- 
sign a character which distinguishes in every case the male 
rom 
