HISTORY OF THE ATOMIC THEORY. 221 



" 1 will relate a few facts very suitable for rendering this 

 view plausible, if they do not confirm it. In analyzing, as I 

 have had occasion to do, some of the oxides of all degrees 

 which Ilinman has announced in calcining steel, iron, and cast 

 iron, I have met with only two known oxides of this metal, 

 mixed in different proportions ; the ores yielded me only an 

 equal mixture of black and red oxide. 



" If we examine the greenish oxides which lead and bismuth 

 give at the commencement of calcination, and those of tin, 

 copper, &c., we only find an oxide at the maximum which 

 envelopes difi*erent portions of the metal. If now we measure 

 these degrees of oxidation by the nitrous gas which they 

 give, we shall be led to believe that the metals oxidize them- 

 selves in all doses. But, no. There are unions of oxygen 

 like those of sulphur, acids, &c. Election (elective aflfinity) 

 and proportion are two poles, round which invariably move all 

 the systems of true combination, whether in nature, or in the 

 hands of the chemist. In a word, oxygen is not one of those 

 bodies which can be mixed; when it combines, it subjects 

 itself to certain proportions, and these proportions are what 

 we have now to study." 



1806. Vol. Ixiii., page 367. 



Defending Klaproth, he says, ^' A combination, according 

 to our principles, Klaproth would tell you, is a sulphuret 

 of silver, of antimony, of mercury, of copper ; it is also a 

 metallic oxide ; it is a combustible body acidified ; it is a 

 privileged production to which nature assigns fixed propor- 

 tions; it is an existence which nature has never created even 

 in the hands of man, except with the balance in her hand 

 pondere et mensura. Know then, he would add, that the 

 characters of true combination are invariable as the propor- 

 tion of their elements. From one pole to the other they are 

 found the same under these two aspects, their physiognomy 

 only may vary according to their mode of aggregation, but 

 never their properties. No differences have been observed 



