LECTURE VI.] HISTORY OF CHEMISTRY. 99 



information about this, and therefore he seeks for other 

 generalisations. He now rejects the apparently most natural 

 assumption of i atom of radical, which he had made in 1819, 

 since it leads him to atomic weights that are not in harmony 

 with Dulong and Petit's law. He finds a new starting-point 

 (except for silver, tellurium, and cobalt) by assuming the pre- 

 sence in compounds of two atoms of the element concerned, 

 and thus he obtains the following series for the stages of oxida- 

 tion of the most of the metals : 



R 2 O, RO, R 2 O :? , RO 2 , R 2 O 5 (R 2 O 7 ), 

 or RO, RO, RO 3 , RO 2 , RO 5 (RO 7 ), 



where he writes RO instead of R 2 O 2 , and RO 2 instead of R 2 O 4 . 



Berzelius advances several grounds which appear to tell in 

 favour of the accuracy of his choice. The most commonly 

 occurring oxides, such as cupric oxide, magnesia, lime, etc., 

 receive the simplest formula, RO ; further, the oxygen com- 

 pounds of nitrogen and of chlorine, in which he knows, from 

 the volumes, the number of atoms, can be made to fit in with 

 his arrangement. On this account, he regards this series as 

 one that occurs generally distributed, calls it the nitrogen series, 

 and contrasts the sulphur series with it. 



Berzelius finds the relative quantities of oxygen that combine 

 with sulphur to be i, 2, 2^, and 3. Hence, he writes the stages 

 of oxidation of this element : SO, SO 2 , S 2 O 5 , and SO 3 . He 

 endeavours to arrange all the oxygen compounds, as far as 

 possible, in the' sulphur and nitrogen series, and assumes, for 

 example, SiO 3 as the formula for silicic acid, corresponding 

 with sulphuric acid, an assumption which afterwards gave 

 occasion for many discussions. 



".- The sulphur compounds (sulphides) are regarded as con- 

 stituted in a manner analogous to the oxygen compounds. He 

 writes sulphuretted hydrogen HS, because water is HO. 



In calculating the atomic weights which he deduces from 

 these considerations, Berzelius starts from O= 100, but, in order 

 to permit of comparison with earlier and with subsequent 

 statements, and since it is merely a question of the relative 



