373 Berzelius on Mr. Proufs theory of Atomic Weights. 



a single atom. You know also that chlorine combines itself 

 with hydrogen only in equal volumes, and that in the great num- 

 ber of isomorphous compounds in which chlorine is substituted 

 for hydrogen, without a change of crystalline form, this substi- 

 tution takes place invariably under equal volumes. We cannot, 

 therefore, avoid the conclusion that two volumes of hydrogen as 

 well as of chlorine, represent two atoms, and consequently an 

 atom of water contains two atoms of hydrogen, just as one atom 

 of cuprous oxyd, raercurous oxyd, and hyposulphurous acid, &c, 

 contain two atoms of the radical. It therefore follows — and not 

 equivocally — that chlorine and hydrogen cannot be compared ex- 

 cept in the relation of equal volumes, and consequently the ato- 

 mic weight of chlorine cannot be a multiple of that of hydrogen. 



In glancing over the table of atomic weights, you will find 

 numerous exceptions to the hypothesis of Mr. Prout. It is true 

 that the most of these numbers have been determined by exper- 

 iments made 30 years since, and at the first development of 

 chemical proportions. They cannot, therefore, have all the ex- 

 actness which they might have with the new analytical processes 

 since discovered. But I am persuaded that they will be found, 

 with few exceptions, to approach quite near the truth, and not to 

 admit of that latitude of correction demanded by Prout's hypoth- 

 esis. I cite, for example, the atomic weight of lead, which has 

 been the subject of reiterated experiments in the course of this 

 discussion, and which has been fixed at 1294-5 — also that of 

 copper 395-7, both lately verified by Messrs. Erdmann and 

 Marchand. To accord with Mr. Prout's hypothesis, the first 

 ought to be 1287-5 or 1300, and the second 387-5 or 400. 



Among the atomic weights there are those which are apparent- 

 ly equal, and others which approximate quite near to being dou- 

 ble of one another in weight — from which it is probable that 

 there is between them a certain relation ; but these pertain only 

 to particular groups of the elements — such as the metals which 

 accompany platinum, — molybdenum and tungsten, — chrome, 

 iron, manganese, &c. But this is a new question, to be discuss- 

 ed only after farther investigation. 



Yours, &,c. Jac. Bekzelius. 



