202 THE ATOMIC WEIGHTS. 



equal in weight, we know from Cooke's experiments that 

 the larger mean is vitiated by serious constant errors. Only 

 in value 5, the result calculated from Dexter's experiments, 

 has the constant error not been pointed out. Cooke consid- 

 ers it probable, however, that the Sb.O^ involved in this 

 work contained traces of some lower oxide, which, if present, 

 would render the atomic weight of antimony apparently too 

 high. Chemically considered, the preponderance of evi- 

 dence is strongly in favor of values 1 to 3, deduced from 

 the experiments of Schneider and of Cooke. These give a 

 general mean of Sb = 119.955, ±: .036; or, if = 16, this 

 becomes Sb = 120.231. 



This we may accept as most nearly the true result, and 

 reject the data of Dexter, Dumas, and Kessler altogether. 



Since this chapter was written, Pfeifer has compared the 

 amount of antimony thrown down electrolytically, with the 

 quantit}^ of silver deposited by the same current in the same 

 time. From rather meagre data he concludes that the atomic 

 weight of antimony, thus determined, may be 121. Addi- 

 tional investigation is promised. The figures thus far pub- 

 lished would weigh little as against Cooke's experiments. 

 (Ann. Chem. Pharm., 209, 161. 1881.) 



BISMUTH. 



Early in the century the combining weight of bismuth 

 was approximately fixed through the experiments of Lager- 

 hjelm.* Effecting the direct union of bismuth and sulphur, 

 he found that ten parts of the metal yield the following quan- 

 tities of trisulphide : 



12.2520 



12.2065 



12.2230 



12.2465 



Mean, 12.2320 

 * Annals of Philosophy, 4, 358. 1814. Results adopted by Berzelius. 



