172 PKOFESSOK H. B. DIXON AND ME. E. C. EDGAR 



The percentage composition of hydrochloric acid according to these deter- 

 minations is : 



Chlorine 97-237 



Hydrogen 2 -763 



100-000 



The number we have obtained for the atomic weight of chlorine is appreciably 

 hio-her than that calculated by F. W. CLARKE from the previous determinations, and 

 is slightly higher than STAS' value : 



After our experiments were completed, we heard that Professor T. W. RICHARDS 

 was engaged on a revision of STAS' work on the composition of silver chloride. 

 G. P. BAXTER quotes the value 35 '467 as having been obtained by RICHARDS and 

 WELLS for the atomic weight of chlorine, a number slightly higher than our own.* 



It would not be difficult to extend our experiments, using larger quantities of the 

 gases, if in the judgment of chemists it were thought desirable, t 



PART II. -DETAILS OF EXPERIMENTS. 

 1. Preparation of Hydrogen. 



For the preparation of hydrogen we employed the electrolysis of a solution of 

 barium hydrate, first proposed by BRERETON BAKER,! as a means f preparing 

 hydrogen free from traces of hydro-carbons. 



Since barium carbonate is quite insoluble in a solution of barium hydrate, any 

 slight action of the carbonic acid of the air ' on the dissolved hydrate during its 

 unavoidable exposure while filling the electrolytic apparatus might be safely neglected. 

 We have to thank Mr. BRERETON BAKER for kindly supplying us with some of his 

 highly purified barium hydrate. It had been re-crystallised fifteen times and was not 

 radio-active. It still contained a very small trace of barium carbonate. 



The arrangement of the hydrogen apparatus is shown in fig. 1. Three preliminary 



* Professor RICHARDS writes (February 13, 1905) that he finds 100.00Q parts of silver yield 132,867 of 

 silver chloride, whereas STAS considered 132,850 the most probable result. This new determination, 

 combined with our value for chlorine, would give silver an atomic weight 107 '90. 



t As further experiments have shown that chlorine can conveniently be burnt in an atmosphere of 

 hydrogen, one of us proposes to make a fresh set of determinations in this way and to condense and 

 weigh the hydrochloric acid formed. July, 1905. 



I ' Jl. Chem. Soc.,' 1902, vol. 81, p. 400. 



