144 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



TABLE V. 



Transference Number of the Anion. 



The agreement between the results presented in the table for the 

 iodides with those for the chlorides is satisfactory. The greatest de- 

 viation occurs in the values for the two halogen acids. Hydrochloric 

 acid,^'' however, is known to be abnormal in its behavior since the 

 transference number passes through a minimum at a concentration 

 below normal, and the same effect might occur in hydriodic acid solu- 

 tion. Furthermore, since the electromotive force measured was so 

 small in the case of hydriodic acid any constant error in the nature of 

 a residual electromotive force might have had a large effect. 



IL Summary. 



In this article an apparatus and procedure have been described for 

 determining transference numbers by the centrifugal method, first 



^^ These values for the transference number at infinite dilution were cal- 

 culated from the latest conductivity data of Kohlrausch, Zt. f. Electrochem. 

 13, 333 (1907). 



2^ This value for the Hittorf transference number of molal HCl is calcu- 

 lated from Buchbock's determination of true transference niunber and hydra- 

 tion by the relation connecting those quantities as developed by ^^'ashburn. 

 Washburn himself gives 0.18 for the Hittorf transference of HCl, a value 

 which was taken from Kohlrausch's tables, but it is probably considerably 

 too high. For HCl, 0.97 molal, Riesenfeld u. Rcinhold. Zts. f. Phys. Chem. 

 68, 440 (1909), obtain the value 0.155, and Hopfgartner, Zts. Phys. Chcni. 25, 

 115 (1898), obtains 0.159 for 0.9 molal HCl. 



" Riensenfeld u. Reinhold, Zts. f. phys. Chem. 68, 440 (1909). 



