TOLMAN. — ELECTROMOTIVE FORCE PRODUCED IN SOLUTIONS. 143 



dient. Owing to their enormous " internal pressure " the properties 

 of li(j[uids are, however, in general but little affected by changes in the 

 external pressure. 



The Effect of the Dissolved Iodine on the Results. 



In the case of the alkali iodides one per cent of the I ion was 

 changed into Is ion by the iodine present. In the HI solution, owing 

 to oxidation about twice as much iodine was present. The transfer- 

 ence number would, however, be only slightly affected by the small 

 admixture of L ion.^o 



10. Comparison of the Results with other Transference 

 Measurements. 



The available data on the transference numbers of iodides are very 

 few, and, of these, many are vitiated by the use of membranes in the 

 apparatus.21 The most satisfactory data for comparison are the values 

 determined by Washburn ^^ for the Hittorf transference numbers of 

 the alkali chlorides (at a concentration 1.2 — 1.3 molal), and the value 

 for the Hittorf transference number of hydrochloric acid (at a concen- 

 tration 1.0 molal) which can be calculated from Buchbock's determina- 

 tion of true transference number and hydration. This comparison is 

 made in Table V. The transference numbers for the chlorides and 

 iodides at infinite dilution, calculated from conductivity data are also 

 given in the table, and finally, values obtained for y^ KI and ^o Nal 

 by Dennison ^3 using the method of Dennison and Steele and a value 

 by Bein ^4= for ^ KI. This last is the only available datum for these 

 iodides obtained by the Hittorf method without the use of membranes. 



^^ For data on the mobility of the I3 ion see Burgess & Chapman, J. Chem. 

 See. Trans. 85, 1305 (1904), Bray & MacKay, Journ. Amer. Chem. Soc, 32, 

 914 (1910). 



21 See McBain (Proc. Wash. Acad. Sci. 9, 1-78 (1907); University of 

 Toronto Studies, Papers from the Chem. Laboratories, No. 07, for a complete 

 collection of the experimental data on transference numbers. 



22 Washburn, Tech. Quart. 21, 164 (1908); Journ. Amer. Chem. Soc, 31, 

 322 (1909); Buchbock, Z. physik. Chem. 55, 563 (1906). 



2' Dennison, Trans. Faraday Soc, 5, 165 (1909). It has been shown by 

 Lewis, J. Am. Chem. Soc. 32, 862 (1910), that the method of Dennison and 

 Steele gives, after applying a calculable correction, the Hittorf transference 

 number and not the true transference number as stated by Washburn. 



2* Bein, Z. physik. Chem. 27, 1 (1898). 



