<o 



254 Van Name and Edgar — Velocities of Certain Reactions 



sharp boundary, and in every case it was the stronger solution 

 in which the diffusion appeared to be most rapid, thus confirm- 

 ing the conclusion drawn above with the aid of the diffusion 

 theory. A confirmation based upon quantitative measure- 

 ments would of course be much more satisfactory, and an 

 attempt will be made by one of us in the near future to obtain 

 quantitative evidence on this question. The result obtained 

 is interesting in that it seems to be an exception to the role 

 of Abegg and Bose* according to which an electrolyte diffus- 

 ing in the presence of a large excess of a salt with like cation 

 tends to assume the velocity of its own anion. As the electri- 

 cally measured velocity of 1/ ion is much less than that of K"*~ 

 ionf, a retardation would be expected in the present case 

 instead of the observed acceleration. 



We have still to consider the application of the diffusion 

 theory to the explanation of the observed reaction velocities 

 of mercury with iodine, bromine, and cupric bromide respec- 

 tively. The values of the iodine and bromine constants, for 

 a potassium iodide or bromide concentration of 3*4 normal 

 and stirring at 240 revolutions per minute, were compared on 

 page 249. A correction for difference in the rate of stirring 

 permits the cupric bromide constants of Table IX to be 

 included in the comparison, which gives, for the approximate 

 ratio of the reaction velocities, iodine ll'O, bromine 12'3, cupric 

 bromide 6*9. Of the three solutions corresponding to these 

 constants the last two would have practically the same vis- 

 cosity ; that of the iodine solution, judging by the viscosities 

 given by Taylor and EankenJ for 3-normal solutions of potas- 

 sium bromide and iodide, would be slightly lower. The differ- 

 ence between the iodine and bromine constants is therefore in the 

 wrong direction to be explained by difference in the thickness 

 of the unstirred layer, so that we must conclude, first, that the 

 rate of diffusion of potassium tribromide is somewhat greater 

 than that of potassium triiodide, at least in the concentrated 

 solutions here used, and second, that cupric bromide diffuses 

 decidedly slower than either. Both of these conclusions can 

 be tested by direct measurement, but at present experimental 

 confirmation is lacking. Both however are plausible, especially 

 the last, which is in full agreement with the slow rates of dif- 

 fusion of copper salts in general as compared with potassium 

 salts. 



We are far from regarding the evidence presented above as 

 in any way conclusive in favor of the cliffusion theory as 

 applied to the reactions in question, but the fact that it 



*Zeitschr. phys. Cliem., xxx, 551. 



f See Burgess and Chapman, Jour. Chern. Soc, lxxxv, 1305. 



JProc. Roy. Soc. Edinburgh, xxv, 231. 



