between Metals and Dissolved Halogens. 253 



and zinc only, but all the five metals investigated react with 

 iodine at practically the same rate. 



Unless we accept the diffusion theory there seems to be no 

 reason why these reaction velocities should be the same. But 

 even if the above result should be found to hold for a large 

 number of metals, it would not constitute conclusive proof 

 that the diffusion of iodine is, as- it appears to be, the deter- 

 mining factor. Evidence of a more direct nature is needed 

 upon this point. 



As shown above, a doubling of the concentration of the 

 potassium iodide accelerates the reaction by about ten per 

 cent. According to the diffusion theory such an acceleration 

 might be due to an increase in (a), the rate of diffusion (here 

 that of the iodine), or to a decrease in (5), the average thick- 

 ness of the unstirred layer, or to both together. While (a) 

 is in general directly measurable, little can be predicted about 

 ■(b) except that it would vary with the properties, especially 

 the viscosity, of the liquid. It seems, however, permissible 

 to assume that in the majority of cases the changes in (a) 

 would predominate over those in (5), and especially here, since 

 the viscosity of strong solutions of potassium iodide at 25° 

 varies but little with the concentration.* According to this 

 view the rate of diffusion of iodine, or more accurately, that 

 of potassium triiodide,f ought to show a distinct increase with 

 the concentration of the potassium iodide. 



As we have been unable to find any published data which 

 either confirm or disprove this conclusion, we have carried out 

 a qualitative test by comparing in two JSfessler tubes of the 

 same dimensions, clamped side by side in a large water bath, 

 the rate at which the color spread from a lower layer of iodine 

 in potassium iodide into an upper layer of pure potassium 

 iodide of the same strength. The potassium iodide solutions 

 in the two tubes contained respectively 25 and 400 grams of 

 iodide per liter, and the iodine concentrations (the same in 

 both tubes) were similar to those used in the previous work. 

 Several repetitions of the experiment gave the same result. 

 After a few hours a difference was visible in the extent to 

 which the brown color had progressed beyond the initially 



* According to W. W. Taylor and C. Banken (Proc. Boy. Soc. Edinburgh, 

 xxv, 231) the viscosities of 1 and 3-normal KI at 25° are 0'467 and 0"459 

 respectively, that of water at the same temperature being 0*501. 



f With the aid of the equilibrium constant of the reaction KI 3 ^^Kl4-I 2 , 

 (k = 0-0014 at 25°), determined by Jakowkin (Zeitschr. phys. Chem. xx, 19), it 

 may easily be calculated that in no experiment in the above tables did the 

 proportion of free iodine exceed a few tenths of one per cent of the total 

 iodine, i.e., that indicated by thiosulphate. A similar calculation based on 

 Jakowkin's data for bromine shows that in the experiments of Table VIII 

 the tribromide bromine formed in all cases more than 98 per cent of the 

 total bromine, so that here too the diffusion velocity would, be practically 

 that of the tribromide. 



