222 Van Name and Bostvorth — Hates of Solution of 



Under B in Table VII are given the mean values of the 

 velocity constants for the experiments with cadmium in the 

 presence of different iodides. Considered from the standpoint 

 of the diffusion theory, these figures must represent, at least 

 roughly, the relative magnitudes of the rates of diffusion of 

 iodine in the different solutions. An exact parallelism would 

 not be expected, since this could only exist if the thickness of 

 the diffusion layer were the same in each case, while in reality 

 the thickness would vary somewhat with the properties of the 

 liquid, especially its viscosity. 



It would be extremely interesting to compare these velocity 

 constants with the directly measured diffusion coefficients of 

 iodine in the different solutions, but as yet none of the nec- 

 essary diffusion coefficients have ever been determined, though 

 it is likely that this deficiency will be remedied before long.* 

 It is possible, however, by making certain assumptions, to 

 obtain a rough estimate of the relative values of these rates of 

 diffusion. For reasons already given (p. 221), they would be 

 expected to depend chiefly upon the diffusion velocities of the 

 triiodide molecule. As a first approximation we may assume 

 that the rates of diffusion of the different triiodides would fol- 

 low the same order as the corresponding iodides. Inspection 

 of the existing data on the rates of diffusion of iodides shows 

 clearly that potassium, sodium, and lithium iodides stand in 

 the order named, while hydrogen iodide, by analogy with 

 hydrogen chloride, should be highest of all. With the divalent 

 iodides it is necessary to base the estimate chiefly upon the 

 migration velocities of the positive ions, which are given in 

 the third column of Table VII, B. The difference in the ioni- 

 zation of divalent and univalent iodides makes it unsafe to 

 base any comparison between the two groups upon the ionic 

 velocities. The diffusion data seem to show that the alkali 

 earth halides diffuse slower than the alkali halides. 



On referring to the values of the velocity constants for cad- 

 mium in the presence of different iodides, as given in Table 

 VII, B, we find the alkali iodides in the order, potassium ; 

 sodium ; lithium ; as would be expected from the above con- 

 siderations. Barium stands above both magnesium and cal- 

 cium, as would be anticipated, but calcium is slightly below 

 magnesium, thus deviating from the order of the ionic velocities. 



There are two values, however, for which the above analysis 

 gives no explanation. One is the high velocity constant for 

 cadmium iodide, which in view of its abnormally low ioniza- 

 tion and its tendency to form complex anions would be 



* Dr. Graham Edgar has kindly undertaken to cany out these measure- 

 ments in connection with other experiments on diffusion which he is conduct- 

 ing at the University of Virginia. 



