Diffusion of Liquids. 431 



my results and those of Kohlrausch; for the latter has shown 

 that the molecular conductivity of a salt in solution depends 

 also on the velocities of its component parts. From this stand- 

 point he has calculated the conducting-powers of the sub- 

 stances for which n is known; and the results thus obtained 

 agree very well with those obtained by experiment. 



It appears, then, that the law of the independent migrations 

 of the ions is as applicable before as during the electrolysis 

 of salt in solution; otherwise what is the meaning of the above 

 analogies ? 



There are also several other phenomena showing a close 

 relation between diffusion and conductivity. Both increase 

 with the temperature, as one might expect ; and from Gra- 

 ham's experiments it appears that the increase in the case of 

 diffusion is about the same as Kohlrausch found in the case of 

 conductivity. The influence of temperature on diffusion has 

 been as yet, however, determined but for few substances; and 

 there is of course more or less uncertainty in these determi- 

 nations. 



Another point of analogy is this. Graham has shown that, 

 in the case of nearly every substance investigated by him, the 

 increase in the rate of diffusion is less as the concentration in- 

 creases. This is also true of the conductivity of the same 

 solutions; and the explanation is the same in both cases. As 

 remarked by Kohlrausch, in concentrated solutions the mole- 

 cules interfere with each other's motions ; and this interfe- 

 rence is manifested in the one case by a decrease in the rate 

 of diffusion, and in the other by a decrease in the conducting- 

 power, both relative of course. 



Indeed, considering all the above data, it will be seen that 

 the most perfect analogy exists. In the one case the number 

 of particles, in the other the force transmitted by the same is 

 the element observed; and that a correlation should be found 

 is not at. all singular. 



Whether the velocity of the salt-molecules is the same 

 during diffusion as during electrolysis I cannot say ; but it 

 seems to be established by the above that, if any increase take 

 place in the latter case, the fraction representing this increase 

 must be the same for all substances of each series. 



But before proceeding any further with this subject, there 

 are several other points which must be investigated. For 

 instance, the rates of diffusion of chlorine and bromine in water 

 must be more exactly determined; and, if possible, more cor- 

 rect determinations of the influence of temperature should be 

 made. 



The alkaline hydrates present also a very interesting field 



