On the Controversy concerning Volta's Contact Force. 455 



too literally, that solution may be a kind of chemical com- 

 bination, and yet all the facts deduced on the free ionisation 

 hypothesis can be true. This compromise is, however, criti- 

 cised by Dr. Larmor in a footnote to page 273, op. cit.; and 

 Boltzmann is able to regard osmotic pressure as due to a true 

 molecular bombardment, seeing that in the liquid slate the 

 molecular velocities should be the same as in the gaseous state 

 at the same temperature, the mean free paths alone being 

 smaller. On the other hand, it has been frequently claimed 

 that any extension of gaseous laws, i.e. of laws deduced from 

 the collisions of free particles, to the extremely hampered 

 condition of a liquid, where the particles are always within 

 each other's molecular range, must be illegitimate. 



Without attempting to decide these moot points, we will 

 here consider the question open whether the statement in 

 terms of free ions is only a mode of expression, or whether 

 it corresponds with the real facts ; corresponds with them, that 

 is to say, not only in result, but in actual detail. Prof. 

 Poynting's paper, for instance, does not decide against the 

 latter possibility, it only attempts to show that it is not 

 really necessary to grant free ions as a physical reality; but 

 Mr. Whetham evidently inclines to the idea that they are at 

 least free of each other, and adduces an argument in favour 

 of that view (pp. 233, 244, B. A. Report 1897) which is worthy 

 of attention. But vv-hile assuming that opposite ions are free 

 of each other, he considers that they are both attached to the 

 solvent, and are by no means really and bombardingly free, 

 as are the molecules of gases. Prof. Poynting appears to 

 agree *; but, as said above, the agreement is not universal. 



The most successful application of Nernst's theory is to the 

 calculation of the feeble E.M.F. of ceils consisting of one 

 metal but two liquids, especially when the two liquids differ 

 only in concentration of one and the same salt. 



The ideas underlying all such calculations are (a) that the ions 

 migrate for two reasons: one their own random motion whereby 

 they naturally diffuse on the average from the more concen- 

 trated to the less concentrated solution, carrying their charges 

 with them, the other an organized motion caused by the pro- 

 pulsive influence on these same charges of the gradient of 

 potential caused by their gradual diffusive accumulation ; and 

 (b) that in the steady state ultimately attained the rate of 

 migration of opposite ions must become equal : otherwise the 

 liquid would be constantly altering its composition in places. 



Indeed the fact that equivalent quantities of anions and 



* Nature, 1896, vol. liv. p. 571 ; vol. lv. pp. 33, 78, 150, &c 



