ON ELECTROLYSIS IN ITS PHYSICAL AND CHEMICAL BEARINGS. 335 



increases, to a limiting' value ; the temperature coefficient beiuj; the same, within 

 narrow limits, for all neutral salts. 



In his previous investigations Kohlrausch had determined the specific conduc- 

 tivities of solutions of varying strength, from saturation dov,ii to a few per cents., 

 and from observations thus obtained had deduced the conductivity in more dilute 

 solutions by extrapolation. In the present memoir he verifies and corrects the 

 results thus obtained by the employment of solutions whose strengths vary from 

 1-0 to O-OOOOl gramme-equivalents per litre. (The strength thus estimated is 

 termed the ' molecular content.') 



In measuring the resistances a Wheatstone bridge is used, with alternating 

 cuiTents and a telephone as indicator (the paper deals at some length with the 

 advantages and difficidties of this method) ; the specific conductivities are expressed 

 in terms of that of mercury, and calculated for a uniform temperature of 18° C. 



In working up the results the conception of 'specific molecular conductivity' is 

 employed, this being defined as the ratio kjm of the specific conductivity k of the 

 solution to the molecular content m. As the specific molecular conductivities are 

 all very small, Kohlrausch tabulates the values of 10** . ki'm, instead of kjin simplj'. 

 The values of this ratio are exhibited graphically, by plotting in a curve, with 7nh as 

 abscissa, this quantity being chosen because it expresses ' the reciprocal of the mean 

 distance, or, in other words, the mean relative pruximity of the molecules ' — i.e., the 

 magnitude on which the specific molecular conductivity is most directly dependent. 



The material thus obtained is amply sufficient to verify the three statements 

 above mentioned, the very important law of independent migration in particular 

 i"eceiving a full confirmation. In addition a large number of interesting relations 

 disclosed themselves, which may be summarised as follows: — 



1. If the solution be sufficiently dilute the specific molecular conductivity A-/;« 

 is, for all neutral salts, independent of the strength of the solution, or each mole- 

 cule conducts the current independently of every other. In other words, the 

 specific molecular conductivity has a limiting value. These limiting values are not 

 the same for aU salts, or even for those of the same base or acid ; their magnitude 

 depends on the nature of both ions. 



2. The velocity with which the ions move past each other in extremely dilute 

 solution, under an electromotive force of 1 volt per linear millimetre, is determined 

 in millimetres per second by multiplying the specific molecular conductivity of the 

 electrolyte by 11000^ (or Kohlrausch's tabulated numbers by 0-00011). The 

 numbers thus obtamed for the various salts examined all he between 0'14 and 

 O'lO mmjsec. 



3. The specific molecvdar conductivity of a salt in dilute solution being propor- 

 tional to the sum of the velocities of the separated ions, the values of these absolute 

 velocities may, as Kohlrausch had shown previously, be determined from the ex- 

 periments in the following way : — 



Let i«i and ?;j be the absolute velocities of K and CI, !/, and v.^ those of Na and 

 Br ; then the 



Specific molecular conductivity, kjm, of KCl = (u^ + r,)/11000 



„ NaCl=(M„ + Vi)/11000 



„ KBr = (w>«..)/11000 



„ „ „ NaBr = (M2-n'.,)/11000 



and so on ; hence the conductivities of KCl, NaCl, KBr, and NaBr, contain all that 

 is necessary for determining the absolute velocities of K, Na, CI, and Br; since the 

 ratios ujv^, u.^jv.., &c., are known from Hittorf 's migration experiments. 



4. As the strength of the solution increases, the specific molecular conductivity 

 slowly diminishes. For salts with monovalent ions it may be approximately re- 

 presented by the formula 



kjm = A - B>«J ; 



and since m-- = the relative mean distance ?• of the molecules, 



kjtn = A — B/r, 



' [This appears to be necessary in order to reduce measures in terms of mercury 

 to absolute measure. — E. F. J. L.] 



