November 4, 1904.] 



SCIENCE. 



585 



and cadmium are pronounced exceptions 

 to the rule. 



A far more extensive material for testing 

 these rules is furnished by the measure- 

 ments made at 25° between the concentra- 

 tions of 1/32 and 1/1024 normal. In the 

 case of the uni-univalent salts, data exist 

 at this temperature and these concentra- 

 tions for 36 inorganic salts, about 65 so- 

 dium salts of organic acids, and about 

 an equal number of hydrochlorates of or- 

 ganic bases. A consideration of all these 

 data shows that, with only three or four 

 exceptions not of a pronounced character, 

 the values of the degree of ionization of all 

 these salts in 1/32 normal solution lie be- 

 tween the limits of 84 and 90 per cent, and 

 are fairly uniformly distributed through- 

 out this range of 6 per cent. For 67 uni- 

 bivalent salts the corresponding limits of 

 the ionization values are 72 and 81 per 

 cent, while for only four such salts do the 

 values lie beyond these limits. For the 6 

 uni-trivalent salts investigated the range 

 is from 67 to 76 per cent. ; for the three 

 uni-quadrivalent salts, from 59 to 63 per 

 cent. ; and for twelve bi-bivalent salts, 

 from 49 to 63 per cent., while three such 

 salts show more considerable variations. 

 The values of the un-ionized fraction cor- 

 responding to the mean of these two limits 

 for the different types of salts at the same 

 equivalent concentration increase some- 

 what more slowly than the product of the 

 valences of the ions. The proportionality 

 becomes a fairly close one, however, when 

 the salts are compared at the same molal, 

 instead of the same equivalent concentra- 

 tion. Thus, with the help of the Kohl- 

 rausch concentration-function, it is calcu- 

 lated from the preceding values that the 

 un-ionized fractions in 1/32 molal solution 

 are as follows : 13 per cent, for the uni-uni- 

 valent salts, 29-| per cent, for the uni-biva- 

 lent salts, 41 per cent, for the uni-trivalent 

 salts, 62 per cent, for the uni-quadrivalent 



salts, 55 per cent, for the bi-bivalent salts, 

 which are seen to be approximately the re- 

 quired miiltiples of the constant factor 14. 



Before leaving this subject it should be 

 stated that the results conform, on the 

 whole, about equally well with the rule that 

 the decrease of equivalent conductivity {in- 

 stead of ionization) is roughly constant for 

 salts of the same type; and when the com- 

 parison is made at the same equivalent con- 

 centration, distinctly better to the rule that 

 the decrease of equivalent conductivity is 

 propo7-tional to the product of the valences 

 of the ions for salts of different types. 

 (When compared at the same molal con- 

 centration, however, this rule does not 

 apply.) These rules were originally stated 

 by Ostwald. They differ not inconsider- 

 ably from those expressing the change in 

 ionization, namely, to an extent corre- 

 sponding to the variations of the conduc- 

 tivities at extreme dilution. The deviations 

 are so irregular, however, that from an em- 

 pirical standpoint the choice between the 

 two pairs of rules is arbitrary. In either 

 form these rules seem to justify the infer- 

 ence that the degree of ionization of salts, 

 unlike that of the organic acids and bases, 

 is not primarily a specific chemical prop- 

 erty determined by chemical affinity, but 

 that it is determined, at least in the main, 

 by the magnitude of the electric charges on 

 the ions. 



The establishment of the principle in re- 

 gard to the ionization of a mixture of salts 

 has a direct bearing on the phenomenon of 

 the effect of. one salt on the solubility of 

 another with a common ion. It has been 

 usually assumed that in a (not too concen- 

 trated) saturated solution the un-ionized 

 molecules of the salt always have the same 

 concentration, and secondly that the prod- 

 uct of the -jow-concentrations (each raised 

 to a power corresponding to the number re- 

 sulting from one molecule) also retains the 

 same value. And the experimental results 



