THE ELECTROLYTIC DISSOCIATION THEORY 389 



point depression — the depression produced by a gram-molecule 

 of the salt x the volume of the solution in litres — gradually 

 diminishes with increase of concentration, reaches a minimum 

 in o*i to 0'5 normal solutions, and beyond this point gradually 

 increases. The molecular elevation of the boiling-point shows 

 the same anomaly, though not so distinctly, and the same is 

 true of the molecular lowering of vapour tension. 1 



A possible explanation of these facts on the modified hydrate 

 theory will be evident. With increase of concentration, the 

 coefficient of ionisation will become less, and consequently the 

 molecular depression will diminish. Further, owing to hydrate 

 formation, the " free " solvent will be diminished, the effective 

 concentration will be much greater than the apparent, and the 

 molecular depression will be abnormally high — this effect will 

 clearly be greatest in concentrated solutions. The molecular 

 depression may thus be expected to be great in dilute solution, 

 owing to practically complete ionisation, and in concentrated 

 solution owing to hydrate formation, and to attain a minimum 

 value at intermediate concentrations — a deduction which exactly 

 corresponds with the experimental facts. 



It is interesting to note that, according to Biltz, 2 potassium 

 and caesium nitrates do not show this anomaly ; they behave 

 as ideal binary electrolytes, showing a gradual increase of 

 the molecular depression with increased dilution. It is well 

 known that strong electrolytes do not obey Ostwald's dilution 

 law, and it occurred to Biltz to test its validity for solutions of 

 caesium nitrate, as the deviations with ordinary electrolytes 

 might possibly be due to hydrate formation. The coefficient 

 of ionisation at different dilutions was determined by freezing- 

 point measurements, and it was found that the results were in 

 complete accordance with the law of mass action ; the agreement 

 is very much better than for any other strong electrolyte. It 

 may be added that the coefficient of ionisation of this salt 

 calculated from conductivity measurements did not agree with 

 the values obtained by the cryoscopic method, and this speaks 

 in favour of the view which has been expressed more particularly 

 by Jahn, 3 that the conductivity method does not give accurate 

 values for the degree of ionisation. 



1 Smits, loc. cit. 



2 Loc. cit. 



3 Zeit. physikal. Chcm. 1900— 1902, 33, 545 ; 35, 1 ; 37, 349 ; 38, 487, etc. 



