i 3 o SCIENCE PROGRESS 



sidered to be an abnormal type, and was quoted (Donnan and 

 Bassett, Trans. Chcm. Soc. 1902, 81, 953) as evidence for the 

 existence in the solution of abnormal complex ions. A study 

 of the phenomena in a large number of solvents and over a 

 wider range of temperature led to the conclusion that the 

 inflected conductivity-temperature curve was the normal type 

 for composite electrolytes (Bousfield and Lowry, Proc. Roy. Soc. 

 1902, 71, 53), the curves for the acids (b negative) being shown 

 to be inflected at lower temperatures (Kunz, Comptes rcndns y 

 1902, 135, 788-90), and those for the salts {b positive) at higher 

 temperatures (Noyes and Coolidge, Zcit. phys. Chem. 1903, 46, 

 323) ; the straight lines given by b = o were merely portions 

 of curves which had become momentarily straight in the 

 neighbourhood of a point of inflection. 



Even this, however, did not exhaust the complexity of the 

 curves, which at higher temperatures exhibited a maximum of 

 conductivity and (at least, in some instances) a second point 

 of inflection. All these peculiarities could be interpreted on the 

 assumption that the two properties of the solution — ionisation 

 and mobility — on which its conductivity depends, varied con- 

 tinuously with the temperature, but in opposite directions and 

 according to different laws. In no one case could the whole 

 curve be studied experimentally ; but the formula 



k=k o (\+bt)" e - at 

 (Bousfield and Lowry, Proc. Roy. Soc. 1904, 72, 281) was found to 

 give expression to all the above peculiarities, and this in spite 

 of the fact that the use of a bent lath or of parabolic formulae 

 would have necessitated the splitting up of the curve into five 

 or six separate sections. 



Curves of a similar type, each containing a point of inflection 

 and a maximum, are obtained by plotting conductivity against 

 concentration in mixtures of nitric acid and water, acetic acid 

 and water, or acetic acid and pyridine ; it is therefore not 

 improbable that these curves "also may be represented by 

 comparatively simple equations. The curves for mixtures of 

 amylamine or butylamine with acetic acid and for aniline with 

 acetic acid (H. E. Patten, Journ. Phys. Chcm. 1902, 6, 554), on 

 the other hand, are obviously distorted owing to the formation 

 of molecular compounds, and must be represented by equations 

 of greater complexity. 



As a further illustration of the danger which undoubtedly 



