RECENT ADVANCES IN SCIENCE 31 



as those marshalled by Dr. Allen. In considering a claim of 

 this kind it must always be taken for granted that the pro- 

 posed model will explain at least one or two facts perfectly. 

 The model has been devised ad hoc. What we have to examine 

 is whether it explains other extraneous facts, and particularly 

 whether it does so without destroying more of the old fabric 

 of science than it builds of new. ... It appears that the ring 

 electron may be welcomed mainly as giving a vivid picture of 

 certain magnetic phenomena. There seems to be no clearly 

 established case in which it successfully explains any pheno- 

 menon outside magnetism, and it is quite out of the question 

 to suppose that the ring electron is going to reconcile the 

 classical dynamics with phenomena which are demonstrably in- 

 consistent with the classical dynamics." 



PHYSICAL CHEMISTRY. By Prof. W. C McC. Lewis, M.A., 

 D.Sc, University, Liverpool. 



Electrolytic Dissociation and the Law of Mass Action. — In an 

 earlier report (Science Progress, April 1916) attention was 

 drawn to an investigation by Schlesinger and Coleman in 

 which the degree of ionisation of certain metallic formates 

 dissolved in anhydrous formic acid was determined. The salts 

 employed were of the univalent type. Anhydrous formic acid 

 is a strongly dissociating solvent, and these electrolytes are 

 largely ionised. Nevertheless the law of mass action, the 

 " Dilution Law," is obeyed with a high degree of accuracy, the 

 extent of ionisation being determined as usual by the electrical 

 conductivity method. 



Recently Schlesinger and Mullinix {J. Amer. Chem. Soc. 

 1919, 41, 72) have extended the investigation to the formates 

 of the divalent metals, calcium and strontium. Again, the 

 degree of ionisation is large. The concentration range ex- 

 amined extended from 0-04 to 0-4 gram-equivalents per litre, 

 It is now found that whilst the law of mass action is obeyed 

 over a considerable range — from o-i to 0-3 gram-equivalents 

 per litre — the law no longer applies to regions above and below 

 these limits. The curve expressing the law consists of three 

 portions. The chief interest attaches to the portion which 

 refers to the most dilute region, for according to these authors 

 the curve shows a sharp break and not a gradual transition 

 in the region of o-i normal. It is suggested that intermediate 



