Surface Tension and G/iemical Interaction. 417 



no contradiction between the two explanations as regards 

 normal colour vision. This should be obvious without the 

 numerical calculation of particular cases. The Newton 

 colour-diagram is common to both explanations, but the 

 Young-Helmholtz theory makes three puints on the colour- 

 diagram unique, and asserts that they represent primary 

 sensations, whereas my explanation gives no pant a 

 preference over any other. The Young-Helmholtz is a 

 "three elements" theory; my theory depends on three 

 quantities, intensity, s, and k 2 , but these three quantities 

 vary continuously. My theory may be regarded, I believe, as 

 the mathematical formulation of Dr. Edridge-Green's views. 



The two theories lead, to a different classification of colour- 

 blindness. It should be possible to decide between them in 

 this way and also by a statistical investigation of colour- 

 vision. According to my view the colour-blind are simply 

 the outliers of a " homogeneous population," whereas the 

 Young-Helmholtz theory appears to require, for example, 

 that the red blind should be a homogeneous population of 

 their own, I have dealt with this point in a previous 

 paper *. 



I am greatly indebted to Prof. A. Schuster for the benefit 

 of his criticism, which has enabled me to considerably 

 improve some parts of the foregoing paper. 



XXXVII. Surface Tension and Chemical Interaction. 



By Prof. G. N. Antonoff f. 



IN a paper published in Phil. Mag. xxxvi. Nov. 1918, a 

 theory of surface tension was developed. It was as- 

 sumed that electrical and magnetic forces are acting between 

 the molecules, which were treated as electrical doublets or as 

 small magnets, the law of attraction in both cases being the 

 same. 



It is difficult to settle the question as to the nature of 

 molecular forces at the present time. Some evidence indi- 

 cates that these forces are purely electromagnetic in nature. 

 Lewis % has shown that the Obach-Walden relation (propor- 

 tionality of dielectric constant and internal pressure) follows 

 from the hypothesis that the molecular attraction is electro- 

 magnetic, not electric, in nature. On the other hand, the 

 fact that this relation is not wholly correct seems not to 



* Proc. Roy. Soc. vol. A, xciv. p. 576 (1918). 



f Communicated by Prof. J. W. Nicholson, F.lv.S. 



% Phil. Mag. xxviii. p. 104 (1014). 



