towards a Dynamical Theory of Solution*. 35 



3. Similar Be suits for all the other chief Physical Properties 

 of Mixtures of Water with Ethyl Alcohol and with the 

 first four of the series of Patty Acids, namely Molecular 

 Re fraction and P)ielectric Capacity, Viscosity, Specific 

 Heat and Heat evolved on mixing, Surface Energy. 



As a rule molecular and atomic refraction are not con- 

 venient indicators of molecular change, for the chief 

 generalization in relation to them is that each molecule and 

 atom carries its molecular or atomic refraction unchanged 

 through all physical and chemical changes. But there are 

 notable exceptions to this rule, and in working out the 

 molecular constitution of water the faet that (n 2 — l)/(?r + 2)p 

 for the D line at 0° C. has for trihydrol the value 0*20963, 

 and for dihydrol 0*20434 proved helpful. Moreover, the 

 average value for hydrol as water of crystallization is only 

 about three-quarters of these. These facts suggest that the 

 study of molecular refraction in the mixtures we are dis- 

 cussing should be expected to yield decisive results at once. 

 But this expectation is disappointed in a very interesting 

 manner. Let us first go back to the fundamentals of the 

 subject. When Maxwell discovered his law K = w 2 , he was 

 confronted with remarkable exceptions in the commonest 

 substances, K for water being of the order 80 and n 2 of the 

 order 2. The substances whose mixtures we are studying 

 furnish marked exceptions to the law of Maxwell. I have 

 suggested that in the case of water the large value of K, the 

 ordinary dielectric capacity, is due to the pairs of electrons 

 #[) uniting the H 2 groups in (H 2 0) 3 and (H" 2 0) 2 . The formula 

 for tetravalent oxygen may be written Ot> 3 ^, in which % and \> 

 may form a doublet £[> of small moment and cause O to become 

 practically divalent 0\}. 2 . But on the other hand, £|? may be 

 opened out so that Z joins with \> from another opened out 

 doublet, and the doublets become effective in building up 

 (H 2 0) 3 and (H 2 0) 2 . When electric force acts upon these 

 doublets, the effect is the same as that of magnetic force on 

 magnetic material, and water has a large dielectric capacity 

 for a reason similar to that which gives iron a large mag- 

 netic permeability. Why then does not this large dielectric 

 capacity of water give it a correspondingly large index of 

 refraction of the order 9 instead of the actual 4/3 ? The 

 reason is that these valency electrons on account of their 

 special chemical relation to the atom are constrained in such 

 a way that they take but little part in the rapidly alternat- 

 ing movements of light. In the same way the other 

 exceptions to Maxwell's law are to be ascribed to the 



D2 



