38 Mr. W. Sutherland on Weak Electrolytes and 



rigidity iv as due to the same electric doublets #[> as cause 

 K-l. By "The Electric Origin of frigidity, &c." (Phil. 

 Mag. [6] xvii. p. 417) i(; = (27r/3K)(eV/2 6 a 6 ), in which K 

 may be put equal to 1. In this case then 



K-l = 3a/», (27) 



and s/2a is only about 0*02. Thus, if the rigidity resisting 

 displacement of the average doublet has its origin in that 

 doublet, K — 1 varies inversely as s, that is, inversely as the 

 electric moment of the doublet, whereas if the rigidity is of 

 independent origin, we saw by (26) that K— -1 is directly 

 proportional to s. In intermediate cases we have an inter- 

 mediate relation between these two. These principles then 

 along with those established in "The Fundamental Constant 

 of Atomic Vibration and the Nature of Dielectric Capacity " 

 give a theory of dielectric capacity in general, whether con- 

 forming to Maxwell's law or exceptional. 



Concerning large exceptional dielectric capacities, great 

 importance seems to me to belong to Abegg's discovery 

 of the law of their variation with temperature, a law to be 

 ranked with the corresponding discovery of J. Curie con- 

 cerning paramagnetism. The law of Abegg (Wied. Ann. lx., 

 Ixv., Ztsclir.f. ph. Ch. xxix.) is that the five alcohols from 

 methyl to amyl, nitrobenzol and water have the same value 

 of — ^K/Kc/T from ordinary temperatures T down to the 

 point of solidification, namely 1/190. Probably a better 

 statement would be that — d(K — 1)/(K— l)dT is constant 

 and the same for these- different substances down to tempera- 

 tures near those of solidification. This law makes it appear 

 that the large exceptional value of K — 1 has a similar 

 origin in these different substances, and this origin I take to 

 be the pair of electrons #j? effective in causing association of 

 their molecules. If we neglect the small variations of a with 

 temperature, we get from (27) that — d(K-l)/(K-l)dT 

 — ds/sdT, which by the law of Abegg must be the same for 

 these different substances. If we regard dT as measuring 

 increment of kinetic energy of the oxygen atoms bound by 

 #t>, then the law of Abegg means that the increase of the 

 electric moment es of #t> due to dt is proportional to s and to 

 dT. This gives a clue to the connexion between the potential 

 energy of the doublets §\) and kinetic energy in the molecule. 

 In applying these considerations to our mixtures we shall 

 begin with 



Molecular Refraction. 



Both (n—l)M/p and (w 2 -l)M/(w 2 + 2> may be used 

 as the measure of this. For present purposes too we may 



