and Two New Types of Viscosity. 21 



• are both large, and vary rapidly with temperature, ionization 

 depends upon the relative magnitude of the two forces, and 

 not on the absolute magnitude of either alone. 



From the results established in Tables V. and VI. it can be 

 shown in the following way that at certain strengths of 

 solutions of NaCl and KC1 the value of X attains a maximum 

 at certain temperatures. For NaCl, from Table V. we 

 can write approximately X ?7 = 1? an d from Table VI. 

 (1 - \Ao)/10 WO* = 0'087t;(* o - t)/K. 



So with the data given in these tables we can calculate X 

 for any concentration, say 0*1 N, at the temperatures of 

 ■Noyes and Coolidge's experiments. In order to go a little 

 beyond their highest temperature, I have calculated that at 

 326° C, at which o=l-53 9 K=5*69, 77 = 0-00083. Thus we 

 have the comparison : 



Temp 18°. 140°. 218°. 281°. 306°. 326°. 



Xexper 92 403 577 656 643 — 



Xcalc 79 397 579 679 705 693 



This shows that the theory gives a maximum at about 306° 

 while the experimental maximum is at about 281°. The 

 theory then, besides giving the characteristic minimum 

 of X?7, yields the characteristic maximum of A, at a certain 

 temperature for a given concentration. This maximum has 

 been found for other solutions. Its occurrence is not due to 

 variation of ionization with temperature or to any other 

 chemical cause. 



We have now to take up the study of (12) as to the radius 

 and dielectric capacity of ions. As X =A 01 +A 02 , then for 

 dilute solutions we can write X in the form 



A 01 + Aq 2 



a, = r -1 ' 



1 + f ;V 1 v 2 (n 1 + n^fn 3 



in which j is constant for a given solvent. For dilute 

 solutions this may be written 



X=(A Ql + A 02 ){l-j^v 2 (n 1 + n 2 fni\. . . . (23) 



This shows that, theoretically, the variation of the part con- 

 tributed to X by an ion with concentration depends upon the 

 other ion of the solute. But Kohlrausch, from the inductive 

 study of his experimental data, has been led to assign a definite 

 variation to each ionic velocity with concentration. For 

 example, he gives the K ion at infinite dilution a velocity 64z'6, 

 at concentration O'OIN 63*3, and at O'lN 55*1. Again 



SU37 



