Diffusion of Liquids. 425 



esting peculiarities. For instance, the chlorides, bromides, and 

 iodides of the alkali metals form a series in which NH 4 stands 

 between K and Na. KC1, KBr, KI, and KCy have nearly 

 the same rate. The same can be said of the corresponding 

 NH 4 and Na salts as far as investigated. 



The chlorides of the dyad metals Ba, Sr, Ca, and Mg are 

 also seen to form a series as to their rates of diffusion. In 

 the case of the alkaline nitrates something similar is seen to 

 exist ; but these and other analogies will be better understood 

 when presented in tabular form below (p. 426). 



If we compare the results in the above Table with those ob- 

 tained by Kohlrausch*, on the electrical conducting-power of 

 liquids, several other interesting relations will be observed. 

 It seems to be true in almost every case that those salts which 

 offer the least resistance to the passage of the galvanic current 

 when in solution are the ones which have the fastest rate of 

 diffusion. These analogies will also be found below. 



Another relation will be found when we compare the rates 

 of diffusion with the molecular weights and the molecular 

 volumes of the salts dissolved. In most cases it will be seen 

 that those salts having the greatest molecular volumes diffuse 

 the best. Another very interesting relation will be found 

 when the rates of diffusion of the anhydrous salts are compared, 

 with their thermal action on being dissolved. Here it appears 

 that those salts which absorb the greatest amount of heat on 

 going into solution — that is, those upon which the greatest 

 amount of work has been expended and which are, as a con- 

 quence, in the finest state of division — are the ones which dif- 

 fuse the most rapidly. 



As these several phenomena have in all probability the same 

 or similar causes, it will be well to present the several series 

 of constants, molecular volume, electrical conducting-power, 

 heat-absorbing power, and rate of diffusion of the various sub- 

 stances in a single Table. The specific gravities of the salts 

 from which the molecular volumes are computed were taken 

 from F. W. Clarke's collection of Tables f , and are the mean 

 values of those there given for the substances in question. 



The values for the electrical conductivity are from Kohl- 

 rausch's papers % cited above. They are referred to that of 

 Hg as unit, and are multiplied by 10 7 . The solutions were 

 also normal — that is, similar to those used by me. 



The calorimetric values are those of Thomsen, Bertheiot and 



* Wied. Ann. 1879, Nos. 1 & 2, pp. 1 & 145. 



t " Constants of Nature," ' Smithsonian Contributions to Knowledge.' 

 Washington, 1873. 



t Wied. Ann. Band vi. pp. 1 & 145. 



Phil. Mag. S. 5. Vol. 9. No. 58. June 1880. 2 H 



