Diffusion of Liquids. 4:27 



iodide of the same metal. It is probable that the rate I have 

 found for the bromide of sodium is somewhat too small, being 

 considerably smaller than that found by Graham for the same. 



In the next series it is seen that the order of the salts is the 

 same as regards rate of diffusion and electrical conductivity, 

 and the same in respect to molecular volume and latent heat 

 of solution. Ammonium nitrate precedes potassium nitrate 

 in regard to rate of diffusion and conducting-power, and fol- 

 lows it in regard to molecular volume and heat-absorption. 



Barium nitrate precedes strontium nitrate in every known 

 respect ; to my knowledge, its conducting-power has not yet 

 been determined. 



In the series of the dyad chlorides the order is the same in 

 respect to rate of diffusion, conductivity, and latent heat of 

 solution as far as is known. The molecular volume of stron- 

 tium chloride seems to be slightly greater than that of barium 

 chloride ; but as the values of the specific gravities from which 

 the volumes were calculated differed considerably among them- 

 selves, it is possible that the mean taken does not represent 

 the true value. 



In the case of the sulphates the order is the same respect- 

 ing rate of diffusion and electrical conductivity. The sul- 

 phates of magnesium, copper, and zinc agree with each other 

 very well in both respects ; and it is interesting to observe 

 that the rates of ammonium and sodium sulphates, while also 

 agreeing pretty well in both respects, are much larger than 

 those of the dyad sulphates. The molecular volumes and heats 

 of solution are here quite irregular; indeed the anhydrous 

 sulphates of magnesium, zinc, copper, and manganese stand 

 in exactly reversed orders in respect to their rates of diffusion 

 and latent heats of solution. But a correspondence in this 

 case could hardly be expected, as these substances, on going 

 into solution, combine with large quantities of water and in 

 variable proportions. 



It will be seen also from the Table that Graham was scarcely 

 correct in stating that no relation exists between molecular 

 weight and rate of diffusion. 



Before seeking for an explanation of these several analo- 

 gies, it might be well to cite a few more data bearing on the 

 subject. Kohlrausch has shown, in his paper referred to, that 

 the conducting-powers of the acids HC1, HBr, HI, and HN0 3 

 are nearly the same. It would be interesting to know how 

 these substances act as regards diffusion. I have not at- 

 tempted an investigation of them; but, fortunately, Graham 

 has experimented upon all four ; and his results, reduced to 

 the same standard as above (that is, relative number of mole- 



2H2 



