Expansion of Salt- Solutions. 397 



maxima, but the other solutions have maxima lying lower on 

 the temperature-scale the stronger the solution ; thus: — 



6NaN0 3 , 90°-95° ; 8Na]N T 3 at 90°. 

 10NaNO 3 , 80°-85° ; 12NaN0 3 at 80°. 

 3KN0 3 , 80°-85°; 5KN0 3 , 75°-80°. 



Again, with 8 and lONaCl and 3, 5, and 7KC1 the volume 

 of the solution at 100° C. is less than that of water ; but such 

 is not the case even with the strongest solutions of the other 

 two salts. Comparing the volumes of the various solutions of 

 the same salt at 100° C., we find that with NaCl and KC1 the 

 stronger the solution the smaller the volume ; but with KN0 3 

 the reverse is the case ; while NaN0 3 forms a connecting- 

 link, the weaker solutions 2, 4, and 6 behaving as those of 

 KN0 3 ; the stronger, 8, 10, and 12, as those of KaCl or KC1. 



This is all that can be gathered from the results in the 

 above form, and it is clear that the conclusions arrived at by 

 former experimenters are in the main correct. The results of 

 Kremers with solutions of the same salts give figures which, 

 where the solutions are the same strength, completely agree 

 with mine at low temperatures, though the discrepancy is 

 marked at high temperatures ; the cause of this being probably 

 in the use of too high a correction for the exposed dilatometer 

 column (see Thorpe, Chem. Soc. Journ. 1880). 



The questions now remain to be considered : Is it permis- 

 sible to compare together equal volumes of solutions of 

 different strengths ? Is it probable that such a comparison will 

 lead to any conclusion that can be in any sense considered 

 general ? A little reflection will show that the answer must 

 be no. For equal volumes of various solutions of a salt con- 

 tain more salt and less water the more concentrated the 

 solution ; and when solutions of different salts are brought 

 into the comparison, the proportions of water and salt vary 

 according to the molecular volumes of the various dissolved 

 salts. The bearing of the expansion of salt-solutions on the 

 question of solution must necessarily be obscure until mole- 

 cular, not unit, volumes are compared ; then, and not till then, 

 can we reasonably expect to gain trustworthy information. 



In order to convert results for equal volumes into those for 

 molecular volumes, it is only necessary to multiply by the 

 molecular volume of the salt-solution at 20° C, which is 



M>Yt= 1800 + nM.W. . 



