(50 On Salt Solutions and Attached Wider. 



(a) In (1), (2), (4), (5) there appears to be no relationship 

 between the salts. Thermally, the mixture may be considered 

 as a mean. The salt of lower solidifying temperature prevents 

 the solidification of the cryohydrate of its associate. The latter 

 in its turn prevents the extreme cold being quite reached 

 proper to the former when alone. In these cases the mixture, 

 when used as a cryogen, gives a temperature a little above that 

 of its more deeply depressive constituent. When jointly forming 

 a cryohydrate, the complete solidification takes place at a slightly 

 higher temperature than when the more strongly temperature- 

 depressing constituent is alone. The variations from this may 

 not be beyond the limit of observational error. 



(b) In (3) and (7) double salts must be formed whose cryo- 

 hydrates demand a lower temperature for solidification than 

 does either of their constituents. 



(c) Finally, in (8) and (9) the temperature of complete 

 solidification is far above the solidifying temperature of that 

 constituent whose solidifying temperature as a cryohydrate is 

 lowest, where again (in 9) probably a double salt is formed, 

 having, if our analogy hold, a greater water-worth than NaCl 

 and a lesser one than K 2 S0 4 . 



Very remarkable is (6). Prone as the sulphates are to form 

 double salts, Ave might indeed expect, as we find, a specific 

 temperature for the cryohydrate. The difference of tempera- 

 ture of a + b (6) when used as a cryogen ( — 16° C.) and when 

 solidifying as a cryohydrate ( — 7°), a difference of 9° C, has 

 at present its only counterpart in iodide of sodium (§§ 65, 68, 

 69), which as a cryogen has the temperature — 26 0, 5, as a 

 cryohydrate —15°, a difference of 11°'5. Some chemists may 

 not be unprepared for the suggestion that iodide of sodium 

 may be in solution a double salt. To a few general considera- 

 tions bearing on this and kindred points I may have to return 

 subsequently. Meanwhile the general conclusion to be drawn 

 from the above experiments is probably this. Nitrates of the 

 alkalies and alkaline earths act together, but not in union (no 

 double salts); chlorides may act in union with one another or 

 with sulphates. Sulphates may act together or with chlorides 

 or even with nitrates. The agreement of the composition of 

 the several crops of cryohydrates of (3) (§ 113), which points 

 to the relation of almost exactly 3KC1 + 4NH 4 C1, shows that 

 in the society of solid water these chlorides have mutual equi- 

 valents, which, if not in simple relation to their so-called atomic 

 weights, are yet equally definite and constant. 



I have again to acknowledge my indebtedness to Mr. E. 

 Cowper, who has been good enough to give me great assistance 

 in the above experiments. 



