106 Frederick Guthrie on Salt Solutions 



matter of the determination of water of crystallization, analysts 

 have for the most part allowed themselves a far greater lati- 

 tude in respect to agreement between the experiment made and 



the conclusion drawn than they have been willing to admit in 

 regard to the other constituents of the salt. Prominent in 

 respect to agreement between experiment and the derived 

 constitution are the investigations of Graham, especially on 

 the water of crystallization of single and double sulphates. 

 Such accord is so rare, that in a very great many instances 

 the experiments actually point to a different water- worth than 

 that adopted by the experimenter. 



That heat is sometimes liberated and sometimes absorbed 

 when a salt is brought into contact with water, not only ac- 

 cording to the nature of the anhydrous constituents, but also 

 to the degree of hydration, has long since shown that there is 

 some essential difference in the tension of the union which is 

 established between the water on the one hand, and the more 

 or less hydrated anhydride on the other. But the statements 

 as to the conditions, under which a salt becomes anhydrous, or 

 exists in combination with a definite relative number of mole- 

 cules of water, are neither definite nor satisfactory. The state- 

 ment that a salt gives up n molecules of water when heated to 

 the temperature E is inexact, (1) unless the hygrometric state 

 of the air is given, (2) unless it is known whether free circu- 

 lation takes place, and (3) unless the pressure on the salt is 

 known. The statement that a salt gives up m molecules of 

 water in vacuo (over a desiccator) is also ambiguous, in so far 

 as it ignores the temperature. 



The salt which I first examined in this respect was chloride 

 of barium, BaCl 2 + 2 H 2 (BaCl + 2 HO), it being a salt easily 

 got quite pure and of a stable nature. The "pure" salt of 

 commmerce was recrystallized, boiled with carbonate of barium, 

 filtered and precipitated, and washed with alcohol. It was 

 then twice recrystallized. It was then finely powdered, and a 

 part A was dried for forty-eight hours in a good vacuum at a 

 temperature of about 17° C. ; a part B was dried between 

 repeatedly renewed bibulous paper in a screw-press for the 

 same time. Two analyses of each w T ere made, the elements 

 being estimated in the usual way. 



Estimation of Barium. 

 Weight of c i i - T5 • Per cent, of 



substance. Sul P hate - Bal ' lum ' barium. 



2-6912 1-57061 56-563 



1-3406 0*78825 56-560 



1-3445 0-79054 56*703 



0-8480 0-49857 56-360 



IS : 



. 2-7768 



. 1-3997 



B(l) • 



. 1-3942 



B(2) . 



. 0-8847 



