Volumes of Salt-Solutions. 



187 



If two reactions take place, each with the formation of a 

 definite quantity of barium sulphate and the simultaneous 

 conversion of a corresponding quantity of a sulphate into a 

 chloride, the volume-change should be the same in each case ; 

 for, as I have shown above, the volume of the metal or the 

 acid-radical is the same within narrow limits in whatever way 

 they are combined. If, however, one of the sulphates involves 

 water of constitution, and the other not, then, if the results in 

 Table V. are correct, we should expect to find a difference in 

 the volume- change. 



I had considerable difficulty in devising a form of apparatus 

 by means of which this volume-change, on precipitation, could 

 be measured; after numerous trials I adopted the form given 

 in fig. 1, and afterwards that in fig. 2. 



The apparatus shown in fig. 1 consists of two tubes, A and B. 

 A has a capacity of about 50 cubic centhn., B about 35 cubic 



FiR-. 1, 



<J 



w 



centim. They are connected by a capillary C, as shown, and 

 the upper ends have capillaries as in a Sprengel tube. In 

 making an experiment the apparatus is inverted, and some of 

 the sulphate solution is sucked into B through D — the quantity 

 to be used has been previously measured from a burette (25 

 cubic centim.) ; the tube is restored to its upright position, 

 and D is dried and the whole exactly weighed : then a mea- 

 sured quantity (18 to 20 cubic centim.) of a solution of barium 

 chloride of known strength is sucked into A through E, the 

 air passing through C and up through the contents of B; the 

 tube, must be inclined so that the solution passes down the 

 side of A opposite to C. The apparatus is again exactly- 

 weighed. The difference in weight gives the quantity of 

 BaCl 2 200 H 2 experimented with. E is then washed with a 

 little water which passes into A, and then paraffin-oil satu- 

 rated with water and of known density is sucked in by E till 



