208 MR J. Y. BUCHANAN ON THE 



The letters and suffixes have the same significance as in Table L 



The numbers in line T show how uniform the temperature was during the period 

 over which the experiments were spread. All the experiments were made between the 

 12th and 22nd of July 1904, with the exception of those on caesium bromide, which 

 were made on 10th August. By that time the anticyclone had begun to break, and 

 the value of T for this salt is 2r4°. For all the other salts, T lies between 22"8'' 

 and 24-3°. 



During the whole of the period the barometer was very steady, varying between 

 758 and 761 millimetres, and the relative humidity of the air in the laboratory varied 

 between 40 and 50 per cent. 



Of the three values D^, Dg, Dg for the specific gravity of the salt, Di is obtained 

 directly from the first portion of the salt, Dg from the sum of the two portions, and 

 Dg is derived from D^ and Dj by subtraction, 



D.2 represents very nearly the mean of D^ and Dg, and is the accepted value for the 

 majority of the salts. It is expressed to three places of decimals, of which units in the 

 second place are exact. 



It will be noticed that in the case of rubidium chloride the value of D^ is accepted. 

 The second determination depends on the approximate weight of the second portion of 

 salt when the tube was being filled, the exact weighing on the balance of precision 

 having been accidentally omitted. The operation was however completed, and the 

 calculation made with the approximate weight was used as a control. The result shows 

 that the value of l)i may be safely accepted. In the case of potassium chloride the value 

 of Dg (1'951) is accepted, and the reason for this is as follows: The first portion of 

 salt was in very coarse powder, and in mixing it with the mother-liquor numerous 

 crystalline particles were observed which contained gaseous enclosures, easily per- 

 ceptible by the naked eye. As was expected, the observed specific gravity proved 

 to be low. The second portion was much more finely powdered and the specific 

 gravity resulting from the two was higher (1"887). But this result is affected 

 to the full extent by the gaseous enclosures in the first portion. We therefore 

 calculate the specific gravity from the second portion alone, which gives 1"951 for 

 the specific gravity. 



It is an advantage of the method just described that it furnishes more than the 

 mere determination of the specific gravity of the salt. Thus, by ascertaining almost 

 simultaneously the specific gravity of the mother-liquor and the displacement in it of 

 the crystals, both being at the temperature of equilibrium, data are obtained for the 

 determination of the relation between the displacement of the salt in crystal and the 

 increment which it produces in the displacement of 1000 grams of water when it is 

 dissolved in this mass of water and forms a saturated solution with it at that 

 temperature. It has not hitherto been permissible to make exact comparisons of this 

 kind, owing to the independence of the observations on the salt and on the solution 

 which have been available. 



