120 MR J. Y. BUCHANAN ON THE 



The numbers in the columns headed B are abstracted from a series of experi- 

 ments made by Mr S. M. Bosworth in March 1911, using hydrometers Nos. 17 and 3. 

 The series from which these values have been extracted was an arithmetical one with 

 the common difference 1/64 gram-molecule, with the exception of the last three, which 

 formed a geometric series. The method of preparation of these solutions was as 

 follows : — 



The first solution was made by dissolving 1/64 gram-molecule KCl in 1000 grams of 

 water. From this the 2/64 gram-molecule solution was obtained by the addition of 1/64 

 gram-molecule KCl to the quantity of the i/64 gram-molecule solution containing 

 1000 grams of water. 



Thus in each case the more concentrated solution was prepared from the more dilute 

 by the addition of the requisite quantity of salt, while in Mr Royal-Dawson's practice 

 the more dilute solution was prepared from the more concentrated by the addition of 

 the requisite quantity of water. 



Mr Royal-Dawson's results obtained with solutions of chloride of potassium, quoted 

 in the table, were among the first which were accepted as final and admitted into this 

 memoir. Before May 1904 Mr Royal-Dawson's work consisted in learning the art of 

 the exact use of the hydrometer and in assisting me in the elaboration of the plan 

 of making and recording the experiments, which has since been adhered to. Mr 

 Bosworth's results were obtained after he had perfected himself in the use of the 

 instrument and after the system had been in continuous operation for seven years. 

 When it is further taken into account that the two experimenters prepared their 

 solutions in almost opposite ways, it will be admitted that the juxtaposition of the two 

 sets of experiments subjects the method to a very stringent test, and that their agree- 

 ment affords the best evidence of the trustworthiness of the experimental method used 

 in this research. 



