272 



MR W. R. BOUSFIELD AND DR. T. M. LOWRY ON THE ELECTRICAL 



to the concentration. This, however, is not the case, as is shown by the differences 

 tabulated in the fourth column of Table VI. Up to 20 per cent, the difference does 

 not exceed O'OOOl, and is for the most part negative in sign ; above this concentration 

 the difference is positive, and increases more or less proportionately from O'OOOl at 

 20 per cent, to 0'0013 at 50 per cent. If the two series of observations had differed 

 only on account of the error introduced by titration, the difference at 25 per cent, 

 should have been half as great as at 50 per cent., and should have amounted to 

 0'00065 ; actually it amounts only to 0'00027, and as the average error of our five 

 observations at this concentration amounted only to '00003, and in the case of 

 PICKERING'S observations to only O'OOOOl, it is clear that the differences are not due 

 exclusively to errors either in the density determinations or in the standardisation of 

 the stock solutions, but must be attributed in part to a difference in purity or some 

 similar cause. 



TABLE V. Density Observations at 18 C. 



NOTES. p = pyknometer, s = sinker. The pyknometer values are the more nearly accurate. The most 

 reliable solutions are those prepared (usually by diluting direct from 50 per cent.) from the mixed solution 

 IV + V + VII. In this series the densities of the solutions were determined by means of two pyknometers. 

 The mean of the duplicate determinations is given above in some cases as if it were a single observation. 

 In no case did these duplicate determinations differ by more than 00003. In series V there was some 

 uncertainty in the dilution from 45 per cent, to 42 '5 per cent., and the concentration of the solution was 

 therefore re-calculated from its density at 40 per cent, and 20 per cent. ; the intermediate observations 

 are of value for interpolation only. 



