[PASEA] RELATIVE BULK OF WEAK SOLUTIONS 29 
regarded as constant throughout a morning’s or afternoon’s work, pro- 
vided the balance had not been used for weighing a much heavier object. 
In all cases the weighings were performed as above, the counterpoise 
method being employed. 
The weights used were a small set obtained from a trustworthy firm, 
and were made up of brass weights amounting to 20 grams, the smaller 
weights being of platinum. They were calibrated according to the 
method advised by Kohlrausch,' the sum of the weights being put as 
correct. Thus in the calculation of specific gravity all errors due to in- 
accurate weights were eliminated. Also as the corrections were small 
and only. required in a few of the weights, I have great confidence that 
the error due to this cause in the determinations of the concentrations 
would lie within the limits of my experimental error. 
The pyknometer having been carefully dried and treated as des- 
cribed was weighed, the mean of 10 weighings being used. To get the 
weight of water contained in the pyknometer, the distilled water sup- 
plied to the laboratory was carefully redistilled. The weights of water 
when corrected for buoyancy showed a maximum deviation of ‘003 per 
cent for their mean value as calculated from 12 observations. Hence in 
the determination of the specific gravity the possible error would be 
"006 per cent. To reduce this error to ‘005 per cent on an examination 
of the 12 values it was found to be necessary to take the weight of a 
solution as the mean of three determinations. This was therefore done. 
Hence the specific gravity of a solution, and therefore its density has an 
error less than 5 in the fifth decimal place. The density was found by 
multiplying the value of the specific gravity by the density of the water 
at 18°, as given by Landolt and Bornstein.’ 
The water used in making up the solutions was the distilled water 
supplied to the laboratory, which would contain some air dissolved in 
it. But it was found that 16 samples of this water, taken from supplies 
received at different times, gave as their average specific gravity referred 
to the redistilled water ‘99998, the values ranging from 1:00003 to 
*99996, which lie within the limits mentioned above. 
In all cases at least two solutions of any one salt were prepared 
directly, the concentrations of which were determined by analysis, the 
sulphate being precipitated in the form of barium sulphate which was 
collected and weighed. The amount of the original salt was deduced 
by means of the chemica equation, the atomic weights of the elements 
being taken as given by Clarke, quoted in Landolt and Bornstein. The 
solution to be analysed was placed in the bath at 18°, and after a cer- 
tain time known volumes were taken out in a pipette, and analysed, 
The amount of barium sulphate precipitated in one case from a solution 

' Kohlrausch, Physical Measurements, p. 35. 
2 Physikalisch-chemische Tabellen, 2 ed., 1894. 
