366 Mr. W. W. J. Nicol on the 



5° above the point at which the vapour-pressure is to be 

 determined. The temperatures at which determinations were 

 made were 65°, 75°, 85°, 95°, as indicated by the thermo- 

 meter B. The temperatures of the bath were therefore respec- 

 tively 70°, 80°, 90°, and 100°. The temperature of the bath 

 having become constant, the stopcock H is closed and D is 

 opened, communication with a water-pump being thus estab- 

 lished. As soon as ebullition has commenced, D is closed and 

 opened occasionally till the reading of B is one degree or so 

 lower than the temperature of experiment. After steady boil- 

 ing for some time, H is momentarily opened ; the temperature, 

 as indicated by B, rises immediately a tenth or so: in this way 

 the temperature of ebullition is slowly raised till the desired 

 point is reached. The heights of F and Gr are then read, and 

 one observation is completed. D is again opened and closed 

 and the whole operation repeated ; and the mean of the two 

 differences in height of the mercury in F and Gr is entered as 

 the pressure of water-vapour at t° from the salt-solution 

 saturated at t°. Two observations are also made with another 

 quantity of the same salt. Thus the results entered in the fol- 

 lowing tables are based on four determinations. The use of 

 the capillary E is to deaden the oscillation of the mercury in 

 F, which would otherwise be produced by the brisk ebullition 

 in A. Thus the reading of F gives the mean of the momen- 

 tary pressures in A, as the thermometer shows the mean of 

 the momentary temperatures. The piece of granulated zinc 

 or tin is added to obviate bumping ; and it is practically un- 

 attacked, as only a trace of dissolved metal could be detected 

 after even four hours' boiling. 



As these experiments were merely tentative, no attempt was 

 made to correct the temperature-readings nor the readings of 

 the barometer ; and necessity for this was to a great extent 

 removed by making a series of observations with pure water 

 and comparing the results with the figures calculated from 

 Begnault's experiments. 



Table I. contains the figures I obtained from two sets of 

 double determinations. As is seen from it, my results are 

 uniformly lower than Begnault's by 7*25 ±*5 ; but as no 

 attempt was made to read closer than *5 millim., the con- 

 cordance is satisfactory; and it is only necessary to apply as 

 corrections the numbers in the difference-column to the 

 pressures of water-vapour of salt-solutions afterwards given, 

 to obtain the true pressures at the true temperatures. 



In Table II. are given the results I obtained corrected as 

 above, and the temperatures to which these pressures of 



