7G 



PUS. J. A. BARKER AND P. CHAPPUIS ON A 



Fig. 18. 



variations in the capillary angle have an appreciable effect on the volume of the gas 

 above the mercury. The extent of the second part on the side of the reservoir is also 

 somewhat uncertain, because of the rapid variation of temperature 

 near its end, but as the capillary tube has only a very small 

 diameter, the influence of this cause of error is not great. 



In order to avoid the uncertainty of any hypothesis concerning 

 the capillary angle under the actual circumstances, we attempted 

 to measure the total volume of the " dead space " directly in the 

 following manner. 



The capillary tube joining the thermometric reservoir to the 

 manometer being straight, the closed branch of the manometer 

 was fixed on its support in the position it afterwards had to 

 occupy (fig. 18), and the open extremity of the reservoir was 

 connected to the mercury pump. The side tube, S, of the 

 manometer terminated in a tube bent downwards, whose lower 

 end was about on a level with the point. The tap, R, was 

 placed in communication with the auxiliary mercury reservoir 

 forming part of the gas thermometer. The tube, m, was first 

 exhausted, and then filled up with mercury to near the steel 

 point, and the tube, s, completely filled ; then air was readmitted, 

 and the mercury was adjusted to the point by slightly dis- 

 placing the reservoir. The taps, R and t, being then shut, the 

 whole was again exhausted. A small, carefully weighed vessel 

 containing mercury was placed under the tube, s, and by opening 

 the tap, t, mercury was allowed to enter the manometric tube 

 and fill all the space above the point, rising to the level, c, 

 which is at the barometric height above the mercury in the 

 weighed vessel. The volume of the " dead space " could then 

 be deduced from the loss of weight of the small vessel. It 

 should be remarked that the pressure in m at the end of the 

 experiment was very nearly the same as at the beginning, and 

 all uncertain corrections were thus avoided. The divergence 

 found between the individual observations given are a fail- 

 measure of the inevitable variations of the " dead space " during Determination of 

 the experiments. the volume of the 



** flpAfl STW1PP '* 



If the point, c, did not coincide exactly with the limits of the 



" dead space," it would be easy to take account of the difference, the volume of the 

 capillary tube having been previously measured. 



The following values for the volume of the " dead space " were obtained by the 

 above method ; 



