508 Sir W. Ramsay and Dr. B. D. Steele on the 



the clip air gradually enters the volume-tube without splash- 

 ing up mercury. When the volume-tube is full of air the 

 cork and thermometer are removed ; any adhering globules 

 of mercury are brushed away from its interior with a hair 

 pencil, and the two pieces of the bulb and the globule of 

 mercury at the top are shaken out on to a clean plate ; these 

 are dried and weighed separately^ and their volume deducted 

 from the volume of the tube. To ensure accuracy the 

 density of the sample of glass used to make the bulb was 

 determined. 



We have given a detailed account of all these operations 

 because unless they are carried out in the manner described 

 the experiment fails. It was only after many failures that 

 the right method of manipulation was gradually evolved. 



Experimental Results. 



A complete record of one experiment is given, with all the 

 necessary corrections, in order that it may be perfectly clear 

 how these are applied. It is, however, unnecessary to quote 

 more than the final figures for the other experiments, because 

 any arithmetical mistakes that might have been made in 

 applying the corrections or in making the calculations were 

 automatically detected on plotting the curves ; many such 

 cases have been found, but on interpreting the results graphi- 

 cally one point was found to lie off the curve, and on again 

 going over the calculations it was always found to be due to 

 a simple arithmetical slip. 



In the annexed tables special sections are devoted to 

 " Compressibility.^'' It will be noticed that the pressures 

 range, in these sections, from 40 mms. to about 200 mms» 

 It would be impossible to weigh with sufficient accuracy, 

 with the volume space practicable in such experiments, a 

 quantity of substance which would yield the small amount 

 of vapour necessary : hence the quantity of substance was 

 only roughly measured. The curve of compressibility repre- 

 senting rise of pv for decrease of pressure, therefore, is not 

 continuous with the density curve ; but if its position be 

 shifted it forms a continuation of that curve. The shift in 

 position is easily effected since the two curves generally 

 overlap. The factor by which the compressibility numbers 

 must be multiplied in order to make the two curves continuous 

 is given in each case. 



