Vapour-Densities of some Carbon Compounds 



513 



impurity was water, the handling of the sample would have 

 reintroduced moisture, and the purification would have been 

 an illusory one. We therefore rely on results obtained with 

 ether and with toluene in order to show that the deviation 

 from what may be conveniently termed '' theory '^ is real, 

 and not due to the presence of traces of impurity. 



In order that the results of different series of measurements 

 shall be comparable (for the pressures are not the same in 

 any two observations) curves were constructed independently 

 from the observations made with each sample, and the values 

 of /)i' at certain definite pressures were read off from these 

 curves, on which the values of /)t'/T were plotted on a very 

 open scale. We give first comparative results wdth two 

 samples of hexane, or rather with two tubes filled at the same 

 time from the same sample; the observations have been made at 

 two temperatures. This serves to show the accuracy of which 

 the method is capable. 



Hexane. 



Temp. 



Pressure. 



'FvriiT. 





A. 



B. 



' DiflFerence. 



129-60° 

 1 99-71° 



540 

 450 

 390 

 300 



480 

 420 

 360 

 300 



712-90 

 715-16 

 716-65 

 718-93 



710-42 

 712-40 

 714-45 

 716-46 



712-93 

 715-20 

 716-70 

 71900 



71040 

 712-42 

 714-50 

 716-53 



003=1 part in 24000 



0-04 = 1 part in 18000 



0-05 = 1 part in 14.500 



' 0-07 = 1 part in 10300 



0-02=1 part in .35000 

 002 = 1 part in 35000 

 005 = 1 part in 14000 

 0-07=1 part in 10000 



The greatest difference between any two measurements is 

 thus 1 part in 10000. With benzene, the results were ob- 

 tained from two different portions, distilled at different times 

 from the same stock. 



Benzene. 



Temp. 



Pres.sure. 



Pi'//(T. 



A. 



B. Difference. 



129-60° 



99-71° 



1 



680 

 560 

 440 

 320 



680 

 560 

 440 

 320 



784-34 

 787-00 

 78965 

 792-32 



779-58 

 783-02 

 786-50 

 789-96 



784-42 0-08 = 1 part in 9S00 

 787-10 0-10 = 1 part in 7900 

 789-77 0- 1 2 = 1 par t i n 6600 

 792-42 0-10=1 part in 8000 



779-82 0-24 = 1 part in 3200 

 783-28 0-26 = 1 part in 3000 

 786-76 0-26 = 1 part in 30(X) 

 790-23 0-27=1 part in 3000 



Here the limit of accuracy is 1 part in 3000. 

 Phil Mag. S. 6. Vol. 6. No. 34. Oct. 1903 



2L 



