PllOFESSOR W. RAMSAY AXD DR. M. TRAVERS 
(W3 
Tlie heavier two-thirds of the light gas, and the lighter two-thirds of the heavy gas 
had the same density. 
As the atomic weigiit of krypton depends upon these determinations, the experi¬ 
mental data are given helo^v ;— 
Volume of bull) . 
Weigiit of gas . 
Pressure . 
Temperature . 
Density 
Eractioi) 34. 
7'152 cub. centims. 
0'01515 gramme. 
470'6 mi Him s. 
18-1° C. 
40-82 
Fraction 35. 
7-152 cub. centims. 
0-01521 gramme. 
474 miliims. 
18-4° C. 
40-73 
The refractivities of the two samples with regard to air were next determined ; 
they were found to agree within the limits of experimental accuracy :— 
Refractivity of fraction 34 . 
1-450 
1-449 
We considered, therefore, that since we had succeeded in sepa,rating by fractional 
distillation from a lighter and from a heavier impurity respectively two samples of 
gas whicli agreed in two distinct physical projDerties, we were justified in assuming 
that we had isolated a definite chemical substance. 
The samples 34 and 35 v^ere mixed, and vull from this point be referred to as pure 
krypton. 
Tlie separation of the xenon from traces of krypton proved to be more tedious than 
difficult; lor though their vapour-pressures at the temperature of liquid air differed 
considerably, the krypton appeared to remain dissolved in the xenon, and coidd only 
be removed by repeated fractionation. Fraction 36, the heavier portion from the last 
ex]3eriment, was introduced into the fractionation bulb, where it of course became 
solid, and the krypton was allowed to volatilise into the pump, vdiile the exhaustion 
Avas continued as long as any gas could be removed. The vessel containing liquid air 
was then lowered, and the coiitents of the ljulb was pumped off and mixed with 
fraction 32, the gas of density 61-7 obtained in the previous experiment (p. 65). 
This operation was repeated several times, Avith the result that the greater part 
of the xenon Avas removed from fraction 36. 
Fraction 32 AAms then sul)mitted several times to tlie iiiAmrse operation. The gas 
was solidified in the fractionation bulb, Avliich AA-as tlien placed in connection AAuth the 
pump and exhausted as far as possible. The litpiid air AA as tlien remoA’ed, and the 
heavy gas AA-as pumped off and collected ; the operation Avas repeated, and the gas, 
after sparking Avith oxygen to remoA'e ti'aces of nitrogen and carbon compounds, Avas 
Aveighed. Its density AAms found to be 62-9. After repeating the same operation tA\'o 
or three times the density was again determined ;— 
