RECENTLY DISCOVERED GASES. 



275 



and rubidium. Such au element should have an atomic weight of 81 

 to 83, which corresponds to a density of 40.5 to 41.5, under the very- 

 probable supposition that, like the other gases of this group, it is mona- 

 tomic. The spectrum of this gas, which we have named " xenon" — the 

 stranger — has many lines; none of these are of marked intensity, and 

 in this respect the spectrum resembles somewhat that of argon. It is 

 also analogous to argon in another particular, that the spectrum under- 

 goes a remarkable change when a Leyden jar is put into the circuit. 

 As with argon, many new blue and green lines appear, while other 

 lines, mostly in the red, either disappear or lose much of their intensity. 

 Further than this we have not proceeded in studying xenon ; for our 

 attenton has been given chiefly to neon, as well as to a problem regard- 

 ing argon. 



We have repeatedly met the question: "Are the properties of argon 

 not appreciably changed by the presence of this new gas ? 7 ' In order 

 to settle this question we have fractioned 25 cubic centimeters of liquid 

 argon several times and have collected separately about 200 cubic centi- 

 meters of the lightest and as much of the heaviest fraction. This 

 operation was repeated three times By this means we hoped to have 

 removed the greatest part of the neon, krypton, inetargou, and xenon. 

 Then we liquefied the argon a fourth time, and as it boiled away col- 

 lected six samples, each after one-fifth of the whole quantity had 

 evaporated. These samples were carefully purified and weighed. The 

 density referred to = 1(3 and the refractivity to air =1 are as follows: 



Fraction. 



Density. 



Refractivity. 



First 



Second 



Third 



Fourth .... 

 Fifth 

 Sixth 



19.65 

 19.95 

 19.95 

 *19. 91 

 19.97 

 19.95 



0.962 

 0.969 





0.968 

 0.966 



'Contained nitrogen. 



The first fraction possesses, as appears from the table, a lower density 

 and also a lower refractivity. The other fractions vary very little from 

 each other. Since these determinations were made by using only 80 

 cubic centimeters, we have weighed 160 cubic centimeters of the fifth 

 and sixth fractions. The first determined density of the fifth fraction 

 was 19.935, but at a pressure of 5 millimeters the spectrum of nitrogen 

 was easily recognizable in a Pliicker tube. After the gas had been 

 again purified by sparking, until all the nitrogen had been removed, 

 the density was 19.957. In two experiments the fourth fraction of gas 

 gave 19.952 and 19.961. We must then accept the true density of argon 

 as not far from 19.96. Independently Lord Eayleigh and I found the 

 density of argon to be 19.94; so it is clear that the impurities of neon 

 and the heavier gases have little influence. The somewhat greater 



