830 REPORT— 1898. 



inconspicuous. There is, however, a line in the green, of approximate wave-length 

 5,030, and another of about 5,400. 



A few words may be said on the other companions of argon. The last fractions 

 of liquefied argon show the presence of three new gases. These are krypton, a 

 gas tirst separated from atmospheric air, and characterised by two very brilliant 

 lines, one in the yellow and one in the green, besides fainter lines in the red and 

 orange ; metargon, a gas which shows a spectrum very closelj' resembling that of 

 carbon monoxide, but characterised by its inertness, for it is not changed by 

 sparking with oxygen in presence of caustic potash ; and a still heavier gas, which 

 we have not hitherto described, which we propose to name ' xenon.' Xenon is 

 very easily separated, for it possesses a much higher boiling-point, and remains 

 behind after the others have evaporated. This gas, which has been obtained 

 practically free from krypton, argon, and metargon, possesses a spectrum analogous 

 in character to that of argon, but differing entirely in the position of the lines. 

 With the ordinary discharge the gas shows three lines in the red, and about five 

 very brilliant lines in the blue; while with the jar and sparli-gap these lines dis- 

 appear, and are replaced by four brilliant lines in the green, intermediate in position 

 between the two groups of argon lines, the glow in the tube changing from blue to 

 green. Xenon appears to exist only in very minute quantity. 



Indeed, all of these gases are present only in small amount. It is, however, 

 not possible to state with any degree of accuracy in what proportion they are 

 present in Atmospheric argon. Of neon, perhnps, we may say that the last fraction 

 of the lightest hundred cubic centimetres from 18 litres of atmospheric argon no 

 longer shows the neon spectrum, and possesses the density of argon ; it may be 

 safe to conclude, therefore, that 18 litres of argon do not contain more than 

 50 cubic centimetres of neon ; the proportion of neon in air must therefore be about 

 one part in 40,000. We should estimate the proportion of the heavy gases at even 

 less. 



It follows from these remarks that the density of argon is not materially 

 changed by separating from it its companions. A sample of gas, collected when 

 about half the liquid argon or about 10 cubic centimetres had boiled off, possessed 

 the density 19-89; the density of atmospheric argon is 19-94. But, of course, we 

 give this density of argon as only provisional ; ' for a final determination the 

 density must be determined after more thorough fractionation. 



With a density of 9-6, and a consequent atomic weight of 19-2, neon would 

 follow fluorine and precede sodium in the Periodic Table ; as to the other gases, 

 further research will be required to determine what position they hold. 



[October 10, 1898. — ^The sample of neon aUuded to above has since been found 

 to contain a small trace of helium. The presence of this light gas has no doubt 

 made the density of neon given in this communication somewhat too low. The 

 actual density has not yet been determined, but the density will obviously not be 

 materially altered. — W. R.] 



2. On the Position of Helium, Argon, Eryjjton, <i:c., in the Periodic 

 Classification of the Elements. Py Professor J. Emersok Reynolds, 

 F.P.S. 



3. Report on the Electrolytic Methods of Quantitative Analysis. 

 See Reports, p. 294, 



4. A new form of Stand Jor Electrolytic Analysis. 

 Py Dr. Hugh Marshall. 



5. Report on the Continuation of the PiUiography of Spectroscopy. 



See Reports, p. 439. 



> July 30, 1898. 



