294 



Prof. W. N. Hartley. On the Spark [Jan. 31, 



blue and ultra-violet rays emitted by argon would be recognised 

 among some of the weaker lines in the spectrum of air. When 

 operating in a tube closed by a plate of quartz with oxygen, carbon 

 dioxide, nitrogen, and other gases, with carbon electrodes and 

 points of aluminium, the variations in the spectrum were considered 

 to be of the same nature as those described in the ' Roy. Soc. Proc.,' 

 vol. 55, p. 344, 1894, but it appears now to be extremely probable 

 that they were in some cases due to the removal of nitrogen and the 

 development of the argon spectrum. Accordingly the spark spectrum 

 of air has been examined* for those lines more refrangible than 4674, 

 since, if it be an elementary form of matter, they must appear there, 

 or some explanation should be given to account for their absence. 



It may here be stated that the photographs were taken in 1882 with 

 a Rutherford diffraction grating with 14,000 lines to the inch and 

 quartz lenses of 36 inches focus. Two sets of measurements were 

 made, one from the photograph of each metal ; where the lines did 

 not agree, sometimes a third set of measurements were made from 

 another plate, but when this was not necessary the mean value was 

 adopted as the wave-length. The maximum error for well-defined 

 lines was believed to be not greater than +0'3 of a tenth metre; for 

 faint lines it is larger, but it seldom rises to more than 0'5. The 

 metre of Angstrom was our standard. 



The following table gives the lines in the red and in the blue 

 spectra of argon, and the corresponding lines in the spectrum of air. 

 With the exception of two lines printed in italics, the latter are all 

 absolute wave-lengths. 



Column I. The spectrum of the red rays of argon. 

 II. The spectrum of the blue rays of argon. 



HI. Hartley and Adeney's spectrum of air, with the non-coincident 

 lines omitted. 



* ' Phil. Trans.,' vol. 175, p. 63, 1884. 



