198 Sir W. and Lady Huggins. On the Spectrum [July 17, 



3814-5 and 3649'6* were not present on the plate. It was clear that 

 the spectrum was not that of the radium molecule when excited by the 

 electric discharge. It was indeed not improbable that if the radiation 

 came alone from the most active molecules, which were suffering loss by 

 material emanations, then if we may accept the analogy from sound, 

 like a filed tuning-fork they would no longer give radiations of the 

 same wave-lengths as before. 



As soon as measures were taken of the lines it was found that several 

 of them agreed in position within the uncertainty of the measures with 

 lines in the spectrum of helium, but not with the most characteristic 

 helium lines in this part of the spectrum. Now Rutherford and 

 Soddy had pointed outf the almost invariable presence of helium in 

 minerals containing uranium. It seemed, therefore, not impossible that 

 we might have to do with helium contained within the radium bromide, 

 and that this gas was being liberated in connection with the active 

 molecules in a state of molecular vibration, analogous to that set up in 

 gaseous helium by an electric discharge. 



In consequence of the wide slit and the small scale of the spectrum, 

 it is not possible to measure with certainty to the fourth figure, but the 

 probable error is, we think, not greater than two units in the fourth 

 place, that is 0'0002. 



[Received August 5. Further examination of the new spectrum, 

 stimulated by the unsatisfactory fact that the strongest lines of helium 

 were not represented in the spectrum from radium bromide, has 

 shown, quite unexpectedly, that if the new spectrum were slightly 

 shifted relatively to the scale, then the seven strongest lines would 

 agree not only in position, but also in relative intensity and character, 

 with bands of the spectrum of nitrogen. 



Now, the wave-lengths of the new spectrum had been found from 

 a fiducial comparison solar spectrum, taken on the same plate. The 

 slit is furnished with three shutters, of which the middle one only 

 was open during the long exposure to the radium bromide. When 

 the exposure was over, the middle shutter was closed, and the two 

 side shutters opened, while the spectroscope was directed for a fraction 

 of a second to the sky. As in both cases the collimator lens was 

 filled with light, it was concluded that the wave-lengths would be 

 correct. It is true that the spectroscope had to be placed in a different 

 position for exposure to the sky, but as the instrument is very short 

 and strongly constructed, it was not considered possible that any 

 shift from flexure could arise. 



The agreement with the bands of nitrogen is so complete that 



* For spark spectrum of radium, see Demar^ay, ' Comptes Kendus,' vol. 129, 

 p. 786, and vol. 131, p. 258 ; Exner and Hasrhek, ' Wien. Akad. Sizber,' vol. 110, 

 July, 1901 ; Eunge, * Astroph. Journ.,' vol. 12, p. 1. 



t ' Phil. Mag./ 1902, p. 582, and 1903, pp. 453 and 579. 



