Royal Society, 537 



by the development of heat in conductors, be augmented a^ in- 

 finitum. 



In this manner an example makes clear how important it is 

 that the analytical expression of the vis viva should contain only 

 positive terms ; and that this condition is not fulfilled by the 

 action at a distance of Weber's law is here exhibited as the last 

 cause of the physically impossible consequences to vvliich it leads. 

 These, at all events, cannot be removed without very bold new 

 auxiliary hypotheses, which must not only vary the actions at 

 molecular, but also those at greater distances. 



In conclusion I have, in the present memoir, endeavoured 

 to clear up the doubts expressed by M. J. Bertrand* respect- 

 ing the structure of the differential equations of the motion 

 of electricity. 



LXIV. Proceedings of Learned Societies, 



ROYAL SOCIETY. 

 [Continued from p. 473.] 



June 20, 1872. — Sir James Paget, Bart,, D.CL., Vice-President, In 



the Chair. 



'T^HE following communication was read : — 



-^ *' On the Spectrum of Nitrogen." By Arthur Schuster, Stu- 

 dent at the Physical Laboratory of Owens College. 



1. Introductory . — The formation of the different spectra which 

 one gas is said to exhibit, when examined under different conditions, 

 still remains one of the most obscure points of spectrum analysis. 

 In 1864, when Pliicker and Hittorf published their researches **0n 

 the Spectra of Ignited Gases and Vapours, with especial regard to the 

 different Spectra of the same elementary gaseous substance "f, they 

 drew attention to the close resemblance in character of the band- 

 spectra which certain metals yield at a comparatively low tempe- 

 rature to the band-spectrum of nitrogen and sulphur. Roscoe and 

 Clifton, in their paper *'0n the effect of increased Temperature upon 

 the nature of the Light emitted by the Vapour of certain Metals 

 or Metallic Compounds" J, rendered it probable that the band-spectra 

 of the metals belonged really to the oxides. The two spectra of 

 nitrogen were not, however, examined from that point of view, but, 

 on the contrary, they were made the starting-point of new investiga- 

 tions by Wiillner, who came to the conclusionQthat certain gases may 

 give even more than two different spectra. Angstrom §, expressing 

 his doubts about the trustworthiness of Wiillner's experiments, says 

 in a note: *'As regards the spectra which are usually attributed to 

 nitrogen, I mention here, as a general fact, that it is my conviction 

 that the fluted bands which are so characteristic of the oxides of 

 metals are never found in spectra of elementary gases." 



I propose to show, in the present communication, (1) that pure 



* Comptes Rendus de VAcad. des Sciences, vol. Ixxiii. p. 968. 



t Philosophical Transactions, vol. civ. p. 1 . 



I Chemical News, vol. v. p. 233. § Comptes Rendus, Aug. 1871, 



