8io 



NA TURE 



[June 24, 1922 



Letters to the Editor. 



{The Editor does not hold himself responsible for 

 opinions expressed by his correspondeiits. Neither 

 can he undertake to return^ or to correspond with 

 the writers of, rejected fnanuscripts intended for 

 this or any other part of Nature. No notice is 

 taken of anonymous communications !\ 



On the Continuous Radiation found in some Celestial 

 Spectra beyond the Limit of the Balmer Series of 

 Hydrogen. 



In the account of his observations of the eclipse 

 of January 22, 1898, pubhshed in the Philosophical 

 Transactions, A 197, pages 389 and 399, Mr. Evershed 

 directed attention to a curious continuous spectrum 

 emitted by the solar chromosphere and prominences. 

 This spectrum begins near the limit or head of the 

 Balmer series of hydrogen lines, and extends with 

 gradually decreasing intensity in the direction of 

 shorter wave-lengths. In describing the phenomenon 

 Evershed referred to the early observation by Huggins 

 of an absorption in the corresponding region in the 

 spectra of Vega and other stars having especially 

 strong hydrogen lines (stars of Class A),^ and advanced 

 the opinion that the spectrum, like the Balmer series 

 which it so curiously supplements, is due to hydrogen. 

 The grounds for this view were afterward strengthened 

 through the discovery by Wood of a continuous 

 spectrum occurring beyond the limit of the sodium 

 series of dark lines, under conditions of laboratory 

 experimentation that favoured the development of 

 the higher members of the sodium series.^ More 

 recently an emission spectrum, apparently identical 

 in character with the one observed by Evershed in 

 the chromosphere, has been found to be characteristic 

 of the planetary nebulae.* The spectrum seems also 

 to occur in the diffuse nebula N.G.C. i499,* and has 

 been a conspicuous feature in the radiation of the 

 novae. It may therefore be regarded as rather a 

 commonplace phenomenon pertaining to the spectra 

 of celestial objects which appear to exist under 

 conditions of strong thermal or electrical excitation. 

 For the purposes of this note I shall use the term 

 outlying spectrum in referring to it in order to dis- 

 tinguish it from the general continuous spectrum of 

 more uniform distribution which is found even in the 

 gaseous or " bright-line ".nebulae. 



The outlying spectrum, as an emission phenomenon, 

 has not, that I am aware, been observed in the 

 laboratory, except possibly by Dufour,^ who found 

 an ultra-violet continuous spectrum which, however, 

 he associated with the secondary spectrum of hydrogen, 

 and not with the Balmer series. In view of its general 

 occurrence in a large group of extremely interesting 

 and important celestial spectra, it would seem 

 desirable that serious effort be directed toward its 

 development in the laboratory, with the view of 

 establishing the circumstances of its origin. As a 

 preliminary step in estimating the conditions likely 

 to prove favourable for its emission it may not be out 

 of place to recall a theoretical explanation that has 

 been advanced to account for it. 



In a paper read two years ago before the American 

 Philosophical Society,* the present writer suggested 

 that the spectrum might be explained on the basis 

 of the Bohr theory, as resulting from the capture of 



• " An Atlas of Representative Stellar Spectra," p. 85. For a more 

 complete investigation of the absorption in the Class A stars, see Hartmann, 

 Physik. Zeit., 18, 429, 1917. 



^ Astrophys. Jour., 29, 100, 1909. 



' Lick Obs. Bull., 9, 54, 1917 ; also Publ. Lick Obs., 13, 256, igiS. 



• Hubble, Publ. Astron. Soc. Pacific, 32, 155, 1920. 

 ' Ann. Chim. et Phys., (8), 9, 361, 1906. 



• Proc, 49, 530, 1920. 



NO. 2747, VOL. 109] 



free electrons by hydrogen nuclei. Although the idea 

 was developed from independent considerations, it 

 cannot be regarded as an original one, since Bohr, in 

 his first paper,' had suggested the reverse process 

 to account for the sodium absorption referred to above 

 as having been found by Wood. Inasmuch, however, 

 as he had not applied the consideration to the explana- 

 tion of an emission spectrum in that region (such 

 as was known to exist in the spectrum of the chromo- 

 sphere), it seemed appropriate at the time to direct 

 attention to that aspect of the theory, and I venture 

 to revert to it here, both on account of its theoretical 

 interest, and because it may possibly prove suggestive 

 of the conditions under which the spectrum should be 

 sought in the laboratory. 



Briefly outlined, the argument is as follows : The 

 mechanism which Bohr sets up for the hydrogen 

 atom is a positive nucleus surrounded by an infinite 

 series of fixed orbits in which it is possible for an 

 electron to revolve. A line of the Balmer series is 

 formed when an electron " jumps " from one of the 

 outer orbits to the second one, and the complete series 

 of lines results from the totality of transfers or jumps 

 from all of the outer orbits. The frequency of 

 vibration is in each case proportional to the energy 

 set free in the transfer from the outer to the second 

 orbit, and the frequency at the head, or limit, therefore 

 corresponds to the energy set free in the fall of an 

 electron from rest at infinity. An electron with an 

 additional velocity of its own will of course give up 

 more energy in its capture than one starting from 

 rest, so it will have to be manifested by a radiation 

 of proportionately greater frequency, that is to say, 

 of less wave-length, than that of the series limit. 

 The exact position of the resulting line will depend 

 upon the amount of the initial kinetic energy, and 

 since this must, in general, vary from one electron 

 to another, the totality of these radiations should 

 make up a continuous spectrum beginning at the 

 Balmer limit and extending into the ultra-violet. 

 It is thus seen that, by this theory, the Balmer 

 series is caused by the falling inwards of electrons 

 forming part of the atomic system, while the outlying 

 spectrum is due to the capture of extraneous electrons. 

 It is scarcely necessary to point out that the reasoning 

 applies with equal force to the Ritz and Lyman series, 

 and to other series of like character. 



Preliminary to discussing this explanation it is well 

 to calculate the initial kinetic energy required to 

 produce the observed spectrum. The outlying 

 spectrum can be followed in a number of planetary 

 nebulae to about 3340A. This is of course an 

 extremely rough estimate, for the spectrum is faint, 

 and on this side fades gradually to invisibility. The 

 energy required to develop a line here may con- 

 veniently be expressed in terms of that needed at 

 the Balmer limit (3646A) as 



365/334 X "1^ = 365/334 X NA/4. 

 where v^ is the frequency at the Balmer limit, N the 

 series constant, and h is Planck's element of action. 

 Of this energy the amount NA/4 has been developed 

 in capture, leaving the remainder as the original 

 kinetic energy of the electron. Calling this 'Ek, we 

 have 



EA = 31/334 xNA/4, 

 substituting the numerical values: h = 6-547x10''" 

 and N = 3-290 X 10+^^, 



EA = 5-00x10-" (i) 



This is the amount of kinetic energy which must have 

 been possessed by the electron before coming under 

 the action of the capturing nucleus in order that 

 it should be able to develop a line at 3340A. 



' Phil. Mag., 26, 17, 1913- 



