From Current Literature

On the Nature of the Northern Lights

1) G. Cario. Über die Entstehung des Nordlichtsspektrums. Naturwissenschaften, 12, 618, 1924 (H. 30).
2) L. Vegard, Über die Entstehung des Nordlichtsspektrums. Naturwissenschaften 12, 673, 1924 (H. 33).
3) J. C. Mc Lennan. On the luminescence of Nitrogen, Argon and other condensed gases at very low temperatures. Proceed. of the Royal Society A., v. 106, 13, 1924.
4) R. d’E. Atkinson. Note on Vegard’s theory of the Aurora. Proceed. of the R., S. A., v. 106, 49, 1924.

Vegard’s work on the cathodoluminescence of solid nitrogen1 might seem decisive, containing an experimentum crucis on the question of the origin of the northern lights. This, unfortunately, is not so, and the investigations cited in the heading contain certain objections which leave Vegard’s hypothesis, at any rate, still open. Let us recall that the basic proposition of this hypothesis is as follows: the northern light, in its principal part (including the famous line 5577 Å), is emitted by crystals of solid frozen nitrogen under the action of electric β-particles of solar origin. Atkinson gives several considerations showing that, independently even of Vegard’s experiment, the hypothesis is wholly unacceptable. In order that small crystals of nitrogen should exist at great altitude in the atmosphere, there must be electric fields of incredibly great magnitude, the origin of which is impossible to explain. The crystals admissible in the auroral region under Vegard’s conditions must contain only about 60 molecules of nitrogen, while electrostatically the existence of no more than one crystal in 4 m³ is possible. The intensity of the northern lights should correspond approximately to the emission of 10,000 quanta per second for every cm³ of the region occupied by the aurora. Consequently, each hypothetical crystal would have to emit 40 billion quanta per second. Moreover, such small crystals should evaporate completely at every encounter with a sufficiently fast β-particle. The green line of the northern light is very narrow; it may be identified with the characteristic line of the night-sky glow, whose wavelength has been measured very accurately by Babcock2: according to Babcock, the wavelength of this line is \(\lambda = 5577.350 \pm 0.005\) Å, and its width is no more than 0.035 Å. Meanwhile, in Vegard’s experiment, as can easily be seen in the photograph, there appears at the corresponding place a rather broad and blurred line, whence in any case—

1 UFN, vol. IV, issue 2–3, p. 205.
2 Babcock, Astroph. Jour. No. 57, p. 209, 1923.

which makes it possible to conclude that, in nature and in the laboratory, the same line is excited under entirely different conditions. At Lindemann’s suggestion, Atkinson carried out a spectral study of the glow of nitrogen, air, and oxygen in the gaseous state. To obtain a high potential, a high current density, and a good glow, a jet of gas enters a good vacuum maintained by a series of pumps. At the entrance to the gas jet a beam of cathode rays is concentrated. Experiments with pure nitrogen revealed no traces of any glow coinciding with the green line of the aurora. In air it was possible to observe a diffuse line \(5582 \pm 4\) Å and, finally, in pure oxygen a very strong line is found with wavelength \(5587.4 \pm 0.2\) Å. The line consists of two components, the wavelength of one of which is \(5578 \pm 1.6\), i.e. practically coincides with the auroral line.

It is possible that the conditions in the upper layers of the atmosphere favor the strengthening of precisely this component. The band spectrum of gaseous oxygen, as the source of the 5577 line, is also indicated by Cario. He attempts to substantiate the view that some auroral lines must be ascribed to the zero lines of the strongest oxygen bands (Bandenmoleküle). The photographs presented in Cario’s article show very close agreement with Vegard’s spectrograms, and Cario explains the latter’s results by an admixture of oxygen. In his response to Cario’s remarks, Vegard points out that the effect he observed appears especially strongly only in absolutely pure nitrogen, and that an admixture of oxygen only weakens the phosphorescence.

In all the works cited, the slight dispersion with which Vegard’s spectrograms were obtained is persistently emphasized and, consequently, the insufficiency of the identification of the corresponding lines. MacLennan and Shrum, independently of Vegard, carried out analogous experiments on cathodoluminescence of nitrogen, argon, and other gases in the solid state. These delicate experiments were performed very carefully, using spectral apparatus of considerably greater resolving power than Vegard’s. The results proved unexpected. In the glow of solid nitrogen a very strong line \(5231\) Å was found, as in Vegard’s case, but there was no green line \(5577\) Å; instead there is a diffuse group of bands with wavelengths 5556, 5517, 5654. In solid argon there is a strong line 5607.4.

It is still difficult to indicate the cause of the discrepancy between these experiments and Vegard’s experiments. Whether it is due to the poor dispersion of Vegard’s spectral apparatus, or to contamination of his solid nitrogen by an admixture of other substances, or, conversely, to the purity of Vegard’s preparations and contamination of MacLennan and Shrum’s preparations—only new experiments can show. In any case, at present Vegard’s hypothesis raises a number of doubts, and the question of the origin of the aurora remains unresolved.¹

S. Vavilov.

¹ MacLennan and Shrum report the following in the March 14, 1925 issue of Nature, by telegraph: “We observed in the spectrum of a mixture of air and helium (with an excess of the latter) the line \(5577.35 \pm 0.15\) Å. A mixture of oxygen with helium gives the same line with an intensity approximately half as great as that of the yellow helium line.

For the experiments a long discharge tube was used, surrounded partly by liquid air. The best results were obtained at a pressure of about 5 mm. The line was not observed in the spectra of pure oxygen, hydrogen, nitrogen, and helium. No other mixtures of these gases, except a mixture of oxygen with helium, reveal the line.

The line is thin, very sharp, and definite. These characteristics, in connection with the wavelength and the conditions under which the line was observed, indicate its identity with the green line of the aurora.”

Thus the question apparently has been definitively resolved, and not in Vegard’s favor.

Note added in proof.