The influence of the magnet in exciting and brightening the glow and spectrum of one gas, while it depresses and extinguishes the glow and spectrum of another gas in the same tube, suggests an explanation of the observed variation in intensity, and difference in number, of the Aurora-lines. Intensity of lines depending on temperature, and this again on resistance, and it appearing that resistance is influenced by the magnetic action, the same effects of brightening or depressing of the spectrum are probably produced in the Aurora, as in the vacuum-tubes placed between the magnet-poles. In the Marquis of Salisbury’s observations, paraffin-vapour gave C and H lines when connected with both poles of the battery, but C lines only when connected with one pole; and in that case the lines were equally sharp on both sides. These observations (repeated in our experiments) may afford an explanation why the hydrogen-lines are not seen in the Aurora-spectrum; although there can be hardly any doubt that the phenomenon usually takes place in air more or less moist. Professor Ångström’s researches on the violet-pole glow are not entirely corroborated by our experiments; and it seems doubtful whether his results in the exhausted flask were not obtained from the negative pole only. One great difficulty in the comparison of the Aurora-spectrum with the violet pole of air-tubes and some other spectra (including oxygen), arises from the presence in the latter of broad bands; and it is difficult to understand how these bands can be aptly compared with the definite, though faint, lines observed by Dr. Vogel and others in the Aurora-spectrum. It must, too, be borne in mind that the conditions under which we may consider the Aurora to obtain, are such as can be only very imperfectly imitated in the laboratory. Auroræ also no doubt differ in density and thickness of layer; and Kirchhoff’s observation must be remembered:—“That if thickness of a film of vapour be increased, the lines are increased in intensity, the bright lines more slowly than the fainter; and it may happen that the spectrum appears to be totally changed when the mass of the vapour is altered.” Were it possible to test with the spectroscope a cloud or film of gaseous vapour corresponding in some degree in density and thickness with an Auroral discharge, we might perhaps get nearer the truth. Mr. Procter also remarks (as we proved in our magnet experiments):—“That frequently very small traces appropriate to themselves the whole of electrical discharges at low pressures, and completely mask the spectra of any other gases present.” The oxygen-spectrum, with its possible variation by the conversion of that gas into the allotropic condition termed ozone, seemed at first to afford a prospect of close relation to the Aurora-spectrum; which, however, disappeared on closer examination. If nitrogen could be modified in some such way as oxygen is converted into ozone, it might perhaps