Chapter XIX. Effects of magnet on discharges _in vacuo_ in larger vessels demonstrated. Ångström’s flask experiment tested; same results not obtained unless one wire only was connected. Experiments demonstrating the action of a magnet on an electric stream, viz. vibration between, and rotation round, poles. Baron Reichenbach’s magnetic flames tested without result. CHAPTER XX. SOME CONCLUDING REMARKS. It is usual, in concluding a work on a special subject, to sum up its contents, and to examine the general results arrived at. This, however, it is not easy to do in the present case. The contents of our volume comprise a short history of the Aurora, its qualities and spectrum; and a statement has been given of the several conclusions at which various observers have arrived as to its character and causes. In the present state of our knowledge of the subject, to add an opinion to these might seem to savour of presumption; and the questions involved may perhaps be better treated as still _sub judice_, and as requiring further and fuller evidence before arriving at a verdict. The following observations must therefore be taken rather as further notes and memoranda, than as conclusions. Apart from the spectroscopic questions involved, the oldest and most received theory of the Aurora—that of its being some form of electric discharge in the more rarefied regions of the atmosphere,—seems to hold its own: and if, as is probable, some form of phosphorescence is involved in the discharge, M. Lecoq de Boisbaudran’s observations on the brightening of the red line under the influence of cold, and the falling of the yellow-green line within a band of phosphoretted hydrogen, come into play; and a connexion, though slight and imperfect, may be in this respect traced between the discharge and its spectrum. The experiments detailed in Part II. seem to have an important bearing, as showing the very marked effect of the magnet on the rarefied glow, as well as on the spark in air at ordinary pressure. The well-defined arc formed by the aura of the spark, the flickering jets which replace the even edge of the arc when partially withdrawn from the magnetic influence, and the streamers formed when the aura is blown away from the spark (Plate XVII. figs. 6, 7, and 8), are certainly highly suggestive of frequent forms of Auroral discharge; and, but for trial and failure, might lead one to expect results from a comparison of the line air-spectrum with that of the Aurora. The experiments with a wire attached to one electrode only, show how the glow may be affected and varied in colour and character when the discharge is interrupted and incomplete. Differences in electric tension may also considerably vary the character of the discharge.