XII. SPECIAL COLLECTIONS. 341 



air of normal pressure, but as soon as it rises to the rarefied air in the Geiss- 

 ler's tubes, there is directly produced a light phenomenon very like the real 

 polar-light. In the apparatus the upper end of the tubes is in union with 

 the earth ; this is by no means necessary, for the light phenomenon is also 

 produced if this union be removed, provided that in such case the tubes be 

 brought a little nearer to the isolated sphere. For the rest, the earth 

 represents here the wide space of rarefied air that we find beyond the limits 

 of the conductor of air, and which serves here as an electric reservoir. 



.Let us now consider how the polar-light on a large scale is formed in 

 nature. As before said, the earth, and the conductor of air, hold to 

 each other the position above mentioned, and the two electricities, the nega- 

 tive electricity of the earth and the positive electricity of the conductor of 

 air, endeavour with a certain force to unite in a belt around the north pole. 

 The isolating power of the denser air prevents this reunion ; but if we assume 

 that the equilibrium is attained, so that the isolating power just cannot be 

 subdued, the reunion will instantly take place as soon as this isolating power 

 is diminished or the electricity on the conductor augmented. The first case, 

 which probably is the most ordinary, happens if a southerly wind carrying 

 a quantity of vapour attains the polar regions ; for instance, the belt, where 

 the vapour, in consequence of the cold, is condensed into a fluid form, re- 

 duces thereby considerably the isolating power of the air and enables the 

 electric current to flow through it. The same thing would occur if a layer 

 of clouds happened to enter into this belt ; the upper end of the cloud 

 would become negatively electric, the lower one positive, and thus the dis- 

 tance between the two conductors would in fact be diminished. The electric 

 current would go from the conductor of air to the cloud, and through this 

 latter to the earth. Similar phenomena are observed in the polar regions, for 

 the upper edges of the clouds are not unfrequently seen shining with a 

 yellowish light stretching considerably upwards, whilst no light is discernible 

 under the cloud because of the air there having attained a density sufficient 

 to prevent the current from producing light. 



For special knowledge of the polar-light and its theory, we refer to essays 

 inserted in the Archives des Sciences Phys. et Natur. de Geneve, 1875 (Sept. 

 and Oct.), and in January 1876, as well as to two essays published in the 

 years 1869 and 1873, in the same scientific journal, all which articles are 

 more or less the result of observations made in the arctic regions. Besides 

 these we may refer also to the works upon polar-light of the American 

 natural philosopher Loomis, Rep. of Smith's Light, 1865, &c. 



