THE AURORA. 203 



The stratum, in which these phenomena occur, must 

 have a certain degree of thickness; its upper surface 

 merging into the region of high vacuum, and its lower 

 surface into that of greater density ; resistance increas- 

 ing upwards and downwards from a central plane. 

 Hence, different phases of electric action must occur at 

 different altitudes; corresponding to the different aspects 

 of electric transmission in high and low vacua, seen in 

 laboratory experiments, as described in Chapter X : 

 which may account for the common auroral appearance, 

 shown in Fig. 58, where the arch seems to form a back- 

 ground for the streamers. And, as there is often a 

 series of concentric arches, as shown in Fig. 61, it is 

 easy to see how streamers might radiate from one arch, 

 across the plane of another arch, at a different altitude. 

 And, if one was below, and the other above the horizon, 

 the appearance would be the same as in Fig. 58. 



Now, since the causes here assigned are in constant 

 operation, we may infer that there should be a constant 

 aurora ; though it does not follow, that it should be 

 everywhere constantly visible. And from the great 

 number of auroras observed in the course of the year, 

 in different parts of the auroral belts, especially in the 

 northern part of the northern belt, it is reasonable to 

 infer, that, with a more perfect system of observation, 

 auroras, of greater or less magnitude, would be seen, at 

 one or more points, every night in the year. 



It is also probable that this electric action may be 

 constant, without being always sufficiently intense to 

 attract attention : and that the aurora is the result of 

 its increased intensity. 



Other atmospheric phenomena, not usually recognized 

 as belonging to the aurora, may also be due to this 



