METEOROLOGY. 423 



indicated au electro-motive force equivaleut in extreme cases to sis 

 Daniell's cells. But a few hundreds of Daniell's cells would be quite in- 

 sufiicient to explain the aui'oral discharge. There is, however, one in- 

 stance of electric phenomena where we have tremendous tensions to» 

 deal with; and atmospheric electricity of tension sufficient, as in ordi- 

 nary lightning", to strike across a mile of air at ordinary density would 

 l^robably be competent to strike across many miles of raretied air. It 

 has long since been remarked that displays of aurorse seem in some way 

 or other in high latitudes to take the place of thunder-storms in low 

 latitudes. Well, then, I will explain what I imagine takes place. I do 

 not enter into any speculation as to the cause of atmospheric electricity. 

 We know as a fact, from its manifestation, that it exists, and tliat is 

 sufficient for my purpose. Suppose, now, that the air, especiallj^ the 

 higher jiortions of the air, over a large tract of country, say to the 

 north of us, were more or less highly electrified — positively or negatively 

 as the case may be — we will suppose positively ; if the electric tension 

 were sufficient, we might, since dense air is a non-conductor, have a 

 discharge taking ]3lace in the higher regions of the atmosphere where 

 the air is rarefied, and accordingly opposes less resistance to the change. 

 "In the figure the great circle P E ^ e is supposed to represent a sec- 

 tion of the earth by a plane i)assing through its center. This faint shad- 

 ing outside represents the atmosphere, the height of course being enor- 

 mously exaggerated. 



Meridian section of the Earth.. 



" Suppose in some way or other a portion of the upper atmosphere, as 

 C, got considerably charged positively or negatively, say positively, it 

 would act by induction on the earth below. The ojiposite electricity. 



