424 SCIENTIFIC RECORD FOR 1882. 



negative in tLat case, would be nccuiiinlated iiuderneatL, as at c, and 

 this portion of tbc earth would form, as it were, a portion of a Leyden 

 jar, the lower atmosphere being the dielectric or glass of the jar, the 

 upper atmosphere being partly the dielectric and j);utly also the charged 

 coatiug. It would be represented more precisely by an imaginary coat- 

 ing outside, com])osed not of tin-foil but of some badly conducting sub- 

 stance. The positive electricity about would be bound down in i)art 

 by the negative electricity about c, which it induces. In another por- 

 tion of the atmosphere — it may be at some considerable distance from 

 the former — you may have the atmosphere charged in an oi)posite way; 

 and of course if this were negative there would be induced positive 

 <3lectricity below, or it might be that the whole of the atmosphere from 

 O to D is charged positively, but at D the negative charge is much 

 feebler than at C. The end result would be the same, but for facility 

 of explanation I will suppose that the upper portion in one place is 

 actually charged with electricity of the opposite kind to what the charge 

 is at the other. If the tension were suflicient then there might be a 

 striking across of the electricity of this name in the atmosphere from 

 C! to D, and in the earth in the reverse direction. Compared with the 

 atmosphere the earth would be an exceedingly good conductor, so that 

 the electro-motive force concerned in sending currents from one part of 

 the earth to another would be, comparatively speaking, trifling, and 

 therefore the electro-motive force represented perhaps by a few scores 

 of the elements of Daniell's battery. Thus, then, in atmospheric elec- 

 tricity we appear to have the tension requisite to send the discharge 

 through a considerable space of rarefied air. Now, if a discharge took 

 place, and if it were night and the sky were clear it would, at least 

 where sufflciently concentrated, be visible to us just the same way as 

 the discharge passing through the exhausted tube is visible by the 

 light it produces. It would produce in fact an aurora. The air is not 

 a comiiaratively good conductor, like a thunder-cloud, from which a 

 great quantity of electricity strikes in one moment, but is a bad con- 

 ductor, so that the electricity can pass only in a spitting sort of way. 

 We may conceive here that we have a sort of double current, yet not 

 forming a complete circuit; nevertheless a discharge would go on nearly 

 of the same nature as if the circuit were complete, and the electro-mag- 

 netic effect of such a discharge on the magnetic needle would be nearly 

 the same as that of a circuit complete. 



" Now, if there only be a suflicient quantity of electricity, we have here 

 the elements necessary for producing a disturbance of the magnetic 

 needle. Moreover, those disturbances, as the instruments show, are of 

 the most fitful and apparently capricious character. They resemble in 

 that the fitful character of electric discharges through air. I need 

 hardly say, according to this theory the earth current consists in tlie 

 return currents produced by the statically-induced charge on the sur- 

 face of the earth, induced by the charged atmosphere above. When 



