502 Prof. E. Edlund on the Magnitude of 



disturbing effect. It is much to be regretted that observa- 

 tions upon the electricity of the air are two few and imperfect 

 to permit of a more detailed comparison between theory and 

 experiment. 



The quantity of electricity which, in accordance with the 

 preceding, should exist in the earth's atmosphere in conse- 

 quence of the unipoiar induction of the earth is sufficient, in 

 my opinion, to account for the indications of the electric 

 condition of the air, given under normal conditions by the 

 instruments employed for this purpose. It is true that the 

 electric tension of the lower layers of the air is not great 

 under normal conditions ; but it increases with the height 

 above the earth's surface, so that the quantity of electricity 

 contained in the atmosphere is considerable in all cases. It 

 is by the induction of the atmospheric electricity that the 

 instruments are affected, and not, or only to a trifling extent, 

 by direct communication with the layers of air in immediate 

 contact. The magnitude of the electric tension shown by the 

 instruments does not depend, then, or at least only to a very 

 small extent, upon the electric tension of the layers of air 

 close to them. The electric tension indicated by the instru- 

 ments may therefore be many times greater than that of the 

 layers of air in question. 



§6. 



The origin of the large quantities of electricity which the 

 atmosphere appears to contain at times of storm may also be 

 explained without difficulty by the unipolar induction of the 

 earth. Although in a lightning-flash the electricity possesses 

 an enormously high tension, it is nevertheless easy to show that 

 the quantity of electricity which takes part in it may be less 

 than there is a disposition to admit in certain quarters*. 



A galvanometer with mirror-attachment, especially con- 

 structed for electric discharges, was combined with a Holtz's 

 electrophorus, of which the two jars had each an exterior 

 and interior covering of 79 square centimetres. With one 

 rotation per second, the galvanometer indicated a mean devia- 

 tion of 7 scale-divisions. Four electric sparks of a length 

 of 1 centimetre passed between the terminals per second. 

 The galvanometer being then connected with the electrodes 



* Compare what is said upon this point by M. W. Siemens {Sitzungs- 

 berichte der Akad. der Wissensch. zu Berlin, May 1883, p. 638 ; Phil. 

 Mag. xvi. p. 173) : — u The sudden appearance of such enormous quantities 

 of electricity, as seen in tropical storms for example, negatives the hypo- 

 thesis that it can have been present in the feeble electric charge of the 

 comparatively small quantity of air which carries the thunder-cloud. 

 There must be more important sources from which they are derived." 



