the Unipolar Induction of the Earth. 503 



of one cell of Daniell's battery, and the resistance of the 

 circuit increased to 5000 ohms, by insertion of a rheostat 

 we obtain a deviation of 37 scale-divisions. Thus in the last 

 case the deflection would have been 7 scale-divisions if the 

 resistance of the circuit had amounted to 22857 ohms. Nov/ 

 if we estimate the electromotive force of one Daniell cell at 

 1*079 volt, it follows that the quantity of electricity dis- 

 charged in each of these sparks of 1 centimetre long was not 

 greater than 0*0000118 coulombs. If it were possible to put 

 so much as one coulomb into the jars in question, and if we 

 suppose also that the length of the spark increased in propor- 

 tion to the electric density, we should obtain, with the quantity 

 of electricity mentioned, a spark about 850 metres long. It 

 would in fact be still longer, for the known reason that the 

 distance of discharge for large quantities of electricity increases 

 more rapidly than the tension. Consequently, if the electric 

 tension is very great, it is possible to have a great distance of 

 discharge, but a relatively small quantity of electricity, for 

 one coulomb only contains in reality the quantity of electricity 

 furnished by 1 volt in 1 second, when the resistance of the 

 circuit is 1 ohm. 



Between the earth and the clouds, whose height we may 

 here estimate at 1000 metres, there exists, in consequence of 

 the earth's unipolar induction, an electromotive force equal 

 to that of 23 Daniell cells. If the moisture of the air is 

 great, and its electrical resistance consequently comparatively 

 small, this electromotive force acting upon a surface of vast 

 extent, and during a time counted by hours instead of seconds, 

 ought to be sufficiently powerful to conduct from the earth 

 to the region of the clouds a quantity of electricity corre- 

 sponding to a considerable number of coulombs. 



But this electricity is only of small density. We may 

 observe, on the other hand, that dry air is almost altogether 

 non-conducting, and that it is the moisture which it contains 

 which renders it conducting. Let us suppose now that we 

 have, in the regions of the clouds, a volume of air nearly 

 saturated with moisture, and containing a quantity of elec- 

 tricity represented by a number of coulombs, greater or less. 

 If this volume of air becomes cooled in any way, its aqueous 

 vapour will liquefy in greater or less quantity. When, for ex- 

 ample, the temperature of the air reaches 20°, the vapour of 

 water with which the air is saturated occupies a volume 57000 

 times greater than thatof the water which furnished the vapour. 

 If the temperature is only 10°, the volume of the vapour is 

 1 05,000 times greater than that of the corresponding quantity 

 of water. When liquefaction takes place the electricity is 



