ATMOSPHERIC ELECTRICITY, 305 



raised or lowered, a permanent and not a transient charge. Further. 

 M. Erman* has remarked that if we cover the stem of a glass tube; 

 closed above, the phenomena remain the same; positive or negative 

 divergence shows itself equally well, according as the direction of the 

 motion is up or down. The same result is still obtained by placing a 

 second tube around the first, in order more effectually to intercept the 

 contact of the air. Finally, to give to this demonstration every pos- 

 sible accuracy, M. Peltier and M. PfafFt reproduced in a room the 

 same eifects as under a clear sky ; the former experimenting under 

 an insulated globe suspended to the ceiling, and positively electrified: 

 the latter, by alternately raising and lowering tlie electrometer in an 

 apartment, in which the upper strata of the air had been strongly 

 electrified by means of points communicating with the electrical 

 machine. 



It hence follows that an insulated conductor, which has been exposed 

 to the air under a clear sky, is in the same state as if it had been sub- 

 jected to A body positively electrified ; it has consequently its natural 

 electricity se])arated, the negative attracted to its upper part, and the 

 positive repelled to its lower part. If, in this state, the lower part is for 

 an instant placed in connection with the ground, the positive elec- 

 tricity escapes and the upper part remains charged with negative 

 electricity; the conductor, as M. Peltier says^ is equipoised. But then 

 the lower part cannot receive a new positive charge, inasmuch as a 

 new decomposition of its natural electricities would take i)lace, which 

 may happen either in raising the conductor in the air or in favoring 

 the escape of the electricity attracted. The same extremity becomes 

 cliarged, on the contrary, with negative electricity, either in lowering 

 the conductor or in modifying the electrical induction to which it is 

 subjected ; such is the natural consequence of the facts which we have 

 related above. If the application be made to electro-atmospheric 

 apparatus, we see that all such apparatus show the great inconve- 

 nience of indicating the differences of one and the same state, and not 

 of absolute quantities, as in other instruments of meteorology like 

 the thermometer and the barometer ; they establish and measure only 

 the changes which occur in their electric state, as they recede from 

 tlie normal condition, which is that of equilibrium. 



From the preceding considerations, it is easy to give a reason for 

 what takes ])lace in the electro-atmospheric apparatus, and to see 

 how their indications may be interpreted. A stationary apparatus 

 may be compared to an insulated conductor which is in equilibrium ; 

 it cannot, on account of its immovableness, afford signs of electricity 

 at its lower end, except by the radiation of the electricity wbich is has 

 previously attracted, an eftect which takes place, slowly if the air is 

 <lry, and rapidly if it is moist or becomes a conductor by any means 

 whatever. This radiation will be promoted also if the apparatus is 

 subjected to the influence of a cloud powerfully electrified, or if it has 

 considerable length; for in this latter case its extent may make up for 

 the weakness of the local radiation produced by the conductibility of 



'■5 Jouin. de Phys., torn. LIX, p. 100. 1S04. 

 f Dictionn. de Gehler, tome YI, p. 507. 



20 



