AERIAL COLUMN AND AEIIIAL ELECTROSCOPE, 



SO 



ference between these tendencies is not sufficient to surmount 

 its weight, and it remains without motion ; but when it is very 

 near each bell, a very small difference of attraction on one 

 side can make it move towards it, whesce it is repulsed. The 

 ditFerence, however, between these attractions may be so small, 

 that the little ball remains undetermined, even at that small 

 distance, though the column has a sensible action. 



21. My plan had been different from tlse beginning, and Tliis did ii^ot 

 thus free from that impediment: it was to obtain a separate ^„ L^^.^, 

 electrometer, formed of a long brass rod with a large ball at the 



bottom, and to suspend at the top, by a conducting thread, a 

 small metallic ball. This small apparatus being connected by 

 its upper part with one side of the coluvin, the litile ball was to 

 diverge ; and I intended to have another large ball in commu- 

 nication eiiher with tjie other side of the column, or 

 ■with the ground, against which the little pendulum should 

 strike, fall, and rise again. This apparatus is represented in 

 the figure annexed to my paper in your Journal for October, 

 1 810. In the same paper, I explained all the difficuhies which 

 I encoutitered, before I could prevent the little bail of the 

 pendulum from sticking to the large ball. At last, however, 

 I succeeded by the means expressed in the figure ; and having 

 determined the distance of the second large laU, at which the 

 pendulum should never cease to strike it by the smallest power 

 of the column, the purpose of the apparatus became to count 

 the numher of the sinkings in z given time; which was the 

 precise indication that I had desired to obtain of the smallest 

 changes happening in the power of theco/.vw??. 



22. This apparatus was ready for observation in the begin- observations 

 ning of April, 1810, and in the rem.alning part of the same ^^'^'^^^ f'-'s ^p- 

 paper I related the phsenomena, which it exhibited daring this 



month and the following month of May. The tables of these 

 observations are composed of five columns : the first indicates 

 the days and parts of the days in which the observations were 

 made. The second, the points at which the barometer stood. 

 The third, the points of the thermometer in the room. I'he 

 fourth that of my hygrometer^. The fifth, the nmnlir of 

 striking^ of the pendulum in determined times. 23. By 



* Tiiis instrumer.t has been taken up by a very jngeulous Hanovciian 



gcntlcr.ian 



