154 Intelligence and Miscellaneous Articles. 



on the density of the earth led rne to think that this apparatus, the 

 construction and regulation of which are so simple, might be put in 

 action by the weakest forces and give very precise measurements. 



The precautions necessary for obtaining good oscillations con- 

 sist in avoiding some perturbations which present themselves, 

 always the same, and the relative influence of which would be 

 the greater the more delicate the phenomenon to be studied. 



Among those perturbations, M. Cornu and I have already 

 pointed out electric influences and trepidations of the ground. 

 These are completely eliminated by surrounding the apparatus 

 with a metallic case attached to the ground, and setting it up in a 

 very quiet place upon solid pillars. 



I have been exposed to another very serious disturbing cause 

 — the influence of heat. It makes itself sensible upon the tor- 

 sion-thread, and upon the air of the case which surrounds the lever. 

 These two effects are easily distinguished from one another. "When 

 the thread alone is heated or cooled, the oscillations are quite irre- 

 gular and disordered ; if the air alone is subjected to the action of 

 the heat, the oscillations remain nearly pendular, but the point of 

 static equilibrium about which the oscillation takes place is more or 

 less quickly displaced, a] ways in the same direction, while the dura- 

 tion of an oscillation is slightly increased. I protected myself from 

 these very tenacious perturbing influences by surrounding the ap- 

 paratus with a thick layer of wood- shavings. 



The apparatus which I employed was composed of a long torsion- 

 thread (270 centims.) of annealed silver, and a lever 50 centims. 

 in length, carrying at each end a ball of gilt copper of 3 centims. 

 diameter. Similar spheres were fixed to the vertices of a rectangle 

 of 20*5 metres, and communicated with each other in pairs diago- 

 nally. The lever, placed at equal distauce from the fixed spheres, 

 communicated, through the intermedium of the torsion-thread, with 

 the positive pole of a determined pile P, the other pole being to 

 earth. 



The charge of the lever was not so constant as I could have 

 desired ; for piles always undergo variations difficult to define. 

 I was also obliged to measure this charge at each observation. I 

 connected the positive pole of the pile P simultaneously with the 

 lever and a pair of fixed spheres, and took the resulting deflection ; 

 I recommenced by taking the deflection on the other side, and so 

 on alternately four times. 



The pile to be measured, X, was composed of ten equal elements ; 

 and I put one of the poles of this pile into communication with 

 the fixed balls, while the lever remained in contact with P, and the 

 other pole of X was to earth ; then I measured the charge of the 

 lever a second time. 



The deflections were read by means of the reflection of an illu- 

 minated glass scale placed at 3*5 metres from the lever. All the 

 elements of the measurement were known ; and Coulomb's static 



