ELECTRICITY. 



447 



tempera- 

 ture. 



Descriptive is proportional to the cube of the weight of the quan- 



Electntity. tj tv of water dissolved in a given volume of ail'. 



*"" "V"' The preceding conclusion, however, cannot be con- 

 sidered as certain, since it depends upon several cir- 

 cumstances which have not been determined with suffi- 

 cient exactness. Coulomb proposed to enclose the balls 

 in different kinds of air, to communicate to the air dif- 

 ferent degrees of humidity, and then to determine the 

 law of the dissipation of the electricity of the balls ; 

 but he found this investigation too arduous, for want of 

 proper instruments for measuring the purity and the 

 humidity of the different gases which he employed. 



The dissi- Although the quantity of water dissolved in a given 

 _iarion of quantity of air increases with the temperature, yet this 

 aty - is not indicated by the hygrometer of Saussure. As 

 " the conducting power of the air, therefore, must in- 

 crease with the temperature, the dissipation of electri- 

 city ought to be greater in warm weather than in cold 

 weather, even when Saussure's hygrometer indicates 

 the same degree of humidity. This conclusion Cou- 

 cmb confirmed by experiment ; but the law of the in- 

 crease depending on the augmentation of temperature, 

 he has not been able to ascertain. Saussure * has pub- 

 lished a Table, shewing the correspondence between the 

 degrees of his hygrometer, and the quantity of water 

 which a cubic foot of air holds in solution at every de- 

 gree of the thermometer. But this Table is not given 

 by its author as correct ; and therefore Coulomb mere- 

 ly concludes, in general, that, in proportion as the tem- 

 perature increases, the dissipation of electricity is not 

 so great as it ought to be, by calculating from Saussure's 

 Table. 



The dissipation of electricity seems also to be affect- 

 ed by some other causes, which it is not easy to dis- 

 cover. On different days, when the barometer, the 

 thermometer, and the hygrometer, indicated that the 

 air was in the same state, the dissipation underwent 

 considerable changes. The only general observation 

 which Coulomb was enabled to make, was, that when 

 there was a sudden change of weather, and when the 

 hygrometer varied sensibly in a few hours from raout 

 to dry, the loss of electricity, relative to its density, re- 

 mained, during some time, greater than it ought to be, 

 according to the degree of dryness indicated by the 

 hygrometer, and vice verta when the hygrometer 

 changes suddenly from dry to moist. If, in 12 or 15 

 hours, for example, the hygrometer passes from moist 

 to dry eight or ten degrees, and that it afterwards 

 remain* stationary at that degree of drvness for several 

 days, it will often happen that the electrical density 

 will decrease the first day after this rise of the hygro- 

 meter, at the rate of T V per minute. Some days after- 

 wards, though the hygrometer is stationary, the elec- 

 tric density will decrease more than -j^ per minute. 

 Coulomb ascribes this variation to the aqueous vapours 

 ixmtractiag aa adhesion to the air, after having conti- 



The dissipa- 

 tion affected 

 by other 



BUECt. 



Method of 

 ascertaining 

 that there 

 was I'.H dis- 

 sipation 

 along the 

 supporting 

 fibre. 



nued in it a certain time ; and he supposes that the hy- nescriprtv 

 grometer will thus only attract to itself the aqueous Electricity^ 

 parts, which are free, and which adhere feebly to the """Y"" 1 

 air. The hygrometer, therefore, upon any sudden 

 change of weather, will indicate only the quantity of 

 aqueous parts in the air that are free, and not the ab- 

 solute quantity of aqueous vapours. This opinion is 

 supported by the fact, that the electrical dissipation 

 is almost always fixed at the end of some hours, rela- 

 tively to the hygrometer, when there is a rapid change 

 from dryness or humidity ; and that it is only in calm 

 weather that the contrary takes place. 



Although Coulomb had satisfied himself that his re- 

 sults were not affected by any dissipation along the sup- 

 porting fibre, yet he made the following experiment to 

 put this point beyond a doubt. He touched the ball by 

 four fibres absolutely similar to those which supported 

 it, so that if there was any electricity lost by the sup- 

 porting fibre, there ought to have been five times as 

 much lost by the five fibres. He could not perceive, 

 however, any dissipation by the five fibres ; so that we 

 are entitled to conclude, that there is no error in the ex- 

 periments arising from imperfect insulation. 



2. On the Dissipation of Electricity along imperfectly 

 insulating Electrics, 



IN the preceding experiments, it was necessary to in- 

 sulate the electrified body in the most perfect manner, 

 in order to obtain the unmingled effect which was due 

 to the contact of the air. In measuring, however, the 

 dissipation arising from imperfect insulation, it is ne- 

 cessary to insulate the balls by electrics so imperfect 

 that the dissipation arising from this cause may be ex- 

 tremely great when compared with that which arises 

 from the contact of the air. But when this relation is 

 very great, the electrified body will lose its electricity 

 so rapidly, that it would be impossible to make seve- 

 ral consecutive observations before the electricity was 

 completely dissipated. Coulomb was therefore obliged 

 to make use of supjwrts, which did not jx^rmit a very 

 rapid diminution, and to determine by calculation the- 

 part which was lost by contact with the air, and that 

 which was lost by imperfect insulation. 



In this way he constructed the following Table, up- PLATE 

 on the same plan as the fonner. The only difference CCXLiT. 

 in the apparatus was, that the ball introduced into the F '6- J - 

 hole m, for the purpose of repelling the ball a of the 

 needle, instead of being insulated as formerly by a 

 small cylinder of gum Lac, from 15 to 1 8 lines long, 

 was supported by a single fibre of silk as it comes from 

 the wonn, and 15 lines in length. The experiment.; 

 were made on the same days with those in the last Ta- 

 ble ; so that, from a comparison of the two Tables, we 

 may readily determine the quantity of electricity dis.ti- 

 puted at every instant along uuporfvvt iusulatoa?. 



On the dis- 

 sipation of 

 electricity 

 along im- 

 perfectly in. 

 sulating 

 electrics. 



fHygnmctrte, chap. T. p. 131. 



