Electric Residue in the Ley den Jar. 319 



ordinates of which, however, have the line be as abscissa. Set- 

 ting, on the contrary, by the observed L^ the corresponding ?-y 

 we obtain a series of points, which, on being united, constitute 

 the line Q^ or v^, which is almost straight. The ordinates of 

 this line, referred to ad as abscissa, represent the quantity Q< oi 

 electricity present in the jar at the time t ; and, when they are 

 referred to be, they represent loss of electricity v^. 



§5. 



The question now naturally presents itself, in what relation 

 does the concealed residue of the same jar stand with respect to 

 the magnitude of the charge first communicated ; for it is soon 

 evident that a stronger charge produces a greater residue. 

 Whether the residue produced in equal times, but with different 

 charges, be proportional to the strength of these charges, may 

 be solved by ascertaining whether the sinking of the disposable 

 charge follows the law of proportionality; for the air, retaining 

 its quality, the electric loss from this source has this property. 

 On this point a single decisive experiment may be cited. 



The method described in Appendix 1. of communicating to a 

 jar instantaneously a known charge, may be easily so applied 

 that the charge in one case is exactly ten times what it is in 

 another case. This was done on two successive serene days, 

 during which the warmed air of the room remained in a constant 

 condition as regards the loss of electricity, a jar being chosen 

 with which the loss was inconsiderable. The results are stated 

 in the following small tables d and e. In table d' the results 

 given in d are reduced to the times of table e. 



The same experiment has been repeated several times with the 

 same degree of coincidence. 



We have thus arrived at the conclusion, that ivith the same jar, 

 the reaidues funned in the same time are proporiiunal to the charge 

 imparted at the commencement. It must be left undecided whether 



