432 Mr. R. Sbida on the Number of Electrostatic 



meters (British-Association Report, 1867), and republished 

 along with his other papers on Electrostatics and Magnetism. 

 In measuring an E.M.F. by this instrument, it is important 

 that the potential of the jar or the guard-ring or disk should 

 be kept constant during the experiment. It was observed, 

 however, that the jar w r as losing its charge, though very 

 slowly, on account of the pieces of ebonite in the replenisher 

 insulating imperfectly. Of course I could keep the potential 

 of the jar the same during the experiment by means of the 

 replenisher; but I found it very difficult to work the replenisher 

 and to take at the same time accurate readings. For this 

 reason I thought it better, when the experiment is conducted by 

 one experimenter (or, I venture to think, even w r hen there are 

 more experimenters than one), to proceed in the following 

 manner: — First, connect one pole (say zinc) to the continuous 

 plate, and the other pole to the outside of the jar, and take a 

 reading ; then reverse the poles and take another reading. 

 Repeat the same operation ; that is to say, take a great num- 

 ber of readings by successive reversals. If the experimenter 

 be well practised, the time each reading will take him will be 

 very nearly the same. Let D l3 D 2 , D 3 , &c. be the readings 

 corresponding to zinc, and D/, D/, D/, &c. be those corre- 

 sponding to copper; then the difference of the two readings 

 of zinc and copper would be the difference between the mean 

 of any consecutive readings of one pole and the reading of the 

 other taken between those two consecutive readings — such, for 



example, as — ^ — - — D/ 3 or — — ^ — D 2 , &c. Thus we 



get many values very nearly the same, if not exactly the same, 

 of the true difference in question. If therefore we take the 

 mean of all these, the error due not only to a small loss of 

 charge, but also to a little inaccuracy in the readings, will be 

 avoided. This is the method I used in measuring the E.M.F. 

 of 30 Daniell cells; and the result I obtained is the mean de- 

 fined as above, = 1 3*283 divisions of the micrometer screw- 

 head. As regards the mathematical calculation, we have 



V-V'=2(D-D')^/ 



F 



R*+m 



where V-V is the E.M.F. of the battery, D-D' the differ- 

 ence of the distances corresponding to the readings of the two 

 poles, F the attracting force of the continuous plate on the disk, 

 Kx the radius of the disk, and R 2 that of the aperture. Since, 

 now, one division of the micrometer-screw-head corresponds 



