the Capacity of Gold-leaf Electroscopes. 253 



Discussion of errors'. Taking the general form of the 

 equation in this method 



K'v 



'' = T> 



where K' is the joint capacity of the outer case and the 

 quadrant system, V the charging potential for the gold-leaf 

 system, and v the final potential, we find that the probable 

 error of the result E c is obtained from the equation 



Taking approximate numerical values 



^ 2 =[A o - G ] 2+ B o - 5 ] 2+ [S°- 002 ] 2 



= (-007) 2 + (-005) 2 + (-002) 2 

 or E c = ±0*009. 



The probable error by this method should thus be about 

 the same as in the previous method. 



A set of twelve observations taken on the 27th October, 

 1918, is given below : — Inserting actual values in (1), c the 

 capacity of the gold-leaf is calculated from the equation 



2 do 

 c x 85 = (39-2 + 37-7) T \ x 1-02, 



di, d 2i and d 3 being electrometer deflexions as in the previous 

 tables. 



Table III. 



No. 123456789 10 11 12 



d x 353-2 351-6 350-0 349'4 348'8 347-4 345-6 344-2 3430 342*2 341'4 340 



d 2 273-0 290-0 278-0 288-2 279-8 264-6 272-1 287-3 2696 263-5 278-5 278-0 



d 3 351-6 350-0 349-4 3488 348-0 346-8 344-2 343-0 342-2 341-4 340-0 3390 



c 071 0-76 0-73 0-76 074 0*70 072 077 072 071 0*75 0*75 



Mean 0*735 ; probable error of a single observation = 

 ±0*015 E.S.U. 



A second set of observations gave the mean value of 0*73 ; 

 two other sets, taken by an M.Sc. student, gave mean values 

 of 0*76 and 0*78, with the same probable error, the difference 

 being due to personal equation in reading the final electro- 

 meter deflexions, the damping being such that in this method 

 the final position is rendered rather indefinite by any leakage 



