the Point of an Electrified Seedle. 



547 



were no serious defects in the arrangements made. The 

 mean values of these series in terms o£ the deflexion for the 

 central position as unity gave the values of o- for various 

 values of 6, from which was plotted a curve of relative 

 density (curve A in fig. 4). 



Fisr. 4. 



7 



G 



$ 

 ii 4 







^sj-^ 





















i 

 > 

 I 







\ 





: 

 ! 



\ 

 \ 







1 1 





i 





; 





\ 





i 





\ II 



i 

 i 

 i 



i 









\ 



\ 



i 



i 

 i 



f 











\ 



V 



o i 



— 9 



J.4. 1.1(> 



The constancy of the potential of the electrometer-needle 

 was at first tested by a frequent application of a Clark cell 

 which was connected to the quadrants through a switch 

 actuated from the outside of the case C by means of a silk 

 thread. This method, however, involved a disturbance and a 

 possible change of capacity of the proof-plane connexions; 

 and it was found preferable, if somewhat less speedy, to take 

 readings for the central position of the proof-plane alternately 

 "with those for other positions, and to compare each of the 

 latter with the mean of the preceding and succeeding readings. 

 If any considerable drift occurred in the needle's potential 

 the readings were discarded. The throw of the needle was 

 taken in preference to steady deflexions partly in order to 

 obtain larger readings and partly to iessen the lapse of time 

 between discharge and reading, and consequently the possi- 

 bilitv of errors due to slight leakage <£e. 



2P 2 



