in  a  Discharge  betiveen  Parallel  Plates. 
737 
No  capacity 
attached  toW. 
Inner  plate  P 
joined  to  W,  and 
outer  plates  Q  to 
negative  electrode. 
Inner  plate  P 
joined  to  W,  and 
outer  plates  Q,  to 
positive  electrode. 
Potential-difference   between 
wire  and  positive  electrode. 
102 
178 
82 
Potential-difference    between 
wire  and  negative  electrode. 
270 
194 
285? 
Sum  
372 
372 
367? 
In  the  last  experiment  the  pointer  of  the  voltmeter  was 
considerably  over  the  scale,  so  the  potential  was  determined 
approximately  with  a  graduated  aluminium-leaf  electroscope. 
The  potential  must  of  course  in  this  case  be  290  (372  —  82), 
since  the  difference  of  potential  between  the  wire  and  the 
electrode  is  not  affected  by  the  voltmeter  as  the  connexions 
to  it  are  made  through  high  resistances. 
These  and  other  experiments  show  that  when  one  set  of 
plates  of  a  small  condenser  is  joined  to  the  wire  W  and  the 
other  set  to  one  of  the  electrodes,  the  potential- difference 
between  the  wire  and  that  electrode  is  diminished.  The 
condenser  has  no  effect  when  the  connexions  to  it  are  made 
through  high  resistances. 
The  effect  may  also  be  observed  by  having  the  leaves  of 
an  electroscope  connected  to  the  wire  and  the  case  to  the 
negative  electrode.  No  change  in  the  divergence  of  the 
leaves  is  produced  when  the  electrostatic  voltmeter  is  con- 
nected up  through  high  resistances,  but  when  the  connectors 
are  of  small  resistance  an  increase  in  the  divergence  of  the 
leaves  is  observed  when  one  terminal  of  the  voltmeter  is 
joined  to  the  wire  in  the  gas  and  the  other  to  the  positive 
electrode.  The  opposite  effect  is  produced  when  the  voltmeter 
terminal  is  joined  to  the  negative  electrode.  Similar  results 
were  obtained  when  air  was  used  instead  of  hydrogen  for 
these  experiments. 
It  is  difficult  to  see  how  these  effects  arise,  but  it  is  most 
probable  that  they  are  due  to  the  unsteadiness  of  the  current. 
A  telephone  placed  in  the  circuit  leading  from  the  battery 
to  the  electrodes  gave  a  rumbling  noise,  which  showed  that 
there  must  have  been  slight  variations  in  the  current.  The 
noise  was  generally  diminished  (without  altering  the  intensity 
of  the  current  as  given  by  a  galvanometer)  by  increasing  the 
potential  of  the  battery  and  the  external  resistance,  but  no 
reliable  method  was  found  of  producing  silence  in  the 
telephone. 
