738  Prof.  J.  S.  Townsend  on  the  Field  of  Force 
It  has  been  suggested  that  the  -unsteadiness  is  due  to 
faults  in  the  connexions  of  the  battery,  but  although  the 
connexions  were  carefully  examined  and  renewed  in  places 
where  there  might  possibly  have  been  a  fault,  still  there  was 
no  diminution  in  noise  in  the  telephone.  It  was  also  found 
that  the  telephone  gave  no  sound  when  the  wires  leading  to 
the  electrodes  in  the  gas  were  connected  together,  and  the 
resistance  of  the  circuit  increased  so  as  to  bring  the  current 
to  its  original  value. 
It  appears  therefore  that  the  slight  variations  in  the 
current  must  be  attributed  to  the  nature  of  the  conductivity 
of  the  gas. 
Experiments  were  also  made  with  a  current  (maintained  by 
a  battery  of  lead  cells)  through  a  gas  at  low  pressure  in  a 
glass  tube  of  diameter  2'4  centimetres.  The  electrodes  were 
flat  aluminium  disks  14  centimetres  apart,  and  a  wire  projected 
into  the  tube  in  the  usual  manner  near  one  of  the  electrodes 
so  as  to  obtain  the  cathode  fall  of  potential.  The  telephone 
showed  that  in  this  case  also  the  current  was  not  perfectly 
uniform,  but  the  effect  of  the  capacity  of  the  voltmeter 
could  scarcely  be  detected,  and  the  ordinary  value  of  the 
cathode  fall  of  potential  was  obtained.  The  potential  of  the 
wire  is  therefore  more  stable  in  a  long  discharge  which  fills 
a  glass  tube,  but  of  course  this  apparatus  would  have  been 
useless  for  these  investigations,  as  one  of  the  principal  objects 
was  to  obtain  small  currents  in  a  uniform  field. 
5.  A  number  of  experiments  were  made  to  determine  the 
potential  of  the  wire  at  different  distances  from  the  large 
parallel  plate-electrodes,  using  large  resistances  in  the 
connexions  to  the  voltmeter.  The  following  curves  represent 
the  results  for  hydrogen  at  different  pressures.  The  three 
curves  shown  in  fig.  2  give  the  potentials  of  the  wire  at 
different  distances  fiom  the  electrodes  for  three  different 
currents.  The  currents  expressed  in  amperes  per  square 
centimetre  of  the  electrodes  corresponding  to  the  curves  1,  t 
and  3  were  1*2  x  10~6,  6  x  10  ~6,  and  3  x  10-5  respectively. 
The  pressure  of  the  hydrogen  wTas  1*37  millimetres,  which 
corresponds  to  the  minimum  sparking-potential  for  the  distance 
of  8  millimetres  between  the  plates.  The  small  difference 
between  the  curves  1  and  2  shows  that  when  the  current  is 
small,  the  potential  of  the  wire  is  not  affected  by  changing 
the  intensity  of  the  current.  It  is  most  likely  in  these  cases 
that  the  charge  on  the  ions  is  too  small  to  affect  the  field  of 
force  appreciably,  and  that  if  the  wire  were  removed  the  field 
would  be  uniform.  The  ratios  of  the  currents  carried  by 
ions  of  different  sign  are  independent  of  the  absolute  value  of 
