Alternating  Current  on  a  Polarizable  Electrode.       331 
explanation  must  account  for  the   very  marked  dependence 
on  the  degree  of  polarization  of  the  electrode. 
We  know,  however,  very  little  indeed  of  the  conditions 
which  prevail  in  an  oxygen-  or  hydrogen-charged  platinum 
electrode,  and  the  hope  of  obtaining  a  satisfactory  explanation 
of  the  above  phenomenon  seems  remote  until  this  gap  in  our 
knowledge  is  filled. 
The  object  of  the  present  paper  is  to  show  that  in  a  polari- 
zable electrode  in  general  (and  the  results  obtained  for  the 
simplest  of  all  cases,  here  treated,  are  applicable  in  principle 
to  all  polarizable  electrodes),  so  soon  as  we  pass  beyond  the 
stage  of  initial  capacity,  the  alternating  current  gives  rise  to 
an  asymmetry,  which  would  appear  as  a  coherer  effect  ana- 
logous to  the  above.  In  order  that  this  asymmetry  may  be 
sufficient  to  account  for  the  coherer  effect  used  in  practice, 
certain  conditions  must  hold  (afterwards  to  be  developed), 
which,  though  not  in  antagonism  with  any  known  fact,  and, 
indeed,  corroborated  to  some  extent  by  what  is  known  of  the 
conditions,  must  be  proved  or  disproved  by  a  greater  ex- 
perimental material  than  I  am  at  present  able  to  bring- 
forward.  With  this  reservation,  however,  I  believe  that  the 
application  to  this  practical  case  is  not  without  interest. 
The  case  here  considered  is  that  of  a  mercury  electrode. 
The  conditions  are  much  better  known  than  in  the  case  of 
platinum,  and  the  changes  of  concentration,  taking  place  as 
they  do,  not  in  the  electrode  but  in  the  liquid,  are  easier  to 
follow.  Another  advantage  that  mercury  offers,  is  that  the 
electrode  can  always  be  set  up  in  a  perfectly  definite  con- 
dition, and  with  comparative  ease  in  a  pure  state. 
The  experiments,  to  be  afterwards  described,  consisted 
essentially  in  the  sending  of  an  alternating  current  of  fre- 
quency varying  between  80-^-  and  5000^  per  second  through 
an  electrolyte  between  two  electrodes,  of  which  one  was  very 
small,  the  other  more  than  a  thousand  times  as  large,  so  that 
the  latter  may  be  considered  as  unpolarizable  in  comparison 
with  the  former.  A  galvanometer  in  the  circuit  showed  that 
the  alternating  current,  before  symmetric,  was  rendered 
asymmetric  by  the  passage  through  the  electrolytic  cell. 
In  other  words,  there  was  superposed  on  the  alternating 
current  a  direct  current,  and  the  production  of  this  direct 
current  involves  the  existence  of  an  asymmetry  in  the 
E.M.F.  of  the  polarization  produced  by  the  alternating 
current. 
This  asymmetry  seems,  however,  to  be  a  natural  conse- 
quence of  the  osmotic  theory  of  Electromotive  Force  in  general 
and  its  application  to  polarization  in  particular.     According 
Z  2 
