Theory  of  Magnetism.  421 
placed  with  its  faces  at  right  angles  to  the  line  of  junction  of  the 
poles  of  a  powerful  horseshoe  magnet,  and  the  terminals  of  a 
galvanometer  were  put  in  contact  with  the  long  edges,  a  galvanic 
current  was  developed  as  soon  as  the  plate  was  caused  to  move 
transversely  to  the  magnetic  current  (Phil.  Trans.  1832,  p.  151). 
Now,  according  to  hydrodynamics,  the  displacement  of  the 
aether  by  the  finite  spherical  atoms  of  the  copper  in  motion 
would  give  rise  to  a  stream  of  aether  in  the  opposite  direction,  due 
to  the  reaction  of  the  unlimited  fluid  mass.  Hence,  conceiving 
the  plate  to  be  held  with  its  faces  horizontal  and  the  long  edges 
parallel  to  the  meridian,  on  moving  it  northward  or  southward  a 
southward  or  northward  current  would  be  produced,  which  would 
coexist  with  the  original  magnetic  current.  The  galvanometer, 
in  fact,  indicates  that  under  these  circumstances  a  galvanic  cur- 
rent is  generated,  which  flows  from  one  of  the  long  edges  of  the 
plate  and  completes  the  circuit  by  entering  at  the  opposite  edge. 
It  follows  from  this  fact,  taken  in  conjunction  with  the  theory, 
that  the  motions  in  rectangular  directions  of  the  two  above- 
mentioned  currents  are  partially  converted  into  dextrorsum  cir- 
cular motion  parallel  to  the  meridian,  and  that  this  effect  is 
attributable  to  the  molecular  constitution  or  arrangement  of  the 
atoms  of  the  copper,  although  theory  is  at  present  incapable  of  as- 
certaining the  exact  modus  operandi. 
36.  It  is  also  found  by  experiment  that  if  the  magnet  be 
moved  and  the  copper  be  stationary,  a  galvanic  current  is  equally 
produced.  This  fact  appears  to  admit  of  the  following  explana- 
tion. The  magnetic  current  and  its  lines  of  motion  neces- 
sarily partake  of  the  motion  of  the  magnet.  Hence  relatively 
to  the  stationary  atoms  of  the  copper  the  current  is  a  composite 
one,  the  horizontal  component  of  which  flows  in  the  direction  of 
the  motion  of  the  magnet  and  with  the  same  velocity.  Now,  by 
hydrodynamics,  this  horizontal  stream  generates,  by  reason  of 
the  contraction  of  channel  by  the  atoms  and  the  inertia  of  the 
fluid  mass,  a  secondary  stream  flowing  in  the  same  direction  as 
the  stream  itself.  Thus,  besides  the  vertical  current,  there  is  a 
horizontal  current  flowing  in  the  direction  of  the  magnet's  actual 
motion,  and,  therefore,  in  the  direction  contrary  to  that  of  the 
virtual  motion  of  the  copper  relative  to  the  magnet.  Conse- 
quently the  conditions  for  generating  galvanic  currents  are  ex- 
actly the  same  in  this  case  as  when  the  copper  was  moved  and 
the  magnet  was  at  rest. 
37.  On  the  principles  thus  established  we  may  proceed  to 
explain  the  whole  of  the  class  of  phenomena  which  depend  on 
the  relative  motion  between  a  magnet  and  a  mass  of  copper. 
But  for  this  purpose  it  will  be  necessary  to  ascertain  previously 
the  direction  of  the  flowing  of  the  current  in  Faraday's  eleraeu* 
