Theory  of  Magnetism,  423 
merit  of  the  hands  of  a  watch.  Then  conceiving  at  first  the 
copper  to  be  at  rest,  let  the  magnetic  streams  which  pass  through 
it,  as  well  those  of  the  magnet  as  the  terrestrial  streams,  be  re- 
solved into  components  parallel  and  perpendicular  to  the  faces 
of  the  plate.  It  is  clear  that  the  vertical  component  of  the 
terrestrial  magnetism  can  have  no  rotational  effect ;  and  the 
same  is  the  case  with  respect  to  the  horizontal  components  of 
both  kinds,  inasmuch  as  these  only  give  rise  to  vertical  galvanic 
currents,  the  effects  of  which,  as  regards  rotation,  neutralize 
each  other.  We  have,  therefore,  only  to  consider  the  vertical 
components  of  the  magnet's  streams.  Now  as  these  streams 
enter  the  needle  at  the  north  end  and  the  needle  is  above  the 
copper,  the  vertical  components  will  flow  upwards  at  that  end. 
Again,  the  copper  being  now  supposed  to  rotate  in  the  direction 
above  stated,  since  the  motion  of  the  parts  under  the  north  end 
is  eastward,  the  secondary  or  induced  stream  will  flow  westward. 
Comparing,  therefore,  these  circumstances  with  the  case  of  the 
elementary  experiment  in  which,  when  the  magnetic  current 
flowed  upwards  and  the  induced  current  southward,  the  resulting 
galvanic  current  flowed  eastward  (see  art.  38),  it  will  appear,  by 
imagining  the  horizontal  directions  to  be  turned  through  90°,  that 
in  the  actual  case  the  galvanic  current  flows  southward.  With  re- 
pect  to  what  takes  place  at  the  south  end,  we  have  to  consider  that 
the  vertical  components  of  the  magnet's  issuing  streams  pass 
through  the  copper  downwards,  and  that  the  direction  of  the  in- 
duced streams  is  changed  from  westward  to  eastward.  Each  of 
these  changes  (as  was  shown  by  Faraday's  experiments)  causes 
a  reversion  of  the  direction  of  the  galvanic  current,  and  conse- 
quently the  direction  is  southward  at  the  south  end  as  well  as  at 
the  north  end. 
After  this  determination  we  have  only  to  inquire  what  effect 
a  southward  galvanic  current  has  on  a  magnetic  needle  above  it. 
The  spiral  motion  being  dextrorsum,  the  circular  motion  will 
conspire  with  the  entering  streams  at  the  north  end  on  the  east 
side,  and  be  opposed  to  them  on  the  west  side.  Hence  the  north 
end  will  be  driven  eastward,  that  is,  in  the  direction  of  the  motion 
of  the  disk.  So  the  circular  motion  will  be  opposed  to  the  issu- 
ing streams  at  the  south  end  on  the  east  side  and  conspire  with 
them  on  the  west  side,  and  the  south  end  will  consequently  be 
urged  westward — that  is,  again  in  the  direction  of  the  motion  of 
the  disk.  These  results  agree  with  the  known  facts  of  the  ex- 
periment ;  and  this  agreement,  it  is  to  be  noticed,  depends  upon 
the  foregoing  inference  respecting  the  direction  of  the  induced 
galvanic  current. 
40.  All  the  other  instances  of  the  mutual  action  between 
a  magnet  and  copper  relatively  in  motion  may  be  referred  to  the 
