264        M.  E.  Edlund's  Researches  on  the  Electromotive 
those  introduced  by  the  pure  hypothesis  perse.  Hence  the  final 
conclusion  may  be  drawn  that,  the  hypothesis  being  at  once  un- 
necessary and  insufficient,  and  its  language  being  in  practice 
ambiguous,  it  is  desirable  that  in  expounding  the  science  of  che- 
mistry, whether  orally  or  in  text-books,  the  hypothesis  and  its 
language  should  not  be  made  to  occupy  the  prominent  and  fun- 
damental position  they  now  fill. 
XXXI.  Researches  on  the  Electromotive  Force  in  the  Contact  of 
Metals,  and  on  the  M edification  of  that  Force  by  Heat.  JB^E. 
Edlund. 
[Concluded  from  p.  223.] 
§5. 
THE  metallic  combinations  the  electromotive  forces  of  which 
have  been  determined  in  the  preceding  pages,  were  further 
investigated  with  respect  to  their  thermoelectric  properties.  As, 
with  the  exception  of  the  platinum-palladium,  there  were  of  each 
combination  two  pairs  of  wires,  I  experimented  first  on  one  pair 
of  each  combination.  Having  investigated  all  these  pairs,  I 
passed  to  the  remaining  pairs,  for  the  double  purpose  of  control- 
ling the  former  determinations  and  of  ascertaining  whether  there 
were  in  the  pairs  of  the  same  combination  any  slight  differences 
from  each  other.  This  was  in  fact  the  case  with  some  of  them ; 
and  these  were  submitted  to  a  new  trial,  effected  on  each  pair. 
The  experiments  were  made  in  the  following  manner : — Near 
the  point  of  soldering  I  bent  each  wire  at  a  right  angle,  so  that 
the  two  wires  were  parallel,  and  the  distance  between  them 
10  millims.  The  point  of  soldering  was  in  the  middle  of  the 
elbow  which  united  them.  As  the  bismuth  wire  could  not  be 
bent,  the  copper  wire  to  which  it  was  soldered  wTas  bent  twice  at 
a  right  angle  near  the  soldering,  which,  therefore,  was  near  the 
elbow  instead  of  in  the  middle  of  it.  The  wires  thus  prepared 
were  passed  through  the  cork  into  a  large  test-glass.  Through 
the  same  cork  a  very  sensitive  thermometer  was  so  placed  that 
its  little  cylindrical  bulb  rested  with  its  middle  against  the  point 
of  soldering.  The  test-glass  was  introduced  through  an  aper- 
ture in  the  centre  of  a  thin  wooden  lid  fitted  to  a  large  glass 
vessel  filled  with  cold  water.  In  order  that  the  warming  of  the 
water  by  the  air  of  the  room  might  be  very  slow,  the  glass  vessel 
was  surrounded  by  a  layer  of  cotton  wadding.  The  free  ends  of 
the  wires,  issuing  from  the  cork  in  a  vertical  direction,  having 
passed  through  the  bottom  of  a  small  wooden  box,  were  there 
united  by  small  screws  to  the  conducting  wires  of  the  magneto- 
meter. In  the  wooden  box  another  thermometer,  exactly  resem- 
bling the  preceding,  was  placed  with  its  bulb  quite  close  to  the 
