Force  in  the  Contact  of  Metals,  265 
points  of  union.     The  holes  in  the  bottom  of  the  box  and  in  its 
upper  surface  were  closed  with  cotton,  to  prevent  the  entrance  of 
the  external  air,  which  would  have  rapidly  altered  the  tempera- 
ture ;  this  was,  besides,  nearly  the  same  as  that  of  the  external 
air,  but  changed  more  slowly.     During  the  whole  of  the  ex- 
periments the  temperature  of  the  point  of  soldering  was  about 
+  10°.      To    investigate  the  galvanic   conductivity,  a  bobbin 
covered  with  copper  wire  had  been  placed  in  the  circuit.     A 
steel  magnet,  placed  in  the  coil,  had  sufficient  space  to  be  moved 
a  certain  distance.     The  inductive  force  thus  remaining  invari- 
able, the  deviations  of  the  magnetometer  obtained  through  the 
inductor  would  be  proportional  to  the  conducting- power.      A 
rheostat  had  also  been  placed  in  the  circuit;  but  it  was  only 
required  for  the  bismuth-copper  combination,  as  the  latter  gave 
deviations  several  times  as  great  as  the  others.     The  magneto- 
meter, the  deviations  of  which  were  read  in  the  usual  manner 
(with  the  aid  of  a  telescope  and  a  scale),  had  a  perfectly  astatic 
system  of  needles,  so  that  the  earth's  magnetism  exerted  no  in- 
fluence upon  it.     The  necessary  directing  force  was  obtained  by 
suspending  the  system,  with  its  mirror,  from  a  thin  silver  wire, 
the  torsion-force  of  which  gave  to  the  system  a  determined  di- 
rection.    This  mode  of  suspension  has  the  important  advantage 
of  being  totally  independent  of  the  variations  of  the  terrestrial 
magnetic  declination.     The  deviations  are  proportional  to  the 
intensity  of  the  current.     When  a  fresh  pair  of  wires  had  been 
introduced,  the  first  observation  was  not  taken  until  the  points 
of  contact  had  acquired  the  temperature  indicated  by  the  ther- 
mometer belonging  thereto,  after  which  the  observations  suc- 
ceeded one  another  at  short  intervals.     With  this  method  of 
observation,  there  was  perhaps  reason  to  fear  that  some  heat 
would  pass  from  the  warmest  to  the  coolest  parts  of  the  wires, 
and  that  consequently  the  points  of  contact  would  not  have  ex- 
actly the  temperature  indicated  by  the  thermometers,  the  indi- 
cations of  which  depended  partly  also  on  the  heat  of  the  sur- 
rounding air.     In  order  to  see  if  this  was  the  case,  the  wires  of 
a  pair  of  combinations  were  joined  to  wires  of  the  same  sort,  so 
that  the  distance,  from  the  wooden  box  above  mentioned,  to  the 
cork  of  the  test- glass  was  more  than  doubled;  but  the  deviation 
remained  unaltered.  The  presumed  possibility  of  errors  in  this  re- 
spect had  therefore  no  existence.     In  the  results  of  experiments 
given  below,  all  the  observations  relating  to  the  same  combi- 
nation have  been  combined  into  a  single  group,  although  they 
may  not  have  been  made  at  the  same  time. 
