386 
DE.  A.  MATTHIESSEX  OX  THE  ELECTEIC 
than  those  found  by  nearly  all  of  the  earlier  experimenters,  are  in  accordance  with 
the  results  lately  made  known  by  Buff. 
The  wires  of  bismuth  employed  for  these  experiments  were  of  a yery  small  diameter 
(0'144  millim.  and  0T85  millim.);  for  when  manipulating  with  thicker  ones,  very  yarding 
results  were  obtamed,  probably  owing  to  the  disposition  of  the  crystals,  as  !Matteucci* 
has  already  observed.  In  fact,  twelve  experiments  made  with  such  wires,  when  the  con- 
ducting power  varied  from  ITl  to  1‘36,  gave  as  a mean  1’20,  which  agrees  pretty  accu- 
rately with  the  number  in  the  above  series. 
Graphite  No.  1 is  the  so-called  pure  Ceylon,  No.  3 the  piuified  German,  and  No.  2 
a mixture  of  both,  all  used  for  making  pencils.  The  specimens  were  pmiiied  by  Beodie’s 
patent,  and  pressed  by  Mr.  Caetmell,  to  whom  I am  indebted  for  the  same.  These,  as 
well  as  the  gas-coke  and  Buxsen’s  battery-coke,  were  shaped  into  thin  rods  by  hand, 
and  the  ends  coppered  galvanoplastically. 
To  form  the  connexion  in  order  to  determine  their  conducting  powers,  one  end  stood 
in  a cup  of  quicksilver,  the  other  cup  being  formed  by  placing  an  india-rubber  tubing 
over  the  other  end,  the  latter  cup  being  of  course  also  filled  with  quicksilver.  As  the 
rods  were  made  by  hand,  the  difference  in  the  results  of  the  same  specimens  is  thus 
accounted  for. 
An  interesting  fact  was  now  observed,  "viz.  that  m all  the  cokes  and  graphites  the  con- 
ducting power  increased  by  heat.  Some  experiments  were  made  with  gas-coke,  and  it 
was  found  that  if  at  0°  the  conducting  power  equalled  100,  it,  from  0°  to  140°,  increased 
0-00245  for  each  degree,  and  between  the  common  temperature  and  a light  red  heat, 
about  12  per  cent. 
For  these  experiments  the  connecting  copper  wires  were  soldered  on  to  the  coppered 
ends  of  the  rods.  The  red  phosphorus  was  likewise  coppered  galvanoplastically,  and  the 
values  determined  after  the  same  method. 
The  following  metals  were  chemically  pure : — Silver,  gold,  zinc,  cadmium,  tin,  lead, 
antimony,  quicksilver,  bismuth,  tellurium ; and  the  undermentioned  pressed,  tiz.  sodium, 
zinc,  magnesium,  calcium,  cadmium,  potassium,  tin,  lead,  strontium,  antimony,  bismuth, 
alloy  of  bismuth-antimony,  alloy  of  bismuth-tin,  and  tellurium. 
No  difficulty  was  found  in  pressing  the  above-mentioned  wii-es,  only  with  antimony 
and  tellurium  care  must  be  taken  in  heating  the  press  of  the  former  to  a dark  red 
heat,  and  of  the  latter  until  it  burns  blue. 
Quicksilver  was  determined  in  glass  tubes,  and  the  alloy  of  antimony-zinc  cast  like- 
wise in  small  glass  tubes,  it  being  too  hard  to  press  and  too  brittle  to  draw.  The 
remainder  of  the  wires  were  drawn. 
In  conclusion,  I will  record  a few  experiments  on  the  alloys  of  lead,  tin,  zinc  and 
cadmium.  The  alloys  made  of  two  of  these  metals  appear  to  conduct  electricity  after  a 
very  simple  law ; namely,  their  conducting  power  is  the  mean  of  the  conducting  power 
of  the  quantity  of  each  metal  employed,  as  the  following  results  show. 
^ Coraptes  Eendus,  vol.  xl.  p.  Sdil. 
