DE.  A.  MATTHIESSEN  ON  THE  THEEMO-ELECTEIC  SEEIES. 
375 
words,  the  thermo-electric  constant  of  commercial  bismuth  (pressed  wire),  =35'81.  All 
the  other  numbers  in  the  Table  have  a similar  meaning,  and  the  metals,  whose  thermo- 
electric numbers  are  calculated  out  of  them,  are  written  in  italics. 
The  coimexion  was,  in  all  cases  where  the  metal  allowed  of  it,  formed  by  soldering. 
Antimony  wii'e,  which  (as  well  as  that  of  tellurium)  is  so  brittle  that  it  will  hardly  bear 
touching,  could  of  course  not  be  directly  soldered,  but  had  to  be  treated  in  the  following 
manner.  The  wire  was  pressed  in  a glass  tube,  in  which  it  was  soldered  to  two  thick 
copper  wires  with  Eose’s  metal ; these,  in  order  to  keep  them  from  moving,  were  made 
fast  in  the  tube  with  gypsum. 
Lithium,  calcium,  strontium,  and  magnesium  were  experimented  with  in  sealed-up 
glass  tubes  filled  with  rock  oil,  as  fig.  1 shows,  the  ends  of  the  wire  being  held  at  a 
Fig.  L 
and  h by  two  steel  wii’e-holders,  whose  jaws  press  so  strongly  together  that  they  flatten 
out  the  ends  of  the  wire.  At  the  other  end  of  the  wire-holders  platinum  wires  were 
soldered  on,  which  passed  out  of  the  tube  at  c and  d.  The  way  in  which  this  piece  of 
apparatus  was  put  together  is  as  follows : — a long  platinum  wire,  soldered  on  to  one  of 
the  wire-holders,  was  passed  through  a glass  tube  (drawn  out  and  bent  as  shown  in 
fig.  2)  aty,  so  that  the  end  of  it  came  out  at  g,  whilst  the  wire-holder  remained  outside 
the  other  end  at/.  The  glass  tube  being  then  laid  in  a trough  filled  with  rock  oil,  the 
wire,  previously  pressed  in  a tube  filled  with  the  same  liquid,  was  then  taken  out  of  the 
tube  in  the  trough,  the  one  end  of  it  scraped  clean  and  fastened  in  the  wire-holder,  and 
drawn  into  the  tube  by  means  of  the  long  platinum  wire,  until  only  the  other  end  of  it 
remained  out ; this  also  having  been  scraped  clean  and  fastened  in  the  second  wire-holder, 
was  drawn  into  the  tube  as  far  as  fig.  2 shows.  The  tube  was  now  lifted  out  of  the 
trough  in  an  inclined  position,  so  that  it  almost  remained  filled  with  the  oil,  and  sealed 
up  at  at  y it  was  then  drawn  out,  the  oil  boiled  to  drive  out  the  air,  and  sealed  up 
MDCCCLVIII.  3 E 
