148     Dr.  H.  L.  Bronson:  Effect  of  High  Temperatures  on 
wires  were  heated  in  a  small  electric  furnace,  made  by 
Di\  C.  A.  limine  of  Berlin.  A  calibration  curve  for  the  furnace 
had  previously  been  made  by  the  use  of  a  platinum-rhodium 
ther mo- junction.  The  wires  were  left  in  the  furnace  different 
lengths  of  time  in  different  experiments,  but  this  seemed  to 
have  very  little  effect  on  the  result.  On  removal  from  the 
furnace,  the  glass  tube  containing  the  wire  was  covered  with 
thin  aluminium-foil  and  placed  in  the  bottom  of  a  cylindrical 
testing  vessel  with  a  central  electrode.  In  this  case  the 
ionization  was  produced  by  the  /3  rays  alone.  In  those  cases 
where  the  wire  was  heated  without  being  sealed  in  a  glass 
tube,  the  wire  itself  was  made  the  central  electrode  of  the 
testing  vessel.  Sometimes  it  was  covered  with  a  sufficient 
thickness  of  lead-foil  to  cut  off  all  the  a.  rays,  but  generally  it 
was  left  uncovered,  in  which  case  the  ionization  was  largely 
produced  by  the  a  rays.  As  the  curves  were  the  same 
in  either  case,  it  will  be  unnecessary  to  mention  which  method 
was  taken  in  any  particular  experiment. 
B,  fig.  3,  is  the  logarithmic  decay  curve,  obtained   when 
1.806 
1.505 
I 
h  903 
5.602 
^.301 
Fig.  3. 
A 
^fi 
^ 
^^V 
N 
100 
125 
150  175 
Time  in  Minutes 
200 
ZZo 
the  active  deposit  was  sealed  in  a  glass  tube  and  heated  to 
900°  C,  and  A  is  the  decay  curve  for  the  active  deposit  which 
has  not  been  subjected  to  a  high  temperature.  These  curves 
are  approximately  parallel,  showing  that  the  rate  of  decay 
was  not  measurably  changed  by  a  temperature  of  900°  CL 
