146     Dr.  H.  L.  Branson:  Effect  of  High  Temperatures  on 
between  the  quadrants.  By  the  use  o£  a  finer  suspension- 
wire  and  a  higher  voltage  on  the  needle,  it  was  possible  to 
increase  this  sensitiveness  nearly  one-hundred  fold.  The 
scale  was  100  cm.  in  length  and  only  180  cm.  from  the 
electrometer,  but  it  was  possible,  by  turning  it  at  a  slight 
angle,  to  make  the  deflexions  over  its  whole  length  propor- 
tional to  the  potential-difference  of  the  quadrants,  with  a 
maximum  error  of  not  more  than  one  half  of  1  per  cent.  In 
case  the  current  to  be  measured  produced  a  deflexion  larger 
than  100  cm.,  a  modification  of  this  arrangement  was  adopted. 
By  means  of  the  battery  B  and  the  two  resistances  1^  and  B2, 
the  potential  of  the  radio-tellurium  plate  was  brought  below 
that  of  the  earth.  This  reduces  the  potential-difference 
between  the  quadrants  by  the  same  amount,  and  thus  makes 
the  deflexions  measurable  on  the  scale.  In  this  way  ionization 
currents  large  enough  to  produce  a  deflexion  of  150  cm.  could 
be  measured.  At  this  point  a  limit  was  reached,  because  larger 
currents  than  this  through  the  radio-tellurium  vessel  were 
not  proportional  to  the  potential-difference  between  the  plates. 
No  doubt  with  a  more  active  substance,  such  as  radium 
bromide,  the  currents  wTould  remain  proportional  through  a 
still  larger  range. 
Both  the  radio-tellurium  and  the  plate  M  above  it  were 
covered  with  very  thin  aluminium-foil,  to  eliminate  as  far  as 
possible  any  contact  potential-difference  ;  and  the  wdiole  was 
enclosed  in  a  metallic  vessel  to  prevent  electrostatic  or  air  dis- 
turbances. In  order  to  correct  for  the  small  contact  potential- 
difference,  which  always  remained,  an  extra  key  was  introduced 
between  the  testing  vessel  and  the  electrometer.  When  both 
keys  were  closed  the  testing  vessel  and  all  four  quadrants  of 
the  electrometer  were  connected  to  earth  ;  when  key  2  was 
open  the  earth  connexion  was  broken  and  the  testing  vessel 
was  connected  with  the  electrometer  :  ■  when  key  1  was  open 
the  earth  connexion  of  quadrants  A  A  was  broken,  and  the 
reading  then  obtained  was  the  true  zero,  and  eliminated  any 
error  due  to  contact  potential-differences.  The  entire  apparatus 
wras  carefully  surrounded  by  a  metallic  screen  to  eliminate 
any  electrostatic  disturbance,  the  keys  being  mechanically 
worked  from  without. 
Results  obtained  by  Curie  and  Danne, 
In  the  course  of  a  careful  investigation  of  the  rate  of  decay 
of  the  active  deposit  from  radium,  an  attempt  wTas  made  to  see 
whether  itwas  permanently  changed  by  high  temperatures.  The 
conclusion  arrived  at,  contrary  to  that  of  Curie  and  Danne*, 
*  Comptes  Rendus,  cxxxviii.  p.  748  (1904). 
