144     Dr.  H.  L.  Bronson  :   Effect  of  High  Temperatures 
deflexions,  when  small,  will  be  proportional  to  the  same 
potential-difference  of  the  quadrants.  This  makes  a  simple 
direct-reading  instrument. 
Fig.  1. 
trimeter: 
/  >w  EUctremet 
Earth 
j 
Testing  Vessel. 
Httjh  Resistance. 
xtt^xx«tt*xxxx*x*xxxx 
Earth. 
Resistances  of  the  order  o£  10u  ohms  made  of  amyl  alcohol 
and  carbon  on  glass  were  tried,  and  gave  results  which  com- 
pared  favourably  with  those  obtained  by  the  "  rate  "  method. 
The  results,  however,  were  not  satisfactory,  because  the  re- 
sistances did  not  remain  sufficiently  constant.  This  was 
probably  due  in  one  case  to  polarization,  and  in  the  other  to 
temperature  changes.  Professor  Rutherford  then  suggested 
the  possibility  of  using  an  ionization  current  in  place  of  the 
conduction  current  through  a  high  resistance.  The  possi- 
bility of  doing  this  is  due  to  the  fact  that  the  ionization 
current  through  a  gas,  subject  to  a  constant  source  of  ioniza- 
tion, is  proportional  to  the  potential-difference  between  the 
plates  when  this  potential-difference  is  small.  Any  strongly 
radioactive  substance  which  will  remain  approximately  con- 
stant during  the  time  of  a  single  experiment  can  be  used  as 
the  source  of  ionization.  In  the  present  case  a  very  radio- 
active bismuth  plate  from  Dr.  Sthamer  of  Hamburg  was 
used.  The  activity  of  this  plate  was  due  to  a  deposit  of  the 
so-called  radio-tellurium  of  Marckwald,  and  has  the  advantage 
of  not  giving  out  any  penetrating  rays.  If,  for  example, 
radium  had  been  used,  it  would  have  been  necessary  to  sur- 
round the  vessel  in  which  it  was  placed  with  a  shield  of  lead 
of  considerable  thickness,  in  order  to  prevent  any  disturbance 
by  the  penetrating  rays. 
