512 
Dr.  F.  Horton  on  the  Electrical 
succeeding  curves  the  value  of  the  ordinate  is  successively 
multiplied  by  ten,  so  that  in  the  second  curve  each  unit  is 
equal  to  10"7  mho,  in  the  third  10"6  mho.  in  the  fourth  10"5 
mho,  and  in  the  last  lO-4  mho. 
These  curves  are  exactly  similar  to  those  obtained  by 
"W  elmelt  *  for  the  relation  between  the  number  of  corpuscles 
given  out  by  lime  (measured  by  the  saturation  current)  and  the 
temperature.  Wehnelt  found  that  the  number  of  corpuscles 
emitted  from  glowing  lime  obeyed  the  law  found  by  0.  "W. 
Richardson  f  and  by  H.  A.  Wilson  J  to  hold  for  the  negative 
b 
corpuscles  from  hot  platinum,  viz..  n  =  k6*e~e  where  n  is  the 
number  of  corpuscles,  6  the  absolute  temperature,  and  A  and 
b  are  constants.  If  the  conduction  is  carried  on  by  the 
corpuscles  set  free  in  the  interior  of  the  oxide  at  high 
temperatures,  we  might  expect  a  similar  relation  to  hold 
between  the  conductivity  and  the  temperature. 
Diagram  II. 
O005  -0007 
Scale    of  9~' 
0009 
In  order  to  test  this  we  may  denote  the  conductivity  bv  c, 
and  write  the  equation  in  the  form 
logkl  c— \  log10  #  =  log10  A- -4346  6-\ 
so   that   on   plotting   the  values  of  log10  c  —  \  log10  0  against 
*  Wehnelt,  Ann.  d,  PJnjs.  xiv.  p.  425  (1904). 
t  O.  W.  Richardson,  Phil.  Trans.  A.  vol.  cci.  p.  497  (]903). 
|  H.  A.  Wilson,  Phil.  Trans.  A.  vol.  ccii.  p.  24-3  (1903). 
