514  Dr.  F.  Horton  on  the  Electrical 
decreasing  rate  until  about  the  same  magnitude  as  the  direct 
current  just  before  reversing.  This  behaviour  may  be 
explained  by  supposing  that  a  layer  of  oxygen  is  formed  by 
electrolysis  against  the  anode,  and  that  this  increases  the 
resistance  of  the  circuit  in  addition  to  causing  the  polarization 
E.M.F.  When  the  circuit  is  completed  for  only  a  short  time 
before  the  key  connected  to  the  lime  is  reversed,  very  little 
gas  can  have  collected  at  the  anode,  and  this  gives  a  back- 
E.M.F.,  but  does  not  greatly  increase  the  resistance.  On 
reversing,  therefore,  a  large  deflexion  in  the  opposite  direction 
is  obtained.  On  the  other  hand,  when  time  is  allowed  for 
much  gas  to  collect  at  the  anode,  then,  on  reversing,  the 
deflexion  in  the  opposite  direction  is  slightly  larger  than  the 
small  value  just  before  reversing,  owing  to  the  back-E.M.F. 
This  increases  as  the  oxygen  recombines  with  the  calcium 
which  is  now  formed  at  that  electrode,  and  the  oxygen  which 
is  being  formed  at  what  is  now  the  anode  combines  with  the 
calcium  already  liberated  there  by  the  direct  current.  Thus 
the  resistance  gradually  decreases  until  all  the  collected  gas 
is  used  up;  a  maximum  current  then  passes  and  the  de- 
flexion afterwards  decreases  as  oxygen  collects  against  the 
other  electrode. 
These  experiments  were  carried  out  at  about  1000°  0.  At 
higher  temperatures  the  current  did  not  die  away  to  so  small 
a  value;  and  on  reversing  the  key  connected  to  the  lime  a 
large  deflexion  in  the  opposite  direction  was  obtained  which 
gradually  died  away,  and  did  not  increase  even  when  the 
direct  current  had  been  flowing  for  a  long  time.  It  is  pro- 
bable that  this  is  due  to  the  oxygen  escaping  more  easily  from 
the  electrode  at  the  higher  temperature  and  not  forming  a 
layer  over  the  anode,  and  thus  increasing  the  resistance  of  the 
circuit  as  it  did  in  the  experiments  at  1000°  C. 
Several  specimens  of  lime  wTere  experimented  on  and  gave 
results  practically  identical  with  those  described  above,  both 
as  regards  the  variation  of  conductivity  with  temperature 
and  the  electrolytic  effects  of  the  current.  It  thus  appears 
that  the  conduction  of  electricity  through  heated  lime  is 
carried  on  partly  by  electrolytic  ions  as  in  a  liquid  electrolyte ; 
but  the  fact  that  the  increase  of  conductivity  with  temperature 
corresponds,  so  closely,  with  the  increase  in  the  rate  of 
emission  of  electrons,  seems  to  indicate  that  most  of  the  con- 
duction is  metallic. 
Magnesia. 
A  piece  of  magnesia  *84  cm.X'83  cm.  and  "090  cm. 
thick  was  used.  The  resistance  of  this  at  the  temperature  of 
the  laboratory  was  too  large  to  be  measured  by  the   method 
