510  Dr.  F.  Horton  on  the  Electrical 
Wheatstone  bridge.  With  some  of  the  oxides  electrolysis 
was  found  to  accompany  the  conduction,  and  with  these 
another  method  had  to  be  used  for  obtaining  the  resistance. 
Kohlrausch's  method  of  using  an  alternating  current  and  a 
telephone  indicator  does  not  work  well  for  resistances  over 
10,000  ohms,  and  the  resistance  to  be  measured  was  usually 
greater  than  this  value.  The  method  of  using  a  rotating 
commutator  to  reverse  the  battery  and  galvanometer  circuits 
together  many  times  a  second  requires  an  exceedingly  well- 
made  commutator.  Unfortunately  the  one  made  for  the 
researches  which  Mr.  Fitzpatrick  carried  out  in  this  laboratory 
was  in  use  at  the  time  and  not  available  for  this  work.  Other 
commutators  were  tried,  but  none  worked  satisfactorily. 
The  method  finally  adopted  was  not  of  great  accuracy, 
but  was  sufficiently  accurate  for  the  object  of  the  present 
research.  It  consisted  in  having  the  galvanometer  per- 
manently connected  up  to  the  Wheatstone  bridge  without 
a  key  in  circuit,  and  noticing  the  direction  of  its  first  kick 
on  momentarily  closing  the  battery  circuit.  The  resistance 
was  connected  to  a  reversing  key  which  was  reversed  after 
each  u  make "  of  the  battery  circuit,  so  that  successive 
momentary  currents  were  sent  in  opposite  directions  through 
the  electrolyte.  This  method  worked  better  than  I  had 
expected,  and,  by  taking  care  not  to  let  a  continuous  current 
pass  through  the  oxide,  results  in  very  good  agreement  were 
obtained.  There  was  sometimes  a  small  thermoelectric  force 
developed  on  heating  the  oxide,  and  consequently  the 
resistance  in  one  direction  was  slightly  greater  than  in  the 
opposite  direction.  The  mean  of  these  resistances  was  always 
taken,  and  the  circuit  through  the  oxide  was  kept  open  except 
just  when  observations  were  being  made.  The  voltage  used 
on  the  battery  circuit  varied,  according  to  the  resistance  to 
be  measured,  from  1  to  500  volts. 
The  oxides  experimented  on  were  lime,  baryta,  magnesia, 
bismuth  trioxide,  plumbic  oxide,  cupric  oxide,  sodium 
peroxide,  and  quartz.  The  results  obtained  will  now  be 
given. 
Lime. 
The  piece  of  lime  used  was  "85  cm.  square  by  *70  mm. 
thick.  Its  resistance  at  the  temperature  of  the  laboratory 
was  oreater  than  100  megohms.  The  following  table  contains 
the  resistance,  and  the  conductivity  calculated  therefrom,  at 
different  temperatures  : — 
