488     Prof.  Wilson  and  Mr.  Gold  :  Electrical  Conductivity  of 
moving-coil  galvanometer  Gr.  The  cell  B  o-ave  about  one 
volt  and  served  to  polarize  the  electrodes.  When  an  alter- 
nating P.D.  was  produced  between  M  and  X,  the  detector  was 
depolarized  and  a  current  passed  through  the  galvanometer. 
The  deflexion  of  the  galvanometer-coil  was  read  by  means  of 
an  incandescent  lamp  and  a  scale,  a  current  of  10-9  ampere 
giving  a  deflexion  of  one  scale-division. 
The  condensers  C1?  C2  each  consisted  of  two  parallel 
circular  disks  10  cms.  in  diameter,  supported  on  ebonite 
rods.  The  distance  between  the  disks  of  C1  was  0*15  cm. 
and  of  Co  0*75  cm.,  in  most  of  the  experiments.  The  con- 
denser C3  consisted  of  two  brass  disks  10  cms.  in  diameter, 
whose  distance  apart  could  be  adjusted  and  measured  by 
means  of  an  accurate  micrometer-screw. 
To  determine  the  conductivity  between  the  flame-electrodes, 
the  rapidly  alternating  P.D.  produced  by  the  Tesla  coil  was 
applied  at  A  and  B  (fig.  2),  and  the  condenser  C3  was  ad- 
justed until  the  deflexion  of  the  galvanometer  was  a  minimum. 
On  starting  the  alternating  current,  the  galvanometer  de- 
flexion increased  to  a  maximum  value  and  then  fell  off 
slowly.  The  alternating  current  was  always  kept  on  for 
15  sees,  and  the  maximum  deflexion  noted,  and  then  after  an 
interval  of  15  sees,  the  current  was  turned  on  again  and  so 
on.  During  the  intervals  the  condenser  was  adjusted,  and 
the  deflexions  corresponding  to  a  series  of  positions  of  the 
adjustable  condenser-disk  were  thus  obtained.  A  curve  was 
then  drawn  on  squared  paper  showing  the  relation  between 
the  galvanometer  deflexion  and  the  condenser-screw  reading, 
and  so  the  screw  reading  for  which  the  deflexion  was  a 
minimum  was  obtained.  The  observations  at  each  position 
of  the  condenser-disk  were  repeated  several  times  and  the 
mean  taken  ;  the  series  of  observations  was  also  repeated 
first  in  one  direction  and  then  in  the  other,  while  various 
intermediate  positions  were  also  tried  so  as  to  make  as  certain 
as  possible  that  the  correct  relation  between  the  deflexion 
and  the  condenser  distance  was  obtained.  These  precautions 
were  very  necessary  because  the  apparatus  was  sometimes 
irregular  in  its  action.  No  great  difficulty  was  experienced 
in  keeping  the  flame  sufficiently  constant  during  the  experi- 
ments, but  it  was  not  easy  to  keep  the  alternating  current 
constant.  This  was  due  partly  to  variations  taking  place  at 
the  spark-gap  and  partly  to  irregularity  in  the  action  of  the 
induction-coil  interrupter.  The  spark-gap  finally  adopted 
consisted  of  two  platinum  spheres  kept  in  an  atmosphere  of 
hydrogen.  In  most  of  the  experiments  the  length  of  the 
gap  was  about  0'2  cm. 
