76  Mr.  Veley  on  an  Apparatus  for  Determination 
the  bottom  of  the  dish.  The  above  mentioned  glass  rod 
passed  through  a  metal  collar  in  which  it  could  be  rigidly 
fixed  by  a  screw  ;  the  collar  was  connected  with  a  sliding 
piece  working  upon  two  upright  struts,  F,  F',  one  smooth 
and  the  other  provided  with  V-groove  ;  at  the  upper  end  of 
the  sliding  piece  was  fixed  a  micrometer-screw,  the  head  of 
which,  divided  into  50,  revolved  once  for  each  '5  millimetre 
on  a  vertical  scale,  so  that  *01  mm.  could  be  read  directly 
and  less  than  half  that  amount  by  estimation  ;  the  vertical 
play  of  the  screw  amounted  to  31  mms. 
The  glass  rod  could  be  fixed  rigidly  in  any  desired  position 
by  a  clamping  screw.  The  vertical  struts  were  fixed  upon  a 
white  marble  slab  by  three  levelling-screws  (two  Gr,  G/ shown) 
arranged  in  a  triangle ;  at  the  other  end  of  the  slab  was 
placed  the  gilt  nickel  dish  kept  in  position  by  three  ebonite 
knobs  (two  H,  H'  shown)  let  into  the  slab. 
The  corrections  for  the  zero  point  and  the  graduation 
errors  of  the  vertical  scale  were  determined  by  taking  away 
the  dish,  placing  the  apparatus  in  a  horizontal  position,  and 
measuring  the  displacement  of  the  edge  of  the  rim  of  the 
disk  by  a  Hilger  travelling  microscope.  A  table  of  correc- 
tions for  each  condenser  was  drawn  out  once  for  all. 
One  of  these  condensers  was  throughout  kept  as  an  air- 
condenser,  while  the  other  was  used  to  contain  the  liquid 
dielectric ;  during  working,  each  condenser  was  closed  with 
an  ebonite  plate  cut  into  half  and  with  a  small  semicircular 
piece  cut  out  of  each  half  for  the  spindle  to  pass  through  ; 
these  plates  served  to  exclude  dust. 
The  two  dishes  were  regarded  as  the  outer,  and  the  two 
disks  with  their  spindles  as  the  inner  plates  of  a  pair  of 
diagrammatic  condensers. 
The  major  portion  of  the  capacity*  of  such  a  condenser 
(or  leyden)  is  made  up  of  three  terms  :  (a)  that  due  to  the  area 
of  the  lower  surface  of  the  disk  and  the  corresponding  area 
of  the  dish,  (b)  that  due  to  the  area  of  rim  of  the  disk  and 
likewise,  (c)  that  due  to  the  area  of  spindle  and  likewise.  Of 
these,  (a)  varies  writh  the  distance  and  nature  of  dielectric, 
(b)  is  constant  according  to  the  method  of  working  adopted, 
(c)  varies  according  to  the  length  of  spindle  immersed  within 
the  dish.  In  actual  working,  the  alteration  of  capacity  due 
to  (c)  varied  about  1  per  cent,  on  an  average ;  as  the  maximum 
*  The  total  capacity  of  such,  a  condenser  is  a  very  complex  matter, 
and  could  only  be  solved  by  the  method  of  conjugate  functions  as  dis- 
cussed by  Clerk  Maxwell,  '  Electricity  and  Magnetism/  Articles  195  et 
seq.,  and  J.  J.  Thomson,  l  Recent  Researches  in  Electricity  and  Mag- 
netism,' pp.  208-250  (Clarendon  Press,  1893). 
