﻿458 
  Dr. 
  J. 
  E. 
  Almy 
  on 
  

  

  being 
  insulated 
  by 
  ebonite 
  ; 
  the 
  second, 
  S, 
  is 
  carried 
  on 
  the 
  

   frame 
  which 
  carries 
  one 
  plate 
  of 
  the 
  interferometer, 
  this 
  

   plate, 
  P, 
  being 
  supported 
  on 
  a 
  post, 
  p, 
  about 
  which, 
  as 
  a 
  

   vertical 
  axis, 
  it 
  is 
  free 
  to 
  rotate 
  but 
  for 
  the 
  stop 
  R, 
  which 
  

   was 
  made 
  as 
  rigid 
  as 
  the 
  other 
  parts 
  of 
  the 
  apparatus. 
  To 
  

   take 
  up 
  " 
  back 
  lash 
  " 
  and 
  any 
  looseness 
  in 
  the 
  adjustment 
  of 
  

   parts, 
  a 
  number 
  of 
  strong 
  indiarubber 
  bands, 
  B 
  B, 
  were 
  

   stretched 
  around 
  the 
  electrodes 
  pressing 
  upon 
  the 
  screw 
  heads, 
  

   ss; 
  a 
  high-power 
  microscope 
  M, 
  placed 
  horizontally, 
  made 
  

   possible 
  the 
  careful 
  alignment 
  of 
  the 
  minute 
  electrodes, 
  and 
  

   enabled 
  one 
  to 
  follow 
  visually 
  the 
  course 
  of 
  effects. 
  With- 
  

   out 
  the 
  rubber 
  bands 
  it 
  was 
  impossible 
  to 
  get 
  the 
  system 
  

   sufficiently 
  rigid 
  so 
  that 
  the 
  electrodes 
  were 
  not 
  visibly 
  

   moved 
  when 
  potentials 
  near 
  300 
  volts 
  were 
  applied 
  to 
  spark- 
  

   gaps 
  of 
  a 
  few 
  wave-lengths 
  of 
  sodium 
  light 
  ; 
  with 
  the 
  bands 
  

   in 
  place 
  the 
  separation 
  of 
  the 
  electrodes 
  follows 
  promptly 
  the 
  

   displacements 
  of 
  the 
  bands 
  of 
  the 
  interferometer, 
  and 
  that 
  the 
  

   rubber 
  bands 
  did 
  not 
  affect 
  the 
  accuracy 
  of 
  the 
  interferometer 
  

   in 
  the 
  measurement 
  of 
  the 
  spark-gap, 
  at 
  least 
  so 
  far 
  as 
  the 
  

   motion 
  of 
  the 
  interferometer 
  plates 
  is 
  concerned, 
  was 
  shown 
  

   by 
  determinations 
  of 
  the 
  wave-length 
  of 
  sodium 
  light, 
  in 
  

   which 
  the 
  same 
  values 
  were 
  obtained 
  with 
  and 
  without 
  the 
  

   rubber 
  bands 
  in 
  place. 
  

  

  The 
  electrical 
  system 
  used 
  was 
  the 
  usual 
  one 
  ; 
  a 
  battery 
  

   of 
  storage-cells 
  gave 
  1200 
  volts 
  potential, 
  this 
  was 
  connected 
  

   through 
  a 
  cadmium-iodide-amyl- 
  alcohol 
  resistance 
  to 
  earth 
  ; 
  

   the 
  spark-gap 
  and 
  a 
  Weston 
  voltmeter, 
  in 
  parallel, 
  were 
  

   shunted 
  on 
  this 
  resistance 
  with 
  movable 
  electrode 
  in 
  the 
  

   iodide 
  solution, 
  so 
  the 
  potential 
  applied 
  to 
  the 
  spark-gap 
  could 
  

   be 
  varied 
  at 
  will 
  and 
  read 
  off 
  directly. 
  For 
  a 
  time, 
  an 
  

   auxiliary 
  circuit 
  with 
  2*2 
  volts 
  and 
  a 
  galvanometer 
  were 
  used 
  

   to 
  detect 
  short 
  circuit 
  of 
  the 
  spark-gap, 
  but 
  the 
  settings 
  

   made 
  in 
  this 
  way 
  proved 
  less 
  definite 
  and 
  capable 
  of 
  repetition 
  

   than 
  those 
  made 
  with 
  from 
  10 
  to 
  20 
  volts 
  on 
  the 
  regular 
  

   spark 
  circuit. 
  The 
  method 
  of 
  procedure 
  was 
  usually 
  this 
  : 
  

   With 
  the 
  electrodes 
  clearly 
  in 
  contact 
  at 
  the 
  start, 
  they 
  were 
  

   moved 
  apart 
  by 
  steps 
  of 
  *1 
  wave-length 
  (Na), 
  testing 
  with 
  the 
  

   10 
  volts 
  for 
  a 
  short 
  circuit 
  after 
  each 
  step 
  ; 
  as 
  soon 
  as 
  open 
  

   circuit 
  was 
  obtained 
  the 
  electrodes 
  were 
  then 
  separated 
  by 
  a 
  

   certain 
  number 
  of 
  wave-lengths 
  and 
  the 
  potential 
  gradually 
  

   raised 
  to 
  a 
  value 
  somewhat 
  less 
  than 
  the 
  " 
  minimum 
  potential"; 
  

   if, 
  after 
  waiting 
  a 
  sufficient 
  interval, 
  this 
  potential 
  failed 
  to 
  

   give 
  a 
  discharge, 
  the 
  spark-gap 
  was 
  decreased 
  by 
  a 
  wave- 
  

   length 
  and 
  again 
  tested 
  ; 
  when 
  this 
  process 
  had 
  been 
  repeated 
  

   until 
  the 
  spark-gap 
  was 
  unquestionably 
  extremely 
  small 
  and 
  

   had 
  been 
  subjected 
  to 
  the 
  high 
  potential 
  several 
  minutes 
  

  

  