﻿Distribution 
  of 
  Energy 
  in 
  the 
  Spectra 
  of 
  Gases. 
  807 
  

  

  point 
  V, 
  it 
  would 
  not 
  return. 
  This 
  happened 
  to 
  a 
  wearisome 
  

   extent 
  in 
  practice. 
  I£ 
  the 
  straight 
  line 
  were 
  made 
  to 
  cut 
  

   the 
  curve 
  in 
  the 
  opposite 
  sense, 
  no 
  unstable 
  regions 
  would 
  

   occur, 
  but 
  the 
  region 
  near 
  the 
  equilibrium-point 
  would 
  then 
  

   degenerate 
  much 
  sooner 
  into 
  a 
  state 
  o£ 
  neutral 
  equilibrium 
  

   as 
  the 
  slope 
  of 
  the 
  curve 
  approximated 
  to 
  that 
  of 
  the 
  

   straight 
  line. 
  

  

  The 
  smallest 
  changes 
  in 
  the 
  earth's 
  magnetic 
  field 
  are 
  

   sufficient 
  to 
  cause 
  a 
  wandering 
  of 
  the 
  spot 
  of 
  light. 
  

  

  As 
  regards 
  the 
  means 
  of 
  reading 
  the 
  mirror, 
  some 
  trouble 
  

   was 
  bestowed 
  on 
  this 
  point. 
  A 
  collection 
  of 
  astigmatic 
  

   lenses 
  was 
  obtained, 
  and, 
  after 
  trying 
  several 
  of 
  these, 
  one 
  

   was 
  found 
  which 
  would 
  give 
  a 
  fairly 
  satisfactory 
  image 
  of 
  

   a 
  silk 
  fibre 
  at 
  a 
  distance 
  of 
  8 
  metres. 
  Nearly 
  all 
  mirrors 
  

   are 
  warped 
  in 
  some 
  plane, 
  which 
  is 
  in 
  general 
  neither 
  

   horizontal 
  nor 
  vertical, 
  making 
  it 
  impossible 
  to 
  obtain 
  an 
  

   image 
  of 
  a 
  vertical 
  thread. 
  The 
  accurate 
  orientation 
  of 
  

   the 
  lens 
  was 
  in 
  the 
  present 
  case 
  of 
  extreme 
  importance, 
  

   and 
  the 
  scale 
  could 
  be 
  read 
  at 
  best 
  to 
  \ 
  mm. 
  

  

  The 
  measurement 
  of 
  the 
  current, 
  which 
  throughout 
  has 
  

   been 
  somewhat 
  difficult, 
  was 
  made 
  by 
  means 
  of 
  a 
  rough 
  

   electro-dynamometer. 
  On 
  passing 
  the 
  discharge 
  through 
  this 
  

   instrument 
  the 
  spot 
  of 
  light 
  which 
  measured 
  the 
  current 
  was 
  

   never 
  steady, 
  and 
  an 
  estimate 
  of 
  the 
  mean 
  current 
  had 
  to 
  be 
  

   obtained 
  by 
  observing 
  rapidly 
  the 
  extremes 
  of 
  the 
  swings 
  

   over 
  4 
  or 
  5 
  of 
  the 
  latter. 
  

  

  The 
  same 
  scale 
  was 
  used 
  for 
  both 
  galvanometer 
  and 
  

   dynamometer, 
  with 
  a 
  Nernst 
  lamp 
  as 
  source 
  of 
  light. 
  

  

  An 
  observation 
  involved 
  the 
  following 
  operations 
  : 
  — 
  

  

  The 
  pressure 
  w 
  T 
  as 
  read 
  first 
  of 
  ail 
  ( 
  p 
  1 
  in 
  schedule). 
  

  

  The 
  current 
  in 
  the 
  di~ 
  charge-tube 
  was 
  turned 
  on 
  and 
  

   allowed 
  to 
  run 
  for 
  a 
  short 
  time. 
  A 
  watch 
  being 
  held 
  in 
  one 
  

   hand, 
  the 
  shutter 
  was 
  opened 
  by 
  means 
  of 
  a 
  long 
  string 
  when 
  

   the 
  second-hand 
  pointed 
  to 
  " 
  60," 
  allowing 
  Ihe 
  rays 
  to 
  fall 
  on 
  

   the 
  thermopile. 
  

  

  At 
  the 
  same 
  instant 
  the 
  zero 
  reading 
  of 
  the 
  galvanometer 
  

   was 
  taken 
  (r 
  in 
  schedule). 
  Readings 
  r 
  , 
  r 
  l9 
  &c. 
  w^ere 
  taken 
  

   every 
  5 
  seconds, 
  beginning 
  with 
  the 
  zero 
  reading. 
  

  

  At 
  the 
  ^-minute 
  these 
  readings 
  were 
  discontinued, 
  and 
  a 
  

   few 
  readings 
  of 
  the 
  dynamometer 
  were 
  jotted 
  down 
  during 
  

   the 
  next 
  ^-minute 
  (<2ia 
  2 
  , 
  bj> 
  2 
  , 
  &c). 
  At 
  the 
  end 
  of 
  1 
  minute 
  

   the 
  galvanometer 
  was 
  again 
  read 
  (r) 
  and 
  the 
  current 
  was 
  

   stopped. 
  

  

  Another 
  minute 
  was 
  allowed 
  to 
  elapse 
  and 
  the 
  zero 
  reading 
  

   again 
  taken 
  (r 
  '). 
  A 
  final 
  reading 
  of 
  the 
  pressure 
  completed 
  

   the 
  observation 
  (/? 
  2 
  ). 
  

  

  