﻿Distribution 
  of 
  Energy 
  in 
  the 
  Spectra 
  of 
  Gases. 
  821 
  

  

  H 
  a 
  appears 
  from 
  the 
  curves 
  given 
  later 
  to 
  have 
  more 
  nearly 
  

   J 
  the 
  total 
  energy. 
  

  

  So 
  far, 
  no 
  spark-gap 
  had 
  heen 
  used 
  in 
  the 
  production 
  of 
  the 
  

   discharge, 
  and 
  the 
  currents 
  did 
  not 
  vary 
  to 
  any 
  great 
  extent. 
  

   Also, 
  as 
  was 
  mentioned 
  before, 
  the 
  discharge-tube 
  was 
  close 
  

   up 
  to 
  the 
  slit, 
  and 
  gave 
  considerable 
  disturbing 
  effects. 
  

  

  It 
  is 
  obvious 
  that 
  if 
  a 
  projecting 
  lens 
  were 
  used 
  to 
  throw 
  

   an 
  image 
  of 
  the 
  capillary 
  portion 
  of 
  the 
  tube 
  on 
  to 
  the 
  slit, 
  

   there 
  would 
  be 
  a 
  great 
  gain 
  in 
  accuracy, 
  although 
  some 
  of 
  

   the 
  energy 
  would 
  be 
  wasted. 
  The 
  conditions 
  which 
  such 
  a 
  

   projecting 
  lens 
  must 
  fulfil 
  are 
  as 
  follows 
  : 
  — 
  If 
  the 
  object, 
  or 
  

   that 
  portion 
  of 
  it 
  which 
  it 
  is 
  desired 
  to 
  observe, 
  is 
  of 
  the 
  

   same 
  size 
  as 
  the 
  slit, 
  the 
  distance 
  of 
  object 
  from 
  lens 
  must 
  be 
  

   equal 
  to 
  the 
  distance 
  of 
  lens 
  from 
  slit, 
  i.e. 
  the 
  ratio 
  of 
  aper- 
  

   ture 
  to 
  focal 
  length 
  must 
  be 
  twice 
  as 
  great 
  in 
  the 
  case 
  of 
  the 
  

   projecting 
  lens 
  as 
  in 
  the 
  case 
  of 
  the 
  collimator 
  lens. 
  It 
  is 
  

   also 
  highly 
  essential 
  that 
  the 
  lens 
  be 
  a 
  good 
  achromatic 
  one, 
  

   in 
  order 
  that 
  no 
  part 
  of 
  the 
  spectrum 
  shall 
  be 
  favoured 
  in 
  

   the 
  energy 
  measurements. 
  If 
  the 
  ratio 
  of 
  aperture 
  to 
  focal 
  

   length 
  be 
  not 
  twice 
  as 
  great 
  for 
  the 
  projecting 
  lens 
  as 
  for 
  

   the 
  collimating 
  lens, 
  some 
  of 
  the 
  surface 
  of 
  the 
  latter 
  will 
  

   not 
  be 
  traversed 
  by 
  light, 
  and 
  a 
  larger 
  lens 
  of 
  the 
  same 
  focal 
  

   length 
  is 
  what 
  is 
  desired. 
  

  

  Now 
  it 
  is 
  difficult 
  to 
  find 
  good 
  achromatic 
  lenses 
  of 
  the 
  

   aperture 
  ratio 
  desired 
  (about 
  ^), 
  especially 
  as 
  a 
  large 
  linear 
  

   aperture 
  is 
  desirable 
  on 
  account 
  of 
  the 
  necessity 
  for 
  re- 
  

   moving 
  the 
  oscillatory 
  circuit 
  as 
  far 
  as 
  possible 
  from 
  the 
  

   galvanometer. 
  

  

  At 
  the 
  beginning 
  of 
  August 
  1912. 
  an 
  achromatic 
  lens 
  of 
  

   3J 
  inches 
  linear 
  aperture 
  and 
  about 
  ^-aperture 
  ratio 
  was, 
  

   through 
  the 
  kindness 
  of 
  Mr. 
  Hincks, 
  borrowed 
  from 
  the 
  

   Observatory 
  at 
  Cambridge, 
  and 
  the 
  curves 
  of 
  distribution 
  

   of 
  energy 
  given 
  later 
  owe 
  their 
  existence 
  entirely 
  to 
  this 
  

   valuable 
  addition 
  to 
  the 
  apparatus. 
  A 
  plan 
  of 
  the 
  new 
  

   arrangement 
  of 
  the 
  whole 
  apparatus 
  is 
  given 
  in 
  fig. 
  11, 
  

   which 
  is 
  drawn 
  roughly 
  to 
  scale. 
  The 
  table 
  to 
  the 
  left 
  has 
  

   the 
  gas-generator, 
  vacuum-pump, 
  &c, 
  while 
  the 
  transformer 
  

   is 
  at 
  least 
  twice 
  as 
  far 
  to 
  the 
  left 
  of 
  the 
  galvanometer 
  as 
  the 
  

   discharge-tube 
  D. 
  

  

  Under 
  the 
  new 
  conditions, 
  even 
  when 
  a 
  large 
  spark-gap 
  

   was 
  placed 
  in 
  the 
  circuit, 
  there 
  were 
  no 
  disturbing 
  deflexions 
  

   greater 
  than 
  005 
  cm. 
  In 
  order 
  to 
  find 
  out 
  what 
  sort 
  of 
  

   constancy 
  was 
  to 
  be 
  expected 
  in 
  the 
  readings, 
  a 
  long 
  series 
  

   of 
  readings 
  at 
  various 
  intervals 
  of 
  time 
  was 
  taken. 
  

  

  The 
  result 
  of 
  these 
  was 
  that 
  the 
  deflexions 
  obtained 
  imme- 
  

   diately 
  after 
  an 
  interval 
  of 
  quiescence 
  were 
  always 
  greater 
  

  

  