﻿Telescopic 
  Vision. 
  971 
  

  

  84. 
  To 
  make 
  use 
  of 
  these 
  diagrams 
  we 
  must 
  remember 
  

   that 
  the 
  ' 
  complete 
  image 
  ' 
  in 
  the 
  pro-telescope 
  is 
  due 
  to 
  the 
  

   simultaneous 
  presence 
  in 
  that 
  telescope 
  of 
  all 
  the 
  partials, 
  

   and 
  that 
  a 
  partial 
  is 
  produced 
  whenever 
  the 
  whole 
  or 
  any 
  

   part 
  o£ 
  the 
  concentration 
  image 
  represented 
  in 
  fig. 
  1 
  is 
  able 
  

   to 
  enter 
  the 
  pro-telescope. 
  Now 
  fig. 
  1, 
  a' 
  is 
  the 
  concentration 
  

   image 
  formed 
  by 
  a 
  stellade 
  of 
  light 
  incident 
  upon 
  the 
  pro- 
  

   Neptune 
  at 
  z' 
  ; 
  and 
  obviously 
  if 
  the 
  direction 
  of 
  the 
  incident 
  

   stellade 
  to 
  which 
  it 
  is 
  due, 
  brings 
  its 
  middle 
  point 
  o 
  anywhere 
  

   within 
  space 
  k, 
  then 
  the 
  whole 
  of 
  the 
  efficient 
  light 
  of 
  that 
  

   concentration 
  image 
  has 
  got 
  into 
  the 
  pro-telescope, 
  and 
  

   therefore 
  the 
  ' 
  partial,' 
  the 
  image 
  of 
  z- 
  which 
  it 
  will 
  form, 
  

   will 
  be 
  an 
  image 
  so 
  good 
  that 
  we 
  may 
  practically 
  regard 
  it 
  

   •as 
  an 
  optimum 
  partial. 
  On 
  the 
  other 
  hand, 
  if 
  othe 
  centre 
  of 
  

   fig. 
  1 
  falls 
  anywhere 
  within 
  space 
  I, 
  only 
  part, 
  although 
  it 
  

   will 
  be 
  the 
  greater 
  part, 
  of 
  the 
  concentration 
  image 
  gains 
  

   admission 
  to 
  the 
  pro-telescope. 
  The 
  resulting 
  ' 
  partial 
  ' 
  

   formed 
  at 
  x 
  l 
  will 
  fall 
  short 
  of 
  being 
  the 
  optimum, 
  the 
  defect 
  

   being 
  slight 
  if 
  o, 
  the 
  centre 
  of 
  the 
  concentration 
  image, 
  

   reaches 
  a 
  part 
  of 
  region 
  / 
  which 
  is 
  distant 
  from 
  the 
  margin 
  

   of 
  the 
  objective, 
  while 
  the 
  defect 
  goes 
  on 
  increasing 
  the 
  

   nearer 
  that 
  o 
  comes 
  to 
  the 
  margin 
  of 
  the 
  objective. 
  This 
  I 
  

   group 
  of 
  partials 
  may 
  be 
  called 
  fair 
  images 
  of 
  z', 
  and 
  will 
  

   ^ome 
  of 
  them 
  be 
  almost 
  good, 
  others 
  only 
  mediocre. 
  

  

  The 
  next 
  stage 
  is 
  when 
  o 
  the 
  centre 
  of 
  the 
  concentration 
  

   image 
  reaches 
  space 
  m. 
  Under 
  these 
  circumstances 
  part 
  

   •even 
  of 
  the 
  light 
  of 
  its 
  central 
  boss 
  is 
  excluded 
  from 
  the 
  pro- 
  

   telescope 
  and 
  the 
  resulting 
  partial 
  will 
  be 
  a 
  decidedly 
  bad 
  

   image 
  of 
  z\ 
  and 
  this 
  defect 
  becomes 
  still 
  worse 
  when 
  the 
  

   centre 
  of 
  the 
  concentration 
  image 
  lies 
  farther 
  from 
  the 
  

   objective 
  within 
  region 
  n. 
  If 
  the 
  centre 
  of 
  the 
  concentration 
  

   image 
  lies 
  farther 
  out 
  than 
  region 
  n, 
  none 
  of 
  its 
  light 
  gets 
  

   access 
  to 
  the 
  pro-telescope 
  and 
  we 
  are 
  therefore 
  not 
  concerned 
  

   with 
  it. 
  Now 
  when 
  the 
  light 
  incident 
  upon 
  z' 
  is 
  diffused 
  

   light, 
  as 
  it 
  must 
  be 
  when 
  we 
  want 
  to 
  represent 
  correctly 
  

   what 
  occurs 
  in 
  nature, 
  this 
  light 
  when 
  analysed 
  will 
  furnish 
  

   stellades 
  that 
  would 
  send 
  the 
  centres 
  of 
  their 
  concentration 
  

   images 
  to 
  all 
  the 
  points 
  represented 
  in 
  fig. 
  3, 
  so 
  that 
  all 
  the 
  

   partials 
  which 
  have 
  been 
  described 
  above 
  are 
  present 
  upon 
  

   disk 
  x' 
  and 
  cooperate 
  to 
  produce 
  there 
  the 
  ' 
  complete 
  image 
  ' 
  

   of 
  z'. 
  It 
  is 
  obvious 
  that 
  in 
  the 
  state 
  of 
  things 
  represented 
  in 
  

   fig. 
  2, 
  b 
  f 
  , 
  this 
  complete 
  image 
  cannot 
  be 
  a 
  very 
  satisfactory 
  

   one. 
  It 
  is 
  to 
  be 
  regretted 
  that 
  when 
  using 
  a 
  telescope 
  we 
  

   are 
  unable 
  to 
  exclude 
  any 
  of 
  the 
  partials. 
  It 
  is 
  convenient 
  to 
  

   divide 
  them 
  into 
  the 
  four 
  groups 
  which 
  we 
  call 
  him 
  and 
  n, 
  

   depending 
  upon 
  the 
  situations 
  on 
  plane 
  Y' 
  (the 
  plane 
  of 
  the 
  

  

  