﻿268 
  Mr. 
  C. 
  C. 
  Paterson 
  on 
  the 
  Proposed 
  

  

  if 
  the 
  same 
  order 
  of 
  accuracy 
  is 
  to 
  be 
  attained 
  that 
  is 
  possible 
  

   with 
  a 
  much 
  smaller 
  number 
  of 
  electric 
  comparisons. 
  It 
  must 
  

   further 
  be 
  remembered 
  when 
  considering 
  the 
  question 
  o£ 
  

   photometric 
  measurements 
  to 
  two 
  or 
  three 
  parts 
  in 
  a 
  thousand 
  

   that 
  the 
  estimation 
  o£ 
  the 
  height 
  of 
  the 
  flame 
  in 
  some 
  lamps, 
  

   or 
  the 
  exact 
  reproduction 
  o£ 
  the 
  standard 
  conditions, 
  may 
  

   not 
  be 
  identical 
  when 
  carried 
  out 
  by 
  different 
  observers. 
  

   Hence 
  it 
  is 
  conceivable 
  that 
  owing 
  to 
  this 
  cause 
  the 
  observa- 
  

   tions 
  in 
  one 
  laboratory 
  on 
  some 
  flame 
  standard 
  may 
  differ 
  

   consistently 
  by 
  a 
  small 
  amount 
  from 
  those 
  in 
  another 
  on 
  the 
  

   same 
  standard. 
  This, 
  however, 
  is 
  not 
  the 
  case 
  when 
  electric 
  

   sub-standard 
  comparisons 
  are 
  made 
  if 
  the 
  electrical 
  measuring- 
  

   apparatus 
  is 
  accurate. 
  To 
  a 
  certain 
  extent, 
  therefore 
  (in 
  

   some 
  cases 
  more 
  than 
  others), 
  a 
  flame 
  standard 
  needs 
  to 
  be 
  

   " 
  interpreted 
  " 
  when 
  its 
  absolute 
  value 
  is 
  desired 
  to 
  a 
  high 
  

   accuracy. 
  

  

  The 
  value 
  of 
  electric 
  sub-standards 
  comparisons 
  thus 
  

   becomes 
  apparent. 
  I£ 
  (as 
  is 
  generally 
  the 
  case) 
  a 
  Laboratory 
  

   has 
  sets 
  of 
  electric 
  sub-standards 
  which 
  have 
  been 
  compared 
  

   at 
  intervals 
  for 
  years 
  with 
  the 
  primary 
  flame 
  standard 
  whose 
  

   value 
  they 
  represent, 
  an 
  opportunity 
  is 
  given 
  for 
  realising 
  

   the 
  absolute 
  value 
  of 
  this 
  unit 
  to 
  an 
  accuracy 
  which 
  could 
  

   hardly 
  be 
  attained 
  with 
  certainty 
  by 
  others 
  who 
  might 
  

   endeavour 
  to 
  reproduce 
  it 
  in 
  a 
  single 
  set 
  of 
  observations, 
  

   however 
  carefully 
  made. 
  When 
  these 
  electric 
  sub-standards 
  

   are 
  intercompared 
  through 
  the 
  medium 
  o£ 
  a 
  travelling 
  set 
  of 
  

   lamps, 
  there 
  is 
  no 
  reason 
  why 
  we 
  should 
  not 
  obtain 
  accurate 
  

   knowledge 
  of 
  the 
  relative 
  values 
  of 
  the 
  different 
  units 
  as 
  

   each 
  is 
  interpreted 
  in 
  the 
  country 
  where 
  it 
  is 
  the 
  recognized 
  

   standard. 
  

  

  The 
  ratios 
  between 
  the 
  four 
  units 
  of 
  light 
  given 
  in 
  the 
  

   Table 
  are 
  the 
  results 
  of 
  measurements 
  which 
  have 
  been 
  made 
  

   at 
  the 
  specified 
  laboratories 
  in 
  the 
  four 
  countries 
  concerned. 
  

   Other 
  determinations 
  were 
  made 
  previous 
  to 
  these 
  *, 
  but 
  the 
  

   standards 
  used 
  £or 
  obtaining 
  the 
  British 
  Unit 
  were 
  o£ 
  several 
  

   different 
  forms 
  and 
  the 
  atmospheric 
  conditions 
  have 
  not 
  

   always 
  been 
  taken 
  into 
  consideration. 
  I 
  have 
  deemed 
  it 
  

   desirable, 
  therefore, 
  to 
  insert 
  only 
  the 
  more 
  recent 
  deter- 
  

   minations, 
  in 
  all 
  of 
  which 
  the 
  10 
  Candle 
  Harcourt 
  Pentane 
  

   Lamp 
  has 
  been 
  used 
  and 
  atmospheric 
  changes 
  have 
  been 
  

   allowed 
  for. 
  

  

  The 
  Table 
  is 
  divided 
  into 
  two 
  portions. 
  Columns 
  1 
  to 
  9 
  

   give 
  the 
  various 
  ratios 
  obtained 
  through 
  the 
  medium 
  of 
  

  

  * 
  For 
  a 
  discussion 
  of 
  these, 
  see 
  J. 
  A. 
  Fleming, 
  '' 
  The 
  Pliotometry 
  of 
  

   Electric 
  Lamps," 
  ref. 
  cit. 
  

  

  