﻿188 
  REPORT 
  OF 
  COMMISSIONER 
  OP 
  PISH 
  AND 
  FISHERIES. 
  [4 
  J 
  

  

  time 
  to 
  time, 
  a 
  sensibly 
  constant 
  temperature 
  will 
  at 
  length 
  be 
  reached, 
  

   at 
  which 
  the 
  gain 
  in 
  temperature 
  of 
  the 
  water 
  in 
  the 
  inner 
  jar 
  by 
  contact 
  

   with 
  the 
  warmer 
  air 
  at 
  its 
  surface 
  is 
  very 
  satisfactorily 
  compensated 
  

   by 
  its 
  loss 
  through 
  the 
  air 
  space 
  between 
  B 
  and 
  C 
  and 
  the 
  badly 
  con- 
  

   ducting 
  walls 
  of 
  B. 
  For 
  temperatures 
  higher 
  than 
  that 
  of 
  the 
  air, 
  the 
  

   water 
  in 
  the 
  outer 
  jar 
  must 
  be 
  warmer 
  than 
  that 
  in 
  the 
  inner. 
  No 
  pos- 
  

   itive 
  rule 
  for 
  differences 
  in 
  the 
  temperatures 
  of 
  the 
  water 
  in 
  the 
  inner 
  

   and 
  outer 
  jar 
  can 
  be 
  established. 
  It 
  may 
  be 
  said, 
  however, 
  in 
  general 
  

   terms, 
  that 
  the 
  greater 
  the 
  difference 
  between 
  the 
  temperature 
  of 
  the 
  

   air 
  and 
  that 
  desired 
  for 
  the 
  comparison, 
  the 
  greater 
  should 
  be 
  the 
  differ- 
  

   ence 
  between 
  the 
  temperatures 
  of 
  water 
  in 
  the 
  outer 
  and 
  inner 
  jars. 
  

  

  On 
  the 
  13th 
  March, 
  1884, 
  seven 
  Negretti-Zambra 
  thermometers 
  were 
  

   compared 
  at 
  every 
  10° 
  from 
  32° 
  to 
  82°, 
  and 
  I 
  find 
  it 
  noted 
  that 
  " 
  the 
  appa- 
  

   ratus 
  worked 
  well, 
  showing 
  no 
  change 
  in 
  forty 
  -five 
  minutes 
  exceeding 
  ' 
  

   0.1° 
  at 
  any 
  temperature 
  below 
  80°." 
  On 
  September 
  10, 
  it 
  is 
  noted 
  that 
  

   " 
  there 
  was 
  much 
  difficulty, 
  temperature 
  of 
  room 
  being 
  80°, 
  in 
  keeping 
  

   the 
  water 
  near 
  a 
  constant 
  low 
  temperature. 
  The 
  change 
  was 
  very 
  reg- 
  

   ularly 
  0.2° 
  every 
  five 
  minutes." 
  On 
  September 
  11, 
  air 
  being 
  S2°, 
  ice 
  and 
  

   water 
  in 
  outer 
  jar, 
  temperatures 
  of 
  water 
  in 
  the 
  inner 
  jar 
  varied 
  as 
  fol- 
  

   lows 
  : 
  At 
  12 
  hours 
  40 
  minutes 
  water 
  in 
  inner 
  jar 
  is 
  38° 
  ; 
  at 
  12 
  hours 
  45 
  

   minutes 
  water 
  in 
  inner 
  jar 
  is 
  38.2° 
  ; 
  at 
  12 
  hours 
  50 
  minutes 
  water 
  in 
  

   inner 
  jar 
  is 
  38.0°; 
  at 
  12 
  hours 
  55 
  minutes 
  water 
  in 
  inner 
  jar 
  is 
  38.0° 
  ; 
  

   at 
  1 
  hour 
  minutes 
  water 
  in 
  inner 
  jar 
  is 
  39° 
  ; 
  at 
  1 
  hour 
  10 
  minutes 
  

   water 
  in 
  inner 
  jar 
  is 
  39°, 
  at 
  which 
  last 
  figure 
  the 
  comparisons 
  were 
  made. 
  

  

  The 
  above 
  will 
  serve 
  as 
  an 
  example 
  of 
  numerous 
  series 
  of 
  observa- 
  

   tions, 
  with 
  widely 
  differing 
  results. 
  Sometimes 
  more 
  than 
  an 
  hour's 
  

   patient 
  watching 
  is 
  required 
  ; 
  at 
  others 
  the 
  constant 
  temperature 
  is 
  

   reached 
  in 
  a 
  few 
  minutes, 
  and 
  may 
  be 
  maintained 
  long 
  enough 
  for 
  prac- 
  

   tical 
  purposes 
  by 
  cautiously 
  adding 
  cold 
  or 
  warm 
  water, 
  as 
  the 
  case 
  

   may 
  be, 
  to 
  that 
  in 
  the 
  outer 
  jar. 
  

  

  To 
  avoid 
  parallax 
  error 
  in 
  reading, 
  the 
  jars 
  are 
  levelled, 
  and 
  readings 
  

   taken 
  by 
  aid 
  of 
  a 
  hand-lens, 
  with 
  the 
  eye 
  and 
  top 
  of 
  the 
  mercury 
  col- 
  

   umn 
  at 
  the 
  level 
  of 
  the 
  top 
  of 
  the 
  outer 
  jar, 
  across 
  the 
  two 
  sides 
  of 
  

   which 
  the 
  reading 
  is 
  ''sighted," 
  the 
  thermometer 
  being 
  held 
  in 
  contact 
  

   with 
  one 
  of 
  the 
  walls 
  of 
  the 
  jar, 
  and 
  parallel 
  with 
  the 
  central 
  spindle 
  of 
  

   the 
  frame, 
  to 
  insure 
  its 
  perpendicularity. 
  Comparisons 
  of 
  readings 
  

   taken 
  in 
  this 
  simple 
  way 
  with 
  readings 
  taken 
  by 
  cathetometer, 
  the 
  

   thermometer 
  being 
  secured 
  in 
  a 
  perpendicular 
  position, 
  show 
  no 
  per- 
  

   ceptible 
  error. 
  

  

  Since 
  none 
  of 
  the 
  thermometers 
  in 
  general 
  use 
  by 
  the 
  Commission 
  are 
  

   pointed 
  to 
  divisions 
  less 
  than 
  1° 
  F., 
  or 
  need 
  to 
  be 
  corrected 
  for 
  its 
  pur- 
  

   poses 
  at 
  temperatures 
  above 
  100°, 
  the 
  contrivances 
  above 
  described 
  

   have 
  been 
  found 
  to 
  afford 
  as 
  great 
  accuracy 
  as 
  is 
  practically 
  required.* 
  

  

  * 
  For 
  a 
  good 
  description, 
  Avitli 
  illustrations, 
  of 
  the 
  more 
  exact 
  methods 
  followed 
  at 
  

   the 
  Kew 
  Observatory, 
  see 
  Mr. 
  Francis 
  Galton's 
  paper 
  in 
  ilie 
  Proceedings 
  of 
  the 
  Royal 
  

   Society 
  of 
  London 
  for 
  March 
  15, 
  1877 
  (vol. 
  xxvi, 
  p! 
  84), 
  entitled 
  "A 
  description 
  of 
  

   the 
  process 
  for 
  verifying 
  thermometers 
  at 
  the 
  Kew 
  Observatory." 
  

  

  