﻿[5] 
  

  

  REPORT 
  ON 
  THERMOMETERS. 
  

  

  189 
  

  

  Corrections 
  are 
  furnislied 
  to 
  the 
  nearest 
  (estimated) 
  0.1°, 
  with 
  a 
  prob- 
  

   able 
  error 
  in 
  estimation 
  not 
  exceeding 
  0.2°. 
  

  

  I 
  have 
  tested 
  a 
  number 
  of 
  different 
  observers, 
  and 
  find 
  that 
  the 
  prob- 
  

   able 
  parallax 
  error 
  in 
  reading, 
  by 
  those 
  who 
  use 
  the 
  thermometers 
  in 
  

   practice, 
  is 
  not 
  far 
  from 
  0.3°. 
  It 
  seems 
  to 
  be 
  a 
  difficult 
  thing 
  to 
  hold 
  a 
  

   thermometer 
  perpendicularly 
  opposite 
  the 
  eye, 
  some 
  observers 
  tipping 
  

   it 
  forward 
  a 
  little, 
  and 
  some 
  backward, 
  with 
  a 
  consequent 
  change 
  in 
  

   the 
  apparent 
  relative 
  positions 
  of 
  the 
  top 
  of 
  the 
  mercury 
  column 
  and 
  

   of 
  the 
  scale 
  behind 
  it. 
  This 
  cause 
  of 
  error 
  applies 
  to 
  all 
  observations 
  

   made 
  previous 
  to 
  last 
  June, 
  when 
  a 
  reading 
  lens 
  was 
  contrived 
  which 
  

   now 
  insures 
  uniformity. 
  (See 
  p. 
  [28].) 
  

  

  There 
  is 
  a 
  probable 
  small 
  inaccuracy 
  in 
  the 
  comparisons 
  of 
  "Miller- 
  

   Casella" 
  thermometers, 
  due 
  to 
  the 
  difficulty 
  of 
  reproducing 
  in 
  laboratory 
  

   comparisons 
  the 
  great 
  pressures 
  met 
  with 
  at 
  considerable 
  depths 
  in 
  the 
  

   sea, 
  which 
  will 
  be 
  discussed 
  more 
  fully 
  hereafter. 
  

  

  Up 
  to 
  May 
  1, 
  1885, 
  185 
  thermometers 
  have 
  been 
  compared 
  in 
  one 
  or 
  

   other 
  of 
  the 
  ways 
  described. 
  Of 
  this 
  number, 
  60 
  were 
  Negretti-Zam- 
  

   bra 
  deep-sea 
  thermometers 
  ; 
  15 
  were 
  Miller-Casella 
  deep-sea 
  thermom- 
  

   eters 
  ; 
  31 
  were 
  Wilder 
  protected 
  water 
  thermometers 
  ; 
  1G 
  Wilder 
  deck 
  

   thermometers 
  ; 
  14 
  were 
  salinometer 
  thermometers 
  : 
  7 
  were 
  Green 
  deck 
  

   thermometers; 
  12 
  were 
  hygrometer 
  thermometers; 
  C 
  were 
  standard 
  

   thermometers 
  ; 
  and 
  24 
  were 
  various 
  patterns 
  no 
  longer 
  in 
  use. 
  

  

  When 
  issued 
  each 
  thermometer 
  is 
  accompanied 
  by 
  a 
  printed 
  blank, 
  

   corresponding 
  to 
  a 
  stub 
  slip 
  in 
  the 
  rating 
  book, 
  and 
  filled 
  out 
  for 
  

   each 
  poiut 
  at 
  which 
  a 
  comparison 
  is 
  made. 
  Following 
  is 
  a 
  copy 
  of 
  one 
  

   of 
  these 
  comparison 
  blanks, 
  with 
  corresponding 
  stub 
  slips, 
  the 
  correc- 
  

   tions 
  being 
  stated, 
  for 
  convenience, 
  on 
  the 
  blank 
  issued, 
  at 
  intervals 
  of 
  

   10°, 
  near 
  the 
  readings 
  actually 
  taken. 
  

  

  U. 
  S. 
  COMMISSION, 
  FISH 
  

   AND 
  FISHERIES. 
  

  

  THERMOMETER 
  RATING. 
  

  

  F. 
  C. 
  No. 
  _. 
  

   Maker's 
  No. 
  

  

  By 
  

  

  CENT1C1KADK. 
  

  

  U. 
  S. 
  COMMISSION, 
  FISH 
  AND 
  FISHERIES 
  

  

  KAHKENnEir. 
  THERMOMETER 
  RATING. 
  

  

  Rated 
  with 
  attachments. 
  

  

  F. 
  C. 
  No. 
  Maker's 
  No. 
  

  

  By 
  , 
  . 
  

  

  Corrections 
  to 
  ho 
  applied 
  to 
  the 
  scale 
  readings, 
  determined 
  by 
  

   comparison 
  with 
  the 
  standard 
  instruments 
  of 
  the 
  Fish 
  Com- 
  

   mission. 
  

  

  Note 
  I. 
  — 
  When 
  the 
  sign 
  of 
  the 
  correction 
  is 
  +, 
  tho 
  quantity 
  

   is 
  to 
  be 
  added 
  to 
  the 
  observod 
  reading, 
  and 
  when 
  

   — 
  , 
  to 
  bo 
  subtracted 
  from 
  it. 
  

   II. 
  — 
  This 
  instrument 
  should 
  bo 
  returned 
  within 
  two 
  

   years 
  to 
  be 
  tested 
  again. 
  

  

  Washington, 
  D. 
  C, 
  

  

  188 
  . 
  

  

  