﻿[11] 
  EEPORT 
  ON 
  THERMOMETERS. 
  195 
  

  

  pression, 
  but 
  the 
  amount 
  of 
  heat 
  developed 
  is 
  dependent 
  in 
  a 
  very 
  curious 
  

   manner 
  upon 
  its 
  original 
  temperature. 
  If 
  compressed 
  at 
  the 
  temperature 
  

   of 
  its 
  maximum 
  density, 
  the 
  water 
  is 
  neither 
  heated 
  nor 
  cooled, 
  but 
  is 
  

   heated 
  when 
  compressed 
  at 
  a 
  temperature 
  above, 
  and 
  cooled 
  when 
  com- 
  

   pressed 
  at 
  a 
  temperature 
  below 
  its 
  maximum 
  density 
  ; 
  and 
  the 
  wider 
  

   the 
  divergence 
  of 
  its 
  original 
  temperature 
  from 
  that 
  of 
  its 
  maximum 
  

   density, 
  the 
  greater 
  is 
  tbe 
  effect 
  produced. 
  Captain 
  Davis 
  combined 
  

   one 
  set 
  of 
  observations 
  taken 
  near, 
  but 
  below, 
  the 
  temperature 
  of 
  maxi- 
  

   mum 
  density, 
  with 
  a 
  number 
  taken 
  near 
  55° 
  F., 
  striking 
  out 
  (unfortu- 
  

   nately), 
  as 
  probably 
  erroneous, 
  all 
  observations 
  which 
  differed 
  much 
  

   from 
  the 
  majority 
  of 
  the 
  others. 
  By 
  an 
  interesting 
  graphic 
  diagram 
  

   Tait 
  shows 
  that 
  the 
  true 
  figures 
  for 
  temperature 
  correction, 
  according 
  

   to 
  Davis's 
  experiments, 
  lie 
  in 
  a 
  line 
  coinciding 
  much 
  more 
  nearly 
  with 
  

   that 
  indicated 
  by 
  his 
  rejected 
  observations 
  than 
  by 
  those 
  which 
  he 
  

   adopted. 
  

  

  Professor 
  Tait 
  found 
  by 
  experiment 
  that 
  a 
  Phillips 
  self-registering 
  

   mercurial 
  thermometer,* 
  wholly 
  inclosed 
  in 
  a 
  sealed 
  glass 
  tube, 
  nearly 
  

   filled 
  with 
  alcohol, 
  as 
  suggested 
  by 
  Sir 
  William 
  Thomson, 
  t 
  was 
  "abso- 
  

   lutely 
  perfect, 
  so 
  far 
  as 
  regards 
  immunity 
  from 
  pressure" 
  (p. 
  7). 
  So 
  that 
  

   the 
  pressure 
  error 
  of 
  the 
  Miller-Casella 
  thermometer, 
  the 
  bulb 
  of 
  which 
  is 
  

   protected 
  in 
  precisely 
  the 
  same 
  way, 
  is 
  due 
  almost, 
  if 
  not 
  quite, 
  entirely 
  

   to 
  pressure 
  upon 
  the 
  stem. 
  For 
  tubes 
  of 
  uniform 
  caliber 
  throughout, 
  

   it 
  was 
  found 
  by 
  experiment 
  that 
  the 
  effect 
  of 
  pressure 
  upon 
  a 
  tube 
  

   similar 
  to 
  those 
  of 
  these 
  thermometers 
  would 
  be 
  an 
  elongation 
  of 
  about 
  

   T 
  _i__ 
  of 
  the 
  length 
  of 
  the 
  column 
  of 
  mercury 
  for 
  each 
  ton 
  weight 
  ap- 
  

   plied 
  to 
  the 
  outside 
  of 
  the 
  tube 
  (or 
  about 
  800 
  fathoms 
  of 
  depth). 
  As 
  

   the 
  elongation 
  will 
  occur 
  iu 
  both 
  legs 
  of 
  the 
  U, 
  and 
  as 
  increase 
  in 
  press- 
  

   ure 
  is 
  in 
  practice 
  (at 
  sea), 
  associated 
  with 
  decrease 
  in 
  temperature, 
  this 
  

   correction 
  for 
  pressure 
  should 
  have 
  been 
  applied 
  also 
  to 
  the 
  minimum 
  

   scale, 
  instead 
  of, 
  as 
  was 
  the 
  fact, 
  to 
  the 
  maximum 
  only. 
  

  

  As 
  the 
  instruments 
  used 
  by 
  the 
  Challenger 
  were 
  actually 
  constructed, 
  

   each 
  leg 
  of 
  the 
  U 
  contained 
  an 
  " 
  aneurism," 
  or 
  small 
  enlargement,! 
  near 
  

   the 
  bend, 
  intended 
  to 
  facilitate 
  recovery 
  of 
  the 
  steel 
  index 
  when 
  it 
  had 
  

   been 
  lost 
  in 
  the 
  mercury. 
  These 
  swellings 
  appear 
  to 
  be 
  larger 
  than 
  they 
  

   really 
  are 
  in 
  the 
  ratio 
  of 
  1.6 
  (the 
  refractive 
  index 
  of 
  glass) 
  to 
  unity, 
  but 
  

   were 
  actually 
  found 
  by 
  Tait, 
  in 
  several 
  instances, 
  to 
  contain 
  five 
  times 
  

   as 
  much 
  mercury 
  as 
  a 
  similar 
  length 
  of 
  thermometer 
  tube, 
  and, 
  conse- 
  

   quently, 
  to 
  produce 
  five 
  times 
  as 
  great 
  an 
  error. 
  

  

  * 
  A 
  mercurial 
  thermometer, 
  self-registering 
  by 
  an 
  index 
  produced 
  by 
  a 
  break 
  in 
  the 
  

   column. 
  Invented 
  by 
  Prof. 
  John 
  Phillips, 
  of 
  Oxford. 
  First 
  used 
  at 
  Kew 
  in 
  1851. 
  

   The 
  principle 
  is 
  now 
  universal 
  in 
  clinical 
  thermometers. 
  

  

  t 
  "The 
  effect 
  of 
  pressure 
  in 
  lowering 
  the 
  freezing 
  point 
  of 
  water." 
  Proc. 
  R. 
  S. 
  E. 
  

   February, 
  1850. 
  

  

  t 
  The 
  "aneurisms" 
  were 
  first 
  added 
  to 
  Sixe's 
  thermometers 
  by 
  Aime", 
  in 
  1844, 
  to 
  pre- 
  

   vent 
  the 
  mercury 
  from 
  passing 
  by 
  the 
  indices. 
  Ann. 
  de 
  CMmie 
  et 
  de 
  Physique, 
  Ser. 
  3, 
  t. 
  

   xv, 
  p. 
  5 
  (1845). 
  

  

  