﻿Inversion 
  Temperature 
  of 
  Kelvin 
  Effect 
  in 
  Hydrogen. 
  539 
  

  

  bridge. 
  The 
  whole 
  of 
  this 
  apparatus 
  is, 
  by 
  means 
  of 
  the 
  

   metallic 
  ring 
  n 
  n, 
  cemented 
  to 
  a 
  thick-walled 
  glass 
  cylinder 
  1 
  1, 
  

   inside 
  which 
  there 
  is 
  a 
  thin-walled 
  beaker 
  m 
  m, 
  which 
  serves 
  

   to 
  receive 
  the 
  cooling 
  agent. 
  

  

  Liquid 
  air, 
  liquid 
  ethylene, 
  and 
  a 
  mixture 
  of 
  solid 
  carbonic 
  

   acid 
  and 
  ether 
  were 
  used 
  as 
  cooling 
  agents. 
  The 
  air 
  and 
  

   ethylene 
  were 
  poured 
  in 
  through 
  the 
  upper 
  opening 
  of 
  the 
  

   T-shaped 
  pipe 
  k, 
  to 
  a 
  level 
  above 
  that 
  of 
  the 
  worm 
  6. 
  The 
  

   opening 
  k 
  serves 
  as 
  escape-tube 
  for 
  the 
  gases 
  of 
  the 
  cooling 
  

   substance, 
  or 
  — 
  the 
  upper 
  opening 
  having 
  been 
  closed 
  — 
  for 
  

   connecting 
  the 
  interior 
  of 
  the 
  apparatus 
  with 
  an 
  air-pump 
  for 
  

   the 
  purpose 
  of 
  lowering 
  the 
  pressure 
  and 
  temperature. 
  In 
  

   the 
  experiments 
  with 
  solid 
  carbon 
  dioxide, 
  this 
  substance 
  was 
  

   introduced 
  into 
  the 
  apparatus 
  before 
  fixing 
  the 
  cover 
  nn. 
  

   For 
  measuring 
  the 
  pressure 
  of 
  the 
  gases 
  coming 
  from 
  the 
  

   cooling 
  substance, 
  a 
  mercury 
  manometer 
  was 
  used, 
  which 
  was 
  

   also 
  in 
  connexion 
  with 
  the 
  pipe 
  k, 
  but 
  which 
  is 
  not 
  shown 
  in 
  

   the 
  figure. 
  

  

  Descrijrfion 
  of 
  the 
  JExperiment8. 
  

  

  In 
  the 
  experiments 
  with 
  liquid 
  air 
  as 
  the 
  cooling 
  substance, 
  

   the 
  initial 
  temperature 
  was 
  about 
  —190°, 
  and 
  the 
  pressure 
  

   of 
  the 
  hydrogen 
  before 
  expansion 
  about 
  170 
  atmospheres. 
  

   The 
  expansion 
  took 
  place 
  slowly 
  and 
  lasted 
  from 
  4 
  to 
  5 
  seconds. 
  

   The 
  cooling 
  was 
  considerable 
  : 
  the 
  galvanometer-deflexion 
  

   amounted 
  to 
  about 
  200 
  mm. 
  scale-divisions. 
  In 
  consequence 
  

   of 
  this, 
  it 
  became 
  necessary 
  to 
  use 
  a 
  less 
  powerful 
  cooling 
  

   agent; 
  for 
  this 
  purpose 
  liquid 
  ethylene 
  appeared 
  suitable. 
  In 
  

   the 
  experiments 
  with 
  this 
  substance 
  the 
  initial 
  temperature 
  

   was 
  about 
  — 
  103°, 
  the 
  initial 
  pressure 
  about 
  150 
  atmospheres. 
  

   In 
  this 
  case 
  also 
  the 
  temperature 
  fell 
  during 
  expansion, 
  but 
  to 
  

   a 
  much 
  smaller 
  extent 
  than 
  when 
  liquid 
  air 
  was 
  employed 
  : 
  

   the 
  galvanometer-throw 
  amounting 
  to 
  about 
  30 
  mm 
  on 
  the 
  

   scale. 
  Thus 
  it 
  appeared 
  that 
  the 
  temperature 
  of 
  liquid 
  ethy- 
  

   lene 
  was 
  too 
  low 
  ; 
  but 
  for 
  the 
  production 
  of 
  higher 
  tempera- 
  

   tures 
  this 
  gas 
  is 
  unsuitable, 
  hence 
  in 
  the 
  third 
  series 
  of 
  experi- 
  

   ments 
  a 
  mixture 
  of 
  solid 
  carbonic 
  acid 
  and 
  ether 
  was 
  employed. 
  

   The 
  initial 
  temperature 
  was 
  —78°, 
  the 
  initial 
  pressure 
  about 
  

   117 
  atmospheres. 
  In 
  this 
  series 
  of 
  experiments 
  25 
  expansions 
  

   were 
  observed, 
  during 
  which 
  the 
  pressure 
  in 
  the 
  steel 
  cylinder 
  

   fell 
  from 
  117 
  to 
  110 
  atmospheres; 
  i. 
  e., 
  during 
  a 
  single 
  expan- 
  

   sion 
  the 
  pressure 
  of 
  the 
  hydrogen 
  was 
  altered 
  by 
  about 
  £ 
  of 
  

   an 
  atmosphere, 
  the 
  volume 
  of 
  the 
  expanded 
  gas 
  (measured 
  

   under 
  atmospheric 
  pressure) 
  amounting 
  to 
  about 
  3£ 
  litres. 
  

   This 
  point 
  is 
  specially 
  mentioned 
  because 
  in 
  some 
  previous 
  

   experiments, 
  not 
  recorded 
  here, 
  in 
  which 
  the 
  fall 
  of 
  pressure 
  

  

  