﻿Chemistry 
  and 
  Physics. 
  299 
  

  

  . 
  6. 
  The 
  Thermal 
  Conductivity 
  of 
  Air. 
  — 
  The 
  specific 
  thermal 
  

   conductivity, 
  k, 
  of 
  a 
  gas 
  is 
  connected 
  with 
  the 
  specific 
  heat 
  at 
  

   constant 
  volume, 
  C, 
  and 
  the 
  coefficient 
  of 
  viscosity, 
  rj, 
  according 
  

   to 
  the 
  relation 
  k 
  = 
  frjC. 
  The 
  value 
  of 
  the 
  constant 
  /, 
  derived 
  

   from 
  purely 
  theoretical 
  considerations, 
  varies 
  from 
  1-25 
  to 
  2-5 
  

   according 
  to 
  the 
  law 
  of 
  force 
  assumed 
  to 
  act 
  during 
  a 
  molecular 
  

   collision. 
  The 
  above 
  equation 
  shows 
  that 
  if 
  k, 
  rj, 
  and 
  C 
  could 
  be 
  

   determined 
  experimentally, 
  and 
  to 
  a 
  sufficiently 
  high 
  degree 
  of 
  

   accuracy, 
  then 
  / 
  could 
  be 
  calculated 
  at 
  once, 
  and 
  a 
  choice 
  between 
  

   the 
  various 
  theoretical 
  hypotheses 
  might 
  be 
  made. 
  Satisfactory 
  

   experimental 
  data 
  for 
  rj 
  and 
  C 
  have 
  been 
  obtained, 
  but 
  the 
  values 
  

   of 
  h, 
  found 
  for 
  the 
  same 
  gas 
  by 
  different 
  observers, 
  are 
  very 
  dis- 
  

   cordant. 
  For 
  example, 
  the 
  value 
  of 
  k 
  for 
  dry 
  air 
  at 
  0°C. 
  

   ranges 
  from 
  4-47 
  X 
  10~ 
  5 
  (Eckerlein) 
  to 
  5-86 
  X 
  10~ 
  5 
  (Miiller), 
  the 
  

   dimensions 
  being 
  gram 
  calories 
  per 
  cm. 
  per 
  sec. 
  per 
  degree 
  C. 
  

   The 
  chief 
  cause 
  of 
  the 
  wide 
  variations 
  seems 
  to 
  lie 
  in 
  the 
  fact 
  

   that 
  nearly 
  all 
  of 
  the 
  investigations 
  have 
  been 
  carried 
  out 
  in 
  

   such 
  a 
  manner 
  as 
  to 
  permit 
  convection 
  to 
  introduce 
  relatively 
  

   large 
  and 
  indeterminate 
  errors. 
  On 
  the 
  other 
  hand, 
  the 
  

   researches 
  of 
  Eucken 
  (1911 
  and 
  1913) 
  on 
  the 
  relative 
  thermal 
  

   conductivity 
  of 
  a 
  fairly 
  large 
  number 
  of 
  gases 
  with 
  respect 
  to 
  

   air 
  as 
  standard 
  led 
  to 
  values 
  which 
  are 
  generally 
  accepted 
  as 
  

   accurate 
  and 
  reliable. 
  It 
  is 
  thus 
  evident 
  that 
  a 
  redetermination 
  

   of 
  the 
  absolute 
  value 
  of 
  k 
  for 
  air 
  is 
  highly 
  desirable. 
  

  

  The 
  solution 
  of 
  this 
  experimental 
  problem 
  has 
  been 
  recently 
  

   undertaken 
  by 
  E. 
  0. 
  Hercus 
  and 
  T. 
  H. 
  Laby. 
  In 
  order 
  to 
  elimi- 
  

   nate 
  convection 
  currents, 
  these 
  investigators 
  took 
  advantage 
  of 
  

   the 
  method 
  suggested 
  by 
  Poynting, 
  which 
  consists 
  in 
  using 
  a 
  

   horizontal 
  layer 
  of 
  gas 
  the 
  upper 
  plane 
  boundary 
  of 
  which 
  is 
  

   maintained 
  at 
  a 
  higher 
  temperature 
  than 
  the 
  lower 
  surface. 
  The 
  

   fundamental 
  parts 
  of 
  the 
  apparatus 
  may 
  be 
  briefly 
  described 
  as* 
  

   follows 
  : 
  There 
  were 
  three 
  parallel 
  circular 
  discs 
  of 
  copper 
  each 
  

   1 
  cm. 
  thick. 
  The 
  middle 
  disc, 
  which 
  was 
  surrounded 
  by 
  a 
  guard 
  

   ring, 
  had 
  a 
  diameter 
  of 
  23- 
  148 
  cm. 
  The 
  outside 
  diameter 
  of 
  this 
  

   ring 
  had 
  the 
  same 
  value, 
  30 
  cms., 
  as 
  that 
  of 
  the 
  bottom 
  and 
  top 
  

   discs. 
  The 
  radial 
  width 
  of 
  the 
  air 
  gap 
  between 
  the 
  central 
  disc 
  

   and 
  the 
  guard 
  ring 
  was 
  0-605 
  cm., 
  half 
  of 
  this 
  number 
  being 
  used 
  

   in 
  computing 
  the 
  effective 
  area- 
  of 
  the 
  central 
  disc, 
  that 
  is, 
  443-1 
  

   cm 
  2 
  . 
  The 
  average 
  distance 
  between 
  the 
  central 
  and 
  bottom 
  discs 
  

   was 
  0-6280 
  cm. 
  The 
  air 
  under 
  investigation 
  occupied 
  the 
  cylin- 
  

   drical 
  space 
  between 
  the 
  effective 
  central 
  disc 
  and 
  the 
  bottom 
  

   disc. 
  The 
  central 
  disc 
  and 
  the 
  guard 
  ring 
  were 
  maintained 
  at 
  the 
  

   same 
  temperature 
  (about 
  35° 
  C). 
  The 
  bottom 
  disc 
  was 
  kept 
  at 
  

   a 
  constant 
  temperature, 
  lower 
  than 
  that 
  of 
  the 
  central 
  disc, 
  by 
  

   the 
  circulation 
  of 
  water 
  through 
  a 
  spiral 
  channel 
  in 
  the 
  zinc 
  base 
  

   of 
  the 
  apparatus. 
  Good 
  thermal 
  contact 
  between 
  the 
  under 
  sur- 
  

   face 
  of 
  the 
  bottom 
  disc 
  and 
  the 
  top 
  of 
  the 
  zinc 
  cooler 
  was 
  effected 
  

   by 
  an 
  intermediate 
  layer 
  of 
  Wood's 
  metal. 
  Heat 
  passed 
  through 
  

  

  