﻿[ 
  598 
  ] 
  

  

  XLIX. 
  The 
  Absolute 
  Thermal 
  Conductivity 
  of 
  Glass. 
  By 
  

   I. 
  Williams, 
  M.Sc, 
  Lecturer 
  in 
  Physics 
  in 
  the 
  University 
  

   of 
  Bristol*. 
  

  

  § 
  1. 
  npHE 
  absolute 
  thermal 
  conductivity 
  of 
  glass 
  has 
  been 
  

   L 
  determined 
  by 
  several 
  independent 
  methods, 
  of 
  

   which 
  the 
  best 
  known 
  are 
  probably 
  the 
  following 
  : 
  — 
  

  

  (i.) 
  The 
  "divided 
  bar" 
  method. 
  

  

  This 
  method 
  was 
  suggested 
  by 
  Lodge 
  and 
  used 
  by 
  Leesf. 
  

   A 
  metal 
  rod 
  is 
  cut 
  in 
  two 
  and 
  the 
  distribution 
  of 
  temperature 
  

   down 
  the 
  whole 
  bar 
  is 
  studied 
  (a) 
  when 
  the 
  cut 
  ends 
  are 
  in 
  

   intimate 
  contact, 
  (b) 
  when 
  a 
  plate 
  of 
  badly 
  conducting 
  sub- 
  

   stance 
  is 
  interposed 
  between 
  them. 
  

  

  (ii.) 
  Voigt's 
  Method. 
  

  

  In 
  this 
  method 
  one 
  surface 
  of 
  the 
  plate 
  is 
  maintained 
  at 
  

   constant 
  temperature 
  by 
  a 
  current 
  of 
  cold 
  water 
  flowing 
  over 
  

   it. 
  The 
  other 
  surface 
  is 
  exposed 
  to 
  hot 
  water 
  in 
  a 
  calorimeter. 
  

   The 
  hot 
  water 
  is 
  thoroughly 
  stirred, 
  and 
  from 
  its 
  rate 
  of 
  

   cooling 
  the 
  amount 
  of 
  heat 
  transmitted 
  through 
  the 
  plate 
  is 
  

   determined. 
  This 
  method 
  has 
  been 
  used 
  by 
  VenskeJ 
  and 
  

   by 
  Focke§. 
  

  

  (iii.) 
  Christiansen's 
  Method. 
  

  

  Paalhorn||, 
  using 
  the 
  well-known 
  three-plate 
  arrangement 
  

   due 
  to 
  Christiansen, 
  has 
  determined 
  the 
  conductivity 
  of 
  glass 
  

   in 
  terms 
  of 
  that 
  of 
  air. 
  

  

  (iv.) 
  Lees's 
  Disk 
  Method. 
  

  

  Leeslf 
  has 
  determined 
  the 
  conductivity 
  of 
  glass 
  and 
  other 
  

   bad 
  conductors 
  by 
  means 
  of 
  a 
  disk 
  method 
  in 
  which 
  the 
  

   heat 
  energy 
  transmitted 
  is 
  measured 
  electrically. 
  

  

  § 
  2. 
  The 
  disk 
  methods 
  just 
  referred 
  to 
  are 
  very 
  convenient 
  

   for 
  comparing 
  the 
  thermal 
  conductivities 
  of 
  two 
  substances, 
  

   particularly 
  when 
  the 
  substance 
  is 
  available 
  in 
  small 
  quantities 
  

   only. 
  In 
  all 
  these 
  methods, 
  with 
  the 
  exception 
  of 
  the 
  second, 
  

   it 
  is 
  usual 
  to 
  secure 
  good 
  thermal 
  contact 
  between 
  the 
  plate 
  

   of 
  copper 
  and 
  the 
  plate 
  of 
  substance 
  being 
  tested 
  by 
  inter- 
  

   posing 
  a 
  thin 
  layer 
  of 
  glycerine 
  or 
  mercury 
  between 
  the 
  

  

  * 
  Communicated 
  by 
  the 
  Author. 
  

  

  t 
  Lees, 
  Phil. 
  Trans, 
  vol. 
  clxxxiii. 
  p. 
  481 
  (1892). 
  

  

  X 
  Venske, 
  Gottingen 
  Nachr. 
  p. 
  121 
  (1891). 
  

  

  § 
  Focke, 
  Wied. 
  Annalen, 
  vol. 
  lxvii. 
  p. 
  132 
  (1898). 
  

  

  j| 
  Paalhorn, 
  Dissertation, 
  Jena, 
  1894. 
  

  

  % 
  Lees, 
  Phil. 
  Trans, 
  vol. 
  cxci. 
  p. 
  399 
  (1898). 
  

  

  