﻿468 
  Mr. 
  J. 
  Perry 
  on 
  the 
  [June 
  10, 
  

  

  expresses 
  the 
  temperature 
  at 
  time 
  t, 
  if 
  when 
  t=0 
  the 
  temperature 
  is 
  

   expressed 
  by 
  

  

  1 
  E« 
  ar 
  

  

  2 
  k 
  cr 
  

  

  Taking, 
  for 
  instance, 
  the 
  copper 
  globe 
  of 
  4 
  centimetres 
  diameter, 
  

   we 
  have 
  

  

  L-^-L 
  oc 
  ; 
  (3) 
  

  

  and 
  we 
  see 
  that 
  in 
  the 
  Glasgow 
  experiments 
  the 
  difference 
  of 
  tempera- 
  

   tures 
  between 
  surface 
  and 
  centre 
  was 
  just 
  jxrW 
  °^ 
  the 
  excess 
  G 
  f 
  either 
  

   above 
  the 
  temperature 
  of 
  the 
  surrounding 
  medium, 
  when 
  time 
  enough 
  

   had 
  elapsed 
  to 
  allow 
  the 
  first 
  term 
  of 
  Eourier's 
  series 
  to 
  be 
  the 
  predomi- 
  

   nating 
  one. 
  Before 
  that 
  time 
  the 
  difference 
  of 
  temperatures 
  must 
  have 
  

   been 
  less 
  than 
  of 
  either, 
  if 
  initially 
  the 
  temperature 
  was 
  uniform 
  

   from 
  surface 
  to 
  centre. 
  The 
  Fourier 
  analysis 
  of 
  the 
  transition 
  from 
  the 
  

   supposed 
  initial 
  uniform 
  distribution 
  to 
  the 
  state 
  represented 
  by 
  (3) 
  is 
  

   exceedingly 
  interesting, 
  but 
  unnecessary 
  for 
  the 
  settlement 
  of 
  the 
  present 
  

   question. 
  

  

  VII. 
  " 
  Preliminary 
  Results 
  of 
  an 
  Investigation 
  on 
  the 
  Electric 
  

   Conductivity 
  of 
  Glass 
  at 
  different 
  Temperatures." 
  By 
  John 
  

   Perry, 
  B.E., 
  Thomson 
  Experimental 
  Scholar 
  in 
  The 
  Natural 
  

   Philosophy 
  Laboratory 
  at 
  Glasgow*. 
  Communicated 
  by 
  Pro- 
  

   fessor 
  Sir 
  "William 
  Thomson, 
  F.R.S., 
  LL.D. 
  Received 
  

   April 
  8, 
  1875. 
  

  

  A 
  quadrant 
  electrometer 
  now 
  in 
  use 
  in 
  the 
  laboratory 
  seems 
  to 
  retain 
  

   its 
  whole 
  charge 
  from 
  day 
  to 
  day 
  ; 
  a 
  week's 
  loss 
  is 
  just 
  perceptible, 
  and 
  

   may 
  be 
  supplied 
  by 
  a 
  few 
  turns 
  of 
  the 
  replenisher. 
  In 
  a 
  guard-ring 
  

   electrometer 
  now 
  in 
  use 
  the 
  charge 
  is 
  almost 
  wholly 
  retained 
  from 
  week 
  

   to 
  week. 
  These 
  qualities 
  are 
  due 
  to 
  the 
  exceptionally 
  great 
  insulation- 
  

   resistance 
  of 
  the 
  glass 
  employed. 
  

  

  At 
  various 
  times 
  experiments 
  have 
  been 
  made 
  in 
  the 
  laboratory 
  at 
  

   Glasgow 
  to 
  determine 
  the 
  insulation-resistance 
  of 
  different 
  kinds 
  of 
  

   glass. 
  Of 
  the 
  specimens 
  hitherto 
  examined, 
  those 
  of 
  flint 
  glass 
  have 
  

   insulated 
  best 
  ; 
  and 
  it 
  is 
  hoped 
  that 
  experiments 
  on 
  flint 
  glass 
  now 
  being 
  

   proceeded 
  with 
  will 
  define 
  the 
  most 
  suitable 
  glass 
  for 
  use 
  in 
  electrometers 
  

   and 
  other 
  electrical 
  instruments. 
  

  

  The 
  method 
  of 
  investigation 
  is 
  essentially 
  the 
  same 
  as 
  that 
  which 
  was 
  

   adopted 
  by 
  Sir 
  William 
  Thomson 
  some 
  years 
  ago. 
  B 
  D 
  A 
  (fig. 
  1) 
  is 
  a 
  

  

  * 
  Now 
  Professor 
  of 
  Engineering 
  in 
  the 
  Imperial 
  College 
  of 
  Engineering, 
  Jeddo, 
  

   Japan. 
  

  

  