﻿44 
  Prof. 
  Chattock 
  and 
  Mr. 
  Tyndall 
  on 
  Changes 
  

  

  shows 
  a 
  tendency 
  to 
  peak 
  formation 
  at 
  the 
  lowest 
  percentages, 
  

   which 
  vanishes 
  further 
  on 
  as 
  though 
  it 
  were 
  overpowered 
  

   by 
  some 
  opposing 
  tendency. 
  

  

  We 
  have 
  also 
  plotted 
  (crosses) 
  the 
  curves 
  for 
  the 
  two 
  

   nitrogen 
  sets, 
  the 
  first 
  being 
  dotted 
  in 
  for 
  clearness. 
  Table 
  3 
  

   gives 
  for 
  the 
  first 
  series 
  the 
  amount 
  of 
  oxygen 
  present 
  in 
  the 
  

   nitrogen 
  up 
  to 
  2' 
  22 
  per 
  cent, 
  of 
  the 
  latter, 
  and 
  beyond 
  that 
  

   we 
  have 
  assumed 
  it 
  to 
  bear 
  a 
  constant 
  proportion 
  to 
  the 
  

   nitrogen 
  of 
  1 
  : 
  600. 
  For 
  the 
  second 
  series 
  Table 
  4 
  shows 
  the 
  

   proportion 
  to 
  be 
  constant 
  at 
  1 
  : 
  770 
  about. 
  

  

  The 
  first 
  nitrogen 
  curve 
  starts 
  with 
  what 
  looks 
  like 
  an 
  

   intention 
  of 
  coinciding 
  with 
  the 
  first 
  oxygen 
  curve. 
  After 
  

   two 
  high 
  readings, 
  however, 
  it 
  falls 
  suddenly 
  to 
  a 
  low 
  level 
  

   and 
  stays 
  there 
  ; 
  and 
  it 
  is 
  after 
  this 
  fall 
  that 
  the 
  peak 
  does 
  

   not 
  again 
  reach 
  a 
  high 
  value 
  in 
  any 
  subsequent 
  curve 
  until 
  

   the 
  heating 
  of 
  the 
  point. 
  Some 
  permanent 
  change 
  in 
  the 
  

   conditions 
  seems 
  to 
  have 
  occurred 
  — 
  possibly 
  connected 
  with 
  

   the 
  large 
  amounts 
  of 
  nitrogen 
  now 
  present 
  (10 
  per 
  cent, 
  to 
  

   14 
  per 
  cent.). 
  

  

  But 
  the 
  beginnings 
  of 
  the 
  curves 
  do 
  not 
  seem 
  to 
  have 
  been 
  

   much 
  affected. 
  All 
  the 
  five 
  curves 
  taken 
  before 
  heating 
  

   the 
  point 
  start 
  off 
  along 
  practically 
  the 
  same 
  line 
  ; 
  so 
  that 
  

   for 
  the 
  lowest 
  percentages 
  — 
  up 
  to 
  about 
  0'003 
  per 
  cent. 
  — 
  

   the 
  condition 
  of 
  the 
  point 
  had 
  little 
  influence 
  on 
  the 
  values 
  

   of. 
  p. 
  

  

  The 
  ends 
  of 
  the 
  nitrogen 
  curves 
  are 
  low. 
  They 
  are 
  also 
  

   steep, 
  and 
  do 
  not 
  seem 
  inclined 
  to 
  follow 
  the 
  directions 
  of 
  

   the 
  other 
  low 
  curves. 
  These 
  facts 
  may 
  perhaps 
  also 
  be 
  

   attributable 
  to 
  the 
  quantity 
  of 
  nitrogen 
  present. 
  It 
  seems 
  

   desirable 
  to 
  repeat 
  and 
  extend 
  these 
  observations 
  with 
  much 
  

   purer 
  nitrogen. 
  

  

  After 
  the 
  point 
  was 
  heated 
  to 
  bright 
  redness, 
  at 
  which 
  

   temperature 
  it 
  was 
  kept 
  for 
  several 
  minutes, 
  the 
  contractions 
  

   went 
  back 
  to 
  their 
  old 
  values 
  for 
  oxygen 
  above 
  about 
  

   0*01 
  per 
  cent., 
  but 
  below 
  that 
  percentage 
  they 
  were 
  com- 
  

   pletely 
  changed. 
  

  

  Curves 
  VIII. 
  show 
  clearly 
  the 
  sharpening 
  and 
  raising 
  of 
  

   the 
  peak 
  that 
  has 
  occurred. 
  The 
  curves 
  for 
  positive 
  discharge 
  

   too, 
  which 
  had 
  so 
  far 
  been 
  practically 
  level 
  up 
  to 
  percentages 
  

   well 
  above 
  0*06, 
  now 
  showed 
  a 
  definite 
  tendency 
  to 
  form 
  a 
  

   minute 
  peak 
  at 
  about 
  0*0005 
  per 
  cent. 
  

  

  Heating 
  Effect 
  of 
  the 
  Discharge. 
  

   It 
  has 
  been 
  shown 
  * 
  that 
  when 
  gas 
  is 
  blown 
  against 
  a 
  

   plate 
  by 
  discharge 
  from 
  a 
  point 
  opposite, 
  the 
  wind 
  produced 
  

   * 
  Chattock, 
  Phil. 
  Mag. 
  vol. 
  v. 
  p. 
  48 
  (1899) 
  ; 
  vol. 
  vi. 
  p. 
  1 
  (1901). 
  

  

  