﻿Ions 
  from 
  Hot 
  Salts. 
  453 
  

  

  have 
  been 
  examined 
  and 
  recorded. 
  The 
  salts 
  which 
  have 
  

   been 
  tested 
  are 
  those 
  which 
  are 
  denoted 
  by 
  the 
  following 
  

   chemical 
  formulae 
  : 
  — 
  Znl 
  2 
  , 
  Cdl 
  2 
  , 
  Cal 
  2 
  , 
  CaBr 
  2 
  , 
  CaF 
  2 
  , 
  Srl 
  2 
  , 
  

   Bal 
  2 
  , 
  Fe 
  2 
  Cl 
  6 
  , 
  and 
  MnCl 
  3 
  . 
  The 
  apparatus 
  used 
  in 
  measuring 
  

   ejm 
  is 
  the 
  same 
  as 
  that 
  previously 
  described 
  ; 
  but 
  in 
  

   many 
  cases 
  the 
  ionization 
  was 
  neither 
  sufficiently 
  regular 
  

   ) 
  or 
  persistent 
  to 
  enable 
  the 
  ratio, 
  of 
  the 
  charge 
  passing 
  

   through 
  a 
  narrow 
  slit 
  to 
  the 
  total 
  charge 
  emitted, 
  to 
  be 
  

   measured 
  in 
  a 
  sufficiently 
  large 
  number 
  of 
  positions. 
  A 
  

   different 
  method 
  was 
  therefore 
  adopted. 
  The 
  electrode 
  

   behind 
  the 
  slit 
  and 
  one 
  of 
  the 
  plates 
  were 
  connected 
  to 
  one 
  

   pair 
  oE 
  quadrants 
  of 
  the 
  electrometer 
  and 
  the 
  other 
  plate 
  to 
  

   the 
  other 
  quadrants. 
  All 
  four 
  quadrants 
  of 
  the 
  electrometer 
  

   were 
  insulated 
  and 
  the 
  position 
  of 
  the 
  slit 
  was 
  determined 
  

   for 
  which 
  the 
  electrometer 
  needle 
  did 
  not 
  deflect 
  when 
  the 
  

   quadr.mts 
  were 
  simultaneously 
  connected 
  to 
  the 
  respective 
  

   plates. 
  Under 
  these 
  circumstances 
  both 
  qua 
  Irants 
  are 
  

   receiving 
  equal 
  or 
  sufficiently 
  nearly 
  equal 
  charges 
  (the 
  

   capacity 
  not 
  being 
  quite 
  the 
  same 
  for 
  the 
  two 
  sides 
  of 
  the 
  

   system), 
  and 
  the 
  position 
  corresponds 
  to 
  the 
  top 
  of 
  the 
  humps 
  

   in 
  the 
  curves 
  previously 
  shown. 
  This 
  position 
  was 
  determined 
  

   for 
  both 
  directions 
  of 
  the 
  magnetic 
  field. 
  The 
  resulting 
  dis- 
  

   placement 
  gives 
  the 
  value 
  of 
  ejm 
  when 
  substituted 
  in 
  the 
  

   formula 
  already 
  deduced. 
  This 
  method 
  has 
  the 
  great 
  

   advantage 
  that 
  a 
  determination 
  of 
  ejm 
  only 
  requires 
  a 
  minute 
  

   or 
  two, 
  so 
  that 
  the 
  variations 
  of 
  this 
  quantity 
  can 
  be 
  followed 
  

   even 
  when 
  it 
  is 
  changing 
  rather 
  rapidly. 
  On 
  the 
  other 
  hand, 
  

   it 
  only 
  gives 
  a 
  kind 
  of 
  average 
  value 
  if 
  more 
  than 
  one 
  kind 
  

   of 
  ion 
  is 
  present 
  ; 
  so 
  that 
  it 
  does 
  not 
  afford 
  so 
  complete 
  an 
  

   analysis 
  as 
  the 
  former 
  method, 
  with 
  which 
  one 
  could 
  tell 
  

   from 
  the 
  shape 
  of 
  the 
  curves 
  whether 
  a 
  mixture 
  of 
  ions 
  or 
  a 
  

   single 
  species 
  was 
  being 
  encountered. 
  I 
  shall 
  refer 
  to 
  this 
  

   as 
  the 
  balance 
  method 
  and 
  to 
  the 
  former 
  as 
  the 
  slit 
  method. 
  

   Whenever 
  the 
  balance 
  method 
  gave 
  results 
  which 
  seemed 
  to 
  

   be 
  of 
  a 
  dubious 
  character 
  they 
  were 
  checked 
  up 
  by 
  measure- 
  

   ments 
  in 
  which 
  the 
  slit 
  method 
  was 
  used 
  if 
  it 
  was 
  feasible 
  

   to 
  do 
  so. 
  I 
  cannot 
  recollect 
  any 
  case 
  in 
  which 
  the 
  two 
  

   methods 
  disagreed 
  with 
  one 
  another 
  ; 
  and 
  in 
  the 
  case 
  of 
  sub- 
  

   stances 
  like 
  potassium 
  sulphate, 
  which 
  give 
  a 
  very 
  definite 
  

   and 
  persistent 
  emission, 
  the 
  two 
  methods 
  give 
  results 
  which 
  

   agree 
  absolutely, 
  so 
  far 
  as 
  one 
  can 
  tell. 
  In 
  all 
  cases 
  the 
  salts 
  

   were 
  heated 
  on 
  a 
  flat 
  strip 
  of 
  platinum 
  which 
  had 
  previously 
  

   been 
  glowed 
  out 
  until 
  it 
  had 
  lost 
  its 
  power 
  of 
  emitting 
  ions 
  

   at 
  the 
  temperature 
  of 
  the 
  experiments. 
  The 
  investigation 
  

   falls 
  rather 
  naturally 
  into 
  two 
  divisions 
  according 
  to 
  the 
  

   si<:n 
  of 
  the 
  ions 
  dealt 
  with. 
  

  

  