﻿Field 
  upon 
  the 
  Electric 
  Discharge 
  through 
  Gases. 
  255 
  

  

  increasing 
  the 
  magnetic 
  field 
  and 
  by 
  diminishing 
  the 
  pressure 
  

   of 
  the 
  gas. 
  Our 
  apparatus 
  was 
  such 
  that 
  the 
  field 
  aud 
  

   pressure 
  could 
  be 
  varied 
  from 
  inferior 
  limits 
  not 
  sufficient 
  to 
  

   produce 
  the 
  rays, 
  to 
  superior 
  limits 
  which 
  destroyed 
  the 
  rays 
  

   in 
  any 
  gas 
  used. 
  By 
  varying 
  these 
  two 
  factors, 
  the 
  virtual 
  

   anode, 
  when 
  it 
  could 
  be 
  found 
  at 
  all, 
  could 
  be 
  made 
  to 
  shift 
  

   its 
  position 
  to 
  practically 
  any 
  part 
  of 
  the 
  column. 
  If 
  the 
  

   virtual 
  anode, 
  accepting 
  Professor 
  Rights 
  definition 
  of 
  it, 
  

   existed 
  near 
  the 
  extremity 
  of 
  the 
  magnetic 
  rays 
  in 
  the 
  various 
  

   gases, 
  why 
  could 
  it 
  not 
  also 
  be 
  shifted, 
  by 
  varying 
  the 
  field, 
  

   to 
  a 
  position 
  where 
  it 
  could 
  be 
  observed 
  ? 
  This 
  objection 
  to 
  

   his 
  hypothesis 
  still 
  seems 
  to 
  stand. 
  

  

  But 
  Professor 
  Kighi, 
  whether 
  the 
  objection 
  is 
  valid 
  or 
  not, 
  

   relegates 
  it 
  to 
  a 
  minor 
  position 
  when 
  he 
  writes 
  : 
  " 
  There 
  is 
  a 
  

   research 
  which 
  ought 
  to 
  have 
  been 
  initiated 
  before 
  formulating 
  

   doubts 
  or 
  objections 
  ; 
  that 
  of 
  examining 
  if 
  the 
  marked 
  

   intermittency 
  produced 
  in 
  the 
  discharge 
  by 
  the 
  magnetic 
  

   field 
  when 
  the 
  tube 
  contains 
  air 
  (which 
  intermittency 
  is 
  

   strictly 
  connected 
  with 
  the 
  formation 
  of 
  the 
  virtual 
  anode), 
  is 
  

   produced 
  or 
  not 
  with 
  those 
  gases 
  in 
  which 
  the 
  induced 
  

   column 
  was 
  not 
  observed 
  by 
  the 
  authors. 
  Evidently 
  if, 
  for 
  

   any 
  reason 
  whatever, 
  known 
  or 
  unknown, 
  the 
  periodicity 
  did 
  

   not 
  exist, 
  the 
  want 
  of 
  the 
  induced 
  column 
  would 
  be 
  the 
  

   natural 
  consequence. 
  " 
  It 
  must 
  be 
  confessed 
  that 
  the 
  writers 
  

   failed 
  to 
  understand 
  that 
  the 
  intermittency 
  was 
  so 
  connected 
  

   with 
  the 
  formation 
  of 
  the 
  virtual 
  anode, 
  or 
  was 
  so 
  vital 
  to 
  the 
  

   theory 
  of 
  magnetically 
  formed 
  doublets 
  and 
  rays. 
  We 
  have, 
  

   naturally, 
  investigated 
  this 
  subject 
  carefully, 
  and 
  shall 
  show 
  

   in 
  this 
  paper 
  that, 
  as 
  the 
  intermittency 
  is 
  not 
  dependent 
  on 
  a 
  

   magnetic 
  field, 
  it 
  cannot 
  be 
  thus 
  vitally 
  connected 
  with 
  the 
  

   formation 
  of 
  a 
  virtual 
  anode, 
  or 
  with 
  Professor 
  Righi's 
  

   theory. 
  

  

  Our 
  excuse 
  for 
  giving 
  this 
  rather 
  long 
  statement 
  and 
  

   discussion 
  of 
  the 
  problem 
  is, 
  that 
  the 
  results 
  about 
  to 
  be 
  

   described 
  have 
  forced 
  us 
  to 
  abandon 
  Professor 
  Righi's 
  

   theory. 
  Our 
  new 
  modification 
  of 
  his 
  apparatus 
  has 
  enabled 
  

   us 
  to 
  obtain 
  results 
  so 
  decisive, 
  that 
  we 
  believe 
  we 
  can 
  now 
  

   offer 
  a 
  fairly 
  accurate 
  and 
  simple 
  explanation 
  of 
  both 
  the 
  

   primary 
  and 
  secondary 
  effects 
  w 
  r 
  hich 
  occur 
  when 
  an 
  electric 
  

   discharge 
  is 
  influenced 
  by 
  a 
  magnetic 
  field. 
  

  

  The 
  principal 
  thing 
  which 
  has 
  hampered 
  the 
  work 
  in 
  the 
  

   past 
  was 
  the 
  use 
  of 
  an 
  induction-coil 
  or 
  an 
  electrostatic 
  

   machine 
  as 
  a 
  source 
  of 
  potential. 
  The 
  potential 
  is 
  then 
  not 
  

   sufficiently 
  uniform, 
  and 
  the 
  rays 
  are 
  rather 
  unstable. 
  The 
  

   potential 
  is 
  too 
  high, 
  it 
  cannot 
  be 
  measured 
  accurately, 
  and 
  it, 
  

   produces 
  too 
  little 
  current. 
  All 
  these 
  defects 
  are 
  avoided 
  by 
  

  

  