﻿754 
  Mr. 
  W. 
  A. 
  Jenkins 
  : 
  New 
  Method 
  of 
  Determining 
  the 
  

  

  were 
  satisfactory 
  ; 
  but 
  such 
  a 
  helix 
  is 
  subject 
  to 
  the 
  following 
  

   faults 
  :— 
  

  

  (a) 
  Leakage 
  of 
  current 
  through 
  the 
  covering 
  of 
  the 
  wire 
  

  

  owing 
  to 
  imperfection 
  of 
  insulation 
  ; 
  

  

  (b) 
  Irregularity 
  in 
  the 
  winding 
  causing 
  uncertainty 
  in 
  the 
  

  

  value 
  of 
  " 
  nf 
  

   Both 
  these 
  errors 
  would 
  be 
  eliminated 
  by 
  the 
  following 
  

   scheme 
  : 
  — 
  Coat 
  the 
  tube 
  with 
  an 
  insulating 
  material 
  and 
  

   turn 
  down 
  to 
  a 
  perfect 
  cylinder. 
  Silver 
  the 
  surface 
  and 
  

   deposit 
  copper 
  electrolytically 
  on 
  the 
  silvered 
  surface. 
  Now 
  

   turn 
  a 
  screw-thread 
  on 
  the 
  surface 
  so 
  as 
  to 
  give 
  a 
  helix 
  of 
  

   copper 
  wire. 
  If 
  a 
  suitable 
  tube 
  could 
  be 
  obtained 
  it 
  would 
  

   be 
  better 
  to 
  use 
  one 
  made 
  of 
  insulating 
  material 
  instead 
  of 
  

   coating 
  the 
  copper 
  one. 
  Owing 
  to 
  its 
  being 
  fragile, 
  the 
  use 
  

   of 
  an 
  ebonite 
  tube 
  is 
  not 
  to 
  be 
  recommended, 
  and 
  as 
  a 
  copper 
  

   tube 
  is 
  (a) 
  easily 
  obtained 
  and 
  turned 
  to 
  a 
  perfect 
  cylinder, 
  

   (b) 
  strong, 
  (c) 
  practically 
  non-magnetic, 
  it 
  was 
  decided 
  to 
  

   use 
  such 
  a 
  tube 
  and 
  try 
  the 
  above 
  suggested 
  method 
  of 
  

   coating. 
  Sulphur 
  was 
  first 
  tried 
  as 
  the 
  insulating 
  material, 
  

   but 
  preliminary 
  experiments 
  with 
  a 
  tube 
  30 
  cm. 
  by 
  2 
  cm. 
  

   showed 
  that 
  it 
  was 
  unsuitable 
  as 
  it 
  w 
  T 
  as 
  (a) 
  extremely 
  

   difficult 
  to 
  apply 
  to 
  the 
  copper 
  tube, 
  (b) 
  difficult 
  to 
  turn 
  a 
  

   good 
  screw-thread 
  on 
  the 
  tube 
  thus 
  coated. 
  Zylonite, 
  a 
  

   white 
  viscous 
  fluid, 
  almost 
  solid, 
  which 
  dries 
  to 
  a 
  hard 
  white 
  

   insulating 
  substance, 
  was 
  next 
  tried. 
  Preliminary 
  experi- 
  

   ments 
  with 
  a 
  small 
  tube 
  were 
  entirely 
  satisfactory, 
  excepting 
  

   for 
  the 
  fact 
  that 
  air-bubbles 
  were 
  continually 
  appearing 
  even 
  

   when 
  the 
  zylonite 
  appeared 
  to 
  be 
  set. 
  Otherwise 
  the 
  zylonite 
  

   satisfied 
  all 
  the 
  conditions, 
  but 
  this 
  defect 
  caused 
  its 
  abandon- 
  

   ment. 
  At 
  the 
  time 
  of 
  writing 
  the 
  zylonite 
  is 
  perfectly 
  hard 
  

   and 
  dry 
  and 
  would 
  probably 
  prove 
  suitable. 
  Experiments 
  

   with 
  ivoride 
  and 
  hard 
  paraffin-wax 
  showed 
  their 
  unsuitability, 
  

   and, 
  consequent 
  upon 
  these 
  failures, 
  it 
  was 
  decided 
  to 
  send 
  

   the 
  tube 
  away 
  and 
  have 
  a 
  layer 
  of 
  ebonite 
  placed 
  upon 
  it. 
  

   Tbis 
  was 
  done 
  and 
  eventually 
  proved 
  satisfactory. 
  Two 
  

   layers 
  of 
  ebonite 
  were 
  placed 
  upon 
  the 
  tube, 
  which 
  was 
  then 
  

   turned 
  down 
  and 
  a 
  screw-thread 
  turned 
  on 
  it. 
  An 
  experi- 
  

   ment 
  was 
  then 
  performed 
  to 
  test 
  the 
  accuracy 
  of 
  the 
  cutting 
  

   of 
  the 
  thread. 
  The 
  tube 
  was 
  fixed 
  up 
  so 
  that 
  it 
  could 
  be 
  

   moved 
  any 
  desired 
  distance 
  in 
  a 
  direction 
  parallel 
  to 
  its 
  

   length. 
  Two 
  microscopes 
  rigidly 
  fixed 
  at 
  a 
  certain 
  distance 
  

   apart 
  were 
  then 
  focussed 
  on 
  the 
  screw-threads, 
  which 
  were 
  

   brightly 
  illuminated. 
  The 
  intersections 
  of 
  the 
  cross-threads 
  

   of 
  the 
  microscope 
  were 
  focussed 
  on 
  the 
  bottoms 
  of 
  two 
  of 
  the 
  

   " 
  Y 
  " 
  shaped 
  screw-threads. 
  The 
  number 
  of 
  screw-threads 
  

   between 
  the 
  cross-wires 
  was 
  counted 
  and 
  found 
  to 
  be 
  41. 
  

  

  