﻿Media 
  Subjected 
  to 
  Torsion, 
  99 
  

  

  To 
  illustrate 
  the 
  character 
  of 
  these 
  observations 
  the 
  readings 
  

   with 
  tube 
  14 
  are 
  given 
  in 
  detail 
  in 
  Table 
  lY 
  and 
  graphically 
  

   in 
  fig. 
  3. 
  The 
  abscissas 
  are 
  azimuths, 
  that 
  is, 
  the 
  angle 
  between 
  

   the 
  initial 
  plane 
  of 
  polarization 
  and 
  an 
  arbitrary 
  radius 
  of 
  the 
  

   jelly 
  cylinder 
  and 
  the 
  ordinates 
  are 
  rotations 
  of 
  the 
  plane 
  of 
  

   polarization. 
  The 
  variation 
  in 
  different 
  azimuths 
  with 
  these 
  

   jelly 
  tubes 
  is 
  very 
  much 
  less 
  than 
  with 
  the 
  glass 
  cylinder. 
  

  

  The 
  effect 
  of 
  torsion 
  in 
  directions 
  perpendicular 
  to 
  the 
  axis 
  

   was 
  studied 
  in 
  the 
  glass 
  bar 
  made 
  by 
  Duboscq 
  and 
  in 
  a 
  jelly 
  

   tube 
  with 
  glass 
  windows 
  on 
  the 
  sides. 
  ]S^o 
  effect 
  was 
  apparent 
  

   if 
  the 
  glass 
  bar 
  was 
  exactly 
  perpendicular 
  to 
  the 
  ray 
  of 
  light 
  

   and 
  with 
  the 
  jelly 
  tube 
  the 
  double 
  refraction 
  was 
  too 
  great 
  for 
  

   satisfactory 
  observations. 
  

  

  These 
  experiments 
  demonstrate 
  that 
  torsion 
  produces 
  rota- 
  

   tory 
  polarization 
  in 
  the 
  opposite 
  direction 
  to 
  the 
  twist. 
  The 
  

   curves 
  of 
  fig. 
  2 
  indicate 
  that 
  the 
  rotation 
  of 
  the 
  plane 
  of 
  polari- 
  

   zation 
  is 
  a 
  function 
  of 
  the 
  twist 
  of 
  deo^ree 
  hio-her 
  than 
  the 
  first. 
  

   It 
  is 
  difficult 
  to 
  find 
  any 
  satisfactory 
  explanation 
  for 
  these 
  

   phenomena. 
  Crystalline 
  plates 
  helically 
  arranged 
  rotate 
  the 
  

   plane 
  of 
  polarization, 
  but 
  these 
  phenomena 
  do 
  not 
  seem 
  analo- 
  

   gous. 
  In 
  the 
  jelly 
  tubes 
  there 
  could 
  not 
  have 
  been 
  originally 
  

   any 
  uniform 
  mechanical 
  difference 
  of 
  structure 
  in 
  directions 
  

   at 
  right 
  angles 
  to 
  the 
  axis 
  of 
  the 
  tube 
  and 
  no 
  such 
  difference 
  

   could 
  be 
  produced 
  by 
  the 
  torsion 
  in 
  the 
  case 
  of 
  glass. 
  Even 
  if 
  

   each 
  cross 
  section 
  was 
  analogous 
  to 
  a 
  crystalline 
  plate 
  the 
  

   helical 
  change 
  of 
  position 
  of 
  the 
  axes 
  would 
  be 
  continuous 
  and 
  

   not 
  by 
  steps, 
  as 
  is 
  required 
  in 
  Mallard's 
  theory 
  of 
  this 
  rotatory 
  

   polarization.* 
  ]^or 
  can 
  the 
  phenomena 
  be 
  explained 
  on 
  the 
  

   basis 
  of 
  a 
  structure 
  of 
  helical 
  fibres, 
  for 
  we 
  cannot 
  imagine 
  

   jelly 
  possessing 
  a 
  fibrous 
  structure, 
  and 
  if 
  we 
  could 
  the 
  twisted 
  

   fibers 
  would 
  be 
  straight 
  and 
  not 
  helical 
  in 
  distances 
  comparable 
  

   with 
  the 
  wave-length 
  of 
  light. 
  

  

  Excessively 
  large 
  twists 
  were 
  used 
  in 
  the 
  experiments 
  with 
  

   jelly. 
  Wertheim 
  demonstrated 
  experimentally 
  in 
  the 
  memoir 
  

   already 
  referred 
  to, 
  that 
  under 
  large 
  torsion 
  a 
  cylinder 
  is 
  short- 
  

   ened, 
  the 
  contraction 
  being 
  proportional 
  to 
  the 
  distance 
  from 
  

   the 
  axis. 
  This 
  longitudinal 
  pressure 
  with 
  the 
  torsion 
  stresses 
  

   would 
  give 
  a 
  component 
  stress 
  in 
  normal 
  sections, 
  but 
  it 
  would 
  

   seem 
  as 
  if 
  the 
  conditions 
  would 
  be 
  the 
  same 
  for 
  either 
  clock- 
  

   wise 
  or 
  counter-clockwise 
  circular 
  vibrations 
  in 
  these 
  normal 
  

   sections. 
  

  

  This 
  investigation 
  substantiated 
  the 
  statements 
  of 
  Yerdet 
  

   and 
  Wertheim 
  that 
  the 
  difficulties 
  encountered 
  in 
  the 
  choice 
  

   of 
  a 
  proper 
  material 
  for 
  experiment 
  are 
  very 
  great. 
  In 
  jelly 
  

   supported 
  in 
  rubber 
  tubes 
  a 
  satisfactory 
  combination 
  was 
  found, 
  

  

  *Ann. 
  des 
  Mines, 
  Til, 
  xix, 
  1891. 
  

  

  