﻿2f€dia 
  Subjected 
  io 
  I'orsion. 
  91 
  

  

  amonnt 
  in 
  different 
  azimuths, 
  there 
  seemed 
  to 
  be 
  an 
  unques- 
  

   tionable 
  rotation 
  of 
  the 
  plane 
  of 
  polarization 
  in 
  the 
  opposite 
  

   direction 
  to 
  the 
  twist, 
  for 
  rajs 
  parallel 
  or 
  nearly 
  parallel 
  to 
  the 
  

   axis 
  of 
  torsion. 
  Before 
  any 
  careful 
  quantitative 
  measurements 
  

   had 
  been 
  made 
  the 
  bar 
  became 
  suddenly 
  very 
  doubly-refract- 
  

   ing, 
  so 
  that 
  on 
  twisting, 
  the 
  only 
  optical 
  change 
  was 
  a 
  clearing 
  

   up 
  of 
  the 
  center 
  for 
  all 
  azimuths, 
  suggesting 
  the 
  phenomena 
  

   observed 
  by 
  Wertheim. 
  Soon 
  after 
  it 
  broke. 
  As 
  this 
  bar 
  

   originally 
  showed 
  no 
  dark 
  cross, 
  more 
  promising 
  results 
  could 
  

   have 
  been 
  expected 
  than 
  Wertheim 
  obtained 
  with 
  his 
  glass 
  

   bars 
  and 
  the 
  observations 
  were 
  certainly 
  very 
  suggestive 
  of 
  

   rotatory 
  polarization, 
  if 
  not 
  decisive. 
  

  

  Four 
  circular 
  glass 
  cylinders, 
  i 
  to 
  6*^°^ 
  long 
  and 
  7 
  to 
  S""'"" 
  

   diameter, 
  were 
  examined 
  between 
  crossed 
  polarizer 
  and 
  anal- 
  

   yzer 
  and 
  all 
  showed 
  the 
  dark 
  cross 
  and 
  rings. 
  One 
  even 
  

   showed 
  the 
  double 
  set 
  of 
  rings 
  and 
  dark 
  hyperbolge 
  of 
  biaxial 
  

   crystalline 
  plates. 
  The 
  best 
  of 
  these 
  cylinders 
  was 
  heated 
  

   twenty-four 
  hours 
  in 
  a 
  furnace 
  and 
  carefully 
  cooled. 
  Although 
  

   one 
  end 
  had 
  been 
  partially 
  fused, 
  the 
  cylinder 
  still 
  showed 
  a 
  

   cross, 
  but 
  fortunately 
  too 
  faint 
  to 
  seriously 
  disturb 
  observa- 
  

   tions. 
  Thin 
  wooden 
  cylinders 
  were 
  cemented 
  to 
  the 
  ends 
  and 
  

   could 
  be 
  grasped 
  in 
  any 
  position 
  by 
  clamps 
  attached 
  to 
  long 
  

   levers. 
  The 
  torsion 
  was 
  applied 
  by 
  weights 
  hung 
  on 
  these 
  

   levers 
  IS^"" 
  from 
  the 
  axis 
  of 
  the 
  glass 
  cylinder. 
  Observa- 
  

   tions 
  were 
  made 
  in 
  the 
  Scheibler 
  sacchari 
  meter 
  for 
  each 
  twist- 
  

   ing 
  moment 
  at 
  azimuth 
  intervals 
  of 
  1:5°. 
  In 
  each 
  azimuth 
  five 
  

   readings 
  were 
  taken. 
  

  

  In 
  most 
  of 
  my 
  experiments 
  considerable 
  double 
  refraction 
  

   appeared 
  on 
  twisting. 
  This 
  would 
  increase 
  in 
  some 
  azimuths 
  

   and 
  decrease 
  in 
  others 
  the 
  effect 
  of 
  any 
  rotatory 
  polarization 
  

   which 
  might 
  be 
  present, 
  so 
  that 
  there 
  resulted 
  some 
  irregu- 
  

   larity 
  in 
  the 
  observations. 
  in 
  different 
  azimuths. 
  Since 
  there 
  

   is 
  nothing 
  in 
  a 
  double 
  refracting 
  substance 
  to 
  distinguish 
  clock- 
  

   wise 
  and 
  counter-clockwise 
  rotation, 
  if 
  the 
  average 
  of 
  the 
  

   observations 
  in 
  many 
  different 
  azimuths 
  indicates 
  a 
  distinct 
  

   rotation 
  of 
  the 
  plane 
  of 
  polarization, 
  the 
  medium 
  must 
  possess 
  

   rotatory 
  polarization. 
  

  

  In 
  Table 
  I 
  are 
  tabulated 
  the 
  observations 
  with 
  this 
  glass 
  

   cylinder. 
  The 
  first 
  column 
  gives 
  the 
  number 
  of 
  the 
  experi- 
  

   ment, 
  the 
  second 
  the 
  length 
  of 
  the 
  cylinder 
  where 
  the 
  twist 
  

   could 
  take 
  place, 
  that 
  is, 
  the 
  distance 
  between 
  the 
  wooden 
  

   cylinders 
  cemented 
  to 
  the 
  ends. 
  After 
  the 
  tenth 
  experiment 
  

   the 
  cylinder 
  was 
  fractured 
  near 
  one 
  end 
  and 
  when 
  repolished 
  

   was 
  about 
  1^"" 
  shorter. 
  The 
  twisting 
  moment 
  per 
  centimeter 
  

   of 
  length 
  is 
  given 
  in 
  the 
  third 
  column 
  in 
  gram-centimeters. 
  

   The 
  fourth 
  column 
  gives 
  the 
  rotation 
  of 
  the 
  plane 
  of 
  polariza- 
  

   tion 
  per 
  centimeter 
  of 
  length, 
  after 
  applying 
  the 
  index 
  error 
  of 
  

   the 
  instrument 
  and 
  reducing 
  the 
  results 
  to 
  angular 
  measure. 
  

  

  