﻿1052 
  Resistance 
  of 
  Mixtures 
  of 
  Xylol 
  and 
  Alcohol. 
  

  

  by 
  the 
  methods 
  here 
  adopted, 
  and 
  no 
  determinations 
  of 
  its 
  

   temperature 
  coefficient 
  were 
  made. 
  

  

  5. 
  As 
  has 
  been 
  said, 
  all 
  these 
  observations 
  were 
  made 
  with 
  

   the 
  object 
  of 
  obtaining 
  information 
  valuable 
  for 
  the 
  con- 
  

   struction 
  of 
  resistances. 
  It 
  would 
  be 
  convenient 
  if 
  precise 
  

   information 
  could 
  be 
  given 
  as 
  to 
  the 
  specific 
  conductivity 
  and 
  

   temperature 
  coefficient 
  of 
  mixtures 
  of 
  various 
  concentrations 
  

   so 
  that 
  a 
  cell 
  of 
  known 
  " 
  conductivity 
  constant 
  " 
  could 
  be 
  filled 
  

   with 
  a 
  mixture 
  so 
  as 
  to 
  give 
  an 
  accurately 
  known 
  resistance. 
  

   And 
  perhaps 
  by 
  very 
  careful 
  investigation 
  of 
  the 
  properties 
  of 
  

   alcohol 
  prepared 
  in 
  various 
  ways, 
  a 
  specification 
  for 
  the 
  pre- 
  

   paration 
  of 
  resistances 
  of 
  accurately 
  known 
  properties 
  could 
  

   be 
  given. 
  Bat 
  it 
  appears 
  that 
  the 
  properties 
  of 
  the 
  mixtures 
  

   depend 
  so 
  greatly 
  on 
  small 
  changes 
  in 
  the 
  properties 
  of 
  the 
  

   alcohol, 
  that 
  it 
  would 
  be 
  much 
  more 
  trouble 
  to 
  prepare 
  a 
  

   specimen 
  of 
  alcohol 
  according 
  to 
  the 
  specification 
  than 
  to 
  

   measure 
  the 
  resistance 
  of 
  a 
  cell 
  when 
  it 
  had 
  been 
  constructed. 
  

   Accordingly 
  I 
  am 
  of 
  the 
  opinion 
  that 
  the 
  easiest 
  way 
  to 
  prepare 
  

   a 
  known 
  resistance 
  is 
  to 
  make 
  the 
  mixture 
  out 
  of 
  commercial 
  

   alcohol 
  and 
  xylol 
  and 
  then 
  to 
  measure 
  its 
  resistance 
  and 
  

   temperature 
  coefficient. 
  If 
  a 
  resistance 
  of 
  a 
  definite 
  number 
  

   of 
  ohms 
  is 
  required, 
  the 
  conductivity 
  constant 
  of 
  the 
  cell 
  can 
  

   be 
  estimated 
  from 
  its 
  dimensions; 
  it 
  can 
  then 
  be 
  filled 
  with 
  

   a 
  mixture 
  of 
  such 
  conductivity 
  (taken 
  from 
  Table 
  I.) 
  that 
  

   the 
  cell 
  will 
  have 
  the 
  desired 
  resistance. 
  In 
  this 
  way 
  a 
  cell 
  

   can 
  easily 
  be 
  prepared 
  with 
  a 
  resistance 
  which 
  does 
  not 
  differ 
  

   from 
  that 
  desired 
  by 
  a 
  factor 
  of 
  3. 
  If 
  a 
  closer 
  adjustment 
  

   is 
  required, 
  another 
  mixture 
  having 
  the 
  desired 
  conductivity 
  

   can 
  be 
  prepared 
  from 
  the 
  same 
  alcohol 
  from 
  the 
  information 
  

   contained 
  in 
  Table 
  I. 
  There 
  is 
  some 
  advantage 
  in 
  con- 
  

   structing 
  the 
  cell 
  so 
  that 
  the 
  mixture 
  which 
  is 
  required 
  to 
  

   fill 
  it 
  is 
  that 
  for 
  which 
  y 
  = 
  0'12 
  ; 
  for 
  the 
  resistances 
  of 
  such 
  

   mixtures 
  do 
  not 
  change 
  much 
  after 
  preparation. 
  It 
  is 
  true 
  

   that 
  such 
  mixtures 
  have 
  a 
  very 
  large 
  temperature 
  coefficient, 
  

   but 
  in 
  any 
  case 
  it 
  will 
  be 
  necessary 
  to 
  measure 
  that 
  coefficient 
  

   if 
  very 
  accurate 
  work 
  is 
  contemplated. 
  If 
  an 
  accuracy 
  of 
  

   ^ 
  per 
  cent, 
  while 
  the 
  temperature 
  of 
  the 
  cell 
  varies 
  3° 
  is 
  

   sufficient, 
  the 
  temperature 
  coefficient 
  of 
  the 
  conductivity 
  of 
  

   such 
  a 
  mixture 
  at 
  15° 
  may 
  be 
  taken 
  to 
  be 
  — 
  0*017. 
  

  

  The 
  best 
  way 
  to 
  determine 
  the 
  absolute 
  value 
  of 
  the 
  

   resistance 
  is 
  to 
  compare 
  it 
  with 
  that 
  of 
  a 
  wire 
  megohm. 
  

   If 
  a 
  resistance 
  of 
  about 
  10 
  9 
  ohms 
  is 
  constructed 
  and 
  an 
  

   accurate 
  voltmeter 
  is 
  available, 
  such 
  a 
  resistance 
  can 
  be 
  com- 
  

   pared 
  with 
  an 
  accuracy 
  of 
  -J 
  per 
  cent, 
  with 
  any 
  resistance 
  

   between 
  10 
  6 
  and 
  10 
  l2 
  ohms 
  by 
  means 
  of 
  the 
  method 
  shown 
  in 
  

   fig. 
  3. 
  When 
  it 
  is 
  desired 
  to 
  measure 
  very 
  small 
  currents 
  

   there 
  seems 
  to 
  be 
  no 
  advantage 
  in 
  choosing 
  a 
  resistance 
  of 
  

  

  