﻿Rowland 
  and 
  Penniman 
  — 
  Electrical 
  Measurements. 
  37 
  

  

  or 
  as 
  a 
  single 
  coil, 
  the 
  coils 
  B^ 
  and 
  B^ 
  being 
  joined 
  up 
  in 
  series 
  

   or 
  in 
  parallel. 
  The 
  dimensions 
  of 
  the 
  coils 
  B^ 
  B2 
  were 
  the 
  same 
  

   as 
  K. 
  Each 
  of 
  the 
  coils 
  B^ 
  and 
  B^ 
  were 
  made 
  up 
  of 
  1600 
  turns 
  

   of 
  No. 
  22 
  B 
  and 
  S 
  single 
  covered 
  cotton 
  copper 
  wire. 
  The 
  

   self-inductance 
  of 
  these 
  coils 
  taken 
  separately 
  when 
  compared 
  

   with 
  P, 
  which 
  was 
  determined 
  absolutely, 
  was 
  nearly 
  1 
  henry. 
  

   On 
  this 
  account 
  B 
  was 
  taken 
  as 
  being 
  1 
  henry, 
  and 
  the 
  other 
  

   coils 
  were 
  compared 
  with 
  it 
  as 
  a 
  standard. 
  

  

  C. 
  Same 
  dimensions 
  as 
  P^. 
  Number 
  of 
  turns 
  ITttT 
  of 
  No. 
  

   22 
  B 
  and 
  S 
  single 
  covered 
  cotton 
  copper 
  wire. 
  Self-inductance 
  

   as 
  determined 
  1'30 
  henrys. 
  

  

  Condensers. 
  — 
  2 
  and 
  3. 
  Two 
  paraffined 
  paper 
  condensers 
  

   that 
  had 
  a 
  capacity 
  of 
  2 
  and 
  3 
  microfarads 
  respectively. 
  

  

  ■J^d 
  Troy. 
  A 
  ^d 
  microfarad 
  standard 
  mica 
  condenser 
  built 
  

   by 
  the 
  Troy 
  Electric 
  Co. 
  

  

  l-d 
  Elliott. 
  A 
  l^d 
  microfarad 
  standard 
  mica 
  condenser 
  built 
  

   by 
  Elliott 
  Bros. 
  

  

  Resistances. 
  — 
  The 
  resistances 
  used 
  in 
  the 
  experiments 
  were 
  

   of 
  two 
  kinds, 
  those 
  wound 
  with 
  double 
  wire 
  so 
  as 
  to 
  have 
  no 
  

   self-inductance, 
  as 
  the 
  ordinary 
  resistance 
  box, 
  and 
  those 
  wound 
  

   on 
  frames 
  or 
  cards 
  which 
  had 
  some 
  small 
  self-inductance, 
  but 
  

   almost 
  no 
  electrostatic 
  capacity. 
  The 
  resistances 
  which 
  had 
  

   self-inductance 
  are 
  called 
  open 
  resistances 
  to 
  distinguish 
  them 
  

   from 
  resistance 
  boxes, 
  and 
  were 
  of 
  different 
  kinds 
  and 
  dimen- 
  

   sions. 
  

  

  Sources 
  of 
  Error 
  and 
  Experimental 
  Difficulties. 
  

  

  In 
  all 
  work 
  with 
  alternatino' 
  currents 
  there 
  are 
  two 
  o:reat 
  

   sources 
  of 
  error 
  that 
  have 
  to 
  be 
  guarded 
  against. 
  These 
  are 
  

   the 
  errors 
  that 
  may 
  arise 
  from 
  the 
  inductance 
  of 
  one 
  part 
  of 
  

   the 
  apparatus 
  on 
  another, 
  as, 
  for 
  example, 
  the 
  direct 
  induction 
  

   of 
  a 
  coil 
  in 
  the 
  circuit 
  on 
  the 
  coils 
  of 
  the 
  electrodynamometer, 
  

   and 
  the 
  effect 
  of 
  the 
  electrostatic 
  capacity 
  of 
  the 
  leads 
  and 
  con- 
  

   nections. 
  In 
  connecting 
  the 
  coils 
  great 
  care 
  had 
  to 
  be 
  taken 
  

   to 
  avoid 
  the 
  effect 
  of 
  electrostatic 
  action 
  of 
  the 
  leads 
  and 
  con- 
  

   nections. 
  For 
  if 
  there 
  was 
  a 
  current 
  of 
  very 
  considerable 
  

   magnitude, 
  the 
  difference 
  of 
  potential 
  between 
  the 
  terminals 
  

   of 
  the 
  coil 
  might 
  be 
  great. 
  If 
  the 
  connections 
  under 
  these 
  cir- 
  

   cumstances 
  were 
  made 
  with 
  double 
  wire, 
  as 
  is 
  customary, 
  a 
  

   great 
  error 
  was 
  introduced 
  due 
  to 
  the 
  electrostatic 
  capacity 
  of 
  

   the 
  leads. 
  The 
  error 
  was 
  sometimes 
  as 
  much 
  as 
  7 
  per 
  cent 
  

   (see 
  method 
  24:). 
  This 
  error 
  could 
  be 
  shown 
  to 
  be 
  due 
  to 
  the 
  

   electrostatic 
  action 
  of 
  the 
  leads 
  by 
  shifting 
  a 
  resistance 
  in 
  cir- 
  

   cuit 
  with 
  the 
  coil 
  in 
  question 
  from 
  one 
  end 
  of 
  the 
  double 
  wire 
  

   to 
  the 
  other. 
  The 
  effect 
  of 
  this 
  was 
  to 
  still 
  further 
  increase 
  

   the 
  difference 
  of 
  potential 
  between 
  the 
  leads, 
  and 
  this 
  increased 
  

   the 
  error. 
  Experiments 
  of 
  this 
  character 
  showed 
  the 
  necessity 
  

  

  