﻿4 
  ON 
  THE 
  DIURNAL 
  VARIATION 
  

  

  this 
  gorge 
  both 
  on 
  the 
  east 
  and 
  west: 
  behind 
  the 
  western 
  hill 
  the 
  sun 
  

   passed 
  about 
  eighteen 
  minutes 
  before 
  the 
  time 
  of 
  sunset. 
  

  

  The 
  meridian 
  line 
  having 
  been 
  determined, 
  the 
  box 
  containing 
  the 
  

   needle 
  was 
  placed 
  upon 
  it, 
  the 
  zero 
  line 
  coinciding 
  with 
  the 
  meridian 
  ; 
  

   the 
  points 
  at 
  which 
  the 
  edge 
  of 
  the 
  box 
  was 
  cut 
  by 
  the 
  line 
  were 
  

   then 
  marked, 
  so 
  that 
  any 
  change 
  of 
  position 
  might 
  be 
  detected. 
  A 
  tem- 
  

   porary 
  inclosure 
  of 
  shingles 
  was 
  next 
  placed 
  around 
  and 
  over 
  the 
  plat- 
  

   form, 
  to 
  defend 
  both 
  the 
  platform 
  and 
  the 
  needle 
  from 
  the 
  sun 
  and 
  

   rain. 
  

  

  The 
  labour 
  of 
  prosecuting 
  hourly 
  observations, 
  by 
  night 
  as 
  well 
  as 
  

   by 
  day, 
  could 
  not 
  well 
  be 
  endured 
  by 
  a 
  single 
  individual 
  for 
  any 
  

   number 
  of 
  successive 
  days 
  ; 
  and 
  mine, 
  though 
  extending 
  through 
  parts 
  

   of 
  ten 
  days, 
  would 
  have 
  met 
  with 
  more 
  interruptions, 
  but 
  for 
  the 
  in- 
  

   telligent 
  aid 
  afforded 
  by 
  my 
  friend 
  and 
  former 
  pupil, 
  John 
  F. 
  Frazer, 
  

   who 
  took 
  a 
  share 
  in 
  the 
  labour 
  of 
  watching. 
  It 
  did 
  not 
  take 
  many 
  

   nights 
  to 
  perceive 
  that 
  the 
  period 
  from 
  1 
  A.M. 
  to 
  seven 
  or 
  eight 
  o'clock, 
  

   did 
  not 
  include 
  any 
  remarkable 
  points 
  ; 
  and 
  when 
  circumstances 
  seemed 
  

   to 
  permit 
  it, 
  I 
  did 
  not 
  continue 
  the 
  observations 
  between 
  those 
  hours. 
  

  

  A 
  few 
  trials 
  enabled 
  me 
  to 
  determine 
  that 
  the 
  space 
  between 
  the 
  

   tenths 
  on 
  the 
  scale 
  was 
  readily 
  divisible 
  by 
  the 
  eye 
  into 
  fourths, 
  with 
  

   accuracy 
  as 
  to 
  the 
  nearest 
  quarter, 
  both 
  by 
  my 
  assistant 
  and 
  myself 
  : 
  

   the 
  observed 
  variation 
  was 
  therefore 
  thus 
  registered, 
  and 
  the 
  limit 
  of 
  

   accuracy 
  of 
  one 
  reading 
  was 
  one 
  minute 
  and 
  a 
  quarter. 
  The 
  observa- 
  

   tions 
  have 
  been 
  turned 
  into 
  minutes 
  and 
  decimals 
  of 
  a 
  minute. 
  To 
  

   show 
  the 
  errors 
  of 
  reading, 
  no 
  better 
  test 
  can 
  be 
  had 
  than 
  to 
  compare 
  

   the 
  differences 
  of 
  two 
  readings 
  of 
  both 
  the 
  north 
  and 
  south 
  pole, 
  when 
  

   either 
  one 
  was 
  at 
  or 
  near 
  the 
  same 
  position. 
  As 
  the 
  most 
  unfavourable 
  

   specimen 
  of 
  such 
  readings, 
  may 
  be 
  taken 
  the 
  time 
  of 
  the 
  first 
  twenty- 
  

   four 
  hours, 
  when 
  the 
  observers 
  were 
  the 
  least 
  practised 
  : 
  such 
  a 
  spe- 
  

   cimen 
  is 
  given 
  in 
  the 
  first 
  table, 
  and 
  it 
  is 
  to 
  be 
  understood 
  that 
  the 
  

   observations 
  of 
  subsequent 
  days 
  presented 
  much 
  smaller 
  errors 
  of 
  ob- 
  

   servation, 
  than 
  the 
  average 
  of 
  those 
  there 
  recorded. 
  The 
  observations 
  

   were 
  begun 
  a 
  few 
  hours 
  before 
  noon, 
  and 
  each 
  table 
  comprehends 
  

   twelve 
  hours 
  of 
  two 
  successive 
  days. 
  

  

  The 
  temperature 
  of 
  the 
  air 
  was 
  noted 
  at 
  the 
  same 
  time 
  with 
  the 
  

   variation 
  ; 
  the 
  place 
  of 
  the 
  thermometer 
  was 
  in 
  the 
  shade 
  near 
  the 
  

  

  