﻿G. 
  O. 
  Comstock 
  — 
  The 
  Secular 
  Variation 
  of 
  Latitudes. 
  479 
  

  

  following 
  table 
  in 
  the 
  preparation 
  of 
  which 
  I 
  have 
  assumed 
  

   that 
  the 
  earlier 
  observations 
  were 
  reduced 
  with 
  sufficiently 
  

   accurate 
  values 
  of 
  the 
  proper 
  motions 
  to 
  require 
  no 
  further 
  

   correction. 
  The 
  error 
  of 
  this 
  assumption 
  will 
  in 
  some 
  measure 
  

   tend 
  to 
  counterbalance 
  the 
  error 
  made 
  above 
  in 
  the 
  same 
  con- 
  

   nection. 
  

  

  Correction 
  to 
  Auwers' 
  6. 
  

  

  Star. 
  

  

  1847. 
  

  

  Obs. 
  

  

  1883. 
  

  

  Obs. 
  

  

  1847-1883 
  

  

  h 
  Androm. 
  

  

  + 
  l"-09 
  

  

  17 
  

  

  — 
  0"-14 
  

  

  28 
  

  

  + 
  l"-23 
  

  

  Gr. 
  1450 
  

  

  + 
  1-12 
  

  

  2 
  

  

  + 
  0-69 
  

  

  24 
  

  

  + 
  0-43 
  

  

  1 
  Leo. 
  Min. 
  

  

  + 
  1-49 
  

  

  2 
  

  

  + 
  0-16 
  

  

  10 
  

  

  + 
  1-33 
  

  

  31 
  Leo. 
  Min. 
  

  

  + 
  1'79 
  

  

  5 
  

  

  — 
  0-31 
  

  

  23 
  

  

  + 
  2-10 
  

  

  17 
  H. 
  Can. 
  Yen. 
  

  

  + 
  1-79 
  

  

  3 
  

  

  — 
  0-15 
  

  

  13 
  

  

  + 
  1-94 
  

  

  7i 
  Herculis 
  

  

  + 
  1-49 
  

  

  5 
  

  

  + 
  0-18 
  

  

  28 
  

  

  + 
  1-31 
  

  

  3- 
  Herculis 
  

  

  + 
  1*78 
  

  

  3 
  

  

  — 
  0-26 
  

  

  19 
  

  

  + 
  2-04 
  

  

  a 
  Lyrae 
  

  

  + 
  1-52 
  

  

  192 
  

  

  —o-io 
  

  

  123 
  

  

  + 
  1-62 
  

  

  10 
  Lacertse 
  

  

  + 
  1-34 
  

  

  8 
  

  

  — 
  0-12 
  

  

  38 
  

  

  + 
  1-46 
  

  

  This 
  comparison 
  may 
  be 
  interpreted 
  as 
  indicating 
  either 
  

   that 
  the 
  latitude 
  of 
  Washington 
  changed 
  to 
  the 
  amount 
  of 
  

   l 
  //# 
  5 
  between 
  1847 
  and 
  1883 
  or 
  that 
  Auwers' 
  proper 
  motion 
  of 
  

   each 
  of 
  these 
  nine 
  stars 
  is 
  too 
  great 
  and 
  that 
  the 
  mean 
  value 
  

   of 
  this 
  error 
  is 
  //, 
  04L 
  I 
  do 
  not 
  at 
  present 
  see 
  how 
  to 
  draw 
  

   any 
  other 
  conclusion 
  and 
  of 
  the 
  two 
  the 
  former 
  appears 
  to 
  me 
  

   the 
  more 
  probable 
  especially 
  as 
  it 
  is 
  confirmed 
  by 
  the 
  Madison 
  

   observations. 
  I 
  therefore 
  adopt 
  for 
  Washington 
  

  

  Annual 
  variation 
  of 
  latitude 
  , 
  + 
  0"*042 
  

  

  I 
  have 
  searched 
  diligently 
  for 
  other 
  American 
  data 
  to 
  com- 
  

   pare 
  with 
  the 
  above 
  but 
  I 
  have 
  found 
  nothing 
  which 
  certainly 
  

   contravenes 
  it 
  and 
  but 
  little 
  which 
  confirms 
  it. 
  A 
  comparison 
  

   of 
  the 
  latitudes 
  determined 
  at 
  Annapolis 
  by 
  Chauvenet 
  in 
  

   1853 
  and 
  by 
  Brown 
  in 
  1883 
  indicates 
  an 
  increase 
  of 
  the 
  lati- 
  

   tude 
  by 
  l 
  //# 
  between 
  these 
  dates, 
  but 
  it 
  is 
  questionable 
  if 
  the 
  

   observations 
  are 
  comparable. 
  

  

  The 
  results 
  at 
  Cambridge 
  are 
  conflicting 
  as 
  is 
  shown 
  in 
  the 
  

   following 
  table 
  taken 
  from 
  vol. 
  xvii 
  of 
  the 
  Annals 
  of 
  the 
  

   Harvard 
  College 
  Observatory, 
  excepting 
  the 
  result 
  for 
  1845 
  

   which 
  I 
  have 
  derived 
  from 
  a 
  rediscussion 
  of 
  the 
  original 
  data 
  

   contained 
  in 
  Peirce's 
  memoir 
  on 
  the 
  Latitude 
  of 
  Cambridge. 
  

  

  Date. 
  

  

  Latitude. 
  Method 
  Employed. 
  

  

  1845-0 
  

  

  42° 
  22' 
  47 
  // 
  -00 
  z 
  b0"'19 
  Prime 
  Vertical 
  Transit. 
  

  

  55'8 
  

  

  47'61rfc 
  -08 
  Zenith 
  Telescope. 
  

  

  85-8 
  

  

  47*64^ 
  -02 
  Almucantar. 
  

  

  In 
  my 
  judgment 
  no 
  conclusion 
  can 
  be 
  drawn 
  from 
  these 
  num- 
  

   bers 
  until 
  the 
  relative 
  errors 
  of 
  the 
  several 
  methods 
  have 
  been 
  

  

  