﻿G. 
  C. 
  Comstock 
  — 
  The 
  Secular 
  Variation 
  of 
  Latitudes. 
  473 
  

  

  first 
  of 
  which, 
  observations 
  of 
  Polaris, 
  are 
  shown 
  in 
  the 
  follow- 
  

   ing 
  table 
  : 
  

  

  Latitude 
  of 
  JPidkowa. 
  Polaris 
  Observations. 
  

   Date. 
  Observer. 
  Latitude. 
  No. 
  of 
  Obs. 
  

  

  1843 
  Peters. 
  59° 
  46' 
  18"-73 
  ± 
  0*"013 
  371 
  

  

  1866 
  Gylden. 
  18-65 
  ± 
  *014 
  236 
  

  

  1872-5 
  Nyren. 
  18-50 
  ± 
  '014 
  155 
  

  

  1882-0 
  Nyren. 
  1.8*40 
  ± 
  *010 
  184 
  

  

  The 
  "first 
  three 
  of 
  these 
  values 
  are 
  taken 
  from 
  Nyren's 
  paper 
  

   on 
  the 
  latitude 
  of 
  Pulkowa, 
  the 
  last 
  one 
  is 
  derived 
  by 
  myself 
  

   from 
  the 
  data 
  given 
  at 
  p. 
  [50J 
  of 
  the 
  volume 
  of 
  Pulkowa 
  

   observations 
  cited 
  above, 
  after 
  correcting 
  the 
  individual 
  results 
  

   there 
  given 
  for 
  the 
  periodic 
  variation 
  of 
  the 
  latitude 
  (Kiistner) 
  

   by 
  means 
  of 
  the 
  formula 
  

  

  — 
  0-"26 
  sin(0— 
  47°) 
  

  

  derived 
  from 
  meridian 
  circle 
  observations 
  at 
  the 
  Washburn 
  

   Observatory 
  and 
  confirmed 
  by 
  the 
  special 
  observations 
  made 
  at 
  

   Berlin, 
  Potsdam 
  and 
  Prague 
  to 
  determine 
  the 
  existence 
  of 
  a 
  

   periodic 
  variation. 
  If 
  this 
  correction 
  were 
  omitted 
  the 
  value 
  

   of 
  the 
  latitude 
  for 
  1882*0 
  would 
  be 
  increased 
  0"-04. 
  A 
  graphi- 
  

   cal 
  treatment 
  of 
  these 
  data 
  indicates 
  an 
  apparent 
  diminution 
  

   of 
  the 
  latitude 
  amounting 
  to 
  0"'005, 
  six 
  inches, 
  per 
  annum. 
  

  

  To 
  this 
  it 
  has 
  been 
  objected 
  by 
  JBruns 
  (V. 
  J. 
  S. 
  vol. 
  xxv, 
  p. 
  15) 
  

   that 
  such 
  a 
  conclusion 
  presupposes 
  that 
  the 
  difference 
  between 
  

   the 
  tabular 
  refractions 
  employed 
  and 
  the 
  true 
  refraction 
  at 
  the 
  

   time 
  of 
  observation 
  is 
  the 
  same 
  for 
  the 
  several 
  epochs 
  and 
  that 
  

   this 
  assumption 
  requires 
  confirmation. 
  This 
  confirmation 
  is 
  

   afforded 
  by 
  an 
  investigation 
  by 
  Nyren 
  contained 
  in 
  vol. 
  xiv 
  

   above 
  cited. 
  From 
  a 
  discussion 
  of 
  the 
  observations 
  of 
  127 
  

   stars 
  north 
  of 
  +45° 
  declination, 
  not 
  including 
  Polaris, 
  treating 
  

   the 
  latitude 
  and 
  a 
  correction 
  to 
  the 
  assumed 
  refraction 
  as 
  

   unknown 
  quantities 
  he 
  finds 
  for 
  the 
  epochs, 
  

  

  1846 
  (?) 
  Latitude 
  = 
  59° 
  46' 
  18"*66 
  

  

  1866 
  ■ 
  " 
  18-546 
  

  

  Annual 
  variation, 
  -0"*006 
  

  

  This 
  astonishingly 
  close 
  agreement 
  between 
  the 
  results 
  of 
  

   two 
  independent 
  series 
  of 
  observations 
  will 
  probably 
  suffice 
  to 
  

   establish 
  the 
  constancy 
  of 
  the 
  refraction 
  at 
  Pulkowa 
  during 
  

   this 
  period. 
  The 
  corrections 
  to 
  the 
  tabular 
  refractions 
  for 
  the 
  

   two 
  epochs 
  differ 
  by 
  less 
  than 
  their 
  probable 
  errors. 
  

  

  There 
  is 
  available 
  another 
  series 
  of 
  Pulkowa 
  observations 
  by 
  

   which 
  to 
  test 
  the 
  reality 
  of 
  the 
  apparent 
  change 
  of 
  latitude. 
  

   I 
  have 
  compared 
  the 
  declinations 
  of 
  fourteen 
  stars 
  observed 
  

  

  