﻿472 
  G. 
  O. 
  Comstock 
  — 
  The 
  Secular 
  Variation 
  of 
  Latitudes. 
  

  

  We 
  have 
  here 
  observations 
  extending 
  over 
  a 
  period 
  of 
  

   nearly 
  two 
  centuries 
  during 
  which 
  the 
  latitude 
  appears 
  to 
  have 
  

   diminished 
  very 
  appreciably, 
  but 
  I 
  do 
  not 
  think 
  that 
  such 
  a 
  

   conclusion 
  can 
  properly 
  be 
  drawn 
  from 
  the 
  data. 
  Dr. 
  Auwers 
  

   informs 
  me 
  that 
  Bradley's 
  latitude 
  may 
  be 
  anywhere 
  from 
  

   half 
  a 
  second 
  to 
  a 
  second 
  in 
  error 
  on 
  account 
  of 
  uncertainties 
  

   inherent 
  in 
  the 
  data, 
  errors 
  of 
  figure 
  and 
  division 
  of 
  the 
  quad- 
  

   rant, 
  errors 
  of 
  the 
  tabular 
  refraction, 
  of 
  the 
  thermometer 
  ex- 
  

   posure, 
  etc., 
  and 
  the 
  same 
  may 
  probably 
  be 
  said 
  of 
  Pond's 
  

   latitude 
  while 
  Flamsteed's 
  is 
  much 
  inferior 
  to 
  either 
  of 
  these. 
  

   If 
  the 
  several 
  values 
  of 
  the 
  latitude 
  given 
  above 
  had 
  all 
  been 
  

   derived 
  with 
  the 
  same 
  instrument 
  and 
  by 
  the 
  same 
  method 
  

   many 
  of 
  these 
  errors 
  would 
  be 
  eliminated 
  from 
  the 
  differences, 
  

   but 
  in 
  fact 
  five 
  different 
  instruments 
  were 
  employed 
  and 
  the 
  

   entire 
  apparent 
  variation 
  of 
  the 
  latitude 
  may 
  fairly 
  enough 
  be 
  

   ascribed 
  to 
  the 
  undetermined 
  errors 
  affecting 
  the 
  results 
  given 
  

   by 
  these 
  instruments. 
  The 
  same 
  facts 
  obtain 
  for 
  much 
  of 
  the 
  

   evidence 
  sometimes 
  cited 
  to 
  show 
  a 
  variation 
  of 
  latitude 
  but 
  

   they 
  are 
  not 
  necessarily 
  true 
  of 
  all 
  of 
  it. 
  

  

  To 
  obtain 
  a 
  reliable 
  indication 
  of 
  a 
  change 
  in 
  latitude' 
  we 
  

   must 
  compare 
  determinations 
  made 
  at 
  a 
  sufficient 
  interval 
  of 
  

   time 
  by 
  the 
  use 
  of 
  the 
  same 
  instrument 
  and 
  the 
  same 
  methods, 
  

   or 
  we 
  must 
  compare 
  determinations 
  made 
  by 
  methods 
  which 
  

   are 
  practically 
  free 
  from 
  systematic 
  error, 
  such 
  as 
  are 
  furnished 
  

   by 
  the 
  zenith 
  telescope 
  and 
  the 
  prime 
  vertical 
  transit. 
  The 
  

   results 
  furnished 
  by 
  these 
  instruments 
  depend 
  upon 
  the 
  adoj)ted 
  

   star 
  places, 
  but 
  by 
  using 
  only 
  observations 
  of 
  the 
  same 
  stars 
  

   made 
  at 
  different 
  epochs 
  the 
  change 
  of 
  latitude 
  may 
  be 
  made 
  

   to 
  depend 
  solely 
  upon 
  the 
  proper 
  motions 
  of 
  the 
  stars 
  and 
  the 
  

   residua] 
  error 
  in 
  these 
  proper 
  motions 
  may 
  be 
  almost 
  indefi- 
  

   nitely 
  diminished 
  by 
  increasing 
  the 
  number 
  of 
  stars 
  employed. 
  

   I 
  assume 
  that 
  absolute 
  determinations 
  of 
  latitude 
  instead 
  of 
  

   being 
  the 
  only 
  data 
  from 
  which 
  a 
  motion 
  of 
  the 
  pole 
  can 
  be 
  

   concluded 
  are 
  in 
  the 
  present 
  state 
  of 
  practical 
  astronomy 
  

   decidedly 
  inferior 
  to 
  differential 
  determinations 
  for 
  this 
  pur- 
  

   pose. 
  If 
  these 
  principles 
  are 
  applied 
  to 
  the 
  data 
  collected 
  by 
  

   Fergola 
  and 
  presented 
  to 
  the 
  International 
  Geodetic 
  Associa- 
  

   tion 
  assembled 
  at 
  Rome 
  they 
  will 
  be 
  found 
  to 
  exclude 
  nearly 
  

   every 
  case 
  of 
  supposed 
  variation, 
  although 
  the 
  general 
  agree- 
  

   ment 
  of 
  the 
  data 
  in 
  indicating 
  a 
  progressive 
  diminution 
  of 
  

   European 
  latitudes 
  must 
  still 
  remain 
  a 
  very 
  striking 
  fact. 
  

  

  Of 
  all 
  the 
  cases 
  in 
  which 
  an 
  apparent 
  variation 
  of 
  an 
  abso- 
  

   lute 
  latitude 
  is 
  shown, 
  the 
  one 
  least 
  open 
  to 
  adverse 
  criticism 
  

   seems 
  to 
  be 
  the 
  discussion 
  of 
  the 
  latitude 
  of 
  Pulkowa 
  published 
  

   by 
  J^yren, 
  see 
  Die 
  Polliolie 
  von 
  Pulkoica 
  and 
  Observations 
  

   de 
  jPoulkova, 
  vol. 
  xiv. 
  There 
  are 
  here 
  two 
  independent 
  series 
  

   of 
  observations 
  made 
  with 
  the 
  vertical 
  circle, 
  the 
  results 
  of 
  the 
  

  

  