﻿JAN. 
  19, 
  1922 
  LAMBERT: 
  LATITUDE 
  OF 
  UKIAH 
  37 
  

  

  rates 
  in 
  the 
  second 
  column, 
  particularly 
  the 
  rates 
  of 
  stations 
  now 
  dis- 
  

   continued, 
  depend 
  on 
  only 
  a 
  few 
  mean 
  latitudes, 
  and 
  offer 
  only 
  an 
  

   insecure 
  basis 
  for 
  conclusions. 
  

  

  The 
  rates 
  of 
  change 
  at 
  Ukiah 
  are 
  not 
  very 
  different 
  from 
  the 
  rates 
  

   found 
  from 
  Dyson's 
  curves, 
  namely, 
  +0".0094 
  by 
  Lawson 
  and 
  

   +0".0081 
  by 
  the 
  author.^ 
  The 
  striking 
  fact, 
  however, 
  is 
  that 
  a 
  rate 
  

   of 
  this 
  size 
  is 
  no 
  longer 
  a 
  solitary 
  phenomenon. 
  There 
  are 
  many 
  rates 
  

   of 
  this 
  order 
  of 
  magnitude 
  and 
  with 
  one 
  exception 
  all 
  rates 
  are 
  positive. 
  

  

  Before 
  seeking 
  an 
  explanation 
  I 
  think 
  it 
  will 
  be 
  wise 
  to 
  rule 
  out 
  the 
  

   results 
  for 
  Tschardjui 
  altogether. 
  If 
  results 
  were 
  weighted 
  according 
  

   to 
  their 
  probable 
  errors, 
  the 
  Tschardjui 
  results 
  would 
  get 
  weights 
  only 
  

   from 
  1/10 
  to 
  1/20 
  as 
  large 
  as 
  those 
  of 
  other 
  stations, 
  and 
  would 
  thus 
  

   have 
  little 
  effect 
  on 
  our 
  final 
  conclusions. 
  

  

  We 
  might 
  explain 
  the 
  positive 
  rates 
  at 
  all 
  stations 
  by 
  a 
  northward 
  

   creep 
  of 
  the 
  surface 
  strata, 
  as 
  Professor 
  Lawson 
  has 
  done 
  for 
  Ukiah, 
  

   but 
  such 
  an 
  explanation 
  is 
  scarcely 
  satisfactory 
  when 
  it 
  must 
  be 
  made 
  

   to 
  apply 
  to 
  so 
  many 
  stations. 
  A 
  better 
  partial 
  explanation 
  is 
  decli- 
  

   nations. 
  The 
  so-called 
  observed 
  latitudes 
  are 
  also 
  computed 
  ones 
  to 
  a 
  

   certain 
  extent, 
  and 
  errors 
  in 
  the 
  declinations 
  of 
  the 
  stars 
  used 
  appear 
  

   with 
  practically 
  full 
  effect 
  in 
  the 
  so-called 
  observed 
  latitudes. 
  It 
  

   would 
  appear 
  from 
  the 
  table 
  as 
  if 
  the 
  average 
  declinations 
  became 
  

   increasingly 
  erroneous 
  with 
  the 
  lapse 
  of 
  time; 
  an 
  error 
  of 
  this 
  kind 
  

   would 
  naturally 
  be 
  looked 
  for 
  in 
  the 
  adopted 
  values 
  of 
  the 
  proper 
  

   motions. 
  But 
  even 
  an 
  error 
  in 
  the 
  proper 
  motions 
  and 
  the 
  consequent 
  

   declinations 
  does 
  not 
  explain 
  all 
  the 
  rates 
  in 
  the 
  table. 
  An 
  error 
  in 
  

   the 
  star 
  places 
  would 
  affect 
  all 
  stations 
  alike, 
  except 
  insofar 
  as 
  bad 
  

   weather 
  might 
  cause 
  the 
  stars 
  actually 
  observed 
  to 
  vary 
  from 
  station 
  

   to 
  station. 
  It 
  is 
  clear 
  that 
  latitudes 
  are 
  increasing 
  much 
  faster 
  on 
  the 
  

   American 
  continent 
  than 
  elsewhere, 
  and 
  for 
  a 
  while 
  in 
  the 
  opposite 
  

   quarter 
  of 
  the 
  world, 
  as 
  at 
  Mizusawa 
  from 
  1900 
  to 
  1911, 
  they 
  were 
  

   actually 
  decreasing. 
  An 
  obvious 
  explanation 
  of 
  an 
  increase 
  in 
  latitude 
  

   on 
  one 
  side 
  of 
  the 
  earth 
  accompanied 
  by 
  a 
  decrease 
  in 
  the 
  other 
  is 
  a 
  

   shifting 
  of 
  the 
  Pole. 
  

  

  I 
  believe 
  that 
  the 
  explanation 
  of 
  the 
  changes 
  of 
  latitude 
  set 
  forth 
  in 
  

   the 
  table 
  will 
  be 
  found 
  in 
  a 
  shifting 
  of 
  the 
  Pole 
  combined 
  with 
  an 
  in- 
  

   creasing 
  error 
  in 
  the 
  declinations. 
  This 
  hypothesis 
  may 
  be 
  tested 
  by 
  

   a 
  least-square 
  adjustment. 
  Let 
  u 
  and 
  v 
  denote 
  the 
  components 
  

  

  ^ 
  The 
  number 
  in 
  the 
  table 
  most 
  nearly 
  corresponding 
  to 
  the 
  result 
  for 
  Ukiah 
  found 
  from 
  

   Dyson's 
  curves 
  is 
  the 
  -f0".0106 
  in 
  the 
  last 
  column. 
  The 
  difference 
  is 
  probably 
  due 
  in 
  great 
  

   part 
  to 
  the 
  declination 
  system 
  used. 
  See 
  below. 
  

  

  