﻿96 
  anintual 
  eeport 
  Smithsonian 
  institution, 
  1929 
  

  

  interest 
  to 
  note 
  that 
  the 
  periods 
  corresponding 
  to 
  the 
  fourth, 
  the 
  

   seventh, 
  and 
  the 
  ninth 
  harmonics, 
  which 
  we 
  find 
  so 
  well 
  marked 
  in 
  

   solar 
  variation, 
  have 
  also 
  been 
  particularly 
  noted 
  by 
  students 
  of 
  

   the 
  march 
  of 
  weather 
  and 
  crop 
  phenomena. 
  

  

  Assuming 
  that 
  the 
  harmonics 
  from 
  the 
  fourth 
  to 
  the 
  thirtieth 
  

   represent 
  all 
  the 
  real 
  regular 
  periodicities 
  in 
  the 
  variation 
  of 
  solar 
  

   radiation, 
  the 
  curve 
  B, 
  at 
  the 
  foot 
  of 
  the 
  diagram, 
  which 
  is 
  their 
  

   summation, 
  represents 
  the 
  march 
  of 
  this 
  periodic 
  part 
  of 
  solar 
  varia- 
  

   tion. 
  Continuing 
  it 
  to 
  cover 
  the 
  years 
  1929, 
  1930, 
  and 
  1931, 
  we 
  are 
  

   led 
  to 
  anticipate 
  features 
  of 
  uncommon 
  interest 
  in 
  the 
  march 
  of 
  solar 
  

   variation 
  in 
  the 
  period 
  just 
  approaching. 
  It 
  will, 
  indeed, 
  be 
  exceed- 
  

   ingly 
  interesting 
  to 
  see 
  to 
  what 
  degree 
  this 
  forecast 
  is 
  verified. 
  

  

  (h) 
  Reduction 
  of 
  Table 
  Blountain 
  observations. 
  — 
  Observations 
  at 
  

   Table 
  Mountain, 
  Calif., 
  which 
  have 
  continued 
  since 
  December, 
  1925, 
  

   have 
  been 
  critically 
  studied 
  at 
  great 
  length 
  during 
  the 
  past 
  year 
  by 
  

   Mr. 
  Fowle 
  and 
  the 
  computers. 
  Mr. 
  Fowle 
  has 
  considered 
  that 
  the 
  

   i-esults 
  might 
  be 
  affected 
  by 
  three 
  variable 
  atmospheric 
  elements 
  — 
  

   the 
  water 
  vapor, 
  the 
  haze, 
  and 
  the 
  ozone 
  which 
  occurs 
  in 
  the 
  very 
  

   high 
  atmosphere. 
  It 
  was 
  easy 
  to 
  arrange 
  the 
  data 
  in 
  groups 
  corre- 
  

   sponding 
  to 
  gradual 
  increase 
  of 
  quantities 
  of 
  atmospheric 
  water 
  vapor, 
  

   for 
  this 
  vapor 
  is 
  readily 
  measured 
  and 
  expressed 
  as 
  total 
  precipitable 
  

   water 
  by 
  Fowle's 
  method 
  which 
  he 
  worked 
  out 
  from 
  spectroscopic 
  

   study 
  in 
  the 
  laboratory 
  many 
  years 
  ago. 
  By 
  such 
  statistical 
  ar- 
  

   rangement, 
  corrections 
  for 
  precipitable 
  water 
  were 
  sought 
  to 
  be 
  

   obtained. 
  

  

  However, 
  there 
  is 
  one 
  obstacle 
  depending 
  on 
  the 
  contemporaneous 
  

   real 
  variability 
  of 
  the 
  sun 
  which 
  hinders 
  immediate 
  estimation 
  of 
  

   water-vapor 
  influence. 
  True, 
  this 
  solar 
  variability 
  might 
  have 
  been 
  

   eliminated 
  by 
  employing 
  the 
  definitive 
  results 
  of 
  Montezuma, 
  but 
  we 
  

   avoided 
  this 
  procedure, 
  since, 
  in 
  the 
  opinion 
  of 
  some, 
  it 
  might 
  not 
  

   have 
  left 
  the 
  Table 
  Mountain 
  observations 
  fully 
  independent. 
  Ac- 
  

   cordingly, 
  the 
  solar 
  variation 
  was 
  roughly 
  estimated 
  from 
  Table 
  

   Mountain 
  pyrheliometry 
  alone, 
  after 
  the 
  method 
  referred 
  to 
  in 
  my 
  

   report 
  for 
  1926, 
  page 
  116. 
  Allowance 
  was 
  thus 
  made 
  for 
  the 
  solar 
  

   variation 
  before 
  determining 
  the 
  water- 
  vapor 
  effect. 
  

  

  When 
  these 
  steps 
  had 
  been 
  taken 
  it 
  became 
  clear 
  that 
  a 
  sudden 
  

   increase 
  of 
  the 
  Table 
  Mountain 
  solar 
  constant 
  values 
  had 
  been 
  indi- 
  

   cated 
  about 
  August 
  12, 
  1927. 
  This 
  change 
  of 
  scale 
  continued 
  with 
  

   apparently 
  increasing 
  departures 
  thereafter. 
  No 
  parallel 
  result 
  hav- 
  

   ing 
  been 
  noted 
  at 
  Montezuma, 
  every 
  contributary 
  element 
  of 
  the 
  

   measurements 
  at 
  Table 
  Mountain 
  was 
  investigated 
  to 
  learn 
  the 
  

   source 
  of 
  the 
  discrepancy. 
  It 
  was 
  soon 
  found 
  that 
  the 
  change 
  was 
  

   due 
  to 
  a 
  large 
  change 
  in 
  the 
  scale 
  of 
  pyranometer 
  measurements 
  of 
  

   the 
  brightness 
  of 
  the 
  sky 
  near 
  the 
  sun. 
  Yet 
  redeterminations 
  of 
  the 
  

   iconstants 
  of 
  the 
  pyranometer 
  itself 
  by 
  observing 
  solar 
  radiation 
  with 
  

  

  