I. 



ON THE ADOPTED VALUE OF THE SUN'S 

 APPARENT DIAMETER. 



By E. S. HOLDEN. 



(Read 1874, January 17.) 



In all, 3639 observations of both horizontal and vertical diam- 

 eters of the sun have been examined, distributed as foUov/s : — 



No. OF Observatioxs. Instruments. 



Hor. Diam. Vert. Diara. 



Greenwich, 1862-1870, 832 905 Transit Circle. 



Washington, 18.i2-1865, 491 430 Transit and MuraL 



1866-1870, 490 491 Transit Circle. 



1813 1826 



These were made by twenty-three different observers. Addenduni 

 "A" gives the corrections to both horizontal and vertical diam- 

 eters as given by all the observations made in one year by each 

 observer. The Greenwich observations are compared with the 

 Nautical Almanac value ; the Washington observations with the 

 value given in the American Ephemeris. 



From Greenwich and Washington observations of horizontal 

 diameter, the concluded value is 32' 2 '.509 



From similar obs. of vertical diameter 32 2 .565 



The concluded value is 32 2 .54 



giving a correction — 1".46 to the Amer. Ephem. 



" 1 .08 to the Naut. Almanac. 



" _|_ .14 to tlie Berlin Jahrbuch. 



Addendum "A" contains the figures upon which this determi- 

 nation depends. Owing to a difficulty of determining from the 

 printed observations the observer with the transit instrument at 

 Washington in the years 1862-65, all the observations of each 

 of these years have been treated as if made by a single observer. 



Each observation of the sun gives an "Apparent error of ephe- 

 meris diameter" (horizontal and vertical) Addendum 'A" con- 

 tains the "error of observer" (by changing signs). If we now 

 subtract from the "Apparent error of ephemeris diameter" the 

 " Personal error of the observer" fpv eacli day, we have a series 

 of "outstanding errors" which are due to some cause exterior to 

 the instrument itself, and from which the influence of the observ- 

 er's personality has been eliminated. 



When thele residuals are tabulated and their means by months 

 taken, as in Addendum "B," it is seen that there is a periodicity 

 in the series. That this periodicity is not wholly or in maia 

 8 ■ (3) 



