126 THE ORBIT OF URANUS. 



way into groups of about a month each, and the mean date and mean correction 

 found for each group. The Paris and Konigsberg results are repeated for the 

 sake of clearness. The small figures show, as usual, the number of observations 

 employed in forming the mean. 



Aa AS 



Date. Observatory. Original. Corrected. 



1827, July 22, Speier, OM6, O'.U 



July 25, Paris, .03, .05 +0".5 3 



September 15, Vienna, _Q .11,, .10 .O u 



October 14, Vienna, 0.18, 0.17 2.2 6 



1828, July 25, Konigsberg, -0 .15 7 0.15 3 .5 7 

 July 29, Vienna, .24 2 .20 1 .4 2 

 August 14, Vienna, _() .13 10 .09 -fl .1 10 

 August 27, Speier, -0 .10 6 0.09 

 September 18, Vienna, -0 .03 9 -fO .01 +1 .0. 

 September 25, Cambridge, .()5 .16 



October 17, Vienna, -0 .13 U .09 .0 14 



October 17, Cambridge, 0.02 1U 0.12 



1829, August 1, Konigsberg, .10. .10 1.0 

 August 6, Cambridge, -f0.1l g 0.08 1 .! 

 August 28, Speier, -0 .04, .04 

 September 23, Cambridge, +0 .21 IO +0 .05 



' November 6, Cambridge, -4-0 .25, -j-0 .09 



Observations from 1830 (o 1872. 



Since the year 1830 heliocentric and geocentric ephemerides of Uranus com- 

 puted from Bouvard's Tables are at our disposal. We make use of those in the 

 Berlin Astronomisches Jahrbuch for the years 1830 to 1833, and of those in the 

 Nautical Almanac from 1834 forward. The system of comparison is the same 

 as that already explained. That is to say, we deduce separately : 



(1) Mean corrections to the geocentric longitude and latitude of Uranus in the 

 ephemeris as derived from observation. 



(2) Mean corrections to the same, given by the provisional theory, as derived 

 from a comparison of the heliocentric positions of that theory with the heliocentric 

 positions in the ephemeris. 



Then (1) (2) is the correction to the provisional theory given by observation. 

 The process of forming (1) and (2) is shown quite fully in the following pages. 

 Each individual printed observation was first compared with the printed ephemeris, 

 and a correction to the latter was thence deduced. When this correction was 

 given with the observations themsejves, it was of course not recomputed, unless in 

 some doubtful cases. The observations were then divided into groups, usually of 

 about a month each, and coinciding in time with the grouping of the Greenwich 

 results. The mean of the dates and the mean of the corrections were then taken 

 separately for each group and each observatory. The separate results are shown 



