T II E R B 1 T O P U R A N U S . 171 



tlie observations. The mean corrections in geocentric lont,ntude for t^ronps of 

 observations are given in cohinins (2), co\uuui (1) sliowing the correction given 

 by the work of the hist chapter. 



Opposition lS(U-2 1SC2-3 



Mean +2 .79 +2.62 -f 2 .65 +2.50 —7.65 —7.82 



A systematic difference of 0".16 woukl seem to be indicated, and on acconnt of it 

 a correction of 0". 10 was applied to the comparisons of the hist few years iu 

 forming the equations of condition. 



In view of the possibility of systematic errors from this sonrce it may be con- 

 sidered that too great relative weight has been assigned to the results of the later 

 observations. If the residuals arise from errors of comparison and of theor}', their 

 probable magnitude is nearly as great at one epoch as at another. It may there- 

 fore be interesting to inquire wliat result we should get if, instead of assigning 

 such different weights to the comparisons at different epochs, we sought only for 

 the best general agreement with observations during the period the planet has 

 been observed. The preceding system of mean residuals will enable us to discuss 

 this question quite easily. In the first solution we shall reject the results from 

 Flamstead's observations, owing to their assured uncertainty, and those from 

 Le Monnier's of 1769, owing to the possible maladjustment of his quadrant. The 

 equations from the remaining residuals will be the following: 



1.1 +4.4 —1.8 —1.2 —0.8 =—0.21 



