184 



C. J. MERFIELD. 



TABLE OF RESIDUALS— continued. 



m 











Residual 





o 



Observatory. 



Date 1899. 



oc 



o -c 



Weight 

 P 



8 



- c 



Weight 

 P 







d. 



p. 





u 







Lyons 



June 30-39 



+ 18-40 



10 



+ 968 



1-0 





Lyons 



„ 30-40 



+ 1821 



1-0 



+ 98-1 



1-0 



M 



Lyons 



„ 30-42 



+ 18-35 



1-0 



+ 993 



10 



> 



Windsor... 



July 295 



+ 18-47 



05 



+ 103-7 



05 



A 



Windsor .. 



391 



+ 18-08 



1-0 



+ 96-5 



1-0 





Bordeaux 



,, 4'44 



+ 17-78 



1-0 



+ 1034 



1-0 



4 



Bordeaux 



„ 540 



+ 1703 



1-0 



+ 100-8 



10 



ft 



Windsor 



5-89 



+ 17-66 



1-0 



+ 96-5 



1-0 





Bordeaux 



„ V-40 



+ 16-53 



1-0 



+ 92-3 



10 





Bordeaux 



„ 8-42 



+ 1606 



10 



+ 87-7 



1-0 





Bordeaux 



9-42 



+ 1595 



10 



+ 90-3 



1-0 







d. 



s. 





// 





< i-5 



Windsor... 



July 12-89 



+ 1602 



10 



+ 95-4 



10 



gU 1 



Bordeaux 



„ 1342 



+ 15-76 



10 



+ 970 



10 



%> 



Bordeaux 



„ 14-43 



+ 1538 



10 



+ 96-2 



10 



fc 



Bordeaux 



„ 1544 



+ 14-79 



10 



+ 86-3 



10 



n = [pn\ and ^ 



Construction op Normals. 

 Normal I. — This normal has been found by the following 

 method : — If we put n, n' etc. to represent the residuals, in the 

 case of either spherical co-ordinate, corresponding to the dates t, 

 t' etc., and as the interval of time between the extreme observa- 

 tions to be combined is small, then 



M 

 W 



The value of n being applied to the ephemeris position for the 

 date t we have the normal place for this date. The weight of 

 the normal will then be [p]. 



The Normals II., III., IV., VIII., have been found in a similar 

 manner. The several values of the definitive co-ordinates will be 

 found in the table, "Residuals for the Equations of Condition." 



Normal V. — The difference between the observed and computed 

 places cannot be considered as varying proportionally to the time 

 in the case of a , so that the error of the ephemeris has been com- 

 puted from an equation of the form 



n a = a + br + cr- 



1 



= T-t 



