( 386 ) 



Pencil of the iiioaii i-esidiials ^iives an equation of conditioii -. 



L6z=^ — ^^M sin (/- ^) (f ; 4- 0.96 ros (/— /^) / rf ^ 



where d i and ö S represent tlie correetiojis to tlie inclination and 

 tlie loniiitnde of the node. As ihe Z-rf may still contain an ontstand ing 

 constant error, the equations were actually put in the form : 

 Z_ Ö = a + h sot {1—6) + i' cos {I- 6). 

 These eqnations were solved substituting in tlieni, 1^^ |1iq mean 

 resnlls of tlie years 1895—1898, 2'^ those of 1899—1902, 3^1 those 

 of tlie 8 years combined (in all cases the mean resnlts as derived 

 from the total sums). 



In this Avay Ave ol»tained : 



h (• <■ corrected 



1895—1898 — 0"25 + 1"11 + 1"23 



1899—1902 +0.62 +0.63 ' +0.79 



1895—1902 +0"18 + 0"86 + 1"00 



Tlic last column contains the vabies of r corrected for the inHueuce 



of the 2'ï term of ' . The corrections actually ai)plied ai-e il> 

 (1). 



products with the mean corrections of the longitnde. 



The two partial results do not agree very Avell, especially those 

 for h, or for the correction of the inclination, and if we compare the 

 corresponding values of L ö from the two four-yeargroups, systematic 

 differences between the t\A'o sets are clearly shown. Considering 

 however my results in connection with those of Xewco.mb there 

 seems to be as yet no sufficient ground to assume a periodic part 

 in the coefficients h and c. 



From the 8 years combined we deriAC : 



fff = — 0".19 



106 = -^ rm 



ö6 = +11".5 



