( 380 ) 



Newcomb, but the öliglitly modified values, which were obtained by 

 taking he — -f 0".45 and h = + 0".26. 



Tlie three normal values obtained thus were : 



1852.6 iV = 200 7 Weight 1 (A — 6. — 9^0 



1868.5 161.9 ') 3 -[- 4.6 



1898.5 15.5 2 — 2.3 



and from them 1 derived a coiTected formula for iV^; I found : 



jVzzz 157°.3 + 19".35 (;f— 1868.5) 

 or taking the mean year as zero-e[)Ocii 



^Y= 3Ü2°.4 + 19°.35 {t — 1876.0). 



The outstanding differences Obs. — Comp. are given above. 



If I had assigned e(|ual \veiglits to the three normal values, I 

 should have found for the annual motion 19°. 45, while by excluding 

 tlie first I should have found 19°. 12, both differing only slightly from 

 the most ])robable value. 



At first when Newcomb's value for the annual variation of N 

 appeared to be too large I had thought that the true value might 

 be equal to the theoretical annual variation of the argument of the 

 Jovian Evection, i.e. 20°. 65. It appears, however, that even the latter 

 is too large to satisfy the observations. 



To judge in how far this is the case a comparison is given below 

 of the values of N for each year as directly derived from obser- 

 vations, first with my formula, secondly with the foi-mula we obtain 

 if we assume the same value of iY for 1876.0, but take as annual 

 variation 20^.65. The two sets of differences are given under the 

 headings JVq — ^^c and JS^o — -^V- 



Epoch W.'ight Xo—Nc No—Xj 



1) With Newcomb's values of N we should have found 200.°5 and 161. °7. 



