( 382 ) 



Hence 



(i a = — 0"3li act'()r(liiiu' lo llic // 

 = + 0.25 „ ' .. .. /• 



Mean value of d « = — 0.05 



The two values for a ol)tai]ie(l iu this wav do jjot aaree satis- 

 factorily. The mean value «=1".45. however, dilfei's little from that 

 deduced above bv assumin,ii' the aujiual \ariatioji of X to be 2(P. 



The values of //,,• and /lV- reuiaiu also uioi-e or less ujicerlaiu. The 

 (ill and (^k show a systematic character cncu to a higher decree than 

 the <i X , but I did not succeed in lijidiug the real law of the discor- 

 dances. If, for instance, we assume that //,.. and /;. vary proporl ion- 

 ally with the time, the agreement does not improve. 



As the most prol)al)le results of my investigation I adopt: 



// = -[- 0".31 — T'.45 sin [302°.4 + 19'. 35 {t — 1876.0)] 

 /;■ = -[- 0".24 -I- i".45 co.s [302^4 + 19\35 (/ — 1876.0)] 



Thence follow as cori'ections of the eccenlricity aiid of the longitude 

 of the j>erigee : 



ÖP = — 0".16 



rri.-T = -[-0".I2 



rf\-r — _^ 2".2 



while an eventual correction of the motion of the perigee remains 

 entirely uncertain. 



The correction of the true longitude of the moon thus l)ecomes: 

 öl= — 0''.31 sin ,j — 0".24 cos ,j 4- 



-f 1".45 Hin [ij -f 212°.4 -f 19".35 (/ — 1876.0)]. 



With this formula we may conijnire the two results, \vliich 

 Battekmann derived from his occnltations and which hold for about 

 1885.0 and 1896.0. 



Battermann found for the total corrections depending on ij U'omp. 

 n°. 5 p. 41, n". 11 p. 52). 



1885.0 fl7 = — 1".14 sin <j -\- 2".67 cos g 

 1896.0 ,/ — — 0".90 mi ij — 1".35 cos g 



