( 384 



1892.5 after being reduced on tlie S3'stem of the 2"^^ 10 veai- cata- 

 logue are greater by + 0."6, -{- 0."1 and -f 0."5 respectively than 

 those of the transit observations at Greenwich ^). 



As to the occultations, it is proved by H. G. v. d. Sandf. Bak- 

 HUYZEN '^) that vahies for tlie moon's longitude derived from them will 

 generally be too small and therefore it is probable that the moon's 

 longitude according to the observations at Greenwich is still in iieed 

 of a positive correction. 



III. Investigation of the errors of tile latitude. 



10. My investigation of the errors of the moon's latitude was 

 based on that of the errors in declination. 



First 1 tried to determine the constant errors in the obser\ations of the 

 moon's declination and to this end I utilized the observations from 

 1895 to 1899. From the diiferences A <ƒ =: Comp.— Obs. I derived 

 mean values for each of the two limbs for each month of the year 

 and from them annual means were derived by taking the mean of 

 the monthly means without regard to their weights. 



In this way I obtained the results given in the following tal)]e. 

 The 2"^^ and 3"^ columns contain the annual means for the north 

 limb and the south limb, the 4*^'^ the means of the two, the 5'^'i their 

 differences i.e. the errors of the moon's diameter, while the G^'u-on- 

 tains this same error derived oidy fi-om siiiiidtaneous obser\ations 

 of the two limbs near full mooji. 



1) If we combine Franz's result from his Königsberg observations witli that 

 which he derived from those at Göttingen, which had been overlooked by me, the 

 last difference, instead of + 0."5, becomes +0."3 {Added 1903 Dec.) 



"') H. G. V. D. Sande Bakhuyzen : The relation between the briglitness of a 

 luminous point and the moments at whicli wc observe its sudden appeaiauce or 

 disappearance. Proc. Acad. Amst. 4. 405. 



