' ( 432 ) 



clironoiiielei' were luuiid l)v direet imerpolalion between tlie observed 

 values as liiveii in the table above. As the teniperatnre varied oulv 

 slow Iv 1 deemed it imnecessarv to aeeount especially for its inlluenee. 

 The riij;hl asceiision and deelination ol'the eoni[>arison stars ^^•ere laken 

 tViun the ephemerides of the nautical almanac ^\ hich for these years 

 are based on Nkwco.mb's Fundamental ( "afaloiiiie and for my purpose 

 could be considered as absolutely cori-ecl. This is not the case ^^ilh 

 the ri;^ht ascension and declination of the moiMi. as the errors in 

 the values compnte<l by means of the tables of IIaxskn-Xkwcomi? 

 jna\' still be ([uite sensible. 



Especially with a \ie\v to my observations, iv V. van ok Sandk 

 Bakhlyzkn has ujiderlaken an iu\estiualion of the errors of those 

 tables^) and 1 could a\ail myself of his results. Koi- the dei'i\atiou of 

 the corrections that must be a}>plied to the moon's places 1, on his 

 ad\ice, [)roceeded as follows. 



]>et us lirst consider the errors of the ion.ii'itude. For the constant part 

 of the correction of the mean lon.uitude 1 a(loi)te(l -[- '^ "-'^^Ml^vKii. 1 p. 

 o7(i and 3(S3) and to this 1 adde(l the coi-rections for the parallactic 

 iiie(pialit\- etc. computed from the tbrmula of üakii. I p. 375. The 

 sums of these two correctiojis \vere transformed into corrections of 

 the true loniiitude, and 1 auain added to these the corrections for 

 the ineipiality in the true loniiitude \\ hich Nkwcomh had been the 

 tirst to tind and which IJvKurvzKX lia<l deternnned anew from the 

 observations. 1 used therefor the \alues of the coefticieJils A and /' 

 as given by the formulae in ISaku. 11 p. 420, but added to either 

 of them the constant coi-rection — ()".3(), in ord(M' better to represent 

 the ol>servalions (comp. I.e. p. 41H). These corrections should be 

 applied to the orbit longitude, but with snfticient ap|)r()ximation might 

 be considered as corrections of the ecliptic longitude. 



As correction of the moon's latitude I a(k)pted (comp. 1>akh. 11 421 ), 



A ,? = — ()".5(S sill u + ().()8(^ [L I — Vr. 1 ) CCS- u 



where // represents die argument of the latitude and LI the total 

 correction of the longitude, while 1 assumed for the correction of 

 the kuigitude of the node for 1902 -|- i2".i. 



From the corrections of the longitude and the latitude hu'med in this 

 way, I derixed those of the riiiht ascensions and the declinatious by 

 means of the tables in Nkwcomh's " I iircsf'Kjdtioir . Another correction, 

 however, had still to be applied to the i-ight ascensions, the reduction 



1) I'l F. VA.x 1)1-: Sandk BAKHrYZKX, iiive.'-tigatiun ot the enois of tlie tables oMke 

 moon of H.wsr.x-XEWco.MB in Uie years iSUÖ-^l'.IOi::. I and II. Tioc. Acad. Amster- 

 dam Der. I'.Mli lUO:'). These two papers I sliall (|nol<' as Üaku. I and liAwn. II. 



