( 417 ) 



Togetlier with the ephemeris I coinnuiiiicated a table containing 

 the variations of the right ascension and the declination by a variation 

 of tlie periiieUon passage of + 4 or — 4 days. In comparing'the 

 above given values — C' witii the numbers of that table, it is evident 

 that by a small negative variation of the perihelion passage, the 

 agreement between observation and computation may be nearly attained, 

 at least in «. The deviations in ö cannot be used so well for that 

 purpose, as the variations of rf, resulting from a variation of T, are 

 always much smaller than those of «, and this is especially the case 

 in the period during which these observations are made. Yet we 

 may conclude from the table for AT= — 4 days that the positive 

 errors in d will not entirely disappear by a variation of T. 



By means of a rough interpolation I derived from the 3 differences 

 — C in right ascension the following corrections for the time of 

 perihelion passage -. 



Observ 



of Aug. 28 

 ,, Sept. 25 

 „ Oct. 10 



AT = 



0.0900 day 

 0.0916 „ 

 0.0896 „ 



In the average AT= — 0.0904 day, which at the rate of a mean 

 daily motion of 51 7 "448 corresponds to an increase of the mean 

 anomalies of 46"8. 



As a first step to correct the adopted elements of the orbit, I 

 therefore computed the 3 places, in the supposition of an increase 

 of the mean anomalies: V by 40", 2° by 50". I interpolated the following 

 sun's co-ordinates (with reference to the mean equinox of 1906.0) 

 from the Naut. Almanac. 



TABLE III. 



For the reduction to the apparent places I added to the mean a 

 of the comet: f -{■ g sin {G -j- «) tg ö, to its mean 6: g cos {G -\- a). 

 The following table contains the computed apparent places in the 

 two suppositions. 



