622 



Proceedings of the Royal Irish Academy. 



Taking tlie mean of all the differences for each hour of Right 

 Ascension, I found — 



0"^ Schj. 



-Lai. = + 



0^-099 from 133 



stars 



1 



+ 



-073 , 



72 



)) 



2 



+ 



-092 , 



80 



;) 



3 



+ 



0-143 , 



142 



J) 



4 



+ 



0-106 , 



121 



;> 



5 



+ 



0-051 



108 



>j 



6 



+ 



0-059 , 



128 



J7 



7 



+ 



-077 , 



112 



}} 



8 



+ 



-042 , 



125 



V 



9 



+ 



-024 , 



135 



n 



10 



4- 



-034 , 



117 



)> 



11 



- 



0-014 , 



132 



M 



12 



_ 



-055 , 



156 



5) 



13 



— 



-023 , 



115 



5) 



14 



+ 



0-032 , 



107 



1} 



15 



- 



-009 , 



137 



J J 



16 



+ 



0-055 , 



149 



)> 



17 



+ 



-076 , 



145 



5) 



18 



+ 



-042 , 



185 



)) 



19 



+ 



-036 , 



176 



7 ? 



20 



+ 



-043 , 



207 



7) 



21 



-1- 



-050 , 



174 



;; 



22 



+ 



0-062 ,, 



140 



J J 



23 



+ 



0-006 ,, 



169 



)) 



As will he seen from the number of stars within each hour, the 

 stars are not quite as uniformly distributed as might have been 

 wished, there being a minimum at l''-2^, and a maximum at 19''-20^. 

 But as these hours must rather be considered as exceptional, and as 

 the single differences within 1'' and 2'^ agree well inter se and with the 

 majority of the remainder, there is no reason to give these hours less 

 weight than the rest. It is not without interest to notice the preva- 

 lence of the plus sign around 3'', and of the minus sign nearly diame- 

 trically opposite. This agrees well with the direction of the motion 

 of the solar system through space. 



The mean result of the 24 hours is — 



Schjellerup-Lalande = + 0^-0459 + 0'-0061. 



From this difference we have now to separate the quantities depend- 

 ing on the constant errors of the fundamental systems, i.e. of Piazzi's 

 Catalogue and the Nautical Almanac. It appeared advisable to reduce 

 these to Newcomb's above-mentioned standard system, which is pro- 

 bably sensibly free from systematic errors ; besides, I could thus make 

 my final result directly comparable with ]S"ewcomb's corrected values. 



