NEAR FORT ST. GEORGE. 5l9 



and in order to have separately the values of" these quan- 

 tities, we have at Paudree 



Latitude E L hy Aldehamn l3o 19' 49" .018 

 ElhyRegidus A\ .340 



Difference L I 7 .67s 



At Trivanaaporam^ 



Latitude Ea by Regidus . 1 lo 44' 49" .329 

 E j^hy a Ononis 41 .67 



Difference K \ 7 .879 



Hence LI \-k ^~ ^''.Qjs + 7.879 m 15.557 as before. 



15. It is therefore evident that, taking the decli- 

 nation of Aldeharan to be corrected, and the observa- 

 tion good, then the declination of a. Orionis should be 

 increased by 15''.557. Again, if we revert to what I 

 have said higher up (Para. 7 *) it may be inferred, that 

 these errors proceed from assigning originally too great 

 a difference of polar distance between the extreme stars 

 from which the mean latitude is deduced (as for ex- 

 ample between Arcturus^ and Procyon :) hence the 

 mean latitude, which is derived from both, w^ill fall 

 somewhat too low ; both on account of the excess of 

 this arc, and also of the error of polar distance oi P D 



16. It follows from this, that the successive small 

 arcs of declination Dy.yv, v d, do (Fig.3) will severally 

 be too great, and this seems to be the case with the de- 

 clinations of the twelve stars registered in Table II. 



17. It is true that in the foregoing example, where 

 I have compared the results given by Aldeharan, Regu- 

 lus and « Orionis, I have selected an extreme case ; but 



* Seeal'jo infra. Pa. I/. 

 LI 4 



