VOL. LIII.] PHILOSOPHICAL TRANSACTIONS. 1^ 



contact happened later than O** 35"' 23', by several seconds. On the whole 

 therefore, we may safely conclude that there is a mistake of one minute in set- 

 ting down the time of the internal contact at the egress at Rodrigues, and that 

 instead of O'' 34'" 47', it should be O'' 35" 47". This sort of mistake has hap- 

 pened several times in the observations of this transit, but they are easily dis- 

 covered. 



Mr. S. now proceeds to compare the observation of the internal contact at the 

 Cape, with the observation of the same contact at Rodrigues, and at 20 places to 

 the north of the line; and also the observation at Rodrigues with the same 20 

 places; and the results of the sun's parallax from these several comparisons are 

 as follow : 



Sun's parallax. Sun's parallax. 



Cape of Good Hope and Rodrigues 8". 54 Rodcigucs and Cape of Good Hope . . 8". 54. 



Paris 8.56 Paris 8.58 



Bologna S.54 Boloona 8 54 



Rome 8.74 Rome 8.94r 



Florence S.plr Florence 9.24r 



Gottingen 8.40 Gottingen 8.30r 



Greenwich 8.40 Greenwich 8.33 r 



Savile-house 8.57 Savile-house 8.5;> 



Leskeard 8.69 Leskeard 8.8 1 r 



Calmar 8.5.i Calmar 8.56 



Hernosand S.51 Hernosand 8.50 



Upsal 8.57 Upsal 8.58 



Stokolm 8.58 Stokolm 8.58 



Abo 8.63 Abo 8.68 



Cajaneburg 8. 56 Cajaneburg 8.57 



Tornea 8.4 1 Tornea SS,"? i- 



Tobolsk 8.64 Tobolsk 8.62 



Calcutta 8.43 Calcutta 8.37 



Madras 8.28 r Madras 8.03 r 



G. Mount 8.70 G. Mount 8.85 r 



Tranquebar 8.60 Tranquebar 8.74 



The mean of the 21 comparisons with the observation at the Cape, gives the 

 sun's parallax = 8''.56. There are only 2 of these 21 comparisons, marked with 

 the letter r, which differ more than -^ of a second from the mean of the whole ; 

 let these be rejected, and the mean of the remaining ig results gives the sun's 

 parallax = 8".56. 



If we select out of these 2 1 comparisons those places whose difference of lon- 

 gitude may be supposed to be the best determined, the mean of these may be 

 regarded as the most exact determination, viz. Paris, Bologna, Greenwich, Sa- 

 vile-house, Upsal, Stokolm, Cajaneburg and Tobolsk; the mean of these gives 

 the sun's parallax = 8". 55, and if we leave out the results of Greenwich and 

 Tobolsk, which differ the most from the rest, the mean of the remaining 6 re- 

 sults gives it = 8". 56, the same as before. 



The mean of the 21 comparisons with the observation at Rodrigues gives the 



E 2 



