VOL. LIII.2 PHILOSOPHICAL TRANSACTIONS. 25 



The mean of these 53 comparisons gives the sun's parallax = (^".61. Reject- 

 ing all those results which differ more than one second from the mean of the 

 whole, the mean of the remaining 45 results gives the sun's parallax = 8".55. 

 Rejecting all those results which differ more than half a second from the mean 

 of the whole, the mean of the remaining 37 results gives the sun's parallax = 

 8".57. And the mean of these three means gives the sun's parallax = 8".58. 



He next compares the observations of the internal contact at the egress made 



at Paris, Greenwich, Savile-house, Bologna, Madras, Grand Mount, and Tran- 



quebar, with those made at Stokolm, Upsal, Tornea, Cajaneburg, Tobolsk, 



Abo, Calmar, Hernosand, and Calcutta. The results are as in the following 



table. 



Stok. Upsal. Torn. Cajan. Tobo. Abo. Calm. Hern. Calcu. ^""'* "'^^" 

 "^ •* par. 



Paris 8".70 8".6"() 7".92 8"-43 8".6o 9".()3 S".63 8".42 7".83 S"A6 



Greenwich .t).66' ;).6v 8.42 9.09 .9.11 10.04 10. 16 9.2O 8.62 9.32 



Savile-house 8..iO 8 50 7.75 S.jO 8.66 9.04 8.50 8.l6 7.5+ 8.36 



Bologna 8.65 8 35 8.23 8.44 8.58 8.71 8.31 S.iiti 8.28 8.43 



Madras 10.33 10 15 8.90 .g.7 1 9.5+ 10.+8 10.80 977 9.32 9.89 



Grand Mount ... . 8.07 777 7.35 8.07 8.34 8.31 7-50 752 (1.9(1' 7.76 



Tranquebar 8.50 8.23 7 6"8 8.36 8.5.i 8.6/ 8-12 7-96 7.7 8.17 



Sun's par. mean .. 8.91 8.75 8.03 8.66 8.77 9-18 8.86 8.48 8.(.0 8.63 



The mean of these 63 results gives the sun's parallax = 8".63 ; and if we re- 

 ject all those which differ more than one second from the mean of the whole, 

 the mean of the remaining 49 results gives the sun's parallax = 8".50. And if 

 we reject all those which differ more than half a second from the mean of the 

 whole, the mean of the remaining 37 results gives the sun's parallax = 8 '.535: 

 the mean therefore of these three means gives the sun's parallax ^ 8".55. 



Thus by the mean of 53 comparisons the sun's parallax is determined tobe = 

 8". 58, and by the mean of 63 comparisons it is determined to be = 8'. 55. The 

 mean of these two means gives 8".565 for the parallax of the sun on the day of 

 the transit. It may be objected, that this determination cannot be depended on 

 to a very great precision, because the greatest difference of the effect of the pa- 

 rallaxes in any of these comparisons does not exceed 3' 31": consequently that 

 this is too small a base, from which we can expect any great exactness in the de- 

 termination of the sun's parallax. But if we consider the great number of com- 

 parisons (no less than 1 1 6), the certainty of the differences of longitude of most 

 of the places of observation, and the small differences in the results themselves, 

 Mr. S. thinks that the force of this objection is in some measure removed; and 

 that this determination of the sun's parallax, by the observations at places on 

 this side of the line only, must be very near the truth. 



In order therefore to remove the force of this objection entirely, let us next 

 consider the observation at the Cape of Good Hope, by which we shall have a 



VOL. xn. E 



