624 ACCOUNT OF ASTRONOMICAL OBSERVATIONS 



Therefore a - 455-24 ; P = 1 1 86 ; Q. = — 12, and Ji; rr 

 S^l andj=^ ^.Px^q X. ^^^ =:13° 4' 3". 1 59, which 



differs only from the mean latitude by the sun (Tabic 

 V.)byO".l69. 



28. It is^ I own, no easy matter 'to give a reason for 

 so great a deviation. The difference between this posi- 

 tion of the sun when in the parallel of the Observatory, 

 and the latitude of that place as given by the stars, being 

 ]0".326, no error in the difference of longitude assumed 

 between the Greentvich and Madras Observatories can 

 accouut for it. For if wc take j' = 13° 4' 13".654 or 

 47053".654 (Table 11.) then resolving the equation wc 

 'have,^=: — ^ — 1 +_^'T^) - 4- ^ .y — « - iM 289 12 



or 1 day 6^^ bO^ IQs .9, which gives a difference of 8 

 minutes and 43 seconds of time for moving through 

 J 0^.326 of a degree in declination. So that if we 

 suppose the sun to be in the parallel given by the stars, 

 the interval of time allowed for the difference of longi- 

 tude^ between Madras and Greenwich must be dimi- 

 nished by that quantity, which is far beyond any un- 

 certainty that may still subsist on that head. 



29. Again ; as to -the time of apparent noon at 

 Madras, the sun's transit was always .observed with the 

 fixed transit instiument, whilst I was observing it with 

 the zenith sector ; and these contemporary observations 

 agreed always to a second of time. There can there- 

 fore be but little irregularity to apprehend from this 

 element, and wc are compelled, though under equal 

 objections and difficulties, to ascribe the error to the 

 declination either as assigned to the sun in the ephe- 

 mcrides, or to the stars in the catalogue for 1802 ; and 

 and I believe the former are the most likely to create 

 suspicion. 



