444 Remarks on Ancient Eclipses. 



Therefore 19 solar years contain 6939 days 14 hours 31 mi- 

 nutes, at the rate of 1 1 minutes per annum for the Julian excess ; 

 and by deducting three whole days in 912 years the n.oon com- 

 pletes her course within one hour and a half of the sun, and con- 

 sequently the cvcle contains (i9ii9 days 13 hours one minute, 

 which makes a difference of tliree whole days in 912 solar years. 



If the moon anticipates the sun three days in 912 solar years, 

 it is evident the moon's motion must he completed in less time 

 than the sun, and therefore the quantity of one hour and a half 

 must be deducted from the solar time as above stated : l)nt the 

 question is, wlicther this equation for the lunar anticipation is 

 any real element in the lunar astronomy; or whether it may not 

 be accounted for on the unknown measures of the moon's true 

 motions. The concurrence of so many examples of eclipses of 

 the sun and moon at fixed periods, stated in the preceding re- 

 marks, affords so many convincing testimonies for the coincidence 

 of their motions, as seems to exclude the idea, that any positive 

 anticipation of either the sun or moon really exists ; and for this 

 plain and self-evident fact, that every eclipse is caused by a uni- 

 son and concurrence of their motions. 



The true measure of a lunar cycle of 19 years may, it is pre- 

 sumed, be best determined from a comparative statement of the 

 motions of the sun and moon with the elements of each at that 

 interval, as computed in the British Nautical Almanack, which 

 is the highest authority to be consulted in so exact a determina- 

 tion. 



1801. 1820. 



D. 

 Sept. 



]) Longi- 



