628 



PHILOSOPHICAL TRANSACTIONS. 



[anno 1775; 



of which the others, taking their numerical denomination from even numbers, 

 are totally destitute. Among the numerous believers in the moon of our days, 

 few, I suppose, retain any confidence in ihe physical powers of the odd numbers. 

 They may imagine therefore, that the apparent inconsistence of Pliny's rule with 

 the truth of things, may be owing to his superstition about the odd numbers, 

 which led him wilfully to deviate from the mean epochs, little apprized (for the 

 Romans never were astronomers) how much they sometimes differ from the 

 true ones, on account of the great and various inequalities of the moon's 

 motions, and how very widely his arbitrary arrangement would in consequence 

 often differ from the times it was intended nearly to represent. 



Instead of Pliny's critical days, I shall now therefore examine the days for 

 which I imagine they were substituted; those I mean of true syzygy, true 

 quadrature, and true octagonal aspect. The following table distinguishes the 

 changes of weather which fell on these days. There were only 22 such, out of 

 all the 69 ; which is scarcely 4 more than their even proportion. And rejecting, 

 as before, on both sides, the alterations of weather which were reversed within 

 the space of 24 hours, there remain, out of 46 changes in all, only 10 on the days 

 of lunar influence, which are 2 less than belong to them on the even chance ; for 

 the days of syzygy, quadrature, and octagonal aspect, in the whole year are 98 ; and 

 365 : 98 = 46 : 12-1- very nearly. 

 It is remarkable that, of these 1 

 changes, 2 only coincide with a 

 new moon ; namely, those of the 

 10th of February and 5th of Sep- 

 tember, and none at all with a 

 full moon. There were indeed 2 

 changes in the year on the day of 

 the full moon ; videlicet, those of 

 the 20th of September and 18th 

 of Nov. but both were reversed 

 within the space of 24 hours. 



Jan. 



Feb. 



March 



April 



May 



June 



July 

 Aug. 

 Sept. 

 Oct. 

 Nov. 

 Dec. 



69 22 



I have added in this table 2 columns, showing the number of changes in each 

 month, and the number out of each agreeing with the moon. I shall only add, 

 that no conclusion must be drawn from the observations of a single year. 



