Ohfervations on the "Zodiac at DenJera. lit 



being lunar, a month of twenty-two and twentv-three days 

 alternately, named Mercedonius, was inferted after the 23d 

 of Febru-irv, to adjuft the lunar reckoning to folar ; but^ as 

 in the year of the Julian reform, the twenty-three days were 

 included, there would of courfe be one day's advance on the 

 calendar computation, which would leave but feven da3's by 

 the lunar account for the fun's place in the fign ; whence the 

 eleven units below would anfwer to the ordinary differences 

 between the lunar year and thefolar; whilR the two rays, 

 deficient in the fourth of the ftars, would point out the com- 

 mencement of the lunar year at Rome on the ift of March, 

 and the folar of Julius on the ill of January; or, in other 

 words, would correfpond to the augmentation of the year, 

 computed backward, which Numa had made. The ftars in 

 fquares, as before, defignating day?, which, to the bend of 

 the leg, amount to fixty-feven, anfwer to the beginning 

 of the year gone back by the Julian computation, while the 

 fevcn in addition correfpond to the fun's place in the fign, 

 which, with the five fupplementary days round the legs, 

 make the twelve days in the laft year of lunar intercalation, 

 and, as the four divifions between them fliow, are coincident 

 again with a quadrcnnium. 



Perfectly congi-uent with the whole is the pyramidal figure, 

 having a globe or fun at its top. It confifts of eleven grada- 

 tions, which anfwer to the eleven days between the folar and 

 lunar year, and the ninety-nine divifions upon it difcrimlnate 

 the ninety-nine days from the winter folftice, by which 

 Julius adjufted his reform, to the commencement of the 

 Roman year on the ift of April. Between, however, the 

 Roman lit of March and the lit of April a month was left 

 out, or, in other words, had gone back a fign. This is here 

 expreffed by the recefs of Cancer from the zodiac, and agrees 

 with Virgil's commencement of the year with Taurus : 

 Candiilus aur^tis afidrit ijoum cornibus annum 

 Taunts — 



the precife time of which is defined by what immediately 

 Follows : 



— cr advfr'b ccdens Canis ocridit aftro. 



If, now, the fctting of the dog-ftar be placed, according to 

 Petaii, on the fourth of the kalends of May, and the fixty- 

 fevcn days which the Roman year had gone back at tlic Julian 

 reform, thence reckoned, we come to the 26lh of February, 

 or ift of Thoth, at the commenccnient of the a^ra of Nabo- 

 naflar, which was that likewife of the Egyptian year. The 

 changes in the Roman year having varied its forui, in rcfpti'>t_ 

 t» its length, as well as the number of its month.'', and the 



