250 



fell on the 21st of March. But Sir W. Hamilton finds that 

 Vince's Solar Tables (or Delambre's, from which those are 

 formed) conduct to about 2^ hours before the Greenwich 

 mean noon of the 20th of March, as the true date of the 

 equinox in that year ; which thus appears to have been as- 

 signed to a wrong day, by some erroneous computation or 

 report, perhaps as long ago as the time of the phenomenon 

 in question. 



As this result is curious. Sir W. Hamilton conceives that 

 it may not be uninteresting to confirm it by a very simple 

 process of calculation, derived from the Gregorian Calendar. 

 According to that calendar, 400 years contain 146097 days, 

 being a number less by 3 than that of the days in four JuHan 

 centuries ; and if the farther refinement be adopted, which 

 some have suggested, of suppressing the intercalary day in 

 each of the years, 4000, 8000, &c., then, in the calendar 

 thus improved, 4000 civil years will contain 1460969 solar 

 days. Assuming then, as a sufficiently near approximation, 

 that such is the real length of 4000 tropical years ; multiply- 

 ing by 3, and dividing by 8, we find that 1500 tropical years 

 are equivalent to 5478G3 days and a fraction ; which fraction 

 of a day, according to this simple arithmetic, would be equi- 

 valent to 9 hours. But 1500 Julian years contain 547875 

 days, that is, 12 more than the number last determined ; and 

 these 12 days are precisely the difference of new and old 

 styles in the present century. If, then, we neglect the frac- 

 tion, the new-style date of an equinox in any year of the 

 nineteenth century ought to be the same with the old-style 

 date of the same equinox in the corresponding year of the 

 fourth century; and in particular the vernal equinox of 325 

 ought to have fallen on the 20th of March, because that of 

 1825 fell on the day so named: while the fraction of a day 

 above referred to, though not entirely to be relied on, ren- 

 ders this result a httle more exact, by throwing back the 

 equinox from the evening to a time more near to noon. 



