Page:Sewell Indian chronography.pdf/96

col. 3, below) we tabulated in col. 17 of Table I. by intervals of alternately 12 m. and 13 m. in order to avoid the use of seconds; so that while in one year the tabulated time is too early by 30 seconds, in the next it is correct. Thus while the actual moment of true Mēsha saṁkrānti in A.D. 499, K.Y. 3601 current, was 2 h. 27 m. 30 s. (but "2 h. 27 m." in Table I.), the actual moment in A.D. 500, K.Y 3602 current, was (2 h. 27 m. 30 s. + 6 h. 12 m. 30 s. =) 8 h. 40 m. exact (the same in Table I.) . This rule holds good for all years A.D., both before and after, A.D. 499 the tabulated time being correct in one year and subject to an addition of 30 seconds in the next.

[Note that in years B.C., owing to the fact that 1 B.C. immediately precedes A.D. 1, a current year of the K.Y. with an even number corresponds to a year B.C. with an odd number; and vice versa. Therefore for years B.C. the time of true Mēsha saṁkrānti (found by use of Table I., col. 17, and calculated backwards as directed in this work) is in whole minutes in odd years B.C. (current even years K.Y.); but, to be accurate, must have 30 seconds added in even years B.C. (current odd years K.Y.). Table XXXVIII.A gives the times of Mēsha saṁkrānti back to 59–58 B.C. on the same principle as in Table I., and its tabulated times can be made absolutely accurate to the second by application of this rule to the figures given in col. 17.]