________________
252
THE INDIAN ANTIQUARY.
been adopted with a view to keeping the day after the full-moon of Wazo, or of second Wazo, as near as is practicable to a suitable time for commencing the vassa or Buddhist Lent (as it has come to be called), the retreat into a fixed abode during the rainy season or the worst part of it. And the result is that, while the Thingyan Tet (mean Mesha-sarkranti) always occurs in Arakan in the first or the second Tagu, in Burma it sometimes occurs in the second month, Kason, answering to the Hindu Vaisakha in which month the Hindu true Mesba-samkranti can never fall. Again, the Hindu lunar month comprises 29 or 30 civil days according to the true movements of the sun and the moon: but the Burmese and Arakanese months have the fixed number of 29 and 30 civil days alternately; except that the intercalated month always has 30 days, and the third month, Nayon, which usually has 29 days, sometimes has 30 days in a year in which a month is intercalated. Also, the waxing or bright fortnight of the Burmese and Arakanese month,-which precedes the waning or dark fortnight bearing the same month-name, just as is the case in Southern India for civil purposes and everywhere in India for astronomical purposes, always comprises 15 days, while in the Hindu month the duration of either fortnight may range from 14 to 16 days. Further, in India, the Chaiträdi year, wherever it is used for civil purposes, changes its number on the day of Chaitra sukla 1, but in Burma and Arakan the lunar year changes its number on the day and at the time of the Thingyan Tet.
+
We would remark, in passing, that we are particularly interested in some of the details indicated above because the mean-time calendar of Burma and Arakan is a surviving illustration, in a general way, of the earlier system that prevailed in India before the period when the Meshādi solar year was established, and the use of true time was adopted, under the influence of the Greek astronomy which was introduced into India about A. D. 400. In that earlier period, as
[SEPTEMBER, 1910.
we know from the Jyotisha-Vedänga, the Iräbmans had a Maghadi lunar year, beginning with the first day of the bright fortnight of Magha, which was bound to a solar year beginning at the winter solstice, in the same way, but not on such strictly scientific lines, as that in which the Chaiträdi lunar year is bound to the solar year biginning at the Mesha-samkranti, which was originally, and still is nominally, the vernal equinox. All the details of the Maghadi year (except perhaps occasionally a crucial new-moon or solstice) were regulated by mean time. And two fixed months-either Ashadha and Pausha, or Sravana and Magha-were intercalated alternately.
Now, in determining the English equivalent of a Burmese or Arakanese date, the practical process is to start with the equivalent of Tagu waxing 1; just as, in determining the English equivalent of a Hindu lunar date on the lines followed in Sewell and Dikshit's Indian Calendar, we start with the equivalent of Chaitra sukla 1, which is given in that book for every year from A. D. 300 to 1900. But, as a result of the differences in detail between the Hindu calendar on the one side and the Burmese and Arakanese calendars on the other side, it is only occasionally that Tagu waxing 1 coincides with Chaitra sukla 1: it usually does not do so; and it may differ from Chaitra śukla 1 by as much as a full month. And, even when Tagu waxing 1 does coincide with Chaitra éukla 1, the same coincidence of days does not necessarily occur, and for the most part will not occur, during the rest of the year. It may be added that, though the Burmese and Arakanese calendars were in close if not actual agreement down to A.D. 1739; there are now considerable discrepancies between them, because the intercalations in the present time do not all fall in the same years.
In illustration of the differences between the Burmese and the Hindu calendars,
3 The Ceylonese Buddhists probably still intercalate in the same manner with the Hindus. But it appears that early in the last century there was an attempt to substitute a fixed intercalation of Aesala, Ashadha Wazo, on the Burmese lines: see Alwis' paper "On the Principles of Singhalese Chronology" in the Journal of the Ceylon Branch of the Royal Asiatic Society, 1855-8, p. 190.
It appears that this rule is so rigid that, if the full-moon day of Waso or of second Wazo, as determined by calculation, falls on a day which would involve allotting one day too much or one day too little to any of the preceding months, then that full-moon day, cardinal as it is, must give way, and must be placed one day earlier or one day later.
5 The statement has been made (JBBRAS, vol. 19, p. 135) that the Saka years must at first have begun with Magha. But all that we have to understand from the Pañobasiddhantika, 12.2 (which is the basis of the statement), is. that one of the five-year cycles of the Paitamaha-Siddhanta (the Jyotisha-Vädänga) began with Magha fukla 1 in (not at the beginning of) Saka-samvat 2 expired. The years of the Saka era were originally regnal years: and one reason for which the era was taken up by the astronomers, and so was perpetuated, apparently was, that they began at some time near the vernal equinox.
See note 9 on page 255 below.
