________________
AUGUST, 1915)
THE CHRONOLOGY OF THE PANDYA MONARCHY
169
Tribhuvana Vikrama Pandya, whose accession he states to have taken place "circa 1280." I do not quarrel with his opinion that a king (or may be a prince) lived at the period of inscription. That is a matter for the historian. I am only considering the case, as he did, from the point of view of chronology
One very good reason why a solitary date, although perfect and regular in itself, should not be accepted as conclusive proof of a reign will be better understood after a short explanation. Unless the number of the solar day of the month is stated, and it is not as a rule stated, all the ordinary details of a Chola or Paudya date will be found often to correspond with about three different days in a century. Thus in the case last mentioned Mr. Swamikannu Pillai shews that the elements of the date and surely also its palæographic character) would equally suit Thurs. 30th June 1278, or Thurs, 1st July A. D. 1305).
Hence in almost all cases, even though the details of the date are found to be perfect and regular for a certain civil day, it has to be steadily borne in mind that the same details will equally suit another day about 30 or 35 years earlier or later, and that paleography will rarely be of any assistance in coming to decision. When, however, the first date is confirmed by another, equally good, the doubt is of course at once removed.
Correction of errors in the original inscription. Those who have engaged themselves in this special line of research constantly have to deal with dates in inscriptions were mistakes appear to have been made either by the original framgr or by the engraver. If, for instance, we find a record belonging to the reign of a known king, of which the calendar-portion (day, week-day, tithi and nakshatra) works out perfectly correctly, but which quotes (gay) the 6th instead of the 7th regnal year as current at the time we should accept it as genuine and as actually appertaining to the given reign, but we should note the error and the fact that the date is not entirely regular. But when we find a mistake in the date-portion itself we have to be careful and to exercise sound judgment. It is often found that a mistake has been carelessly made in describing the lunar fortnight, the other details being correct; the numeral of the tithi is sometimes wrongly copied, or wrongly calculated ; and so on. A careful chronologist like Prof. Kielhorn will in such circumstances note the defect and state his reason for accepting the date.
But it is manifest that much greater caution has to be observed in the case of a record which cannot be assigned to the reign of any known king, and which is desired by the computer to establish the reign of a king of whom hitherto nothing has been heard. In such case it is clearly dangerous to correct the original and then build up a theory on the result.
Again, it seems hardly safe to alter more than one of the details given in the dive and then to build history upon it. Even if it were allowed in the case of a known reign, such a date should never be accepted as a sound basis for finally entering a new and previously unheard-of ruler on the historic list of kings.
Let me give a few instances. If the date of an inscription belonging apparently to the reign of a known king and certified by the Epigraphist to be approximately of that period, mentions the 3rd regnal year when the 2nd or 4th regnal year was current ; or if a dark fortnight is quoted instead of a light one; or if (say) a 6th tithi is quoted when by all known practise a 5th or a 7th tithi was the correct ono; or if a solar month should be quoted which is one place wrong; or the same with the position of the moon in the nakshatras; or with the week-day-if one such error occurs in a date otherwise satisfactory and regular we may assume a computer's or a copyist's or an engraver's error, and pass the date
