________________
No. 29] A NOTE ON THE PONNUTURU PLATES OF GANGA SAMANTAVARMAN 173 māsa-divasē trin satimē, i.e., the 30th day of Māgha. Mr. Sarma supposes that the eclipse occurred on the day mentioned at the end, but of this there is no indication in the plates. As Mr. Sarma has himself observed, 'the charters in some cases at least were issued long after the actual date of the grant. This shows that the donee was in possession of the land or agrahāra given to him from the date of the actual grant and the royal charter recording the gift was given him some time later-after a lapse of some months or even years)'. The same seems to have happened in the case of the Tekkali plates of Dēvēndravarman. According to the epoch of A. D. 498, about the general correctness of which there should now be no doubt, the Gn. year 192 should correspond to A. D. 690. In that year there were two lunar eclipses-one in Jyēshtha and the other in Märgaśīrsha, but none in Māgha. The expression Māgha-trinsatima need not therefore signify Māgha pürnimā; it may as well denote Magha amāväsyā. This date does not therefore indicate that the month of Māgha mentioned in the Tekkali plates was pūrnimānta.
(4) Siddhāntam plates of Dēvēndravarman n . 195. These plates record in line 14 a grant made on the occasion of the Dakshinayana or Karkataka-sankranti. No tithi is mentioned in connection with it; but at the end in 1. 29 is mentioned Srāvana-krishna-dina-panchame as the date on which the charter was made over to the donee. There is no indication in the record that this was also the date of the Dakshiņāyana. There is thus no basis for Mr. Sarma's statement that the Dakshinayana referred to in the text of the grant must have occurred on the 5th day of the dark fortnight of the purnimanta Srāvana in 195 G. E.' As in the the case of the Pondutūru plates of Samantavarman dated Gn. 64 and the Achyutapuram plates of Indravarman II, dated Gn. 87, the Siddhantam plates of Dēvēndravarman dated Gn. 195 may have been issued some days after the grant was made. According to my epoch of the Ganga era, the Dakshiņāyana in the expired Gn. Year 195 occurred 3 h. 20 m. on the 22nd June A. D. 693 and the amanta Srāvana kțishna 5 ended on the 28t1 July A. D. 693. This date does not therefore go against the epoch fixed above ; nor does it indicate that the month Srāvaņa mentioned in it was pūrnimānta.
There are thus only two dates of the Ganga era which are recorded in the purnimānta months. As against this, in three dates cited before, the amanta reckoning is unmistakably noticed. This mixture of amanta and pürrimanta months in the dates of the Ganga era is not surprising; for the same thing is noticed in the case of other eras also. Kielhorn has, for instance, observed after examining several dates of the Vikrama era that the southern (Karttikādi) year of that era was joined with the pūrnimānta as often as with the amanta scheme. The dates of the Saka era are generally in amänta months, but Kielhorn noticed that in the case of one date, (viz., the Hyderabad plates of Pulakēģin II), the pūrnimānta scheme had been used. I have shown elsewhere that the months of the Kalachuri year were generally amanta in Mahārashtra and Gujarāt and purnimānta in Central India and Chhattisgarh, but in exceptional cases the other scheme also is seen to have been used. The same seems to have occurred in the case of the Gānga era also.
It will be noticed that the purnimänta scheme has been used in two of the earliest grants of the Ganga era, while the amānta scheme has been adopted in some later ones. The reason for this is
1 Ind. Cult., Vol. IX, p. 148, n. 1, * Above, Vol. XIII, pp. 213 f.
* Ind. Cult., Vol. IX, p. 147. If the Dakshiņāyana occurred on the pirnimanta Srāvans vs. di. 5 in A. D. 899, it was a mere coincidence. The purnimänta scheme is wholly inapplicable in the case of the dates cited above, on p. 172.
These plates were granted on the Udagayana, but were issued on Chaitra amávesyd. The Udagayans or Uttarayana could not have occurred in Chaitra.
. Ind. Ant., Vol. XXV, p. 272. .A.B.O.R. 1. Vol. XXVII, pp. 22 f.