________________
74
EPIGRAPHIA INDICA.
(VOL. XXI.
animal. The idea underlying the emblem seems to be that the tampering with the seal is 88 heinous as the killing of a bull or violating the dharma.
As to the text of the legend, there is hardly any difference from what was given by Fleet. One might say, it rather confirms his reading and is, practically, alike in all the three specimens. It runs as given below; the slight difference which is seen here and there is pointed out in the foot-notes.
TEXT. 1 Chatus-samudr-ätikkränta-kirttih pratāp-anurag-panat-anya-rāja(0) Varnn-asrama
vyavastha2 pana-pravsitta-chakkraś=Chakkradhara iva prajánām=arttihara[ b*sri - Mahārāja-Hari
varmmā[l*] Tasya 3 puttras-tat-pad-ānuddhyāto Jayasvāmini-bhattārikā dēvyām®=utpannaḥ śri
Mahārāj. Adityava4 rmmā[ |*] Tasya puttras-tat-p[ā?]d-anuddhyāto Harshaguptā-bhattārikā-dēvyām
utpannah sri-Mahārā-8 5 Esvaravarmma (IM) Tasya puttras-tat-päd-anuddhyata Upagupta-bhattärika
dôvyām=utpanno 6 Maha(ā) rājādhiraja-sri-12 Isänavarmmă[ 11* Tasya puttras-tat-pād-änuddhyāto
Lakshmiva-13 7 ti-bhattarika-Mahādēvyām"=utpannah=paramamahësvarős Ma8 härājadhiraja-sri-Sarvvavarmnā Maukhari) [II]
The genealogy shown in these seals up to Isānavarmmā is identical with that given in the Haraha inscription which I published in this journal long ago.16
The seals of Harsha or Harshavardhana. · I may publish here only one of the seals of Harsha of Thānēsar. The Sonpat seal which Fleet reproduced in 1888 must have belonged to some plate which is not forthcoming. The Nālandā specimens so far recovered seem to be the impressions taken from some independent moulds. As we see even nowadays, the originals are only the negatives which, when pressed at the time of sealing documents, give the positives. Several specimens of the seal of Harsha
+ Like the Asirgadh seal both these specimens give rāja forrijo. * The vicarpa is left out in these seals sloo.
I think it is tri in all the three documenta. c. the symbol in line 8. • The letters putera and the t of ata are lost in Al. . It is clearly devyām in both these seals. • Letters rum fanya are missing in Al. * The length mark is obliterated in both the seal . Both these seals give Mahi. • Tbo akaharasjóvarava are missing in Al. u In both these seals, too, the symbol after u is more like ma than pa ; of. Fleet, C. 1. I., Vol. II, p. 220,
The longth mark is not to be seen in both the sesls. The initial lotters maha(a)ra are lost in Al.
Buth these soole give bi, of, the bus symbol in line 8. The histus in bri-Itana® is intentional, evidently.
u The La symbol is clear in both these seals and the reading Lakshmivati is pretty certain. Fleet's conjeotural reading of this name is therefore correct.
# A distanotly gives Mahao though the length mark is not so clear in Al. 10 mark is clear in both.
Vol. XIV. pp. 110 ff. #0.1. I., . Vol. III, pp. 231 f. and plato XXXII-B.