The Pandyan Kingdom from the Earliest Times to the Sixteenth Century c8: Chapter IV. The Transition to the First Empire: The Kaḷabhras

The Pāṇḍyan Kingdom from the Earliest Times to the Sixteenth Century c8

Language: English

Chapter IV. The Transition to the First Empire: The Kaḷabhras

WE have no information as to the exact steps by which the transition was brought about from the conditions reflected in the Śangam literature to those of the Age of the First Empire, as we propose to call it,—an age comprising roughly three centuries from, say, the beginning of the seventh century to the beginning of the tenth. And the great danger at this point of the story is the temptation to make hasty reconstructions by piecing together fragments from literature and epigraphy which at the first blush seem to have a connection with one another, but on closer scrutiny fail to support the superstructure they have been made to bear. Great impediments to a proper understanding of the records result sometimes from chance suggestions thrown out by authoritative scholars which are often repeated by their followers without the reservations with which they were originally given and occasionally even palpable errors are handed down as gospel.¹ In the present account no attempt will be made to carry the story beyond what is exactly warranted by the state of our sources and the distinction will be carefully maintained between facts as such and subjective impressions. If in consequence we get an account which is discontinuous, that is unfortunately a result which in the present state of our knowledge cannot be helped.

c8.4

¹ The brilliant epigraphist, Mr. Venkayya, did much good work on the Pāṇḍyas. He always stated his results with caution. When he wrote, some mistakes were unavoidable in the study and interpretation of old records

c8.5

which he handled for the first time. But his successors make a mistake when they fail to test his results independently for themselves before accepting them as established facts. But (1) Mr. Venkayya must in some measure be held responsible for starting the habit of confusing different kings with similar names. Thus he said, ‘The names Neḍumāṛan, Neḍuñjeḷiyan and Neḍuñjaḍayan are quite similar, and one is almost tempted to think that they must have denoted the same individual’, and though he took care to add ‘Beyond this similarity of the mere names we possess no materials for their identification’ (*I.A.*, vol. xxii, p. 65), some of his successors have not exercised the same caution—e.g. the so-called Śeḷiyan of the Veḷvikkuḍi grant has been identified with Neḍuñjeḷiyan by Mr. K. V. S. Aiyar (see chap. iii. above). (2) Mr. Venkayya also misread some of the names of the kings. He called Kaḍungōn’s grandson Śeḷiyan Śēndan (*A.R.E.*, 1907-08, p. 51) in summarizing the Tamil portion of the Veḷvikkuḍi grant. Again, in the same summary, he called the father of Parāntaka by the name Tērmāṛan; it is strange that Mr. Krishna Sastri should have followed this reading of the king’s name and perpetuated the mistake in his edition of the grant in *E.I.*, vol. xvii. The king’s name is only Māṛan; *tēr* should be read along with the preceding *mān* as a compound adjective to the king’s name. (See K. G. Sankara, *I.A.*, vol. li, p. 214). By a similar mistake he recognized a Tēr Varōdayan besides Varōdayan in the *Iraiyanār Ahapporuḷ* stanzas and identified the hero celebrated in those verses with his Tērmāṛan (*A.R.E.*, 1907-8, p. 57). In this, however, Mr. Krishna Sastri takes care not to follow him (*E.I.*, vol. xvii, p. 297). (3) Lastly he committed a somewhat serious error in the collation of the data from the larger Sinnamanūr and the Veḷvikkuḍi grants and working up the genealogy of the Pāndyas mentioned in these grants (*A.R.E.*, 1907-8, pp. 54-55). This has been set right in a large measure by Mr. Krishna Sastri giving up the consideration which was the most vital in Venkayya’s account of the matter, viz., ‘that the Neḍuñjaḍayan of the Veḷvikkuḍi grant cannot be identical with his namesake of the Madras Museum plates, but that the former must be earlier than the latter’ (see also T. A. Gopinatha Rao in *Sen Tamil*, vol. vi, pp. 437-39). Gopinatha Rao’s genealogy at p. 154 of *Trav. Arch. Series*, vol. i, is inaccurate and insufficiently documented.

c8.6

The chief records which furnish data on the genealogy and chronology of the Pāṇḍyas of the First Empire are—

c8.7

1. The Veḷvikkuḍi grant of Parāntaka Neḍunjaḍayan.
2. The Smaller Śinnamanūr plates.
3. The Larger Śinnamanūr plates of Rājasimha.
4. The Madras Museum plates of Jaṭilavarman.
5. The two related Ānamalai stone inscriptions of Māraṇjaḍayan (453 of 1906) and Parāntaka (454 of 1906), the latter dated in the Kaliyuga era and yielding A.D. 770.
6. The Aivarmalai Inscription of Varaguṇa dated in Śaka 792, i.e. 870 A.D. corresponding to the eighth regnal year of the King.

c8.8

The first thing to do with these records is to settle the genealogy of the dynasty with the aid of the copper plate grants which give particulars enabling us to do this and to assign these grants themselves to the respective kings to whose reigns they belong. It has been generally accepted that the Sanskrit portion of the Veḷvikkuḍi grant mentions only the last four names in the list contained in the Tamil portion which extends over seven generations and the last king Parāntaka *alias* Jaṭila of the Sanskrit part is identified with Neḍunjaḍayan the last king of the Tamil part and corresponding identifications made in regard to the three preceding generations. It has also been generally accepted that this Parāntaka Neḍunjaḍayan of the Veḷvikkuḍi grant is the same as Jaṭila, the second king of the Sanskrit portion of the larger Śinnamanūr plates—whose name has been unaccountably passed over in their Tamil part. But this arrangement leads to a duplication of kings with the same name for which there seems to be no warrant or explanation. On the other hand, if we identify Parāntaka Neḍunjaḍayan of the Veḷvikkuḍi grant with Varaguṇa Mahārāja of the larger Sinnamanūr plates, this difficulty is avoided. And we have ample support for this course. Parāntaka is coupled with Neḍunjaḍayan by the Veḷvikkuḍi grant and with Māranjaḍayan i.e. Śaḍayan the son of Māran by the Ānamalai records; and the Trichinopoly inscriptions (413 and 414 of 1904) of the reign of Māranjaḍayan call the king Pāṇdyādhirāja Varaguṇa.¹ By this identification of Parāntaka Neḍunjaḍayan with Varaguṇa, not only is the perplexing duplication of Rājasimha avoided, but Varaguṇavarman and Parāntaka Vīranārāyaṇa become the grandsons of Parāntaka alias Varaguṇa Mahārāja which seems quite natural. It also seems to me—this is only a personal impression—that the campaigns indirectly referred to in the Ambasamudram inscription of Varaguṇa Mahārāja² are best ascribed to Parāntaka Neḍunjaḍayan of the Veḷvikkuḍi grant, who according to that grant fought and won battles on the banks of the Kāvēri early in his reign. One apparent objection to this course is found in the chronological indications given by our records and this has been held³ to be fatal to the arrangement of the genealogy of the Pāṇdyas suggested here; but we shall see presently that the difficulty, if there is one, is not insuperable. It may be well at this stage to give the genealogy of the kings as fixed by the identifications proposed hitherto.

c8.9

¹ There can be no doubt that the ājñapti of the Veḷvikkuḍi grant is the same as the excavator of the Ānamalai temple and that consequently these two records belong to the same king. Mr. Venkayya was inclined to ascribe the Trichinopoly records to Varaguṇavarman, the elder son of Srīmāra Srīvallabha, but he stated no reasons for his view which he said was only provisional. (A. R. E., 1907, p. 53, para 21.)

c8.10

² Edited by Venkayya, E.I., vol. ix, pp. 84 ff.

c8.11

³ See Mr. K. V. S. Aiyar, Ancient Dekhan, pp. 103-4.

c8.12

c8.13

Before taking up the chronology of the period for discussion, something must be said about the smaller Śinnamanūr and the Madras Museum plates. These two sets are engraved by apparently the same scribe and this justifies the assumption that the records must be assigned to the same reign, especially as the writing in the smaller Śinnamanūr inscription is held to resemble that of the Madras Museum plates of Jaṭilavarman.¹ But these considerations are not as conclusive as one would wish. It must be noticed that while the smaller Sinnamanūr plates make Arikēsari the son of Jayantavarman, the Vēḷvikkuḍi grant is not so specific in the Tamil portion and indicates the relation between Arikēsari and Sēndan by the vague phrase ‘மற்றவர்க்குப் பழிப்பின்றி வழித்தோன்றி’,² and the Sanskrit portion of the Vēḷvikkuḍi grant gives no help here as it begins only with Arikēsari. After some discussion, it has become clear that the Madras Museum plates of Jaṭilavarman too must be ascribed to Parāntaka Neḍunjaḍayan of the Vēḷvikkuḍi grant. Palæographical considerations which were once held very strongly to militate against such a view have not stood the test of further criticism.¹ On the other hand, several weighty considerations can be urged in support of the view now taken of this grant. *First* several surnames of the ruling king occur in both

c8.14

¹ See A.R.E., 1907, p. 52 and K. V. S. Aiyar (op. cit., pp. 103-4). Nothing can be stated finally about the smaller Sinnamanūr grant. The grant is incomplete and the text is still unpublished. It is not known how many plates are missing. There is however just a possibility that the grant belongs to Kōccaḍayan (No. 5 of the genealogy in the text), the king who fought at Marudūr, in which case the engraver may be taken to be the grandfather of the engraver of the Madras Museum plates. But there can be no doubt that Venkayya was strangely misled by vague palæographical considerations into ignoring the probability of the identity of Sēndan and Jayanta.

c8.15

² Mr. Krishna Sastri gets over the difficulty by saying வழித்தோன்றல் means ‘son’; I am afraid that this gives no help in interpreting the expression quoted which is not வழித்தோன்றலாகி, but வழித்தோன்றி. Nor is Mr. K. G. Sankara’s suggestion that Arikēsari was the son of a daughter of Sēndan easy to accept. If, as is not unlikely, the smaller Sinnamanūr plates are earlier than the Vēḷvikkuḍi grant and belong to the reign of Raṇadhīra, then their evidence must be preferred to that of the Vēḷvikkuḍi grant. But after all, it may be that the Vēḷvikkuḍi grant itself is not in conflict with the Sinnamanūr plate. For some reason the Sanskrit part of the Vēḷvikkuḍi grant begins only with Arikēsari and the Tamil part, for the same reason, may be taken to make a fresh start with this king, Arikēsari, though he was the son of the immediately preceding Sēndan (E.I., vol. xvii, p. 365, n. 5 and K. G. Sankara, I.A., vol. li, p. 213 and Q.J.M.S., vol. x, p. 178). It may be noted in passing that nothing has turned up to justify Mr. Venkayya’s suspicion that the kings of the smaller Sinnamanūr and the Madras Museum plates may not belong to the main line of the Pāṇdyas (A. R. E., 1908, p. 55).

c8.16

¹ See A.R.E., 1908, pp. 50 and 55 and H. K. Sastri in E.I., vol. xvii, p. 293. Mr. Venkayya said, ‘ The characters of the Sanskrit portion (of the Veḷvikkuḍi grant) are older than those of the Madras Museum plates of Jaṭilavarman and of the two Śinnamanūr ones.’ Again, ‘ the numerals which are marked in the Veḷvikkuḍi grant are very old, while those of the Jaṭilavarman plates bear a close resemblance to the corresponding symbols used in the larger Śinnamanūr plates.’ Mr. Krishna Sastri in editing the Veḷvikkuḍi grant enumerates the differences in the Grantha (Sanskrit) portions of the two sets of plates; but a careful study of them does not seem to justify his conclusion that the Tamil portion in both the grants formed the original ‘grants proper in both’ cases, and ‘the insertion of the Grantha portion in the Veḷvikkuḍi grant might have been somewhat earlier than that in the Madras Museum plates.’ The only proof for this statement that is furnished by Mr. Sastri is in ‘the remark that the Sanskrit portion (of the Veḷvikkuḍi grant), by its brief notice and the very meagre historical material which it supplies in the form of a general introduction, could not have been contemporaneous with the Tamil portion.’ As against these statements it may be pointed out : (1) a glance at the E.I. and the I.A.) does not support the view that the Sanskrit parts are later insertions ; (2) the differences in the Sanskrit portion enumerated by Mr. Sastri are not serious and may be only due to the facts (a) that different scribes were employed to engrave the two sets of plates and (b) there is an interval of at least fourteen years between them so that in any case the Sanskrit portion of the Veḷvikkuḍi grant will be ‘somewhat earlier than that in the Madras Museum plates’; (3) without the Sanskrit portion, the Veḷvikkuḍi grant will begin with கொல்யானை and the Madras Museum plates with அண்ணனாகிய, and we may take it that no inscriptions could have begun so inauspiciously or so abruptly; (4) the meagreness of the historical particulars in the Sanskrit part is only to be expected because that part was intended only to be a general introduction to be read together with the Tamil portion; it should be noted there is not merely a Sanskrit introduction but also a Sanskrit epilogue to the two sets of plates; (5) lastly, arguments from palæography are always open to doubt and much more so in the case of early South Indian palæography owing to the scarcity of the records available for comparative study.

c8.17

The Tamil parts of the two sets show only a single perceptible difference to which Mr. Krishna Sastri has called attention; but this ‘need not show that the two grants must belong to different periods’.

c8.18

that of the Madras Museum plates of Jaṭilavarman.¹ But these considerations are not as conclusive as one would wish. It must be noticed that while the smaller Sinnamanūr plates make Arikēsari the son of Jayantavarman, the Vēḷvikkuḍi grant is not so specific in the Tamil portion and indicates the relation between Arikēsari and Śēndan by the vague phrase ‘மற்றவர்க்குப் பழிப்பின்றி வழித்தோன்றி’,² and the Sanskrit portion of the Vēḷvikkuḍi grant gives no help here as it begins only with Arikēsari. After some discussion, it has become clear that the Madras Museum plates of Jaṭilavarman too must be ascribed to Parāntaka Neḍunjaḍayan of the Vēḷvikkuḍi grant. Palæographical considerations which were once held very strongly to militate against such a view have not stood the test of further criticism.¹ On the other hand, several weighty considerations can be urged in support of the view now taken of this grant. *First* several surnames of the ruling king occur in both

c8.19

c8.20

² Mr. Krishna Sastri gets over the difficulty by saying வழித்தோன்றல் means ‘son’; I am afraid that this gives no help in interpreting the expression quoted which is not வழித்தோன்றலாகி, but வழித்தோன்றி. Nor is Mr. K. G. Sankara’s suggestion that Arikēsari was the son of a daughter of Śēndan easy to accept. If, as is not unlikely, the smaller Sinnamanūr plates are earlier than the Vēḷvikkuḍi grant and belong to the reign of Raṇadhīra, then their evidence must be preferred to that of the Vēḷvikkuḍi grant. But after all, it may be that the Vēḷvikkuḍi grant itself is not in conflict with the Sinnamanūr plate. For some reason the Sanskrit part of the Vēḷvikkuḍi grant begins only with Arikēsari and the Tamil part, for the same reason, may be taken to make a fresh start with this king, Arikēsari, though he was the son of the immediately preceding Śēndan (E.I., vol. xvii, p. 365, n. 5 and K. G. Sankara, I.A., vol. li, p. 213 and Q.J.M.S., vol. x, p. 178). It may be noted in passing that nothing has turned up to justify Mr. Venkayya’s suspicion that the kings of the smaller Sinnamanūr and the Madras Museum plates may not belong to the main line of the Pāndyas (A. R. E., 1908, p. 55).

c8.21

¹ See *A.R.E.*, 1908, pp. 50 and 55 and H. K. Sastri in *E.I.*, vol. xvii, p. 293. Mr. Venkayya said, ‘The characters of the Sanskrit portion (of the Veḷvikkuḍi grant) are older than those of the Madras Museum plates of Jaṭilavarman and of the two Śinnamanūr ones.’ Again, ‘the numerals which are marked in the Veḷvikkuḍi grant are very old, while those of the Jaṭilavarman plates bear a close resemblance to the corresponding symbols used in the larger Śinnamanūr plates.’ Mr. Krishna Sastri in editing the Veḷvikkuḍi grant enumerates the differences in the Grantha (Sanskrit) portions of the two sets of plates; but a careful study of them does not seem to justify his conclusion that the Tamil portion in both the grants formed the original ‘grants proper in both’ cases, and ‘the insertion of the Grantha portion in the Veḷvikkuḍi grant might have been somewhat earlier than that in the Madras Museum plates.’ The only proof for this statement that is furnished by Mr. Sastri is in ‘the remark that the Sanskrit portion (of the Veḷvikkuḍi grant), by its brief notice and the very meagre historical material which it supplies in the form of a general introduction, could not have been contemporaneous with the Tamil portion.’ As against these statements it may be pointed out that (1) a glance at the Veḷvikkuḍi and Madras Museum grants (in the plates published in the *E.I.* and the *I.A.*) does not support the view that the Sanskrit parts are later insertions; (2) the differences in the Sanskrit portion enumerated by Mr. Sastri are not serious and may be only due to the facts (a) that different scribes were employed to engrave the two sets of plates and (b) there is an interval of at least fourteen years between them so that in any case the Sanskrit portion of the Veḷvikkuḍi grant will be ‘somewhat earlier than that in the Madras Museum plates’; (3) without the Sanskrit portion, the Veḷvikkuḍi grant will begin with கொல்யானை and the Madras Museum plates with அண்ணணகிய, and we may take it that no inscriptions could have begun so inauspiciously or so abruptly; (4) the meagreness of the historical particulars in the Sanskrit part is only to be expected because that part was intended only to be a general introduction to be read together with the Tamil portion; it should be noted there is not merely a Sanskrit introduction but also a Sanskrit epilogue to the two sets of plates; (5) lastly, arguments from palæography are always open to doubt and much more so in the case of early South Indian palæography owing to the scarcity of the records available for comparative study.

c8.22

c8.23

the grants. Such are Paṇḍitavatsalan, Vīrapurōgan, Vikramapāragan, Parāntakan, Neḍunjaḍayan and also Śrīvaran. The *biruda*, Kaṇṭakaniṣturan of the Vēḷvikkuḍi grant is echoed by the phrase கண்டக சோதனை தான் செய்து of the Museum plates. It is extremely unlikely that *all* these different *birudas* were common to two different kings. Secondly among the subdonees of the Vēḷvikkuḍi grant is a Mūrti Eyinan who is specially mentioned (l. 136). The *ājṇapti* of the Madras Museum plates is a Dhīrataran Mūrti Eyinan who was a *mahā-sāmanta* of the king. There is reason to think that these two references are to the same person who may have been, as has been suggested on the strength of the Ānamalai records, a brother of Mārangāri who was the *āṇatti* of the Vēḷvikkuḍi grant and the king's *uttara-mantri* when he excavated the Ānamalai cave.¹ Lastly the Vēḷvikkuḍi grant ascribes an important victory against the Pallavas to Parāntaka's father Māran, or Māravarman Rājasimha as he is called in the Sanskrit portion; and the Museum plates ascribe to Jaṭilavarman's father Māravarman the *biruda* Pallavabhanjana; and this surely may be taken to furnish a confirmation of the probability suggested by the two considerations urged above. It is thus clear that though there may be some scope for doubt as to the place of the smaller Śinnamanūr plates, it may be accepted as settled that the Madras Museum plates of Jaṭilavarman belong to the reign of the same king as the Vēḷvikkuḍi grant.

c8.24

¹ See *E.I.*, vol. xvii, pp. 295-6.

c8.25

We may now turn to the chronology of the period. The best starting-point is furnished by the Ānamalai inscription which is dated in the year 3871 (expired) of the Kaliyuga era = A.D. 770.¹ We thus get a definite date in the reign of Parāntaka I alias Varaguṇa Mahārāja (No. 7 in the table given above).² Another datum which is equally definite is derived from the Aivarmalai record (705 of 1905) which gives Saka 792 = 870 A.D. as the eighth regnal year of a king Varaguṇa who must have come to the throne in A.D. 862-3. This Varaguṇa must have been Varaguṇavarman (No. 9) the grandson of Varaguṇa Mahārāja. And this has been held to be a serious objection to the scheme of identifications on which we have based the genealogy of these rulers. It has been said that ‘ we have only one sovereign between Neḍuñjaḍayan of about A.D. 770 and Varaguṇavarman who ascended the throne in A.D. 862, and we are obliged to give him a reign of nearly one hundred years which is absurd on the very face of it.’³ But the situation is nothing so absurd; the interval is exactly ninety-two years; the Ānamalai record may be ascribed to an early regnal year of Varaguṇa I and the reigns of Varaguṇa and his successor Śrīmāra Śrīvallabha may have been exceptionally long yielding an average of forty-six years or a little more for each of the reigns. If a precedent is needed in support of this position, we may turn to Prof. Dubreuil who allows 113 years between the accession of Nandivarman Pallavamalla and Nandi of Tellāru separated by only one reign, viz. that of Dantivarman.¹ There is thus no serious chronological absurdity involved in the assumptions we have made regarding the identity of the kings in these records. It may be accepted that Rājasimha II (No. 11) the last king in our genealogical table was the Rājasimha Pāndya who was defeated by Parāntaka I Chola early in the tenth century A.D.² as this fits in well with the chronology of the age as derived from the Aivarmalai record. If we calculate backward from the Ānamalai record allowing, say, twenty-five years on the average for each generation we arrive at some date at the beginning of the seventh century A.D. for the accession of Kadungōn from whom our genealogy begins. Even the end of the sixth century may be accepted as possible. We thus see that the epigraphs we have been discussing at such length relate to the history of the three centuries from the beginning of the seventh century A.D. to the beginning of the tenth and this period we might call the Age of the First Empire. The period begins with a restoration and witnesses a rather wide extension of Pāndya power at the expense of the Pallavas who apparently had succeeded in dispossessing the Cholas of their ancestral dominions even before the Pāndya expansion began. It is marked by repeated contests between the Pāndyas and the Pallavas which is carried on right through these centuries with varying fortunes. It ends with the revival of Chola power under Vijayālaya and his successors and may be said to close with the attack delivered at the heart of the Pāndyan Empire by Parāntaka I, the grandson of Vijayālaya, who began his rule early in the tenth century A.D. It now remains for us to narrate in some detail the story of the age as we are able to reconstruct it, and then bring together such particulars about the social and political life of the period as we can gather from the contemporary records of the age. We shall attempt to do this presently, but, before doing so, we must give some consideration to the vexed question of the Kalabhra occupation of the Pāndyan country.

c8.26

¹ See *E.I.*, vol. viii, pp. 318 and 320.

c8.27

² It is rather strange that Mr. T. A. Gopinatha Rao (writing sometime in 1910-13) should ascribe this Ānamalai record to Kōccayan Raṇadhīra (No. 5 in our list), the grandson of Jayantavarman; this is simply impossible because the Māraṅgāri of the Ānamalai temple appears as the *ājñapti* of the Veḷvikuḍi grant in the fifth generation after Jayantavarman (Sēndan). After this start, he finds himself compelled (1) to ignore the Veḷvikuḍi grant in the learned discussion of the history of Mārañjaḍayan which follows, (2) to oppose on palæographical considerations Mr. Venkayya's suggestion that the Trivandrum Museum inscription which he edits must be ascribed to Varaguṇa Mahārāja, and (3) to enter upon an imaginary reconstruction of the Pāṇḍyas before Jayantavarman to arrive at the date of Gnānasambandar and his contemporary Neḍumāran in the middle of the seventh century A.D. (See *Trav. Arch. Series*, vol. i, pp. 153-7). But the only proper solution seems to be to identify Varaguṇa, Mārañjaḍayan and Parāntaka—a course which fits in with palæography and the place of Māraṅgāri in the king's reign as indicated by the Ānamalai and Veḷvikuḍi records.

c8.28

³ K. V. S. Aiyar, *Ancient Dekhan*, p. 103.

c8.29

¹ Dubreuil, *The Pallavas*, p. 66. See also T. A. Gopinatha Rao, *Trav. Arch. Series*, vol. i, p. 19, n. 5.

c8.30

² See *S.I.I.*, vol. ii, p. 383, verses 9-11 and vol. iii, p. 386.

c8.31

The Vēlvikkudi grant only says (ll. 39-40), ‘Then a Kali king named Kalabhran took possession of the extensive earth driving away numberless great kings (adhirājas)’ and tells us no more about it, although it refers to the Kalabhras in the plural and their brave ocean-like army (l. 111).¹ Mr. Krishna Sastri is inclined to accept the suggestion that Kali was the name of a dynasty of kings, Kalikula;² but nothing is known of such a dynasty yet and the mention of the Kalabhras in the plural in the Vēlvikkudi grant itself and elsewhere seems to point to a military tribe rather than a dynasty of rulers. We can only say that the Kalabhras overran the Pāndya country sometime after Mudukuḍumi’s time; how long after we cannot say. ‘How the Pāndyas were overcome by the Kalabhras, how long the sovereignty of the latter lasted and how they were driven back are points on which no information is at present forthcoming' (Venkayya). It may however be noted that the Pallava king Simhaviṣṇu who stands at the beginning of an important line of Pallava rulers, just as Kaḍungōn does in the case of the Pāndyas, and whose accession has been placed at *c.* A.D. 575 by Dubreuil, also claims to have conquered the Kaḷabhras; and the date we have arrived at for Kaḍungōn is the beginning of the seventh century A.D. or even the end of the sixth century. This raises the presumption that the Kaḷabhra occupation was a danger which threatened the independence of both the Pāndya and the Pallava dynasties and that these powers, either independently or in co-operation with each other, managed to throw off this incubus before they started on their long careers of expansion and success which so often brought them into collision with each other and lasted for close upon three centuries until the newly risen power of the Cholas of the Vijayālaya line gave them a check at the close of the ninth and beginning of the tenth centuries.¹ be justified.’ ‘Koccadyan’s victory was also over the Kalabhras.’ ‘The Kalabhras are Kamasas.’ He thinks also that the downfall of the Kalabhras was due to Simhavisnu and not to Kadungon and says ‘Perhaps before he had succeeded in settling the land, the Pandyan prince Kadungon came out from the dark and occupied his ancestral land.’ Mr. K. G. Sankara ingeniously suggests (*I. A.*, vol. li, p. 213) that the expression *Alavariya Adhirajarai* in the Velvikkudi grant means ‘countless Pandyas through their last representative’: but it seems simpler to make it refer to other rulers besides the Pandyas, and understand the phrase *Alavariya* as containing a rather natural exaggeration of the number of dynasties displaced.

c8.32

¹ The translation of ll. 39 and 40 is Mr. Krishna Sastri’s. He seems slightly to misunderstand the word துளக்கமில்லா in l. 111.

c8.33

² E.I., vol. xvii, p. 306, n. 2; contra, Hultzsch, E.I., vol. xviii, p. 260 who translates Kalikula into the people of the Kali age.

c8.34

¹ Mr. Venkayya made an attempt to clear up the story of the Kaḷabhra occupation from Mūrti Nāyanār Purāṇam in the *Periyapurāṇam* and the *Tiruvilaiyāḍal* (*A.R.E.*, 1908, p. 53). But there is nothing in the *Purāṇa* except the mention of a Karnātaka king of Jaina persuasion ruling in Madura which can connect it with this age. (See K. G. Sankara, *Q.J.M.S.*, vol. x, p. 178). Mūrti, not Kaḍungōn, succeeds the childless foreign ruler in the *Purāṇa*. Judging by results, Venkayya committed a more serious mistake in suggesting the identity of Kaḷabhra with the Erumaiyūran, one of the opponents of the Pāndyan king in the battle of Talaiyālangānam, and saying, 'it may be that Neḍuñjeḷiyan drove out from the Pāndya country the Kaḷabhras' (*A.R.E.*, 1908, p. 53). This has led others to identify Neḍuñjeḷiyan with Jayantavarman, as we have seen before. Mr. T. N. Subramanian (*Q.J.M.S.*, vol. xii, pp. 304-6) makes a number of statements which are not easy to support from the evidence at our disposal. He says 'From the analogy (in the Vēlvikkudi grant) it appears that the Pāndya line was unknown to the world while the Kaḷabhras ruled there. Thus the evidence of the *Periyapurāṇam* that there was no Pāndya prince (!) left to succeed when a Kaḷabhra king died might

c8.35

It is not easy to follow what Mr. T. A. Gopinatha Rao says in a footnote (n. 1 at p. 49 of *E. I.* xv) to his edition of the Anbil plates of Sundara Chola At any rate, it may be pointed out that the transition from Kalvara (which is a form assumed by Mr. Rao for the word Kalvar) to Kalabhra seems impossible; (the analogy Valavan—Valabha is not equal) and that Suvaran Maran could not have been at once a contemporary of Nandivarman Pallavamalla and the Kalabhra king of the Velvikkudi grant if we follow the chronology of the Pallavas that has been established by modern research. At the same time it must be observed that Dr. S. K. Aiyangar derives Kalabhra from Kalavara through Kanarese Kalabharu and holds that the southern invasion of the Kalabhras was due to the expansion of Satavahana power. (His *Univ. Lectures on the Sangam Age and Pandya Charters*; also R. Gopalan, *The Pallavas of Kanchi*, n. at p. 85.)

c8.36

Yet another, and a very plausible, suggestion is made by Pandit M. Raghava Aiyangar, in his forthcoming work on *Epigraphy and Tamil Literature*. He brings together much recondite evidence from classical Tamil Literature and seeks to establish: (1) that the Kalabhra king who displaced the ancient rulers of the Tamil land was Acyuta who perhaps ruled from Chidambaram and was probably identical with the Acyuta referred to by Buddhadatta; (2) that the expression *Kaliarasan* of the Velvikkudi grant applied to the Kalabhra King taken along with the story of Kurruva Nayanar as given by Sekkilar may lead one further to identify the Kalabhra King with Kurruva Nayanar mentioned by Sundaramurti in his *Tiruttondattogai*; and (3) that the Kalabhras figure in Tamil literature as Kalappar or Kalappalar and were akin to the Vellalas. It seems very likely that further study on the lines indicated by the learned Pandit will yield results of great value for the history of the period before the accession of Kadungon.

Linked texts (6)

பெரிய புராணம் — Cēkkiḻār, Tamil
Dhanaidaha Plate, Year 113 GE — Sanskrit
Vēḷvikkuṭi grant, time of Neṭuñcaṭaiyaṉ, year 3 — Tamil
Indian Antiquary
South Indian Inscriptions
Sundhā Hill Inscription of Chāchigadēva; [Vikrama-]Saṁvat 1319 — Sanskrit