Pramāṇa · Search · About · API

History of the Pallavas of Kanchi c7: Chapter II. Theories of the Origin of the Pallavas: Pre-Pallava History of Kāñchīpura

History of the Pallavas of Kanchi c7

Language
English

Contents

Chapter II. Theories of the Origin of the Pallavas: Pre-Pallava History of Kāñchīpura

c7.2

THE age of the Pallavas in South India may be taken to extend roughly from the third century A.D. to the end of the ninth. Many facts connected with their origin and early history are, in the existing state of historical research, uncertain. Their chronology is yet in the process of construction, and the genealogical order of succession, in spite of the comparative abundance of copper-plate characters, cannot be regarded as having been finally settled. The question of their origin has baffled investigators, European and Indian alike. Ingenious suggestions and theories have been offered for the solution of this enigma by various scholars who occupied themselves with this work.

Various Theories

c7.s1.1

The theory that held the field until recently, almost undisputed, is known as the Persian or Parthian origin of the Pallavas, and was adopted by the late Vincent Smith in the first edition of his *Early History of India*.¹ According to the late Mr. Venkayya, who developed this theory elaborately :—

c7.s1.2

¹ *E.H.I.* (1904), p. 348.

c7.s1.3

‘The Pallavas . . . may, until their origin is satisfactorily established by indisputable evidence, be supposed to be identical with the Pahlavas, Palhavas, and the Pahnavas of the Purāṇas. This identification is based on etymological grounds and supported by the fact that the Palhavas formed a distinct element in the population of Western India early in the second century. Their movement from Western India to the

c7.s1.4

East Coast is not only possible but rendered likely by known historical facts.'¹

c7.s1.5

¹ A.S.R., 1906-7, p. 221.

c7.s1.6

This theory, however fascinating to the imagination, rests altogether on the superficial verbal resemblance of the words Pahlava and Pallava.² There is no evidence whatever, as Mr. Venkayya himself admits, that the Pahlavas of the west coast moved into the east anytime during the second century A.D. At any rate there is no reference or reminiscence whatever in the large numbers of Pallava copper-plate inscriptions of any such migration. Mr. Venkayya presumes that it was the war of the Western Satrap Rudradāman, A.D. 150, with the Andhra king Gautamīputra-Śātakarṇi and his successor, that probably brought about in some manner, now unknown, the eastward movement of the Pahlavas in his service. It is indeed true that the minister of Rudradāman is called Suviśāka and described in the Junagadh inscription as a Pahlava. But no evidence of a positive character is available to connect him or his descendants with the Pallava kings of Southern India. For one thing this Pahlava minister of Rudradāman was not a ruling king. He is not moreover mentioned in any of the ancestral genealogical lists of the Pallava kings. In the second edition of his *Early History of India* Vincent Smith gave up this theory of Pahlava origin which he had adopted in the earlier edition of *Early History of India*, with the remark that it is more likely that the Pallavas were a tribe, clan, or caste which was formed in the northern part of the Madras Presidency, possibly in the Vengi country.³ Mr. Venkayya's objection against this view of indigenous origin of Pallavas is the apparent improbability of two tribes, the Pahlava in Western India and the Pallava in the delta of the Godavari, bearing the same name. Some of his other arguments against the indigenous origin of the Pallavas are ingenious and may be stated here. One is that the Pallavas never refer to themselves in their records as belonging either to lunar or solar race like other dynasties of South India such as the Chōḷas, Pāṇḍyas, Chāḷukyas, etc. Another is that none of the kings mentioned in the Purāṇas are to be found in the Pallava genealogy such as Manu, Śibi, Ikshvāku, etc. All these facts according to him, raise a presumption that the Pallavas of Southern India were not an indigenous tribe in the sense that the Chōḷas, the Pāṇḍyas and the Cheras were.¹

c7.s1.7

² It is to be noted however that in Sanskrit orthography Pahlava with a *h* can never be confounded with the word Pallava. In the Purāṇas, as well as in Persian and even in Prakrit the *h* is distinct. The two words appear to indicate two distinct people.

c7.s1.8

³ V. A. Smith, *E.H.I.* (1908 ed.), p. 423.

c7.s1.9

¹ A.S.R., 1906-7, p. 219, foot-note.

c7.s1.10

*Views of L. Rice.*—This view of foreign origin was also shared by L. Rice. In his *Mysore and Coorg from Inscriptions* he points out that the Pallava kings of South India were identical with the Pahlavas who, with the Yavanas and Śakas, were destroyed by Gautamīputra-Śātakarṇi and that the word Pahlava is a Prakṛit form of Parthava meaning Parthian, especially the Arsacidan Parthians.² As we have already pointed out this conjecture that the Pallavas of the south are identical with the Pahlavas of the west and ultimately with the Parthians and the Persians is based entirely on the similarity of the words Pahlava and Pallava, whereas everything known about the Pallava kings of the south, inscriptions, traditions and monuments indicate that they were a peninsular race; they are probably identical with the class of people described as Toṇḍaiyar in Tamil literature and also familiar as Kāḍavas or foresters. Vincent Smith totally rejected this theory of foreign origin in the third edition of his *Early History of India*, and has expressed his conclusion on the subject in the following terms:—

c7.s1.11

² *Mysore and Coorg from Inscriptions*, p. 53.

c7.s1.12

‘The name Pallava resembles Pahlava so closely that most writers have been disposed to favour the hypothesis that the Pallavas and the Pahlavas were identical, and that consequently the southern Pallava dynasty of Kāñchī should be considered as ultimately of Persian origin. But recent research has failed to adduce any historical facts in support of

c7.s1.13

that notion, and it seems more likely that the Pallavas were an indigenous tribe, clan, or caste.’¹

c7.s1.14

¹ *E. H. I.* (third edition), p. 469.

c7.s1.15

One important piece of evidence among others ² which strongly militates against this theory of the foreign origin is that furnished by the poet Rājaśēkhara, the great critic, in his account of the geographical division of the people of his times.

c7.s1.16

² Cf. C. V. Vaidya’s views, *Mediæval Hindu India*, vol. i, p. 280.

c7.s1.17

This is brought out in full relief by Dr. S. Krishnaswami Aiyangar in his dissertation on the Pallavas, wherein discussing the origin of the Pallavas he observes as follows :—

c7.s1.18

*Rājaśēkhara on the Pallavas.*—‘The words “Pahlava” and “Pallava” are philologically one, but we have good authority for taking it that the two terms refer historically to different peoples, thereby illustrating that the partition that separates philology from history is not always very thin. The poet Rājaśēkhara lived in the courts of the Gūrjara sovereigns Mahendrapala and Mahīpāla about the end of the ninth and the commencement of the tenth centuries A.D. just the period when the Pallavas were passing out of existence as the dominant South Indian power. He is the author of a geographical work named *Bhuvanakōśa* to which he actually refers for further information in ch. xvii of his *Kāvyamīmāṃsa*. In this work he makes a division of India into five, and allots to each division, the peoples, towns and rivers that belonged to it. In that section he allots the Pallavas to the southern division, or Dakshiṇāpatha, lying beyond Mahishmatī, while the Pahlavas he allots to the division Uttarapātha lying beyond Prithūdaka.³ This last name the “great water” probably stands for the Indus, and the people Pahlava are found associated with the Huns, Kambhoja, Bahlika, etc. In the estimation, therefore, of Rājaśēkhara who seems a much travelled man, and has really much interesting information to give of different parts of India in respect at any rate of the cultivation of Sanskrit learning, the Pahlavas and the Pallavas were distinct peoples, one of them belonging to the south and the other to the frontier on the other side of the Indus. Rājaśēkhara, no doubt is a late authority, but undoubtedly gives expression to the prevailing opinion of his time in regard to these two peoples.'¹

c7.s1.19

³ This has been since satisfactorily identified with Pehoa in the Panipet District, and this is in complete accord with the details given in Rājaśēkhara.—*Ed.*

c7.s1.20

¹ J. I. H., vol. ii. pp. 25-26.

c7.s1.21

Prof. Jouveau-Dubreuil's Views.—The view held by Prof. Jouveau-Dubreuil is only slightly different. Although he does not go the whole extent of advocating the theory of a Parthian origin of the Pallavas he is inclined to connect the early Pallava kings of Conjeevaram with the Pahlavas of Western India. We read for instance in his chapter on the Pallavas in his latest work, *The Ancient History of the Deccan*.

c7.s1.22

‘Up to the present day we have not found any document which proves a Parthian invasion of Southern India. We may say that this theory makes the Pallava mystery still more mysterious.’²

c7.s1.23

² *Ancient History of the Deccan*, p. 47.

c7.s1.24

According to him the Pallavas were alien to South India and the early members of this dynasty were not kings. They were neighbours of the Nāgas and Chūṭu-Nagas in the days of Rudradāman in the second century. The first Pallava king inherited the throne of Kāñchī on marrying the daughter of a Nāga chief by name Śivaskanda-Nāga. On pages 55-56 he elaborates this view and establishes a connection between the Pahlavas of the west coast with the Pallavas of Kāñchī. On page 55 of this work he says :—

c7.s1.25

‘. . . We learn from the Girnar inscription that Suviśāka who was the minister of Rudradāman in A.D. 150 was a Pahlava.’³

c7.s1.26

³ Junagadh Insc., *Ep. Ind.*, vol. viii, p. 37, plate line 19.

c7.s1.27

Thus the word which was written Palhava in A.D. 78 was written Pahlava in A.D. 150 and we note when we see the Girnar inscription, that the compound letter ‘hla’, resembles the double ‘ll’. Besides in Mayidavolu plates written in Prakrit this question of letters has been solved in a very simple manner, they wrote Pālava. This can correspond to Pallava for in Prakrit the consonants are not doubled. In the Hirahadagalli plates it is written as Pallava though they are also in Prakrit. Later on when the documents were written in Sanskrit the word Pallava would be understood to mean ‘sprout’.

c7.s1.28

From all this the following conclusion is drawn:—

c7.s1.29

‘The identity of names leads us to think that the ancient kings of Kāñchī belonged to the same family as the minister of Rudradāman. He lived in A.D. 150 and we know Pallava kings of Kāñchī reigning about A.D. 225.’

c7.s1.30

*Objections to the Theory.*—This theory which connects the Pallavas of Kāñchī with the Pahlavas of the west coast presents difficulties, some of which have already been pointed out before. In the first place it was seen that there is no evidence whatsoever of an eastward movement of the Pahlavas. There is moreover no evidence of a Pahlava prince marrying the daughter of a Nāga chief on which Prof. Jouveau-Dubreuil relies for his theory of the attainment of political sovereignty by the Pahlavas. The Vēlurpāḷayam plates which mention a Nāga alliance¹ do not say that it was the first Pallava king who so attained power. In this record which belongs to the ninth century, he is called Virakūrcha and not Śivaskaṇḍa. Then again this record mentions many earlier kings such as Aśokavarman who ruled before the days of Virakūrcha a fact which militates strongly against the view of Prof. Jouveau-Dubreuil. The theory does not explain why they came to be called Pallavas. It fails to take into consideration the existence of a people known as Toṇḍaiyar, synonymous with the Sanskrit Pallava. Moreover early Tamil literature does not support the view that the Pallavas originally emigrated from Surāṣṭra. Finally there is no mention in any of the numerous charters of the Pallava kings and other contemporary kings in South India of Suviśāka the supposed ancestor of the Pallava kings who lived in Rudradāman’s court. If he was really connected with the Pallava family of Kāñchī how is it that he does not figure even in one of the lists of their ancestors contained in the Pallava records? It is not again definitely known whether the authority of Śivaskaṇḍanāga extended as far south as Toṇḍaimaṇḍalam. No inscription or copper-plate of this chief or others have been found proving that his authority extended in this region. These considerations make us hesitate to accept the theory of Prof. Jouveau-Dubreuil that the kings of Conjeeveram were descended from the Pahlavas of Rudradāman’s time, and that the first Pallava king obtained the throne of Kāñchī by a marriage alliance with the Nāgas.

c7.s1.31

¹ S.I.I., vol. ii, Part 5, p. 510, verse 6.

c7.s1.32

Yet another View.—Mr. Rasanayagam who does not agree with the views and conclusions arrived at by the Professor seeks to find the home of Pallavas in the south in the region known as Maṇipallavam to the early Tamils.¹ According to him Kiḷḷi’s son Iḷam Tiraiyan through the Nāga princess mentioned in the early Tamil work Maṇimēkhalai was really the first Pallava king. In this work he is described as the son of Kiḷḷi by Pilivaḷai, a daughter of Vaḷaivaṇan, the Nāga king of Maṇipallavam. He was lost in a ship-wreck and was found washed ashore with a coil round his ankle of a twig of the Toṇḍai-creeper (Caphallandra Indica). Hence his name Toṇḍamān-Iḷam-Tiraiyan. Iḷam-Tiraiyan was therefore the first to become an independent king and progenitor of the Pallava dynasty. His period is fixed as the third quarter of the second century A.D. The dynasty was called the Pallava designated after the mother’s native place Maṇipallavam. The titles, such as Ankura, and Pōttaraiyan which were assumed by Pallava kings are synonymous with the word Pallava and remind us of the origin of the name from Maṇipallavam.

c7.s1.33

¹ Ind. Ant., vol. lii, pp. 75-80.

c7.s1.34

Consideration of the Theory.—According to this theory the Pallavas were of Chōḷa-Nāga origin and were indigenous to the southern part of the peninsula and Ceylon. They had nothing to do with Western India, Persia, the Chūṭu-Nāgas, Suviśāka and others. But there are certain difficulties in accepting this view as correct. The earliest members of the Pallava family that we know from copper-plate inscriptions do not include either Toṇḍamān-Iḷam-Tiraiyan or any member of the Chōḷa-Nāga dynasty. On the other hand, they describe themselves as members of the Bhāradvāja-gotra and as the performers of Aśvamēdha, Vājapēya and other sacrifices.

c7.s1.35

Moreover they addressed their charters invariably in Prakrit and Sanskrit, and never in the Tamil language. Such of the charters that contain any allusion to the Nāga marriage mention neither a Chōḷa ancestry nor Maṇipallavam their alleged original home. Their records on the contrary show that they are different from the southern Tamil rulers, and hailed from the north as representatives of Aryan or Sanskrit culture. The hostile relationship of the early Pallavas with the early Tamil rulers, one of the Pallava kings being called the ‘submarine-fire to the oceanlike army of the Chōḷas,’ ¹ seems to be opposed to this Chōḷa-Nāga origin of the Pallava dynasty. For these reasons it looks as if that Toṇḍamān-Iḷam-Tiraiyan had no connection with the Pallava family of Kāñchīpura known to us from charters, and probably was a Tiraiyar-chief ² who preceded the Pallavas and was a subordinate of the contemporary Chōḷa ruler in the second century A.D.

c7.s1.36

¹ S.I.I., vol. ii, Part 5, p. 510.

c7.s1.37

² He is actually described as given the name of Tiraiyan because he was brought over by the waves of the sea. This seems but a learned attempt at explaining an obscure folk-name Tiraiyar.—Ed.

c7.s1.38

*Other Views.*—A more or less similar view is expressed by Mr. M. S. Ramaswami Aiyangar in his recent work *Studies in South Indian Jainism*.³ According to him the early Pallavas, especially those kings in the fourth century A.D., were called Tiraiyar only and not as Pallavas. In support of this view he mentions the reference ⁴ to Viṣṇugōpa in Samudragupta’s Allahabad pillar inscription as king of Kāñchī only (*Kāñchēyaka Viṣṇugōpa*) and not as the king of the Pallavas.

c7.s1.39

³ *Studies in South Indian Jainism*, p. 143.

c7.s1.40

⁴ ‘The Allahabad pillar inscription of Samudragupta merely speaks of Viṣṇugōpa as the ruler of Kāñchī. The name Pallava does not occur there. From this it is evident that in the fourth century A.D. the word Pallava was little used by them. . . . In Sangam literature the rulers of Kāñchī are spoken of as Tiraiyan and Toṇḍaimān. . . . Viṣṇugōpa, who was ruling Kāñchī at the time of Samudragupta’s invasion must therefore be a Tiraiyan. “The *Toṇḍamaṇḍalapaṭṭayam* gives an account of the various branches of the Tiraiyar. This has been noted by Kanakasabhai Pillai in his book, *Tamils 1800 Years Ago*. As one of the main branches of Tiraiyar we have Pallava Tiraiyar showing thereby the connection between Pallavar with Tiraiyar. It is therefore natural to suppose that the Pallavas were known to the early Sangam literature by their group name of Tiraiyar but as their power and influence increased in the land

c7.s1.41

This explains according to him that in the fourth century A.D. the word Pallava was little used by them. Another ingenious suggestion made on similar lines is that of Mr. Swaminatha Aiyar, based upon the apparent similarity of the words Tiraiya, Draya and Darya in Persian. ‘Draya’ and ‘Darya’ according to this writer mean the sea in Persian, and Tiraiyar is apparently a Persian-derived name meaning seamen. He would thus ascribe to the Tiraiyar a Persian or Pahlava origin. As the fact seems to accord with the Persian origin of the Pallavas held by certain scholars it was thought by the author as giving the most satisfactory solution of the question. But as was already stated, the Persian origin of the Pallavas rests on the doubtful philological resemblance of words to which the addition of Tiraiya and Draya are now sought to be added. Apart from the resemblance of the words, is there anything positive which supports the identification of these different people Tiraiyar and the Pallavas? The evidence at our disposal on the other hand is wholly against it, and in the existing state of research it is not possible to regard the identification as established. The Pallava kings of the copper-plate charters must therefore be looked upon as different from the Tiraiyar, and the early members of the family such as Śivaskaṇdavarman, Budhavarman, etc., have very little to do with the kings and chiefs known as Tiraiyan and Iḷam-Tiraiyan figuring in early Tamil literature. Another consideration which militates strongly

c7.s1.42

their branch name Pallava-Tiraiyar assumed greater importance. Hence the absence of the word Pallava in the fourth and the fifth century literature and the greater frequency with which this name occurs in such works as *Nandikkalambakam*.”

c7.s1.43

There is no evidence to prove that the Tiraiyar of the early Tamil literature were the same as the Pallavas, and the testimony of *Toṇḍamaṇḍalapaṭṭayam* is invalidated as the work is only a work of comparatively modern times. It is not correct to say that in the fourth century the word Pallava was little used by them. In the copper-plate charters of the fourth century as well as the following centuries they are called only as Pallavas and never as Tiraiyan. If Viṣṇugōpa was a Tiraiyan it is strange that he is not mentioned as such in the Allahabad pillar inscription.’

c7.s1.44

against the identification of the Tiraiyar chiefs and the Pallava kings is the attitude of these kings towards Tamil literature and poets. While we find, that the chiefs of the Tiraiyar like Toṇḍamān-Iḷam-Tiraiyan not only encouraged and fostered Tamil poets but are actually reputed to have been poets themselves. We do not find that any of the early Pallavas were similarly devoted to the encouragement of Tamil. That Toṇḍamān-Iḷam-Tiraiyan was himself a poet is evident from Puraṉāṉūṟu 185, and verses 94, 96 and 106 of Naṟṟiṇai. The explanation is that the Pallavas were essentially a northern rather than a Tamilian race, steeped in Aryan or Brahmanic culture which through their encouragement took the leading place in the Tamil country. They were aliens to the Tamil land in this sense that their culture and outlook were totally different from those of the Tamils, and hence did not give themselves to actively patronizing Tamil literature like the Tamil kings of the south. Throughout their history we find them almost exclusively encouraging Sanskrit culture and literature. With the exception of the Āḻvārs and Aḍiyārs and the author of Nandikkalambakam, no Tamil poet appears to have enjoyed the patronage of the Pallavas. On the other hand, Sanskrit was actively fostered at the Pallava court, and a number of Sanskrit poets found hospitable reception in the Pallava court such as Bhāravi, Daṇḍin, etc., and some of the kings themselves were noteworthy writers in Sanskrit.¹ When all these considerations are taken into account it becomes difficult to sustain the identity of the Tiraiyar chiefs with the Pallava kings.

c7.s1.45

¹ It has to be noted in this context that Kāñchī was famous as a centre of Sanskrit learning and was noted as ‘a Ghaṭikā of the Brahmans’ in the Velūrpālayam Plates. It was a notable centre of Sanskrit learning when the Kadambas were coming into the view of history as in the story of Mayūra Śarman.—Ed.

c7.s1.46

*Kurumba-Pallava identification.*—Certain writers are inclined to consider that the Pallavas were the same as the people described as Kuṟumbar figuring in some of the accounts of the Mackenzie manuscripts. As regards these people there is a great deal of vague and misleading ideas in the minds of most writers. Who are the Kurumbas? According to one of the Tamil *Nigaṇṭus*¹ certain hill chiefs known as *Kuṟunila-mannar* were known as the Kurumbas. But in the Mackenzie accounts these people are described as the earliest inhabitants of Toṇḍamaṇḍalam and said to have been given the name for their cruel ways. They are also in some of the accounts believed to have been responsible for the division of the region into twenty-four *koṭṭams*, most of which figure in Pallava and Chōḷa inscriptions. The story of the conquest of the region by Ādoṇḍaichōḷa, which will be described later on, is believed by Sir Walter Elliot, Sewell and others to have been accomplished in the eleventh century in the days of Kulōttunga Chōḷa, a supposition for which there is no evidence. It is believed by some that these Kurumbas were the same as the Pallavas. But there seems to be absolutely nothing which supports such a notion. There is no reference in the numerous inscriptions of the Tamil land that there was at any time a dynasty of the Kurumbas, or that the Pallava kings are identical with the Kurumbas. Hence it appears to be preposterous to accept the identification.

c7.s1.47

¹ *Divākaram.*

c7.s1.48

*Still another View.*—According to Alexander Rea the references making the Pallavas ‘hold the sprout’ are pedantic Pandit’s puns, or a play upon the double meaning of the word ‘Pallava’. ‘I would suggest,’ he says, ‘that it might be derived by somewhat free translation from the two Tamil words *Pal* and *Avil* (to pull)? referring to their being agriculturists or milk-drawers like the Gopalas of the northern Dekkan.’² He also throws out another suggestion: ‘the masculine singular termination in Tamil is *avan*, and the plural *avar*; now prefixing the word *Pal* (milk), would give Pālavan; or Pālavar as the honorific plural. Pālavan in that case would mean milkman.’ Both these suggestions, however appear more fanciful than possible as the derivations have not even grammatical justification in their favour.

c7.s1.49

² Rea, *Pallava Architecture*, p. 2, foot-note 5.

c7.s1.50

In the present state of research definite conclusions on the question seem unattainable. It is, however, quite plain that the theory of foreign origin of the Pallavas, especially that which gives them a Parthian or Persian home, is without foundation. The theory that contents itself with connecting the Pahlava minister with the Pallavas of Kāñchī although having an air of plausibility is hardly acceptable as no evidence is forthcoming of the eastward migration of the Pahlavas in the second century. The next alternative is to regard them as indigenous to the country. In doing so we have to establish first whether the Pallavas should be regarded as a family or a tribe. Those who contend that the modern Kallar and the Pallis are descended from the Pallavas are labouring under the impression that the Pallavas constituted a big tribe; there seems to be little evidence for the assumption. In the Pallava inscriptions themselves we scarcely come across with the word Kurumba as a dynastic surname of any of the Pallava kings. It is therefore improbable that the identification of the Kurumbas with the Pallavas postulated by certain scholars rests on sound foundation.

c7.s1.51

That the rule of the Pallavas of Kāñchī was preceded by that of the Chōḷas is made clear not only by the evidence of Tamil literature, but also from the testimony of the Mackenzie manuscript accounts to which reference had been made already. These accounts although belonging to the early part of the nineteenth century, being the traditional account furnished by the village elders in the days of Col. Colin Mackenzie, profess to relate the early history of the region known as Toṇḍamaṇḍalam which corresponded to the region ruled over by the Pallava kings. ‘According to the *Ancient History of Toṇḍamaṇḍalam*, one of the accounts in the Mackenzie collection, the region was occupied by the Kurumbas who are said to have been responsible for the erection of the twenty-four forts, and the division of the region into twenty-four districts. These are also said to have carried on a sea-trade with the merchants of Kāvēripaṭṭinam. It is from these Kurumbas that one Ādoṇḍachakravarti, said to be an illegitimate son of the contemporary Chōḷa king, is described as having conquered the region and acquired it for the Chōḷas setting up his own rule at Kāñchī after uprooting the Kurumbas. The region which was till then known as Kurumbabhūmi is said to have been renamed Toṇḍamaṇḍalam atter the name of the conquering Chōḷa prince. If we properly interpret the traditional account in the light of the facts regarding the occupation of the region around Kāñchīpuram by Iḷam-Tiraiyan, it is easy to understand that in this account we find preserved the earliest reminiscence of the conquest of the region around Conjeevaram by the Chōḷas. When did this happen? Early scholars who noticed the account have been misled into thinking that this conquest was made in the eleventh century.¹ They have been led to think so by the reference that the contemporary Chōḷa king in these accounts was Kulōttunga, and there was a Kulōttunga ruling in the eleventh century. But all the other details in the account are against this date. The conditions of the territory and the state of civilization at the time of Ādoṇḍaichōḷa, the first Chōḷa representative, are too primitive to apply to a period like the eleventh century by which time the region had become fully civilized under the Pallavas who are known to have beautified Toṇḍamaṇḍalam, and whose sway over the region had already lasted for eight centuries. It looks therefore probable that the traditional account has mistaken the actual name of the contemporary Chōḷa king at the time of the subjugation of the Kurumbas which must have been made by some member of the family of the early Chōḷa Karikāla in the second century A.D.

c7.s1.52

¹ Sewell, *Lists*, vol. i, p. 177.

c7.s1.53

Pre-Pallava History of Kāñchī.—Kāñchī which is one of the seven great holy cities of India is believed to have had a history much anterior to the beginning of the Christian era. From the Chinese pilgrim Hiuen Tsang we learn that in the days of Buddha in the fifth century B.C. he frequented the region around Kāñchīpuram and made conversions. From the same source we learn also that Aśokarāja built several stupas none of which has survived to the present day. It would be remembered that Dharmapāla, the celebrated Buddhist savant and scholar at Nalanda, hailed from Kāñchīpuram. Whether Buddhism made any progress at Kāñchī at the time of Buddha is doubtful as there is not much evidence that Buddhism had advanced into the heart of the Tamil land in this period. Whether Aśoka ordered stupas to be built at the city of Kāñchīpura or not, it is not possible to say. The presumption that he might have done so gains strength in view of the discovery of Aśoka's inscriptions as far south as Siddāpura in the region of Mysore, and from the fact that his inscriptions mention the Chōḷas, the Pāṇḍyas and the Kēraḷaputras. At any rate Hiuen Tsang says that he noticed a stupa about 100 feet high built by Aśokarāja.¹

c7.s1.54

¹ Beal, *Rec.* II, p. 230.

c7.s1.55

During the reign of Karikāla who has been assigned, with acceptable authority by several scholars to the age of the Tamil Śangam in the beginning of the Christian era, Kāñchīpuram appears to have been one of the northern outposts of the kingdom. At any rate it is probable that one of Karikāla's exploits, or those of his generals, must have been the transfer of power from the Āndhra-Śātavāhanas, whose ascendency in this region is indicated by the discovery of ship-coins in profusion in the region between the Pennar and Pālar into the hands of the Chōḷas. At any rate in the period immediately following the city of Kāñchī is found in the possession of Toṇḍamān-Iḷam-Tiraiyan as stated in the *Perumbāṇāṟṟuppaḍai* of Rudran Kaṇṇanār. According to the Mackenzie MSS. one of the achievements of Aḍoṇḍachōḷa, by whom probably we are to understand Toṇḍamān-Iḷam-Tiraiyan, is the uprooting of the the Kurumbas by which term the unruly chiefs of the region were probably known. Under the rule of Toṇḍamān-Iḷam-Tiraiyan of Kāñchī, presumably as the Chōḷa viceroy, the city appears to have been in a flourishing condition. The *Perumbāṇāṟṟuppaḍai* of Rudran Kaṇṇanār, the contemporary of Toṇḍamān-Iḷam-Tiraiyan gives some idea of the city as it was in those days:—‘In the heart of the town were the Brahmān quarters, where neither dog nor the fowl could be seen. They were flanked on the one side by the fishermen's streets, and on the other by those of the traders, and these were surrounded by the Cheris of the Maḷḷas and the toddy-drawers. Then far removed from them were situated at one extremity of the city the paḷḷis of the Iḍaiyans, and beyond these lay the isolated paraichēris of the Eyinas. Next to the Maḷḷa streets were the temple of Tiruveḥkā and the palace of King Iḷam-Tiraiyan.’¹

c7.s1.56

¹ Tam. Stud., p. 76.

c7.s1.57

The history of Kāñchīpuram is lost in obscurity almost from the days of Karikāla to its occupation by the Pallava kings under Śivaskandavarman. It is probable that during this period it was in the hands of the Chōḷa princes some of whom are mentioned in the Manimēkhalai to have built Buddhist temples.² It is here mentioned that there were Buddhist chaityas in Kāñchī constructed by Kaḷarkilli and Iḷam-Kiḷḷi whose actual position in the Chōḷa family of this period is not known. It would be an interesting study to examine the evidence contained in the Tamil works regarding the connection of the early Chōḷa princes with Kāñchīpuram. There are several members of the Chōḷa family bearing the name of Kiḷḷi mentioned in early Tamil literature. One of these is Kuḷamurṟattu-tuñjina Kiḷḷivaḷavan, Kiḷḷi who died at Kuḷamurṟam, of whom many poets have sung. Then again Kovūr-Kiḷār mentions in Puṟam 44 and 45 a certain Kāriyaṟṟu-tuñjina (who fell or died at Kāri River) Neḍumkiḷḷi of Uraiyur and of his friend Iḷam-Tattan. Elsewhere we find the poet singing of Kurāpaḷḷi-tuñjina (who died at Kurāpaḷḷi) Kiḷḷivaḷavan,³ one of whose exploits was the destruction of Karuvūr. His Pāṇḍyan contemporary was apparently the Ugrapperuvaḷudi who died at Veḷḷiyambalam. Other pieces in the same collection speak of Soḷan Ilavandigaippaḷḷi-tuñjina Naḷangiḷḷi Śētchenni, Ēnādi (General) Tirukkiḷḷi, etc.

c7.s1.58

² See chap. 28, Account of Manimēkhalai's visit to Kāñchī.

c7.s1.59

³ Puṟam 373.

c7.s1.60

According to the account of *Maṇimēkhalai*, the Chōḷa King Veḷvēṛkiḷḷi, who is probably one of the Kiḷḷis enumerated above, married a Nāga princess called Pilivaḷai, the daughter of Vaḷai-Vaṇan, and became the father of the prince known afterwards as Toṇḍamān-Iḷam-Tiraiyan whom certain scholars look upon as the ancestor of the Pallavas of Kāñchī. This account although it seems to set up a connection between the Chōḷas and the Pallavas has not been corroborated by the evidence of the Pallava charters.

c7.s1.61

Śengaṇṇan.—Among the successors of the Chōḷa king Karikāla, one of the most important was Ko-Śengaṇṇan who is mentioned not only in the early, Tamil classics¹ but also in the *Tēvāram* hymns of the Śaiva saints Appar and Sambandar.²

c7.s1.62

¹ வெள்ளி வெண்ணஞ்சிலான் ஞால முழுவனபோல் எல்லாக் களிறு நிலஞ்சேர்ந்த பல்வேற் பிணை முழங்கு போர்த்தானைச் செங்கட் சினமால் கணைமாரி பெய்த களத்து— *Stanza from ‘Kaḷavaḷi.’*

c7.s1.63

² Appar, *Tev. Tirukurukkai Patikam*; Sambandar, *Tiruvānaikkāval Patikam.*

c7.s1.64

We learn from Tamil literature that he defeated his Chera enemy at a place called Kaḷumalam, took his enemy prisoner and confined him at a place called Kuḍavāyil-koṭṭam variously identified with Koḍavāśal and places in the Cochin State. Śengaṇṇan’s sway must have extended throughout Toṇḍamaṇḍalam. The *Kaḷavaḷi-Nāṛpadu*, a poem celebrating the powers of this king as a warrior, gives a picturesque account of the battle of Kaḷumalam and early Tamil modes of warfare. The age of Śengaṇṇan may be ascertained if we know the correct date of Poygaiyār who was his contemporary. According to certain scholars Poygaiyār is identical with Poygai-Āḷvār the Vaishṇava saint whose age is that of the third Śangam or somewhere near it. According to others Poygaiyār is assignable to the sixth or the seventh centuries A.D. This latter date is impossible if we bear in mind that Śengaṇṇan is mentioned already in the seventh century A.D.

c7.s1.65

by Appar as a well-known Śaiva saint. He is said to have worshipped at Tirunaraiyūr near Kumbhakonam by Tirumangai Āḻvār, the Vaiṣṇava saint who also notes that he built as many as seventy temples dedicated to Śiva. If Ko-Śengaṇṇan could become so popular as early as the seventh century to be mentioned by the Śaiva saints of the period in their hymns a considerable time must have elapsed between his period and the period of the Śaiva saints. It would not be wrong to assign him to the period succeeding the age of Karikāla (third century A.D.) and immediately preceding the advent of the Pallavas of Kāñchīpuram.

Linked texts (9)