Pramāṇa · Search · About · API

The Rashtrakutas and their Times c8: Chapter II. The Rise of the Rāṣṭrakūṭas: Indra I, Dantidurga, and Kṛṣṇa I

The Rāṣṭrakūṭas and their Times c8

Language
English

Contents

Chapter II. The Rise of the Rāṣṭrakūṭas: Indra I, Dantidurga, and Kṛṣṇa I

c8.1

The Rise of the Rāṣṭrakūṭas

c8.3

A bird's-eye-view of the political condition of India in general, and the Deccan in particular, in the first half of the 8th century would be necessary to understand properly the rise of the Rāṣṭrakūṭa house. The history of Kashmir need not be considered, for after the death of Lalitāditya Muktāpīḍa that kingdom did not come into contact with the kingdoms in the rest of India for a long time. We do not know who was occupying the imperial throne of Kanauj after Yaśovarman. It would appear from the *Chachnāmā*¹ that Rai Hari Chander was ruling there in c. 715 A.D., but this king is yet a mere name to us. The same is the case of Vajrāyudha, who is supposed to have been the predecessor of Chakrāyudha by some writers. Bengal was just recovering from anarchy under the leadership of Gopāla I. In the Deccan itself the Chālukya empire was powerful, but its strength was being wasted in hereditary wars with Pallavas. The precise extent of the Chālukya kingdom at the time of its overthrow is not easy to determine. Its southern boundary line must have been a changing one, as the wars with the Pallavas were being continuously fought with varying success. A subordinate branch, related by blood-relationship to the main line, was no doubt ruling at Vengi; but it had become practically independent of the main line at this time, though the imperial title Mahārājādhirāja Parames'vara Parama-bhaṭṭāraka was not assumed by its rulers before the time of Vijayāditya II. The northern boundary of the Chālukya kingdom was probably the Kim in southern Gujrat; beyond that river stretched the kingdom of the Gurjaras of Bharoch (or rather of Nāndīpurī, to be precise) which embraced the territory between the Kim and the Mahi. How much further east of the Kim valley the sway of the Chālukyas extended is difficult to determine, for the history of the Central Province of this period is still shrouded in mystery. Portions of this province must have passed under the Chālukya sway with the overthrow of the Kalachuri prince S'ankaragaṇa by Mangalis'a towards the end of the 6th century A.D., but whether they continued to be ruled by the Chālukyas by the beginning of the 8th century is not known. Portions of Berar were being ruled by the ancestors of Dantidurga; the Multai and Tivarkhed records² do not mention the name of the suzerain ruler, but it is almost certain that he must have been the Chālukya king upon the throne. From the Udayendiram plates of doubtful authenticity³ it would appear that there was a fairly powerful king in the Vindhya regions at this time named Pṛthvīvyāghra, who is claimed to have been defeated by the Pallava ruler Nandivarman. Who this Pṛthvīvyāghra was, and how far he had encroached upon the Chālukyan kingdom, we do not know. It is, however, possible that Pṛthvīvyāghra may have been another name of Jayavardhana of S'ailodbhava dynasty who, in his Ragholi plates, which on palæographical grounds have been assigned to the middle of the 8th century by Rai Bahadur Hiralal, (4) styles himself *Paramamāhes'vara Sakalavindhyādhipati Mahārājādhirāja Parames'vara*. Whether this identification is true or not, it is clear that the Chālukyas must have lost a portion of their dominions in the north by the rise of Jayavardhana. In the eastern Kos'ala, there was kingdom at Sirpur near Raipur, where king Udayana was ruling by the middle of the 8th century. In Rājputānā and Mālvā there ruled two Gurjara houses, one at Bhinmal and another at Ujjain. Both these, along with the kingdoms of Valabhi and Nandīpurī, were being considerably harassed by the raids of the Arabs of Sindh in the second quarter of the 8th century A.D.

c8.4

¹ Elliot, I, pp. 207-208.

c8.5

² I. A., XX, p. 100.

c8.6

³ I. A., VII, pp. 275 ff.

c8.7

Such then was the political condition in India, when Indra I began his political career. The Sanjan plates inform us that he had married the Chālukya princess Bhavanāgā by *rākshasa* form of marriage at Kaira.(5) This event must have taken place some time after 722 A.D. Kaira was then still under the rule of the Valabhi house, as the Gondal plates of Silāditya V show.(6) Very soon thereafter Kaira and Panchmahal districts were lost by the Valabhi kingdom; no record belonging to that dynasty comes from this area during the next 29 years. This negative evidence is confirmed by the Kavi plates of Jayabhaṭṭa III of Bharoch,(7) where that ruler claims to have defeated the king of Valabhi by the skill of his sword.

c8.8

⁴ Hiralal, *List of Inscriptions in C. P. and Berar*, p. 18.

c8.9

⁵ E. I., XVIII, pp. 235 ff.

c8.10

⁶ J. B. B. R. A. S., XI, p. 112.

c8.11

⁷ I. A., V, p. 112.

c8.12

It would seem that Jayabhaṭṭa III had made an alliance with the Gujrat Cālukya ruler Maṅgalarasa, and that the allied forces had wrested away the continental districts from the Valabhī king in c. 725 A.D. Indra I seems to have, very probably, served in this campaign as one of the feudatories of the Cālukya king; it was in that capacity that he may have been present in Kaira, when he carried away the Cālukya princess by force from the marriage pandal there. Bhavanāgā may have been a daughter of Maṅgalarasa or Pulakeśin. Hindu marriage is indissoluble and, therefore, the estrangement caused by this *rākshasa* marriage could not have lasted long. The Cālukya ruler may have soon reconciled himself with the accomplished fact; the marriage may have also added to the prestige and patrimony of Indra in Berar.

c8.13

The Kairā and Panchmahāl districts were not held for a long time by the Gurjara-Cālukya forces. From Al Biladuri we learn that Junaid, the Arab governor of Sindha, had sent expeditions to the kingdoms of Marwar, Bharoch, Ujjain, Mālvā and Bhinmal.⁸ The account of the Muslim chronicler is confirmed by the Naosari plates of Pulakeśiraja, dated 739 A.D.,⁹ which state that the Muslim army, which had defeated the kings of Sindh, Kachchha, Kathiawar and the Chayda, Maurya, and Gurjara rulers, was repulsed by Pulakeśin. Since the grant is dated 739 A.D., and since Pulakeśin's accession took place in 731 A.D., the Muslim raid must have taken place between these two dates. The havoc of the Muslim raid and the effort to repulse it must have weakened the states in Gujarat and Malva, and Dantidurga must have decided to fully exploit this situation when he began his career in c. 745 A.D. at the age of about 22 or 23. His imagination must have been fired to some extent by his descent from the Cālukyas on his mother's side.

c8.14

⁸ Elliot, I, p. 109.

c8.15

⁹ B. G., I. i, p. 109.

c8.16

Our sources of information about Dantidurga are two contemporary records, the Samangad plates ¹¹ dated 754 A.D.

c8.17

¹¹ I. A. XI, pp. 111 ff.

c8.18

11. I. A. XI, pp. 111 ff. The genuineness of this record has been recently called into quetion by Drs. Sukhtankar and Bhandarkar (E. I. XIV, pp. 121-2; Ibid, XVIII, p. 236). Dr. Sukhtankar's conclusion, which is mainly based on palæographical grounds, is that the plates are a few decades later than the alleged time of their issue. The palæographical test is, however, a hardly convincing one when the difference is only of a few decades. If we compare the Talegaon plates of Krishna I with the Samangad plates, we no doubt find that the duct of the former is much more archaic than that of the latter; but the s'a in the latter record is more archaic than the s'a in the former. The sign-manual of Dadda Pras'antaraga in his Kaira Plates of 669 A.D. is in characters at least two centuries later than that date. (J. R. A. S., 1864, p. 205). The same is the case of the Mathura Jain inscriptions of the Kushana period, whose characters are much in advance of their age. It is clear from the last two cases that the current hand was considerably in advance of the monumental writing, and it is not impossible that the characters of the Samangad plates appear a few decades later than the date of the record, not because it is not genuine but because it is written in the current, rather than in the monumental script of the period. The names of the villages have been tampered with in this record, and if we have to suppose that the record is not genuine, it will be a case of double forgery. Palæographical evidence being inconclusive in this matter, attention may be drawn to a few points which would show that the record is a genuine one:

c8.19

i. The verse :-

c8.20

महीमहानदीरेवारोभोभित्तिविदारणम् ।
लोका विलोकयन्त्युच्चैः कृतं यज्जयकञ्जरैः ॥

c8.21

does not occur in any other Rashtrakuta record; and it will be soon shown that the armies of Dantidurga had operated in the vallies of these rivers. If the record were not genuine, or based upon a genuine record, and if it were forged by the kramavid Brahmanas of Karad, the donees of the grant, the rivers mentioned in connection with the exploits of Dantidurga would have been big rivers like the Ganges, the Jumna etc. and not insignificant ones like the Mahi, the Mahanadi and the Reva. The fact that these comparatively unimportant rivers are mentioned would show that the grant is either genuine or based on an original document issued by Dantidurga.

c8.22

and the fragmentary Ellora Daśāvatāra cave inscription, ¹²

c8.23

¹² [^cont]: text of the note (Continued from last page)

c8.24

ii. The writer of the Talegaon plates of Kṛṣṇa 1 is the same individual as the writer of the Samangad plates issued 15 years earlier. It is very likely that Indra, who composed the grant, was serving both under Dantidurga and Kṛṣṇa. When it is remembered that the Talegaon grant is not based or modelled on the Samangad one or vice versa, the identity of the composers can be explained only on the assumption that the record is a genuine one. Forged records are not likely to be correct on such minute points.

c8.25

iii. The date of the Antroli-Chharoli record, 757 A.D., is not the date of Karka, the grandfather of Dantidurga, as shown in the last chapter. So there is nothing impossible in Dantidurga being a ruling prince in 753-4 A.D. when the Samangad grant was issued.

c8.26

It may be pointed out that even if we accept the view of Dr. Sukhatankar, the plates will have to be pronounced to be a few decades later than the alleged date of their issue. If such is the case, we will have to assume that the present plates are a later copy of an originally genuine grant, and therefore their value as a record describing contemporary incidents cannot be much diminished. The Samangad plates in that case will stand on the same level as the Konnur Inscription of Amoghavarsha. The latter record was no doubt not written in S'aka 782, but it has been shown recently by Dr. D. R. Bhandarkar ( E. I., XVIII, p. 235 ff. ) that the record is a genuine copy of an earlier grant.

c8.27

12. Kielhorn seems to assign this record to the reign of Amoghavarsha I, ( E. I. VII, Appendix, p. 13 ) but this view is not correct. The record does not refer to any king later than Dantidurga but merely supplies S'arva as an epithet of that king. This becomes quite clear from an analysis of the record. V. 23 describes Dantidurga's victory over the rulers of Lāta, Mālva, Badāmi, etc., V. 24 compares him to a number of gods according to the usual notions on that subject, and V. 25 after describing his bravery concludes with the line:—

c8.28

बीभत्सुर्दुष्टनागो जयिनमिव परं श्रीमहाराजशर्वः ॥ ‘S'rī Maharaja S'arva was terrible like a mad elephant to the enemy who was ambitious of conquest.’

c8.29

Then immediately follows the line:—

c8.30

दत्तं येनोज्जयिन्यामपि नृपतिमहादानमाश्चर्यभूतम् । ‘Who had made in Ujjayini in a wonderful way the great gift prescribed for kings.’

c8.31

and the notices about him in the records of his successors. From these we learn that the rulers of Kanchi, Kalinga, Srisaila, Kosala, Malava, Lata, Tanka, and Sindh were defeated by Dantidurga.¹³ Our records, however, do not supply any clue as to whether the overthrow of the Chalukyas preceded or followed the defeat of other kings mentioned in them.

c8.32

¹³ Cf. Ellora Inscription V. 23, Samangad grant lines, 21-22, Begumra plates of Indra III, E. I. IX, pp. 24 ff. The Das'avatara record gives the name of the last mentioned king as Sandhubhupa but this is obviously a mistake for Sindhubhupa.

c8.33

But before we proceed to determine the probable chronology of these wars, we have to consider a serious discrepancy

c8.34

This line obviously refers to हिरण्यगर्भ महादान which Dantidurga had given at Ujjayini, as we know from the recently published Sanjan plates of Amoghavarsha I. It is, therefore, clear that Maharaja S'arva who is mentioned in V. 25 of the Das'avatara record must be obviously Dantidurga, whose exploits are the topic of eulogy from V. 23. If we assume that it is Amoghavarsha I and not Dantidurga who is referred to in the last line of V. 25, which refers to king S'arva, we shall have to suppose that the record passes over, entirely unnoticed, as many as four rulers, viz. Krishna I, Govinda II, Dhruva and Govinda III. This is extremely improbable, since the author of the Das'avatara inscription has devoted several verses to describe the imaginary careers of the predecessors of Dantidurga. He would have waxed ten times eloquent in describing the all-India exploits of Dhruva and Govinda III; Buhler, who had edited the record, had realised that king S'arva could not be identified with Amoghavarsha I; he had proposed to regard him as a brother or a minister of Dantidurga. (A. S. I., V, pp. 186 ff.) This view, however, overlooks the fact the V. 25 describes the महादान ceremony of S'arva as something which other kings could not have thought of emulating, even in dream. This statement would have been altogether inappropriate and wide of the mark, if S'arva were a younger brother or a minister of Dantidurga. To conclude, the last ruler mentioned in this record is Dantidurga S'arva and therefore the record may well be a contemporary document. But as it is incomplete, one cannot be positive on the point.

c8.35

in the Rāṣṭrakūṭa records about the overthrow of the Chālukyas. Some of the records ascribe the defeat of Kīrtivarman II to Dantidurga and others to Kṛṣṇa I. There can, however, be no doubt that Kṛṣṇa I only completed the work of his nephew. It is no doubt true that the Wani-Dindori¹⁴ and Radhanpur¹⁵ plates of Govinda III, dated 808-9 A.D., Baroda plates of Karka, ¹⁶ dated 812 A.D., and the Kāpadwanj plates of Kṛṣṇa II, ¹⁷ dated 910 A.D., give the credit of the overthrow of the Chālukyas to Kṛṣṇa I, but this is obviously due to the fact that these records altogether pass over Dantidurga, because he was only a collateral ancestor. The Sanjan plates of Amoghavarsha I, ¹⁸ dated 871 A.D., and the Cambay¹⁹ and Sangli²⁰ plates of Govinda IV, dated 930 and 933 A.D. respectively, no doubt mention Dantidurga, but give the credit of the Chālukya overthrow to Kṛṣṇa I. But these are late records and their testimony will have to be rejected in favour of the undoubtedly earliest and almost contemporary records, viz. the Bhārata Itihāsa Samshodhaka Mandala²¹, Talegaon²² and Bhanduk²³ plates of Kṛṣṇa I dated 758, 770 and 774 A.D. respectively. These records were issued by Kṛṣṇa I himself, and they not only do not claim the credit of the Chālukya defeat for him, but actually give it to Dantidurga. There can, therefore, be no doubt that it was Dantidurga, and not Kṛṣṇa I who gave the first decisive blow to the Chālukya house. The later records ascribe that feat in some cases to Kṛṣṇa I, partly because they pass over Dantidurga altogether, and partly because Kṛṣṇa I completed the work of his nephew and humbled the Chālukyas of Vengi as well.

c8.36

¹⁴ I. A., XI, pp. 157 ff.

c8.37

¹⁵ E. I., VI, pp. 242 ff.

c8.38

¹⁶ I. A., XII, pp. 158 ff.

c8.39

¹⁷ E. I., I, pp. 52 ff.

c8.40

¹⁸ E. I., XVIII, pp. 235 ff.

c8.41

¹⁹ E. I., VII, pp. 36 ff.

c8.42

²⁰ I. A., XII, pp. 249 ff.

c8.43

²¹ B. I. S. M. J., VIII, 165-8.

c8.44

²² E. I., XIII, pp. 275 ff.

c8.45

²³ E. I., XIV, pp. 101-02.

c8.46

Dantidurga seems to have begun his career by attacking his eastern neighbours in Kos'ala,(24) Udayana of Sirpur near Raipur and Jayavardhana of Srīvardhana near Ramtek, who was probably also known as Pṛthvīyāghra.(25) If the Udayendiram plates,(26) issued by Nandivarman II in his 21st year, are geniune or based, like the Konnur inscription of Amoghavarsha I, on a genuine record, it would appear that Dantidurga and Nandivarman II were co-operating with each other in this expedition: for the Udayendiram plates claim that Nandivarman II had captured Udayana, the king of the Sabaras, and defeated Pṛthvīyāghra, another S'abara chief, who was presumptuous enough to perform an As'vamedha sacrifice. Since several records of Dantidurga claim for him also the credit of defeating the Kos'ala ruler or rulers, it seems very likely that he had made an alliance(27) with the Pallava ruler Nandivarman, the natural enemy of his prospective

c8.47

24. Udayana's great-grand-son Chandragupta was defeated by Govinda III in c. 810 A.D. It is, therefore, very likely that Udayana himself was on the throne at this time. See E. I., XVIII, pp. 240 and XI, pp. 185 ff.

c8.48

25. Jayavardhana is called परमेश्वरमहाराजाधिराजसकलविन्ध्याधिपतिः। in the Ragholi plates; he may, therefore, well have been the same as the S'abara chief, Prithvīyāghra, who attempted to perform an As'vamedha according to the Udayendiram plates. With reference to Nandivarman he was a northerner.

c8.49

²⁶ S. I. I. II., pp. 365 ff.

c8.50

²⁷ Dubreuil has suggested that Dantivarman cemented this alliance by giving his daughter S'ankhā in marriage to Nandivarman (Pallavas, p. 75) There are several difficulties in accepting this theory. In the first place it may be pointed out that the Bahur plates simply mention that S'ankhā was a Rāshṭrakūṭa princess; the name of her father is not given. (S. I. I., II, p. 515). We have seen already how there were several Rathi or Rāshṭrakūṭa families in Andhra and Karnatak; it is not impossible that S'ankhā may have belonged to one of these and may have been altogether unconnected with the house of Dantidurga. In the second place it may be pointed out that Nandivarman, who had married S'ankhā, was the son of Tantivarman (See V. 12 of the Bahur plates; S. I. I., II, p. 515), while Nandivarman, who was the ally and [P. T. O.

c8.51

opponent Kīrtivarman II. It is no doubt true that a number of records, like the Daśāvatāra one, state that Dantidurga had defeated the king of Kanchi as well; but this statement is not against the theory of alliance. For Nandivarman I was not the son of his predecessor Parameśvaravarman, as the Udayendiram plates were supposed to prove. He was born in a collateral branch and seems to have supplanted by force either Parameśvaravarman or his successor.²⁸ Dantidurga may probably have helped Nandivarman in this war against his predecessor. Hence his records may well claim that he had defeated the Pallava king. It was probably when Dantidurga was cooperating with the forces of Nandivarman that he may have defeated or come into hostile contact with the ruler of Śrīśaila in Karnul district.

c8.52

²⁸ See Kasakudi plates of Nandivarman, S. I. I., II, No. 78, and Velurapalaiyam plates of Vijayanandivarman III, S. I. I., II, No. 98.

c8.53

After strengthening his position by the defeat of his eastern neighbours and by his alliance with Nandivarman, Dantidurga seems to have attacked the Gurjaras of Bharoch and the Chālukyas of the Gujrat branch, and annexed their dominions. These kingdoms were already exhausted by the Muslim raid and its repulsion, and Dantidurga's task may not have been very difficult.

c8.54

Dantidurga's successes must have made a collision with Kīrtivarman II inevitable, and it seems to have taken place before January 754 A.D. when the Samangad plates were issued. We are told that the Chālukya emperor was easily defeated merely by the frown without even any weapons.

c8.55

( Continued from last page ) contemporary of Dantidurga, was the son, either of Parameśvaravarman, according to the earlier view based upon the Udayendiram plates, or of Hiraṇya, according to the present view based upon the Kasakudi and Velurapalaiyam plates. No record, however, designates his father as Tantivarman. It must, therefore, be admitted that there is no conclusive evidence to prove that S'ankhā, the Rāshṭrakūṭa princess, was a daughter of Dantidurga married to Nandivarman II.

c8.56

being raised;(29) this statement, if not an exaggeration, would show that the overthrow of the Chālukya emperor was brought by strategem or treachery. Dantidurga was enabled by this victory to occupy Khandesh, Nasik, Poona, Satara and Kolhapur districts, as the findspot and the villages mentioned in the Samangad plates would show. Kīrtivarman, however, continued to hold sway in Karnatak right up to Sholapur even after this defeat. In 757 A.D. we find him encamped with his army at Bhadara-Gaviṭṭage,(30) a village on the northern bank of the Bhima in Sholapur district. Either Dantidurga or his successor Kṛṣṇa I soon completed the Chālukya overthrow. The second defeat, the whereabouts of which are still unknown, was decisive; for not only do we not find any record of Kīrtivarman subsequent to 757 A.D. but the records of the later Chālukyas themselves state expressly that the glory of the Chālukyas set with Kīrtivarman II.(31)

c8.57

Dantidurga's attacks on the rulers of Tank, Sindha and Malva now remain to be considered. Tanka has not been so far identified. Arabs of Sindh were now and then attacking their western neighbours and Dantidurga seems to have defeated some of these raids. Some time after the first defeat of Kīrtivarman, Dantidurga seems to have led an expedition to Ujjain either to help the Gurjara-Pratihāra ruler Devarāja, or to take advantage of the defeat that was inflicted on him by the rival Gurjara-Pratihāra king, Siluka, ruling at Bhinmal.³² This becomes quite clear from the Das'avatāra record which states that Dantidurga was occupying a palace of the Gurjara ruler³³ and from the Sanjan plates where we are informed that when Dantidurga performed the *Hiranya-garbhadāna* ceremony at Ujjain, the Gurjara ruler acted as his door-keeper.

c8.58

²⁹ Cf. सभूविभङ्गंमगृहीतनिशातशस्त्रम् । अश्रान्तमतिहताज्ञमपेतयत्नम् ॥ यो वल्लभं सपदि दण्डबलेन जित्वा । भृत्यैः कियद्भिरपियः सहसा जिगाय ॥ I. A., XI, pp. 111.

c8.59

³⁰ Vakkeri plates of Kīrtivarman II, E. I., V, pp. 202 ff.

c8.60

³¹ तद्भवो विक्रमादित्यः कीर्तिवर्मा तदात्मजः । येन चालुक्यराज्यश्रीरन्तरायिण्यभूद्भुवि ॥ Yewoor Inscription, I. A., VIII, pp. 11 ff.

c8.61

³² The line: 'पूर्वीं श्रीमदवन्तिभूभृति नृपे वत्साधिराजे' in the colophon of *Harivans'a* of Jinasena has to be construed to mean that Vatsarāja was ruling at Avanti; the lines in the Sanjan plates

c8.62

³³ यस्याखण्डितविक्रमस्य कटकेनाक्रम्य तीरक्षितिम् ।

c8.63

After the conquest of Gujrat, Dantidurga appointed Karka II of the Antroli-Chharoli record as his governor of the province. Karka II was a relative of his as shown already. The Antroli-Chharoli record does not claim that either Karka, or any of his ancestor had conquered southern Gujrat. Since a number of records ascribe that conquest to Dantidurga, it is clear that Karka owed his position in Gujrat to that king. It may be pointed out that Karka's titles do not show that he was an independent ruler.

c8.64

A review of Dantidurga's career will show that he was a leader gifted with political insight and possessed of great organising capacity. He could see how the Chālukya empire had become weakened by its incessant wars with the Pallavas and how the petty states of Gujrat, exhausted by the Muslim forays, could no longer hold their own against a resolute conqueror. Forming wise alliances, proceeding cautiously step by step, utilising the services of his youthful nephews and mature uncles, he gradually enlarged his kingdom until it included southern Gujrat, Khandesh, Berar, and northern

c8.65

हिरण्यगर्भे राजन्यैरुज्जयिन्यां यदासितम् ।
प्रतिहारीकृतं येन गुर्जराधिराजकम् ॥

c8.66

make this construction most plausible. For the defeat of Devarāja, see Jodhpur Inscription of Pratihara Bauka and Dr. Majumdar's remarks thereon; E. I., XVIII, pp. 87 ff.

c8.67

सौधेऽस्मिन्कृतगुर्जरेन्द्ररुचिरे... ॥

c8.68

Mahārāṣṭra. The theory that he oppressed his subjects and was overthrown by his uncle Kṛṣṇa I is, as will be shown in the next section, no longer tenable. Like most other founders of new dynasties he seems to have been a wise and able ruler. In religion he shared orthodox Hindu beliefs. When he was at the sacred tīrtha Ujjain, he performed there the Hiraṇyagarbhamahādāna; he must have obviously believed that the ritual would endow him with a divine body. On the rathasaptamī day of 754 A.D. he weighed himself against gold and distributed the money among the Brahmanas. He showed his devotion to his mother by bestowing, at her request, lands in several villages on worthy Brahmanas.³⁴ The precise date of his death is not known but since his successor was on the throne in 758 A.D., it is clear that he must have died not long after 754 A.D., when he could hardly have been much above 30. When we remember how young he died, his achievements will have to be pronounced as unique.

c8.69

³⁴ Samangad plates, I. A., XI, pp. 111 ff.

Kṛṣṇa I

c8.s1.1

Dantidurga was succeeded by his uncle Kṛṣṇa I probably because he left no male issue. This reason is no doubt given by the Kadba plates which were issued in the reign of Karka II, but a record from Chitaldurg confirms it. This record³⁵ informs us that when Dantidurga died without a son, Kannara, his junior uncle, succeeded him.

c8.s1.2

³⁵ E. C., XI, Chitaldurga No. 49, Alas plates of Yuvarāja Govinda refer to मानावलोकरत्नवर्षश्रीविजयादित्य, son of Dantivarman, as the person at whose request the grant was made. But since Dantivarman of this record is not the son of Indra, it will not be possible to identify him with Dantidurga.

c8.s1.3

The view that Dantidurga was deposed by his uncle Kṛṣṇa for oppressing the subjects was based upon a misunderstanding of the verse:—

c8.s1.4

तस्मिन्दिवं प्रयाते वल्लभराजेऽकृतप्रजाबाधे । श्रीकर्कराजसूनुर्महीपतिः कृष्णराजोऽभूत् ॥

c8.s1.5

occurring in the Begumra plates of Kṛṣṇa II. The reading of the last part of the first line is ‘Kṣataprajābādhaḥ’ in the Talegaon plates of Kṛṣṇa I himself, and the Paithan plates of Govinda III, ‘Kṛtaprajāpālaḥ’ in the Daulatabad plates of S’aṅkaragaṇa, and ‘kṣitau prajāpālaḥ’ in the Alas plates of Kṛṣṇa’s son Govinda II. It is, therefore, clear that this line either praises Kṛṣṇa I for removing the miseries of the subjects or observes that Dantidurga was also a ruler of the same category; the correct reading of the last word of the first line in the Begumra plates is clearly, ‘akṛtaprajābādhe’ and not kṛtaprajābādhe. (36) A verse in the Baroda plates of Karka of the Gujrat branch no doubt says that Kṛṣṇa I had ousted a relative, who had gone astray, but the Bharata Itihasa Samshodhaka Mandala, (37) Talegaon (38) and Bhanduk (39) plates of Kṛṣṇa I himself make it abundantly clear that this relative could not have been Dantidurga. As Dr. D. R. Bhandarkar observes, (40) Kṛṣṇa I would not have tolerated a eulogy of his murdered nephew in his own grant, and added at the end that when the victim of his machinations had departed from this world, he ascended the throne. The relative ousted by Kṛṣṇa was very probably Karka II of the Antroli-Chharoli record who, we have seen, was a distant cousin of his. The Antroli-Chharoli record itself shows that Karka II was even then entertaining imperial ambition. Karka there styles himself ‘Samadhigatapanchamahāś’abda-parama-bhaṭṭāraka-mahārājādhirāja-parameś’vara : It will be seen that here Karka is trying to smuggle cleverly for himself the imperial title after first using the feudatory one to which alone he was entitled. After the death of Dantidurga he may have openly declared himself emperor. It is, therefore, very probable that the relative ousted by Kṛṣṇa was Karka II, he was certainly not Dantidurga.

c8.s1.6

³⁶ I. A., XIII, pp. 66 ff.

c8.s1.7

³⁷ B. I. S. M. J., VIII, p. 165.

c8.s1.8

³⁸ E. I., XIII, pp. 275 ff.

c8.s1.9

³⁹ E. I., XIV, pp. 121 ff.

c8.s1.10

A verse occurring in four records of the Gujrat branch⁴¹ states that Kṛṣṇa obtained the title of *Rājādhirājaparameśvara* after defeating Rāhappa who had become inflated with pride about his strength. This Rāhappa must have been different from the relative, whom Kṛṣṇa had ousted, for he is nowhere described as a relative of Kṛṣṇa. He cannot be, therefore, Karka of the Antroli-Chharoli record. The present state of our knowledge, however, does not enable us to identify this potentate with certainty. A ruler of Mewar, bearing that name, is known to have flourished in c. 1200 A.D.; an inscription from the same province, dated 725 A.D., refers itself to the rule of *Mahārājādhirāja Parameśvara* Śrī-Dhavalappadeva.⁴² It may thus appear that in some of the ruling families of Mewar, names ending in *ppa* were current, and it is not impossible that Rāhappa, the opponent of Kṛṣṇa I, may have been a ruler in Mewar, possibly a hitherto unknown son or successor of king Dhavalappa just referred to. It has also to be remembered that the last part of the name Rāhappa bears a Canarese appearance, and since Kṛṣṇa I is said to have obtained the imperial position after defeating Rāhappa, it is not impossible that Rāhappa may have been either another name of Kīrtivarman, or a *biruda* of the contemporary Vengi ruler Viṣṇuvardhana defeated by his heir-apparent. It must, however, be confessed that in the present state of our knowledge no identification, that can be unreservedly accepted, can be proposed.

c8.s1.11

⁴¹ Kavi Plates of Govinda, 827-8 A. D., I. A., V, pp. 145 ff. Begumra plates of Dhruvarāja II, 867-8 A. D., I. A., XII, pp. 181 ff. ,, ,, ,, Kṛṣṇa, 888-9 A. D., I. A., XIII, pp. 46 ff. Surat plates of Karka, 821 A. D. (being edited by the writer in E. I.)

c8.s1.12

⁴² E. I., XII, pp. 11 ff.

c8.s1.13

Kṛṣṇa was a mature man of about 45 when he succeeded his nephew in c. 756. He had very probably participated in many of the campaigns of his nephew, and it is very probable that the final defeat of Kīrtivarman was his achievement. For Kīrtivarman was holding his own in Karnatak as late as 757 A.D., i. e. three years after the last known date of Dantidurga and one year before the first known date of Kṛṣṇa I. The Bharata Itihasa Samshodhak Mandala plates of Kṛṣṇa, issued in September 758, do not refer to his defeat of the Chālukyas; it is probable that the event was yet to take place then. But not much later than that date he must have completed the Chālukya overthrow, and annexed the southern part of that kingdom.

c8.s1.14

He must have spent some time in consolidating his position; but that was not much. Soon he undertook an offensive expedition against Gangavadi; for his Talegaon plates (43) show him encamped in 768 A.D. at Maṇṇe in Mysore state during the course of that expedition. An echo of this invasion of Gangavadi is heard in a lithic record from Tumkur district, (44) immortalising the memory of a hero, who had fallen in the war caused by the rising of the Raṭṭas against Gangavadi. Since this record refers itself to the reign of Srīpurusha, it is clear that this aged ruler was on the throne of Gangavadi when Kṛṣṇa invaded it after annexing the Chālukya dominions. The issue of Talegaon plates from Maṇṇe, the Ganga capital, shows that Kṛṣṇa was successful in his expedition. Kṛṣṇa's expedition against the Gangas was not known before the publication of his Talegaon plates.

c8.s1.15

The Chālukyas of Vengi were next attacked, and this expedition was under the charge of Yuvarāja Govinda whom we find encamped on the confluence of the Musi and the Krishna—hardly 100 miles from Vengi itself—in 770 A.D. while conducting that expedition (45). Viṣṇuvardhana IV, who was then upon the Vengi throne, was defeated by Govinda. As a result of this victory a major part of the modern state of Hyderabad must have passed under the Rāṣṭrakūṭa sway. The recently published Bhanduk plates (46) of Kṛṣṇa show that in 772 A.D. practically the whole of Marathi C. P. was then under his rule.

c8.s1.16

⁴³ E. I., XIII, pp. 275 ff.

c8.s1.17

⁴⁴ E. C., XII, Maddagiri, No. 99.

c8.s1.18

Kṛṣṇa I had also brought under his sway southern Konkan. From the Kharepatan plates of Raṭṭarāja, we learn that Saṇaphulla, the founder of the line, had obtained the territories between the Sahya and the sea through the favour of Kṛṣṇarāja. (47) S'ilāhāras were for a long time very loyal feudatories of the Rāṣṭrakūṭas; allowing about 25 years per generation for the ancestors of Raṭṭarāja, we find that Saṇaphulla has to be placed in the latter half of the 8th century It is therefore evident that Kṛṣṇarāja, who had placed Saṇaphulla in charge of Konkan, must be Kṛṣṇa I.

c8.s1.19

The date of the death of Kṛṣṇa I is not known, but it must have taken place some time between the 23rd of June 772 A.D., when the Talegaon plates were issued and October 775 A.D., when the Pimperi plates (48) were issued by his son Dhruva, which do not refer to Kṛṣṇa I as living or ruling.

c8.s1.20

Kṛṣṇa was undoubtedly an able ruler and a skilful general. During his short reign of about 18 years he enlarged the kingdom he had inherited to three times its original size by annexing Konkan, Karnatak and the major portion of Hyderabad state to his empire. He had humbled down the Ganga and Vengi rulers and the expression ‘Kāñchīguṇālaṅkṛtā vis'vambharā nijavaniteva sā tena bhuktā’ occurring in his Talegaon plates, may perhaps be referring to the defeat of the Pallava ruler of Kanchi as well, though we have no other evidence on the point. It is clear that by his conquests Kṛṣṇa I had secured for his house a dominating position in the Deccan paving thereby the way of his successors to participate in the politics of the north. He was also a great builder and caused to be excavated the Ellora Kailasa temple which is one of the architectural wonders of the world, since the whole structure is hewn out of solid rock.

c8.s1.21

⁴⁵ Alas plates of Yuvarāja Govinda, E. I., VI, pp. 208 ff.

c8.s1.22

⁴⁶ E. I., XIV, pp. 121 ff.

c8.s1.23

⁴⁷ E. I., III, pp. 292 ff.

c8.s1.24

⁴⁸ E. I., X, pp. 81 ff.

c8.s1.25

The overthrow of the Cālukyas was complete and decisive, and it is only rarely that we come across even Cālukya feudatories in the first century of the Rāṣṭrakūṭa rule. Only three such feudatories come to our notice. (i) Kaṭṭiyira of Diḍgur inscription (49) was possibly a Cālukya chief but he was a very petty ruler. Kaṭṭiyiradeva, who is mentioned as an ancestor of later Cālukyas in Managoli inscription, (50) may possibly be this Kaṭṭiyira. (ii) Kadba plates of Govinda III (51) disclose another Cālukya feudatory named Yaśovarman, but since he had to requisition the services of the Ganga chief to secure the permission of Govinda III for certain grants it is doubtful whether he was a ruler of even a mediocre feudatory status. (iii) Mahāsāmanta Budhavarsha of the Torkhede plates of Govinda III (52) was possibly of Cālukya extraction, but his feudal sway extended over only 12 villages. Arikesarin was a powerful Cālukya feudatory, but he belongs to the eve of the rise of the later Cālukyas. It seems clear that Kṛṣṇa had not permitted any member of the dethroned dynasty to occupy a position of any political importance in his empire.

c8.s1.26

It was only for the sake of convenience that the house of Dantidurga was designated as the Malkhed Rāṣṭrakūṭa house in the present and the preceding chapter. Malkhed became the capital in the time of Amoghavarsha I, and the pre-Malkhed capital of the house cannot yet be determined with certainty. On the strength of a verse in the Kathākosha, it was once suggested that Malkhed was the capital of the Rāṣṭrakūṭas from the beginning, since the verse in question mentions Śubhatuṅga, which was one of the *birudas* of Kṛṣṇa I, as ruling at Malkhed. This view, however, does not seem to be correct, since Śubhatuṅga is known to have been a *biruda* of Kṛṣṇa II as well. Besides a verse in the Karda plates of Karka II distinctly states that Amoghavarṣa I built Mānyakheṭa, which could put to shame even the capital of gods. ⁽⁵³⁾ It would, therefore, seem that before the time of Amoghavarṣa I some other city than Malkhed was the capital of the empire, especially as no record earlier than his time refers to Malkhed as capital. It has been suggested that Mayūrakhiṇḍi or Morkhind, a fort in the district of Nasik, may have been the pre-Malkhed capital, since Wani-Dindori, Radhanpur and Kadba plates of Govinda III were issued from that place. But the expression used in these records is *Mayūrakhiṇḍīsamāvāsitena mayā* and not *Mayūrakhiṇḍīvāstav-yena mayā*. It is, therefore, quite possible that at the time of the issue of these records Govinda III may have temporarily encamped in this fort. In the Wani-Dindori plates the village granted is described as included in Vaṭanagara *viṣaya* of Nāsika *deś'a*. If Morkhinda of Nasik district were really the capital at this period, it is strange that the *viṣaya* or *deś'a* in which it was situated should not have been designated after it. It would, therefore, appear that Mayūrakhiṇḍī was only a military station and not a capital. This same circumstance would go against the view that Nasik was the capital. The Dhulia ⁽⁵⁴⁾ and Pimperi ⁽⁵⁵⁾ plates besides show that Nasik was the seat of a viceroy, and not of the central Government, in the 8th century. Lāṭūra was the original home of the family, from which it had migrated in Berar a few centuries earlier; and it was also included in the Rāṣṭrakūṭa kingdom in the reign of Kṛṣṇa I. But there is no evidence to show that the earlier rulers of the family had made their ancestral home once more their capital. Paithan also was not the capital as the Paithan plates of Govinda III would show. As it is, we have to confess that the early capital of the house cannot be determined with certainty. It is, however, not improbable that Elichpur in Berar, where the ancestors of Dantidurga were ruling before the rise of the house, may have continued to be the capital, till it was changed to Malkhed. Cousens has recently suggested ⁵⁶ that the Rāṣṭrakūṭa capital may have been located somewhere in the vicinity of the Elora caves in the time of Dantidurga. He thinks that the place called as 'Sooloobunjun' in the Indian atlas sheets, which is just above the plateau near the Elora caves, may have been the actual locality of the capital, since what look like the remains of a town and a very large tank are still to be seen in the place. This view is a probable one, but the identification of the pre-Malkhed capital with Sooloobunjun can be accepted as proved only when more substantial evidence can be adduced. As it is, we have to admit that we have no definite knowledge as to the precise site of the capital before it was shifted to Malkhed.

c8.s1.27

⁴⁹ E. I., VI, pp. 251 ff.

c8.s1.28

⁵⁰ E. I., V, pp. 16 ff.

c8.s1.29

⁵¹ I. A., XII, pp. 13 ff.

c8.s1.30

⁵² E. I., III, pp. 54 ff.

c8.s1.31

⁵⁴ E. I., VIII, pp. 182 ff.

c8.s1.32

⁵⁶ Cousens, *The Chalukyan Architecture*, p. 1.

Linked texts (16)