Pramāṇa · Search · About · API

The Rashtrakutas and their Times c7.s1: Earlier Rāṣṭrakūṭa Ruling Families

The Rāṣṭrakūṭas and their Times c7.s1

Language
English

Earlier Rāṣṭrakūṭa Ruling Families

c7.s1.1

Ancestors of Dantidurga excepted, Abhimanyu of Mānapura, Nannarāja of the Multai grant and Karkkarāja of the Antroli-Chharoli record are the only Rāṣṭrakūṭa rulers, whose houses are known to have flourished earlier than the house of Dantidurga. A number of other rulers, however, have been considered as belonging to the Rāṣṭrakūṭa stock by some earlier writers; it will be first shown why these views are untenable.

c7.s1.2

(1) While editing the Kadaba plates¹, Hultzsch had suggested that Akālavarṣa Śubhatuṅga, mentioned in Merkara plates² as having flourished in the 5th century A.D., might be a Rāṣṭrakūṭa prince, possibly the father of Indra who, according to the Kauthem plates, was defeated by the early Cālukya prince Jayasiṃha. Akālavarṣa Śubhatuṅga no doubt looks like a Rāṣṭrakūṭa name, but we have to omit this king altogether from our consideration, as the Merkara plates have been shown to be a forgery.

c7.s1.3

¹ I. A., XII, pp. 13 ff.

c7.s1.4

² I. A., I, p. 363.

c7.s1.5

It is no doubt true that a few records of the later Cālukya dynasty, viz., the Kauthem plates of Vikramāditya V(3), Yewoor(4) and Nilgunda(5) plates of Vikramāditya VI, and Sonavade(6) and Miraj(7) plates of Jayasiṃha refer to a Rāṣṭrakūṭa dynasty said to have flourished earlier than that of the Cālukyas of Badāmi. We are told in these records that after ruling for 59 generations at Ayodhyā, the Cālukyas migrated to the south and ruled there for 16 generations. Then their glory was eclipsed for a time owing to the ascendancy of the Rāṣṭrakūṭas. But soon there arose the valiant Jayasiṃha, who conquered the Rāṣṭrakūṭa king Indra, the son of Kṛṣṇa, and reestablished the fortunes of his family. If this version of history be true, no doubt we have evidence for the existence of a Rāṣṭrakūṭa dynasty, ruling prior to c. 500 A.D., somewhere in central Mahārāṣṭra or northern Karṇātak. But there is ample evidence to show that the defeat of the Rāṣṭrakūṭas, attributed in these records to Jayasiṃha, is not based upon any reliable tradition. In the first place, the records in question are full of mythological details; a Cālukya dynasty ruling for 59 generations at Ayodhyā is not known to sober history. The most conclusive reason, however, for ignoring the theory, that Jayasiṃha had really defeated a Rāṣṭrakūṭa king Indra in c. 500 A.D., is the fact that not a single record of the early Cālukyas of Badāmi refers to this incident. This silence will appear the more significant when we remember that the Cālukya records mention a number of petty rulers like the Kadambas, the Mauryas, the Naḷas etc., who were supplanted by the early rulers of the dynasty. It is inexplicable why the Cālukya rulers of Badāmi should have conspired to condemn to oblivion the most significant achievement of Jayasinha, whom they were claiming as the founder of their dynasty. Fleet's theory in this respect appears as the most plausible one; the defeat of Rāṣṭrakūṭa Indra, son of Kṛṣṇa, attributed to Jayasinha in c. 500 A.D., is probably due to the fact that the historians of the later Cālukyas, like some of their present-day successors, believed that history repeats itself. They knew that Taila II, the founder of the later Cālukya empire, had defeated Rāṣṭrakūṭa king, Indra IV, the (grand-) son of Kṛṣṇa III; they attributed an exactly similar feat to the founder of the early Cālukya dynasty as well. Coins of a king Kṛṣṇa have been discovered in the Central Provinces which seem to belong to a fairly early date. But there is nothing to connect the king Kṛṣṇa of the coins with Kṛṣṇa, the father of Indra, said to have been defeated by Jayasinha. The Kṛṣṇa of the coins may be perhaps the Kalachuri ruler Kṛṣṇarāja, whose grandson Budharāja was defeated by Mangaliśa. (8)

c7.s1.6

³ I. A., XVI, pp. 151 ff.

c7.s1.7

⁴ I. A., VIII, pp. 11 ff.

c7.s1.8

⁵ E. I., XIII, pp. 12 ff.

c7.s1.9

⁶ B. I. S. M. J., X, pp. 87 ff.

c7.s1.10

⁷ E. I., XII, pp. 142 ff.

c7.s1.11

(2) Fleet had suggested that king Attivarman, mentioned in a grant hailing from Guntur district (9), may probably have been a Rāṣṭrakūṭa prince (10). This suggestion also is to be ruled out of order, because there is hardly anything to support it. The provenance and the characters of the grant show that it is very probably a Pallava record. The mythological descent from Hiranyagarbha, which Attivarman claims, would also suggest that he was a Pallava, rather than a Rāṣṭrakūṭa ruler. Rāṣṭrakūṭa rulers of no branch whatever make any mention of Ānandamarshi with whom Attivarman claims to be connected. The only ground for suggesting that Attivarman may have been a Rāṣṭrakūṭa is the expression 'Kandharanrpatikulodbhūtena', used with reference to Attivarman. But this expression is hardly sufficient to prove that Attivarman was a Rāṣṭrakūṭa prince; for, the name Kṛṣṇa is used by several non-Rāṣṭrakūṭa rulers as well. Besides the Canarese *apabhraṃśa* of Kṛṣṇa is Kannara and not Kandhara.

c7.s1.12

⁸ E. I., V, p. 295.

c7.s1.13

⁹ I. A., IX, p. 102.

c7.s1.14

¹⁰ Dynasties of the Canarese Districts, p. 334.

c7.s1.15

(3) The Komaralingam grant of Ravidatta (11) mentions that the founder of the grantor's dynasty was a king called Rāṣṭravarman. Hultzsch had placed this record, which is undated, in the 5th century A.D., and suggested that the founder's name, Rāṣṭravarman, may suggest Rāṣṭrakūṭa overlordship over Punnaḍu Viṣaya (modern Salem and Coimtore districts). If this suggestion were to be accepted, it would follow that there was a Rāṣṭrakūṭa kingdom in the 5th century, which extended right up to Coimtore. This view is, however, untenable; the grant has been proved to be a forgery (12), probably of about the 9th century A.D.; and even if it were genuine, the mere occurrence of the term *rāṣṭra* in the founder's name, Rāṣṭravarman, is hardly sufficient to prove that he was a feudatory of the Rāṣṭrakūṭas.

c7.s1.16

Having shown how a number of records, which were regarded by some of the previous writers as possible sources of information about earlier Rāṣṭrakūṭa rulers, do not supply any information in that connection, I propose to discuss the information we possess about Abhimanyu, Nannarāja and Karkkarāja, whom their records expressly describe as Rāṣṭrakūṭas. Let us see whether any of these rulers can be connected with the predecessors of Dantidurga.

c7.s1.17

The earliest of these rulers is king Abhimanyu, one of whose grants (13) describes his great-grand-father Mānāṅka as the most prominent among the Rāṣṭrakūṭas. Mānāṅka's son was Devarāja, and grand-son Bhavishya, and the grantor Abhimanyu was a son of the latter. Unfortunately the grant is not dated, nor is its findspot known. On palaeographical grounds we can place it in the 7th century A.D. About the locality of the territory over which Abhimanyu was ruling, we can get some idea from the statement in the plates that he was gracing Mānapura by his presence, when he made the grant of village Uṇḍivāṭikā to the Śiva temple at Peṭhapāṅgārikā, which is recorded in the charter. Dr. Bhagwanlal's view that Mānapura is the same as Mānyakheṭa or modern Malkhed¹⁴ has to be rejected; for, if the later Rāṣṭrakūṭa capital was really known as Mānapura in the 7th century, it is difficult to understand why the epithet pura of Mānapura should have been changed into kheṭa, especially since this change was calculated to belittle the importance of the place. For, pura denotes a city or a capital, while kheṭa is used only in connection with small towns. Further, it must be remembered that Abhimanyu and his ancestors were petty rulers, and if we identify their Mānapura with Malkhed, their dominions, we shall have to suppose, were very extensive. For, Pāṅgārikā, the Śiva temple of which received the grant, has been identified by Hultzsch¹⁵ and Fleet¹⁶ with Pagārā, 4 miles north of Panchmarhi in Houshangabad district, C. P., and Uṇḍivāṭikā, the village given, with one of the two villages called Oontia in the same locality. Abhimanyu would thus be ruling over a kingdom at least 400 miles in length, which would hardly have been the case, since such a kingdom in the 7th century would have cut across the dominions of the early Chālukyas. Fleet's suggestion that Mānapura, which was probably founded by Mānāṅka, may be Mānapura in Malva, the chief town of Mānapura sub-division, 12 miles southwest of Mhow, seems more probable, for it is only about 100 miles west of the village granted.

c7.s1.18

¹¹ I. A., XVIII, p. 366.

c7.s1.19

¹² E. I., I, p. 5.

c7.s1.20

¹³ E. I., VIII, pp. 163 ff.

c7.s1.21

¹⁴ J. B. B. R. A. S., XVI, p. 89.

c7.s1.22

¹⁵ E. I., VIII, p. 163.

c7.s1.23

¹⁶ I. A., XXX, p. 503.

c7.s1.24

The Rāṣṭrakūṭa house of Abhimanyu was thus ruling over the Mhow-Houshangabad tract in Central India. The question whether it can be connected with the house of Dantidurga cannot be settled at present.

c7.s1.25

Abhimanyu is lion, whereas that of the Rāṣṭrakūṭas of Malkhed is either Śiva or Eagle. Nor is there any similarity in the names of the members of the two families. The territory over which Abhimanyu was ruling was, however, immediately to the north of the kingdom of the Rāṣṭrakūṭa king Nannarāja who was, as will be presently shown, very probably either a direct or a collateral ancestor of Dantidurga. Since on palaeographical grounds, the grant of Abhimanyu can be placed in the middle of the 7th century, Mānāṅka, Devarāja, Bhavishya and Abhimanyu become the contemporaries, as will be shown below, of Durgarāja, Govindarāja, Svāmikarāja and Nannarāja respectively of the Tivarkhed and Multai plates. The kingdoms of these Rāṣṭrakūṭa families were also contiguous to each other. It is not unlikely that the two houses may have been connected with each other by blood relationship; but definite evidence to establish such connection is still wanting.

c7.s1.26

The next definite mention of Rāṣṭrakūṭa kings is to be seen in the Tivarkhed and Multai plates. The name of the grantor of the Multai plates has been read by Fleet as Nandarāja(17) but an examination of the facsimile published by him makes it absolutely certain that the name of the king is Nannarāja and not Nandarāja. Fleet has mistaken the partially faint lower na for da; a comparison of this letter with nda in ll. 2 and 6 will show clearly that the letter in question is nna and not nda.

c7.s1.27

If we compare the genealogies in the Tivarkhed and Multai plates, we shall find that they are absolutely identical. In both the plates the grantor is Nannarāja and his father, grand-father and great-grand-father are Svāmikarāja, Govindarāja and Durgarāja respectively. But the Tivarkhed plates were issued by Nannarāja in Śaka 553 or in 631-2 A.D. (18), whereas the Multai plates purport to be issued by the same ruler in Śaka 631 or 709-710 A.D.

c7.s1.28

¹⁷ I. A., XVIII, pp. 230 ff.

c7.s1.29

¹⁸ E. I., XI, pp. 276 ff.

c7.s1.30

Multai plates were correct, it was possible to argue that Nandarāja was a younger brother of Nannarāja and, therefore, may have been on the throne even 78 years later than the elder brother. But it is shown above that the grantor of the plates in question is Nannarāja, and not Nandarāja, and therefore identical with the ruler who had issued the Tivarkhed plates. In his grant issued in 631-2 A.D., Nannarāja claims that unlike his ancestors he was a ruler entitled to the dignity of the pañchamahās'abda, which he had personally won. He must be at least 25 at this time; could he be then still on the throne 78 years later? We have got the case of Nizam-ul-Mulk, who died when he was 104; so this is not impossible. But we must confess that such a long reign, as distinguished from life, is very rare, and it is therefore almost certain that the date of one of the two records is wrong.

c7.s1.31

There are good grounds to hold that the date 709-10 A.D. supplied by the Multai plates may not be genuine. The genealogy of the Multai plates starts in verse, but after the first verse there is a sudden break. A sentence in prose follows but the concluding portion of it, tasyātmavānātmajo, is again the fragment of a verse. The record, therefore, does not seem to be genuine, at least it is not carefully drafted or copied; and the date it supplies to us may not be genuine.

c7.s1.32

The genealogy of this Rāshṭrakūṭa house, as we gather it from these two records, is as follows:—

c7.s1.33

(1) Durgarāja.

c7.s1.34

(2) Govindarāja, son of No. 1.

c7.s1.35

(3) Svāmikarāja, son of No. 2.

c7.s1.36

(4) Nannarāja Yudhāsura, son of No. 3. Known dates, 631-2 A.D. and 709-10 A.D. (?)

c7.s1.37

Since the reign of Nannarāja commenced earlier than 631 A.D., we may assign his predecessors to the following periods, assuming an average reign of 20 years.

c7.s1.38

Durgarāja 570-590 A.D. Govindarāja 590-610 A.D. Svāmikarāja 610-630 A.D. Nannarāja, known dates, 631 A.D. and 710 A.D. (?)

c7.s1.39

Can we establish any connection between the Rāṣṭrakūṭas of the Tivarkhed and Multāi records with the predecessors of Dantidurga? It may be at once admitted that direct evidence to connect the two houses is not yet forthcoming: but the available data make it extremely probable that Dantidurga was either a direct or a collateral descendant of Nannarāja of the Tivarkhed and Multai records. The seal of the two plates issued by him is Garuḍa or Eagle, which was also the seal of Dantidurga and his descendants. The names of Nannarāja and his ancestors are either similar to, or identical with, the names of many of the predecessors and successors of Dantidurga. One of the latter's uncles, who was probably a younger brother of Kṛṣṇa I, and was governing the territory round Daulatabad in 793 A.D. (19) was Nannarāja. The formation of the name Svāmikarāja is similar to the biruda Pṛchhakārāja which was borne by Indra (20). Govindarāja, the name of Nannarāja's grand-father, is repeated four times in the main Rāṣṭrakūṭa line ruling at Malkhed, and once in the Gujrāt branch. The great-grand-father of Nannarāja was Durgarāja and his name may have paved the way of the formation of the name Dantidurga. This close similarity in the names of the members of the two houses can hardly be explained, except on the assumption that the two families were connected with each other. Indra, Karka, Govinda, Dhruva, Akālavarṣa Subhatunga, Akālavarṣa Kṛṣṇa and Dantivarman were the names of the different rulers in the Gujrat Rāṣṭrakūṭa branch, whose blood relationship with the house of Malkhed cannot be doubted; and all these names are repetitions of the names of the earlier rulers of the main line. The names of four out of five rulers of the Gujrat branch of the Chālukyas, viz. the two Jayasinhavarmans, Vinayāditya, Maṅgalarasa and Avanijanāśraya Pulakeśin are borrowed from those of the main Chālukya line. If the names of Nannarāja and his ancestors are identical with or similar to those of the predecessors and successors of Dantidurga, the presumption is possible that there was a close family connection between the two houses.

c7.s1.40

¹⁹ E. I., IX, pp. 195 ff.

c7.s1.41

²⁰ E. I., XVIII, p. 235.

c7.s1.42

It will be presently shown that the early exploits of Dantidurga and his father, Indra, e. g. the latter's feat of carrying away by force (*rākshasa vivāha*) a Chālukya princess from Kaira, Dantidurga's occupation, at the beginning of his career, of Gujrāt and northern Mahārāshtra and the defeat of the kings of Sindh, Mālvā and Kos'ala, would indicate that the family must have been ruling in the feudatory capacity somewhere in the central or western portion of Central India, prior to the rise of Dantidurga in c. 745 A.D. We learn from the Tivarkhed and the Multāi grants, that Nannarāja was also ruling in a portion of the territory above indicated. The Tivarkhed plates were issued from Achalapura, which is the same as Elichpur in Berar; Tivarkhed itself was granted away by the charter, and it is only 55 miles from Elichpur. The villages referred to in the Multāi plates have not yet been identified, but Multāi, where the plates were discovered, is only about 20 miles from Tivarkhed. Nannarāja and his ancestors were ruling in Berar and their capital was probably at Elichpur. The exploits of Dantidurga also can best be explained if we assume that his patrimony was somewhere in Berar. This province occupies a central position with reference to Kaira, Ujjain, Sirpur and Satara, where we know that his forces were operating at the beginning of his career.*

c7.s1.43

And finally, the known chronology of the two houses does not only not go against the proposed connection between them but does support it. Assuming an average reign of 20 years (except where the succession was not from father to son), and working back from the known dates of Kṛṣṇa I, we get the following genealogical and chronological table:—
(1) Durgarāja, c. 570-590 A.D.
(2) Govindarāja, son of No. 1, c. 590-610 A.D.
(3) Svāmikarāja, son of No. 2, c. 610-630 A.D.
(4) Nannarāja Yudhāsura, son of No. 3, c. 630-650 A.D. Known dates, 631 A.D. and (?) 709 A.D.
(5) Dantivarman, probably son of No. 4, 650-670 A.D.
(6) Indra Pṛchhakarāja, son of No. 5, c. 670-690 A.D.
(7) Govindarāja, son of No. 6, c. 690-710 A.D.
(8) Karka I, son of No. 7, c. 710-730 A.D.
(9) Indra I, son of No. 8, c. 730-745 A.D.
(10) Dantidurga, son of No. 9, c. 745-756 A.D. Known date, 754 A.D.
(11) Kṛṣṇa I, uncle of No. 10, c. 756-775 A.D. Known dates, 758, 768 and 772 A.D.

c7.s1.44

∗ See the adjoining map.

c7.s1.45

N. B.—In the case of Indra I and his son Dantidurga, a reign of 15 years only has been assumed, as Nanna Guṇavaloka, a younger brother of Indra I, was still alive in 792 A.D., as the Daulatabad plates show ²¹.

c7.s1.46

²¹ E. I., III, pp. 195 ff.

c7.s1.47

If we reject the date 709 A.D., supplied by the Multai plates, as not genuine, it will appear very probable that Nannarāja was the predecessor, and very likely the father, of Dantivarman, the earliest known ancestor of Dantidurga. If, on the other hand, that date has to be accepted as genuine, the probability would be that Dantivarman was a younger brother of Nannarāja Yudhāsura, ruling somewhere in Khāndesh, over part of the dominions of his elder brother. Nannarāja may have had no sons, or they and their descendants may have been eclipsed altogether by the successors of Dantivarman.

c7.s1.48

To sum up, the similarity and identity of the names of the members of the two families, the close contiguity of the places where they were ruling, the identity of the seal design, and the striking manner in which the known dates of the members of the two families can be worked up into a mutually adjusting chronological and genealogical table,—all these make it extremely probable, if not almost certain, that the ancestors of Dantidurga were ruling somewhere in Berar, and were either the direct or collateral descendants of the Rāṣṭrakūṭa king Nannarāja Yudhāsura, who was ruling at Elichpur in Berar in the middle of the 7th century A.D.

c7.s1.49

There remains to be considered the Rāṣṭrakūṭa family disclosed by the Antroli-Chharoli plates of Karkarāja II, dated 757 A.D.²² The following genealogy is supplied by the record:—

c7.s1.50

²² J. B. B. R. A. S., XVI, pp. 106 ff.

c7.s1.51

(1) Karkarāja I, c. 690-710 A.D.
(2) Dhruva, son of No. 1, c. 710-730 A.D.
(3) Govinda, son of No. 2, c. 730-750 A.D.
(4) Karka II, son of No. 3, c. 750-770 A.D. Known date 757 A.D.

c7.s1.52

From this record we learn that the village of Sthāvarapallikā, *i. e.* modern Antroli-Chharoli, was granted away by Karka II in 757 A.D. Since Antroli-Chharoli is 10 miles to the north-east of Surat, and since the donee hailed from Jambusāra in the Bharoch District, it is clear that Karka II was ruling over Surat and Bharoch districts. It is, however, almost certain that the immediate ancestors of Karka II were not living in this locality, or if they were doing so they were wielding no ruling powers. For, in the early decades of the 8th century A.D. the districts in question were being governed partly by the Gujrat Cālukyas, and partly by the Bharoch house of the Gurjaras. The rule of Karka II in southern Gujrat, therefore, must have been of recent origin.

c7.s1.53

The seal of Karka II of the Antroli-Chharoli record is the same as that of the main Rāṣṭrakūṭa branch, *viz.* eagle; the names of the ancestors of the donor are those, which frequently figure in the main Rāṣṭrakūṭa line. Karka II was a contemporary of Dantidurga and was ruling over a portion of Lāṭa, which the latter claims to have conquered. It is therefore quite clear that the house of Karka must have been closely connected with that of Dantidurga, but what precisely that relationship was we do not know. Bhagwānlāl Indraji has suggested²³ that Dhruva of the Antroli-Chharoli record may be assumed to be another brother of Indra I, the father of Dantidurga. On this hypothesis the genealogy of the main line with known dates will be as under:-

c7.s1.54

²³ J. B. B. R. A. S., XVI, pp. 106 ff.

c7.s1.55

Karka I

c7.s1.56

| . . . . .
| | | |

c7.s1.57

Dhruva Indra I Kṛṣṇa I Nannarāja

c7.s1.58

| | 758, 768 and 772 A.D. 793 A.D.

c7.s1.59

Govinda Dantidurga

c7.s1.60

| 754 A.D.

c7.s1.61

Karka II 757 A.D.

c7.s1.62

There is nothing impossible in the above genealogy and chronology. Nanna, the youngest brother of Karka I, we may assume, was born in c. 715 A.D. and Dhruva, his eldest brother, may have been his senior by about 20 years. Dhruva's son Govinda may have been born in c. 715 A.D. and his grandson Karka II, in c. 735 A.D. Indra may have been two years younger than Dhruva, and his son Dantidurga may have been born in c 720 A.D. Indra seems to have been the most ambitious of the four brothers and his son Dantidurga may have begun his military career in c. 745 A.D. when he was only about 25. Shivaji and Babar are known to have begun their careers at even an earlier age. Dantidurga's nephew may have actively assisted him in his conquests when he was only a youth of 20; the uncle may have rewarded the services of the nephew by appointing him the governor over the newly conquered province of Lāṭa. After the death of Dantidurga in c. 760 A.D. Kṛṣṇa, his eldest surviving uncle, may have succeeded him, being the most senior member of the house. The drawback in this theory is the necessity of assuming that Dhruva, Govinda and Karka II were, all of them, the oldest sons of their parents, and even after making this assumption we find that Karka II has to be assumed to have joined his uncle, when he was only just a boy of 18 or 20. It is also difficult to understand why a youth of about 20 should have been preferred by Dantidurga for the Gujrat viceroyalty to his two mature uncles.

c7.s1.63

Recently²⁴ Dr D. R. Bhandarkar has tentatively advanced the view²⁴ that Karka II and his father Govinda of the Antroli-Chharoli record may be identified with Karka and Govinda, father and grand-father respectively of the Indra I of the main Rāṣṭrakūṭa house. He points out that the genuineness of the Samangad plates of Dantidurga is not above suspicion and therefore we may ignore the date 754 A.D., supplied by this record for Dantidurga.

c7.s1.64

²⁴ E. I., XVIII, p. 238.

c7.s1.65

This view also is not free from difficulties. If Karka and his father Govinda of the Antroli-Chharoli record are to be identified with the father and grand-father of Kṛṣṇa I of the main line, Dhruva and Karka I of the former record will have to be identified with Indra I and Dantivarman of the main house. We shall have to assume that either Dhruva and Karka were the *birudas* of Indra I and Dantivarman or vice versa. Rāṣṭrakūṭa kings have used several *birudas*, but there is not a single instance of any of the names Dhruva, Karka, Indra or Dantivarman being used as a *biruda*. So far these names have been known only as personal names.

c7.s1.66

In the second place the known dates of Dantidurga and Kṛṣṇa I militate against this view. Even supposing it to be proved that the Samangad record is a later forgery, the chronological difficulties involved in this theory are not solved. The new plates²⁵ published by the Bhārata Itihāsa Samshodhaka Mandala now supply 758 A.D. as the earliest date for Kṛṣṇa I, and if we accept 757 A.D. as the date of Dantidurga's grandfather Karka, the reigns of Dantidurga and his father Indra will have to be crammed in the incredibly short space of about a year! The reigns of both these rulers were eventful; Indra I had come into hostile contact with the Gujrat Chālukyas, and Dantidurga had in the course of a brilliant career defeated the kings of Mālvā, Sindh, Kośala, Lāṭa and Śris'aila, besides annexing the northern dominions of the early Chālukyas. How could all this be achieved in the course of a single year? Even if we assume that Indra had predeceased his father, the difficulty would not be got over, for one year is too short a period for the achievements of Dantidurga.

c7.s1.67

²⁵ B. I. S. M. J., VIII, pp. 165-8.

c7.s1.68

A third possible view in this connection is to regard Karka I of the Antroli-Chharoli record as a younger brother of Indra Pṛhakarāja of the main line, and to make his descendants Dhruva, Govinda and Karka II contemporary cousins of Govinda, Karka and Indra of the main line. This view presents no chronological difficulties like those in the first two theories; Karka II, according to this theory, becomes an elder contemporary of Dantidurga and he may quite possibly have rendered him valuable assistance in his military plans, in return for which Dantidurga may have appointed him his deputy in southern Gujrat. It must, however, be admitted that the assumption that Karka I was a younger brother of Indra I is based upon mere conjecture. It has to be confessed that the precise relation of the main line with the Rāṣṭrakūṭas of the Antroli-Chharoli record still remains to be discovered. That the two houses were related in some way is fairly certain; the ill-behaved relative, whom Kṛṣṇa I ousted soon after his accession, was very probably either Karka II or his successor.

c7.s1.69

The Naravana plates of Vikramāditya II, dated Saka 664, (January 743 A.D.)²⁶ state that the village Naravana in Chiplun taluka of Ratnāgiri district was given to certain Brāhmanas by that Chālukya king at the request of Rāṣṭrakūṭa Govindarāja, son of Śivarāja. The charter was issued when the king was encamped at Ādityavāṭika or modern Aitavade in Satara district. It is not possible to connect this Rāṣṭrakūṭa Govindarāja either with the main line or with the ancestors of Karka II of the Antroli-Chharoli record. He cannot be Govinda of the main line, for the latter was great-grand-father of Dantidurga and he could not have been possibly alive in 743 A.D., when the Naravana plates were issued. Nor can we identify him with Govinda; the father of Karka II of the Antroli-Chharoli record. It is true that no chronological difficulty arises in connection with this identification; for the known date of Karka II, viz. 757 A.D., fits in well with the known date of Govinda of the Naravana plates, viz. 743 A.D. The father of Govindarāja may have been an officer under Vikramāditya II in northern Konkan and his son may have carved out a principality in southern Gujrat at the downfall of the Chālukyas. There is, however, an almost fatal objection to this view; the father of Govindarāja of the Naravana plates is Śivarāja, whereas that of Govindarāja of the Antroli-Chharoli record is Dhruvarāja. There is no evidence so far forthcoming to indicate that either Dhruvarāja or Śivarāja was ever used as a biruda by the Rāṣṭrakūṭas. It may be also pointed out that Govindarāja of the Antroli-Chharoli record was using feudatory titles, whereas Govindarāja of the Naravana plates has not even the title of a Sāmanta. The latter was probably a mere district officer ruling over some portions, either of Satara or of Ratnagiri district. It is, therefore, very likely that he was in no way connected with the Rāṣṭrakūṭa family that later ousted the Chālukyas.

c7.s1.70

²⁶ B. I. S. M. J., X, pp. 9 ff.

c7.s1.71

Having indicated the probable relationship of the main Rāṣṭrakūṭa line with earlier and contemporary Rāṣṭrakūṭa ruling families, let us now consider the question of the stock and nationality of the house of Dantidurga. We may refer only very briefly to mythological or semi-mythological views in this connection. Later Rāṣṭrakūṭa records claim that the dynasty was descended from the race of Yadu. Bhagwānlāl Indraji had conjectured that this theory was started in c. 930 A.D., to explain the change in the emblem of the Rāṣṭrakūṭa seal from lion to eagle, which is the vāhana of Viṣṇu. This suggestion, however, is not happy; lion was the emblem of the Rāṣṭrakūṭas of Mānapura and not of Malkhed. The seal-emblems on the early Rāṣṭrakūṭa records, e g. the Alas plates of Yuvarāja Govinda, the Paiṭhan plates of Govinda III, are of Garuḍa or eagle. On some of the later records the emblem is that of seated Śiva but the lion emblem figures nowhere.

c7.s1.72

The earliest date, now known, when the Rāṣṭrakūṭas are seen claiming descent from Yadu is 871 A.D., when the Sanjan plates were issued. 63 years earlier, however, the Rāṣṭrakūṭas had not even dreamt of claiming this descent. The Wani-Dindori plates of Govinda III, dated 808 A.D., while referring to the birth of that monarch, observe that when he came on the horizon the Rāṣṭrakūṭa race became invincible like the race of the Yadus when Murāri was born in it. (27) It is fairly certain that the birth of Govinda III in the Rāṣṭrakūṭa family would not have been compared to that of Kṛṣṇa in the Yadu race, if the Rāṣṭrakūṭas had at that time thought of claiming connection with it by descent. It is probably the simile in this verse, (28) which suggested to the later kings and poets the idea of claiming a descent from the Yadus, especially since it seems fairly probable, from the seal emblem of the eagle, that the earlier rulers were, like the Imperial Guptas, Vaiṣṇavas.

c7.s1.73

R. G. Bhandarkar had suggested that the Rāṣṭrakūṭas had probably sprung from a family that was known by the name of Tuṅga (29), since in the Karhad (30) and Deoli plates (31) of Kṛṣṇa III it is stated that the family had its origin from a king named Tunga. The above plates, however, mention in the immediately next line Raṭṭa as a descendant or perhaps a son of Tunga, and observe that the Rāṣṭrakūṭa family was known after him. Both Tunga and Raṭṭa are imaginary rather than real rulers, and even if we suppose that the family was descended from them, we find ourselves in no better position about the solution of the question of the stock and the original home of the Rāṣṭrakūṭas.

c7.s1.74

²⁷ I. A., XI, pp. 157 ff.

c7.s1.75

²⁸ Cf. Yasmin sarvaguṇāś'raye kṣitipatau S'ri-Rāṣṭrakūṭānvaye| Jāte Yādavavans'avanmadhuripāu āsīdalanghyah paraih ||
29. E. I. IV, pp. 278 ff.

c7.s1.76

³⁰ E. I., V, pp. 192 ff.

c7.s1.77

Fleet has suggested (32) that since the names Rāṭhor and Rāṭhod are to be derived from the term Rāṣṭrakūṭa, we may connect the Rāṣṭrakūṭas with the Rājputānā-Kanauj country, which seems to be the original habitat of the Rāṭhor clans of the Rajputs. But the Rāṭhors come to our notice much later than the southern Rāṣṭrakūṭa families, and it is quite possible that the Rajput Rāṭhors may be the descendants of some members of the Deccan Rāṣṭrakūṭa families, left behind in northern India during the northern campaigns of Dhruva I, Govinda III, Indra III and Kṛṣṇa III.

c7.s1.78

Burnell was inclined to hold that the Malkhed Rāṣṭrakūṭas were Telgus and were of the same stock as the Reddies of the Andhra land. (33) This view, however, does not bear close scrutiny. The Reddies are at present scattered even over Tamil country and north-eastern portions of Mysore State; but their original home and present stronghold is the Andhra country. The Reddies of Mysore are undoubtedly of Telgu origin, (34) and those in Tamil districts still speak a broken Telgu dialect, which clearly proves their northern origin. If we suppose that the Rāṣṭrakūṭas were the ancestors of the modern Reddies, their original home will have to be located somewhere in the Krishna-Godavari doab. In that case it is reasonable to expect that they would have first come into prominence in that locality. As it is, not only did the Rāṣṭrakūṭa expansion not begin from the Telgu-speaking area, but most of it was never included in the Rāṣṭrakūṭa dominions. The Chālukya rulers overthrown by them were those of the main and the Gujrat lines. The Vengi kings continued to defy them down to their fall. The mother-tongue of the Rāṣṭrakūṭas was Canarese and not Telgu, as will be presently shown. The Reddies were a class of traders and cultivators and they are not known to have distinguished themselves by military exploits in any period of ancient Indian history. It is but once that they are known to have founded a kingdom, and that was after the fall of the Gaṇapatis of Warangal, when for about a century, c. 1350-1450 A.D., they were holding portions of Kṛṣṇā and Rājmahendri districts.³⁵ The change of Rāṣṭra into Raḍḍa or Reddies is also not possible in the Telgu dialect, though it can take place in the Tamil one.³⁵ Under these circumstances it is not possible to identify the Rāṣṭrakūṭas with the ancestors of modern Reddies.

c7.s1.79

³² B. G. I. ii, p. 384.

c7.s1.80

³³ South Indian Palaeography, p. 10.

c7.s1.81

³⁴ Imperial Gazetteer, XVIII, p. 191.

c7.s1.82

³⁵ *Ibid*, VII, p. 158.

c7.s1.83

Mr. C. V. Vaidya holds that the Rāṣṭrakūṭas of Malkhed were a Marathi-speaking family, and were the ancestors of the modern Marathas.³⁶ This view also seems to be untenable, for it can be shown that the family belonged to the Canarese stock and its mother-tongue was Canarese.

c7.s1.84

³⁶ *His. ry of Medieval Hindu India*, II p. 249 (Marathi Edition).

c7.s1.85

A survey of earlier history is necessary in order to decide whether the Rāṣṭrakūṭas of Malkhed hailed from Mahārāṣṭra or Karṇāṭak. It is as early as the time of Aśoka that we first hear of the Raṭhikas, a term to which we have to trace the Raṭṭas and Rāṣṭrakūṭas of later centuries. There is no philological difficulty in postulating this derivation; the name is spelt as Riṣṭika at Girnar, Raṣṭika at Shahabazgarhi, and Raṭraka at Mansera, showing that all these terms are to be derived from the Sanskrit term Rāṣṭrika. It is true that there is no trace of the sibilant of the original word in the terms Raṭṭika or Raṭṭa, but the Mansera form Raṭraka shows that the term Rāṣṭrika assumed a form altogether devoid of any sibilant in some dialects. As a matter of fact the form Raṭraka of Mansera supplies us the link between Rāṣṭrika and Raṭṭika or Raṭṭa. Pischel besides quotes several cases of a Sanskrit ṣṭa becoming a simple ṭṭa in Prakrits. Cf. Skt. Iṣṭa; M., J. M. and Amg. Iṭṭa. Skt. Auṣṭrika; J. M. Uṭṭiya. Skt. Sandaṣṭa; M. Sandaṭṭa. (37)

c7.s1.86

It will be seen that this tendency to change the original ṣṭa into ṭṭa is to be mainly seen in the Mahārāṣṭrī and Jaina-Mahārāṣṭrī, and it is precisely in the south, where these dialects were spoken, that we find the change of original Rāṣṭrika into Raṭṭa.

c7.s1.87

In my opinion the various Raṭṭa or Rāṣṭrakūṭa families of our period were the descendants of some of the Rathika families, that were ruling over small tracts in the feudatory capacity since the time of Aśoka. Aśoka inscriptions describe Rathikas as westerners, but they also associate them with the Bhojakas, which will show that they were occupying portions of Mahārāṣṭra and Berar as well. The next reference to them is in the Nānāghaṭ inscription of Queen Nayanikā, where we learn that she was the daughter of Mahārathi Traṇakayira, whose statue was erected by the side of his daughter at Nānāghaṭ. There existed at this time numerous feudatory rulers known as rathikas. In connection with the western expedition of king Khāravela his Khaṇḍagiri inscription tells us that he carried away the wealth and crowns of all the rathikas and bhojakas. There are two records at Kārli (38) belonging to a little later period, recording the benefactions of Mahārathi Gotiputa Agimitanaka and Mahārathi Vasithiputa Somadeva. The latter grants a village which shows that he was a ruling chief. A Bhājā record (39) discloses the existence of a Mahārathi Viṇhudatta, and a Kaṇheri one that of a Mahāraṭhinī Nāgamūlanikā, who was the daughter of a mahārāja and sister of a mahābhoja. This record again shows, like that of king Khāravela, that raṭhis and mahāraṭhis were feudal chiefs, and not mere generals or local district officers.

c7.s1.88

³⁷ Grammatik der Prakrit-Sprachen, Section 304.

c7.s1.89

³⁸ J. B. B. R. A. S., V, pp. 152-3.

c7.s1.90

³⁹ Ib., p. 62.

c7.s1.91

It is usually supposed that the Raṭhis and Mahāraṭhis were in power in Mahārāṣṭra only, but there is definite evidence to show that some of them were occupying portions of Karṇāṭak as well. Lead coins bearing the legend ‘Sādakani-Kaḷalāya-Mahāraṭhi’ have been found near Chitaldurga.⁴⁰ These coins belong to the middle of the 3rd century A.D. The Hirahadagalli grant of Dharmamahārājadhirāja Śivaskandavarman⁴¹ is addressed, among others, to Raṭhikas. We further find that some of the Mahāraṭhis were closely connected with Canarese families. The cumulative evidence of two Kaṇheri records shows that Nāgamūlanikā, who was married to a Mahāraṭhi, was the daughter of Hāritiputra Viṣṇu-kada Chūtu-Sātakarṇi, who was a Canarese prince ruling at Banavasi.⁴² Some of the Mahāraṭhis were Nāga-worshippers; the name of the queen of Sātakarṇi I, Nayanikā, or Nāganikā, suggests that her Mahāraṭhi father was a Nāga-worshipper. One of the lion-pillars at Karli was the gift of a Mahāraṭhi, Agimita-naka or Agimitra-nāga, whose name also shows that he was a Nāga-worshipper. The donor of one of the Kaṇheri inscriptions, who was the wife of a Mahāraṭhi, is Nāgamūlanikā or Nāgamūlanikā and her son’s name was Skandanāga.⁴³ One of the inscriptions of her father Viṇhu-kaḍa-chuṭu, discovered at Banavasi, is upon a stone slab bearing the representation of a five-headed cobra. Nāga-worship was extensively practised by the early inhabitants of Mysore,⁴⁴ and since some of the Mahāraṭhis were Nāga-worshippers and connected by marriage with Canarese families, we are justified in concluding that Raṭhis and Mahāraṭhis were in power also in parts of Karṇatak, especially since coins of Sadakana Kaḷalāya Mahāraṭhi are discovered in the heart of Karṇatak near Chitaldurga. In the face of these facts it can no longer be maintained that the Raṭhi and Mahāraṭhi families were confined only to Mahārāṣṭra.

c7.s1.92

⁴⁰ Rapson, *Catalogue*, pl 57, Plate VIII, No. 233.

c7.s1.93

⁴¹ E. I., I, P. 2.

c7.s1.94

⁴² Rapson, *Catalogue*, p. LIII, Luders, *List*, No. 1021.

c7.s1.95

⁴³ A. S. W.I., V, p. 86.

c7.s1.96

⁴⁴ Rice, *Mysore and Coorg from Inscriptions*, p. 202.

c7.s1.97

We can now take up the question as to whether the Rāṣṭrakūṭas of Malkhed originally belonged to Mahārāṣṭra or Karṇatak. We have shown already how the main line, that was later established at Malkhed, has to be connected with the Rāṣṭrakūṭa family ruling at Elichpur. But Dantidurga and his ancestors were not natives of Berar. Canarese was their mother-tongue. It was Canarese and not Mahārāṣṭrī literature that flourished at their court.⁴⁵ Amoghavarsha I of the line was either himself the author, or at least the inspirer, of the oldest Canarese work on poetics. The sign-manuals of Karka and Dhruva of the Gujrat branch of the house in the Naosāri plates of 816 A.D.⁴⁶ and the Baroda grants of 812⁴⁷ and 835⁴⁸ A.D. are in the south-Indian proto-Canarese characters, whereas the records themselves are to be seen inscribed in the usual script of the locality of the period in question. If the home of the Malkhed Rāṣṭrakūṭas were in Mahārāṣṭra, it is difficult to explain how they could be using the script of Karṇatak as their mother script. It is true that much earlier than 812 A.D. the Rāṣṭrakūṭa empire had embraced the whole of Karṇatak, but if the family had originally belonged to Mahārāṣṭra, its members deputed to rule over southern Gujrat could not be seen using for their sign-manual a script that was current neither in southern Gujrat, nor in Mahārāṣṭra but in Karṇatak. The fact that the recently published Jura inscription of Kṛṣṇa III, found in Bundelkhand, should be using the Canarese language to describe his achievements can also be explained only on the assumption that Canarese was the mother tongue of the Rāṣṭrakūṭas of Malkhed.

c7.s1.98

⁴⁵ See Supra, chapter XV.

c7.s1.99

⁴⁶ J. B. B. R. A. S., XX, p. 135.

c7.s1.100

⁴⁷ I. A., XII, p. 157. Plate III a.

c7.s1.101

⁴⁸ Ibid, XIV, p. 197.

c7.s1.102

The use of the Canarese script and language by the members of the family of Dantidurga is not, however, inconsistent with the theory, here advanced for the first time, that Dantidurga's ancestors were direct or collateral descendants of Nannarāja of Elichpur. It is shown already how a number of Raṭhi families were long domiciled in Karṇāṭak even before the 3rd century A.D. Under the Cālukyas of Badāmi, a Canarese-speaking dynasty was founded in southern Gujrat, and another in the Telgu-speaking Andhra country. The predecessors of Dantidurga may similarly have carved a domain in Berar away from their home in Karṇāṭak.

c7.s1.103

There is also evidence available to show that the home of the house of Dantidurga was in a Canarese-speaking locality. A number of Rāṣṭrakūṭa records describe the rulers by the epithet *Laṭṭalūrapuravarādhīśa* 'lord of Laṭṭalūra, the best of towns.'⁴⁹ This epithet is changed into *Laṭṭalūrapuravinirgata* 'emigrants from Laṭṭalūra, the best of towns' in some of the records of the Raṭṭas of Saundatti.⁵⁰ Laṭṭalūra, to which the Rāṣṭrakūṭas belonged, need not have been necessarily included within the jurisdiction of Nannarāja and his ancestors. Maṇalera, a general of Kṛṣṇa III, has been described as *Valabhipuravarādhis'a*, 'lord of Valabhi, the best of towns' at a time when it is clear that he had no control over that city.⁵¹ The Kalachurya kings of Kalyāni, the Gutta princes of Guttavolal, and the Yādava rulers of Devagiri called themselves lords of Kālañjara, Ujjayinī, and Dvārakā respectively,—places which were never included in their dominions. The identification of Laṭṭalūra is not yet definitely certain. Fleet had once suggested that Laṭṭalūra may be Ratanpur in Vilaspur District of the Central Provinces, pointing out how most of the Rāṣṭrakūṭa families belonged to Central India. But later on he himself abandoned this theory, when it was found from a 12th century inscription from Ratanpur itself, that the name of the place was originally Ratnapura and not Laṭṭanūrapura. His later view that Laṭṭalūra may be Lāṭūr in Bedar District of Hydrabad State seems to be hitting the truth.⁵² Phonetically the change of Laṭṭalūra into Lāṭūr is quite natural. The second *la* in Laṭṭalūra, being dropped by haplology, the original form will assume the garb of Laṭṭaūra in Prakrit; this form will later change into Lāṭūra, the loss of the double consonant *ṭṭa* being compensated by the doubling of the preceding vowel.

c7.s1.104

⁴⁹ Sirur and Nilgund inscriptions of Amoghavarṣa I, I.A., XII, p. 218; and E. I., VI, p. 102.

c7.s1.105

⁵⁰ J. B. B. R. A. S., X, pp. 167 ff. In two of the Saundatti Raṭṭa records Kṛṣṇa the founder of the house is described as *Khaṇḍarapuravarādhīśa*. This, however, cannot show that Kandharapura was the home of the Rāṣṭrakūṭas, since the vast majority of records describe them as *Laṭṭalūrapuravarādhīśa*. It is likely that Kandharapura was only an imaginary city since we do not hear of it anywhere else. If at all it was a real city, Kṛṣṇa III alone may have been associated with it; it was not the home of the family.

c7.s1.106

⁵¹ E. I., VI, pp. 56 ff.

c7.s1.107

⁵² E. I., VII, pp. 186 ff.

c7.s1.108

Since there is no phonetic difficulty, there is nothing to stand against the identification of Laṭṭalūra with Lāṭūr. As a matter of fact all the known facts of history can be explained only on this hypothesis. It has been shown already that the Raṭhis and the Mahāraṭhis were spread over portions of Karṇāṭak as well, in the early centuries of the Christian era. So there may, quite possibly, have been a Raṭhi family holding local sway at Lāṭūr, which is almost due east of Poona and south of Berar. This family may have later migrated to Elichpur or some other place nearby in Berar, where Nannarāja was ruling in 631-2 A.D. Elichpur is only about 150 miles north of Lāṭūr, and the emigration is quite possible. And by assuming that Dantidurga's ancestors in Berar were immigrants from Lāṭūr, we can well understand how they were using Canarese language and proto-Canarese script. Lāṭūr is now a Canarese-speaking town and such was also the case during our period; for we learn from the Kavirājamārga, which was composed in the 9th century A.D., that Canarese was then spoken between the Godāvari and the Kāverī.⁵³ The immigrant rulers naturally stuck to their language and script during their stay in Berar. When we remember how the Maratha families ruling in Baroda, Indore, and Gwallior still use their mother tongue and script in personal matters, there is nothing impossible in the supposition that the Canarese Rāṣṭrakūṭa family, ruling in Berar, which was and is just contiguous to Canarese-speaking area, preserved its mother tongue and script.

c7.s1.109

⁵³ Kavirājamārga, I, 46.

c7.s1.110

If we assume with Fleet and Vaidya that the home of the Malkhed Rāṣṭrakūṭas was either in Mahārāṣṭra or somewhere in Central India, we cannot explain how they were using the Canarese language and the southern script. If we assume that they were local chiefs somewhere in central or southern Karnatak, we cannot understand how Dantidurga's exploits were most of them performed in Gujrat, Malva, Central India, and northern Mahārāṣṭra, and how the Chalukyas of Badāmi continued to hold the southern portions of their dominions down to the reign of Kṛṣṇa I. If on the other hand, we assume that they were immigrants in Berar from Lāṭūr, we can understand why they call themselves Laṭṭalūra-puravarādhīśa, why they were using the proto-Canarese script and the Canarese language, why their early exploits were performed in Central India, Gujrat and northern Mahārāṣṭra and why the early Chalukyas continued to hold their own so long in the southern portion of their kingdom.

c7.s1.111

It will have become already clear from what has been said so far, that the terms raṭhi, mahāraṭhi, raṭhika, rāṣṭriya, rāṣṭrapati and rāṣṭrakūṭa possessed merely political and administrative significance; none of them was used to denote any tribal or ethnical stock. R. G. Bhandarkar has argued that the name Mahārāṣṭra is due to the province being occupied in the early days by the raṭhis and the mahāraṭhis,(54) but with due deference to the views of that learned scholar, it must be confessed that there is no evidence to show that the raṭhis and the mahāraṭhis were the names of any tribes, Aryan, non-Aryan or mixed. Mr. C. V. Vaidya holds that the Rāṣṭrakūṭas represent the captains of the Aryan army, who parcelled out the districts in Mahārāṣṭra among themselves. This view again ignores the presence of the Rāṣṭrakūṭas in Andhra country, and of the raṭhis and the mahāraṭhis in Karnatak right upto Chitaldurga. The Nāga worship, which prevailed extensively among the raṭhi and the mahāraṭhi families, is surely no argument to prove their Aryan origin. We must, therefore, conclude that the facts so far known do not support the view that the Rāṣṭrakūṭas belonged to any particular stock, either Aryan or non-Aryan.

c7.s1.112

A point of minor importance remains to be considered, viz. what was the real and original name of the Rāṣṭrakūṭa family. After an elaborate analysis of the epigraphical evidence upon the point,(55) Fleet has come to the conclusion that 'while Rāṣṭrakūṭa was the formal appellation which it was customary to apply to the kings of Malkhed in ornate language, the real practical form of the family name was Raṭṭa.' He then points out that the form Raṭṭa was not derived from the form Rāṣṭra or Rāṣṭrakūṭa, but that the reverse was the case.

c7.s1.113

This controversy as to the original form of the word Rāṣṭrakūṭa is partly due to a misunderstanding as to what we mean by the term 'original.' Fleet, while maintaining that Raṭṭa was not derived from Rāṣṭra, does not assert that Raṭṭa is a des'ī or Dravidian word of non-Aryan parentage. We have already pointed out a number of cases, where original ṣṭa in Sanskrit becomes simple ṭṭa in Mahārāṣṭrī and Jain- Mahārāṣṭrī.⁵⁶ There is, therefore, no phonetic difficulty in deriving the term Raṭṭa from the Sanskrit term Rāṣṭra or Rāṣṭrakūṭa. As to the question which of these was the original form, the answer will entirely depend upon what we mean by ‘original’. From the 3rd century B.C. to about the 4th century A.D., Prakrits were usually used for epigraphical purposes, and perhaps for daily intercourse as well. The form raṭhi was therefore the original or earlier form in vogue, when the terms Rāṣṭrakūṭas and Rāṣṭrapatis began to be generally used on the use of Sanskrit becoming common for epigraphical purposes. In this sense Raṭhi or Raṭṭa is no doubt the earlier form, but philologically and historically the term Raṭṭa has to be admitted as a derivation from and contraction of the term Rāṣṭrakūṭa.

c7.s1.114

⁵⁴ B. G., I, ii, p. 143.

c7.s1.115

⁵⁶ Ante, p. 19.

c7.s1.116

The Raṭhis and the Mahāraṭhis were, as we have seen above, small local rulers. The Hindu imperial theory and practice were generally against the policy of annexation, as will be shown in Chapter XII, Section B; the local chiefs were often converted into district officers and vice versa, according to the changes in the political situation. The terms Raṭhika, Rāṣṭriya, Rāṣṭrapati and Rāṣṭrakūṭa were used to denote local chiefs, district officers, and big landlords. From the Girnar inscription of Rudradāman we learn that the Mauryan governor in Kathiawar was styled Rāṣṭriya. From the Hirahadagalli grant of S’ivaskandavarman we find that district officers were known as Raṭhikas.⁵⁷ Numerous Vengi records show that the term Rāṣṭrakūṭa denoted important landlords in Andhra country during our period.⁵⁸ A number of Gurjara and Valabhi plates show that Rāṣṭrapati was the title of the district governor.⁵⁹ In the recently published Naravana plates of Vikramāditya II,⁶⁰ Govindarāja, who is called Rāṣṭrakūṭa in the record, may have been either a district officer or a big landlord, who wielded a fair amount of influence with the Cālukya emperor.

c7.s1.117

⁵⁷ E. I., I, pp. 2 ff.

c7.s1.118

⁵⁸ E. I., V, p. 122, 133 ; I. A., VII, pp. 186 ff. etc.

c7.s1.119

⁵⁹ I. A., X, p. 284 ; XIII, p. 15; p. 77 etc.

c7.s1.120

⁶⁰ B. I. S. M. J., X, pp. 9 ff.

c7.s1.121

Whether the Rāṣṭrakūṭas of Malkhed were originally local rulers or district officers or mere landlords is difficult to judge. Many of the Raṭhis of the 3rd century B.C. must have been reduced to the position of district governors at the rise of the Andhra empire. Some of them must have again become petty local rulers with the fall of that empire. Their status in many cases must have been again reduced by the early Cālukyas, when they rose to power. With reference to the ancestors of Dantidurga we notice that Yudhāṣura Nannarāja was the first to claim the Pañcamahāś’bdas. His predecessors, therefore, may have been either local officers or mere landlords. They rose to power first under Nannarāja, and it is possible that his descendants may have continued to enjoy the feudatory status, won for the family by Nannarāja.

First 121 of 137 passages. Read on

Linked texts (19)