Pramāṇa · Search · About · API

The Dynastic History of Northern India (Early Mediaeval Period) v1.c20: Chapter X. The Later Gurjara-Pratīhāras of Kanauj

The Dynastic History of Northern India (Early Mediaeval Period) v1.c20

Language
English

Contents

Chapter X. The Later Gurjara-Pratīhāras of Kanauj

v1.c20.2

The Kānyakubja empire of the Puṣpabhūtis did not long survive the death of Harṣavardhana in about 647 A.D. His death introduces us to one of those periods of turmoil which inevitably follow the disruption of a strong empire and precede the foundation of a new one in Indian history. It is very difficult to give any connected account of political happenings in Northern India for about two hundred years after this event. The Tibetan invasion and control of the Ganges valley, the attempted revival of the Gupta empire by Ādityasena and of the Kanauj empire by Yaśovarman are some of the more important incidents of the first half of this period. The hope of an early revival of the hegemony of Kanauj however was shattered by the ambition of the Kārkoṭa kings of Kashmir. Then followed what may be described as a triangular contest between the Pālas of Bengal and Bihar, the Gurjara-Pratīhāras of Malwa and Western India, and the Rāṣṭrakūṭas of the Deccan for the possession of the Ganges-Jumna valley and the city of Kanauj. After a struggle of about half a century, in which fortune showed her fickleness by alternately favouring each of the combatants, she at last became the bride of the Gurjara-Pratīhāras. Some time before the year 836 A.D. they captured Kanauj, and, thanks to the achievements of two of their ablest chiefs, Bhoja and Mahendrapāla, at last succeeded in re-establishing an empire in India that in extent rivalled, if it did not exceed, that of the Guptas and the Puṣpabhūtis. There is unquestioned epigraphic evidence to show that the last of these two princes ruled over an area which extended from the Karnal district in the Punjab to Bihar and from the Kathiawar peninsula to Northern Bengal.¹ The strength and power of the Gurjara-Pratīhāras are amply attested by the account of the contemporary Arab writers.¹ The Arab satraps of Sind soon found their way blocked by the mighty arm of these rulers, and came to regard them as the greatest foe of the Muḥammadan faith. Though there is at present little evidence, it may be surmised that but for the powerful aid of the Rāṣṭrakūṭa rulers of the Deccan, the Arab power in the last half of the 9th century would probably have been crushed in Sind.

v1.c20.3

¹ The important inscriptions and dates for the history of the Gurjara-Pratīhāras up to the reign of Mahendrapāla are as follows :
(1) Jodhpur stone-inscription of Pratīhāra Bāuka, V. S. 894, *EI*, XVIII, 87-99.
(2) Ghaṭayāla stone-inscription of Pratīhāra Kakkuka, V.S. 918, *JRAS*, 1895, 513-21.
(3) 4 Ghaṭayāla pillar stone-inscription of Pratīhāra Kakkuka, V. S. 918, *EI*, IX, 277-81.
(4) The five records of the Gurjaras of Broach : (*i-ii*) Two Kaira grants of Dadda II, K.S. 380 and 385, *IA*, Vol. XIII, pp. 82-91. (*iii*) Sankheda grant of the time of Dadda II, K.S. 391, *EI*, Vol. II pp. 20-21. (*iv*) Nausari grant of Jayabhaṭa III, K.S. 456, *IA*, Vol. XIII, pp. 77-81. (*v*) The second plate of the Kavi grant of the same, K.S. 486, *IA*, Vol. V, pp. 113-15.
(5) The Sankheda Gurjara grant (second half only) of (Dadda I?), K.S. 346, *EI*, Vol. II, pp. 19-20.
(6) The three records of the Gurjaras of Broach held to be spurious by Bhagwanlal Indraji and Fleet : (*i*) Umeta grant of Dadda II, *Śaka*. S. 400, *IA*, Vol. VII, pp. 61-66. (*ii*) Bagumra grant of the same, *Śaka*. S. 417, *IA*, Vol. XVII, pp. 183-200. (*iii*) Ilao grant of the same, *Śaka*. S. 417, *IA*, Vol. XIII, pp. 116-19.
(7) Hansot plates of the Cāhamāna Bhartṛvaddha of the time of Nāgavaloka, V.S. 813. *EI*, Vol. XII, pp. 197-204; Nāgāvaloka is identified with Nāgabhaṭa I, *IA*, Vol. 40, 1911, p. 240.
(8) Osia stone-inscription of Vatsarāja, *JRAS*, 1907, Part II, pp. 1010-11.
(9) The passage in *Jaina-Harivaṁśa* giving the date *Śaka*. S. 705 for Vatsa king of Avanti; see *IA*, XV, pp. 141-43; *BG*, Vol. I. Part II, p. 197 and fn. 2; *EI*, Vol. VI, pp. 195-96.
(10) Buchkala stone-inscription of Nāgabhaṭṭa II, V.S. 872, *JRAS*, 1907, Part II, p. 1011. The date is wrongly given by Marshall as 892 but see *EI*, Vol. IX, p. 4, fn. 1.
(11) Barah copper-plate of Bhojadeva, V.S. 893, *EI*, Vol. XIX, pp. 15-19.
(12) Daulatpura plate of the same, V.S. 900, *EI*, Vol. V, pp. 208-13; see for the correction of the date, *JBRAS*, Vol. XXI, pp. 410 ff.
(13) Deogadh pillar-inscription of Bhoja, V. S. 919 and *Śaka* S. 784, *EI*, Vol. IV, pp. 309-10. (14-15) The two Gwalior Caturbhuja temple inscriptions of the same, V.S. 932 and 933. *EI*, Vol. I, pp. 154-62.
(16) Gwalior (Sāgar-Tāl) stone-inscription of the same, *EI*, Vol. XVIII, pp. 99-114; *ASI*, 1903-04, pp. 277-85.
(17) Ahar stone-inscription of the same, *Harṣa*. S. 259. *EI*, Vol. XIX, pp. 52-62.
(18) Delhi fragmentary stone-inscription of Bhoja *RMR*, 1924, p. 3.
(19) Pehowa Garībnāth temple stone-inscription of the same, *Harṣa* S. 276, *EI*, Vol. I, pp. 184-90.

v1.c20.4

¹ *Elliot*, Vol. I, p. 4. See also ‘Dynastic History of Sind,’ *supra*, pp. 10-11 and 15-16.

v1.c20.5

Mahendrapāla died some time after 907-08 A.D. The chronological arrangement of the princes who succeeded him is

v1.c20.6

(20) Barton Museum, Bhavnagar fragmentary stone-inscription of the time of Bhoja (?), *EI*, Vol. XIX, pp. 174-77.
(21-22) The two Una grants of the time of Mahendrapāla, *Valabhī*. S. 574 and V.S. 956, *EI*, Vol. IX, pp. 1-10.
(23) Dighwa-Dubauli plate of the same, V.S. 955, *IA*, Vol. XV, pp. 105-13; *JBRAS*, Vol. XXI, pp. 405 ff.
(24) Siyadoni stone-inscription of the same, V.S. 960 and 964, *EI*, Vol I, pp. 162-79.
(25) Pehowa *praśasti* of the same, *EI*, Vol. I, pp. 242-50.
(26) Ram-Gaya stone-inscription of the same, year 8, *MASB*, Vol. V, No. 3, p. 64.
(27) Guneria stone-inscription of the same, year 9, *MASB*, Vol. V, No. 3, p. 64.
(28-29) Two British Museum stone-inscriptions of the same, years 2 and 6, *NKGWG*, 1904, pp. 210-11.
(30) Itkhori stone image inscription of the same, *ASI*, 1920-21, p. 35.
(31) Paharpura (N. Bengal) votive stone pillar-inscription of the same, 5th year, not yet edited; noticed in *ASI*, 1925-26, p. 141.

v1.c20.7

*Non-Gurjara Inscriptions* :

v1.c20.8

(1) Aihole inscription of Pulakeśin II, *Śaka*. S. 556, *EI*, Vol. VI, pp. 1-12.
(2) Radhanpur plates of Govinda III, *Śaka*. S. 730, *EI*, Vol. VI, pp. 239-51.
(3) Wani grant of the same, *Śaka*. S. 730 (for 728), *IA*, Vol. XI, pp. 157-63.
(4) Sanjan plates of Amoghavarṣa, *EI*, Vol. XVIII, pp. 235-57.
(5) Pathari pillar-inscription of Parabala, V.S. 917, *EI*, Vol. IX, pp. 248-56 and *IA*, 1911, pp. 239-40.
(6) The Baroda grant of Karka-Suvarṇavarṣa, *Śaka*. S. 734, *IA*, Vol. XII, pp. 156-65.
(7) The Nilgund inscription of Amoghavarṣa, *Śaka*. S. 788, *EI*, Vol. VI, 98-108.
(8) Khalimpur plate of Dharmapāla, *EI*, Vol. IV, pp. 243-54.
(9) Monghyr grant of Devapāla, *EI*, Vol. XVIII, pp. 304-07.
(10) Bhagalpur plate of Nārāyaṇapāla, *IA*, Vol. XV, pp. 304-10.
(11) Badal pillar-inscription of Guravamiśra, *EI*, Vol. II, pp. 160-67.
(12) Kahla plate of the Kalacuri Soḍhadeva, V.S. 1134, *EI*, Vol. VII, pp. 85-93.
(13) Bilhari stone-inscription of the rulers of Cedi, *EI*, Vol. I, pp. 251-70.
(14) Benares copper-plate of Karṇadeva, K.S. 793, *EI*, Vol. II, pp. 297-310.
(15) Chatsu inscription of the Guhila Bālāditya, *EI*, Vol. XII, pp. 10-17.

v1.c20.9

somewhat difficult. The Asiatic Society’s plate of Vināyakapāla, dated in V.S. 988, informs us that Mahendrapāla had a son named Bhoja II by his queen Dehanāgā-devī, and another son Vināyakapāla by the queen Mahī-devī. The latter is represented as meditating on the feet of his father, as well as of his illustrious half-brother. The Asni stone pillar-inscription of Mahīpāla, dated in 974 V.S., tells us that Mahīpāla meditated on the feet of Mahendrapāla. That Mahīpāla was a son of Mahendrapāla is also clear from the plays of Rājaśekhara. The Siyadoni stone inscription mentions Kṣitipāla after Mahendrapāla, though the relationship between the two is not specified. The inscription supplies the date V.S. 964 for Mahendrapāla and V.S. 1025 for Devapāla, the son of Kṣitipāla. The Partabgarh inscription of Mahendrapāla II, dated in V.S. 1003, mentions Vināyakapāla as meditating on the feet of Mahendrapāla I. The Khajuraho inscription of Dhaṅga, dated in V.S. 1011, tells us that the Candella Yaśovarman received an image of Vaikuṇṭha from the *Hayapati* Devapāla, the son of Herambapāla. The latter had received the statue from the Sāhi king of Kīra in return ‘for a force of elephants and horses.’¹ The late Dr. Kielhorn assumed the identity of this Devapāla with the Devapāla of the Siyadoni inscription,² and thereby identified his predecessors Kṣitipāla and Herambapāla. As *Kṣiti* and *Mahī* mean the same thing, he further assumed the identity of the princes named Kṣitipāla, Mahīpāla, and Herambapāla.³ It remained for Prof. D. R. Bhandarkar to show that Vināyaka and Heramba also mean the same deity and hence to propose the identification of Vināyakapāla with the three mentioned above.⁴ The chronology of the immediate successors of Mahendrapāla, according to this view would be as follows:

v1.c20.10

¹ > *Kailāsād-Bhoṭanāthaḥ suhṛditi ca tataḥ Kīrarājaḥ prapede* *Sāhi-stasmādavāpa dvipaturagabalēnānu Herambapālaḥ.* *tatsūnor-Devapālattamatha Hayapateḥ prāpyaninye pratiṣṭhāṁ* *Vaikuṇṭhaṁ kuṇṭhitāriḥ kṣitidharatilakaḥ śrī-Yaśovarmmarājaḥ.* *EI*, Vol. I, pp. 129 and 134. V. 43. Scholars are not unanimous about the interpretation of this verse.

v1.c20.11

² *EI*, Vol. I, p. 124.

v1.c20.12

³ *Ibid.* pp. 170-72.

v1.c20.13

⁴ *JBRAS*, Vol. XXI, pp. 406-07.

v1.c20.14

```
(1) Mahendrapāla (last known date 907-08 A. D.) Dehanāgā-devī = | | = Mahī-devī.¹

v1.c20.15

¹ According to the Partabgarh inscription the name is *Mahā-devī*, see *EI*, Vol. XIV, p. 176.

v1.c20.16

(2) Bhoja (II) |
(3) Mahīpāla (914, 917 A. D.) alias Kṣitipāla, alias Vināyakapāla (931 A. D.) alias Herambapāla. ```

v1.c20.17

Recently some Indian scholars have taken objection to this arrangement.² They do not accept the identification of the *Haya-pati* Devapāla of the Khajuraho epigraph with the Gurjara prince of that name, on the ground that *Hayapati* was ‘never the accepted title of the Pratīhāra kings of Mahodaya.’ Secondly, as the dates of Mahīpāla and Vināyakapāla never overlap, they see no reason to justify their identification. They would therefore reject the identification of Herambapāla with Vināyakapāla and of the latter with Mahīpāla and propose to arrange the princes as follows :—

v1.c20.18

² *EI*, Vol. XIV, pp. 178-80. *JL*, Vol. X, p. 54, fn. 6.

v1.c20.19

```
(1) Mahendrapāla (907-08 A. D.).

v1.c20.20

| . . . . .
| | |

v1.c20.21

(2) Mahīpāla, alias (3) Bhoja II. (4) Vināyakapāla Kṣitipāla (914-17). (A. D. 931). ```

v1.c20.22.1

The situation however has been further complicated by the discovery of the name of Vināyakapāla as ‘protecting the earth’ at the end of the Khajuraho inscription of Dhaṅga, dated in 954 A.D. A critical examination of the passage leaves no doubt that the name of Vināyakapāla was mentioned as that of the sovereign ruler. Now if this Vināyakapāla is identified with the Vināyakapāla of the Gurjara inscriptions, the attempt to identify Devapāla of the Khajuraho inscription at once falls to the ground. The identification of the two Vināyakapālas is however difficult on account of the date of the Khajuraho inscription (954 A.D.); for we know from the Partabgarh inscription that Mahendrapāla, the son of Vināyakapāla, was already on the throne in V. S. 1003 (A.D. 945-46). The suggestion that the name of Vināyakapāla occurred in the original record of Yaśovarman some time before 954 A.D. and that after the latter's death the record was set up in his son's reign in 954 A.D. with some verses ‘added at the end to describe the martial exploits of Dhaṅga’ seems to me rather improbable.¹ It is unlikely that in an official document a dead king should be referred to ‘as protecting the earth.’ Thus if we accept the suggestion that the Vināyakapāla of the Khajuraho inscription was a Gurjara prince, we are driven to assume the existence of a second Vināyakapāla who ruled after Mahendrapāla II.

v1.c20.22.2

Another complication has been introduced by the recent discovery of the Bayana ‘Ukha-Mandir’ stone-inscription of *Mahārājādhirāja* Mahīpāla, dated in V. S. 1012 (A.D. 956). The fact that the neighbouring feudatory princes of the locality where the inscription was found acknowledged the sovereignty of the Gurjaras as late as 1016 V. S. (960 A. D.), and the difficulty of identifying this prince with the Pāla or the Paramāra prince of the same name, seem to confirm the suspicion that he may have belonged to the Gurjara-Pratīhāra line of Kanauj. If this is admitted, we shall have to assume the existence of another Mahīpāla after the reign of Vināyakapāla of the Khajuraho inscription mentioned above.² Now in summing up the whole discussion, I beg to point out that once we accept the principle that synonymous names may be identified by identifying Mahīpāla with Kṣitipāla, and further grant the separate existence of the Vināyakapāla of the Khajuraho inscription, there is no really serious objection against the acceptance of the arrangement of succession advocated by Kielhorn and Bhandarkar. It is true that the title *Haya-pati* does not occur among the epithets of the Gurjara princes so far discovered; but at the same time the title “lord of horses” occurs as a title of kings during this period on the records of the princes of northern India.¹ It is therefore likely that Devapāla was a king and not merely a cavalry leader.

v1.c20.22.3

Again, it is clear from the assistance he gave to the Kīra prince that Herambapāla was also most probably a king. It is of course not absolutely impossible that these two princes were separate from the Gurjara princes bearing the same names; but it is improbable that there should be two sets of princes in the same period both in close association with the Candellas and both bearing the same names. If the fact that the dates of Mahīpāla and Vināyakapāla do not overlap be urged against their identification, the same argument may also be applied against the identification of Kṣitipāla and Mahīpāla, for so far there is nothing in their dates derived from their own inscriptions or those calculated from the records of their successors or predecessors to suggest any overlapping. It is possible to suppose for instance that Kṣitipāla like Bhoja II, may have had a brief reign between Mahendrapāla and Mahīpāla or Mahīpāla and Vināyakapāla; or to assume that Kṣitipāla, as the eldest son, first reigned for a brief period and was then ousted by Mahīpāla and the latter by Bhoja, who was in his turn peacefully succeeded by Vināyakapāla.

v1.c20.22.4

Such an assumption of internal dissensions in the family and enmity between Mahīpāla on the one hand and Bhoja and Vināyakapāla on the other, would furnish satisfactory reasons for the omission of Mahīpāla’s name from the records of Vināyakapāla and also explain some of the causes that may have operated to bring about the decline of the Gurjara Pratīhāras, so soon after the death of Mahendrapāla. But as this theory does not satisfactorily explain the absence of the name of Kṣitipāla from the inscription of Vināyakapāla, I would tentatively adhere to the arrangement of Kielhorn and Bhandarkar. According to this theory Mahendrapāla was succeeded some time after A.D. 907-08 by his son Bhoja II. Nothing is known about this prince excepting that he had a short reign and that he must have ceased to rule some time before 914 A.D., the first known date of Mahīpāla. There is reason to think that Mahīpāla during the first part of his reign succeeded in preserving the mighty empire that was handed down to him by his predecessors. Thus Rājaśekhara in the introduction of his *Bālabhārata* or *Pracaṇḍapāṇḍava* describes the victories of Mahīpāla in the following terms: "In that (lineage of Raghu), there was born the glorious Mahīpāladeva, who has bowed down the locks of hair on the tops of the head of the Muralas, who has caused the Mekalas to suppurate; who has driven the Kaliṅgas before him in war; who has spoilt the pastime of (the king who is) the moon of the Keralas;

v1.c20.22.5

who has conquered the Kulūtas; who is a very axe to the Kuntalas, and who by violence has appropriated the fortunes of the Ramaṭhas." In the same passage we are further told that the play was acted before an assembly of guests invited by Mahīpāla, ‘the pearl-jewel of the lineage of Raghu the *Mahārājādhirāja* of *Āryāvarta*.’ Another passage of the same work tells us that the play was performed by the king’s orders at the great city of Mahodaya.¹ The list of Mahīpāla’s conquests given by Rājaśekhara may possibly be exaggerated, after the manner of court poets. But an examination of it does not reveal any unsurmountable obstacles to our acceptance of it as mainly true. Thus a king who ruled over the Karnal district might well wage war with the tribes in the Kangra valley and the neighbouring regions.² Again a prince who ruled over Bundelkhand, Bihar, and Northern Bengal might well have engaged in hostilities with the rulers of the upper Narbada and the lower Mahanadi. Nor does a struggle with Kuntala present any difficulty. The rulers of Kuntala or the Kanarese districts of the Bombay Presidency were certainly the Rāshṭrakūṭas of Mānyakheṭa.

v1.c20.22.6

In describing her lineage, the *nāyikā* in the *Karpūramañjarī* gives the name of the ruler of Kuntala in the Deccan as *Vallabha-rāja*.³ The title *Vallabha-rāja* was assumed by the Rāshṭrakūṭa Dantidurga and the designation Balharī (بلهري) of Mānkīr (مانكير) applied by the Arabs to the Rāshṭrakūṭa sovereigns of the Deccan was only a corruption of this epithet.⁴ The Rāshṭrakūṭas and the Gurjara-Pratīhāras were warring with each other since the days of Dhruva Nirupama and Vatsa, and it is quite likely that the hostilities continued up to the time of Mahīpāla. The Gurjaras were at this time at the height of their power, and it is not at all improbable that Mahīpāla may have even carried on a raid into the heart of his enemies’

v1.c20.23

¹ *JL*, Vol. X, pp. 60-61.

v1.c20.24

² *ASI, WC*, 1919, pp. 43-44; *IA*, 1928, pp. 230 ff.

v1.c20.25

¹ The title *Aśvapati* occurs in the inscription of the successors of Lakṣmaṇasena in Bengal, see *JASB*, 1896, Vol. LXV, pp. 6 ff.; *ibid*, 1914, Vol. X (N.S.), pp. 97 ff. For the title amongst Candella kings, see *IA*, Vol. XVII, pp. 231 ff. The Gāhaḍavālas and the Haihayas also assumed this title, see *JASB*, Vol. XXXI, pp. 123-24; *EI*, Vol. XI, p. 321; *JRAS*, 1927, pp. 695 ff.; *EI*, Vol. XI, pp. 139 ff.; *ibid*, Vol. XII, pp. 205 ff.

v1.c20.26

¹ *Namita-Murala-mauliḥ pākalō Mēkalānāṁ,* *raṇa-kalita-Kaliṅgaḥ keli-tat Kēral-ēndoḥ,* *Ajani jita-Kulūtaḥ Kuntalānām, kuṭhāraḥ,* *haṭha-hṛta-Ramaṭha-Śrīḥ Śrī-Mahīpāladēvaḥ.* *Tēna ca Raghu-vaṁśa-muktāmaṇinā-Āryāvarta-mahārājādhirājena* *Śrī-Nirbhayanarendra-nandanenādhikṛtāḥ sabhāsadaḥ sarvān......etc.* *Nirṇaya Sāgar Press Ed.*, I, 7-8. The Muralas may have inhabited the banks of the river Muralā which is identified by some with the river Narbada. (*Trikāṇḍaśeṣa*, Chapter I). But in the *Raghuvamśa* (IV. 54, 55) the river is placed after the Kerala country. The Mekalas probably lived in the region round about the Maikal Range (Amarkantak) in the Central Provinces. The Kulūtas were taken by Cunningham to be the inhabitants of Yuan Chwang's *Kiu-lu-to* "which corresponds exactly with the position of the district of Kullu in the upper valley of the Byās river." The position indicated is roughly the modern Kangra district. The Ramaṭhas "must be a neighbouring people, as they are placed with the Pāñcanadas in the western division in the *Bṛhat Saṁhitā* and with the Kulindas in the Northern division in the *Vāyu Purāṇa*." Kuntala, Kerala, and Kaliṅga are well known geographical divisions of ancient India, and require no identification. For the identifications see *GDI*, Luzac, 1927; *JL*, Vol. X, pp. 63-64.

v1.c20.27

¹ For these and all relevant passages on the date of Rājaśekhara, see V. S. Apte, *Rājaśekhara, his Life and Writings*, 1886; Fleet, *The Date of the Poet Rājaśekhara, IA*, 1887, Vol. XVI, pp. 175 ff.; Wilson, *Theatre of the Hindus*, Vol. II, pp. 361-62.

v1.c20.28

² For identifications of the tribes conquered, see above, fn. 1, p. 576.

v1.c20.29

³ *Asti ettha Dakkhiṇāvade Kuntalesuṁ saalajanavallaho Vallaha-rājo nāma rāā*. Harvard Oriental Series, 1901, p. 32, Trans., p. 240.

v1.c20.30

⁴ *IA*, Vol. XII, p. 182, V. 10; *BG*, Vol. I, Part II, pp. 387-89.

v1.c20.31.1

territory, penetrating as far south as Kerala. Such sudden raids by enemy kings were quite familiar features of Indian history up to comparatively recent times. The evidence of Rājaśekhara gains substantial support from the *Murūj ul-Zahab*, a work by the contemporary Arab traveller Mas‘ūdī (*c.* 890-956 A.D.).¹ The following relevant passages may be collected from his work. After referring to the Balharī kings of Mānkīr, he says; “one of the neighbouring kings of India, who is far from the sea, is Ba’ūrah, (بَوْرَة) who is lord of the city of Qanūj (قنوج). This is the title given to all the sovereigns of that kingdom. He has large armies in garrisons on the north and on the south, on the east and on the west, for he is surrounded on all sides by warlike kings.² After criticising Jāḥiẓ, who supposed that the river Mihrān (Indus) came from the Nile, Mas‘ūdī says that that author “did not know that Mihrān of Sind comes from well-known sources in the highlands of Sind, from the country belonging to Qanūj in the kingdom of Ba’ūrah and from Kashmīr, Kandahar, and aṭ-Ṭāfin……”³ Further on, in Chapter XVI, he tells us that “the king of Qanūj, who is one of the kings of Sind, is Ba’ūrah; this is the title common to all kings of Qanūj…This king has four armies according to the four quarters of the wind. Each of them numbers 700,000 or 900,000 men. The army of the north wars against the prince of Multan and with the Musulmāns his·subjects, on the frontier.

v1.c20.31.2

The army of the south fights against the Balharī, king of Mānkīr. The other two armies march to meet enemies in every direction.”⁴ It is clear from the passages quoted above that the territories of the king of Kanauj included some portions of the Punjab and Sind, and bordered on the dominions of the Rāṣṭrakūṭas of the Deccan. As the date of Mas‘ūdī’s stay in India makes it almost certain that the king of Kanauj was Mahīpāla, his account gives us some idea of the power and prestige of the Gurjara-Pratīhāras during the period 912-16 A. D.¹ In waging a two-fold struggle with the Arabs of Sind and the Rāṣṭrakūṭas of the Deccan, Mahīpāla was only carrying on the traditional policy of his family.² The evidence of Rājaśekhara and Mas‘ūdī is also corroborated by archaeological testimony. The find-spots of the Asni (in Fatehpur dist., U. P.), Haddala (in Eastern Kathiawar), Rakhetra (near Chanderi in Gwalior) and the Asiatic Society’s plate (which grants land in Vārāṇasī-*Viṣaya*) show that the dominions of Mahīpāla at least extended from Kathiawar to the borders of Bihar. Thus it is clear that Mahīpāla during the first part of his reign could with some exaggeration be called the *Āryāvarta-mahārājādhirāja*. By successive stages the Gurjara-Pratīhāras had succeeded in occupying an almost supreme position amongst the rulers of Northern India. But the position was no bed of roses.

v1.c20.31.3

The incipient hostility of the Pālas in the east and the powerful combination of the Rāṣṭrakūṭas and the Arabs were perennial sources of danger. As Mas‘ūdī has pointed out, they had to maintain four mighty armies to meet “enemies in every direction.” The economic consequences of continuing such a struggle would have sapped the foundation of any empire, and the Gurjara empire was no exception to the operation of inexorable economic laws. The crash appears to have come soon after 915 A. D. Verse 19 of the Cambay plates of the Rāṣṭrakūṭa Govinda IV contains the following passage about the northern expedition of Govinda’s father Indra III. “The courtyard (of the temple of the god) Kālapriya became uneven by the stroke of the tusks of his rutting elephants. His steeds crossed the unfathomable Yamunā, which rivals the sea (*Sindhu-prati-spardhinī*). He completely uprooted (*nirmmūlaṁ unmūlitam*) the hostile city of Mahodaya, which is even to-day greatly renowned among men by the name Kuśasthala.”¹ As “the god Kālapriya is generally identified with Mahākāla” of Ujjayinī, the passage suggests that Indra III proceeded in his northern expedition by way of Malwa. The statement of this inscription is further supported by some passages of the *Vikramārjuna-vijaya* of the Kanarese poet Pampa usually known as *Pampa-Bhārata*. This poem was composed in about *Śaka* 863 (941 A. D.) under the patronage of the Cālukya chief Arikesarin, a feudatory of the Rāṣṭrakūṭa Govinda IV.

v1.c20.31.4

Pampa gives us the following information about the military achievements of Narasiṁha, the father of Arikesarin, who must have been a contemporary of Indra III. We are told that “when preparing for victory he captured the champion elephants which marched in front, and penetrating and putting to flight the army of the Ghūrjjara-rāja, secured the victory and eclipsed Vijaya (or Arjuna),—this Narasiṁha. Terrified at the army of this Naraga, which fell like a thunderbolt, Mahīpāla fled in consternation, not stopping to eat or sleep or rest. His own horse he bathed at the junction of the Ganges and the sea (*Gaṅgā-vārdhiyoḷ*), thus becoming celebrated; and by his own friend destroying the qualities and character of (?) Saṅga, established with pride the victory of his arm.”² Though the meaning and reference in the latter part of this passage are not clear, yet it is evident that the poet was apparently referring to the achievements of his patron's father, who must have accompanied his sovereign Indra III when the latter overran northern India. The exact date of this victory of the Rāṣṭrakūṭas cannot be settled; but it appears to have happened some time between 915 and 918 A.D. For the Nausari grant of Indra III which is dated in 915 A.D., contains no reference to these victories; and the Daṇḍapur stone inscription of his son Govinda IV shows that he was dead sometime before 23rd December, 918.¹ The defeat of the Gurjara-Pratīhāras appears to have been complete;

v1.c20.31.5

but the Rāṣṭrakūṭas could not take full advantage of their victory because of the confusion in which they were involved after the death of Indra III. Amoghavarṣa II, who succeeded his father, was probably killed and dethroned by his younger brother Govinda IV, and the latter in his turn, "being caught in the chains of the eyes of women," took to "vicious courses" and met his destruction.² It is however likely that Govinda IV retained some hold on the Gaṅgā-Yamunā valley till about the 10th May, 930 A.D., the date of his Cambay plates. For in verse 28 of that inscription "the Gaṅgā and Yamunā are mentioned as doing service in the palace of Govinda IV." But there is unquestioned epigraphic evidence to show that the Gurjara-Pratīhāras recovered a substantial portion of their dominions, probably by taking advantage of this weakness of the Rāṣṭrakūṭa power. The Asiatic Society's plate of Vināyakapāla (*alias* Mahīpāla ?), which was issued from Mahodaya, shows that *Vārāṇasī-Viṣaya* in the *Pratiṣṭhāna-Bhukti*, was still under the Gurjaras in 988 V.S. (931 A.D.). The Rakhetra stone inscription of the same prince shows that the Gwalior region was under him in V.S. 999-1000 (A.D. 942-43). In his attempts to recover his dominions Mahīpāla appears to have been helped by a number of his feudatories.

v1.c20.31.6

One of the Candella fragmentary stone inscriptions of Khajuraho claims that either Harṣa or his son Yaśovarman placed Kṣitipāla on the throne.¹ The Guhila prince Bhaṭṭa, who appears to have been a contemporary of Mahīpāla, is credited in V. 26 of the Chatsu inscription of Bālāditya with having defeated the armies of the kings of the south at the behest of his overlord.² It is not unlikely, as has been suggested, that "the kings of the south were no other than the chiefs of the Rāṣṭrakūṭa army by defeating whom king Mahīpāla regained his territories."³ There is therefore some incontestable evidence that Mahīpāla succeeded in recovering a portion of his dominions with the assistance of his feudatories. There is nothing to shew what the limits of his revived empire were. But there is every reason to believe that the blow inflicted by the Rāṣṭrakūṭas helped to bring about a new relationship between the imperial government and its vassals which in the end proved fatal to it. Thus, though many of the subordinate kings still acknowledged the supremacy of the imperial power in their official documents, they resisted by physical force every attempt of the sovereign to exert any real authority over them. The best example of this state of things is found in the records of the Candellas. Though Dhaṅga retains the name of Vināyakapāla, apparently a Gurjara king, in his inscription even as late as A.D.

v1.c20.31.7

954, he none the less described his father Yaśovarman as having been "a scorching fire to the Gurjaras."⁴

v1.c20.32

¹ Al Mas'ūdī was born in Bagdad towards the close of the 3rd century A. D. and died in Egypt in 345 A.H. (956 A.D.). He visited Multan and Manṣūrah, *c.* A.H. 300 (A.D. 912) and Cambay *c.* 301 A. H. (A. D. 916). See, *Les Prairies d'or*, Texte et Traduction par C. Barbier de Meynard, Paris, 1861, Tome I, p. 111.

v1.c20.33

² *Elliot*, Vol. I, p. 21, *Les Prairies d'or*, Tome I, p. 178. Meynard transcribes بَوْرَدْ as Baourah.

v1.c20.34

³ *Elliot*, Vol. I, p. 21; *Les Prairies d'or*, Tome I, p. 207.

v1.c20.35

⁴ *Elliot*, Vol. I, pp 22-23; *Les Prairies d'or*, Tome I, pp. 372 and 374.

v1.c20.36

¹ Dr. R. C. Majumdar first noticed the importance of these passages in his paper on the Gurjara-Pratīhāras in *JL*, Calcutta University, Vol. X, pp. 64-66. He may be right in his suggestion that the word Ba'ūrah (بَؤُرَة) is an “Arabic corruption of the word Pratīhāra or its Prakrit form Paḍihāra.” See also *supra*, ‘Dynastic History of Sind,’ p. 4, fn. 3, p. 16, etc.

v1.c20.37

² See Gwalior (Sāgar-Tal) stone inscription of Bhoja, V. 4, *EI*, Vol. XVIII, p. 167, also fn. 10; Radhanpur grant of Govinda III, V. 8, *EI*, Vol. VI, pp. 239 ff.

v1.c20.38

¹ *EI*, Vol. VII, p. 38.

v1.c20.39

² Ed. by B. Lewis Rice, in the *Bibliotheca Carnatica*, Bangalore, 1898, pp. 3-4.

v1.c20.40

¹ *JBRAS*, Vol. XVIII, pp. 257 ff.; *IA*, Vol. XII, pp. 222 ff.; *JL*, Vol. X, pp. 66-67.

v1.c20.41

² *BG*, Vol. I, Part II, pp. 204-05 and 416-17; *EI*, Vol. VII, pp. 34-35; XIII, p. 328. *JBRAS*, Vol. XVIII, pp. 250-51.

v1.c20.42

¹ *EI*, Vol. I, pp. 121-22, line 10; the name of the prince who helped Kṣitipāla is lost. Kielhorn, who edited the inscription, took this prince to be Harṣadeva, while Hoernle thought that he was probably his son Yaśovarman; see *JRAS*, 1904, p. 654 and fn. 1.

v1.c20.43

² *EI*, Vol. XII, pp. 12 and 16, V. 26.

v1.c20.44

³ *JL*, Vol. X, p. 68.

v1.c20.45

⁴ *EI*, Vol. I, pp. 127 ff., V. 23.

v1.c20.46

Mahīpāla-Vināyakapāla ruled at least up to about 942-43 A.D. The following records may be referred to his reign:

v1.c20.47

(1) *Haḍḍālā grant.*—This was discovered near "Haḍḍālā, a large village on the old road from Dholka to Dhandhūkā, but belonging to Eastern Kathiawar.'' It consists of 52 lines, and is written on two semi-elliptical plates. On the first plate "it shows the rude conventional representation of the moon and the sun, while it is blank on the second.'' There are no holes for rings. The inscription begins with an invocation to (Siva) Dhandeśvara. It then proceeds to give the legendary origin of the Cāpas from the *cāpa* (bow) of Śambhu. In the *Cāpa-vaṁśa* was born *Nṛpa* Vikramārka. His son was *Rājā* Aḍḍaka; his son was king Pulakesi; his son was king Dhruvabhaṭa; his younger brother was king Dharaṇīvarāha, resident in Vardhamāna (mod. Wadhwan in E. Kathiawar), who was *Samadhigat-āśeṣa-mahā-śabda-Mahāsāmantādhipati*, and who through the favour of *Rājādhirāja-Parameśvara* Śrī-Mahīpāla-deva ruled the Aḍḍāṇakadeśa named after his own grandfather. This prince granted to Maheśvarācārya on the day of the winter solstice the village named Vimkala, which is connected with the Kanthikā-*sthalī*. It is dated in *Śaka*. S. 836 and was written by *Sāṁdhivigrahika* Mahindaka. Taking the date of the inscription into consideration, Bühler calculated the date of Vikramārka, the founder of the feudatory line, as about 800 A.D. The same scholar also identified Aḍḍāṇa with the modern village of Haḍḍālā.¹

v1.c20.48

¹ Bühler edited the inscription in 1883 from a facsimile, a paper rubbing, and an impression on lead of the second half of the grant in *I.A*, Vol. XII, pp. 190-95. The date of the record was read by Fleet, *ibid*, 1899, Vol. XVIII. p. 90.

v1.c20.49

(2) *Asni stone pillar-inscription.*—It is recorded on one of the faces of a square sandstone pillar that was found at Asni, a village about 10 miles north of Fatehpur, the chief town of the district of the same name in U.P. The inscription consists of 14 lines of incorrect Sanskrit prose written in northern Devanāgarī of the 10th century. The record opens with the sign ९ and then mentions *Pb.-M.-P* Mahīṣa(ndra?)pāladeva-*pādānudhyāta Pb.-M.-P.* Mahīpāladeva. In his reign, in the (V) year 974, this inscription was set up in a certain *caitya* of the god Yogasvāmin. It made some arrangements for the worship of the god by Brahmans and ascetics of the locality. The record was written by Suvarṇabhadra.¹

v1.c20.50

¹ Edited by Fleet in *IA*, Vol. XVI, pp. 173-75, in 1887. The reading of the name of Mahīpāla's predecessor is not certain. Fleet read it as Mahīgapāla; Mr. R. D. Banerji (*MASB*, Vol. V, No. 3, p. 65) read the name as Mahindrapāla from an impression of this inscription in the Allahabad exhibition of 1910. This reading however is not supported by the plate given by Fleet. The pillar-inscription is now in the Municipal garden attached to the Town Hall of Fatehpur.

v1.c20.51

(3) *Bengal Asiatic Society’s plate.*—The find-spot of this plate, sometimes spoken of as the ‘Benares plate’ from the fact that it mentions in line 10 Vārāṇasī-*Viṣaya*, is not known. It consists of 17 lines, and is written on one side of a single plate. “On to the proper right side of the plate there is soldered a thick and massive seal with a high raised rim all round it;……it is rectangular, except that the top of it is raised into an arched peak with a slight depression on each side of it. In the arch thus formed, there is a standing figure facing full front of a goddess, doubtless the *Bhagavatī*,” who is mentioned in the inscription. Below this, across the surface of the seal, there is the legend in 16 lines in relief. The seal and legend are exactly similar to those in the Dighwa-Dubauli plate of Mahendrapāla. The additional information in the legend of this inscription is as follows:—*Mahendrapāla-devastasya puttras tat-pādānudhyātaḥ Śrī-Dehanāgā-devyām utpannaḥ Parama-vaiṣṇavo Mahārāja-Śrī-Bhojadevas tasya bhrātā Śrī-Mahendrapāladeva-puttras-tayōḥ-pādānudhyātaḥ-Śrī-Mahī-devī-devyām utpannaḥ Paramādityabhakto Mahārāja-Śrī-Vināyaka-pāladevaḥ*. The inscription begins with *Om svasti*, and then tells us that it was issued from the *Skandhāvāra* situated at Mahodaya. Then it again repeats the genealogy given on the seal. Next it records the gift of ‘*Pratiṣṭhāna-Bhuktau Vārāṇasī-Viṣaya-sambaddha-Kāśī-pāra-Pathaka-Pratibaddha-Ṭikkarikā-grāma*,’ to the Bhaṭṭa Bhullāka, a student of the Atharvaveda, by Vināyakapāla. The village of Ṭikkarikā has been identified with the modern Tikari, 4 miles due south of Benares. The date (*V*) *Saṁvatsara* 988, comes last.¹

v1.c20.52

¹ This inscription was first noticed by Dr. R. L. Mitra in 1848 in the *JASB*, Vol. XVII, part I, pp. 70 ff. In 1862 Dr. F. Hall noticed it in *JASB*, Vol. XXXI, pp. 1 ff. In 1864 a lithograph of the record was published in the same journal. Fleet next edited it from the original plate in 1886 in the *IA*, Vol. XV, pp. 138-41. He however read the date as (Harṣa) S. 188. It was corrected by Prof. D. R. Bhandarkar in the *JBRAS*, Vol. XXI, pp. 405 ff.

v1.c20.53

(3) *Rakhetra stone-inscription.*—This inscription was found on the right bank of the river Orr within the limits of the village of Rakhetra, not far from the old site of Chanderi in Gwalior. It is written in Sanskrit and incised on a rock-tablet. It is dated in V.S. 999 and 1000, and apparently records the construction of some sort of waterworks at a cost of 95 or 96 crores (of coins) by Vināyakapāla. As there is no overlapping of dates, we may with some certainty identify him with Mahīpāla *alias* Vināyakapāla.²

v1.c20.54

² *ASI*, 1924-25, p. 168.

v1.c20.55

Mahīpāla-Vināyakapāla was succeeded by his son Mahendrapāla II by his queen Prasādhanā-devī, some time before V.S. 1003. The existence of this prince was revealed by the discovery of the *Partabgarh stone-inscription*. This stone was found affixed to a platform at Partabgarh, the capital of the state of that name in Southern Rajputana. The inscription contains 35 lines of writing and is naturally divided into four parts. It records a series of grants in favour of various shrines attached to the monastery of Hari-Ṛṣīśvara, who originally lived in Daśapura (mod. Mandasor). The inscription opens with 2 verses invoking the sun-god, followed by two verses in praise of the goddess Durgā. Then comes the statement that the inscription was issued from the *Skandhāvāra* situated at Mahodaya. Next is given the genealogy of the Gurjara-Pratīhāras, which is the same as in the Asiatic Society's plate of Mahīpāla as far as Mahendrapāla. *Tasya-putras tatpādānudhyātaḥ Śrī-Mahādevyām-utpannaḥ Paramāditya-bhakto Mahārāja-Śrī-Vināyakapāla-devas* *tasyaputras tatpādānudhyātaḥ Śrī-Devathāddhi(?)-nāma-nija-kula-Prasādhanādevyām-utpannaḥ Parama-māheśvaro-Mahārāja-Śrī-Mahendrapāladevaḥ*. This last named king grants the village of Kharpara-padraka in the holding of Tala-vargika Hariṣaḍa and situated in the vicinity of Ghoṇṭā-varṣikā, in the western *Pathaka* of Daśapura, to the goddess Vaṭa-Yakṣhiṇī Devī, whose temple was connected with the *maṭha* of Hari Ṛṣīśvara. The grant was made in V.S. 1003 at the request of Dhana-Sūra. The record was written by *Purohita* Trivikramanātha under the orders of Jajja-Nāga. It ends with *Svahasto'yam Śrī-Vidagdhasya* (line 13).

v1.c20.56

The second grant opens with a praise of the princes of the *Cāhamānānvaya*, who, we are told, were a source of great pleasure to Bhojadeva. Then comes the name of Govindarāja, of this lineage; his son Durlavarāja; his son Indrarāja built the great temple of the Sun-god known (as Indrādityadeva at Ghoṇṭā-varṣikā). This is an entirely new Cāhamāna family, which appears to have risen to importance in the service of Bhoja I (*c.* 836-90). Next we are told that there was one Mādhava, who was *Mahāsāmanta-Danḍanāyaka* in Ujjayinī, and Śrī-Śarman appointed by Kokkaṭa, who was *Balādhikṛta* serving at the feet of *Parameśvara* (*i.e.*, Mahendrapāla II?) at Maṇḍapikā (mod. Mandu) (line 20). The aforesaid Mādhava (here called *Tantrapāla-Mahāsāmanta-Mahādaṇḍanāyaka*) (line 21) having come to Ujjayinī on business and having bathed at the temple of Mahākāla, granted on the Mīna-Saṁkrānti day the village of Dhārā-padraka for repairs to and maintenance of daily services to the temple of Indrāditya deva, at the request of *Cāhamānānvaya-Mahāsāmanta-Śrī-Indrarāja*, son of Durlabharāja. In line 26 is recorded a grant of a field. It comes to an end with *Svahasto'yam Śrī-Mādhavasya Svahasto'yam Śrī-Vidagdhasya*.

v1.c20.57

Part three opens with the date (*V*) *Saṁvat* 999, which marked the grant of a field named Vavvūlika by the side of the river Nandyā in the village of Palāśa-kūpikā to Indrāditya-deva of Ghoṇṭā-varṣi by *Mahārājādhirāja* Bhartṛ-paṭṭa, son of Khommāṇa. G. H. Ojha has identified this prince with Bhartṛpaṭṭa II, son of Khommāṇa III of Mewar, of the Guhila family. The same scholar has suggested the identification of Palāśakūpikā with Parasia, about 15 miles south of Mandasor.

v1.c20.58

Part four of the inscription records about four minor grants to various deities connected with the same *maṭha*. Last comes the date (V) *Saṁvat* 1003.¹

v1.c20.59

¹ Edited by G. H. Ojha, in *EI*, Vol. XIV, pp. 176-88.

v1.c20.60

The facts revealed by the inscription mentioned above show that the Gurjaras in 946 A.D. held the Malwa region, and it is not unlikely that their power was still acknowledged by the feudatories in South and Eastern Rajputana and Central India. Mr. R. D. Banerji therefore is certainly wrong when he asserts that Mālava had never been reoccupied after the northern campaign of Indra III.∗ However nominal the allegiance paid by these feudatories stationed at Mandu, Ujjayinī, and the places near about Partabgarh and Mandasor, it is clear that the occupation of Ujjayinī by the Rāṣṭrakūṭas was not continuous till the time of Kṛṣṇa III (*c.* 940-70 A.D).

v1.c20.61

∗ *JBORS*, 1928, December, p. 486.

v1.c20.62

Mahendrapāla II, appears to have been succeeded by his brother or half brother Devapāla on or before 948-49 A.D. The existence of this prince was revealed by the *Siyadoni stone-inscription* discovered about 10 miles N. N. W. of Lalitpur near Jhansi, at Siron Khurd, in the precincts of the Jain temple of Sāntinātha. The inscription contains 46 lines of incorrect Sanskrit influenced by the local vernacular. It is divided into two parts. The first part (lines 1-39) is written in prose, and records a large number of donations made at different times from the (V) year 960 down to the (V) year 1025 by some feudatories and private individuals in favour of the god Viṣṇu and various other Brahmanical gods at Siyadoni (mod. Siron). In line 1, which is much damaged, we have with the date (V) S. 960 the names of the *Pb.-M.-P.* Mahendrapāladeva (meditating on the feet of *Pb.-M.-P.* Bhojadeva). In line 4, with the date (V) S. 964, the same two princes are mentioned in the same order. In line 28, with the date (V) S. 1005, occurs the name of *Pb.-M.-P.* Devapāladeva, meditating on the feet of *Pb.-M.-P.* Kṣitipāladeva. As line 40 speaks of a ruler of Mahodaya who granted some land to certain Brahmanas of Siyadoni, Kielhorn rightly assumed that these rulers were Gurjara-Pratīhāra kings of Kanauj.¹

v1.c20.63

¹ The inscription was first noticed by F. Hall in *JASB*, Vol. XXXI, pp. 6-7. This was based on a very imperfect copy of a transcript of the record. Kielhorn then edited it in *EI*, Vol. I, pp. 162-79.

v1.c20.64

It is significant that, unlike the Bengal Asiatic Society's plate of Vināyakapāla, which mentions his predecessor and brother Bhoja, this inscription does not give the name of the reigning king's brother and predecessor Mahendrapāla II. It is not unlikely therefore that the succession of Devapāla was not entirely peaceful, and that he may have omitted to mention the name of one whom he looked upon as a mere usurper. If this was so, then these internal dissensions must have hastened the decline and break-up of the Gurjara dominions. Another important cause that appears to have helped towards the dismemberment of the Gurjara empire was the revival of the Rāṣṭrakūṭas' attacks on Northern India. The Karhad plates of Kṛṣṇa III, dated in *Śaka* 880 (A.D. 959), refer to his victories in the North.² V. 25 of this inscription, in which he claims to have "conquered Sahasrārjuna," has been rightfully interpreted by Dr. R. G. Bhandarkar as referring to his victories over some Cedi rulers of Northern India. In V. 31, we are told that "all the feudatories from the eastern to the western ocean and from the Himalayas to the island of Siṁhala bowed to him out of fear of severe punishment, though he himself was obedient (*i.e.*, subordinate to his father)." Such vague claims in Indian records of course mean nothing. But V. 30 of the same inscription contains more definite information. It runs as follows : "On hearing of the conquest of all the strongholds in the southern region simply by means of his angry glance, the hope about Kālañjara and Citrakūṭa vanished from the heart of the Gurjara.''¹ These claims of conquests in Northern India are supported by the recent discovery of several inscriptions of Kṛṣṇa III in the Ahmedabad district of Gujarat and the Maihar State in Baghelkhand Agency (C.I.). Two of these were found at the village of Harasola in Ahmedabad, and record grants of land in V.S. 1005 (A.D. 949) by the Paramāra Sīyaka, apparently a feudatory of *Pb.-M.-P.* Akālavarṣa (Kṛṣṇa III), the son of Amoghavarṣa.² The third inscription was discovered on a stone slab at the village of Jura in the State of Maihar. It is written in Kanarese, and "contains the name and the different titles or *birudas* of the Rāṣṭrakūṭa king Kṛṣṇa III, Akālavarṣa, the son of Amoghavarṣa III, *alias* Baḍḍega.''³ These inscriptions show that some time before the middle of the 10th century Kṛṣṇa III had attacked the Gurjara dominions from the south-west and the south-east, and it would seem that he even succeeded for a time in capturing the famous forts of Citrakūṭa (Chitor) and Kālañjara from the Gurjara-Pratīhāras.⁴ The Deoli and the Karhad plates seem to indicate that most of these victories of Kṛṣṇa occurred during the lifetime of his father, who died some time before 940 A.D.⁵ We have seen that the attack of Indra III proceeded by way of Ujjayinī. The Partabgarh inscription, dated in 945-46 A.D. however shows that the Gurjaras had recovered Malwa. Their possession of the fort of Mandu in that year probably led the Rāṣṭrakūṭas in the reign of Amoghavarṣa III to avoid the Malwa route and attempt a flanking movement from the south-west and the south-east. The inscriptions mentioned above show that their strategy was brilliantly successful. Another interesting fact to be noted in this connection is the mention of the Gurjara dominion in connection with the forts of Chitor and Kālañjara. This shows that the Guhilots and the Candellas were as late as 959 A.D. regarded still as feudatories of the Gurjara-Pratīhāras. It is therefore not difficult to understand why Dhaṅga should mention the name of Vināyakapāla as his overlord in his Khajuraho inscription, dated in 954 A.D. Thus though the Candellas claimed victories over the Gurjaras before 954 A.D. they none the less must have retained in their outward relations an attitude of formal subordination to the effete imperial government at Kanauj.¹

v1.c20.65

² *EI*, Vol. IV, pp. 278 ff.

v1.c20.66

¹ *EI*, Vol. IV, p. 284; V. 30 also occurs in the Deoli plates of the same king, see *JBRAS*, Vol. XVIII, p. 247, lines 7-8 of the second side of plate 11.

v1.c20.67

² *PTOC*, Madras, 1925, pp. 303-08; *EI*, Vol. XIX, pp. 236 ff. *JBORS*, December, 1926, pp. 479 ff.

v1.c20.68

³ *MASI*, No. 23. *The Haihayas of Tripuri and their Monuments*, pp. 11 and 117; *JBORS*, December, 1928, pp. 476 ff.

v1.c20.69

⁴ Mr. R. D. Banerji suggests that the Citrakūṭa of these inscriptions should be identified with the place of the same name near Kālañjara, in the Banda District. See *JBORS*, 1928, p. 481. But I think that the famous fort of Chitor is here meant. The two forts were apparently the objective of the two expeditions which penetrated the Gurjara territory from the S.W. and S.E.

v1.c20.70

⁵ *EI*, Vol. IV, p. 279.

v1.c20.71

¹ Dr. R. C. Mazumdar (*JL*, Vol. X, p. 69) has in this connection pointed out the relations of the rulers of Oudh and the later Mughuls of Delhi. A more interesting instance is probably that of the Buwayhids of Ray and the later Abbasid Caliphs of Baghdad.

v1.c20.72

The end of the reign of Devapāla brings us again to a period of confusion in the chronology of the Gurjara-Pratīhāras. I have already pointed out the difficulty of accepting the identification of the Vināyakapāla of Dhaṅga's Khajuraho inscription (dated in V.S. 1011)² with the Mahīpāla-Vināyakapāla whose known dates range between *Śaka* S. 836 and V.S. 1000. If the former is to be accepted as a separate ruler of the Gurjara-Pratīhāra line of Kanauj, how are we to connect him with that line? The only suggestion so far advanced is based on the custom sometimes found in Indian royal families of naming the grandsons after the grandfather. According to this suggestion, Vināyakapāla of the Khajuraho inscription may have been a son of Mahendrapāla of the Partabgarh inscription.¹ Another difficulty is the identification of the *Mahārājādhirāja* Mahīpāla of the Bayana stone-inscription of Citralekhā, dated in V.S. 1012 (*c.* A.D. 956). This inscription consists of 22 lines of writing, and records the erection of a temple of Viṣṇu by a queen named Citralekhā during the reign of Mahīpāla. Citralekhā was the daughter of one Rājayika and married Maṅgalarāja, who may perhaps be identified with the Kacchapaghāta prince of the same name mentioned in the Gwalior Sāsbahū temple inscription of Mahīpāla, dated in V. S. 1150. The inscription records the grant of two villages to the god Nārāyaṇa. Of the villages mentioned Gograpura has been identified with modern Gogera in tahsil Weir, Nāgapallī with modern Nāvali in tahsil Bayana, and Hāḍhapallī with mod. Hadholi a village in the district of Hindaun in Jaipur state.² Mr. R. D. Banerji's suggestion that the Mahīpāla of this record is a prince of the imperial Pratīhāra line of Kanauj, is certainly plausible, and if accepted will give us a Mahīpāla II. As the Bayana inscription gives no genealogy of Mahīpāla we can only accept the tentative suggestion that he may have been a son of Devapāla and grandson of Mahīpāla I. The situation is further complicated by the discovery of a fragmentary stone-inscription at Osia of the Pratīhāra Vatsarāja, dated in V.S. 1013. The inscription was found incised on a slab built in the wall of the *maṇḍapa* of a Jain temple.³ As an epigraph of Vatsa the father of Nāgabhaṭa II was found at Osia, it may be assumed that the place was still situated within the Gurjara dominions. Was this Vatsa then a Pratīhāra of Kanauj? If so, where are we to place him on the genealogical table? His date of course places him after the Mahīpāla (II?) of the Bayana inscription. The acceptance of this Vatsaraja as a member of the imperial family of Kanauj would then involve us in the supposition that three princes, Vināyakapāla II, Mahīpāla II, and Vatsa II, ruled in rapid succession, during the period V.S. 1011-13. This of course is not impossible, but at the same time the suspicion that the last two rulers may have been representatives of local branches of the family cannot entirely be avoided. The Gurjara-Pratīhāra empire was certainly declining since the capture of Mahodaya by Indra III. The renewal of the invasions under Kṛṣṇa III helped to disintegrate the empire and give rise to local lines, some of whom at least may have been related to the imperial stock. In the present state of our knowledge however it is better to keep an open mind and await the discovery of fresh material before formulating any definite opinion on the point.

First 82 of 196 passages. Read on

Linked texts (22)