The Early History of India, from 600 B.C. to the Muhammadan Conquest, including the Invasion of Alexander the Great c32: Appendix O. The Origin and Chronology of the Sena Dynasty

The Early History of India, from 600 B.C. to the Muhammadan Conquest, including the Invasion of Alexander the Great c32

Language: English

Appendix O. The Origin and Chronology of the Sena Dynasty

c32.2

The strong interest taken by many of my readers in the early history of Bengal induces me to devote considerable space to the justification of the statements in the text concerning the Sena dynasty, which differ widely from those made in the second edition, when much material now available was not at my disposal.

c32.3

The Sena kings succeeded one another from father to son. The names and order of succession are established by inscriptions beyond dispute as being (1) Sāmantasena, (2) Hemantasena, (3) Vijayasena, (4) Vallālasena or Ballāl Sen, (5) Lakshmaṇasena, (6) Viśvarūpasena. Nos. 1 and 2 were merely local chiefs in Orissa, and No. 6 was a ruler of small power in Eastern Bengal. The general history of India is interested only in Nos. 3, 4, 5, who governed dominions of large extent and took rank among the greater powers.

c32.4

In supersession of the view adopted in the third edition of this work, it now seems desirable to accept the hypothesis that there were two Lakshmaṇasenas, and that Lakshmaṇasena of the inscriptions is to be distinguished from Rāe Lakhmaneya who was driven out of Nūdīah (Nuddea) by Muhammad the son of Bakhtyār, as described in the *Ṭabaḳāt-i-Nāṣirī*. Another matter definitely settled by the labours of the late Professor Kielhorn, as confirmed by subsequent researches, is the beginning of the era known by the name of *Lakshamaṇasena*. The first day of that era was October 7, A. D. 1119, and the first current year as reckoned from that era was A. D. 1119–20. It is also clear that a Sena king who bore the name of Lakshmaṇa, or perhaps the epithet Lakshmaneya, a descendant of Lakshmaṇa, and who was posterior to the three sons of Lakshmaṇasena of the inscriptions, was driven out of Nūdīah by Muhammad the son of Bakhtyār at some date subsequent to the taking of Delhi by the Muhammadans in A. H. 589, which is practically equivalent to A. D. 1193, and prior to Muhammad's expedition into the hills of the NE. frontier, called Tibbat (Tibet) by the author of the *Ṭabaḳāt*, which took place in A. H. 601 (Aug. 1204–Aug. 1205).

c32.5

But considerable difference of opinion exists as to the exact date of the raid on Nūdīah, which is not stated in the *Ṭabaḳāt*, our only authority for the details. That work, it may be noted, was closed in A. H. 658, practically equivalent to A. D. 1260. The author, commonly called Minhāj-i-Sirāj, expressly states that in A. H. 641 (June, A. D. 1243–June, 1244) he obtained an account of the operations of Muhammad the son of Bakhtyār against Bihār town from two of his surviving soldiers (Raverty, *transl.*, p. 552). His account, therefore, has almost the authority of a contemporary narrative so far as that event is concerned. But he does not seem to have been so well informed about the raid on Nūdīah.

c32.6

In the briefest possible summary, the historian's narrative is as follows. Muhammad, son of Bakhtyār, a man of the Turkish Khalj tribe, failed to obtain employment from Kutb-ud-dīn after the capture of Delhi in A. H. 589. When some time, apparently a considerable interval, had elapsed, he acquired a certain amount of military power and obtained a fief in the Mīrzāpur district from which he was 'in the habit of making incursions into Manēr (in the north-west of Patna District)¹ and Bihār', until he collected 'ample resources in the shape of horses, arms, and men'. We are further told that he 'used to carry his depredations into those parts' until he organized a final attack upon the fortified city of Bihār. He captured the city, as related in the text, and brought great booty to the presence of Kutb-ud-dīn, who was, perhaps, then at Mahobā in Bundelkhand. The favour with which he was received excited jealousy, which was not allayed until Muhammad justified himself by defeating a furious elephant. After that incident he departed for Bihār. Meantime, many of the inhabitants of Nūdīah became alarmed and deserted their king, Rāe Lakhmaniya, or Lakshmaṇasena. 'The following year after that, Muhammad-i-Bakhtyār caused a force to be prepared, pressed on from Bihār, and suddenly appeared before the city of Nūdīah', as described in the text.²

c32.7

¹ *J. B. & O. Res. Soc.*, iv, 266.

c32.8

² Since the passage above was written, Mr. S. Kumar has published the opinion that the testimony of the *Ṭabaḳāt* should be disregarded. He is inclined to believe that Lakshmaṇasena ascended the throne about A. D. 1119 and was dead long before the Muhammadan raid (*Ind. Ant.*, 1913, p. 188). This is also the opinion of R. D. Banerji, as briefly stated in *Ep. Ind.*, Oct. 1917, vol. xiv, p. 157. After further consideration their view has been adopted. The two systems of chronology differ by just fifty years.

c32.9

Now, on reconsideration of the evidence, I agree with Blochmann that it is impossible to date the attack on Nūdīah, as Raverty did, in A. H. 590. The operations of Muhammad above detailed must have occupied several years after A. H. 589, when Delhi was taken. On the other hand, Minhāj-i-Sirāj tells us (Raverty, p. 560) that 'after some years had passed away', Muhammad organized his expedition to 'Tibbat'. That disastrous operation took place in A. H. 601 (Aug. A. D. 1204–Aug. 1205). The capture of Nūdīah, therefore, must be dated several years after A. H. 589, and 'some years' before A. H. 601, say in or about A. H. 595 (Nov. A. D. 1198 to Oct. 1199).

c32.10

But the story told by Minhāj-i-Sirāj enables us to fix the date with a little more precision. He was informed that Rāe Lakhmaniya had then been on the throne for eighty years, reckoned from his birth. That assertion, which is supported by an anecdote, manifestly legendary, is in itself highly improbable. The longest recorded Indian reign is that of Choṛaganga of Orissa, which extended to seventy-one years complete (A. D. 1076–1147); and, so far as I know, a reign of eighty years cannot be traced in the history of any country. Raverty supported his belief in the eighty years' reign by quoting a statement made by Munshi Shiām Parshād in an account of Gauṛ, written for Major Francklin, that Lakshmaṇasena reigned from A. H. 510–590, eighty lunar years. But it does not appear what authority the Munshi had for his statement. Another argument on the same side is that Muhammad died in A. H. 602, and according to certain historians had reigned or ruled for twelve years in Lakhnautī or Gauṛ. Twelve years back from A. H. 602 bring us to A. H. 590. It is possible, however, as Bābū Monmohan Chakravarti suggests, that the rule of Muhammad may have been reckoned from a time prior to the attack on Nūdīah. On reconsideration, I agree with Blochmann in rejecting both the alleged eighty years' reign and the date A. H. 590 for the attack on Nūdīah.

c32.11

I now accept the suggestion made long ago by Professor Kielhorn (*Ind. Ant.*, vol. xix (1890), p. 7) that the legend of the eighty years' reign is due to a misunderstanding, the Nūdīah raid having really taken place in the year 80 of the Lakshmaṇasena era. Dates in that era were usually calculated as expired years, but occasionally as current years. On the supposition that the year was 'expired', the year 80 would be A. D. 1119-20 plus 80 = A. D. 1199-1200 (October to October). If the current year should be understood, the date would be A. D. 1198-9 (November to October). Probably the event took place during the cold season of 1199-1200, that is to say, late in A. D. 1199, early in A. H. 596. We may be confident that it occurred in either A. H. 595 or 596, not in A. H. 590, as I formerly believed.

c32.12

Kielhorn's view that the conquest of Nūdīah must have taken place in the eightieth year of the Lakshmaṇasena era is supported by the Jānibīghā inscription of the year 83 (A. D. 1202) of the same era.¹ It is tolerably clear that the era commenced with the reign of the king whose name it bears, and who, as shown in the text (pp. 421, 422 *ante*), was a widely respected king and patron of literature.

c32.13

¹ *J. B. & O. Res. Soc.*, vol. iv, part iii (1918), pp. 266 ff. and pp. 273-80,

c32.14

For Vijayasena we have three synchronisms. He is described as 'the friend of Choṛaganga (*Choṛaganga sakhaḥ*)'. Choṛaganga had an exceptionally long reign of more than seventy-one years, from A. D. 1076 to 1147. The earlier part of it coincides with thirty-eight years of the reign of Vijayasena, according to the chronology followed by R. D. Banerji, which seems to me now to be correct. The other two synchronisms are vague and imperfect. An inscription records that Vijayasena made captive four kings, namely, Nānya, Vīra, Rāghava, and Vardhana. We are also told that he 'impetuously assailed the lord of Gauḍa, put down the prince of Kāmarūpa, and defeated the Kalinga'. Unluckily, the record does not join the names of the kings and the countries. Nānya may mean Nānyadeva of Tirhūt, who, according to tradition, founded Simraun in A. D. 1097 and afterwards established the Karṇāṭaka dynasty in the valley of Nepāl.² I cannot positively identify either Vīra or Vardhana. One of them presumably must have been the Rāja of Kāmarūpa or Assam.

c32.15

² Nānya certainly was a Karnāta king of Mithilā and contemporary with Vijayasena n the eleventh and beginning of the twelfth century, and probably also with Jayachchandra of Kanauj (Monmohan Chakravarti in *J. & Proc. A. S. B.*, 1915, pp. 409–11).

c32.16

I conclude this dissertation by a notice of the origin and rise of the Sena royal family.¹ The ancestors were of southern origin, from the Deccan, and are described both as Karṇāṭa Kshatriyas, and as Brahmakshatras. The meaning of the latter term, misunderstood by Professor Kielhorn, has been elucidated by Mr. D. R. Bhandarkar. His observations, which throw much light on the history of caste, deserve to be quoted textually :—

c32.17

¹ The genealogy of the Sena dynasty given by R. D. Banerji is :—

c32.18

‘We have already seen that a Chāṭsū inscription speaks of a Guhilot king Bhartṛibhaṭṭa as *Brahma-Kshatr-ānvita*, which I have translated by “possessed of both priestly and martial energy”, but a footnote has been added below saying that what is also implied is that Bhartṛibhaṭṭa was a Brahmakshatrī, i. e. belonged to the Brahmakshatri caste. Bhartṛibhaṭṭa is not the only ancient king of India who is so called. In the Deopārā inscription of Vijayasena, of the well-known Sena dynasty of Bengal, Sāmantasena is described as *Brahma-kshatriyānāṁ kula-śiro-dāma*, which expression was rendered by Prof. Kielhorn by “head-garland of the clans of the Brahmans and Kshatriyas”, but which ought to have been rendered, I think, by “head-garland of the Brahma-kshatrī family”. That the latter is the correct translation is shown by the term *Brahma-kshatra* used with reference to the Sena kings in the *Ballāla-charita* [Bibl. Ind.].

c32.19

Now, there is a caste called Brahmakshatrī, corresponding to this *Brahmakshatra*, the members of which are found all over the Panjāb, Rājputānā, Kāthiāwāṛ, Gujarāt, and even the Dekkan. In my opinion, as already stated, they were originally the Brāhmaṇa classes of new tribes afterwards turned Kshatriyas, before their final mergence into the Hindu society.’

c32.20

The author then cites the case of the Bandhārā weavers and

c32.21

``` Vīra

c32.22

Sāmanta

c32.23

Hemanta

c32.24

Vijaya

c32.25

Ballāla

c32.26

Lakshmaṇa

c32.28

Mādhava Viśvarūpa Késava ```

c32.29

(*J. & Proc. A. S. B.*, 1914, p. 98). The *lakshana*, or emblem of the dynasty, was then *Sadāśiva-mudrā*, a seal with a seated figure of the 10-armed form of Śiva, called Sadāśiva (ibid., p. 99).

c32.30

dyers in the Jodhpur State, who originally were Nāgar Brahmans, and proceeds:—

c32.31

‘Here then we have an instance of a Brahmakshatrī caste, the people of which say that they were originally Nāgar Brāhmaṇas. This clearly explains how the Guhilots, who were also originally Nāgar Brāhmaṇas, became Brahmakshatrīs or Khatrīs, and also strengthens my theory that the various castes of the Brahmakshatrīs were originally the Brāhmaṇa classes of foreign tribes, which after the process of fusion had set in, but before it was complete, exchanged their priestly for martial pursuits.’¹

c32.32

¹ Note that the Kanaujiya Brahmans, who supplied many sepoys to the old army, used to say, if irritated, *Ham kshatriya-Brahman hain*, as a boast (J. Wilson, *Indian Caste*, ii, 151).

c32.33

Mr. Bhandarkar is perfectly right. Consequently, the ancestor of the Sena kings must have been a Brahman from the Deccan, probably employed in the natural office of a Brahman as a minister. When he passed from ministerial to ruling functions, he became a Brahmakshatrī, his descendants being accepted as full Kshatriyas, capable of intermarriage with other ruling families reckoned as Kshatriyas. Most likely Sāmantasena had been in the service of the king of Kalinga or Orissa. The establishment of Sāmantadeva as a semi-independent chief in northern Orissa may have occurred somewhere about the middle of the eleventh century. Possibly he may not have been a ruling chief. His son, Hemantasena, may have been the first to act as Rāja.

c32.34

According to the authority quoted below, the earliest actually known seat of the Senas was at Kāśīpurī, the modern Kasiāri, on the Suvarṇarekhā river, in the Mayūrabhanja State, the most northerly of the Orissan Tributary States, adjoining the Midnapore District. I quote from the admirable *Archaeological Survey Report* of Bābū Nagendranāth Vasu.

c32.35

‘We have read in the genealogical history of the Paśchātya Vaidika of Bengal, written on palm-leaves and about three hundred years old, that the royal Sena dynasty reigned in a place called Kāśīpurī and situated on the banks of the Suvarṇarekhā. Two sons were born to Vijayasena, one of the rulers of this place; the elder being named Malla and the younger Śyāmala. It was the latter that conquered eastern Bengal and made the city of Vikramapura his capital. According to the Paśchātya Kulamañjarī, Śyāmalavarma’s sway in Vikramapura commenced in Śaka 994, i. e. 1072 A. D. . . . There is no doubt that the ancient name of Kāśīpurī has now degenerated into Kāsiārī.

c32.36

The matter, however, is not free from doubt.²

c32.37

² See ‘Earliest Seat of the Senas’ by S. Kumar (*Ind. Ant.*, Dec. 1915, p. 270 ff.), and also by the same author, ‘The Inscriptions of Asokashalla’ (ibid., p. 215). It is not easy to see how Kāśīpurī could become Kāsiārī. An alternative synonymous name, Kāśīwārī, may have existed. ‘The Senas, who replaced the Pālas in the twelfth century, are believed on acquiring Varendra, to have made their capital at Bijayanagar near Godāgāri in the south-west of the tract, and to have subsequently moved to Lakshmaṇāvatī, the town which afterwards took the name of Gauḍa’ (*J. R. A. S.*, 1914, p. 101). Varendra, also spelt Varendrī, the modern Barind, may be defined as 'comprising the Districts of Rājshāhi, Māldā, Dinājpur, Rangpur, Bogra, and part of Pabna—in fact almost the whole of the Rājshāhi Division' (*Ep. Ind.*, xiii (1916), p. 285). Godāgāri is a busy mart on the Ganges, where the Calcutta and Mālda road crosses the river. Gauḍa is the Sanskrit way of writing Gaur.

c32.38

I cannot follow out the problems of local history suggested by that passage, and the observations which follow in the work cited.

c32.39

In order to save the necessity of a multitude of foot-notes the principal references are appended in a classified form.

c32.40

References

c32.41

The following classified list gives the authorities on which the statements in the text and appendix concerning the Senas are based. Obsolete publications are not cited.

c32.42

It is difficult to interpret the account of 'the four Senas' by Tāranāth (Schiefner, pp. 252–7). He gives the names of the kings as (1) Lavasena, (2) Kāśasena, (3) Maṇitasena, (4) Rāthikasena; observing that although he was unable to fix the duration of each reign, all four together ruled for not more than about eighty years. His account of the Turushka king Chandra, who conquered all Magadha, destroyed Vikramaśīla, and slew many clergy in Otantapurī (Bihār town), seems intended to describe the raid of Muhammad the son of Bakhtyār, but why that personage should be described as Chandra I cannot say. He proceeds (p. 256) to enumerate the later Senas, viz. (1) Lavasena II, (2) Buddhasena, (3) Haritasena, and (4) Pratītasena, princes of small power, subordinate to the Turushkas or Muhammadans.

c32.43

In the third edition the identity of Lakshmaṇasena of the inscriptions with Rāe Lakhmaniya of the *Ṭabaḳāt-i-Nāṣirī* was assumed. Fresh light has now been thrown on the subject by K. P. Jayaswal in *J. B. & O. Res. Soc.*, vol. iv, pp. 266–72 and by H. Panday (ibid., pp. 273–80, with facs.). It seems reasonable to suppose that Lakshmaṇasena of the era was dead long before Muhammad's raid, and that Lakshmaṇasena II, the Lavasena of Tāranāth, came to the throne after the three sons of Lakshmaṇasena I, shown in the genealogy on p. 435 *ante*. This thesis is supported by the copper-plates of Viśvarūpa-Sena and Keśava-Sena, whose capital was at Gaur. They must have preceded Muhammad, who made the same place his capital. The dates of the plates are respectively the 14th and 3rd regnal years. Lakshmaṇasena therefore was dead before A. D. 1182 (1199 less 17), and a good deal before that date, as the elder brother reigned first. The victory over the Muhammadans claimed by the sons of Lakshmaṇasena must have occurred before the raid of 1199, i. e. it must have been won against the Ghorī armies, which had advanced as far as Benares.

c32.44

Blochmann, *J. A. S. B.*, part i, vol. xliv (1875), p. 275 ; Raverty, in reply, ibid., vol. xlv (1876), p. 320, and transl. *Ṭabakāt*, App. D ; Monmohan Chakravarti, ‘Appendix on Sena Kings’, *J. & Proc. A. S. B.* (N.S.), vol. i, 1905, pp. 45–50 ; and ‘Certain Disputed or Doubtful Events in the History of Bengal, Muhammadan Period’, ibid., vol. iv, 1908, p. 151.

c32.45

In addition to the above papers—Nagendranāth Vasu, *J.A.S.B.*, part i, vol. lxv (1896), pp. 6–38 ; Babu Akshay Kumar Mitra, ibid., vol. lxix (1900), p. 61 ; Kielhorn, *Ind. Ant.*, xix (1890), p. 6 ; and *Ep. Ind.*, i, 306 ; Beveridge, *J. A. S. B.*, part i, vol. lvii (1888), pp. 1–7 ; R. D. Bandyopādhyāya, ‘Mādhāinagar Grant of Lakshmaṇasena’, *J. & Proc. A. S. B.*, vol. v (N.S.), 1909, p. 467 ; and Prof. N. K. Bhattasali, ‘King Lakshmaṇa Sena of Bengal and his Era’ in *Ind. Ant.*, vol. xli (July 1912), pp. 167–9.

c32.46

Monmohan Chakravarti, ‘*Pavana-dūtam*, or Wind-Messenger, by Dhoyīka, a court poet of Lakshmaṇa-sena, king of Bengal,’ *J. & Proc. A. S. B.* (N.S.), vol. i (1905), p. 41 ; ‘Supplementary Notes on the Bengal Poet Dhoyīka and the Sena Kings,’ ibid., vol. ii (1906), p. 15 ; ‘Sanskrit Literature in Bengal during the Sena Rule,’ ibid., p. 157.

c32.47

Monmohan Chakravarti, ‘Chronology of the Eastern Gaṅga Kings of Orissa,’ *J. A. S. B.*, part i, vol. lxxii (1903), p. 14, quoting *Vallāla-charitaṁ* of Ānanda Bhaṭṭa.

c32.48

For Rāghava, Monmohan Chakravarti, *J. & Proc. A. S. B.* (N.S.), vol. i (1905), p. 49. For Nānya, S. Lévi, *Le Népal*, t. ii, p. 198 ; Kielhorn, *Ep. Ind.*, i, p. 313, note 57. For kings of Assam named Vīra, Gait, *Report on the Progress of Historical Research in Assam*, Shillong, 1897, pp. 11, 19.

c32.49

Nagendranāth Vasu, *Archaeological Survey of Mayūrabhanja*, published by the Mayūrabhanja State, 1911, p. 122.

c32.50

D. R. Bhandarkar, ‘Guhilots,’ *J. & Proc. A. S. B.* (N.S.), vol. v, 1909, pp. 167–87, especially p. 186 ; an exceptionally valuable and original essay.

Linked texts (3)

Vallālacarita — Ānandabhaṭṭa 1510, Sanskrit
Pavanadūta — Dhoyī, Sanskrit
Ṭabaqāt-i Nāṣirī — Minhāj al-Dīn Jūzjānī