The Ancient Geography of India/Eastern India

From Jatland Wiki
Jump to navigation Jump to search
Wikified by:Laxman Burdak, IFS (Retd.)

Back to Index of the Book

The Ancient Geography of India: I.
The Buddhist Period, Including the Campaigns of Alexander, and the Travels of Hwen-Thsang
Sir Alexander Cunningham
Trübner and Company, 1871 - India

Eastern India

[p.499]: In the seventh century the division of Eastern India comprised Assam and Bengal proper, together with the Delta of the Ganges, Sambhalpur, Orissa, and Ganjam. Hwen Thsang divides the province into six kingdoms, which he calls Kamarupa, Samatata,


Tamralipti, Kirana Suvarna, Odra, and Ganjam[1] and under these names I will now proceed to describe them.

1. Kamarupa

From Paundra Varddhana, or Pubna, in Middle India, the Chinese pilgrim proceeded for 900 li, or 150 miles, to the east, and crossing a great river, entered Kia-mo-leu-po, or Kamarupa, which is the Sanskrit name of Assam.[2] The territory is estimated at 10,000 li, 1667 miles, in circuit. This large extent shows that it must have comprised the whole valley of the Brahmaputra river, or modern Assam, together with Kusa-Vihara, and Butan. The valley of the Brahmaputra was anciently divided into three tracts, which may be described as the Eastern, Middle, and Western districts, namely, Sadiya, Assam proper, and Kamrup. As the last was the most powerful state, and also the nearest to the rest of India, its name came into general use to denote the whole valley. Kusa-Vihara was the western division of Kamrup proper ; and as it was the richest part of the country, it became for some time the residence of the rajas, whose capital, called Kamatipura, gave its name to the whole province.[3] But the old capital of Kamrup is said to have been Gohati, on the south bank of the Brahmaputra. Now, Kamatipura, the capital oi Kusa-Vihara, is exactly 150 miles, or 900 li, from Pubna,[4] but the direction is due north ; while Gohati is about twice that distance, or say 1900 li, or 317 miles, from Pubna, in a north-east direction. As the position of the former agrees exactly with the distance recorded

  1. See Map No. I.
  2. Julien's ' Hiouen Thsang,' iii. 76.
  3. ' Ayin Akbari,' ii. 3. " Kamrup, which is also called Kamtah."
  4. See Map No. I.

[p.501]: by the pilgrim, it is almost certain that it must have been the capital of Kamrup in the seventh century. This would seem to be confirmed by the fact that the language of the people differed but slightly from that of Central India. It was therefore not Assamese, and consequently I infer that the capital visited by Hwen Thsang was not Gohati, in the valley of the Brahmaputra, but Kamatipura, in the Indian district of Kusa-Vihara. The great river crossed by the pilgrim would therefore be the Tista, and not the Brahmaputra.

On the east Kamrup touched the frontiers of the south-western barbarians of the Chinese province of Shu; but the route was difficult, and occupied two months. On the south-east the forests were full of wild elephants, which is still the case at the present day. The king was a Brahman, named Bhaskara Varmma, who claimed descent from the god Narayana, or Vishnu, and his family had occupied the throne for one thousand generations. He was a staunch Buddhist, and accompanied Harsha Varddhana in his religious procession from Pataliputra to Kanoj, in A.D. 643.

2. Samatata

The capital of the kingdom of Samatata, or San-mo-ta-cha, is placed at from 1200 to 1300 li, or from 200 to 217 miles, to the south of Kamrup, and 900 li, or 150 miles, to the east of Tamralipti, or Tamluk.[1] The first position corresponds almost exactly with Jasar, or Jessore, which is most probably the place intended. The bearing and distance from Tamluk would take us to the uninhabited part of the Sundari-vana, or Sundarbans,

  1. Julien's 'Hiouen Thsang,' iii. 81. See Map No. I.

[p.502]: between the Huranghata river and Bakarganj. But in a country so much intersected by watercourses as Lower Bengal, the road distance is about one-fourth greater than the direct distance, measured on the map. Thus, Jessore, which is 103 miles from Dhakka, and 77 miles from Calcutta by road, is only 82 and 62 miles distant from them by direct measurement. Accordingly, Hwen Thsang's distance of 150 miles by route will not be more than 120 miles by direct measurement on the map, which is only 20 miles in excess of the actual direct distance between Jessore and Tamluk. But as Tamluk is not approachable by land from the east, the pilgrim must have travelled at least one-half of the route by water, and his distance of 150 miles may be accepted as a fair estimate of the mixed route by land and water, which could not be actually measured. The name of Jasar, or " The Bridge," which has now supplanted the ancient name of Murali, shows the nature of the country, which is so completely intersected by deep watercourses, that before the construction of the present roads and bridges, the chief communication was by boats. Murali, or Jasar, is most probably the Gange regia of Ptolemy.

The country of Samatata is mentioned in the inscription of Samudra Gupta on the Allahabad pillar,[1] in which it is coupled with Kamrup and Nepal. It is mentioned also in the geographical list of Varaha Mihira, who lived in the beginning of the sixth century.[2] According to Professor Lassen, the name signifies "bas pays littoral," which accords exactly with

  1. Journ. Asiat. Soc. Bengal, vi. 793 ; line 19 of inscription,
  2. Dr. Kern's ' Brihat-Sanhita,' xiv. 6.

[p.503]: Hwen Thsang's description of it as a low, moist country on the seashore. The inhabitants were short and black, as is the case at the present day with the people of Lower Bengal. From all these concurrent facts, it is certain that Samatata must be the Delta of the Ganges ; and as the country is described as 3000 li, or 500 miles, in circuit, it must have included the whole of the present Delta, or triangular tract between the Bhagirathi river and the main stream of the Ganges.

Hwen Thsang mentions several countries lying to the east of Samatata, but as he gives only the general bearings and not the distances, it is not easy to identify the names. The first place is Shili-cha-ta-lo, which was situated in a valley near the great sea, to the north-east of Samatata[1] This name is probably intended for Sri-Kshatra, or Sri-Kshetra, which M. Vivien de Saint-Martin has identified with Sri-hata, or Silhat, to the north-east of the Gangetic Delta. This town is situated in the valley of the Megna river, and although it is at a considerable distance from the sea, it seems most probable that it is the place intended by the pilgrim. The second country is Kia-mo-lang-kia, which was situated beyond the first, to the east, and near a great bay. This place may, I think, be identified with the district of Komilla, in Tipera, to the east of the Megna river, and at the head of the Bay of Bengal. The third country is To.lo.po.ti, which was to the east of the last. M. Julien renders the name by Dwaravati, but he makes no attempt to identify it. I would, however, suggest that it may be Talaingvati, that is, the country of the Talaings, or Pegu. Vati is

  1. Julien's 'Hiouen Thsang,' iii. 82.

[p.504]: the common termination of the names of the Burmese districts, as Hansavati, Dwayavati, Dinyavati, etc. The next name is I-shang-na-pu-lo, which was to the east of the last ; then still further to the east was Mo-ho-chen-po, and beyond that to the south-west was the kingdom of Yen-mo-na-cheu. The first of these names I take to be the country of the Shan tribes, or Laos ; the second is probably Cochin China or Anam; and the third, which M. Stanislas Julien renders by Ya-mana-dwipa, is almost certainly Yava-dwipa, or Java.


The kingdom of Tan-mo-li-ti, or Tamralipti, is described as 1400 or 1500 li, about 250 miles, in circuit.[1] It was situated on the seashore, and the surface of the country was low and wet. The capital was in a bay, and was accessible both by land and water. Tamralipti is the Sanskrit name of Tamluk, which is situated on a broad reach or bay of the Rupnarayan river, 12 miles above its junction with the Hughli. The district probably comprised the small but fertile tract of country lying to the westward of the Hughli river, from Bardwan and Kalna on the north to the banks of the Kosai river on the south.

From Tamalitti, the Pali form of the name, came the classical Tamalites.

4. Kirana Suvarna

Hwen Thsang places the capital of Kie-lo-na-su-fa-la-na, or Kirana Suvarna, at 700 li, or 117 miles, to the north-west of Tamralipti, and the same distance to the north-east of Odra or Orissa.[2] As the capital of

  1. Julien 's 'Hiouen Thsang,' iii. 83.
  2. Julien 's 'Hiouen Thsang,' iii.84 and 88. See Map No. I.

[p.505]: Orissa in the seventh century was Jajipur on the Vaitarani river, the chief city of Kirana Suvarna must be looked for along the course of the Suvarna-riksha river, somewhere about the districts of Singhbhum and Barabhum. But this wild part of India is so little known that I am unable to suggest any particular place as the probable representative of the ancient capital of the country. Bara Bazar is the chief town in Barabhum, and as its position corresponds very closely with that indicated by Hwen Thsang, it may be accepted as the approximate site of the capital in the seventh century. The territory was from 4400 to 4500 li, or from 733 to 750 miles, in circuit. It must, therefore, have comprised all the petty hill- states lying between Medinipur and Sirguja on the east and west, and between the sources of the Damuda and Vaitarani on the north and south.

This large tract of country is now occupied by a number of wild tribes who are best known by the collective name of Kolhan or Kols. But as the people themselves speak various dialects of two distinct languages, it would appear that they must belong to two different races, of whom the Munda and the Uraon may be taken as the typical representatives. According to Colonel Dalton,[1] "the Mundas first occupied the country and had been long settled there when the Uraons made their appearance;" and "though these races are now found in many parts of the country occupying the same villages, cultivating the same fields, celebrating together the same festivals and enjoying the same amusements, they are of totally distinct origin, and cannot intermarry without loss of caste." This

  1. Journ. Asiat. Soc. Bengal, 1866, p. 154.

[p.506]: difference of race is confirmed by the decisive test of language, which shows that the Uraons are connected with the Tamilian races of the south, while the Mundas belong to the hill men of the north, who are spread over the Himalayan and Vindhyan mountains from the Indus to the Bay of Bengal.

The various tribes connected with the Mundas are enumerated by Colonel Dalton[1] as the Kuars of Elichpur, the Korewas of Sirguja and Jaspur, the Kherias of Chutia Nagpur, the Hor of Singhbhum, the Bhumij of Manbhum and Dhalbhum, and the Santals of Manbhum, Singhbhum, Katak, Hazaribagh, and the Bhagalpur hills. To these he adds the Juangas or Pattuns (leaf-clad) of Keunjar, etc. in the Kataka tributary districts, who are isolated from " all other branches of the Munda family, and have not themselves the least notion of their connection with them ; but their language shows that they are of the same race, and that their nearest kinsmen are the Kherias. The western branches of this race are the Bhils of Malwa and Kanhdes, and the Kolis of Gujarat. To the south of these tribes there is another division of the same race, who are called Suras or Suars. They occupy the northern end of the eastern Ghats.

According to Colonel Dalton, [2] the Ho or Hor tribe of Singhbhum is "the nucleus of the Munda nation." He calls it "the most compact, the purest, the most powerful and most interesting division of the whole race, and in appearance decidedly the best-looking.

  1. Journ. Asiat. Soc. Bengal, 1866, 158. I write Santal in preference to Sonthal, as I believe that the short o is only the peculiar Bengali pronunciation of the long ā.
  2. Journ. Asiat. Soc. Bengal, 1866, 168.

[p.507]: In their erect carriage and fine manly bearing the Hos look like a people that have maintained and are proud of their independence. Many have features of sufficiently good cast to entitle them to rank as Arians; high noses, large but well-formed mouths, beautiful teeth, and the facial angle as good as in the Hindu races. . . . When the face of the Munda varies from the Arian or Caucasian type, it appears rather to merge into the Mongolian than the Negro. . . . They are of average stature, and in colour vary from brown to tawny yellow."

In the different dialects of the Munda language Ho, Hor, Horo, or hoko is the term for "man." The assumption of this name by the people of Singhbhum is a strong confirmation of Colonel Dalton's description of the tribe as the most powerful division of the Munda nation. But they also call themselves Larakas, or the "warriors," which points to the same conclusion that they are the leading division of the Munda race.

Colonel Dalton gives no explanation of the name of Munda ; but as I find that the head men of the villages are called Munda or Moto amongst the Hors of Singhbhum and other divisions of the Munda race, I conclude that the Mundas or Motos must once have been the ruling division of the nation. The name of Munda is found in the Vishnu Purana[1] as the appellation of a dynasty of eleven princes who succeeded the Tusharas or Tokhari. In the Yayu Purana, how-ever, the name is omitted, and we have only Marunda, which is most probably the variant form of another name, Murunda, as found in two inscriptions

  1. Wilson's translation, edited by Hall, book iv. 24, and vol. iv. p. 203.

[p.508]: of the second and third centuries.[1] Ptolemy has Marundai as the name of a people to the north of the Gauges ; but to the south of the river he places the Mandali, who may be the Mundas of Chutia Nagpur, as their language and country are called Mundala. This is only a suggestion ; but from the position of the Mandali they would seem to be the same people as the Monedes of Pliny, who with the Suari occupied the inland country to the south of the Palibothri.[2] As this is the exact position of the country of the Mundas and Suars, I think it quite certain that they must be the same race as the Monedes and Suari of Pliny.

In another passage Pliny mentions the Mandei and Malli as occupying the country between the Calingae and the Ganges,[3] Amongst the Malli there was a mountain named Mallus, which would seem to be the same as the famous Mount Maleus of the Monedes and Suari. I think it highly probable that both names may be intended for the celebrated Mount Mandar, to the south of Bhagalpur, which is fabled to have been used by the gods and demons at the churning of the ocean. The Mandei I would identify with the inhabitants of the Mahanadi river, which is the Manada of Ptolemy. The Malli or Malei would therefore be the same people as Ptolemy's Mandalae, who occupied the right bank of the Ganges to the south of Palibothra. Or they may be the people of the Rajmahal hills who are called Maler, which would appear to be

  1. Samudra Gupta, about A.D. 125 ; and a copper-plate dated in 214 or A.D. 293.
  2. Hist. Nat. vi. c. 22. " Ab iis (Palibothris) in in teriore situ Mo-nedes et Suari, quorum mons Maleus," etc.
  3. Hist. Nat. vi. c. 21. " Gentes : Calingae proximi mari, et supra Mandei Malli, quorum mons Mallus, finisque ejus tractus est Ganges."

[p.509]: derived from the Kanarese Male and the Tamil Malei, a "hill." It would therefore be equivalent to the Hindu Pahari or Parbatiya, a "hill man."

The Suari of Pliny are the Sabarae of Ptolemy, and both may be identified with the aboriginal Savaras, or Suars, a wild race of wood-cutters, who live in the jangals without any fixed habitations. The country of the Savaras is said to begin where that of the Khonds ends, and to extend as far south as the Pennar river. But these Savaras or Suars of the eastern Ghats are only a single branch of a widely-extended tribe, which is found in large numbers to the south-west of Gwalior and Narwar, and also in southern Rajputana. The Savaris or Saharias of the Gwalior territory occupy the jangals on the Kota frontier to the westward of Narwar and Guna. They are found along the course of the Chambal river and its branches, where they meet the Rajputana Surrias of Tod. The name is preserved in the Sorae Nomades of Ptolemy, who are placed to the south of the Kondali and Phillitae, or the Gonds and Bhils. They must therefore be the Suars or Savaras of central India, who occupy the wild hilly country about the sources of the Wain Ganga, and who are also found along the valley of the Kistna river. As Kiraṇa means a "man of mixed race," or barbarian, it seems probable that the name of Kirana Suvarna may be the original appellation of the barbarian Suvaras, or Suars.

In the beginning of the seventh century the king of this country was, or Shashanka, who is famed as a great persecutor of Buddhism. [1]

  1. 'Hiouen Thsang's Life,' i. 112 and 235. Also ' Travels ' ii 349 422, and 468.

[p.510]: I found a gold coin inscribed with the name of this prince at full length in the ' Payne Knight Collection' of the British Museum, and there are a few specimens in other collections.

5. Odra or Orissa

The kingdom of U-cha. or Oda, corresponds exactly with the modern province of Odra, or Orissa. By a reference to the ' Biography of Hiouen Thsang,'[1] it would appear that the capital of Odra was 700 li to the south-west of Tamralipti, and as this bearing and distance agree with the position of Jajipura, I think that the pilgrim must have returned to Tamluk from Kirana Suvarna before proceeding to Odra. In the travels of the pilgrim[2] the bearing and distance are taken from Kirana Suvarna; but this is perhaps a mistake, as they are usually referred to the capital, which, whether we place it at Jajipur or at Katak, is due south of Kirana Suvarna.

The province was 7000 li, or 1167 miles, in circuit, and was bounded by the great sea on the south-east, where there was a famous seaport town named Che-li-ta-lo-ching, or Charitrapura, that is, the " town of embarkation " or " departure." This was probably the present town of Puri, or "the city," near which stands the famous temple of Jagannath. Outside the town there were five contiguous stupas with towers and pavilions of great height. I presume that it is one of these which is now dedicated to Jagannath. The three shapeless figures of this god and his brother and sister, Baladeva and Subhadra, are simple copies of the symbolical figures of the Buddhist triad, Buddha,

  1. Julien, i. 181. See Map No. I.
  2. Julien, iii. 88.


Dharma, and Sangha, of which the second is always represented as a female. The Buddhist origin of the Jagannath figures is proved beyond all doubt by their adoption as the representative of the Brahmanical Avatar of Buddha in the annual almanacs of Mathura and Banaras.

The political limits of Orissa, under its most powerful kings, are said to have extended to the Hughli and Damuda rivers on the north, and to the Godavari on the south. But the ancient province of Odra-desa, or Or-desa, was limited to the valley of the Mahanadi and to the lower course of the Suvarna-riksha river. It comprised the whole of the present districts of Katak (Cuttack) and Sambhalpur, and a portion of Medinipur. It was bounded on the west by Gondwana, on the north by the wild hill-states of Jashpur and Singhbhum, on the east by the sea, and on the south by Ganjam. These also must have been the limits in the time of Hwen Thsang, as the measured circuit agrees with his estimate.

Pliny mentions the Oretes as a people of India in whose country stood Mount Maleus ;[1] but in another passage he locates this mountain amongst the Monedes and Suari; and in a third passage he places Mount Mallus amongst the Malli. As the last people were to the north of the Calingae, and as the Monedes and Suari were to the south of the Palibothri, we must look for the Oretes somewhere about the Mahanadi river and its tributaries. The Monedes and Suari must therefore be the Mundas and Suars, as already

  1. Hist. Nat. ii. 75. " In India; gente Oretum, mons est Maleus no-mine." See also vi. 22, "Monedes et Suari, quorum mons Mallus and VI. 21, " Malli, quorum mons Mallus."

[p.512]: noticed, and the Oretes must be the people of Orissa. Male is one of the Dravidian terms for a mountain ; and as the Uraons, or people of west Orissa, still speak a Dravidian dialect, it is probable that Mallus was not the actual name of the mountain. May not this have been the famous Sri-Parvat of Telingana, which gave its name to the Sri-Parvatiya Andhras ?

The ancient metropolis of the country was Katak on the Mahanadi river, but in the early part of the sixth century Raja Jajati Kesari established a new capital at Jajatipura on the Vaitarani river, which still exists under the abbreviated name of Jajipura. The same king also began some of the great temples at Bhuvaneswara ; but the city of that name was founded by Lalitendra Kesari. The language and pronunciation of the people is said to have differed from those of central India, which is still true at the present day.

To the south-west there were two hills, on one of which, called Pushpagiri, or the "hill of flowers," there was a monastery of the same name and a stupa of stone, and on the other to the north-west only a stupa. These hills I take to be the famous Udayagiri and Khandagiri, in which many Buddhist caves and inscriptions have been discovered. These hills are situated 20 miles to the south of Katak and 5 miles to the west of the grand group of temples at Bhuvaneswara. The stupas were said to have been built by demons ; from which I infer that the origin of the great caves, and other Buddhist works on these hills was quite unknown at the period of Hwen Thsang's visit.

6. Ganjam.

[p.513]: From the capital of Odra the pilgrim Xuanzang proceeded to the south-west for 1200 li, or 200 miles, to Kong-yu-to[1] This name has not been identified; but M. Vivien de Saint-Martin has, I think, indicated its true position in the neighbourhood of the Chilka lake. The capital was situated near a bay, or "junction of two seas," which can only be intended for the great Chilka lake and the ocean, as there is no other great sheet of water along this surf-beaten coast. Ganjam itself must therefore be the old capital. But as Ganjam is only 130 miles from Jajipur in a direct line measured on the map, or about 150 miles by road, I conclude that the pilgrim must have visited the hills of Udayagiri and Khandagiri and the town of Charitrapura, or Puri, on his way to Ganjam. By this route the distance would be increased to 165 miles by direct measurement, or about 190 miles by road, which agrees with the estimate of the Chinese pilgrim.

The Chinese syllables Kong-yu-to are rendered by M. Julien as Konyodha ; but there is no place of this name that I am aware of. I observe that M. Pauthier[2] writes the name Kiuan-yu-mo, which would seem to be intended for a transcript of Ganjam, of which the derivation is unknown. Hamilton proposes ganjam, "the depot," but this term is never used alone, so far as I am aware, but always in combi-

  1. Julien's ' Hiouen Thsang,' iii. 91. See Map No. I.
  2. 'Journal Asiatique,' 1839, p. 404. U-cha, or Oda, in eastern India, is said to be also named Kiuan-yu-mo, at which time, therefore, A.D. 650 to 684, it must have become dependent on Odra or Orissa.

[p.514]: nation, either with the founder's name or with the name of the principal article sold in the place, as Ram-ganj, or " Rama's market,"Thithar-ganj, the "brazier's market," etc. The district was only 1000 li, or 167 miles, in circuit, which shows that the territory was confined to the small valley of the Rasikulya river. But though the domain was small, the state would appear to have been of some consequence, as Hweng Thsang describes the soldiers as brave and bold, and their king as so powerful that the neigh-bouring states were subject to him, and no one could resist him. From this account I am led to infer that the king of Ganjam, at the time of Hwen Thsang's visit, must have been Lalitendra Kesari of the Orissa annals, who is said to have reigned for nearly sixty years, from A.D. 617 to 676. The pilgrim visited Ganjam in A.D. 639, when this prince was at the very height of his power. But only four years later, when the pilgrim revisited Magadha, he found the great King Harsha Varddhana of Kanoj[1], had just returned from a successful expedition against the king of Ganjam. The cause of the war is not stated, but as Harsha Varddhana was a staunch Buddhist, while Lalitendra was a devoted Brahmanist, the difference of religion would easily have furnished a sufficient pretext for war. It seems probable that Ganjam was then annexed to the dominions of the Kanoj king, and formed part of the province of Orissa.

Hwen Thsang notes that the written characters of Ganjam resembled those of central India, but that both the language and the pronunciation were different. This statement proves that the same alphabetical

  1. Julien's 'Hiouen Thsang,' i. 236,

[p.515]: characters were still in use over the greater part of India at as late a date as the middle of the seventh century. It also serves to show that the intercommunications of the Buddhist fraternities throughout India were not yet broken, although they must already have been much restricted by the steady progress of Brahmanism.