Geography of Early Buddhism

Chapter II: The Uttarāpatha or Northern India


[48] Nowhere in Brahmanical or Buddhist literature is mentioned the four boundaries of the Uttarāpatha. According to the Brahmanical tradition as recorded in the Kāvyamīmāṁsā (p. 93), the Uttarāpatha or Northern India lay to the other, i.e., the western side of Prithudaka (Prithudakāt parataḥ Uttarāpathaḥ) or Pehoa, about 14 miles west of Thāneswar.

Other Brahmanical sources, e.g., the Dharmasūtras of Vaśiṣṭha, Baudhāyana and Manu, purport to furnish practically the same evidence, i.e., the Uttarāpatha lies to the west of the place where the Saraswatī disappears.

But our knowledge of the eastern boundary of Uttarāpatha is derived only in connection with the boundaries of the Madhyadeśa as given in the texts referred to above. There is nowhere any independent evidence of the boundaries of Uttarāpatha as such.

It is interesting to note that the Brahmanical definition of Āryāvarta excludes the greater portion of the land of the Rigvedic Aryans, which, however, is included in the Uttarāpatha. Thus the entire Indus valley which was the cradle of the Rgvedic culture and civilisation is practically outside the pale of Manu’s Madhyadeśa or Baudhāyana’s Āryāvarta, but is included in Uttarāpatha according to the Kāvyamīmāṁsā.

The Buddhist northern division is also to be located, as in Brahmanical texts, to the west of the Brahman district of Thūna (Sthūna) or Thaneswar as recorded in the Mahāvagga and the Divyāvadāna. There too the boundaries of Uttarāpatha as such are not recorded; its eastern boundary alone can be derived from the western boundary of the Majjhimadesa.

There are numerous references to Uttarāpatha in Pāli literature. In the Hāthigumphā inscription of King Khāravela, we are told that King Khāravela was able to strike terror into the heart of the King of Uttarāpatha. He compelled King Bahasatimita of Magadha to bow down at his feet.

Khāravela’s Uttarāpatha probably signifies the region including Mathurā in its south-eastern extension up to Magadha.

From the prologue of Book V of the Suttanipāta (p. 190), it appears the Dakkhiṇāpatha lent its name to the region through which it passed – the whole tract of land lying to the south of the Ganges and to the north of Godāvarī being known, according to Buddhaghosa, as Dakkhiṇāpatha, or the Deccan proper (VT., Mahāvagga, V, 13; Cullavagga, I, 18, p. 362).

Uttarāpatha too may supposed to have been originally a great [49] trade route – the northern high road, so to speak, which extended from Sāvatthī to Takkasīlā in Gāndhāra, and have lent, precisely like the southern high road, its name to the region through which it passed, i.e., the region covering, broadly speaking, the north-western part of the United Provinces, and the whole of the Punjab and the North-western Frontier Provinces. But this definition of Uttarāpatha is nowhere explicitly stated in Pāli literature. It is, therefore, not at all improbable that Uttarāpatha in Pāli literature might have also signifled the same region, i.e., the entire northern India from Aṅga in the east to Gandhāra in the north-west and from the Himalayas in the north to the Vindhyās in the south as under stood by its later and wider sense (i.e., the whole of Āryāvarta), e.g., in the Cālukya inscriptions of the 7th and 8th centuries A.D.

Bānabhaṭṭa, the author of Harsha-Carita, however, uses the word Uttarāpatha in its narrower sense and seems to include within the region so named the western part of U.P., the Punjab and the North-western Frontier Provinces. According to Chinese Buddhist writers, northern India ‘comprised the Punjab proper including Kashmīr and the adjoining hill states with the whole of eastern Afganisthan beyond the Indus, and the present Cis-satlej States to the west of the Saraswatī river’ (CAGI, p. 13).

Two Mahājanapadas

(i) Gandhāra:

In the Aṅguttara Nikāya, Gandhāra is included in the list of the sixteen Mahājanapadas (AN., 1., p. 213; IV, pp. 252, 256, 260). The Gandhāras were a very ancient people.

Their capital Takshasīlā is also mentioned in the Mahābhārata in connection with the story of King Jātamejaya who is said to have conquered it. ‘The Purāṇas represent the Gandhāra kings as the descendants of Druhyu (Matsya, 48. 6; Vāyu, 99. 9). This king and his people are mentioned several times in the Ṛgveda. In the Vedic Index (I, 385) it is stated that from the tribal grouping it is probable that the Druhyus were a north-western people. Thus the Puranic tradition about the connection of the Gandhāras with Druhyu accords with Vedic evidence.’ (PHAI., 93.)

The kingdom of Gandhāra included Kashmīr and the Takshasīlā region (PHAI., p. 93) We find it otherwise in Jāt., III, 365.

Gandhāra comprises the districts of Peshawar and Rawalpindi in the northern Punjab as we find in the Mahāvaṁsa (Geiger’s tr., p. 82, n. 2) wherein it is stated that after the dissolution of the Third Buddhist Council, Moggaliputtatissa thera sent Majjhantika thera to Kāsmīra-Gandhāra for propagation of the Buddhist faith. Dr. Raichaudhuri points out (PHAI., p. 93) that the inclusion of Kāshmīr in tho Gandhāra kingdom is confirmed by the evidence of Hekataios of Miletos (B.C. 549–486) who refers to Kaspapyros = Kaśyapapura, i.e., Kashmīr (cf. Rājataraṅginī, I, 27) as is Gandharic city. Gandhāra thus comprised the whole [50] of the districts of Peshawar and Rawalpindi in the northern Punjab.

Takkasīlā or Taxila was the capital city of the Gandhāra kingdom, and according to the Jātakas (Telapatta Jātaka, 96, Susīma Jātaka, 163) it lay 2,000 leagues from Benares.

In the time of Nimi, King of Videha, Durmukha, King of Pañchāla. and Bhīma, King of Vidarbha, the throne of Gandhāra was occupied by Naggaji or Nagnajit (Kumbhakāra Jātaka; Aitareya Brāhmaṇa, VII, 34; Sat. Brāhmaṇa , VIII, 14.10). PHAI, p. 93.

In the Kumbhakāra Jātaka we are told that Naggaji’s capital was Takkasīlā.

The Jātakas testify to the evidence of trade relations between the Kashmīr-Gandhāra kingdom and Videha (Jāt., III, pp. 363–369).

In the Niddesa we are told (P.T.S., Vol. I, p. 154) that in Taxila people used to flock in the wake of trade and commerce to earn money.

The king ruling in Gandhāra contemporaneously with King Bimbisāra of Magadha was Pukkusāti who is said to have sent an embassy and a letter to his Magadhan contemporary as a mark of friendship. He is also said to have waged a war on King Pradyota of Avanti who was defeated.

The Behistun inscription of Darius (C. 516 B.C.) purports to record that Gadara or Gandhāra was one of the kingdoms subject to the Persian Empire; it, therefore, appears that some time in the latter half of the 6th century B.C., the Gandhāra kingdom was conquered by the Achaemenid kings. In the time of Asoka, however, Gandhāra formed a part of the empire of the great Buddhist Emperor; the Gandhāras whose capital was Takkasīlā are mentioned in his Rock Edict V.

(ii) Kamboja:

Kamboja is mentioned along with Gandhāra in the Aṅguttara Nikāya (I, p. 213; Ibid., IV, pp. 252, 256, 261) as one of the sixteen great countries of India. In the Paramatthadīpanī on the Petavatthu (P.T.S., p. 113) Dvārakā occurs along with Kamboja. But it is not expressly stated if Dvārakā was the capital of the Kamboja country. Dvārakā, in fact, was not really a city of Kamboja; nowhere in early or later Pāli literature is there any mention of the capital city of the Kamboja people, ‘We learn from a passage of the Mahābhārata that a place called Rājapura was the home of the Kambojas (Mahābhārata, VII, 4, 5; “Karṇa Rājapuraṁ gatvā Kāmboja nirjitā stvayā “). The association of the Kambojas with the Gandhāras enables us to identity this Rājapura with the Rājapura of Yuan Chwang which lay to the south or south-east of Punch (Watters, Yuan Chwang, Vol. I, p. 284). The western boundaries of Kamboja must have reached Kafiristan, and there are still in that district tribes like “Caumojne”, “Camoze” and “Camoje” whose names remind us of the Kambojas.’ (PHAI., p. 95.) nor of the location of their country, though it is certain that Kamboja must be located in some part of north-west India not far from Gandhāra. [51] Nandipura seems to be the only city of the Kambojas that is known from Luder’s Inscriptions, Nos. 176 and 472.

In the Sumaṅgalavilāsinī (I, p. 124), we are told that Kamboja was the home of horses.

The Commentary on the Kunāla Jātaka (Jāt., V, p. 446) gives us to know how the Kamboja people caught horses in the forest.

In one of the Jātakas (Jāt., Cowell, VI, 110 note) we are informed that the Kambojas were a north-western tribe who were supposed to have lost their original Aryan customs and to have become barbarous.

In the Bhūridatta Jātaka (Jāt., VI, p. 208) we are told that many Kambojas who were not Aryans told that people were purified by killing insects, flies, snakes, frogs, bees, etc. The Jātaka tradition is corroborated by that contained in Yāṣka’s Nirukta as well as in Yuan Chwang’s account of Rājapura and the adjoining countries of the north-west. The Nirukta would have us believe that in Yāṣka’s time the Kambojas had come to be regarded as a people distinct from the Aryans of India proper, speaking a different dialect.

Speaking of Rājapura, Yuan Chwang says, ‘From Lampa to Rājapura the inhabitants are coarse and plain in personal appearance, of rude violent disposition . . . they do not belong to India proper but are inferior peoples of frontier (i.e., barbarians) stocks’ (Watters – Yuan Chwang, I, pp. 284 ff).

It is stated in the Sāsanavaṁsa (P.T.S. 49) that in the 235th year of the Mahāparinibbāna of the Buddha, Mahārakkhita thera went to the Yonaka Province and established the Buddha’s sāsana in Kamboja and other places. The Kambojas are mentioned in the Rock Edicts V and XIII of Asoka.

They occupied roughly the province round about Rajaori, or ancient Rājapura, including the Hazārā district of the North western Frontier Province.

Janapadas, Nigamas, Puras, Gāmas, etc.


The Mahāvaṁsa (Geiger’s tr., p. 194) refers to the town of Alasanda which was the chief city of the Yona territory. Geiger identifies Alasanda with the town of Alexandria founded by Alexander near Kabul in the Paropanisadae country.

In the Milindapañho, however, Alasanda has been described as an island where in the village of Kalasigāma King Milinda was born (Trenckner, Milindapañho, pp. 82 and 83; CHI., p. 550).


From the Sivi Jātaka (Jāt., IV, p. 401) we know that Ariṭṭhapura was the capital of the Sivi kingdorm. Several Jātakas mention (e.g., Nimi Jātaka, No. 541) a king named Usīnara and his son Sibi; but whether this prince Sibi had anything to do with the Sibi people or their country, it is difficult to ascertain.

In a passage of the Ṛgveda (Vll, I8. 7) there is a mention of the Siva, people along with the Alinas, Pakthas, Bhalānasas and Viśānina.

Early Greek writers also refer to a country in [52] the Punjab as the territory of the Siboi.

It is highly probable that the Śiva country of the Ṛgveda, the Sibi country of the Jātakas (Ummadanti Jātaka, No. 527; Vessantara Jātaka, No. 547) and the Siboi country of the Greek geographers are one and the same.

Patañjali mentions a country in the north called Śiva-pura (IV, 2, 2) which is certainly identical with Sibipura mentioned in a Shorkot inscription (Ep. Ind., 1921, p. 6.)

The Siva, Sibi or Siboi territory is, therefore, identical with the Shorkot region of the Punjab – the ancient Sīvapura or Sibipur. ‘The Mahābhārata (III, 130–131) refers to a rāṣṭra of the Śivis ruled by King Usīnara, which lay not far from the Yamunā. It is not altogether improbable that the Usīnara country was at one time the home of the Śivis. We find them also in Sind, in Madhyamikā in Rājputānā (Vaidya, Med. Hindu India, I, p. 162; Carmichael Lectures, 1918, p. I73) and in the Dasakumāra-Carita, on the banks of the Kāverī.’ (PHAI., pp. I55–56, also f.n., No. 2.).

Besides Ariṭṭhapura there was another city of the Sibi kingdom called Jetuttara near Chitor (Vessantara Jātaka, No. 547).

Asitañjana Nagara:

In the Ghata Jātaka (Jāt., Vol. IV, p. 79) we are told that a king named Mahākaṁsa reigned in Uttarāpatha, in the Kaṁsa district, in the city of Asitañjana which, however, is difficult to be identified.


Uttarakuru is often mentioned in Pāli literature as a mythical region. It has also been mentioned in Vedic and later Brahmanical literature as a country situated somewhere north of Kashmīr.


Kalasigāma was the birth place of King Milinda (Milindapañho, p. 83); it was situated in the Island of Alasanda or Alexandria.


According to a Jātaka story (No. 406) the kingdom of Kāsmīr was included in the Gandhāra Kingdom.

It is stated in the Mahāvaṁsa that after the dissolution of the Third Buddhist Council, Moggaliputta Tissa thera sent Majjhantika thera to Kāsmīra-Gandhāra for propagation of the Buddhist faith. (See ante: Gandhāra).

During the reign of Asoka, Kāsmīra was included in the Maurya dominion. This is proved by the testimony of Yuan Chwang (Watters, I, pp. 267–71).


The Dīpavaṁsa (p. 16) refers to the Kurudīpa which, however, may be taken to be identical with Uttarakuru.


Takkasīlā (Sans. Takshasila) was the capital city of the Gandhāra kingdom, and according to the Jātakas (Telapatta Jātaka, No. 96; Susīma Jātaka, No. 163) it lay 2,000 leagues from Benares as already pointed out.

In Pāli literature Takkasīlā has been frequently mentioned as a great seat of learning in Ancient [53] India.

In the Vinaya Piṭaka (Mahāvagga, pp. 269–270) it is stated that Jīvaka, the royal physician received his education in medicine and surgery there.

In the Jātakas (I, p. 259; V, pp. 161, 210, 457) we are told that princes from various kingdoms went to Taxila for education.

In one of the Jātakas (Jāt., I, p. 447) it is stated that a young man of the Lāḷa country went to Taxila for education.

In another Jātaka (Jāt, II, p. 277) a very beautiful picture of the student life of those days has been drawn.

From the Cittasambhūta Jātaka (Jāt, IV, p. 391) we learn that education was eligible for upper classes alone, the Brāhmaṇas and khattiyas. Of the subjects taught, the first three Vedas and eighteen Vijjās are mentioned. Some of the Vijjās taught at Taxila are also mentioned in the Jātakas, e.g., the art of archery (Jāt., I, p. 356), the art of swordsmanship and the various arts (Jāt., V, p. 128.)

The Susīma Jātaka (Jāt, II, p. 4.-7) tells us that Bodhisatta, the son of a priest who was a Hatthimaṅgalakāraka to the King of Benares, travelled a distance of 20,000 yojanas and went to Takkasīlā to learn Hatthisuttaṁ.

References to Ālambanamantaṁ (mantaṁ for charming snakes) and Nidhi-uddharaṇamantaṁ as taught in Taxila are made in the Campeyya Jātaka (Jāt., IV, p. 457) and the Vrahāchatta Jātaka (Jāt, III, p. 116) respectively.

From the Divyāvadāna (p. 371) it appears that Takkasīlā was included in the empire of Bindusāra of Magadha, father of Asoka. Once when during his reign there was a rebellion in Takkasīlā, he sent his son Asoka to put down the rising. From the minor Rock Edict II of Asoka it seems that Takkasīlā was the headquarters of the Provincial Government at Gandhāra and was placed under a kumāra or viceroy.

According to the Divyāvadāna, a rebellion again broke out in Takkasīlā during the reign of Asoka, and the latter sent his son Kunāla to put down the disturbances.

Takkasīlā is identified with Taxila in the district of Rawalpindi in the Punjab.


In the Samantapāsādikā (p. 179) there is a reference to Uttarakuru and its city Tidasapura.


Maddaraṭṭha is not mentioned in the list of the sixteen Mahājanapadas.


In the Milindapañho we are told that King Milinda (Menander), a powerful Graeco-Bactrian King, ruling over the Madda country with Sāgala as his capital became a convert to Buddhism (S.B.E.,Vol. XXXV, p. 6).

That Sāgala or Sākala (modern Sialkot in the Punjab) was the capital of the Madra country is also attested to by the Mahābhārata (ll, 32, l4) – ‘Tataḥ Sākalamabhyetva, Madrānāṁ putubhedanam’, as also by several Jātakas (e.g., the Kāḷiṅgabodhi Jātaka., No. 479); the Kusa Jātaka, [54] No. 531).

The Madras had a monarchical constitution and their territory may be said to correspond roughly to Sialkot and its adjacent districts which were known as late as the 18th century as the Madradeśa.

In one of the Jātakas (Cowell’s Jātaka, V, pp. 146–147) we are told that King Okkāka had a son named Kusa who married a daughter of the King of Madda. It is further stated that King Okkāka went with a great retinue from Kusāvatī, his capital, to the city of Sāgala, capital of the Madda King.

From the Kāliṅgabodhi Jātaka (Cowell’s Jātaka, IV, PP- 144–145) we know that a matrimonial alliance was established between the King of Madda and the King of Kāliṅga. Another matrimonial alliance of the Madda King was made with the royal house of Benares (Chaddanta Jātaka – Cowell’s Jātaka, V, p. 22).

The Mahāvaṁsa (p. 70) tells us that in Sīhapura, on the death of King Sīhavāhu, his Son Sumitta became king, and married the daughter of the Madda King and had three sons by her.


It is referred to in the Rock Edicts V and XIII of Asoka. The Nabhapantis of Nābhaka In the Rock Edicts V and XIII of Asoka, the Yonas, Kambojas, Gāndhāras, Rāshtrikas-Pitinikas, Bhojas Nābhapantis, Andhras and Pulindas are mentioned. We have to take these names as those of subject people, forming some of the frontier districts of Asoka’s Empire. must be looked for somewhere between the North-west Frontier and the western coast of India.

Yona or Yonaka:

The Yonaka or Yona country was visited, according to the Dīpavaṁsa and Mahāvaṁsa (Chap. XII) by the Thera Mahārakkhita.

According to the Sāsanavaṁsa (p. 12) the Yonakaraṭṭha is the country of the Yavana or Yona people.

The Rock Edicts V and XIII of Asoka mention the Yonas as a subject people, forming a frontier district of Asoka’s Empire. The exact situation of the Yonaka country is difficult to be determined.

According to the Mahāvaṁsa, its chief city was Alasanda identified with Alexandria near Kabul in the Paropanisadae country (Mahāvaṁsa, tr., p. 194; Trenckner, Milindapañho, p. 82).

Rivers, Lakes, Tanks, etc.


Anotatta has been mentioned as a lake in the Aṅguttara Nikāya (IV, p. 101) and is included in the list of the seven great lakes in the Himalayas (Dv. and Mv.). Buddha is said to have visited the lake many a time. It is generally supposed that the Anotatta or Anavatapta lake is the same as Rawanhrad or Langa. But Spence Harmy considers it to be an imaginary lake (Legends and Theories of the Buddhists, p. 129).


The river Uhā is stated in the Milindapañho (p. 70) to have been located in the Himavanta.


[55] In the Milindapañho (p. 114) we are told of the five hundred rivers that issued forth from the Himavanta mountain. Of these rivers ten are said to be important: Gaṅga, Yamunā, Aciravatī, Sarabhū, Mahī, Sindhu, Sarassatī, Vetravatī, Vitaṁsā and Candabhāgā. The Candabhāgā (Sans. Candrabhāgā) is the Chināb, the Acesines of the Greeks or the Asiknī of the Ṛgveda, a tributary of the Indus or the Sindhu.


Vītaṁsa (Milindapañho, p. 114) represented by the Sanskrit Vitastā is the river Jhelum, the Hydaspes of the Greeks.


It has been described in the Kunāla Jātaka (Jāt, Vol. V, p. 415) as a lake in the Himavanta. Tiyaggala has been described in the same Jātaka to be another lake in the Himavanta.


Of the five hundred rivers referred to in the Milindapañho as issuing from the Himavanta (p. 114), Sindhu is one of the most important. It is the river Indus, the Sintu of the Chinese travellers.

Mountains, Hills, Precipices, etc.


Añjana has been described in the Sarabhaṅga Jātaka (Jāt.,Vol. V, p. 133) as a mountain situated in the Mahāvana or Great Forest. It is the Sulliman range in the Punjab.

Anoma, Asoka, and Cāvala:

These are mountains not far from the Himavanta (Apadāna, pp. 342, 345 and 451 respectively).


In the Abbhantara Jātaka (Jāt., II, p. 396) we are told that the Kañcana pabbata is in the Himavanta. From the Nimi Jātaka (Jāt, VI, p. 101) we know that it is in the Uttara Himavanta.


The Nisabha pabbata is not far off from the Himavanta (Apadāna, p. 67). It is the mountain which lies to the west of the Gandhamādana and north of the Kabul river called by the Greeks Paropanisos, now called the Hindukush.


The Nandamūlappabhāra is in the Uttara Himavanta (Jāt., II, p. 195).