Pramāṇa · Search · About · API

Cunningham's Ancient Geography of India c15: Appendix I

Cunningham's Ancient Geography of India c15

Language
English

Appendix I

c15.1

PURANIC NINE DIVISIONS OF GREATER INDIA.

c15.2

Sir Alexander Cunningham has pointed out, in his Ancient Geography of India (p. 7), that the *Mahâbhârata*, the *Purânas* and *Bhâskarâchârya*, the astronomer, have given names of a Nine-Division of India and that the names are *Indra*, *Kaserumat*, *Tâmraparna*, *Gabhastimat*, *Kumârika*, *Nâga*, *Saumya*, *Vâruna* and *Gândharva*. No clue has been given, remarks he, to their identification. But he has suggested that *Indra* was the eastern division, *Vâruna* the western and *Kumârika* the middle, while *Kaseru* must have been the northern one.

c15.3

Alberuni also has quoted this Purânic account and has added the following description:—(1) *Indradvîpa*—the middle, (2) *Kaserumat*—eastern, (3) *Tâmraparna*—south-eastern, (4) *Gabhastimat*—southern, (5) *Nâga*—south-western, (6) *Saumya*—western, (7) *Gândharva*—north-western, (8)—————, (9) *Nagarasamvritta*—north-eastern.

c15.4

Thus the two authorities differ not only as to the location of the divisions, but as to one name also—the *Kumârika* of Cunningham appearing as *Nagarasamvritta* in the list of Alberuni who has not mentioned the 8th name.

c15.5

Let us now refer to the original sources. The majority of the *Purânas** agree in reading, in the

c15.6

* भारतस्यास्य वर्षस्य नव भेदान् निबोध मे ।
समुद्रान्तरिता ज्ञेयास्ते त्वगम्याः परस्परम् ॥
इन्द्रद्वीपः कशेरुमान् ताम्रपर्णी गभस्तिमान् ।
नागद्वीपस्तथा सौम्यो गान्धर्वो वारुणस्तथा ॥
अयं तु नवमस्तेषां द्वीपः सागरसंस्तृतः ।
योजनानां सहस्रं वै द्वीपायं दक्षिणो रात् ॥
पूर्वे किराता यस्यान्ते पश्चिमे यवनास्तथा ।

c15.7

*Bhuvanakosha* section [*Mārkandeya* LVII; *Matsya* CXIV; *Vāyu* XLV, etc.], the first seven names of Alberuni but without giving any direction. Their eighth name is *Vāruna*. As for the ninth, they read

c15.8

“अयं तु नवमस्तेषां द्वीपः सागरसंवृतः।” “And this sea-girt isle is the ninth of them.” Alberuni’s *Nāgarasamvritta* is thus a corruption of *Sāgarasamvrita* meaning ‘Sea-girt’. The Purānas do not give any name to it and refer to it as a *dvīpa* too well-known to require the mention of its name. Rājasekhara, however, in his *Kāvyamīmāmsā* (Gaekwad Oriental Series No. 1.) named it *Kumārī* and has supplied a clear clue to locate it by stating:—

c15.9

“तत्रेदं भारतवर्षम्। अस्य च नव भेदाः। रम्बूद्वीपः [etc.] . . . . . कुमारीद्वीपस्यायं नवमः। ......अत्र च कुमारीद्वीपे

c15.10

विन्ध्यश्च पारियात्रश्च सुक्तिमानृक्षपर्वतः। महेन्द्रस्सह्यमलयाः सहैते कुलपर्वताः॥ Page 92

c15.11

And the “Sea-girt ninth *dvīpa*” has also been thus described in the Purānas: “This *dvīpa* is a thousand yojanas from S. to N. At its east end are the Kirātas and at the west end are the Yavanas.” [*Mārkandeya*, LVII, vv. 5-11.]

c15.12

By combining these two accounts we find that the “Sea-girt ninth *dvīpa*” called *Kumārī* was the tract peopled by the Yavanas (Greeks) at its west end and by the Kirātas (Mongoloid tribes) at its east end and having the mountain-chains now known as the Vindhyas, Western and Eastern Ghats and the Nilgiris in it. It is thus the whole of India, or almost the whole of it. And it has been described as one of the nine divisions of *Bhāratavarsha*. But it is absurd to take the whole as equal to its part. Hence either *Kumārī-dvīpa* cannot be the whole of India or the term ‘*Bhāratavarsha*’ has been used here in a wider sense. But as the description of *Kumārī* is very clear, we cannot but take it as equal to almost the whole of India. So ‘*Bhāratavarsha*’ is used here in a wider sense, in the sense of Greater India, i.e.

c15.13

India proper and her colonies, eight *dvīpas* not far from it. That the *dvīpas* were separated from the main land of India by water is clear from the following Purānic account: "Hear from me the nine divisions of this country of *Bhārata*; they must be known as extending to the ocean, but as being mutually inaccessible." [*Mārkaṇḍeya*, LVII, 5.]. It is also to be added in this connection that the word *dvīpa* has been derived by Pāṇini as *dvi*+*ap*. It thus means land having water on two of its sides. Thus *dvīpa* is not identical with 'island'. It includes peninsulas and sometimes *doabs* also.

c15.14

As for the identification of the other *dvīpas*, it requires no comment to take *Tāmraparṇa* as Ceylon. It is to the south of India. *Indradvīpa* is to be located to the east. For India is the eastern *Dikpāla*. The Purāṇas also corroborate it in a passage which describes the courses of seven rivers rising from बिन्दुसरस् in the Himalayas—the three western rivers, *Sitā*, *Chakshu* and *Sindhu*, the *Gaṅgā*, and the three eastern rivers, *Nalinī*, *Hlādinī* and *Pāvanī*. One of these eastern rivers is described as rising from बिन्दुसरस् in the Himalayas, flowing to the east and then to the south and then emptying its water into the Ocean near *Indradvīpa* [इन्द्र-द्वीप-समीपे तु प्रविष्टा लवणोदधिम् *Matsya*, CXXI. 57.] Thus *Indradvīpa* was Burma. And this conjecture seems to be supported by so great an authority as Ptolemy. While describing India beyond the Ganges, Ptolemy (M'Crindle, p. 219) mentions the country of *Kirrhadia* [=the किराताः placed, in the Purāṇas, to the east end of the ninth '*dvīpa*'—the 'Sea-girt' *Kumārī*] producing *Malabathrum*; then he locates the Silver country [Arakan] and then the "Gold country" [the *Suvarṇabhūmi* of Buddhist literature and the *Soṇāparānta* of Burmese documents]. And again he remarks (M'Crindle, p. 221), between the ranges of Bepyrrhos and Damassa the country furthest north is inhabited by the Aninakhai [occupying the mountain region to the north of the Brahmaputra, corresponding to a portion of Lower Assam—M'Crindle's note, p. 222] ; to the south of these Ptolemy places the Indaprathai. Thus in the *dvīpa* or peninsula of Burma and just to the south of Lower Assam we hear the name of Inda or Indra. *Prathai* is to be connected with *Prastha* meaning a plain level country.

c15.15

Indra-*dvīpa* was, thus, Burma and it was to the east and *Tāmraparna* (Ceylon) to the south of India. Hence Kaserumat which is mentioned, in the Purānas, between them is to be located to the S.-E. of India. The word means 'abounding in excellent Kaserus' (called Kesur in Bengali and Kaseru in Hindi) for which Singapur is famous. So I propose to identify Kaserumat with the Malay Peninsula in the Wellesley district of which was discovered a fourth century A. D. Pillar inscription of the Buddhist Sea-Captain Mahānāvika Budha-gupta of Raktamr(i)ttikā (in Murshidabad district) showing that the Hindus were acquainted with it.*

c15.16

The only other *dvīpa* which I can identify with certainty is *Gāndharva*. It is identical with *Gāndhāra*, the valley of the Kabul with a small tract of land to the east of the Indus. Its position in the Puranic list of eight *dvīpas* [Indra (E.), Kaserumat (S.E.), *Tāmraparna* (S), *Gabhastimat* (S.W.), Nāga (W.), Saumya (N.W.), *Gāndharva* (N.) and Vāruna (N.E.)†] would suggest that it is the northern *dvīpa* (=doab) and Indian geographers placed *Gāndhāra* to the N. (and not N.W.) of India.‡

c15.17

∗ Kern's *Verspreide Geschriften*, III (1915), p. 255. † Varuna is the lord of West and so Varuna ought to be located to the west. But the order of the *dvīpas* as mentioned in the Purānas would suggest that it is in the N.-E. ‡ See the Bhuvanakosha of the Purānas—मार्कण्डेय, LVII; मत्स्य CXIV; वायु, XLV—and the कूर्मविभाग of the बृहत्संहिता.

c15.18

That the country of Gândhâra was also known as the Land of the Gandharvas is clear from the following verses of the Rāmāyana :—

c15.19

अयं गन्धर्वविषयः फलमूलौपशोभितः ॥
सिन्धोरुभयतः पार्श्वे देशः परमशोभनः ।
तं च रक्षन्ति गन्धर्वाः सायुधाः युद्धकोविदाः ॥
[Rāmāyana, Uttarakānda, CXIII, 10-11.]

c15.20

तच्च तक्षशिलायां तु पुष्कलं पुष्कलावतं ।
गन्धर्वदेशे कर्षि गान्धारविषये च सः ॥
[Uttarakānda, CXIV, 11.]

c15.21

These verses mean: This exceedingly charming country on both the banks of the Sindhu [Indus] decorated with fruits and roots is the land [ विषय ] of the Gandharvas. It is protected by the Gandharvas who are expert in fighting. [They were defeated by Bharata, the brother of Rāma ; their country was divided into two provinces, each of which was governed by a son of Bharata.] He (Bharata) installed his son Taksha at Takshasilâ and [his other son] Pushkala at Pushkalâvatî [identified with modern Charsada ; Peucelaotis of classical writers] in the charming Gandharva-country (also called) Gândhâra-vishaya (district). We thus see that Gândharva was Gândhâra. It was, as Yuan Chwang has aptly remarked, the border-land of the Barbarians who were Indians in culture and religion (i.e. Buddhism). So it was considered as a separate dvīpa included within Greater India and not as a part of India proper.

c15.22

As for the four other dvīpas, a search is to be made for them keeping in view their directions as suggested by the order of their names in the Purânic list. Gabhastimat, Nâga, and Saumya are to be located in S.-W., W. and N.-W. respectively. And we have Laccadive, Maldive or Ernaculam in the S.-W., Salsette, Elephanta (meaning the same as Nâga or Elephant), and Kathiawar in the W., and Cutch in the N.-W. (according to the direction of Kûrmavibhâga and Bhuvanakosha). Vâruna of N.-E. seems to be the Indian colony in Central Asia the exploration and research in connection with which by Sir A. Stein and a host of Russian, French, German, English and Japanese scholars are supplying new light on Indian culture.

c15.23

The above are my suggestions for the location of the eight dvîpas of the Purânas. As for the location of Indra, Tâmraparna and Gândharva there cannot be any doubt. Tâmraparna has long ago been correctly identified. The two others I identify—Indru on the authority of the Purânas and Ptolemy and Gândharva on the authority of the Râmâyana. As for the location of others I offer suggestions only. But what I have pointed out is enough to show that the Purânic nine divisions of Bhâratvarsha are not so many provinces of India but of Greater India.

c15.24

[Reprinted from my article in the Journal of the Bihar and Orissa Research Society 1922, March.]

Linked texts (9)