Geographical and Economic Studies in the Mahabharata: Upayana Parva c5: II

Geographical and Economic Studies in the Mahābhārata: Upāyana Parva c5

Language: English

II

c5.1

The date of the *Mahābhārata* remains still a matter of contention; Dahlmann's theory⁸ placed the *Mahābhārata* in the fifth or sixth century before Christ. His contention that it was the work of a single man has long been given up. The *Mahābhārata* is not the work of a single hand and as pointed out by Professor Hopkins the crystallisation of its present form may be put between the 4th century B.C. and 4th century A.D., though the latter limit seems to be too late. We are not concerned in our present study with the external and internal evidences which have been impressively marshalled to prove the date of the *Mahābhārata* or more correctly a particular part of the *Mahābhārata*;

c5.2

⁸ *Das Mahabharata als Epos und Rechtbuch*, (1895) and *Genesis des Mahabharata* (1899).

c5.3

⁹ The Cambridge History of India, Vol. I, p. 258.

c5.4

neither we are concerned here with the revival of Brahminism apparent from the *Mahābhārata* with its insistence on the worship of Viṣṇu. The only purpose of this section is to show by the internal evidences from the *Sabhāparva* or to narrow it down still from the *Digvijaya* and the *Upāyana-parvas* which give a mass of geographical informations useful to determine the probable date of the composition of the *Sabhāparva*. Every piece of evidence will be taken up separately and discussed on its own merits, and then we shall try to reach a certain conclusion. The chronology has always been rather a controversial point in Indian history before the pre-Christian era and no theory or conclusion has escaped criticism. Whatever views I propose to place before the scholars are certainly far from dogmatic and I would certainly welcome corrections and criticisms.

c5.5

The most important evidence which throws a considerable light on the date of the *Sabhāparva* comes from Arjuna's expedition (MB, II, 23-25). His campaign as we shall see later on could be divided into two parts or may be three parts. We are only concerned here with that section where Arjuna having conquered the Daradas with the Kāmbojas (II, 24.22) proceeded towards the north and having conquered the robber tribes, subdued the Lohas, Parama-Kāmbojas, the Ṛṣikas and Paramṛṣikas. I give the full text below noting the variant readings where necessary

c5.6

गृहीत्वा तु बलं सारं फल्गुनोत्सृज्य पाण्डवः
दरदानसह काम्बोजैरजयत्पाकशासनः 22
प्रागुत्तरां दिशं ये च वसन्त्याश्रित्य दस्यवः
निवसन्ति वने ये च तान्सर्वानजयत्प्रभुः 23
लोहानपरमकाम्बोजानृषिकाननुत्तरांनपि
सहितां स्तानमहाराज व्यजयत्पाकशासनः 24

c5.7

The variants for परमकाम्बोजान् are:—लोकान्, मौकान् होलान्, लहानपरमकाम्बोजान्, पण्ड़यानपरमकाम्बोजान्, लम्बकान्यपरमकाम्बोजान् etc. The variants for ऋषि are इषि or इषी।

c5.8

ऋषिकेषु तु संग्रामो बभूवातिभयंकरः
तारकामयसंकाशः परमर्षिक पार्थयोः 25
शुकोदरसमुख्यन्धान् समानयत्
मयूरसदशान्न्यातुभयानेवचापरान् 26
स विनिर्जित्य संग्रामे हिमवन्तं सनिष्कुटम्
श्वेत पर्वतमासाद्यन्यवसत्पुरुषर्षभः 27

c5.9

In these verses we get some very important information about the situation of the Ṛṣikas and Paramarṣikas. To locate them properly we should try to follow the route taken by Arjuna. After having conquered the Bāhlīkas (M.B. II, 23, 21) or the Bactrians in Northern Afghānistān, he subdued the combined Darada and Kamboja forces (M.B., II, 23, 22). The crux of the problem is the proper identification of Kamboja to which we shall come later on. It is not the Chhibhal country or Kabul but as proposed by Pandit Jayachandra Vidyālaṃkāra Badakshān and the Galcha speaking part of the Parmirs. Now what route Arjuna took from the Bāhlīka country? The key to this problem lies in the proper identification of Valgu. This may be identified with the river Baghlan. In their explorations of the Oxus countries Wood and Lord investigated the route which lay directly south from Kunduz by the river of that name to its junction with the Baghlan. Thence folllowing Baghlan to its head they crossed by the Murgh Pass into the Valley of the Andarab and diverging eastward they adopted the Khawak Pass to reach the Panjishir Valley and so to Kabul. No great difficulties were encountered on the route, involving only two passes between the Oxus and the Kabul, the Murgh (7,400 ft.) and Khawak (11,650 ft.) and it undoubtedly possesses many advantages as the modern popular route between Kabul and Badakshān.¹⁰ Arjuna probably followed this route on his return journey to the Śvetaparvata which may be identified with the Safīd-koh whose rugged ilex-covered spurs centre on the giant peaks of Pirghal and Shuidar, overlooking the plains of Afghānistān towards Ghaznī.¹¹ Arjuna however in his march towards the Paramakāmbojas and the Ṛṣikas left the Baghlan route which might have led him to Kabul and proceeded towards the north, and in the ensuing battle defeated the Kāmbojas and the Daradas, who probably came to help their allies through the Dora Pass, which is the chief pass over the Hindukush, directly connecting India (through Chitral) with Badakshān.¹² The next stage in his campaign took him to the north-easternly direction (*prāguttarām diśam*) (M.B., II, 24, 23) where the robber tribes (*dasyavaḥ*) having obtained asylum (*āśritya*) lived, and also the tribes living in the forest. These were conquered. Apparently these robber tribes were the descendants of the Eastern Iranian speaking ancestors of the Wakhānis, Shighnis, Roshanis and the Sarīkolīs of the Pāmīr plateau. Then comes the most important section of the campaign—the conquest of the allied forces (*sahitān*) of the Lohas, Paramakāmbojas and the Ṛṣikas of the north or the Great Ṛṣikas (M.B., II, 24). The Paramakāmbojas have been identified by Prof. Jayachandra Vidyālaṃkāra¹³ with the Galcha speaking Yāghnobis who live in the Valley of Yaghnob at the head water of the Zarafshān river, a tract of country considerably to the north of the Pāmīrs and separated from them by the hill states subordinate to Bokhārā,¹⁴ though he has not stated his reasons for this identification. The same writer has suggested the identification of the Yüe-chī's with the Ṛṣikas.

c5.10

¹⁰ Holdich, *The Gates of India*, pp. 434-35.

c5.11

¹¹ *Imperial Gaz.*, Vol. I, p. 10.

c5.12

¹² *Holdich, loc. cit.*, p. 427.

c5.13

¹³ *Bhāratbhūmi aur uske nivāsī*, p. 313. V.S. 1987.

c5.14

¹⁴ *Linguistic Survey of India*, Vol. X, pp. 455-56.

c5.15

The problem of identifying the Ṛṣikas with the Yüe-Chi is not a new one and is closely connected with the problem of Ārśi the language of the Śakas in the Central Asian texts, the Ārśi Kanta, i.e., Ārśi tongue.¹⁵ It has been also suggested to connect Yüe-chi with Ārśi on certain phonetic peculiarities in the Chinese which used üe to understand ri, ūr and o in the transliteration of Indian words.¹⁶ Klaproth derived the Yü-chï from Yetes. Franke takes them to be the people of Yet or Get; Baron von Stael Holstein infers a pronunciation Kurshi or Gurshi, and F. W. K. Müller maintains that Yüe-chï is probably a rendering of the same word which we have learned to know as Ārśi as a designation of the language of the Tocharians the view to which Franke has subsequently agreed.¹⁷ The battle royal over Ārśi has not yet finished. H. W. Bailey¹⁸ derives Ārśi from the Sanskrit Araya. Pelliot¹⁹ rejects the derivation of Ārśi from Ṛṣika as Pandit's etymology and the same author²⁰ finds the rejection of Ārśi as convincing.

c5.16

¹⁵ Sieg, SBAW, 1918, pp. 560 ff. quoted by Sten Konow (CII, II, p. viii, fn. 3).

c5.17

¹⁶ Hirth, Journal of the China Branch of the Royal Asiatic Society, 1886, p. 220.

c5.18

¹⁷ CII, II, p. lix.

c5.19

¹⁸ Ttaugara, BSOS, VIII, 1936, pp. 883, 905 sqq.

c5.20

¹⁹ JA, 1934, p. 23.

c5.21

²⁰ To'ung-Pao, XXXII, 1936, p. 265.

c5.22

Before we advance our own views about the Ṛṣikas it would be better to state briefly the facts already known about the migration of the Yüe-chï, and then try to reconcile it if possible with the information supplied by the *Mahābhārata* about the Ṛṣikas.

c5.23

The Yüe-chï first appear in history in Kansu province in the north-west of China, where they had apparently been living for some time. According to Chang Ki'en before they were defeated by the Hiungnu they lived between Tun-huang (now Sha-Chou) and K'i-lien (a hill south-west of Kan-chóu-fu).²¹ A struggle between the Yüe-Chï and the Hiung-nu, the progenitors of the later Huns, culminated in 176 or 174 B.C. in the complete defeat of the Yüe-chï. They were compelled to quit Kan-su and set out on their long journey westward. Part of the horde called by the Chinese Siao-Yüe-Chï, in contrast to the larger body the Ta Yüe-Chï unwilling to hazard and the long journey turned southward and settled into the Tarim Valley.²² The main horde going westward fell on the Wu-sun, killed the king, occupied their grazing ground and were again driven away by the Hiungnu. Still going westward they attacked Sai-Wang some time before 160 B.C., about the lake Issyk-Kul and the plain northward of the Alexandrovsky range and the Sai-Wang fled southward. Their subsequent movement will be examined later on. But in or just before 160 B.C. the Yüe-Chï were again attacked by the son of Wusun King with the help of the Hiung-nu and were driven out of the Sai-Wang country. The main body of the Yüe-Chï again went westward. After 160 B.C. the Yüe-Chï disappear for a generation reappearing shortly before 128 B.C.

c5.24

²¹ Hirth, JAOS, 1917, pp. 96-97.

c5.25

²² Ib. 1917, p. 97.

c5.26

What ultimately turned them southward is unknown. It seems that the intervening years between 160—128 B.C. were spent in fighting to settle down somewhere. Between 141—128 B.C. they crossed Jaxartes westward of Ferghana went southward and finished off the Greek Kingdom of Bactria. They were living north of the Oxus²³ when Chang K'ien visited them not having yet moved across the Oxus into Bactria though they had conquered it. The chronological sequence in the great movement of Yüe-Chï has been arranged by Hirth.²⁴ In the following sequence : In 176 the Yüe-Chï were defeated by the Hiung-nu for the second time; in 165 B.C. Lau Shang, Mau-tun's successor annihilated the Yüe-Chï and the Yüe-Chï fled westward; in 164 B.C. they settled down near Issyk-kul driving away Sai-Wang who migrated southward and became the rulers of Kipin; in 160 B.C. the Wusun drove away the Yüe-Chï and occupied their territory near Issyk-kul. It is necessary to remember these dates and the Yüe-Chï settlement near Issyk-kul to which we will have to return presently.

c5.27

²³ Hirth, p. 97.

c5.28

²⁴ Ib. pp. 133-34.

c5.29

Now let us return for a moment to the nomad conquest of Bactria. It has been a practice with the older writers to attribute the fall of Bactria to Sakas, though the statement of Chang K'ien who attributes the conquest to Ta Yüe-Chï should have been taken as final. This mistake according to W. W. Tarn²⁵ who has made the latest contribution to the problem of the nomad conquest of Bactria, arose from a simple passage of Strabo (XI, 511) in which he says that the Sakas occupied Bactria. But as the context proves he is talking of the Achaemenid times or probably the 7th century B.C., when the great Saka invasion well-known from the Assyrian sources which had played its part in the fall of Ninevah and had penetrated as far as Cappadocian Pontus.

c5.30

²⁵ *The Greeks in Bactria and India*, p. 283; London, 1938.

c5.31

Apollodorus attributes the conquest of Bactria to four nomad peoples namely Asii, Pasiani, Tochari and Sacarauli (Strabo. XI, 511). The 'Trogus Source' (Trogus Prol. XLI) formally attributes it to two—Asiani and Saraucae, though subsequently he mentions Tochari. Taking the Trogus source first one of the two names must represent Chang K'ien's Yüe-Chï and as Saraucae are out of the question the Asiani should represent the Yüe-Chï (Tarn, *loc. cit.*, p. 284). The form Asiani is an Iranian adjectival form of Apollodorus' Asii which is the substantival form; the Asii are therefore Yüe-Chï. This identification of Yüe-Chï with Asii lead to the great controversy.²⁶ From 1918—36 it was further believed that the Ārśi of the Central Asian texts was the language spoken by the Tocharians; its very existence was also denied. It has, however, been shown (Tarn, *Loc. cit.*, p. 285) that the Hellenistic world knew of a people called Ārśi even if Central Asia did not. The name occurs in a curious list of people in Pliny.²⁷ This only proves that Pliny was acquainted with the original name of the Asii but not knowing where to place them put them in the refuge list.

c5.32

²⁶ Sten Konow, CII., II, pp. iviii Sq; T. Pelliot, JA, 224, 1934, p. 25, n2.

c5.33

²⁷ VI, 16, 48, sqq.

c5.34

Another name which must be considered, as it has definite connection with the Paramarṣika of Arjuna’s campaign (M.B., II, 24, 25) is Pasiani. In the conquest of Bactria Apollodorus speaks of a Saka tribe Pasiani. As Asiani is the adjectival form of Asii so Pasiani should be a similar adjectival form of a name Pasii or Pasi; and there is no doubt that this name is Parsii of Greek geographers.

c5.35

Tarn locates the Pasii west of the Arius, Tapuria and Traxiane²⁸ and tries to identify them with Paras-Parsua—the Persian tribes who played an important rôle in the history of Iran. As the house of the ancient Persians was Eranvez identified with Khawarizm Tarn suggests that the Parsii-Parsua stayed behind in Eranvez, which their kinsmen migrated south and later on played their part in the eviction of the Greeks from Bactria. This argument is rather speculative.

c5.36

²⁸ Tarn, *loc. cit.*, p. 292.

c5.37

Now let us examine at some length as to what the *Mahābhārata* has to say about the Ṛṣikas, though unfortunately its information is rather meagre. In the *Ādiparva* (M.B., I. 61. 30) the homage of the eponymyous Ṛṣika King is traced from Candra and Diti. In this connection it is interesting to note Prof. Charpentier’s²⁹ suggestion that the word Yüe-Chï could be translated as ‘the moon clan.’ It is difficult to trace the connection of the Ṛṣikas with the Moon God, except the *Mahābhārata*. The Ṛṣikas again appear in the *Udyogaparva* (V. 4. 15) where they are mentioned in the company of the Śakas, Pahlavas, Daradas, Kāmbojas, and Paścimānūpakas. It is worthy to note that here as well they are placed side by side with the Kambojas. In the variant text of Kāmboja (D’jā) *ṛṣika-rājñamca* the adjectival form of *Ṛṣika* is also found. We shall discuss its import later on. Then we meet the Ṛṣikas in the north-easterly direction of Kamboja country (M.B., II, 24, 24-23) or Badakshān. In the Bhandarkar Oriental Research Institute’s edition of the *Sabhāparva* the Prākṛt form of the Ṛṣika namely Iṣi and Iṣī have been noted among the variants. These forms are very important as we shall presently see that the Greek historians were acquainted with both the Sanskrit and Prākṛt forms of the Ṛṣika. In another couplet (M.B., II. 24, 25) describing a battle between Paramarṣika and Arjuna, the adjectival form of the substantival form Paramārṣika is found. Thus we find that the *Mahābhārata* knew of the Ṛṣika, its adjectival form Ārṣika, its Prākṛt forms Iṣika and Iṣīka, and the adjectival form of Paramarṣika—Paramārṣika.

c5.38

²⁹ ZDMG, 71, 1917, p. 375.

c5.39

Now coming to the Greek forms of one of the nomad tribes who conquered Bactria and who have been identified with the Yüe-Chï by Tarn, we meet Asii in Apollodorus.³⁰ Asiani in the Trogus Source³¹ is the adjectival form of the Asii of Apollodorus. Tarn has also hunted out the original Arsi from Pliny (VI, 16, 48 Sqq). It is not difficult now to see that the Greek Asii is from Sanskrit Isi or Iṣi, and probably the Greek Arsi may be derived from Sanskrit Ārṣika.

c5.40

³⁰ Strābo, XI, 511.

c5.41

³¹ Trogus Prol. XLI.

c5.42

Now we come to Pasiani, another tribe who invaded Bactria according to Apollodorus. Tarn as we have already shown by a long winded argument tries to prove that they were Parsa—Parsua who played such an important part in the history of Iran. The Pasiani according to him were the remnant of these of tribes in the Eranvej. But we have in the Sabhāparva a tribe of the Paramarṣikas whose adjectival form Paramārṣika (II,24,25) has been mentioned. Could we not identify the Pasiani of Apollodoros with them? Apparently they were a separate tribe being the member of the Yüe-Chï clan with whom they fought against Arjuna.

c5.43

Now reverting to the campaign of Arjuna we find that the first part of his campaign was devoted to the reduction of the Kāmbojas of Badakshān. Then he proceeded in the north-easterly direction and reduced the robber tribes (M. B. II, 24, 23) and the Lohas, Paramakāmbojas and finally the Ṛṣikas (II. 24, 24). A glance at the map of eastern Afghānistān and adjacent countries should convince us that the Lohas and Kambojas and the robber tribes must have been settled in the country which is represented now by the Tadzhik Soviet Social Republic which till recently was divided in the Russian Wakhan, Shighnān, Roshan and Darwānz etc. It is known that the Yüe-Chï were in 160 B.C. or thereabout in the region of the Lake Issyk-Kul from where they were driven out by the Hiung-nu and Wu-sung. It seems probable that the author or authors of the Sabhāparva have very cleverly transposed the doing of the Hiung-nu to Arjuna. The union of the Yüe-chï and the Eastern Iranian speaking republics on the Oxus was natural in the face of common danger; moreover there were close ethnic relation between Tajiks and the Yüe-chï both of whom came from the common Iranian stock; this must have further cemented the bond of friendship.

c5.44

Another very important point which should not escape our attention is the adjective Uttara used in connection with the Ṛṣikas (Uttarān-ṛṣikān, II, 24, 24), which denotes here the sense of superiority or greatness and which is an exact equivalence of the Ta Yüe-Chï as opposed to the Sieou Yüe-Chï or little Yüe-Chï. After all the Sanskrit geographers were not so fantastic or ignorant as it is generally supposed.

c5.45

In the *Upāyanaparva* (*Sabhāparva*, Chapters 47 & 48) the allusions to the Śakas, Tukhāras, Kaṅkas, Cīnas, Hūṇas and the order in which they are described, and their relative positions determinable by the Chinese sources also throw considerable light on the date of the *Sabhāparva*. In a couplet (*M.B.*, II, 47, 19) the following order is given :—Cīna, Hūṇa Śaka Oḍra, the inhabitants of the mountainous country (*parvātāntaravāsinaḥ*); at another place Śaka, Tukhāra, Kaṅka (*M.B.*, II 47, 26) in their respective order are mentioned; at a third place Śauṇḍika, Kukkura and Śaka (*M.B.*, II 48, 15) appear. But before we take up the discussion any further it would be better to know something about the history of the Śakas and other tribes mentioned above.

c5.46

We need not bother ourselves with the earlier references to Śakas in Herodotus pointing their home in the plains east of the Caspian and north of Jaxartes in the Pamir country, to the north of Hindūkush and east of Bactria and Sogdiana and later on Seistan, and the allusions to the Saka *Tigrakhauda*, Haumavarka and Tardarya the last living on the east of the Caspian Sea as mentioned in the inscriptions of Darius³² as these Sakas lived at a time with which we are not concerned.

c5.47

³² Konow, CII. II, pp. XVII—XVIII.

c5.48

In the Chinese annals they are known as Sai and in the oldest sources they are spoken of as Sai-wang. Sometimes before 160 B.C. they were driven off from their own home by the Yüe-Chï who were in their turn driven out by Wu-sun whose settlements according to Prof. Franke, extended from Urumchi to the west of Issik-Kul, from the Dzungarian desert and down towards Tarim.³³ The Ts'ien Han Shu³⁴ speaks of several Sai states. It seems that they were greatly affected by the great nomadic movement in the second century B.C.

c5.49

³³ *Ib.*, p. XIX.

c5.50

³⁴ A. Wylie, *Notes on the Western Regions*, J. R. Anthro So., Vol. X, p. 34.

c5.51

A laconic statement in Ts'ien Han Shu states that the Sai-wangs after their defeat by the Yüe-Chï went southwards and made themselves masters of Kipin.³⁵ The routes over which they passed was the Hien-fu (the hanging passage) which according to M. Chavannes is the Bolar route through the Yasin Valley by which the travellers went to Wakhan, then to Indus and further to Kashmir or Udyāna.³⁶ According to Sten Konow however, the Kipīn country comprised the Swat valley and it extended westward towards Arachosia; whether it extended to Kabul is uncertain. The Saka occupation of Kipin is further supported by the description of Sha-mi, the present Mastuj by Yuan Chwang which mentions that the king of that country was of Śākya race. Sten Konow also quotes *Abhidhānacintāmaṇi* (V. 960) of Hemacandra in which the Lampākās are said to be *muruṇḍas* which shows that even at a comparatively late period the memory of the rule of the Saka Muruṇḍas had been preserved. This theory of the Saka occupation of Kipin has been rejected.³⁷ Their main movement, impeded by the Yavana power in Kabul, according to the *Cambridge History*, would naturally be westwards in the direction of Herat and thence southwards to Seistan.

c5.52

³⁵ Wylie, *loc. cit.*, p. 34.

c5.53

³⁶ CII, II, p. XXIII.

c5.54

³⁷ *Cambridge History of India*, Vol. I, p. 567.

c5.55

A novel suggestion has been put forward by Tarn³⁸ which deserves our consideration. According to him the Sai-wang flight southward, which more correctly should mean south-west, would take them across Jaxartes to Ferghana. At this stage, they must have ceased to be a hoard as the Chinese literature does not mention them. It is probable that some joined Ka'ng-kiu whose grazing ground was the Tashkant plain, but those who went to Ki-pin probably joined the Sacaraucae, who occupied Khojend and the Steppes west of it, for it does not appear how otherwise they could have reached India. The remainder of the Sai-wang horde settled in the Greek provinces of Ferghana—their name mixing up with the name of the Province which in Chinese was Ta-yuan. There they set up the Saka or rather nomad government which Chang K'ien found in 128 B.C.; they are represented as distinct from the Kang-kiu, but on good terms with them.

c5.56

³⁸ Tarn, *loc. cit.*, p. 278.

c5.57

The Tocharians have also received much attention. The scholars have generally occupied Marquart's identification of Ta-hia with Tu-ho-lo a designation met in the history of Northern Wei³⁹ (A.D. 386-556). But many scholars from Richtofen to Herzfeld have held that the Tochari were the Yüe-Chï. The word has been identified with the Thagouroi of Ptolemy (VI, 16, 5) on the silk route; the Chinese knew of remnants of the Togara in Kan-su in the second century B.C.; the name occurs in or about Kan-su in Tibetan texts; and the same name *taugara* for a town in Kan-su is found in a document in the Khotan Saka with date equivalent to as late as 800 A.D.⁴⁰ It follows that the Tochari in Bactria as mentioned by Apollodorus came from Kan-su, but the Chinese historians say that Yüe-chï came to Bactria at that time. Again we know from Apollodorus and Pliny that while there were no Tocharis in the Tarim basin in the reign of Euthydemus (died 190 B.C.) there were Tochari there later at the date of some source of Pliny (VI. 55), and here again we learn from the Chinese sources that the little Yüe-Chï came and settled there sometime not long after 174 B.C.; and the Indian writers called them by the same name Tokhāra.⁴¹ Finally Ptolemy unknowingly locates Tochari at several places where the Yüe-Chï are known to have been on their journey. Thagourai in Kan-su, Takoraioi north of Imaos, Tagouraioi near lake Issyk-Kul, Tachoroi in Sogdiana and Tocharoi in Bactria which alone should be conclusive.⁴² The Yüe-Chï hoard therefore was composed of two different peoples who appear in the Greek sources as Asii or Tochari. The Trogus source gives out the relationship where he says that the Asii are lords of the Tochari.⁴³

c5.58

³⁹ CII, II, Ivii.

c5.59

⁴⁰ Tarn, loc. cit., pp. 285-86.

c5.60

⁴¹ Sylvain Lévi, J.A., 1897, I, p. 10; 1933, p. 26.

c5.61

⁴² Tarn, *loc. cit.*, 286.

c5.62

⁴³ *Trogus Pro.* XLII, Tarn, *loc. cit.*, 286.

c5.63

The race and the language of the Tochari is a difficult problem. It was once supposed that they brought it from Europe and spoke the Centum language with the Italo-Celtic affinities discovered in Chinese Turkestan. Today two dialects said to be A and B the languages of two states in northern part of Chinese Turkestan A of Agni-Karachar (Turfan) and B of Kucha are known, and that none of these languages could be the language of historical Tochari, who invaded Bactria as their name is aspirated while dialect A and B have no aspirates.

c5.64

The Kaṅkas (M.B. II, 47, 26) may probably be identified with the Ka'ng-Kiu⁴⁴—the inhabitants of Sogdiana, placed in the company of Śakas and Tukhāras. The Ka'ng-kiu according to Chang-kien were under the political influence of Yüe-Chï⁴⁵ on the south and on the east under that of Hiung-nu.

c5.65

⁴⁴ Hirth, *loc. cit.*, p. 96.

c5.66

⁴⁵ Ib. *loc. cit.* p. 96.

c5.67

The relative position of the Śakas, if we take them to be settled in Ta-yuan, and the Kankas is settled as their country was conterminus.⁴⁶ The Tokharas who were probably a component tribe of the Great Yüe-Chï were perhaps encamped further to the south. It is therefore clear that the arrangement followed in the *Sabhāparva* in placing the Śakas, Tukhāras and Kaṅkas gives their relative position in the second century B.C. It is significant that the Ṛṣikas are missing in the list. The only conclusion which could be drawn from this is that after their defeat in 160 B.C. when they had moved towards the west, the Tukhāras were sent as vanguards and the information contained in this particular passage could be dated between 160 B.C. and 128 B.C. the year in which Yüe-Chi defeated the Greek power of Bactria. A provisional date of 150 B.C. may be assigned to this state of affairs.

c5.68

⁴⁶ Ib.

c5.69

Another passage which is quoted below also confirms our view that the *Sabhāparva* was probably composed in the second century B.C. The couplet in question is from the *Digvijayaparra* in which Sahadeva having conquered the Pāṇḍyas, the Draviḍas, the Coḍras, Kerlas, Andhras, Lavaṇas, Kaliṅgas and the Uṣṭrakarṇikas (M.B., II, 28, 48) subdued Antākhi and Romā by the despatch of the political missions only.

c5.70

अन्ताख्यं चैत रोमांच यवनानां पुरं तथा
दूतैरैव वशे चक्रे करं चैनानदापयत् M.B., II, 28, 49

c5.71

Here Antākhī is substituted by the editor as correct text, and probably he is right.

c5.72

Before coming to Sahadeva’s embassy to Antioch it is better for us to know certain broad facts in the history of the Seleucid kings. Alexander died in 323 B.C. and Seleucus I assumed the satrapy of his eastern empire. It was not till 311 B.C. that he was free to deal with his distant provinces. The date of his Indian expedition is stated to be 304 or 305.⁴⁷ He was defeated by Candragupta Maurya and had to transfer to his adversary the Satrapies of Arachosia (Kandhar) and the Paropanisadae (Kabul), with at least some portion of Gedrosia (Balūchistān) and Aria (Herat). In 293 he became joint king with his son Antiochus.⁴⁸ He was assasinated in 283 B.C. Seleucus was the founder of the city of Antioch—a city in Syria described as Epidaphnes or "on the Orantes" to distinguish it from the fifteen other Greek towns which like itself owed their foundation to Seleucus Nicator, and their names to his father Antiochus.⁴⁹ The probable date of the foundation of the city is 300 B.C.⁵⁰

c5.73

⁴⁷ *The Cambridge History of India*, Vol. I, p. 430.

c5.74

⁴⁸ *Ib.* p. 432.

c5.75

⁴⁹ *Encyclopaedia Br.*, Vol. I cf. Antioch.

c5.76

⁵⁰ *JAOS*, 58, p. 265.

c5.77

In the time of Antiochous I (283—261) and even earlier the friendly relation with the Mauryas continued. Athenaeus (1, 32, 18 D) has preserved the story of certain drugs sent as present by Candragupta I to Seleucus I and it is to the same writer (XIV, 67, 652 I and 653A) that we an anecdote of how Chandra-gupta's son Bindusāra wrote to Antiochus I requesting him to buy and send him some wine, figs and a sophist to teach him to argue. Antiochus replied sending the figs and wine but not the sophist who, he wrote, were not saleable. The intercourse was not only confined to these civilities. We know of Megasthenes' missions to the court of Chandragupta, and Daimachus of Plataea also went on mission or missions from Antiochus to Bindusāra. Pliny (His Nat, VI, 58) also speaks of a mission of certain Dionysius sent by Ptolemy Philadelphus (285—246 B.C.), which was sent probably to Bindusāra or Aśoka.⁵¹

c5.78

⁵¹ *The Cambridge History of India*, Vol. I, pp. 432-33.

c5.79

On the Indian side Aśoka (274—237—6 B.C.) as mentioned in the Thirteenth Rock Edict sent embassies to Antiochus II (261 B.C.), Ptolemy (Philadelphus of Egypt (285—247 B.C.), Antigonus (Gonatas of Macedon 278—239), Magas (of Cyrene, died 258) and Alexander (of Epirus 278—258?) probably in 258 B.C.⁵² Later on the contact was renewed when Antiochus III (221—187 B.C.) in 206 B.C. in his short campaign in India met the Mauryan Sophagasenus or Subhāgasena⁵³ who presented him elephants. Antiochus IV (175—163 B.C.), though his direct contact with India is not mentioned, was the moving spirit behind Eucratides who overthrew Demetrius.⁵⁴

c5.80

⁵² *Ib.* p. 502.

c5.81

⁵³ *Ib.* p. 512.

c5.82

⁵⁴ Tarn, *loc. cit.*, Chap. V.

c5.83

Now coming to the embassies sent by Sahadeva we are faced with a dilemma as to the period in which the embassy to Antioch was sent. Antioch was probably founded in about 300 B.C. and since then it had continuous friendly relation with the Mauryas which we have already described. The embassy was peaceful and not in connection with any war. Does it therefore reflect the embassy of Aśoka to Antiochus II in 258? But the date is too early on the strength of the evidences which we have already produced about the probable date of the Sabhāparva, though of course it is possible that in this couplet reference is being made to some previous event. Another point which should be marked is that at the time of the embassy the independent states of Antioch as the capital of free Seleucid state was recognised. This independence was greatly reduced after the battle of Magnesia in 187 B.C. when the Romans won over the forces of Antiochus III. Antiochus IV, though he was pulled sharply from time to time by the Romans to remind him of his allegiance to Roman State tried to reassert the waning power of the Seleucids. After his death in 163 B.C. Seleucid power declined rapidly. Antiochus VII (138 B.C.—129) the last strong representative of the old royal house fell fighting with the Parthians⁵⁵ in 129. After him follow a long line of rival kings fighting over what remained of the Seleucid Kingdom. The possibility is therefore that the embassy referred to may be dated either in the reign of Antiochus III (221—187 B.C.) or Antiochus IV (171—163 B.C.)

c5.84

⁵⁵ Cambridge Ancient History, Vol. VIII, pp. 529-30.

c5.85

The reference to Rome or its exact Latin feminine form Romā is somewhat difficult to explain in the 2nd century B.C. as no Indian embassy reached the Romans until Augustus in the 1st century B.C.⁵⁶ It seems however possible that the Indians who had frequent intercourse with the Seleucids of Syria were acquainted with the name of Rome and perhaps also with its growing power, and the author of the Mahābhārata could not restrain himself from adding the name of Rome in the conquests of the Pāṇḍavas. But this is a pure suggestion and must be treated as such.

c5.86

⁵⁶ *The Commerce between the Roman Empire and India*, pp. 35—38.

c5.87

Another allusion which is of some importance in determining the date of the Sabhāparva refers to the conquest of the Vāṭadhāna Brāhmaṇas in Madhyamikā by Nakula in his expedition to the west (M.B., II, 29, 7). The first line of the verse (Ib.) says, तथा मध्यमिकांश्च वाट धाना- न्द्विजानथ ‘and in Madhyamikā he conquered the Vāṭadhāna Brāhmaṇas’. The statement looks quite simple but its import is of considerable importance. The siege of Madhyamikā by the Yavanas was such a famous event in the second century B.C., in the reign of Puṣyamitra Śuṅga, as to merit its being cited as a grammatical example by Patañjali. We shall return to the events presently. In this connection a few words about the Śuṅgas would not be out of place here.

c5.88

The origin of the Śuṅgas is obscure, the word Śuṅga which denotes fig tree may perhaps be tribal. According to Pāṇini (IV, I, 117) they claimed to be the descendants of Bharadvāja, the Purohita of Divodāsa, the king of Titsus. They probably lived in the countries which under the Mauryan Empire were included in the viceroyalty of Ujjain.⁵⁷

c5.89

⁵⁷ *The Cambridge History of India*, I, p. 518.

c5.90

We are not concerned here with the various events of Puṣyamitra Śuṅga’s life. We are only concerned here with the invasion of India by Demetrius with his generals Apollodotus and Menender and the siege of Madhyamikā by the Yavanas in his time.

c5.91

This siege of Madhyamikā is supported to some extent by the grammarian Patañjali (a contemporary of the Śuṅga King, Puṣyamitra) who to illustrate the use of the imperfect tense (to denote an event which had recently happened) quotes the example.⁵⁸ ‘The Yavana was besieging Sāketa: The Yavana was besieging Madhyamikā.’ The siege of Madhyamikā is attributed to Apollodotus by Tarn⁵⁹ who concludes that the country was conquered and the Śibis⁶⁰ whose coins have been found at Nagarī near Chittor and whose original country was about Jhang in the Southern Panjab were brought and settled there by Apollodotus.⁶¹ This is however a pure speculation. The possibility is that the turmoil which must have followed in the wake of the Greek invasion compelled the Śibis to seek refuge elsewhere and their settlement at Madhyamikā probably was affected after the retirement of the Greeks. All these events must have happened by 175 B.C.⁶²

c5.92

⁵⁸ Kielhorn, *Ind. Ant.* VII, p. 266.

c5.93

⁵⁹ Tarn, *loc. cit.*, p. 150.

c5.94

⁶⁰ Ib. p. 151.

c5.95

⁶¹ Tarn, p. 151.

c5.96

⁶² Ib. p. 156.

c5.97

Now let us examine the information we get in the conquest of Madhyamikā by Nakula. The first important point is that the city must have been defended by the Vāṭadhāna Brāhmaṇas and that they were conquered. There is every possibility that the event here points to the siege of Madhyamikā, a town which has been identified with Nagarī near Chittor in Southern Rājputānā⁶³—the action being cleverly transferred from the Yavanas to Nakula. The probability is that the Śuṅgas were the Vāṭadhāna Brāhmaṇas as Vaṭa and Śuṅga both in Sanskrit mean Banyan tree. It is curious to note, however, that according to the *Manu-Smṛti* the Vāṭadhānas were the offspring of an outcaste Brāhmaṇa by a Brāhmaṇa mother (X, 21).

c5.98

⁶³ For references, Allan, *Coins of Ancient India*, pp. cxxii-cxxv.

c5.99

Lastly there is another list (M.B., II, 47, 19) in which the names of the Cīnas, Hūṇas, Śakas and Oḍras appear in a descending order. The appearance of the Hūṇas may at once make us jump to the conclusion that the text must be of the 5th century or later when the Indians to their cost came to know of the Hūṇas. But these are not the Hūṇas of the 5th century and neither their presence within the boundaries of India should be sought for. They probably represent the Hiung-nu settled on the borders of China, who drove away the Yüe-Chï and whose depredations caused the Han kings to build the Great Wall as a protection against their inroads. The order in which the tribes are mentioned here is quite appropriate. First come the great Chinese people then the Hūṇas or Hiung-nu from the Mongolian side, then the Saka tribes settled near Issyk-kul in the beginning of the second century B.C.⁶⁴ and then finally the Oḍras or the inhabitants of Swat about whom we shall have to say something later on. The mention of the Oḍras immediately after the Śakas probably points to the route which a section of the Śakas took after their defeat by the Yüe-Chï which landed them in Swat and thence to Ki-pin. In this list the author of the *Sabhāparva* seems to emphasise the ethnic importance of the races in their proper order.

c5.100

⁶⁴ CII, II, p. XIX.

c5.101

The foregoing discussion so far has tended to prove by various internal evidences to throw light on the probable date of the *Sabhāparva*. The geographical positions of the Ṛṣikas, Śakas, Tukhāras, Kaṅkas, Hūṇas, Cīnas, etc., have been discussed already. The light which is thrown by the allusions to Antakhī and Romā (which show connection with the Greek and the Roman world) on the probable date of the *Sabhāparva* has been further increased by the mention of Madhyamikā. Weighing all the evidences it may be said that the events referred to in the *Sabhāparva* range between 184 B.C. to 148 B.C., *i.e.*, the period of Puṣyamitra Śuṅga.

c5.102

That was the period of the Brāhmanical revival. It is perhaps in this period that the Pāṇḍus the heroes of the Mahābhārata first appear in a *vārtika* or supplementary rule to Pāṇini (IV, I, 44) attributed to Kātyāyana (180 B.C.). It is also in the second century B.C. that we find unmistakable allusions to what we may call an epic poem in the account of the *Mahābhāṣya*, which alludes to the poetic treatment of the epic and speaks of the epic characters.⁶⁵ It seems that this was the formative period of the epic, though nobody need have any doubt that the story existed in some form or the other long before it as mentioned in *Āśvalāyana Gṛhyasūtra* (III, 3, 1) which mentions Bhārata.

c5.103

⁶⁵ *The Cambridge History of India*, Vol. I, p. 253.

Linked texts (6)

Mahābhārata — Vyāsa, Sanskrit
Aṣṭādhyāyī — Pāṇini, Sanskrit
Mānava Dharmaśāstra — Manu & Bhṛgu, Sanskrit
Naturalis historia — C. Plinius Secundus, Latin
Āśvalāyanagṛhyasūtra — Āśvalāyana, Sanskrit
Abhidhānacintāmaṇi — Hemacandra, Sanskrit