Pramāṇa · Search · About · API

The Greeks in Bactria and India c10.s2: Chapter IV. Demetrius and the Invasion of India

The Greeks in Bactria and India c10.s2

Language
English

Chapter IV. Demetrius and the Invasion of India

c10.s2.2

G REAT changes had taken place in India since Alexander's day. He had found a number of disconnected states and peoples in the North-West, and had had no relations with, even if he had heard of, the most powerful of the Indian kingdoms, that of Magadha on the Ganges. Soon after his death the Maurya Chandragupta had seized the crown of Magadha, and, perhaps by 312, had extended his rule to embrace all India north of the line of the Vindya mountains and the Nerbudda river. He was succeeded first by his son Bindusāra and then by his grandson Asoka, under whom the Mauryan empire was expanded to include a considerable part of peninsular India; but the southern conquests were only temporary and were apparently lost after Asoka died, and the empire was essentially a North Indian empire; the capital was Pātaliputra on the Ganges. The Seleucids and the Mauryas were always on friendly terms, and Greeks knew a good deal about the Mauryan empire as it had been under Chandragupta through the account of it given by Megasthenes, Seleucus' ambassador at his Court; probably they knew as much about it as they had known about the Persian empire in Xenophon's day, while Indians in turn knew a certain amount about the Greeks of the Seleucid East, whom they called Yavanas or Yonas (p. 417). It is however of some importance to the subsequent story to note that the Mauryan empire as most Greeks knew it was that of Chandragupta and not that of Asoka, that is, it was an empire of Northern India. Asoka made one other very great change in India. He became a convert to Buddhism, and through his encouragement and missionary efforts that religion attained a position in India such as it never held again, though Brahmanism remained strong; in particular, he successfully evangelised a good deal of the North-West. Had the Mauryan empire continued powerful it might perhaps have done something to create a sense of Indian nationality in the loose complex of subordinate states and peoples which went to form it, but after Asoka's death it began to suffer the common lot of Oriental empires and gradually to decline; little however is really known of his successors, and it is not even certain whether the whole empire remained in one hand, or whether the two extant lists of names mean that the dynasty had divided into two lines,¹ one ruling in Pāṭaliputra and one in the North-West, or merely reflect the fact that one list is Brahman and one Buddhist. Certainly the Sophagasenos whom Antiochus III met in the Paropamisadae (p. 101) was no local rajah but a Maurya,² a powerful ruler³ with whom he renewed the traditional friendship of the two houses. It was the ultimate break-down of the Mauryan empire which gave Demetrius his opportunity.

c10.s2.3

¹ CHI pp. 511-12; de la Vallée-Poussin pp. 163-8.

c10.s2.4

² Hemchandra Raychaudhuri, JASB 1920 pp. 305, 310. Polyb. XI, 34, 11 calls him τὸν βασιλέα τῶν Ἰνδῶν, which on Greek usage (p. 154) ought to mean a Maurya.

c10.s2.5

³ Cf. J. Allan in Camb. Shorter Hist. of India 1934 pp. 54, 63.

c10.s2.6

Demetrius, when he crossed the Hindu Kush, was the third foreign conqueror whom north-west India had seen in historical times, not counting the unrecorded tribes, proto-Bactrian and other, who prior to the Achaemenid period had made their way over the passes and settled in the country. Darius I had conquered Gandhāra, Sind, and part of the Punjab; whether he had any plan beyond the enlargement of his empire is not known, but there seems to have been a good deal of Iranian blood in the North-West, which may have had some bearing on his actions. These Indian provinces were finally lost in the reign of Artaxerxes II; Artaxerxes III (Ochus) was very hazy about the geography of the Indus,⁴ and Alexander met no Persian officials east of the Hindu Kush. Alexander himself had a double plan: to conquer what Darius I had held, which he achieved, and to reach the Eastern Ocean which he thought quite close, a thing now known to have been impossible. His success was far more evanescent than that of Darius; a few years after his death the only traces left of his rule, not counting the Paropamisadae, were two or three of the cities he had founded, islands now in an Indian sea. Demetrius’ invasion was a different matter. It followed a plan which neither Darius nor Alexander had known enough about India even to dream of, and employed methods which Alexander had indeed dreamt of but had only begun, very tentatively indeed, to practise when he died, and which might have provided possibilities of permanence in advance of previous attempts; in distances traversed, in territory acquired, the Bactrian Greeks far surpassed both the Persian and the Macedonian, and came near to success in an undertaking hardly less ambitious and far-reaching than had been Alexander’s conquest of the Persian empire. What this plan was will have to be elucidated by events. But it was a plan which could only have originated in some definite man’s brain, and that man was quite certainly Demetrius.

c10.s2.7

⁴ Aristot. Liber de inundacione Nili, Rose³ fr. 248.

c10.s2.8

One thing however must be noticed here which will be elaborated later. The Greek ‘conquest’ of India was hardly a conquest in the ordinary sense of the word, the sense in which Alexander conquered Persia. But in the earlier part of this chapter I shall for convenience use the conventional language of conquest, and shall consider in the latter part what it was and what it meant.

c10.s2.9

That Demetrius was quite consciously (up to a point) copying Alexander—that he regarded Alexander not merely as his supposed ancestor (App. 3) but as his model—comes out clearly from his coins, and is of the first importance for the story. On his own coins¹ he wears the elephant-scalp. As elephants live in India, it was inevitable that the elephant-scalp should have been taken to refer to his Indian conquests; but it is certain that it does not, for not only does it appear on his Bactrian coinage from the beginning of his reign, but it had been used as a head-dress for Alexander on early coins of both Ptolemy I² and Seleucus I;³ yet Ptolemy I had no connection of any kind with India—he neither ruled it nor aspired to rule it—and Seleucus had ceded all his Indian possessions to Chandragupta. The tradition behind this portrayal of Alexander is unknown, but the elephant-scalp itself must be a symbol of power—power far extended, as his had been; for both Ptolemy and Seleucus had every object in representing themselves as successors of the man who had reached the summit of human greatness. The representation of Demetrius in the elephant-scalp then means that he had himself portrayed in the guise of Alexander;⁴ and in fact, apart from the general resemblance of his portrait (features excepted) to that of Alexander on Ptolemy’s coins, the elephant-scalp on the two is identically treated,⁵ as opposed to its later treatment in art. There will be more to say about the elephant-scalp later (pp. 189, 206); but meanwhile this suffices. Again, Demetrius, presumably after crossing the Indus, took the title ἀνίκητος, ‘the Invincible’; it has already been mentioned that he is so called on the bilingual Indian tetradrachm of Demetrius II and on the Demetrius coin of Agathocles’ pedigree series, and the same title occurs on those rare bilingual copper coins of ‘King Demetrius the Invincible’ which have been supposed to be his copper coinage for India.¹ No king anywhere before him had assumed this title. It is a poetical word, known in Hesiod and the tragedians, but it is occasionally used in prose and was so used in a famous story: when Alexander visited the oracle of Delphi, the Pythia hailed him ἀνίκητος,² and this story must be the origin of Demetrius’ title. He wore then the symbol of Alexander’s power and used the title conferred upon him by Apollo; he was to be a second Alexander.

c10.s2.10

¹ *BMC* pp. 6, 163, Pl. II, 9–12; see Plate no. 3.

c10.s2.11

² *BMC Ptolemies*, pp. 1–3, Pl. I nos. 1, 2, 5, 6, 8.

c10.s2.12

³ On some anonymous double staters (Head² 756) and on a rare copper coin in the collection of E. T. Newell, figured by M. Rostovtzeff, *Seleucid Babylonia* Pl. VI, 3.

c10.s2.13

⁴ Cf. Rostovtzeff *op. cit.* p. 53.

c10.s2.14

⁵ Best seen in *BMC Ptolemies* Pl. I no. 1.

c10.s2.15

¹ *BMC* p. 163 no. 3, Pl. XXX, 3.

c10.s2.16

² Plut. *Alex.* 14, ἀνίκητος εἶ, ὦ παῖ. See Diod. XVII, 93, 4; *Anth. Pal.* VII, 239.

c10.s2.17

Before considering the course of the invasion, one must fix the chronology, as near as may be. It has been seen that Demetrius’ conquest of the Seleucid provinces in eastern Iran, which naturally antedates the invasion of India, could not have been begun till after the battle of Magnesia, 187 being the most probable year; how long it took cannot be said, but Demetrius cannot have crossed the Hindu Kush till very distinctly later than 187. The other terminal point is given by the account in the Yuga-purāṇa³ of the Gārgī Samhitā, which says that, after the occupation of Pātaliputra, the Greeks would not stay in the Middle Country (say roughly the district between Mathurā and Pātaliputra) because of a terrible civil war which would break out among themselves; the reference is of course to the invasion of Eucratides (Chap. v), because there is no other civil war to which the words ‘an awful and supremely lamentable strife’ can refer. It was therefore Eucratides’ invasion which caused the abandonment of Pātaliputra. I must anticipate here what will be proved in the next chapter, that Eucratides’ dates are certain within very narrow limits; he set out most probably in 169, though early in 168 may be possible, and had conquered everything west of the Hindu Kush by the end of 167; the most probable date for the evacuation of the Middle Country is therefore some time in 168, with a possible year’s margin either way. The statement that the Greeks will not stay in the Middle Country means of course that they will not stay long; and reasons will be given later (p. 156) for supposing that a date of c. 175 for the occupation of Pāṭaliputra cannot be far wrong.

c10.s2.18

³ Translations and discussion of the material sections of this work are given in App. 4. Being embedded in an astrological work, it is given in the form of a prophecy; but the Yavana sections appear to reproduce an older document of the nature of a chronicle.

c10.s2.19

That gives the end of the conquest; it remains to date the beginning a little more closely. The date given in the Purāṇas for the end of the Maurya dynasty, 184, has been generally accepted by historians of India;¹ that is the year in which Pushyamitra the Sunga, hereditary ruler of Vidiśā (East Malva) and general of the last Maurya king, assassinated his master and seized the vacant throne. Whether the Maurya dynasty had split into two lines and, if so, what were their relationships is too obscure a matter to warrant any deductions; we can only take 184, the year of Pushyamitra’s accession, as signifying the end of the Mauryan empire.² In the tradition (p. 177) Pushyamitra proceeded to make his power felt, first near the capital, and then at Sāgala (Sialkot) in the eastern Punjab, subsequently Menander’s capital, which must imply some intermediate steps; the Greeks then did not take Sāgala for an unknown period after 184, say two or three years at the least. On the other hand, there are reasons, which need not be anticipated here, for connecting Demetrius’ enterprise with the end of the Mauryan empire and the accession of a usurper in 184, a thing which fits very well with the dates already obtained for Demetrius; if then it be supposed that he crossed the Hindu Kush about 183 or 182, that date cannot be very far out. In any case, the whole of the events to be recorded down to the death of Demetrius (Chap. v) must lie between 184 and 167 as their terminal points.

c10.s2.20

¹ There seems to be a variant, 185.

c10.s2.21

² For later descendants of the Mauryas see *CHI* p. 513.

c10.s2.22

The story has been rendered meaningless by the custom of dating Menander either in the second half of the second century or, even worse, about 125–95 B.C.³ One of the many merits of the late E. J. Rapson’s work in the Cambridge History of India was to place Menander in his correct period;¹ this has not been followed by subsequent writers,² but is so obviously right that it is needless to argue it afresh;³ everything that follows will bear it out. Menander’s chronology, like that of the Victory of Samothrace, has been an instructive instance of the danger of dating historical events by considerations drawn from artistic style. Because his coins are much inferior in style to those of Demetrius and his successors in Bactria, who could be approximately dated from Polybius, it was concluded that he must be late, so as to give time for the art to become ‘debased’; whereas in reality it means that the artists at his disposal in the rather remote eastern Punjab were inferior in skill to those who worked in Bactria. It is as though some historian in the distant future should place the reign of George V in the Aurignacian period on the strength of some of Epstein’s sculpture.

c10.s2.23

³ Von Gutschmid’s date (*Gesch. Irans* p. 104), though he himself called it an unsafe calculation. He was going on the list in the *Vayū-purāṇa* which gives eight Greek kings of India—Demetrius, Eucratides, Apollodotus, Strato I, Strato II, Zoilus, Menander, Dionysius—and puts Menander two generations after his great-grandson. The sooner this worthless list is allowed to die the better. Even later dates for Menander have been suggested; see Winternitz, Eng. Tr. II p. 174 n. 2. I need not consider them.

c10.s2.24

¹ *CHI* pp. 543 *sqq.*: contemporary with Demetrius.

c10.s2.25

² The editor of V. A. Smith⁴ pp. 229, 239 (his invasion 156-3); Grousset p. 39 (155 B.C.); Przyluski, *Açoka* p. 166 (150 B.C.). No evidence exists for such a dating; see p. 146 on Patañjali.

c10.s2.26

³ Apollodorus makes him contemporary with Demetrius, Trogus with Apollodotus, and some coin indications (*CHI* p. 551) with Eucratides.

c10.s2.27

Demetrius had to make arrangements for the government of Bactria during his absence, and he left his eldest son Euthydemus II as king in Bactria-Sogdiana; Euthydemus II put his own name and portrait on his silver coins,⁴ but on his nickel and bronze issues he used a Seleucid type, the head of Apollo and a tripod-lebes.⁵ The common Hellenistic practice when a king was absent had been to leave a son merely as governor (p. 218 n. 1), but a parallel to the kingship of Euthydemus II can be found later among the Ptolemies: when Ptolemy VI Philometor invaded Syria and expected to be in that country for some time he left his son Ptolemy Neos Philopator as king in Egypt.⁶ As Bactria was Demetrius’ home kingdom, it is probable that Euthydemus II was not a sub-king—that would hardly have suited the circumstances—but a full joint-king with his father on the Seleucid model; this would agree with what happened later (p. 221). The western provinces of the empire in Iran, as already noticed, were under the rule of Antimachus in the north and Apollodotus in the south.

c10.s2.28

⁴ *BMC* p. 8, Pl. III, 3, 4; *CHI* pp. 447-8; J. N. Svoronos, *J.I.d’A.N.* xv, 1913, p. 186.

c10.s2.29

⁵ *BMC* p. 8, Pl. III, 5, 6.

c10.s2.30

⁶ Otto, *Zeit d. 6. Ptolemäers* p. 128 n. 4.

c10.s2.31

Demetrius took to India with him his second son Demetrius (II), and also his general Menander, of whom much will be heard later. It is just possible that the Paropamisadae were his already (p. 101); anyhow he took Gandhāra, crossed the Indus, and occupied Taxila, which had been Alexander’s advanced base and must have been his also. It may be taken as certain that he occupied Taxila himself, because the line of conquest there bifurcated, and had he left Taxila to be occupied by Menander, it and not Sāgala must have become Menander’s capital.

c10.s2.32

Gandhāra,¹ the country between the Kunar river and the Indus, comprising the modern Bajaur, Swat, Buner, the Yusufzai country, and the country south of the Kabul river about Peshawur, was to be one of the strongholds of Greek power; it has been called a kind of new Hellas.² Asoka had converted much of the country, and it became to Buddhists a second Holy Land, where rose three of the four great stupas³ which recorded Buddha’s charity with his own body in earlier incarnations, those of the Body-gift at Manikyala, the Flesh-gift at (probably) Girarai in the hills between Peshawur and Buner, and the Eye-gift; this last may have towered aloft on the acropolis of what was to be the Greek capital, Pushkalāvatī (Charsadda), rendering, as has been said, ‘still more striking its resemblance to its more famous Athenian counterpart’.⁴ But Pushkalāvatī, like Taxila, was only partially Buddhist; Siva was still powerful enough there for his humped bull to become the coin-type of the Greek mint,⁵ while the Greeks were to worship Artemis as their city goddess.⁶ But she was not the Greek Artemis; she was Anaïtis (Anahita) of Bactra, for Anaïtis and her crown of rays appear as Artemis radiate on coins of the Saca king Maues⁷ which are shown by the humped bull on them to have been struck at Pushkalāvatī, just as in Bactra itself she had appeared as Artemis radiate on a coin of Demetrius (p. 115). Unfortunately it is not known whether the Greek invaders brought her with them from Bactra, which would throw light on Greek relations with Asiatic deities, or whether she had arrived long before with one of the earlier streams of invaders whom Indians comprehensively called Bāhlīkas, *i.e.* Bactrians (p. 169); the latter possibility would, if correct, imply an Iranian element at Pushkalāvatī, again as at Taxila. But, unlike Taxila, Pushkalāvatī became a Greek *polis* (doubtless somewhat of the type of Susa, p. 27), as is shown by the Fortune of the city on kings' coins;¹ the solitary coin of the city itself² which exists to prove that it was once for a time completely independent (p. 336) shows, beside Siva's bull, the Fortune of the ‘city of lotuses’ with her mural crown, holding in her hand the lotus of Lakshmī. Evidently Pushkalāvatī, when a Greek *polis*, was no less proud of her alien deities than was Ephesus of her alien Artemis, and Siva's bull is a parallel to Artemis' bee on the coins of the Ionian city. Pushkalāvatī stood at what was probably then the junction of the Swat and Kabul rivers,³ and as it and not Purushapura (Peshawur) became the Greek capital, the regular Greek line of communication westward probably did not run through the Khyber pass but by the route which Alexander had followed more to the northward; it seems unlikely that the Khyber was in *regular* use till the Kushans made Peshawur their capital.⁴

c10.s2.33

¹ In the *Jātakas* Gandhāra includes Taxila; but in this book I use the term in its strict sense. On Gandhāra see Foucher, *Gandhāra*; R. Grousset, *Sur les traces de Bouddha*, 1929, chap. vi.

c10.s2.34

² Grousset *ib.* p. 96.

c10.s2.35

³ A stupa was a Buddhist shrine, circular and domed, usually but not always enclosing a relic. Buildings, even of stupa form, which did not enclose a relic were usually called chaityas: de la Vallée-Poussin p. 149. See on stupas, archaeology apart, Foucher I chap. I, and the long study by P. Mus, ‘Barabudur’, *BEFEO* XXXII, 1932, pp. 269–439, XXXIII pp. 577–980, XXXIV pp. 175–400.

c10.s2.36

⁴ Foucher, *Gandhāra* p. 15.

c10.s2.37

⁵ This certain fact (*CHI* p. 557) is confirmed by Siva being known to Greeks as the god of Gandhāra: Hesychius, Γάνδαρος · ὁ ταυροκράτης παρ’ Ἰνδοῖς.

c10.s2.38

⁶ Copper coins of Peucolaos, obv. Artemis, rev. the Fortune of some city with mural crown; *Lahore Cat.* p. 324 no. 20, see *CHI* p. 558. What identifies the city of these coins with Pushkalāvatī is Maues’ coins (next note).

c10.s2.39

⁷ *BMC* Pl. XVI, 4; *Lahore Cat.* Pl. X, 10; *ASI* 1929–30 p. 89 nos. 24, 25.

c10.s2.40

¹ Those of Peucolaos (above). On the Greek name of Pushkalāvatī see p. 237 n. 5. Many of the city Fortunes on the Saca coinage (p. 353 n. 1) are probably Pushkalāvatī.

c10.s2.41

² *CHI* p. 587 and Pl. VI, 10.

c10.s2.42

³ Foucher, *Gandhāra* p. 11.

c10.s2.43

⁴ See Foucher, *BSOS* VI, 1930-2, pp. 344-5, and plan p. 343. A correspondent, however, sent me a sketch of the masonry of some old block-houses above and commanding the Khyber pass, which he suggested was Greek. It looks to me more like the Kushan masonry at Taxila; but it is a matter which requires investigation on the spot by an archaeologist.

c10.s2.44

With Gandhāra in his hands, Demetrius would be well informed of Buddhist feeling, a matter which was to be of great importance; but from the military and political point of view the acquisition of Taxila was of more moment. The great city was even more important than it had been in Alexander's day, for it had long been the seat of the Mauryan governor of the North-West; though near it stood the fourth great stupa, that of the Head-gift, it was only partially Buddhist⁵

c10.s2.45

⁵ There are Buddhist inscriptions of the Saca period, *CII* nos. II, XIII, XXVII, XXXI, XXXII.

c10.s2.46

Vishnu was strong there (p. 406); with its famous University, of which the buildings have been excavated,¹ sought by students from many quarters, its merchant guilds who struck their own city coinage,² the Iranian element in its population with their Towers of Silence,³ its balance of religions, its feeling of independence which had led it to withstand Porus and to revolt against the Maurya, it seemed destined to be the capital of the foreign invaders. So Demetrius thought. The city he found is now represented by the latest stratum of ruins on the Bhir mound; he presently built a new city on Sirkap, now buried beneath the remains of the later Parthian city (p. 179). To it he transferred the population of Old Taxila, as Hellenistic kings in the West would transfer the population of some Greek town to one of their new foundations, and the city on Bhir came to an end;⁴ the Taxila henceforth mentioned throughout this book is the city on Sirkap, which will be described later.

c10.s2.47

¹ Sir J. Marshall, *A guide to Taxila* 1918 p. 72.

c10.s2.48

² E. J. Rapson, *Indian coins* 1897 p. 14; C. J. Brown, *The Coins of India* 1922 pp. 15–19; J. Allan, *BMC India* pp. cxxv, cxxviii, and see *post* p. 161 n. 1.

c10.s2.49

³ Aristobulus saw there corpses exposed to vultures, which he saw nowhere else in India, Strabo xv, 714. The Aramaic inscription found there (L. D. Barnett, *JRAS* 1915 p. 340; A. Cowley, *ib.* p. 342), though much earlier, may support this; and a tutelary Yaksha in the region of Taxila had an Iranian name, S. Lévi, *JA* 1915 p. 75.

c10.s2.50

⁴ Sir J. Marshall thinks that Bhir came to an end with the Greek conquest, the two latest strata being Mauryan, *ASI* 1930–4 p. 149. Sirkap was therefore certainly Demetrius’ foundation, even if he did not finish it, for there is nothing beneath the two Hellenistic strata.

c10.s2.51

Demetrius left his son Demetrius II as his sub-king to govern the Paropamisadae and presumably Gandhāra also, that is, all the country between the Hindu Kush and the Indus; his also must have been the task of securing and perhaps improving the communications with Bactria. That he had the royal title is shown by his putting his own portrait on the bilingual tetradrachm already referred to (p. 77). It is certain enough (p. 158) that his seat was Alexandria-Kapisa, the capital of the Paropamisadae (App. 6), from which Gandhāra also could be governed, as it was from Kapisa in Hsüan Tsiang’s day. Many reasons contributed to the importance of the capital beside its wonderfully fertile plain, which has led to it being called a little Kashmir without the lake.⁵ It was near the silver mines of the Panjshir valley and was thus well suited to be the principal mint of the province; Kapisa was the outlet for Kafiristan, the land of the Kambojas, who were possibly a valuable support to the Greeks (p. 170)—indeed it has been thought probable that Kapisa and Kamboja are the same word;¹ and the dual city was nearer to Bactra than any other important city and commanded the three routes. The reverse type on the bilingual tetradrachm of Demetrius II is Zeus holding a thunder-bolt. Zeus, one of the three deities of the Alexander coinage, had not before been used by any Bactrian king, and it is almost certain, from the types on the silver coinages of Pantaleon and Agathocles (p. 158), that the Zeus of this tetradrachm is meant for the elephant-god of Kapisa. A few years later the god of Kapisa began to be regularly represented as Zeus enthroned;² but the reason for representing him as *enthroned* (p. 213) was due to other circumstances which had not yet arisen, while the reason for representing him as Zeus was a compelling one. For the elephant-god had his abode on the mountain Pilusāra;³ and to Greeks a god who lived on a mountain-top could not well become anything but Zeus.⁴

c10.s2.52

⁵ Foucher, *Afghanistan* p. 266 (of Kapisa). But as he thought Kapisa is represented by Begram, I fancy that the plain he describes must be that of Alexandria.

c10.s2.53

¹ S. Lévi, *JA* 1923 ii p. 52.

c10.s2.54

² Eucratides’ coin, see p. 212.

c10.s2.55

³ *CHI* p. 556 (from Hsüan Tsiang). The mountain appears on Eucratides’ coin.

c10.s2.56

⁴ Zeus Kasios is perhaps the best-known instance, but there are many; see A. B. Cook, *Zeus* II, App. B, ‘The mountain cults of Zeus’.

c10.s2.57

Demetrius II then ruled and coined in Alexandria-Kapisa. But he coined for his father, not for himself, as Antiochus I had once done:⁵ this is shown by his putting his father’s title ‘Of King Demetrius the Invincible’ on his tetradrachms, while on his square bilingual copper coins (which were struck by him and not by his father)⁶ he put not only his father’s title but his father’s head, the well-known head wearing the elephant-scalp. The tetradrachms would circulate principally among Greeks, who understood the position; hence his own head. But the copper coins would circulate, or so it was hoped, among Indians, who might not understand; hence his father’s head. But the real matter was the introduction on the coinage of a Prakrit legend, written in Kharoshthi, beside the Greek legend. The great importance of this step will be considered later; here I need only say that this radical development in policy could only have been due to Demetrius himself, not to any subking, and proves yet again that Demetrius II was coining to his father’s instructions. Demetrius himself struck no coins in India; his coins nearly all come from Iran, and are practically never found east of the Indus,¹ though one has come from the excavations at Taxila.²

c10.s2.58

⁵ Coins with Antiochus’ legend and Seleucus’ head: E. H. Bunbury, *NC* 1883 pp. 67-71; Head² p. 758.

c10.s2.59

⁶ Because the reverse type (*BMC* p. 163 no. 3) is the winged thunder-bolt, which appears in Zeus’ hand on the tetradrachm and symbolises Zeus, and is therefore the type of Demetrius II, not of his father.

c10.s2.60

¹ Cunningham, *NC* 1869 p. 141; *BMC* p. xxv; Whitehead, *NNM* p. 15.

c10.s2.61

² One from Bhir, *ASI* 1920–1 Part I p. 21. Also one of Euthydemus from Sirkap, *ib.* 1927–8 p. 60. In the Pearse collection in the Indian Museum, Calcutta, is a silver coin ascribed to Demetrius, *ASI* 1928–9 p. 139 no. 4, Pl. LVI, no. 4: obv. youngish head of king, diademed and uncovered; rev. Apollo on omphalos, with legend βασιλέως Δημιτρί[ου] [σώ]τηρος. It is obviously the Seleucid Demetrius I.

c10.s2.62

One word as to Demetrius’ communications. The Hindu Kush, which has never prevented anyone from invading India who had a mind to, is said to be a less formidable barrier than it seems; and it has been pointed out that the whole story of the Greeks in India presupposes fairly easy communication between Taxila and Bactra.³ There were three routes across the Hindu Kush into Bactra,⁴ all of them commanded by Alexandria-Kapisa at the junction of the Panjshir and Ghorband rivers. The central route, over one of the lofty Kaoshan group of passes, does not come in question; it rises too high, though local tradition believes that Alexander used it for one of his crossings. The north-eastern route commanded by Kapisa, up the Panjshir and across the longer but lower Khawak pass, had been used by Alexander on his other crossing; but though it may have occasionally been used by the Greeks, it led primarily to Badakshan, and made the road to Bactra itself very long. The south-western route commanded by Alexandria, generally used to-day, furnished the most direct road between the capital of the Paropamisadae and Bactra; it runs up the Ghorband by Bamyan and across the Kara Kotal pass to the Darrah, the river of Bactra, thus turning the Hindu Kush rather than crossing it; the road crosses three passes, but all are much lower than the Khawak.⁵ This was the regular route in Hsüan Tsiang’s day, though the pilgrim himself, perhaps for variety, went home by the Khawak; and the great Buddhist sculptures found at Bamyan attest the importance of the place subsequently to the Greek period.

c10.s2.63

³ A. Foucher, *CR Ac. Inscr.* 1927 p. 117.

c10.s2.64

⁴ Cunningham, *Geog.* p. 28. On the routes of the Paropamisadae see also E. Trinkler, *Afghanistan*, Petermanns Mitt. Supp. Bd. 196, 1928, pp. 57 sqq.

c10.s2.65

⁵ On the Bamyan route see Foucher, *Afghanistan* p. 257 and his Map 3 facing p. 278.

c10.s2.66

The French archaeological mission had no doubt that the Bamyan route was the usual Greek route,¹ though apparently no archaeological remains, such as foundations of block-houses, were found. I do not know their reasons, but there are two pieces of evidence which are very strong: one is the passage from Varro to be presently cited, and the other the fact that Pliny, speaking as though approaching from the north, names first the Bamyan-Ghorband eparchy and then Opiane (p. 97 n. 1), that is, he speaks from the point of view of someone approaching Alexandria by the Bamyan route. If this were the usual route, it would explain why Alexander founded Alexandria on the opposite bank of the river to Kapisa instead of utilising the latter city, and would explain two other things also: the small importance in Greek times of Kabul, cut off from this road by the Koh-i-baba range between Kabul and the Ghorband valley, and the tradition of the hardships endured by Alexander's army in crossing the Khawak; for though Persian armies had invaded India before him, if they had used the Bamyan route his crossing of the Khawak in force may have been pioneering work. It may be supposed that the Greek kings did all they could—road improvement, shelters, depôts of provisions suggest themselves—to make the route between Alexandria and Bactra as easy as might be, and there is one curious bit of evidence that they succeeded. When Pompey called for a report on the feasibility of making a trade route from India via Bactria and the Caspian to the Cyrus river, Varro says that the report stated that goods could be brought in seven days from India (presumably Alexandria-Kapisa) to the river of Bactra;² one need not insist on the seven days, but it shows that the transit was considered tolerably easy, and also that the regular route was that by Bamyan to the Darrah river.

c10.s2.67

¹ Foucher, *Afghanistan* pp. 280 sqq.; cf. *CR Ac. Inscr.* 1927 p. 117.

c10.s2.68

² Pliny VI, 52; see App. 14.

c10.s2.69

Once in possession of Taxila, Demetrius had two possible lines of advance, on either side of the Indian desert: one south-eastward along the great road across the Punjab and by the Delhi passage to the Ganges and the Mauryan capital Pātaliputra, the other southward (at first southwestward) down the Indus to its mouth and whatever might lie beyond. Alexander had attempted the two lines successively; Demetrius took them concurrently. His own sons, as their portraits show, were as yet too young to lead a great advance; but he was fortunate in commanding the services of two lieutenants who must have been very able men, his brother or kinsman Apollodotus and his general Menander. The two are twice coupled in the classical tradition,¹ which indicates some close connection between them,² but in each case Apollodotus is named first, which suggests that he was the more important; doubtless the reason is that he was connected with the royal house, while Menander was not. Who Apollodotus was has already been considered (p. 76); for Menander I must refer to the Excursus (pp. 420 sq.). He was a Greek from the Paropamisadae, and certainly a commoner; his birth in a village might mean that his father was a great landowner, successor to one of those Iranian barons who figure so largely in the Alexander story,³ and that he was born in his father's stronghold to which the village was ancillary, but it more probably means that he was merely the son of a cleruch in a military settlement; in either case he had risen by his own abilities. As he died between 150 and 145 (p. 226), and as his latest coins (so far as portraiture can be relied on in Graeco-Indian coins) show a man of advanced years,⁴ he must have been nearer to the generation of Demetrius than to that of his sons; and he had certainly seen fighting, for Demetrius would not have put an unproved man in command of the advance to Pāṭaliputra. His portraits,⁵ for what they are worth, confirm the fact that he was not a Euthydemid; he has a different type of face, and the Euthydemid bull-neck is conspicuously absent. An Indian writer remarked later that among the Yavanas slaves could rise to be kings;⁶ doubtless he used 'slaves' in the Persian sense of everyone not royal (p. 355 n. 6), and was thinking primarily of the career of the most famous of the Yavana monarchs. But it must be emphasised that at the time of the invasion Menander was only Demetrius' general, a fact, it would seem, better understood by Indian writers of the period (p. 166) than by modern scholars.

c10.s2.70

¹ Trogus Prol. XLI; Periplus 47.

c10.s2.71

² As E. J. Rapson has noticed: *Ancient India* 1914 p. 128; *CHI* p. 547.

c10.s2.72

³ See p. 32. It is not however known that this ever happened in the Farther East.

c10.s2.73

⁴ *Lahore Cat.* p. 59 n. 1.

c10.s2.74

⁵ Whitehead *NNM* Pl. VII, 1 is about the best.

c10.s2.75

⁶ Lévi, *Quid de Graecis* p. 23.

c10.s2.76

Our two primary Greek sources, taken together, ascribe the conquest of Northern India to three men, Demetrius, Apollodotus, and Menander. At first sight indeed it looks as if Apollodorus ascribed the conquest to Demetrius and Menander,⁷ 'Trogus' source' to Apollodotus and Menander;⁸ but these brief notices in a fragment of Apollodorus and in Trogus’ prologue to a chapter which Justin did not excerpt are inclusive, not exclusive; they mean, not that Apollodorus excluded Apollodotus or ‘Trogus’ source’ Demetrius, but that these were the three men who between them carried the conquest through. As however the secondary sources, Strabo and Trogus, while one selected Demetrius for mention and the other Apollodotus, both name Menander, and as there were certainly two lines of advance, we are justified in taking it to mean that one line of advance was Menander’s and that the other was shared by Demetrius and Apollodotus; it will appear that the evidence agrees with this.

c10.s2.77

⁷ Strabo XI, 516.

c10.s2.78

⁸ Trogus Prol. XLI: Indicae quoque res additae, gestae per Apollodotum et Menandrum reges eorum.

c10.s2.79

Demetrius himself was responsible for the conquest of Sind. A scholion to the grammarian Patañjali¹ (p. 146) mentions a town Dattāmitrī among the Sauvīras and says that it was founded by Dattāmitra, who is named in the Mahābhārata as king of the Yavanas and Sauvīras and is undoubtedly Demetrius; and the existence of this Demetrias in Sind is confirmed by an inscription.² It was certainly not the Arachosian Demetrias between Seistan and Ghazni (see App. 9), for the Sauvīra-Sindhus had nothing to do with Arachosia; at this time they were on the lower Indus and occupied its Delta (p. 171). This Demetrias is not likely to have been a completely new city. Alexander had begun to build great docks at Patala and must have left a colony there; what Demetrius found there is unknown, but Patala was the natural port and centre, and undoubtedly Demetrias was Patala re-founded and renamed;³ Demetrius may have had in mind the creation of a port on the Indus which should correspond to that of Seleuceia on the Tigris. Demetrius then followed Alexander’s track down the Indus to the sea. Alexander had gone by water; Darius I before him had sent a fleet down the Indus; Demetrius too must have followed the easy and natural course of going by water, which would mean that on reaching the sea there was a fleet at his disposal. The trident on one of his coins⁴ must imply naval power, but it is a Bactrian coin and not likely to refer to a fleet on the Indus; probably it is connected with the symbolism of Antimachus’ coins (p. 90) and refers to the squadron which every Bactrian king must have maintained on the Oxus as part of the country’s system of defence.

c10.s2.80

¹ Given by A. Weber, *Indische Studien* v p. 150 n.

c10.s2.81

² No. 19 of the Nasik cave inscriptions (p. 257 n. 3). Weber *ib.* pointed out that the term Dattāmitriya used in another scholion to Patañjali for an inhabitant of Dattāmitrī is only the Sanscrit form of the Prakrit Damtāmitiyaka of the inscription, for which he suggested Datāmitiyaka; since then E. Senart has in fact read the word in the inscription as Datāmitiyaka, *Ep. Ind.* VIII, 1905-6, p. 90 no. 18. See also on the identification of the towns of the scholion and the inscription N. R. Ray, *IHQ* IV, 1928, p. 743. It seems free from doubt.

c10.s2.82

³ This will be confirmed by the section on the pepper trade in chap. IX (pp. 370 *sqq.*).

c10.s2.83

⁴ *BMC* p. 7 no. 14.

c10.s2.84

Demetrius himself can have gone no farther. Like the Antigonids, the Euthydemids were tied to their northern frontier; as Macedonia was the shield of Greece against the barbarism of the Balkans, so Bactria was the shield of Iran against the nomads who, as Euthydemus had told Antiochus III, were perpetually threatening her; and who were one day to overwhelm her; no Bactrian king, for his own safety, dare neglect this responsibility. Demetrius, even though he had left a young son to guard Bactria, had taken some risk in going himself to Sind. He had done what Alexander had done; he must now have handed over the command of the advance southward to Apollodotus and returned to Taxila. Apollodotus, coming from the Arachosian Demetrias, may have joined him on his way down the Indus, or may have been annexing eastern Gedrosia, which was seemingly governed from Sind (p. 94).

c10.s2.85

Menander’s advance to the south-east is attested both from the Greek and the Indian side. Some writers indeed, with no clear idea of the two lines of advance, have ascribed all the Indian conquests to Menander, a thing which time, space, and Trogus’ mention of Apollodotus alike forbid. It is a proof of Cunningham’s penetration that he saw something of the truth as long ago as 1870,¹ when he said that the campaigns of Apollodotus and Menander were contemporary but distinct, that of Apollodotus being directed from Sind against Rajputana; but nothing came of his illuminating suggestion, because he put both kings much too late and numismatists subsequently saw that Apollodotus, who still coined on the Attic standard and some of whose coins were overstruck by Eucratides, must be a very early king. The first thing is to consider exactly what Apollodorus does say, before coming to the Indian account.

c10.s2.86

¹ NC 1870 p. 85.

c10.s2.87

He says in one passage that the Greeks conquered more of India than the Macedonians (Alexander) had done,² and in another that they became (imperfect tense; that is, they were for a time) masters of ‘the Indians’; they overthrew more peoples than Alexander had done (i.e.

c10.s2.88

² Strabo xv, 686: πλείω τῆς Ἰνδικῆς ἐκείνους (the Greeks) ἢ Μακεδόνας καταστρέψασθαι λέγων.

c10.s2.89

Alexander in India) and most of all Menander, some himself and some Demetrius.¹ As the words ‘they overthrew’ (aorist; that is, one point of time) apply to *both* men, we get two facts: that Demetrius and Menander were acting in concert, and that Menander went farther than Demetrius. Strabo adds to this excerpt a note of his own, showing that (like some moderns) he found it hard to believe: ‘at least if Menander really crossed the Hypanis (Beas) toward the east and went as far as the Isamos’,² which implies that Apollodorus had said he did, and incidentally implies that Demetrius did not go so far and did not cross the Hypanis. Most of the Alexander-historians call the Beas, where Alexander turned back, the Hyphasis; but one of them, Aristobulus, preferred the form Hypanis,³ and that is the form always used by Strabo. For the unknown name Isamos the most usual conjectures are the Iomanes (Jumna) or the Soamos (Sōn); if there really be a Prakrit name Issumai for the Jumna⁴ it settles the matter, but it is not very material. For there is one more passage of Apollodorus, or rather of Strabo paraphrasing Apollodorus in his own words, which has too often been overlooked:⁵ it says that those who came after Alexander advanced beyond the Hypanis to the Ganges and Pātaliputra.⁶ The language used imports a military expedition and imports also that Pāṭaliputra was taken; Strabo could not have put it in that form had Apollodorus said that they had tried to take the capital and failed.

c10.s2.90

¹ Strabo xi, 516: τῆς τε ᾿Αριανῆς ἐπεκράτουν καὶ τῶν ᾿Ινδῶν, ὡς φησιν ᾿Απολλόδωρος ὁ ᾿Αρτεμιτηνός, καὶ πλείω ἔθνη κατεστρέψαντο ἢ ᾿Αλέξανδρος, καὶ μάλιστα Μένανδρος, τὰ μὲν αὐτὸς τὰ δὲ Δημήτριος.

c10.s2.91

² Ib. εἴ γε καὶ τὸν ῞Υπανιν διέβη πρὸς ἕω (Menander) καὶ μέχρι τοῦ ᾿Ισάμου προῆλθε.

c10.s2.92

³ In xv, 686 Strabo contrasts Apollodorus with some unnamed writer who uses the form Hypanis and exaggerates city numbers in round thousands (Alexander had 5000 cities between Hydaspes and Hypanis). In 693, a named fragment of Aristobulus (= F. Gr. Hist. fr. 35, 19), a similar exaggeration of city numbers in round thousands occurs (the shifting of the Indus made over 1000 cities desert). Therefore the unnamed writer of 686, who uses the form Hypanis, is Aristobulus, though the passage is not given as his in F. Gr. Hist. It is morally certain that the Hypanis of 700 is from Aristobulus also. Strabo took the form from him.

c10.s2.93

⁴ K. H. Druva, JBORS xvi, 1930, p. 34 n. 25. But I cannot make out if Issumai be a real name or if the writer is only suggesting that it would be the Prakrit form of Isamos.

c10.s2.94

⁵ It is not given among the fragments of Apollodorus in FHG iv, p. 308, but there can be no question about it. Apollodorus is not yet given in F. Gr. Hist.

c10.s2.95

⁶ Strabo xv, 698: we know India within the Hypanis καὶ εἴ τινα προσειστόρησαν οἱ μετ’ ἐκεῖνον (Alexander) προελθόντες μέχρι τοῦ Γάγγου καὶ Παλιβόθρων. The word προελθόντες shows that a military expedition is meant and excludes the possibility of the reference being to Megasthenes, who anyhow could not be classified under Alexander’s successors (οἱ μετ’ ἐκεῖνον).

c10.s2.96

The advance of the Greeks to Pāṭaliputra is recorded from the Indian side in the Yuga-purāṇa (p. 132); translations of the material sections are given in full in Appendix 4, with such discussion as is necessary. It remains to take the outline (we cannot get more) of Menander’s advance and see the way in which the Greek and Indian sources agree and supplement each other, a conclusive proof that the story is true.

c10.s2.97

In the tradition (p. 177) Pushyamitra’s power reached anyhow to Sāgala (Sialkot between the Chenab and the Ravi); it is possible, as will be seen, that the halt at Taxila, while the ground won was being consolidated and Demetrius’ fleet was being built, was used to prepare Menander’s way with a little propaganda (p. 178). Menander first occupied Sāgala, known from the Milindapañha to have been his capital later (see Excursus), and then, as Apollodorus says, crossed the Beas, where Alexander had turned back. The Yuga-purāṇa then mentions the Yavanas at Mathurā (Muttra) on the Jumna; here comes in Apollodorus’ statement about the Isamos, if it be the Jumna. The Yuga-purāṇa then records the Yavanas at Sāketa (in Oude) and in the Pañchāla country (the Jumna-Ganges doab), which is followed by Apollodorus’ statement that the Greeks reached the Ganges. Finally both Apollodorus and the Yuga-purāṇa record the occupation of the capital. The latter document says that the Greeks first took Kusumadjava, which is Kusumapura, the old name of Pāṭaliputra, but which at this time must have been separate from, or a suburb of, the Mauryan town, and then took the Mauryan capital itself, which was defended by a mud wall, necessitating the use of their siege train, as Alexander had had to use his siege train against the high mud wall of Cyropolis; it is said that the excavations at Pāṭaliputra have brought to light a mud wall of the Mauryan period 14 feet thick and flanked with wooden palisades.¹ The Yuga-purāṇa subsequently treats the Greeks as masters of the country: they command, and the kings disappear.

c10.s2.98

¹ K. P. Jayaswal, JBORS XIV, 1928, p. 417.

c10.s2.99

One point in this account, the taking of Sāketa, is further confirmed from the Indian side by a statement of the grammarian Patañjali (made merely to illustrate the right tense to use for an event which has just happened), ‘The Yavana was besieging Sāketa.’¹ Patañjali’s date has generally been put about 150 B.C. on the strength of his supposed reference to Pushyamitra’s horse sacrifice as a contemporary event, and the dating so reached has been used to date Menander’s advance to about 150. There is nothing in this, for it is generally admitted that Patañjali’s grammatical examples are, or in any particular case may be, not necessarily his own composition but traditional examples, put together before his own time;² in fact a recent authority, I venture to think conclusively, puts him much later than 150.³ What Patañjali does show is that the Greek invasion produced such an impression that it could be used as a commonplace illustration in grammars.

c10.s2.100

¹ Cited in many works: see Lévi, *Quid de Graecis* p. 16; *CHI* p.

c10.s2.101

Before passing on, one or two points in connection with the Greek advance to Pāṭaliputra must be noticed. One need not waste time over the belief of some writers that the Greek kings were condottiere and their conquests raids, beyond hoping that such writers have clear ideas of what a ‘raid’ from Rawul Pindi upon Patna would mean; but the view held in defiance of Apollodorus, that it was Demetrius and not Menander who led the advance south-eastward, must be considered. It was first put forward as a guess in 1911 by Professor D. R. Bhandarkar,⁴ because he very properly saw that the advance must have taken place much earlier than the late date which he believed to be that of Menander; his reasoning was sound, but now that Menander’s true date is known it has no further application.⁵ Subsequently in 1923 Dr Sten Konow⁶ based a similar theory upon a passage in the Hāthigumphā inscription of Khāravela, which is supposed to state that Demetrius withdrew (from Pāṭaliputra) to Mathurā, and this has found some acceptance. It is fully discussed in Appendix 5; apart from some uncertainty as to what the inscription does say on the matter, it is abundantly clear that it has no bearing at all upon the Greek invasion or who led it, though it may be important for later events (p. 166). Let me assume that the passage means what Dr Konow and the late Dr Jayaswal say it means (though they do not altogether agree) and then put the matter in general terms. Some king, carrying on widespread operations over an enormous territory, is said to have abandoned a certain conquered province. This does not imply that he conquered the province in person or abandoned it in person or ever set eyes upon it; it may, and probably will, refer only to his orders to his generals or governors. *Qui facit per alium facit per se.*

c10.s2.102

⁵⁴⁴ Weber, *Ind. Studien* XIII p. 304, pointed out that the verb in the sentence, *arunad*, means ‘besiege’ and nothing else. ² Weber *op. cit.* XIII pp. 312, 315, 319; de la Vallée-Poussin p. 200. ³ De la Vallée-Poussin pp. 199–202, based on Patañjali’s mention of the Sacas. See on these Sacas the theories of Bhandarkar, *Indian Culture* I, 1934, p. 275, and Konow, *ib.* II, 1935, p. 189, with de la Vallée-Poussin’s reply, *ib.* II, 1936, p.

c10.s2.103

⁵⁸⁴ His argument is unanswerable, unless the mention of Sacas in Patañjali be a later interpolation, which no one has suggested. ⁴ *Ind. Ant.* XL, 1911, p. 11 n. 5. ⁵ The same thing applies to the adoption of this theory by H. Raychaudhuri, *The Political History of Ancient India* 1923 pp. 204 sqq., 209. I have been unable to see this book, and take the information from L. D. Barnett, *Calcutta Review* X, 1924, p. 250. ⁶ *Acta Orientalia* I, 1923, p. 27.

c10.s2.104

We may now leave Menander at Pāṭaliputra and return to Sind and Apollodotus. The evidence for what he did is not so clear as that for Menander, the Yuga-purāṇa being only concerned with Pāṭaliputra, but rather more is known about the ultimate consequences. Doubtless he first conquered the coastal provinces where a fleet could help him, and the notice of him in southern Rajputana comes later; we may suppose that he went round the Rann of Cutch, then possibly a sea-gulf, and so southward. Ptolemy (VII, 60) knows of a Greek city Theophila in those parts (p. 234), and his co-ordinates, for what they are worth, place it eastward of the Indus delta; he seems to have figured it as on the route from Patala to Ujjain,¹ and its foundation—it must have been an Indian town refounded—would belong to Apollodotus' activity. A woman's name means a dynastic name, which implies a Greek *polis*; Theophila might have been Demetrius' queen, but the likeliest guess would be that Apollodotus *was* Demetrius' youngest brother (p. 76) and that Theophila was their mother, the half-Seleucid consort of Euthydemus. Apollodorus says of this advance that not only did the Greeks occupy Patalene (the Indus delta country) but also, of the rest of the coast land, the kingdom called of Saraostos and the kingdom of Sigerdis.² The former has long been certain; it was Surāṣṭra, Kathiawar.³ There seems to have been a foreign element in the province, at any rate in the seaports, a thing in favour of the Greeks; Asoka's governor in Surāṣṭra had had an Iranian name, Tushaspa, and had been called a Yavana,¹ and Ceylonese tradition knows of a missionary, Dhammarakkita, sent by Asoka to Aparānta (Gujerat), who is called a Yona.² The other kingdom, that of Sigerdis, is unknown, but can only mean the country between Patalene and Surastrene, including Cutch;³ the provinces however will be discussed later (pp. 233 sqq.) and I only want here to get the outline of the conquest.

First 103 of 297 passages. Read on

Linked texts (31)