Pramāṇa · Search · About · API

The Greeks in Bactria and India c10.s4: Chapter VI. Menander and His Kingdom

The Greeks in Bactria and India c10.s4

Language
English

Chapter VI. Menander and His Kingdom

c10.s4.2

M ENANDER¹ was the most famous of the Yavana kings, and his legend attests the impression he made upon the world about him; and a sketch must now be attempted of the kingdom of the man who for a little while had held Asoka’s capital and whose conquests were exalted by a Greek historian even above those of Alexander. The deaths of Demetrius and Apollodotus and the return of Eucratides to Bactria left him master of the position in India, and thenceforth to his death he, the one man who had successfully resisted Eucratides, ruled the whole of the territory still remaining to the Greeks in that country, excluding the Paropamisadae; if he had not the royal title before (p. 167) he must have taken it, presumably by a vote of his army, when Demetrius was killed. The growth of his legend, and the establishment of his coinage in Barygaza, postulate for him a reign of reasonable length; at the same time the fact that his son, Strato I, was too young to rule alone when he died sets a definite limit to that length. He legitimised his rule by marrying Demetrius’ daughter Agathocleia; the evidence that she was his queen seems conclusive.² If cadets of the house of Euthydemus still survived,³ they must have accepted his rule as the only security against Eucratides and his line; he never had any civil war—at least in all his vast coinage no coin seems known which has been overstruck by anyone else or upon anyone else’s money.

c10.s4.3

¹ Menander of India is not included among the numerous Menanders in PW.

c10.s4.4

² Evidence: E. J. Rapson in *CHI* p. 552 n. 1. It is conclusive to myself, as to de la Vallée-Poussin, p. 236; this chapter I hope will show how other things make it inevitable.

c10.s4.5

³ For Antimachus II see p. 229, and on the question of Apollodotus II p. 318.

c10.s4.6

It is not possible to get an accurate chronology for Menander’s reign, but one must approximate as nearly as possible. Assuming that Agathocleia was Demetrius’ daughter and not a daughter of Apollodotus (p. 78), she may have married her father’s general directly after Demetrius’ death and his own return from Pāṭaliputra, *i.e.* in 166, or she may not have married him till the death of Apollodotus left him the sole repository of power in India and sole effective support of the Euthydemid cause, say *c.* 161 at latest; a later date than 166 seems probable, but it can hardly be supposed (since unmarried kings normally married on their accession) that Menander did not marry till he was rid of Eucratides in 159. On his death Agathocleia became regent for their son (or eldest son) Strato I; he was therefore a minor, and as it may be assumed that he would take the power at 18 at latest, he was under that age. Now Agathocleia’s first coins as regent bear her own portrait alone, with Strato named only in the Kharoshthi legend; her second series shows jugate busts of herself and Strato; then comes Strato’s coinage alone.¹ As she was regent long enough to issue two series of coins, and long enough for Heliocles, who overstruck some of her coins, to invade India during her regency, it appears that Strato cannot have taken the power till at earliest the third year after his father’s death; that is, he was certainly not over fifteen when Menander died and may of course have been two or three years younger. If now we take one extreme—that Menander married in 166 and that Strato was only twelve when he died—it is possible to put Menander’s death as early as 153, though that is very unlikely. If we take the other extreme—that Menander married about 161 and that Strato was fifteen when he died—it is possible to put Menander’s death as late as 145. To say that Menander’s son might not have been born till some years after the marriage (unlikely in that age) and to bring Menander’s death down to *c.* 140 seems to me impossible, for the association of Strato’s grandson Strato II with Strato as king cannot be later than *c.* 100 or soon after (Chap. VIII). It would seem then that Menander’s death must lie between the limits I have indicated, preferably in the latter half of the period; I shall therefore, with all necessary reserves, treat his death as occurring between 150 and 145. This would give him an undivided reign of probably not less than twelve years and perhaps two or three years more; that suffices. It cannot be very far out; for the portraits on his silver *δίκαιος* coins show an elderly man (p. 262) and, as has been seen, he must have proved himself as a general before Demetrius invaded India (see Addenda).

c10.s4.7

¹ For the coins of Agathocleia and Strato see Rapson, *Corolla Numismatica* 1906 p. 245; *CHI* p. 553.

c10.s4.8

There is no difficulty about these dates from Agathocleia’s side. It was found possible in Chapter IV to give the approximate dates at which Demetrius’ four sons were respectively given office as kings; they show that the eldest, Euthydemus II, cannot have been born very long after 204, say *c.* 202 at latest (it is known from Polybius that Demetrius was not yet married in 206), and that the youngest, Agathocles, cannot have been born too long after 195; 190 would hardly be possible. Agathocleia, from her name, should come next to Agathocles, whether before or after; her birth would then fall somewhere from *c.* 195 to *c.* 190. Again taking the two possible extremes, she might have been about twenty-four when she married or she might have been about thirty-four; probably she was not much, if at all, under thirty. There is precedent enough for Hellenistic princesses not marrying till thirty, or even later: putting aside the first Berenice, Ptolemais and Berenice II at once occur to the mind, even if the marriage of Cleopatra VII, who was thirty-three when she married Antony, be hardly in point; in Agathocleia’s case it may well have happened that, unlike her sisters in the West, there was no one of suitable standing for her to marry. It is generally safe to assume that a princess marrying a commoner was not in her first youth.

c10.s4.9

A modern writer has stated comprehensively that Menander’s empire extended from Mathurā (Muttra) in the east to Barygaza (Broach) in the west,¹ and this is substantially correct, allowing that in form it was a horse-shoe; but the details have to be considered. There is not much doubt about his south-eastern boundary. When he was recalled from Pātaliputra he abandoned a great stretch of the valleys of the Ganges and the Jumna and formed a new frontier to the south of Mathurā,² which remained the farthest town of importance toward the south-east held by the Greeks (p. 245); the Mathurā-Delhi country has produced many Greek coins.³ Whether Pātaliputra was reoccupied after Menander’s withdrawal by Pushyamitra has been disputed, but it seems probable that he reoccupied Ayodhya in Oude,⁴ and if so he was, as his legend states (p. 177), in possession of the capital. It seems uncertain whether Ayodhya and Sāketa were different towns or merely two names for one place, perhaps a double town;¹ but Menander had to take Sāketa from Pushyamitra on his way to the capital and the Sunga king would certainly reoccupy it if he could. In the story of Pushyamitra’s horse-sacrifice his grandson Vasumitra, who was guarding the horse, came to the south bank of the Sindhu and had a brush with some Yavana cavalry who were patrolling the northern bank;² this implies that the Sindhu was part of Menander’s frontier. It may be uncertain whether the Sind or the Sindhu tributary of the Chambal be meant; probably the Chambal, south of Mathurā, was the frontier west of the Jumna. On the Purāna chronology Pushyamitra died in 148, and as it was his grandson who guarded the horse, the incident must belong to the very end of his life; it is usually put about 150. As Menander’s death falls between 150 and 145, he and his opponent Pushyamitra must have died very close together. If Plutarch’s story that Menander died in camp be true,³ the reference must be to further fighting on this frontier, which may have had to be kept by the strong hand.

c10.s4.10

¹ Przyluski, *Açoka* p. 167.

c10.s4.11

² On the historical importance of the ‘Jumna march’—the Mathurā-Delhi district—as a natural frontier see A. J. Toynbee, *A Study of History* II, 2nd ed. pp. 130-1.

c10.s4.12

³ For Mathurā itself, Whitehead *NNM* p. 45. For the great hoard from Sonipat near Delhi, Cunningham, *NC* 1872 p. 159; 883 coins seen. For the hoard found in Bundelkhund south of the Jumna, including 40 coins of Menander, V. A. Smith, *Ind. Ant.* XXXIII, 1904, p. 217.

c10.s4.13

⁴ The much discussed Sunga inscription from Ayodhya (see in the last place D. R. Sahni, *Ep. Ind.* XX, 1929-30, p. 54) shows that a descendant of Pushyamitra, probably a son, was in possession of that town; it is probable therefore that Pushyamitra re-occupied Oude (see Grousset, I p. 39) which would entail possession of the capital.

c10.s4.14

¹ Rhys Davids, *Buddhist India* 1903 p. 39. So the old name of Pātaliputra was Kusumapura; but the *Yuga-purāna* (App. 4) shows that after the Mauryan Pātaliputra was built Kusumapura continued to have a separate existence.

c10.s4.15

² The story comes from Kālidāsa’s drama *Mālavikāgnimitra* (*CHI* p. 520) and is late, but most writers (not all) have accepted it as history; it makes little difference in any case, for it is certain that Menander held Mathurā, both from Ptolemy (p. 245) and from a hoard of 96 coins of his son Strato I found there (S. P. Noe, *A bibliography of Greek coin-hoards*, *NNM* 1925 p. 126), beside his own coins (p. 227 n. 3). Some have maintained that Kālidāsa’s Sindhu was the Indus, which historically is nonsense; also the Indus has no south bank, though R. C. Mazumdar, *IHQ* 1925 pp. 214 *sqq.*, says that we must suppose it had in the second century B.C.

c10.s4.16

³ *Mor.* 821 D, ἀποθανόντος ἐπὶ στρατοπέδου.

c10.s4.17

In the north, Menander’s treaty with Eucratides had left Alexandria-Kapisa to the latter and presumably therefore the Paropamisadae (p. 217). Certainly numbers of Menander’s coins are said to come into Begram,⁴ but the amount of trade passing through that gateway of the West would account for a good deal, and with a coinage so enormous and widespread as his the mere presence of used coins is little guide; had he ruled wherever his coins have been found he would have been king in Pembrokeshire,⁵ and his coins from Begram cannot compare with the vast numbers of those of Eucratides collected in that district. It is plain enough that he ruled from Mathurā to the Upper Indus; and his rule over Gandhāra is now proved also (see Addenda).

c10.s4.18

⁴ J. Hackin, *JA* 226, 1935, p. 290. Curiously, there was no coin of his in a hoard of 97 tetradrachms, of ten kings, found in the Paropamisadae in 1917: Whitehead, *NC* p. 315.

c10.s4.19

⁵ On his coin found at Tenby see V. A. Smith, *Ind. Ant.* xxxiv, 1905, p. 202. Warmington, p. 301, thinks it was brought to Wales by a trader; but more probably

c10.s4.20

There never was much doubt that Gandhāra was part of Menander’s realm, for 200 tetradrachms of his in mint condition have been found in Swat¹ and 721 drachmae, showing little signs of circulation, in the great Bajaur hoard;² coins in this condition have nothing to do with trade. At the same time he seems not to have used the Pushkalāvatī mint, for Siva’s bull never appears on his money, so probably he ruled Gandhāra through a sub-king; presumably he had some sub-kings, but only in the case of Gandhāra is it possible to identify one.³ The discovery of the Bajaur hoard put one obscure Graeco-Indian king, Antimachus II Nikephoros, in a new light, for the mass of almost new coins of his found there in conjunction with almost new coins of Apollodotus and Menander⁴ proves that he was an early king contemporary with Menander⁵ and that his kingdom must have been in Gandhāra. There can therefore be little doubt now that he was a son of Antimachus (not a grandson) who escaped to India from Eucratides’ attack and fought for Apollodotus or Menander or both against Eucratides, and that, after Eucratides retired, Menander made him his sub-king in Gandhāra; if he hated Eucratides as heartily as did his cousin Demetrius II, he was the right man to guard the frontier province. His title Nikephoros, the Victorious, must refer to some success gained against Eucratides in Menander’s war with him, the only occasion on which there could have been a Euthydemid success; this, and the fact that the treaty must have given Menander Gandhāra, are reasons for supposing that Menander may have recovered that province from

c10.s4.21

¹ Whitehead, *NC* p. 312.

c10.s4.22

² M. F. C. Martin, *JASB* XXIII, 1927 (pub. 1929), Num. Supp. XL p. 18.

c10.s4.23

³ Zoilus Dikaios *may* have been another, see p. 319.

c10.s4.24

⁴ Martin saw 969 coins (drachmae)—95 of Apollodotus, 721 of Menander, 152 of Antimachus II, and one of Zoilus—and believes that at least 1200 reached Peshawar. The man who *formed* a hoard was not necessarily the man who *buried* it; the Zoilus coin may have belonged to the subsequent owner of the hoard who buried it.

c10.s4.25

⁵ This could almost have been deduced from a hoard found in 1877 in a village in Bundelkhund south of the Jumna (V. A. Smith, *Ind. Ant.* XXXIII, 1904, p. 217) which contained 34 coins of Apollodotus (Soter), 40 of Menander, and 21 of Antimachus II, together with 3 of Eucratides.

c10.s4.26

it was a legionary’s mascot, just as Eucratides’ unique gold stater was once an Afghan officer’s signet. For some strange instances of mascots see S. Yeivin, *Ann. Serv.* XXXII, 1932, p. 152.

c10.s4.27

Eucratides by force of arms, for Eucratides’ recall from India was so urgent that it would seem that his treaty with Menander can only have been one hastily arranged on the basis of each keeping what he had. What clinches the matter is that all Antimachus’ four monograms are said to occur frequently on Menander’s coins:¹ he was therefore Menander’s sub-king, and his money, as was not unusual with sub-kings, was struck for him in his suzerain’s mints. His two characteristic types are the Gorgon-head, used by Demetrius and Menander,² and a new type, ‘King on prancing horse’;³ this latter was imitated by several later kings,⁴ who might therefore perhaps be his descendants. He was only a younger son of a younger son, who can have brought Menander little but his sword; juridically he had no claim to rule India which could compare with that of Menander’s wife Agathocleia, and it is likely enough that the suggestion that he was content to serve Menander as the one remaining bulwark against Eucratides⁵ is correct. Like Apollodotus, he never put his own portrait on his coins: whether it is conceivable that we have his portrait from Seleuceia is discussed in Appendix 8.

c10.s4.28

¹ Martin *op. cit.* p. 19.

c10.s4.29

² Antimachus, *BMC* p. 55; Demetrius, *ib.* p. 7 no. 14; Menander, *ib.* p. 49 nos. 59–62.

c10.s4.30

³ *Ib.* p. 55.

c10.s4.31

⁴ Philoxenus, Hippostratus, Nicias, and the Hermaeus-Calliope coins: *CHI* p. 586. See Chap. VIII.

c10.s4.32

⁵ Martin *op. cit.* p. 19.

c10.s4.33

In the south-west, Menander certainly ruled in Barygaza for a number of years, which may be taken to mean from Apollodotus’ death to his own; the evidence is precisely the same as for the rule of Apollodotus (see p. 149) and need not be repeated. This implies that he ruled the whole of the Indus country and southward to Gujerat. But Ujjain and Avanti were now Pushyamitra’s (pp. 164, 165 n. 1), which means that if they had belonged to Apollodotus they had been lost or abandoned just as Pāṭaliputra had been abandoned; if so, Madhyamikā must have been abandoned also and Greek rule confined to the coastal provinces. But here the important evidence is that of Claudius Ptolemy, which has been several times alluded to; and I must now turn to that writer.

c10.s4.34

We possess in Ptolemy some fragments of what must once have been a complete list of the provinces into which the Greek empire in India was divided.⁶ These fragments have never been noticed, since it was not possible to identify them till the -ηνn names in the Seleucid empire and its Succession states had been examined, and this was not done till 1929.¹ This matter has already been fully explained,² and it may be taken that, east of the Euphrates, names in -ηνn and -ιανη, the forms which provide a touchstone, are practically never used for anything but the Seleucid eparchies or the satrapal (i.e. primary) provinces of some kingdom which had either possessed the Seleucid organisation or was copying that organisation, whether at first or at second hand. In particular, the usage of Ptolemy himself with respect to the -ηνn names is both strict and consistent; and if therefore he locates a group of -ηνn names in India, a country where Greek had never before been used, it does not seem open to doubt that they are the provinces of a Greek kingdom (see further p. 240), especially as for Indian districts which were never ruled by Greeks he uses a different form of ending in -κη, as Larike, Anariake, Prasiake.³

c10.s4.35

⁶ Ptol. VII, 1, 55 and 42; the latter includes two fragments.

c10.s4.36

¹ Tarn, *SP Stud.* §IV.

c10.s4.37

² P. 3 and App. 2.

c10.s4.38

³ So in Thrace (III, 11, 8–9) all his provinces end in -κη, because it never received the Seleucid organisation.

c10.s4.39

A word must be said about Ptolemy’s sources and methods.⁴ As a geographer he was extremely painstaking but devoid of critical and historical instinct; he collected or had at his disposal a vast mass of material from every kind of source and of many different dates and set himself conscientiously to work everything in, somehow or other; he illustrates what a nuisance learning divorced from criticism can be. Many of his sources were itineraries compiled by merchants or other people; these he plotted out on his maps, deducting a regular percentage for exaggeration of distances, and located the places named accordingly. But itineraries were seldom exact, and different men may exaggerate differently or not at all; also they were of different dates and might call the same place by different names; he sometimes therefore arrived at the result of the same city existing in two places, and as he knew nothing about inner Asia himself he made two cities of it. Thus he makes Arsacia-Europus⁵ and Bactra-Zariaspa⁶ in each case two different cities, things which have taken in some modern writers; he puts Isagouros near Taxila but the Isagouroi on the Pamir;¹ his finest example is in the well-known Transjordania, where Rabbath-ammon and Philadelphia have become different cities located in different subsections of his list, which shows that he was using two documents of different periods.² Naturally boundaries of provinces or kingdoms varied greatly at different periods, but all periods with Ptolemy go in on one flat plane; so we get things like the Astauenoi located both in Aria and Hyrcania,³ because *at different times* that may have been true. One has therefore, over each separate bit of information, to seek the date of the ultimate source he is using. A good example is his extraordinary Margiane (p. 89), where the first question is, at what date can we find a province or kingdom which may have given rise to his account? Having plotted out the position of his cities from itineraries or what not, he then proceeded to determine the latitude and longitude of the position chosen; in inner Asia, at any rate, the only value of his co-ordinates is to indicate that a place probably stood *somewhere* in that locality or to give very roughly its *relative* position. The idea that he had at his disposal a number of towns whose latitude and longitude had been determined by somebody on the spot, because Hipparchus had expressed a hope that one day this might be done, is, at any rate for inner Asia, quite out of the question.⁴

c10.s4.40

⁴ I am speaking throughout only of the *ultimate* sources of the work we call Ptolemy’s, without prejudice to the question of how much he took from Marinus and Marinus’ painstaking collection of material (see Honigmann, *Marinos von Tyros* in PW) or of how the ultimate sources reached him; such questions do not affect my object, which is to bring out the Hellenistic material in the work.

c10.s4.41

⁵ VI, 2, 16 and 17. Both are names of Rhagae, Strabo XI, 524.

c10.s4.42

⁶ VI, 11, 7 and 9; see p. 114 n. 5.

c10.s4.43

¹ Lévi, *JA* 1915 p. 85, who adds that in the Hindu-Kush region Ptolemy often takes a name from two different itineraries without seeing that it is the same.

c10.s4.44

² He gives Philadelphia under the Syrian Decapolis (v, 15-23) but ‘Ραβαθμών under Arabia Petraea (v, 17, 3), *i.e.* from a document of the period of Nabataean rule.

c10.s4.45

³ vi, 9 and 17.

c10.s4.46

⁴ Berthelot believed this, p. 117, and called it ‘une vaste enquête’.

c10.s4.47

Consequently, as regards the Greek provinces in India, the first thing to consider is the date of Ptolemy’s information. Some of these provinces he locates in ‘Indo-Scythia’,⁵ the kingdom of the Sacas in India, which stops part of the way up the Indus, Abiria being the northernmost province (I shall take the several provinces presently). His source here then dates from a time when the Sacas, who went *up* the Indus from Abiria, had not yet reached the Punjab; and this intermediate stage in the Saca advance must have been stable long enough for it to find its way into history. Now ‘Indo-Scythia’ belongs to the generation which centres on 100 B.C.; the Saca invasion of India may have begun any time after about 120, and the Sacas reached Taxila before c. 77 B.C. but only shortly before, since on the Taxila copper-plate of that date Maues is not yet, as he is on the majority of his coins, Great King of Kings;¹ one may say therefore that the latest date for the advance northward from Abiria would be c. 80 B.C. Ptolemy’s kingdom of ‘Indo-Scythia’ will thus belong in the period c. 110 to c. 80 B.C., and c. 80 is the latest possible date for his ultimate source for the North-West, since in his north-western India the Sacas have not yet appeared. The Greek names of the provinces in his ‘Indo-Scythia’ were obviously not given to them by Sacas but by Greeks before the Sacas came; the names are therefore second century B.C., and in one case, Abiria, this can be proved from Patañjali (p. 235). But if the Greek names continued to be used in the Saca period, as they evidently were, then they were well-established names, that is, they go back some way into the second century. Now as the three fragments in Ptolemy which we possess belong to widely separated districts—Sind with Kathiawar, Gandhāra, and the eastern Punjab—there must once have been in existence a complete list of the Greek provinces in India, something like the list given by Hieronymus of Cardia of the satrapies of Alexander’s empire (p. 46); and as the names in the list, as has been seen, go well back into the second century B.C., there is no reasonable doubt that the list, whatever the date of its compilation, referred to the flourishing period of Greek rule, whether before the death of Demetrius or during the reign of Menander. Whether the original was an official or a historian’s list cannot be said, but there can be little doubt that the writer who reproduced the list if it was official or compiled it if it was not, and from whom bits of it ultimately reached Ptolemy, was the historian whom I have called ‘Trogus’ source’ (see p. 46) and who completed his history round about 85 B.C. and included in it an account of the reign of Menander.

c10.s4.48

⁵ Ptol. vii, 1, 55.

c10.s4.49

¹ See on the invasion Chap. VIII, pp. 320 sqq. and App. 16, especially p. 501.

c10.s4.50

The first fragment gives the Greek names of certain provinces in Ptolemy’s ‘Indo-Scythia’.² It begins with Patalene, the Indus Delta country, a province whose name and location are well known from Strabo. North of Patalene lies Abiria. South of Patalene, past the mouths of the Indus and round the ‘Kanthian Gulf’, lies Surastrene. Ptolemy knows absolutely nothing of the coast line here; he does not know of the Rann of Cutch, of Cutch, or of Kathiawar; his ‘Kanthian Gulf’ extends in a flat curve from the Indus mouth to the coast of Gujerat,¹ and his Surastrene apparently extends northward to adjoin Patalene, which would make a province of unmanageable size and cannot possibly be right. Surastrene, Apollodorus’ ‘kingdom called of Saraostos’ (p. 147), included both Kathiawar (Surāṣṭra) and, as the name has been preserved in the modern Surat, part of Gujerat also, *i.e.* the eastern coast of the Gulf of Cambaye with Surat and Barygaza (Broach); indeed the Periplus,² about the middle of the first century A.D. (p. 148 n. 4), says explicitly that Surastrene included part of Gujerat, and though, strictly speaking, this may not be evidence for the state of things two centuries earlier, Ptolemy leaves little room for doubting that the Sacas retained the Greek provinces. But Apollodorus mentions a second kingdom, also on the coast, that of Sigerdis (pp. 147 *sq.*), as held (imperfect tense, *i.e.* held for some time) by the Greeks; there must have been another province corresponding to this kingdom, and as it was on the sea it must have lain between Patalene and Surastrene and have included among other things Cutch; the only alternative would be to place it south of Surastrene, which need not be seriously considered. Now it was traditional at a later time that Sind should be divided into four provinces³—Upper and Middle Sind along the Indus, Lower Sind which was the Delta country and corresponded nearly enough to the Greek Patalene, and Chach (Cutch).⁴ So Hsüan Tsiang found it; but Cutch, according to the dimensions he gives, was far larger than the peninsula of that name and was supposed by Cunningham to have also included the country north of the Rann,⁵ and this supplies a *prima facie* reason for supposing that there was a Greek province whose name is lost between Patalene and Surastrene, roughly corresponding to the later province of Cutch. It has been seen that, beside Demetrias in Patalene, another Greek city with a (probably) dynastic name, Theophila, stood somewhere east of Patalene (p. 147): Greek cities in India were so scarce that Theophila must have been the capital of a province, and as it was much too far north to be the capital of Surastrene I take it to have been the capital of the province corresponding to the later Cutch. This does not necessarily mean that Ptolemy was wrong; for the form of his sentence about Surastrene is such that the name of a province could easily have dropped out.¹ Theophila can no more have been a completely new Greek foundation than Demetrias in Patalene; it would be an Indian town resettled and was probably the later Minnagara, ‘Min town’,² a capital of the Parthians, which also stood somewhere eastward of the Indus Delta;³ it is noteworthy that in the twelfth century an Arab writer recorded that a ‘descendant of Alexander’ was ruling in Minnagara,⁴ which might be an Islamite legend but might also mean that Minnagara had once been Greek. The Greek names of both Patalene and Surastrene lasted long, as can be seen from Strabo and the Periplus; both provinces were therefore under Greek rule long enough for the names to become well established, and this is confirmed by the curious discovery that the women of Surastrene continued to use as a form of greeting the Greek χαίρειν or χαῖρε.⁵ Whether Junagarh, the old capital of Kathiawar, be Yonagarh, ‘Greek-town’, as has been suggested, I do not know.

c10.s4.51

² VII, 1, 55: ’Ινδοσκυθία, ταυτῆς δ’ ἡ μὲν παρὰ τὸν διαμερισμὸν τῶν στομάτων (of the Indus) Παταληνή, καὶ ἡ ὑπερκειμένη αὐτῆς ’Αβιρία, ἡ δὲ παρὰ τὰ στόματα τοῦ ’Ινδου καὶ τὸν Κανθικὸν κόλπον Συραστρηνή (Renou’s text).

c10.s4.52

¹ See the map in Berthelot facing p. 312 and cf. p. 319.

c10.s4.53

² Periplus 41, τὰ δὲ παραθαλάσσια (of his Ariake) Συραστρηνή.

c10.s4.54

³ Cunningham, Geog. p. 285.

c10.s4.55

⁴ Chach means ‘sea-coast land’, Mazumdar in Cunningham, Geog. p. 696.

c10.s4.56

⁵ Op. cit. p. 347, but see Mazumdar’s note. See also on Chach T. Watters, On Yuan Chwang’s travels in India II, 1905, p. 245.

c10.s4.57

¹ This is not apparent in Renou’s text (set out p. 233 n. 2), based on X (on the MSS see App. 13). But the note in Renou’s apparatus, p. 25, shows that Γ and ω read ἡ δὲ περὶ τὰ στόματα τοῦ Ἴνδου for παρὰ of X (which cannot be right, as Patalene was περὶ) and that Γ also inserts a second ἡ περὶ, thus making a reading ἡ δὲ περὶ (read παρὰ) τὰ στόματα τοῦ Ἴνδου καὶ ἡ περὶ τὸν Κανθικὸν κόλπον Συραστρηνή, which certainly suggests the loss of a province-name after Ἴνδου. The name may have fallen out early and the lacuna been smoothed over in X and ω.

c10.s4.58

² Cunningham, Geog. p. 334, suggested that Min nagara was connected with Isidore’s Min in Sakastene. Min should be Saca, while Minnagara was a capital of the Parthian Gondophares; but it may have been a Saca capital first. Herzfeld, Sakastan p. 4, notices the connection, but does not elucidate Min.

c10.s4.59

³ Periplus 38, to one going eastward along the coast from the Indus Minnagara is μεσόγειος. Not to be confused with the other Minnagara between Barygaza and Ujjain, Periplus 41, Ptol. VII, I, 62.

c10.s4.60

⁴ Cunningham, Geog. p. 336.

c10.s4.61

⁵ A. Weber, Indische Studien IV pp. 269–70, 349–50; IX p. 380; Berlin SB 1890 p. 911; he cites a Shikshā, but their dates vary enormously (Winternitz, Eng. Tr. I pp. 282–5) and I have no means of ascertaining the date.

c10.s4.62

Ptolemy’s third province, Abiria, northward of Patalene, is also a good Greek form and is named from the Abhīras. Patañjali locates this people, about the end of the second century B.C. (p. 146 n. 3), in Sindhudeça,⁶ the ‘country of the Indus’, which supplies an additional proof that the list of provinces behind Ptolemy’s sources refers to the second century, for at a later time the Abhiras moved much farther southward and founded a kingdom in Avanti.¹ Ptolemy’s Abiria must have corresponded in main outline to the later province of Middle Sind. But this only takes us part of the way up the Indus, and there must have been another Greek province between Abiria and the confluence of the Punjab rivers with the Indus, corresponding to the later province of Upper Sind. The name of this province has, I think, been preserved by Pliny (VI, 71), doubtless ultimately from the same list which lies behind Ptolemy’s sources: he says that the Indus formed two islands, a very large one called Prasiane and a smaller one called Patala; and Prasiane, a properly formed eparchy name, should be the name of the Indus province north of Abiria. The explanation of this ‘island’, which only means a piece of land between two pieces of water,² is given by Aristobulus,³ who says that when he was there the Indus had recently changed its course into ‘the much deeper channel’ to the eastward—doubtless the Hakra channel—and that an enormous tract of country was perishing for want of water. Presumably therefore, at the date of Pliny’s information, the Indus was running in *both* channels and a long tract of country between them was fertile; Prasiane means ‘the eastern country’, *i.e.* east of the Indus channel proper, but the province would include far more than the actual ‘island’. The strange view sometimes put forward,⁴ that during the Greek period the Indus and its tributaries ran precisely as they do to-day, hardly requires mention; Aristobulus and Pliny’s source are evidence enough of changes within the period itself, which is all that concerns me here. Prasiane has of course nothing to do with the Prasii, the ‘Easterners’ of Magadha on the Ganges, whose country Ptolemy (VII, 1, 53) calls Prasiake.

c10.s4.63

⁶ See A. B. Keith, A History of Sanskrit Literature 1928 p. 33, who however misplaces Patañjali’s Sindhudeça; its natural meaning is in agreement with Ptolemy, and what Kālidāsa may have said some six centuries later does not affect the matter.

c10.s4.64

¹ Indian evidence in V. A. Smith⁴ p. 290. They had moved southward before the time of the *Periplus* (41). The Paikuli inscription also puts them near Avanti, Herzfeld *Sakastan* p. 90 no. 2, and one of the Nasik cave inscriptions mentions an Abhira king, *Arch. Survey of Western India* IV, 1883, p. 103 no. 12.

c10.s4.65

² Cunningham, *Geog.* p. 251 and map IX, thought that the Indus sometimes made a true island just north of Patalene; but this could not be Pliny’s island, as it would be in Abiria, while Prasiane is an eparchy name itself. The word ‘island’ here must ultimately represent the Sanscrit *dvipa*, *doab*, which means any tract of land between two rivers (*cf. CHI* p. 550), a usage common everywhere; Arab geographers called Mesopotamia *el jezireh*, ‘the island’ (J. Kirste, *Wien SB* 182, 1918, Abh. 2 p. 72), and compare the long peninsula in Scotland called the Black Isle of Cromarty.

c10.s4.66

³ Strabo XV, 693, ἐκτραπομένου (the Indus) εἰς τὸ ἕτερον ἐν ἀριστέρᾳ κοιλό-τερον πολύ. The definite article shows that two channels were well known. Jacoby in *F. Gr. Hist.* no. 139 fr. 35 prints a bad conjecture, τι for τό.

c10.s4.67

⁴ E.g. Berthelot pp. 269, 271-3, and Ernst Meyer on the Beas, *Klio* XXI, 1927, p. 183. On the known changes see Cunningham *Geog.* pp. 253-6; R. B. Whitehead, *Ind. Ant.* 1932 p. 162; and for the Indus especially, H. Cousens, *Memoirs of the Arch. Survey of India* XLVI, 1929, who (pp. 3-6) concludes that there were many different channels, often altering, and that we do not really *know* anything. Doubt-

c10.s4.68

South of the Punjab, then, the Greeks made five provinces, substantially according with later divisions of the country, their names from north to south being Prasiane, Abiria, Patalene, (Cutch), Surastrene, the Greek name of the fourth being lost. What sort of provinces they were will be considered later.

c10.s4.69

The second fragment of the list (VII, 1, 42) gives two names of provinces in the Gandhāra kingdom, Souastene (Swat) and Goruaia, which is the alternative form of Goryene (cf. p. 3). Souastene must be Lower and perhaps Middle Swat; how far up the river Greek rule extended cannot be said,¹ for though Ptolemy describes Souastene as below the sources of the Swat river, it does not follow that anyone knew where those sources were. Goruaia must be the province between the Gouraios (lower Swat river) and the Kunar, the modern Bajaur.² Omitting Gandaritis, which *might* only mean Gandhāra generally,³ one other province of Gandhāra is known from Arrian, Peucelaïtis,⁴ the province along and north of the Kabul river with the capital Peucela⁵ (Pushkalāvatī, now Charsadda).

c10.s4.70

¹ Sir A. Stein, *On Alexander’s track to the Indus* 1929, found Graeco-Buddhist art motives still freely employed in wood carvings, not only in Middle Swat (p. 64) but as far north as Branial (p. 93). But this does not mean that Greeks ever ruled there.

c10.s4.71

² Berthelot, p. 279, would derive the name Goruaia not from the river but from Ptolemy’s town Gorya, the Gorys of Strabo xv, 697, which he places on the Kunar. The derivation seems probable; but Strabo is here most confused and it is impossible to be sure that his ‘Choaspes’ on which Gorys stands means the Kunar; most probably it is meant for the united Panjshir-Ghorband river (p. 97 n. 2). But Goruaia certainly extended to the Kunar, for Ptolemy makes it include Nagara-Dionysopolis, the site of which has been identified (near Jalalabad).

c10.s4.72

³ See p. 445 for two cases of eparchy names improperly used.

c10.s4.73

⁴ In Arr. *Ind.* 4, 11 nearly all the MSS give Πευκελαΐτιδι (for the district), which on the analogy of a number of province-names in -ιτις is preferable to the Πευκελαῶτις of Arr. *Anab.* IV, 22, 7. Various other spellings have been proposed. It was not of course called Peucelaïtis when Alexander arrived; Arrian, as so often, is using the later Greek name proleptically.

c10.s4.74

⁵ The MSS of Arr. *Ind.* 1, 8 give the Greek name of the capital as Πεύκελα or Πεύκελλα, and later writers give for the inhabitants the forms Πευκαλεῖς and in

c10.s4.75

less V. A. Smith went too far the other way; but it is strange that no one has discovered the Aristobulus passage.

c10.s4.76

We know then three of the Greek provinces: Goruaia (Bajaur), Souastene (Swat), and Peucelaïtis (between Swat and the Kabul river); there must have been two others at least, Buner and the Peshawur country, whose names are lost, though conceivably one was Gandaritis. Still, taking the western part of Menander’s realm from the Paropamisadae border to Surastrene, we are not badly informed about the province-names and can see the outline of the organisation.

c10.s4.77

But little is actually recorded of the country east of the Indus. A third fragment in Ptolemy (VII, 42) gives the names of two provinces in Menander’s home kingdom east of the Jhelum: Kaspeiria,¹ which Ptolemy calls the upper valleys of the Jhelum, Chenab, and Ravi, and which would thus have corresponded to southern Kashmir; and Kulindrene, named from the Kulindas or Kunindas,² which he calls the upper valleys of the Beas, Sutlej, Jumna, and Ganges, a statement which cannot be correct as it stands. We can see from some of the other provinces, as Abiria (the Abhīras) and Patalene (the Sauvīras), that there was a natural tendency to make the province coincide with the tribe or people, precisely as in the Seleucid empire; and the Kunindas were only one of several peoples in the eastern Punjab who lay between Sāgala (the Madras) and Mathurā and who must have been included in the Greek empire, not only because of their geographical position but because they started coining at the time which saw the end of Greek rule and the establishment of their independence (pp. 324 sq.). This will be considered later; it need only be said here that there must have been other provinces south-east of Sāgala beside Kulindrene—the whole country down to Mathurā could not have been included in one province—though whether each of these peoples formed a separate

c10.s4.78

¹ Stephanus’ Κάσπειρος · πόλις Πάρθων προσεχὴς τῇ Ἰνδικῇ is fictitious; Herzfeld, *Arch. Mitt. aus Iran* I p. 94, has pointed out that it does not come from Herodotus, as Stephanus says, but from the *Bassarica* of Dionysius.

c10.s4.79

² Kulinda usually in literature, Kuninda always on the coins: *BMC India* p. cxxxix. On the interchange of *l* and *n* see Lévi, *JA* 1915 p. 101.

c10.s4.80

Latin Peucalei = Ποκλάεις of Ptol. VII, 1, 44; references collected in the Index to Renou, *La géographie de Ptolémée: l’Inde* s.v. Ποκλάεις (Peucolis of Pliny VI, 94 can however hardly be the same place). When Arrian, *Anab.* IV, 28, 6 (in the MSS), speaks of πόλιν Πευκελιῶτιν he is transferring the name of the province to the city, possibly because the Indian name for the city was Pushkalāvatī, which had become graecised as the province-name. But the Greek form Peucela finally affected the Indian form; on the unique autonomous coin of the city the Kharoshthi legend gives Pakhalavadi, *CHI* p. 587. To avoid confusion, I call the city Pushkalāvatī throughout, as that is usually done.

c10.s4.81

province cannot be said. Unfortunately the topographical indications in the Brihat Samhitā are too generalised to be of use here, and these peoples can only be located by the find-spots of their coins—an unsatisfactory method, but there is nothing better to be had.

c10.s4.82

The find-spots show that the most northerly people were the wealthy Audumbaras,¹ whose country was on the upper Beas in the Gurdaspur and Hoshiarpur districts, with its centre perhaps about Pathankot; their location shows that Ptolemy has not only made Kulindrene too large but has placed it too far to the north. This people manufactured a fine cotton cloth, kotumbara, which they traded to Sāgala;² apparently they were Buddhists,³ and some of their coins after Greek rule ended imitated Greek types, the ‘elephant and humped bull’ type of Apollodotus I⁴ and the regular type of Demetrius, Heracles standing and crowning himself,⁵ so freely imitated by the Sacas. Southward of the Audumbaras were the Trigartas,⁶ between the Ravi and the Sutlej with their centre about Jullundur, and the Kunindas⁷ eastward of the Trigartas, somewhere between the Sutlej (Ludhiana district) and the Jumna at Saharanpur in the United Provinces; their centre may have been about Ambala. The coins of the powerful Yaudheyas⁸ have been found in many places in the Punjab, but especially between the Sutlej and the Jumna; large finds have been made at Sonipat a little north of Delhi, and many types are said to have come from near Saharanpur. Geographically, their territory can only have lain southward of the Kunindas, with whose coins some of their own have affinities,⁹ and of the Trigartas, with whom they are connected in literature;¹⁰ it may perhaps be called the Delhi country with a large extension northward and north-eastward. Lastly come the Ārjunāyanas,¹¹ somewhere between Delhi and Agra; they must have marched with the Yaudheyas, with whom they are connected both in literature and by coin-legends, and were presumably Mathurā’s nearest neighbours.

c10.s4.83

¹ J. Przyluski, *JA* 1926 p. 1; *BMC India* pp. lxxxiii *sqq*.

c10.s4.84

² Przyluski *op. cit.* pp. 21–2.

c10.s4.85

³ The regular type on their coins of the first century B.C. is a stupa, *BMC India* pp. lxxxiii, 122 *sqq*.

c10.s4.86

⁴ *Ib.* pp. 123 nos. 12, 13; 125 no. 23. I take ‘Apollodotus Philopator’ on p. xv to be a slip.

c10.s4.87

⁵ *Ib.* pp. cxxxiv–v.

c10.s4.88

⁶ *Ib.* p. cxxxix.

c10.s4.89

⁷ *Ib.* p. ci.

c10.s4.90

⁸ *Ib.* pp. cxlvii *sqq*.

c10.s4.91

⁹ *Ib.* p. cxlix.

c10.s4.92

¹⁰ *Ib.* p. cxxxix.

c10.s4.93

¹¹ *Ib.* p. lxxxii.

c10.s4.94

Which, if any, of these peoples represent the Oxydracae of the Alexander-story, who lived between the Beas and the Sutlej, or whether the Oxydracae had gone south to Malva with their allies the Malli and lost their identity, cannot be said.

c10.s4.95

These peoples probably cover the whole of Menander’s domain between the Madras and the frontier south of Mathurā.¹ No doubt it is unfortunate that Ptolemy’s record is so imperfect, and not only in the east; we know nothing of the provinces, which must have existed, in the southern Punjab and in the important kingdom of Taxila. However, when one considers his complete ignorance of the Seleucid organisation and province-names in parts of eastern Iran, one must be very grateful that he has preserved as much as he has done about India.

c10.s4.96

¹ There is no means of knowing if or how far his rule extended into the northern part of the United Provinces. Ptolemy’s reference to the Ganges *may* mean that it did.

c10.s4.97

Of the nine province-names east of the Paropamisadae thus recovered, six—Surastrene, Patalene, Prasiane, Souastene, Goruaia (Goryene), and Kulindrene—have names formed in the usual fashion of Seleucid eparchies, as have the four provinces known in the Paropamisadae (pp. 96 *sq.*); and the other three—Abiria, Kaspeiria, Peucelaïtis—have names of types regularly found among the provinces of Seleucid Succession states, probably also, in their time, Seleucid eparchies. That the Indian conquests therefore were organised in imitation of the Seleucid system, more or less, is obvious; but it is also obvious that these great provinces were not eparchies, that is, subdivisions of satrapies, but were full satrapies, with governors responsible only to the king. The Greek kings of Bactria had already turned the eparchies of Bactria-Sogdiana into satrapies of the Bactrian kingdom (p. 113), and to Greeks from Bactria, like the Euthydemids, it would be the natural thing to give eparchy names to the new satrapies they formed, just as was subsequently done by every Succession state; in both respects Bactria provided the model which Parthia and every other Succession state was ultimately to copy. But in one way these Indian satrapies were a new departure. The Achaemenid satrapies had often been too large to be properly controlled from the centre of the kingdom, and there are indications that Alexander had had some idea of breaking them up.² This was not actually done; but the Seleucids had subdivided these satrapies into eparchies, and it was the eparchies, not the satrapies, which speaking generally survived to be the basis of the organisation of Asia when the Seleucid realm broke up. But in India we meet the great satrapies again without the threefold Seleucid subdivision, though the population was probably denser than the Iranian. It looks like a retrograde step; but I do not think that is the explanation. As the organisation of the Seleucid empire was necessarily, from its size, looser than that of Ptolemaic Egypt, so the Greek organisation of India was looser again than that of the Seleucids; and the reason, apart from the lack of sufficient Greeks, must be sought in the retention in large measure of the existing native organisations (p. 259). In both empires the satrapies increased in size as the distance from the centre increased; a Seleucid satrapy in western Asia Minor was a more compact affair than Media or Bactria-Sogdiana, and the Greek satrapies in Gandhāra were of more manageable size than Surastrene. A modern parallel can be seen in the great size of the Highland counties of Scotland as compared to the Lowland.

c10.s4.98

² *CAH* VI p. 426.

c10.s4.99

The Indian satrapies, as one would expect, were governed by generals, *strategoi*, like those of the Seleucids; this is clear from the use of the word *strategoi* on the coins of both Saca and Parthian kings in India.¹ It is perhaps not certainly known what term the Parthians themselves may have used for the governors of their satrapies, but Greek and Jewish writers and scribes as a rule called them satraps, though occasionally they translated the Parthian term by *strategos*, a word borrowed from the Seleucid organisation which they knew;² if therefore in the Kharoshthi legends on Saca and Parthian coins in India their moneyers used *stratega* as a translation, they too must have taken it from something *they* knew, which could only be the Greek organisation in India. This conclusion is not affected by the fact that the Sacas, who stepped into the shoes of the Greeks, also used the term satraps,¹ as an Iranian people naturally would. Whether there was any and what subdivision of the more remote satrapies cannot be said, but there was subdivision in the North-West, for inscriptions have revealed the existence there of meridarchs, ‘governors of fractions’; of the two known, one² was a Greek and belonged to the later period of Greek rule in Swat, and the other,³ who was certainly an Indian, belonged to the Taxila kingdom and seemingly held office under some Saca king, another proof of how closely the Sacas followed the Greeks. Meridarchs are not heard of in the Seleucid empire except in Palestine, and there only in Jewish writers; it is impossible to say what they mean.⁴ In Ptolemaic Egypt the Arsinoite nome, the Fayûm, was, unlike others, divided into three *merides*,⁵ but the purpose of this is obscure. It is strange therefore to meet meridarchs in India, and one cannot explain them more closely than as subordinate governors of parts of a satrapy; I imagine that the Greek kings adopted this word because, as they had not the triple Seleucid division of satrapy, eparchy, hyparchy, they could not use the Seleucid terms and only wanted some indefinite word which would express the subdivision of the satrapy into districts.⁶ Incidentally it shows that they were acquainted with the Seleucid organisation in Palestine.

c10.s4.100

¹ The *strategoi* Aspavarman and Sasas are known from coins, the former under Azes and Gondophares, the latter under Gondophares and Pacores; see *CHI* p. 577; de la Vallée-Poussin p. 271; Herzfeld, *Sakastan* pp. 92, 101. The word is copied exactly, the Indian form (in the genitive) being *strategasa*. Aspavarmasa *strategasa* also in an inscription, *ASI* 1929-30 p. 62 no. 46.

c10.s4.101

² See the discussion in M. Rostovtzeff and C. B. Welles on *Doura Pg.* 10, *Yale Class. Stud.* 11 pp. 46 sq. and add to the references for ‘satrap’ Josephus, *Ant.* xx, 24, 54; *SEG* vii, 13 l. 6, on which see Fr. Cumont, *CR Ac. Inscr.* 1931 p. 245; and the Gotarzes inscription (text best in Herzfeld, *Sakastan* p. 80). But I feel no certainty that the *strategos* of *Doura Pg.* 10 was a satrap rather than a military commander in a group of frontier provinces.

c10.s4.102

¹ List of Saca satraps in de la Vallée-Poussin p. 268; see also *CII* p. xxxiv. It seems however that the Saca Great Satraps in Malva and Mathurā were really viceroys, corresponding perhaps to the Greek sub-kings.

c10.s4.103

² Theodorus the meridarch, *CII* no. 1; see F. W. Thomas, *Festschrift Ernst Windisch* 1914 p. 362, and Chap. ix p. 388. The name has recently been read as Theüdata = Theodotus (C. C. Dāsa Gupta, *JRAS* 1933 p. 403); Thomas however (ib. p. 405) says that Theodorus still seems to him certain.

c10.s4.104

³ *CII* no. 11, from Taxila; the name of the meridarch, a Buddhist by birth, is lost. For his nationality see p. 358 n. 1.

c10.s4.105

⁴ Jos. *Ant.* xv, 216; xii, 261, 264; *I Macc.* x, 65. The usual view is that the meridarch governed some subdivision of a satrapy, but the difficulties are great; see U. Kahrstedt, *Syrische Territorien in hellenistischer Zeit* p. 54, whose own view, that the meridarch was head of the *Selbstverwaltung*, raises other difficulties in turn. It really looks as if καὶ μεριδάρχης in *I Macc.* x, 65 is an unintelligent gloss.

c10.s4.106

⁵ A possible third-century meridarch of the *Πολέμωνος μερίς* was suggested by B. A. van Groningen, *Aegyptus* xiii p. 21; see however U. Wilcken, *Arch. f. Pap.* xi, 1935, p. 125.

c10.s4.107

⁶ There was an indefinite use; see Stephanus *s.v.* *’Ατροπατία*, where Atropatene is called the second *μερίς* of Media as opposed to Media Magna, and cf. Strabo xvi, 749.

c10.s4.108

It will have been noticed that for the titles of the Greek officials we have largely to go to inscriptions of the Saca period. The Sacas in fact seem to have been no less imitative than their Parthian kinsfolk; though they came up the Indus instead of down, they followed where the Greeks had led, kept the Greek provinces, the Greek system of officials, the Greek coin-names (p. 85 n. 8), the Macedonian month-names (p. 359), and coined in the Greek mints with Greek moneyers.¹ This is not without its importance. The Saca satraps in Mathurā and Surastrene confirm the fact that Greeks had ruled there before them; and their western Great Satraps who, starting from the coastal provinces, at one time conquered and held Ujjain,² furnish an additional reason for supposing that Apollodotus, after securing the coastal districts, may have occupied that city, if only for a short time. Though the Saca Great Satraps became practically independent, the Saca system in theory was that of a king ruling in the North-West, with his seat apparently at Taxila, and two Great Satraps (viceroys), one in Mathurā for the east and one in Surastrene for the west; the resemblance of this arrangement to Demetrius' scheme of government as deduced in Chapter IV is patent, and I venture to think furnishes valuable confirmation of the correctness of that deduction.

First 107 of 290 passages. Read on

Linked texts (30)