Pramāṇa · Search · About · API

The Greeks in Bactria and India c10.s7: Chapter IX. The Greeks and India

The Greeks in Bactria and India c10.s7

Language
English

Chapter IX. The Greeks and India

c10.s7.2

THE end of Greek rule east of the Hindu Kush closely and rather dramatically coincided in date with two linked events in the Mediterranean world, the end of the last Hellenistic kingdom in the West, Egypt, and the establishment of the Roman empire of Augustus; as Augustus entered Alexandria on the first of August 30 B.C., it is conceivable that one or even both of the two remaining Greek kingdoms in India was actually the last Hellenistic kingdom to survive in independence. The reign of Augustus, though he had nothing much to do with the world east of the Euphrates, forms a convenient line of division both in India and in the Parthian empire, though not in the same way. For the Greeks of Parthia it marks, very roughly, the time when the period of achievement passes over into the period of decay, though in some places, as Seleuceia on the Tigris and Seleuceia on the Eulaeus (Susa), the latter process was perhaps not very marked till the second century A.D. In India, however, the period of achievement seems to have ended with Menander’s death, though that may be an illusion due to the fact that the succeeding half century is to us a blank; but in any case it will appear that the process of Indianisation of the Greeks cannot have well begun later than the early part of the first century B.C. The division which Augustus’ reign makes in India is a hard and fast division in the nature of our evidence. The knowledge, principally of Northern India, obtained by the Greek invaders is carried no further, but is replaced by information about the coasts and the ports derived from merchants of the Roman empire, men of every nationality, who now began to visit the harbours of Southern and of Farther India and even of China, giving Greek names to the little nameless islets where they found anchorage or water;¹ they learnt much about the coasts, but little or nothing about the interior of Northern India, with which they were not concerned. Much information about their doings survives in the Periplus, in Pliny, and in Ptolemy, and has been worked up in modern books; this has to be separated carefully from the fairly abundant Hellenistic information in Ptolemy and Pliny, and the Hellenistic historian must avoid drawing on it unless he has some definite reason, for it has nothing to do with the story of the Bactrian Greeks in India.

c10.s7.3

¹ Such names are too common throughout Ptolemy book VII to list.

c10.s7.4

As in this chapter I shall sometimes have to refer to Kanishka and other Kushan kings, it may be convenient to some readers to have an indication of the dating. The five Kushan kings (or principal kings) in order are Kujula Kadphises I, Wima Kadphises II, Kanishka, Huvishka, Vasudeva. The reign of Kujula, the conqueror of the Paropamisadae, is from somewhere after A.D. 25, the date at which the Hou-han-shu begins, to c. A.D. 50. Professor F. W. Thomas tells me that his chronology in the Cambridge History of India for the next two reigns will be Wima (the conqueror of Northern India) c. 50 to 78 (the so-called Saca Era),¹ Kanishka 78 to c. 103. The date of Kanishka’s accession has been one of the most disputed points in history, four theories having found support:² the Vikrama Era, 58 B.C.;³ A.D. 78; c. A.D. 100; second century A.D. To-day the question seems to lie between 78 and 128 (Dr Sten Konow’s date).⁴ I do not believe in 128 myself; even if it should turn out that, as some believe, 78 is a Malva Era (see p. 335) and not the Era of Kanishka by which the later Kushans dated, that would not make Kanishka second century. To avoid begging questions I generally (after Kujula) refer to reigns, not dates.

c10.s7.5

¹ Sir J. Marshall, ASI 1929-30 (pub. 1935) pp. 55-7, has recently given reasons for believing that the Kushans, i.e. Wima Kadphises, took Taxila between A.D. 60 and 64.

c10.s7.6

² A convenient list of the writers who have supported each of these theories will be found in Winternitz, Eng. trans. II pp. 612-13.

c10.s7.7

³ I had supposed this view was obsolete; but D. N. Mukherji, Indian Culture I, 1935, p. 477, has adduced some astronomical calculations in its support.

c10.s7.8

⁴ In CII. A specific argument against this date, breaking fresh ground, is given by L. Bachhofer, Ostasiat. Z. XVI, 1930, p. 9.

c10.s7.9

Little remained in India which could in any sense be called distinctively Greek after about the middle of the first century A.D., or say roughly the establishment of the Kushan empire which replaced in the North-West the rule of the Parthian Gondophares and his successor Pacores. Such cities as were Greek poleis—they were very few—lived on under Saca and Parthian rule in the same sort of quasi-autonomous relationship to their Iranian overlords as can be traced in the relationship of Seleuceia to the Arsacid monarchs; the unique coin of Pushkalāvatī (p. 336) shows that that city, like Seleuceia for a time in the first century A.D., was at some period completely independent. The number of representations of the Fortune of some city (i.e. the city itself) with her mural crown on Saca coins¹ tells its own story; most illuminating is a coin of Zeionises,² nephew and satrap of the Parthian Gondophares (A.D. 19 to 45 plus), which shows the Fortune of some city, undoubtedly Pushkalāvatī, confronting the king as a separate and obviously quasi independent entity. Literature gives little help here. The Periplus in the middle of the first century A.D. mentions Bucephala,³ which probably, though not necessarily, means that it was still surviving as a city; the second part of the Milindapañha twice mentions Alexandria of the Caucasus (p. 421 n. 4), but that document may not be later than the end of Greek rule, and in any case the name may simply have been borrowed from the first part. Fortunes of cities never appear on the Kushan coinage, and there can be little doubt that the cities, if not destroyed, had by the end of the first century A.D. become to all intents and purposes Indian, though that does not mean that some people may not still have understood the Greek language. This general result is borne out by the Greek monograms on the coins, for as long as they persist they show that Greek moneyers were still operating or controlling the city mints. On the coins of Azes and his dynasty Greek monograms are so numerous and varied that it seems as if Azes must have opened some fresh mints. They occur on the coins of Gondophares (who was still reigning in A.D. 45) and of his nephew and sub-king Abdagases, but not on those of Gondophares' successor Pacores, or later; that is, we meet no Greek moneyers, as we meet no Greek cities, after about the middle of the first century A.D.

c10.s7.10

¹ Maues: *BMC* pp. 68 no. 4 and 70 nos. 12, 13, Pl. XVI, 3, 9; *NC* 1890 pp. 131 no. 7 and 134 no. 21; *Lahore Cat.* p. 99 no. 15; *ASI* 1914–15 p. 46 no. 12; *NC* 1923 p. 340 no. 66. Azes: *BMC* pp. 82 nos. 109–11 and 90 nos. 191–9, Pls. XVIII, 10, 11, XX, 1. Azilises: *ib.* p. 94, Pl. XX, 9–11; Cunningham *NC* 1890 Pl. VIII, 6; *ASI* 1914–15 p. 31 no. 26; an unpublished type *ASI ib.* no. 24; and see *CHI* p. 588. Whether the seated figure on a Saca coin, Cunningham, *NC* 1890 p. 154 no. 6, be really the Fortune of a city may be doubted; it is difficult to see the turreted crown on the plate.

c10.s7.11

² *BMC* p. 110 nos. 1, 2, Pl. XXIII, 4. An inscription, *CII* p. 81 no. xxx, shows that Zeionises (Jihonika) also ruled in Taxila; but Taxila was not a *polis*, and the city on his coins is certainly Pushkalāvatī, cf. *CHI* pp. 582 n., 588.

c10.s7.12

³ *Periplus* 47. The writer, who knew nothing about Northern India, wrongly put it in Προκλαίς, which here and in 48 is obviously not Pushkalāvatī, as it is in Ptol. VII, 1, 44, but the province, Peucelaïtis.

c10.s7.13

The question of whether Greek remained a living tongue in India as late as the reign of Kanishka was argued at a symposium of the Royal Asiatic Society in 1913, and the supporters of the negative view, led by Professor F. W. Thomas, had the best of the argument.¹ I am not going over that ground again; but several things are clear to-day which were unknown or not clear in 1913, and they are worth looking at. I must premise that the *sh* sign, *p*, on coins of Kanishka and his successors has nothing to do with it; that sign had appeared long before on the coins of Miaos and Spalirises (App. 17, 18). Greek was alive in the time of Kujula Kadphises; the square letter-forms and the different mistakes in the legend on his coins prove that much. But the coins of Wima Kadphises exhibit correct Greek written in the orthodox way except for the lunate sigma; *prima facie*, his moneyers were copying a dead language. Are there any signs of life? The significance of the verbal form τυραννοῦντος on the coins of Miaos, contemporary with Hermaeus, is noticed elsewhere (App. 17), and on the coins of Gondophares’ sub-king Abdagases the verbal form occurs again as βασιλεὺς βασιλευόν, ‘under the rule of king Abdagases’;² and as the use of βασιλεύοντος on the pedigree coins of Antimachus and Agathocles (App. 3) shows that it was regarded as the proper word to express the position of a sub-king, it follows that in Gondophares’ reign there were Greeks who still understood very well the distinction between βασιλεύς and βασιλεύων.³ This in itself takes us no further down than the Fortune of a city on Zeionises’ coins has already done; but βασιλεύων occurs again in the regular legend on the widespread coinage of the ‘Nameless king’. ⁴ This shows that to some of his moneyers Greek was still alive, and also shows that he was a sub-king of some sort; and as the Hou-han-shu says that Wima Kadphises governed his Indian possessions through a viceroy, and as it has usually, though not universally, been supposed that that viceroy can only have been the Nameless king,⁵ this may take us down nearly to the reign of Kanishka.

c10.s7.14

¹ Reported at length in *JRAS* 1913 pp. 627 *sqq*., 922 *sqq*.

c10.s7.15

² *BMC* pp. 107–8.

c10.s7.16

³ In Philostratus’ *Life of Apollonius* II, 32, Gondophares himself is made an Indian who talks Greek and reads the *Heracleidae* of Euripides. Any Parthian king might have known Greek; but Philostratus’ story was merely invented for the sake of Apollonius’ remark, II, 33, about the ‘return of the Heracleidae’.

c10.s7.17

⁴ *BMC* pp. 114–16: βασιλεὺς βασιλεύων σωτὴρ μέγας.

c10.s7.18

⁵ J. Kennedy, *JRAS* 1913 p. 664; de la Vallée-Poussin p. 312; Konow in *CII* pp. lxviii–ix. Cunningham possessed a coin which connected the two, *NC* 1892 pp. 55, 71 no. 14. It is noteworthy that a gold coin of Wima is now known (*NC* 1934 p. 232, *BM Quarterly* VIII p. 73) which gives him the title βασιλεὺς βασιλέων σωτὴρ μέγας.

c10.s7.19

For Kanishka’s reign itself the question must turn on his Greek title βασιλεὺς βασιλέων Κανέρκου, ‘Kanishka king of kings’, where a nominative, βασιλεύς, is in agreement with a genitive Κανέρκου. To Professor Thomas in 1913 this appeared to be an unintelligent copying of a dead language;¹ Dr Sten Konow in 1928 compared it with an idiom from Chinese Turkestan,² which I understand obtained in more than one language there. This no doubt is a possible origin, especially if Kanishka, as some believe, came from Chinese Turkestan; but there does seem now to be a simpler explanation. Inscriptions from Doura have given us some knowledge of the sort of Greek, alive enough but vulgarised, spoken by the less educated classes in the decay of what had once been a Hellenistic city, and its most marked feature is the substitution of genitive for nominative and the use of the two in agreement.³ We may believe that in Kanishka’s time educated Greeks and educated Greek were alike extinct in India; but some of his moneyers could evidently still speak Greek in a vulgarised form, whatever else they spoke also—incidentally an argument for not dating this king too late. As the coins of the Kushan kings after Kujula Kadphises show no moneyers’ marks, neither Greek monograms nor Kharoshthi letters, a thing which must reflect royal orders, we are deprived of our best means of judging what Kanishka’s moneyers were. There were still people with Greek names, and the father of Menander the wrestler⁴ must have had *some* Greek feeling if he named his son after the great Yavana king. The ‘slave Agesilas’⁵ who was the architect of Kanishka’s stupa near Taxila and made his relic casket may have been anything from a skilled Greek slave imported from the West to a subject of Kanishka⁶ with little Greek about him but his name; as he worked in the Gandhāra style the latter is more probable. After Kanishka no question arises.

c10.s7.20

¹ *JRAS* 1913 pp. 1013–14.

c10.s7.21

² *CII* p. liii.

c10.s7.22

³ Cumont, *Fouilles* p. 351. His instances are confined to γυναικός for γυνή, but instances of other words from later finds are *SEG* VII, 683, 685, 688, 696–8, nominative and genitive in agreement; 374, 488, genitive for nominative. So on a coin of Gotarzes of Parthia, A.D. 41–51, *BMC Parthia* p. 165.

c10.s7.23

⁴ *CII* p. 134 no. LXX, late Kushan period, from Peshawur. It is remarkable that the name, Minamdra, keeps the Greek nu; but *l* and *n* could run parallel in Prakrit (p. 238 n. 2), and the *n* does not prevent the inscription being late.

c10.s7.24

⁵ *Ib.* p. 135 no. LXXII, Agisala.

c10.s7.25

⁶ If ‘slave’ be used either in the Persian (Achaemenid) sense of any subject not royal, or in the Graeco-Parthian sense of any subject not *āzāt*, a member of the privileged Parthian nobility (Rostovtzeff and Welles, *Yale Class. Stud.* II p. 52; H. W. Bailey, *BSOS* VI, 1932, p. 953), which in Greek was rendered by ἐλεύθερος; hence the strange idea in some Roman and modern writers that Surenas took an army of slaves against Crassus.

c10.s7.26

Greek legends cease to appear on the coins; Saca written in Greek letters continues till the reign of Vasudeva, and then even Greek script vanishes from India.

c10.s7.27

We can now return to the period of Greek rule and gather up what indications remain of what the Greeks in India did. The provincial organisation of the Greek empire in India at its fullest extent has already been examined (Chap. VI), and Indian citizens of one Greek *polis* at least, Demetrias in Sind, have been traced at a date which cannot be much later than 50 B.C. (p. 257 n. 3). The Kharoshthi letters on some of the Greek coins,¹ which from their relation to the Greek monograms are clearly the initials of an Indian moneyer, enable us to confirm these Indian citizens. These letters are for some time confined to the group of kings who ruled east of the Jhelum, the kings of the houses of Euthydemus and Menander, which for all purposes of policy were identical in fact as, except for Menander himself, they were in blood; Kharoshthi letters are common on the coins of Strato I, Zoilus II Soter, and Apollodotus II, and occur on those of Apollophanes and Dionysius, but are never found on coins of Menander and his contemporaries, *e.g.* Antimachus II, or for a little while after his death (Agathocleia and Zoilus I Dikaios); we may say roughly that they begin with the last quarter or last third of the second century B.C. They never occur on coins of any king known to be of the house of Eucratides, or on the coins of Lysias, who, though a Euthydemid, coined in Antialcidas’ mints (p. 315). Except on the late base silver of Strato I they are, prior to Hippostratus, confined to the copper currencies, and the usual rule is that they never appear alone but always in association with a Greek monogram. In the Seleucid mint at Antioch it was the rule for two moneyers to sign the coins, one being a permanent mint-master and the other a changing city magistrate.² I do not know whether an examination of every coin with a Kharoshthi initial on it (which I cannot make) would show if these Indians were ever magistrates; but there is one simple case in which the Kharoshthi letter certainly represents a permanent official of the mint, and that is the ‘city’ coins of Hippostratus, minted in Bucephala (p. 245), a Greek *polis*. All his ‘city’ type coins, both bronze and silver, bear the same very distinctive Greek monogram and the same Kharoshthi letter; these were his two mint-masters at Bucephala. It is known from Plutarch that Perseus had two mint-masters at Pella (p. 437), and I have traced the two in Bactra in the great period (App. 1); what is not known is whether, when there were two, they were colleagues with equal authority, or whether one was subordinate; this in Hippostratus’ mint would be the Indian. The latter is more probable, for we see the Indian moneyer at first signing copper money only before being at the end promoted to signing the silver also. But in one mint we find an Indian mint-master already in control by about the beginning of the first century B.C.: on some late silver of Strato, called semi-barbarous, and on some bronze of Strato, Zoilus II, and Apollodotus II, there is no Greek monogram at all but only two Kharoshthi letters, showing that the mint in question had two Indian mint-masters; the mint in question was probably Sāgala, which, unlike Bucephala, was not a Greek *polis*. The fact that Kharoshthi letters are comparatively rare on coins of Maues¹ is one more sign of his early date, for throughout the Saca and Parthian (Pahlava) coinages from Azes to Pacores such letters are common. Very rarely, and very late, one Kharoshthi letter appears alone; this occurs only on some ‘uncertain’ bronze coins, which I have thought might belong to kinglets driven into the hills (p. 323 n. 1), and on the second series of the ‘Hermaeus’ coins of Kujula Kadphises (App. 17); he was an invader issuing proclamations, and had to get his propaganda money struck as best he could.

c10.s7.28

¹ *BMC* under the various kings here mentioned; cf. Whitehead, *NC* p. 314. On the monograms generally see App. 1.

c10.s7.29

² E. T. Newell, *The Seleucid Mint of Antioch* 1918 passim.

c10.s7.30

¹ Instances are Cunningham *NC* 1890 pp. 133 no. 19, 135 no. 24.

c10.s7.31

That these Indian mint-officials who signed the coinage were citizens where they occur in a Greek *polis* is certain, for they could not possibly have been anything else. And as it cannot be supposed that Sāgala under Strato or Bucephala under Hippostratus was incapable of supplying two Greeks to manage the mint, we must see in this admission of Indians to office a deliberate act of policy. The evidence of the coins takes us a good deal further back than that of the cave inscriptions, and no doubt by the last quarter of the second century the Euthydemids, whether from choice or by force of circumstances, were coming to rely more and more upon their Indian subjects; but the coins do not supply evidence for the time of Demetrius and Menander. Seeing, however, that the rule of these two kings was based on the idea of a sort of partnership, Indians must from the first have been admitted to citizenship; indeed it is not apparent on what other lines a place like Demetrias in Sind could ever have been a *polis*, as it certainly was from its dynastic name. The kings of Eucratides’ line, like Eucratides himself, may have held different views; we do not know. But the only king west of the Jhelum on whose coins Kharoshthi letters occur, and the only king therefore to admit Indians to office, was Hermaeus, and it is quite uncertain if he did belong to the house of Eucratides; in any case he may have turned to his Indian subjects, as at the end he turned to China, in an attempt to stave off the inevitable.

c10.s7.32

That is as far as one can get with the cities, though it seems obvious that if there were Indian officials of the mints they were not likely to be the only officials who were Indians; it has already been pointed out that the great loosely hung satrapies could only have been worked by the retention of much of the existing Indian arrangements. The one Indian governor mentioned, an unnamed meridarch, seems to belong to the early Saca period;¹ but undoubtedly it must be supposed that, while the Seleucids occasionally fitted an Iranian official into their own Greek system, the Euthydemid provincial administration in India was essentially Graeco-Indian. Still it must not be forgotten that the undivided empire of Demetrius and Menander was in form a Hellenistic kingdom, one that had originally branched off from the Seleucid kingdom though it had taken a somewhat different complexion; and the Greek satrapies, with their *strategoi* and meridarchs, became so firmly impressed on the country that the Sacas took over the Greek provincial organisation with even less change than the Parthians made when they took over the Seleucid, while the four satrapies into which the Greeks had divided Sind lasted, with whatever boundary modifications, far down into Indian history as the four traditional divisions of that land. Other things too, for a time, were firmly impressed by the Greeks upon that part of India which they ruled. The bilingual coinage, with legends in Greek and Kharoshthi, was continued by Sacas, Parthians, and the earlier Kushans, though with the passing of Greek rule its meaning and significance had passed also; the Kharoshthi legends were first discontinued by Kanishka.²

c10.s7.33

¹ *CII* p. 4 no. 11, from Taxila, dated by Konow about the second half of the first century B.C. It reads: ‘By... the meridarch, together with his wife, the stupa was established in honour of his mother and father, for the presentation of a respectful offering.’ The name is lost; but it is said that Indian Buddhist texts always say ‘women and men’, putting the woman first (Rhys Davids 1 p. 83 n., 11 p. 127 n. 1), and in fact the phrase ‘in honour of mother and father’ is very common in Kharoshthi inscriptions, *CII* nos. xx, xxiii, xxvii, xxxvi no. 6, lxxxv, lxxxvi; also mother comes before father in other phrasing, *ib.* nos. xiii, xxxi, xcii. This makes this Buddhist meridarch an Indian, for it is incredible that a Greek would have named his mother before his father.

c10.s7.34

² A solitary bilingual coin of Kanishka is known, *BMC* p. 175 no. 1.

c10.s7.35

Another thing was the Greek calendar. It has already been mentioned (p. 64) how that great invention the Seleucid Era, model for the Parthian and for the Roman provincial Eras, swept Asia west of India; and it is certain that the half-Seleucid Euthydemids used it in Bactria, for they could not possibly have used anything else, in view of the trouble they took to proclaim that they were Seleucids; as to the house of Eucratides, Plato’s dated coin is definite proof (p. 209). Demetrius of necessity took the Seleucid calendar (in its Macedonian, not its Babylonian or semi-Babylonian, form)¹ to India with him, and there it gave birth to many other Eras;² kings or dynasties of alien blood might desire to set up Eras of their own, but they were all copies; the *idea* of reckoning time from a date fixed once for all came to India with the Greeks. Sacas and Kushans copied so closely that nothing differed but the initial year, and their calendars have been called Indo-Macedonian;³ they kept the same subdivisions, the Macedonian months, and the Macedonian names of these months in place of the Indian names often appear in the datings of Kharoshthi inscriptions of the Saca and Kushan periods.⁴ Indeed, though Brahmi inscriptions normally use the Indian month-names, a Sanscrit inscription recording a donation is now known, written in Brahmi and found as far afield as Mathurā, which is dated in the reign of the Kushan Huvishka in the 28th year of the Era of Kanishka on the first day of the Macedonian month Gorpiaios.⁵ It has however been suggested that the donor must have come to Mathurā from the North-West.⁶

c10.s7.36

¹ I must refer back to pp. 64 *sq*. It should be obvious now, apart from the evidence given p. 47 n. 2, that the Euthydemids could only have used the same form as the Seleucids themselves.

c10.s7.37

² Konow’s discussion of the Eras, *CII* pp. lxxxii *sqq*., overlooks this obvious fact.

c10.s7.38

³ Konow, *CII* p. lxxxix.

c10.s7.39

⁴ *CII* nos. XIII, LXXIV, LXXIX, LXXXII, LXXXVI; one Saca period, four Kushan.

c10.s7.40

⁵ Jayaswal, *JBORS* XVIII, 1932, p. 4; Konow, *Ep. Ind.* XXI, 1931 (pub. 1934), p. 55; H. K. Deb, *IHQ* 1932 p. 117, who points out that the Indian and Macedonian months were readily interchangeable, both being lunar. The Macedonian months could as well be fitted to the Indian year as they were to the Roman year at Nineveh, *CIG* 4672.

c10.s7.41

⁶ Konow, *IHQ* IX, 1933, p. 147.

c10.s7.42

The one town in the Greek sphere which has been excavated, Taxila, was unfortunately not a Greek *polis*. Hellenistic Taxila has already been noticed (p. 179); but the town on Sirkap now so well known from the excavations is essentially the Parthian town of Gondophares, one of his capitals, and the bulk of the jewellery and small objects found, which are very numerous, illustrate Parthian culture in the middle of the first century A.D.¹ Sir John Marshall’s comparisons of the details revealed by excavation with the account of Taxila in Philostratus’ Life of Apollonius² have shown that Philostratus, however much he may have romanced in other ways, had before him a tolerably accurate account of the external aspect of Taxila itself, given by some traveller who visited it in Gondophares’ reign; and though the ground-plan of Hellenistic Taxila was fairly regular, Philostratus remarks on the narrow and irregular streets of the Parthian town, which are compared to those of Athens.³ The palace, a plain unadorned building in the middle of the town, probably Śaca or Parthian, is frankly an Assyrian palace; it has been compared to Sargon’s palace at Khorsabad.⁴

c10.s7.43

¹ *ASI* 1929–30 p. 57.

c10.s7.44

² *ASI* 1912–13 pp. 8, 30, 34, 37 and part I p. 13; the passages cited are from Philostratus II, 20, 23, 25.

c10.s7.45

³ *ASI* 1912–13 p. 8; Philostratus II, 23, φασὶ δ’ ὡς ἀτάκτως τε καὶ ’Αττικῶς τοὺς στενωποὺς τέτμηται. The narrow winding streets of Athens, never remodelled, were long famous in a world which had adopted Hellenistic town-planning.

c10.s7.46

⁴ Sir J. Marshall, *ASI* 1913–14 part I p. 16; *A Guide to Taxila* pp. 68–70.

c10.s7.47

The one definitely Greek public building which has been discovered did not belong to the Hellenistic city. It is a little Ionic temple of the Parthian period at Jandial,⁵ which in form is a Greek temple; the column drums were ground together in the Greek way, but it was covered with the common shell stucco and built on to it was a square tower; it is thought that this may have been a Tower of Silence and that the temple was built by a Greek architect for the Iranian community. There is a square Parthian tower also at Tel Umar (Seleuceia on the Tigris),⁶ the so-called ‘Ziggurat of Opis’; what the relation, if any, between the two may be I do not know. Jandial attests the presence, shortly after the Christian Era, of a Greek or semi-Greek architect who still understood Greek methods of work; this naturally suggests the former existence of Greek architecture in Bactria, but it does not follow, and nothing can be known till excavation be possible in that country. But one other building at Taxila reinforces my conclusion that *c.* A.D. 50 is the period after which we cease to find anything which can really be called Greek. A little Buddhist building of the late first century A.D. (early Kushan period), inside the Kunāla stupa, shows that its architect had heard of, and had tried to apply, the Greek principle of *entasis*, that is, giving a slightly convex curve to parts of the building instead of a flat line in order to correct certain optical illusions; but he did not really understand the matter, and though he knew that there ought to be curves he has made his own curves concave instead of convex, thus increasing the optical illusion which the curves were intended to correct.¹ In other words, understanding of this principle of Greek architecture was dead; all that remained was the knowledge that there once had been some principle.

c10.s7.48

⁵ *ASI* 1912–13 pp. 35–9; 1913–14 part I pp. 16–17; 1914–15 part I p. 14. Patches of the stucco, λίθου κογχυλιάτου of Philostratus, still remained. Some remarks on Jandial in Benjamin Rowland Jr., *AJA* xxxix, 1935, pp. 493–6; he thinks Ionic capitals disappeared from India with the Greeks, to be replaced by Roman Corinthian. His chronology however is not always correct.

c10.s7.49

⁶ L. Waterman in *Second Preliminary Report upon the Excavations at Tel Umar* pp. 75–8, cf. Preface p. vi.

c10.s7.50

¹ Sir J. Marshall, *ASI* 1914–15 part I pp. 13 *sq.*

c10.s7.51

In addition, so far as Greeks go, Taxila has supplied a few pieces of Greek sculpture, the best known being the head of a Dionysus and the child with its finger to its lips;² various small objects of art and also seals, some with Greek engraving and two or three bearing Greek names written in Kharoshthi (p. 389); and a great number of coins, including new ones of the Parthian and Kushan periods,³ sometimes of value for chronology because of the stratification of the finds. But the great value of the excavations to the Greek historian lies in their revelation of the fact that the capital which Demetrius built for himself was an Indian city and not a Greek one (p. 179).

c10.s7.52

² *ASI* 1912–13, Pl. xx; *CHI* Pls. xxxi, xxxii. Often figured.

c10.s7.53

³ Published year by year in *ASI*; many are noticed throughout this book. On one group see *CHI* pp. 580–1.

c10.s7.54

Something must now be said about trade. Trade between India and the West did not start in the reign of Augustus;⁴ a considerable trade had been going on since the Greek conquest, and though it is only now and then that details can be obtained, the main fact is unquestionable. Certain things have already been noticed; the new outburst of prosperity at Seleuceia, which coincided almost exactly in date with the great period of Greek rule in India from Demetrius’ conquest to the death of Menander, is the most notable (p. 261), but almost equally so are the enormous quantities of Indian ivory and spices exhibited by Antiochus IV in his triumph at Daphne in 166.¹ Parthia grew wealthy on the land trade which poured through her, and towards the end of the second century B.C. Chang-k’ien, or rather perhaps Ssu-ma Ch’ien (p. 281), referred to the long distances both by land and water which the merchants of that empire (i.e. Parthian subjects, whatever their nationality) were accustomed to travel.² Even towards the end of Seleucid rule in Syria, when politically the kings appear to us to have been little but leaders of mercenaries and northern Syria to have been in a state not far removed from anarchy, trade between Syria and the East still went on, as is attested by the number of late Seleucid coins found in the hoards from Mandali,³ east of the Tigris, and from Teheran;⁴ but the commercial revival already noticed under the last Greek rulers Hippostratus and Hermaeus was connected with the establishment by Rome in 64 B.C. of settled rule in Syria, for Crassus’ invasion of Parthia in 53 was too local and transitory to disturb the main trade route from Seleuceia to Syria for more than a year or two.

c10.s7.55

⁴ For the trade between the Roman Empire and India, which is a separate matter, see Warmington (the fullest account); also M. P. Charlesworth, *Trade routes and commerce of the Roman Empire* Chaps. iv, vi.

c10.s7.56

¹ Polyb. xxx, 25, 12, 800 tusks; ib. 26, 1-2, spices, of which the cinnamon and nard anyhow came from India.

c10.s7.57

² Hirth p. 97 §35.

c10.s7.58

³ E. T. Newell, NC 1924 p. 141.

c10.s7.59

⁴ M. Dayet, Aréthuse 1925 p. 131.

c10.s7.60

These facts suggest that what was important to the Greeks of Bactria and India was trade with the Seleucid empire through Seleuceia, for there do not appear to be any corresponding facts with regard to Egypt. This trade was essentially a trade with Seleuceia, which passed the goods on overland to Syria and the Phoenician ports; the great city which had taken the place of Babylon was clearing-house and middleman by whatever route goods came from the Far East, whether by the main land road which ran Bactra-Hecatompylos-Ecbatana-Seleuceia, or by the southern land route through Seistan to the Gulf of Ormuz (App. 12) and so up the Persian Gulf, or directly by sea from Barygaza and Demetrias-Patala to the Gulf of Ormuz (App. 12). More will be known about the trade between India and Seleuceia if the Greek centre and mart on that Gulf, Harmozia-Omana, the ‘harbour of the Macedonians’, be ever located and excavated; its merchants were certainly important as middlemen, for Seleucid coins have rarely been found in India, and those apparently only certain pieces of Antiochus I and II.⁵

c10.s7.61

⁵ Sir G. Macdonald, JHS xxiii, 1903, p. 108; S. K. Chakrabortty, IHQ xi, 1935, p. 244, who argues that certain specimens lighter than the Attic standard were struck on an Indian standard for trade purposes; sed quaere.

c10.s7.62

Some of the Hellenistic information in Ptolemy with regard to India and the Farther East generally, which I have attempted to separate from that part of his information which dates from the Roman empire, was ultimately derived from the itineraries of merchants who went thither from the West, and the amount he knew witnesses to the number of his informants, that is, to a very lively trade intercourse. I may add that if an embassy from India came to Augustus as early as 26 B.C.,¹ rulers in India must for some time have been accustomed to follow the course of events in the Mediterranean world.

c10.s7.63

¹ Warmington p. 36.

c10.s7.64

Of the spread of trade eastward in the Greek period we do not get too many indications. With the nomad conquest of Bactria and the loss of Bactra, whose equipment of markets was duly recorded by Chang-k'ien,² the Greeks lost control of the half-way house where met the land routes from India and China and the trade was gathered up, though it is evident from the coinages of Hippostratus and Hermaeus that the Yueh-chi no more interfered with trade with the West than did the Parthians at this time. But the loss of Bactria meant that when in 106 B.C.³ the Chinese sent their first caravan through from China to Parthia *via* Bactra, the Greeks, now confined to India, had no part in this through trade or anything to do with the import of Chinese silk which grew up along what was to be known as the Silk Route. The most definite things which can be traced to the time when Greeks still held Bactria are the import of nickel from China at the end of the reign of Euthydemus (p. 87), an Alexandrian glass vase with a head of Athena of the second century B.C. found in Honan,⁴ and Bactrian and Syrian cloth, perhaps as early as the late second century, found by the Koslov expedition at Urga in Mongolia,⁵ which had travelled along the route to Mongolia of which something has already been said (p. 109); whether it was at this time or later that the Greek word diadem reached Mongolian⁶ through Sogdian I do not know, nor do I know the date of the famous white bronze mirror of the Han period with Graeco-Bactrian designs.⁷ The failure of the people of Ferghana to get into touch with China and their anxiety to do so, recorded by Chang-k’ien in 128 B.C. (p. 87), may show that the difficulties had been very considerable till China opened up the way; and the things which China received from the Graeco-Iranian world—the pomegranate and other ‘Chang-k’ien’ plants (p. 311), the heavy equipment of the cataphract,¹ the traces of Greek influence on Han art²—are probably all later than 106 B.C., as were her own exports of silk, furs,³ and the high-class iron which came to Parthia through Merv;⁴ and if China got the vine (p. 311) and the great war-horse (p. 309) earlier, it was still not till after the fall of the Greek kingdom of Bactria; the Greeks lost Bactria too soon for the Chinese trade. Doubtless in the first century B.C. the Greeks in India imported for themselves silk and other articles from China, for Chinese jade⁵ and a Japanese scallop-shell⁶ (the earliest known indication that Japan existed) were found at Taxila, and later we hear of a regular trade route from Bactra to Barygaza;⁷ but what the ancient world really found lucrative was transit trade,⁸ and it cannot be supposed that silk, instead of going to Seleuceia, would travel all the way to Barygaza to be shipped to the West till direct voyages to Egypt became usual, and by that time the Greeks had lost Barygaza and their Indian ports.

c10.s7.65

² Hirth p. 98 § 51.

c10.s7.66

³ Hirth's date, p. 133.

c10.s7.67

⁴ M. Rostovtzeff, *Social and Economic History of the Roman Empire* p. 513.

c10.s7.68

⁵ W. P. Yetts, *Burlington Magazine* 1926 p. 168; M. Rostovtzeff, *Mon. Piot* XXVIII, 1925-6, pp. 171-2 (some Bactrian, some Syrian, end of second or first century B.C.), and *The Animal Style in South Russia and China*, 1929, Pl. XXIV A and p. 85.

c10.s7.69

⁶ B. Laufer, *Sino-Iranica* p. 573.

c10.s7.70

⁷ In the Victoria and Albert Museum. See S. W. Bushell, *Chinese Art* 1,² 1910, fig. 60.

c10.s7.71

¹ Laufer, *Chinese clay figures* 1914 p. 217; M. Rostovtzeff, *Mon. Piot* XXVI, 1923, p. 135; *The Animal Style* p. 107.

c10.s7.72

² Hirth, *Ueber fremde Einflüsse in der chinesischen Kunst* 1896; Rostovtzeff, *Mon. Piot* XXVIII, 1925-6, pp. 171-3 and references, *Inlaid bronzes of the Han dynasty* 1922 pp. 58 sqq., who thinks (*L’art gréco-iranien*, Rev. des Arts Asiatiques 1933 p. 218) that it was exercised through that form of Graeco-Indian art which he calls *Greco-Sakian*, *Seminarium Kondakovianum* 1933 p. 169. Contra, O. Fischer, *Die chinesische Malerei der Han-dynastie* 1931 (not seen): Han art purely indigenous.

c10.s7.73

³ Pliny XXXIV, 145; *Periplus* 39.

c10.s7.74

⁴ Seric iron was called Margian in 53 B.C. because it reached Parthia through Merv, Orosius VI, 13, 2 combined with Plut. *Crassus* 24. The widespread theory that Seric iron came from the Cheras of southern India (Warmington pp. 157 sq.) cannot stand against this fact. Pliny XXXIX, 15 says that Seric iron was better than that produced in Parthia.

c10.s7.75

⁵ *ASI* 1919-20 part I p. 19.

c10.s7.76

⁶ *Ib.* 1926-7 p. 118 no. 14.

c10.s7.77

⁷ *Periplus* 48, 64.

c10.s7.78

⁸ Tarn, *Hell. Civ.*² p. 220.

c10.s7.79

Two slight indications may exist of communication, not of course direct, between the Greeks of India and the Indo-Chinese peninsula. One, already noticed, is the occurrence of Menander’s name in Indo-Chinese tradition (p. 267); the other is an object dug up at Taxila (Sirkap), the head and shoulders of a gorilla-like figure.⁹ Sir John Marshall has pointed out that it is hollow like many Hellenistic statuettes, and that the modelling is much superior to that of objects from the Saca-Parthian and Kushan strata; and as it came from one of the Hellenistic strata it is probably Greek work. No Greek in India can have known the gorilla, but the figure certainly suggests that the artist had seen, or had had a description of, an orang; and if a polar bear reached Alexandria,¹ an orang may well have come to Taxila. In addition, the Greek word for mustard, σίναπι, is supposed to have come from Farther India,² but that has nothing to do with the period under consideration, for the word was known at Athens in the fifth century B.C.;³ though how it got there is a mystery, as it is said not to occur in Persian.

c10.s7.80

⁹ *ASI* 1929-30 p. 91 no. 111.

c10.s7.81

¹ Athen. v, 201 c, ἄρκτος λευκὴ μεγάλη μία.

c10.s7.82

² J. Przyluski and C. Régamy, *BSOS* VIII, 1936, p. 703.

c10.s7.83

³ Aristoph. *Eq.* 631, in the form νᾶπυ; see Athen. ix, 367 a.

c10.s7.84

What India, whether Greek or Indian, did try to do was to get into touch with China through Khotan, independently of Bactria. We have seen that the Greeks were trying to do this in the reign of Hermaeus (p. 338); in the latter half of the first century B.C. Indian-speaking traders had got as far as the Chinese limes,⁴ and in the first three centuries of the Christian era there appear to have been settlements of Indian or Indian-speaking traders in Chinese Turkestan, more especially in the Khotan district,⁵ with a view to the Chinese trade; but the communications of that country with India, whether with Gandhāra or Kashmir, were across difficult passes, and it is not known how much it really means. The Greek words found in documents of this period from the Khotan country have already been considered (pp. 85 *sq.*); but to these settlements must belong the clay sealings found at Niya and other places near Khotan⁶ which occasionally show Greek or Graeco-Roman types, sometimes on the same documents as Chinese sealings; a number of clay sealings of the sort, once also attached to documents, have been found in India at Besnagar.⁷ Sir Aurel Stein said comprehensively in 1907 that all these classical seals from the Khotan country were Roman work,⁸ of the third or at earliest the second century A.D., and certainly most of them came from the Roman empire; some correspond to types on Roman Imperial coins, as the supposed Eros to the *Genius populi Romani* on coins of Diocletian,¹ or Ganymede feeding the eagle of Zeus to a similar type on coins of Marcus Aurelius and Commodus.² But it is also possible that this or that one might have come from Seleuceia, or those with a Kharoshthi legend, like the Athena Promachos,³ from India, or some might even be copies made locally; dating by seals is rather delusive work, for a signet ring might pass from father to son for generations. It is certain however that the Turkestan seals are essentially later than the period treated in this book, and I have only mentioned them to bring out that point, since they are rather famous.

c10.s7.85

⁴ Sir A. Stein, *Asia Major* (Hirth Anniv. volume) 1923 pp. 367, 372.

c10.s7.86

⁵ De la Vallée-Poussin pp. 321-2 and references.

c10.s7.87

⁶ Published by Sir A. Stein in *Ancient Khotan* 1907, and *Serindia* 1921.

c10.s7.88

⁷ *ASI* 1914-15 part 1 p. 20; see D. R. Bhandarkar, *ib.* 1914-15 p. 77.

c10.s7.89

⁸ *Ancient Khotan* p. 357.

c10.s7.90

¹ L. C. Woolley in *Serindia* I p. 216; figured Pls. XX, XXVII (N. XIII. ii. 10).

c10.s7.91

² Woolley in *Serindia* III pp. 1250, 1260; Pl. V (F. VII. 002).

c10.s7.92

³ *Serindia* I p. 258, Pl. XX; frontispiece to *Ancient Khotan*. Several impressions of her were first given by Stein in *Sandburied ruins of Khotan* 1903. She is Greek enough.

c10.s7.93

With trade between India and the West⁴ in the period of Greek rule we reach firmer ground. Earlier times need not detain us here. The export from India of ebony and other woods,⁵ of peacocks, and doubtless of some spices, was very old, and in the third century under the Mauryas there must have been a regular export trade to the Greek West, though it cannot be said how large it was: Indian ivory continued to reach the Aegean till Ptolemy II threw enough African ivory on the market to break the price and secure the trade for himself;⁶ Indian spices came to Egypt; Ptolemy II exhibited Indian dogs and cattle in his triumph.⁷ So long as Egypt held the Phoenician ports the overland commerce from the East through Seleuceia partly enured for her benefit; to this period may belong the epitaph, once seen at Harūnabad near Kermanshah in Media, of a Greek merchant from Samaria in the Fayûm who was working the land-route.⁸ But what needs considering here is the increase of trade after Northern India passed under Greek rule, and especially certain things which happened round about 100 B.C.; the caravan which came through from China in 106 B.C. was not the only phenomenon of importance at the turn of the century. One must first get clear about the sea-route.

c10.s7.94

⁴ Most of the literature on the Indian trade in the Hellenistic period relates to Egypt. See M. Chwostow, *History of the eastern trade of Graeco-Roman Egypt* 1907 (in Russian; results in M. Rostovtzeff, *Arch. f. Pap.* IV, 1908, p. 298); H. Kortenbeutel, *Der ägyptische Sud- und Osthandel in der Politik der Ptolemäer und römischen Kaiser* 1931; Rostovtzeff, *Foreign Commerce of Ptolemaic Egypt*, *Journ. of Econ. and Business Hist.* IV, 1932, p. 728 (largely the Arabian part); E. Leider, *Der Handel von Alexandreia* 1933 pp. 57 sqq.

c10.s7.95

⁵ Warmington p. 213.

c10.s7.96

⁶ Tarn, *Hell. Civ.*² p. 226 (references).

c10.s7.97

⁷ Callixenus in Athen. V, 201 b, c.

c10.s7.98

⁸ Fr. Sarre and E. Herzfeld, *Iranische Felsreliefs* 1910 p. 226, reproduced from an old copy, which, though very imperfect, shows the earlier form of sigma, Σ.

c10.s7.99

Indian vessels were coasting along Gedrosia to Arabia and the Persian Gulf before the time of Alexander, for Nearchus got a guide in Gedrosia who knew the coast as far as the Gulf of Ormuz;¹ but from those points Arab peoples had a monopoly.² The chief Indian export for some time must have been spices, which Arabs distributed to the western world along with their own spices; the Arabs of Gerrha on the Persian Gulf supplied Seleuceia, so far as the trade did not come overland, and after the Ptolemies lost their Phoenician ports at the end of the third century they were thrown back for spices exclusively upon their Arab middlemen—Sabaeans, Minaeans, Lihyanites; Alexandria knew so little about the trade that even in the second century B.C. she believed that cinnamon came from Arabia or Somaliland. Egyptian records testify to the importance of the spice trade at this time, and when the Greek Eudoxus of Cyzicus made the first (Greek) through-voyage from Egypt to India c. 120 B.C. (p. 370) it was with spices and precious stones that he loaded his ship for the return journey;³ and Demetrius may for a time have held the spice-land of eastern Gedrosia (p. 94). But the wealth of Antiochus IV in Indian products shows not only that the Seleucids, after they acquired all Phoenicia, probably began to gain on the Ptolemies, but that there was now a large increase in the ivory trade: the 800 elephant tusks from India which that monarch exhibited in his triumph at Daphne were presumably meant to overtrump, in the eyes of the world, the 600 African tusks which had been carried in the triumph of Ptolemy II.⁴ The position in Menander’s reign has already been described (pp. 260 sq.): ships from the Greek ports in India were following Nearchus’ route along the Gedrosian coast to the Greek centre on the Gulf of Ormuz (App. 12), whence the goods went by water to Seleuceia; what is not known is when the Greek centre on that gulf began to compete with Gerrha as intermediary between India and Seleuceia,⁵ and how Gerrha, which was badly placed, stood the competition. This route, in which ships from Barygaza hugged the coast the whole way, calling at Demetrias-Patala, is the first of the four stages described by Pliny. It is now necessary to date those stages, for the second and third have usually, in defiance of Pliny’s own language, been put far too late; and as I shall need these stages to explain the Yavana merchants of the cave inscriptions, I must not at this point use those inscriptions.

c10.s7.100

¹ Arr. Ind. 27, 1.

c10.s7.101

² For what follows see generally Tarn, Hell. Civ.² pp. 213 sqq. and references.

c10.s7.102

³ Poseidonius in Strabo 11, 98, ἀρώματα καὶ λίθους πολυτελεῖς.

c10.s7.103

⁴ Callixenus in Athen. v, 201 a.

c10.s7.104

⁵ There may already in Persian times have been a centre on this gulf intermediate between India and Babylonia, for Nearchus changed guides there, his second guide being a Persian, Arr. Ind. 37, 2.

c10.s7.105

Pliny’s four stages¹ are stages of the voyage from Egypt to India, the gradual utilisation of the south-west monsoon; no information remains as to the use of the north-east monsoon for the voyage westward, but I shall assume that when ships began to cross open sea on the way to India they soon learnt to cross it on the way back also. As Nearchus was told of, and waited for, the north-east monsoon,² Indians had known of it long before Greeks, though they had only used it for coasting voyages;³ had they learnt to cross the open sea, Greeks in the second century must have copied them. Pliny’s four stages are (1) coasting all the way;⁴ (2) cutting across a bit of open sea from a point in Arabia which he calls Syagros to Demetrias-Patala;⁵ (3) cutting out Demetrias (this is the important matter) by sailing direct from the Arabian coast (and *a fortiori* from the Persian Gulf) to a port he calls Sigerus, somewhere south of Barygaza (and naturally therefore to Barygaza also);⁶ (4) cutting out Arabia altogether and sailing direct from the Somali coast to Muziris and other ports in the extreme south of India.⁷ He gives certain time-indications. The fourth stage was in use when he wrote but had not been in use very long;¹ doubtless the usual date for its commencement, A.D. 40–50, is approximately correct. The second stage was very short, for he says that the next generation, *secuta aetas*, passed to the third; the second then did not last more than thirty years. By contrast, he says that the third stage, after the *secuta aetas* adopted it, lasted a long time, *diu*, and ‘a long time’, as specifically contrasted with a generation, cannot mean less than a century and might mean more; from the way he puts it, he is clearly describing what, to himself, had seemed to be the permanent way of the world till quite recently. Taking then a century for the third stage, which would make it begin *c.* 60–50 B.C., and thirty years for the second stage, we get from Pliny a date of *c.* 90–80 B.C. for the first use of the south-west monsoon (otherwise than for coasting). But we happen to know from other sources that *something* important connected with the Indian Ocean had happened before 78 B.C. and after *c.* 111 B.C., say *c.* 100–80 B.C., and this cannot well be anything but the first use of the south-west monsoon.² The two lines of evidence then agree as well as may be; that means that the third stage cannot have begun later than *c.* 50 B.C. and may have begun *c.* 70 or even a little earlier. It does not seem possible to put the beginning of the second and third stages in the Roman Imperial period.

c10.s7.106

¹ Clearly done in Warmington, pp. 45–6 and map; but in dating stages 2 and 3 he has not considered what Pliny says. See also (more briefly) E. Kornemann, *Janus* 1 (*Festschrift für C. F. Lehmann-Haupt*), 1921, p. 56, who puts the third stage in Augustus’ reign. The possible bearing of the cave inscriptions on the matter has not been considered.

c10.s7.107

² He left Patala when the ‘Etesian winds’, the S.W. monsoon, stopped, late September or early October, Arr. *Ind.* 21, 1. He waited a month in ‘Alexander’s harbour’ till he got the N.E. monsoon, November; Arr. *Anab.* VI, 21, 3, παρέμενε τὴν ὥραν τοῦ παράπλου, shows that he knew it was coming.

c10.s7.108

³ Evidence for Indian voyages, W. H. Schoff, *Periplus* pp. 228 sq.; but the use of a ‘shore-sighting bird’ does not prove ocean voyages. I am speaking only of Greeks and Indians at one particular period; Polynesians and Malays have crossed great expanses of ocean in their canoes both before and since.

First 107 of 340 passages. Read on

Linked texts (19)