Short URL for this page:

[image ALT: Much of my site will be useless to you if you've got the images turned off!]
Bill Thayer

[image ALT: Cliccare qui per una pagina di aiuto in Italiano.]

[Link to a series of help pages]
[Link to the next level up]
[Link to my homepage]

This webpage reproduces an article in the
Journal of the American Oriental Society
Vol. 37 (1917), p240‑249

The text is in the public domain.

This page has been carefully proofread
and I believe it to be free of errors.
If you find a mistake though,
please let me know!

 p240  Navigation to the Far East under the Roman Empire
Wilfred H. Schoff
Commercial Museum, Philadelphia

Communication between Mediterranean lands and the Far East, which had been growing in importance since establishment of political contact in the conquests of Alexander and the consequent opening of the overland caravan routes, became exceedingly active between the first and third centuries of the Christian era through the discovery of the periodicity of the trade winds and the opening of active maritime traffic. There was, however, among writers in the Roman world considerable confusion because of their assumption that the land and sea routes had the same destination. This confusion, due partly to primitive misconceptions of geography, was greatly enhanced by the surprising misinterpretation of reports of various travelers upon which Ptolemy based his geographical calculations.

In tracing the caravan route it is impossible to go far astray because of limitations imposed by mountains, deserts, and water-courses. Richthofen (China, 1.10) and others have followed the whole route between the Pamirs and Sera metropolis, which may quite surely be identified with the ancient Chinese capital Singan‑fu. This was the great trade route of the silk merchants, and that trade was already of importance in the second century before the Christian era. The sea route was opened first to the west coast of India and Ceylon, where contact was made with another sea route leading further east known to the natives of India as the 'golden route,'​1 and its eastern termini as the 'golden and silver islands' — whence silk was also obtained. Inland from these islands (or shores, either interpretation being possible), was a metropolis Sina Sinorum, known to the Roman world by hearsay only, and assumed to be identical with the earlier known Sera metropolis, so that both caravan and maritime routes were supposed to have reached the same trade centers.

 p241  Although the various ports of call along this ocean route have been reasonably identified by comparison of place names, consideration of sailing conditions may still yield matter of interest.

Graeco-Roman navigation in the Indian ocean, as we know from ample evidence from coins, painting, and sculpture, was carried on in open craft, felucca rigged, very similar to the Arab dhows of our own time. Sailing before the wind was preferred: considerable effort was required to hold a course with the wind abeam, and tacking against the wind was not attempted out of sight of land because neither log nor compass was available for calculating or holding a course. The trade winds, which blew from southwest to northeast between April and October, and in the reverse direction between October to April, had doubtless long been used by Arab and Dravidian mariners, but vessels from Egypt to reach these ocean winds depended also on the so‑called etesian winds blowing from north to south, and setting in usually about mid-summer, to carry them on the long journey down the Red Sea.​2 The time for their voyage was therefore very limited. We learn from Pliny (Hist. Nat. 6.26) that vessels set sail from Egypt at mid-summer and in about thirty days arrived at the Straits. There they met the trade wind, which they gave the name Hippalus from the first of their countrymen to discover its use, and after exchanging cargo, set sail for India and reached the Malabar coast in forty days. Owing to depredations of pirates, they were obliged to steer with the wind partly abeam so as to make a landfall at the Tamil ports near the southern extremity of India. There they again exchanged cargo and set sail on their return voyage in December, using the southeast trade wind, which, owing to climatic influence due to the Sahara and Arabian deserts, is projected up the Red Sea as a southerly wind. Only by adopting this sailing schedule could the return journey be made in the same season. The voyage from the protects of the Dravidian kingdoms and Ceylon to the Far East was undertaken under similar conditions, except that, since no secondary passage had to be made through a body of water like the Red Sea, the whole period of each monsoon could be utilized. It was easy for the Chola mariners to reach Farther India and the South China Sea with a longer period for exchange  p242 of cargoes and to return the same season. For vessels hailing from Red Sea ports this was impossible. To the Greek or Roman merchant who ventured to the Far East this was a venture indeed, not to be undertaken unless under exceptional conditions of charter or cargo to be secured. There is, however, good reason to believe that the eastern sea trade of India exceeded its western trade and that the commercial activity of Rome in its prosperous period was but the reflection of greater activity in the capitals of India from Madura to Palibothra.

Comparison of the evidence available from the records of Rome, India, and China is of interest. Professor Hirth​3 has made the Chinese Annals available, and we find that although the southern coast of China proper, including the ports of Kwang‑Tung and Fo‑Kien provinces, had not as yet been made part of China, the province of Tong‑King had been over-run by the Chinese B.C. 214, was incorporated into the Empire B.C. 111, and remained a Chinese possession until A.D. 263.​4 This conquest followed that of the province of Yun‑Nan, still one of the richest provinces of Chinese in metals, in forestry and agriculture. From the capital of Yun‑Nan situated on an inland lake, vividly described by Marco Polo (2.48), and evidently reflected through the accounts of the Roman traders, there was a well-defined trade route down the Yang‑tse river and overland through central China to the capital Singan‑fu on the watershed of the Yellow River; and from the adjoining plateau of eastern Tibet through parallel valleys within a few miles of each other flow not only the Yang‑Tse and the Red River of Tong‑King, but the Me‑Kong of Cochin China, the Salwin and Irawadi of Burma. Adjoining Yun‑Nan in the water-sheds of these Burmese rivers was located the kingdom of the Shans, then an important tribal federation, and there is reason to believe that an active trade existed out of China through Yun‑Nan with the Shans as intermediaries. A Chinese record dating from A.D. 120 informs us that 'the king of the country of Shan sent an embassy to the Chinese Emperor offering musicians and jugglers,' whose accomplishments suggest the juggling of India, and  p243 who said: 'We are men from the west of the sea. The west of the sea is the same as Ta‑Tsin. In the southwest of the country of Shan one passes through to Ta‑Tsin.'​5 While Ta‑Tsin was the name given by Chinese to the eastern lands of the Roman Empire, it cannot always be given that meaning, and in this case seems to mean merely people coming from the West. The route is, however, clear; the embassy came by sea to the southwest of the Shan country, that is, the Gulf of Martaban, the shores immediately east of the modern Rangoon, and proceeded inland up one of the river valleys. The modern rail route leaving Rangoon follows the valley of the Sittaung river to Mandalay, thence up the Irawadi. At Bhamo, the head of navigation on that river, the overland route to Yun‑Nan began crossing the parallel gorges of the other rivers by suspension bridges. The earlier route probably ascended the Salwin passing the Shan capital Theinni and crossing the other rivers a little lower down, both routes having as their destination Yun‑nan‑fu, Cheng‑tu‑fu, and finally Singan‑fu. Another Chinese record informs us that in A.D. 166 the king of Ta‑Tsin, 'An‑tun,' who may of course be identified with Marcus Aurelius, 'sent an embassy with tribute from the frontier of Jih‑Nan,' and that 'merchants of this country frequently visit Fu‑Nan, Jih‑Nan, and Kiao‑Tsi,' but that 'few of the inhabitants of those southern frontiers ever went to Ta‑Tsin.'​6 We have here evidently still a confusion of Burma, southern India, and the Roman East. The name of Antoninus suggests a stray Roman subject, but the merchants 'frequently visiting the southern states,' which we may identify with the modern Siam, Annam, and Tong‑King, must have come from the ports of India or Ceylon. Another record dating from the fourth century gives us the route from the Chinese capital to its Tong‑King seaport, and the routes down the other rivers as follows: 'southeast you come to Kiao‑Tsi; there is also connection by water [in fact by both river and ocean routes] with the principalities of Yun‑Nan and Yung‑Chang [near Bhamo; that is, through Burma].' Chinese interest in distant lands is reflected in this same record in its concluding observation: 'Although in that country, Ta‑Tsin, sun and moon and the constellations  p244 are quite the same as in Chinese, former historians say that going a hundred li west of Tiao‑Chih [mouth of the Euphrates] you come to the place where the sun sets. This is far from being true.'7

From the hills of Yun‑Nan came gold, silver, and precious stones, silk, and the fragrant cinnamon bark so greatly prized in Rome. The upper Yang‑Tse in Chinese speech is still the 'river of the golden sands,' and a recent traveler refers to a neighboring river valley as being called the 'silver shore.'​8 The overland route from Yun‑Nan to the upper Irawadi was used by conquering Chinese troops in the 18th century and was by them called the 'gold and silver route.'​9 The southern port of China mentioned in the record as southeast from Yun‑Nan, that is, down the Song‑Koi or Red River of Tong‑King and named Kiao‑Tsi, we may safely follow Richthofen in identifying as the Kattigara (or Katti-nagara, from some Prakrit-speaking pilot?) of Ptolemy and other Roman writers. This gave the Chinese Empire an outlet to the southern seas, the sailing course from which, being within the tropics, was steered by the southern cross and not the north star. The south seems to have been the cardinal direction with the Chinese. The magnetic needle having already been known to them for centuries, although apparently not put to practical uses for navigation, was also called the south-pointing chariot.10

Indian records of Farther India show very active communication at this same time. There was missionary activity of all creeds then held in India — Brahmin, Buddhist, and Jain — and there was active sea trade. The 'golden route' of the Chinese was known in India as the Golden Coast, Suvarṇa bhūmi, and near one of the mouths of the Ganges was an important port of India named Suvarṇa Grāma, the Golden Port, better known in the days of Arab trade as Sonargaon.​11 Not only from the  p245 kingdoms of the Ganges was there navigation across the Bay of Bengal: from southern India the sea trade was so important that the Andhra kings struck numerous coins bearing the impression of capacious two-masted vessels used in that service and evidently regarded as the source of national power and prosperity. The Tamil poem Paddinappalai12 gives us a vivid description of a busy port of the Chola Kingdom, Kaviripaddinam, which was built on the northern bank of the Kaviri river, then a broad and deep stream into which heavily laden ships entered from the sea without slackening sail. At the beach were raised proofs and warehouses where cargoes were stored. The goods were stamped with the royal tiger stamp after payment of customs duty and then released to the merchants. Close by were settlements of the Yāvana merchants, which name included not only Ionians or Greeks, but Graeco-Bactrians and Parthians. Here were quartered foreign traders from other lands beyond the seas, and precious cargoes of many kinds were brought from all directions — from the northern mountains, the western coast, the valley of the Ganges, Ceylon and Burma. There were lighthouses built of brick and mortar which exhibited flashing lights at night to guide ships to port. Among the workmen on the Chola palace in that city were not only artisans from all parts of India, but carpenters from Yāvana, that is, probably Greeks from Egypt or Syria. Another Tamil poem describes the 'seaport of Muchiri on the west coast near the mouth of the Periyar where the beautiful large ships of the Yāvanas bringing gold come splashing the white foam on the waters of the Periyar which belongs to the Cherala and return laden with pepper.'13

An early Sanskrit play of India, 'The Little Clay Cart,' describes the same activity. One of the characters is a gentleman 'dressed in silken raiment glittering with rich ornaments.' In one of the scenes appears a row of jewelers' shops 'where skilful artists are examining pearls, topazes, sapphires, beryls, rubies, lapis lazuli, coral, and other jewels; some set rubies in gold; some work with gold ornaments on colored thread; some string pearls; some grind lapis lazuli; some pierce shells and some cut  p246 coral.'​14 In this list there are gems from all four points of the compass.

Meager enough are the Roman references to this sea trade. The conditions of the journey given by Pliny have already been mentioned. The author of the Periplus mentions the three kingdoms of southern India: Chera, with its port of Muziris, the Muchiri of Tamil poets; Pāṇḍya, the capital of which, Madura, Pliny reports as Modiera; and the 'coast country,' that is, Chola, with its capital called Argaru, that is Uragapura, Uraiyūr, the modern Trichinopoly on the Kaviri, while the port of Kaviripaddinam he mentions as Camara, and says that there were in that port not only the large single-masted vessels peculiar to the Malabar coast, but others very much larger which made the voyage to Chryse and the Ganges. He mentions Chryse as 'an island opposite the Ganges and under the rising sun' and tells us that beyond Chryse the sea comes to an end, and that to the north was 'a land called This with an inland city called Thinae from which silk was brought overland through Bactria to the Gulf of Cambay and by way of the Ganges to the ports of Damirica,' that is, Tamil Land, the Tamilakam of their poets (Peripl. Mar. Erythr. 63, 64).

There is apparent confusion here between the overland Turkestan route to North China and the combined sea and land route to South China. Thinae, the eastern metropolis, may be the same in name as Theinni, the Shan capital;​15 though the Turkestan caravans never reached it, and it is not north of the Gulf of Tong‑King, which is the place where the outer sea was thought to come to an end. Burma, Yun‑Nan, and China proper were brought into one peninsula by as late a geographer as Edrisi.

The sea route to Kattigara, according to Marinus of Tyre, was twenty days coasting south from the Golden Chersonese to a place called Zabai, whence, sailing 'a little to the left of south' one came in 'some days' to Kattigara. How many days he could not tell, but thought 'some,' as reported to him by mariners, meant 'many.' Ptolemy (1.14) criticized him severely, and said 'some' meant few,' and that Kattigara was therefore close  p247 by the Golden Chersonese. He plotted it on his map, fixed the latitude and longitude, and then asserted that the coast beyond trended westward, joining Africa below Zanzibar. But, calculations notwithstanding, he was guessing, just like Josephus who said Solomon's ships went for gold 'to the land that of old was called Ophir, but now the Golden Chersonese, which belongs to India' (Ant. Jud. 8.2).

Marcian of Heraclea, generally considered a mere compiler out of Ptolemy, gives further details of the far eastern voyage. The unknown land east of the Sinae, and the unknown land south of the sea called Prasodes (this name being identical with the Green Sea of the Arab geographers and the Erythraean or Red Sea of the Greeks, although apparently derived from Sanskrit prasāda, 'pacific') came together, making 'a sort of angle near the Gulf of the Sinae.' Above the Sinae, he said, was the region of the Seri and their metropolis; the unknown eastern land was dotted with 'stagnant lakes, in which great reeds grow, so closely crowded together that men cross over the lakes by walking upon them' — which may readily refer to the lakes of Yun‑Nan and the bamboos used for bridges there. He mentions Thinae, the metropolis of the Sinae, as 'the border between the known and the unknown land'; but the sailing course beyond Kattigara, he says, 'cannot be set down in stages or figures,' because 'there are no witnesses to point out the course unless it be some God who knows.'16

But it is noteworthy that Marcian's sailing distance, ignoring his directions, is reasonably accurate between his 'Point of Departure' at the westernmost mouth of the Ganges, and the mouth of the Cottiaris river by Kattigara, if we take that to be the Song‑Ka or Red River of Tong‑King.​17 And from that point it is true that 'the unknown land to the east' makes 'a sort of angle' with a land reaching westward, the Island of Hai‑Nan. Marcian's sailing course also is correct, holding southward from the Golden Chersonese, if we take that as the Martaban coast — the modern Moulmein — and not the lower tip of the Malay Peninsula. There, however, his knowledge ends.

 p248  There is a triple confusion in all these sailing courses of the Roman period. Mediterranean courses were set by the north star, and 'the right-hand coast' would be east. Red Sea courses were set by the wind and the coast-line, never far distant, and the direction being south, the western or African shore was the 'right-hand coast' and the eastern or Asiatic shore the 'left-hand coast.' At the Horn of Africa, Cape Guardafui, the course was set by the trade winds, and connection was made with the active shipping of India, where the cardinal direction was east. This may have led to Pliny's and Ptolemy's failure to realize the southern extension of India, which was well known to the author of the Periplus. At the Tamil ports in Southern India, connection was made with shipping bound across the Bay of Bengal to the Golden Chersonese, also a course steered by the trade winds. There connection was made with Malay or Chinese shipping bound to Kattigara, but once past the Straits of Malacca, directions were reversed, and an actual east-and‑north course was reported as south-and‑west. This may have been due merely to the different point of view of the steersman. The Greek southbound was steering backward, as it were; and the Chinaman forward, south being his cardinal direction.

But finally we have the correct description of the trend of the coast at the head of the Gulf of Tong‑King, indicating personal observation by some navigator who was neither Tamil, Hindu, nor Chinese, and the wholly unwarranted assumption that the coast of Hainan extended westward all the way to Cape Prasum in Africa (Peripl. Mar. Ext. 40). A like assumption was made by the author of the Periplus for Ceylon (Peripl. Mar. Erythr. 61). Both were due to the notion of a southern continent or Antichthones, conceived by both Greeks and Romans as necessary to counterbalance the Eurasian continent and so prevent it from sliding off toward the ultimate north.18

The inferences as to Roman enterprise by sea to the far east are negative. That Roman shipping frequented the ports of the  p249 Tamil Kingdoms and Ceylon is undoubted. But of the great beyond, they brought back hearsay. The author of the Periplus, like Tavernier in the 17th Century, gives us a summary out of the Rāmāyaṇa and the Purāṇas. Marinus of Tyre gives the accounts of a few other mariners, on which Ptolemy makes specious calculations. Tamil literature and coinage alike testify to maritime enterprise eastward, and Chinese annals refer to a few visits of people coming 'from the west of the sea' — which may mean Tamil Land and Ceylon, though in one case the mention of An‑Tun seems to mean a Roman subject. All point to the same conclusion, that Roman ships in the Bay of Bengal and the China Sea were so rare that two or three in a century might tell their tale.

But what of the other voyager out of the west, who came to Kattigara and, being conducted to the Chinese Court, gave an account of the lands to the west of the sea, which the Annalist warns us are not at the place where the sun sets? His name comes to us as Tsin‑Lun;​19 that may be no more than Lun, the Tsin, or Roman; and may not Lun also be some attempt of the Chinaman to pronounce Romanus? Here was an earlier Marco Polo who should have imprisoned by some enemy, that the world might be the richer for his memoirs. For of the vast Pacific the only witness to the Roman world might be, as Marcian put it, 'some God who knows.'

The Author's Notes:

1 Nundo Lal Dey, Notes on the History of the District of Hugli or the Ancient Rada, JASB new series, 6, no. 11, 1910.

2 Peripl. Mar. Erythr. 57.

3 China and the Roman Orient, passim, from which references herein are quoted.

4 Richthofen, op. cit. 1.509.

5 Hou‑han‑shu, c. 86.

6 Ibid. c. 88; Liang‑shu, c. 54.

7 Wei‑shu, c. 102.

8 Johnston, From Peking to Mandalay, p35, 44, 104, 255.

9 Cordier's Yule's Marco Polo, 2.67‑76.

10 Hirth, Ancient History of China, p126‑134.

11 Nundo Lal Dey, Notes on Ancient Anga or the District of Bhagalpur, JASB new series, 10, no. 9, 1914; cf. the Mahājanaka Jātaka, where a single ship from Chāmpā to Suvarṇabhūmi had on board seven caravans with their beasts.

12 Quoted from Mookerji, History of Indian Shipping, p135‑6; see also Pillai, The Tamils 1800 Years Ago, p16, 24‑26.

13 Erukkaddur Thayan Kannanar‑Akam; quoted from Mookerji, op. cit. p135.

14 Mṛcchakaṭika, Act 4, tr. Ryder, Cambridge, 1905, p70.

15 Kingsmill, The Mantse and the Golden Chersonese, JRAS China Branch, 35.76‑101.

16 Peripl. Mar. Ext. 44, 46: fifth mouth of Ganges to frontier of Sinae, 45,350 stadia; thence to Cottiaris R., 12,650 stadia.

17 Ibid. 44. Concerning the 'Point of Departure' cf. Nundo Lal Dey, The Vikramasila Monastery, JASB new series, 5, no. 1, 1909.

18 Cf. Pomponius Mela, De Situ Orbis, 1.9: quod si est alter Orbis, suntque oppositi nobis a meridie Antichthones, ne illud quidem a vero nimium abscesserit, in illis terris ortum amnem, ubi subter maria caeco alveo penetraverit, in Nostris rursus emergere et hac re solstitio accrescere, quod tum hiems sit unde oritur. Alia quoque in his terris mira sunt . . .

19 Liang‑shu, c. 54.

[image ALT: Valid HTML 4.01.]

Page updated: 2 May 20