You are on page 1of 13

The Persian Gulf or the Red Sea?

Two axes in ancient Indian Ocean trade, where to go


and why
Author(s): Eivind Heldaas Seland
Source: World Archaeology , SEPTEMBER 2011, Vol. 43, No. 3, THE ARCHAEOLOGY OF
TRAVEL AND COMMUNICATION (SEPTEMBER 2011), pp. 398-409
Published by: Taylor & Francis, Ltd.

Stable URL: https://www.jstor.org/stable/41308507

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/41308507?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to
World Archaeology

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
The Persian Gulf or the Red Sea? Two
axes in ancient Indian Ocean trade,
where to go and why

Eivind Heldaas Seland

Abstract

From the first century BCE until the third century CE, Roman trade with the Indian Ocean passed by
way of two major axes: the Red Sea-Nile and the Persian Gulf-Syrian Desert. The overall distance
by way of the Persian Gulf-Syrian Desert route was considerably shorter, but the overland part of
the journey was almost four times longer, and goods had to pass through politically tense border
regions between the Parthian and Roman empires. Although the relative importance of the two
routes probably varied, both were in operation at the same time and for a prolonged period. How
can this be explained? This article explores the passages from India to the Mediterranean in the
Roman imperial period and argues that a relatively straightforward answer is to be found in the
annual rhythms of movement responding to ocean winds, desert weather and river floods.

Keywords

Palmyra; Indian Ocean; Red Sea; Persian Gulf; maritime trade; caravan trade.

Introduction

In the late months of the year 467, era of Seleukos (= 156 CE), a group of merchants from
the Syrian city of Palmyra were waiting out the monsoon in the harbour of Barbarikon at
the mouth of the Indus. As the north-easterly winds stabilized in the early months of the
new year, which in Roman Syria was reckoned to start in October (Butcher 2003: 122),
they would leave on the ship of Honaino son of Haddoudan, for the Persian Gulf port of
Forat, near modern Basra, from which they would proceed to Spasinou Charax, the
capital of the Mesopotamian kingdom of Mesene. In Charax, goods would be loaded on
camels and carried to Palmyra, a month-long journey across the Syrian Desert. The
merchants had to bring their provisions for the entire journey and depended on protection

Routledge World Archaeology Vol. 43(3): 398^09 The Archaeology of Travel and Communication
K Routledge Taylor & Francis croup © 2011 Taylor & Francis ISSN 0043-8243 print/1470-1375 online
http://dx.doi.org/10.1080/00438243.2011.605844

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
The Persian Gulf or the Red Sea? 399

against or good relations with the nomadic population and the patronage of members of
the Palmyrene elite for water, guides and safety (Sommer 2005: 213-18; Teixidor 1984: 15-
19; Will 1957, 1992: 58-63; Yon 2002: 100-6; Young 2001: 150-7). From Palmyra goods
were then transported to the Mediterranean for customs clearance and onwards
distribution to the many cities of the Roman Empire, where Chinese silk, Indian cotton
and Afghan gems were in demand. An inscription set up in the agora of Palmyra in March
468/157 ( Inv . X 96/ PAT 1403), records that the merchants arrived safely in Palmyra. The
route they may have followed can be reconstructed by way of the c. thirty other
inscriptions referring to the trading activities of Palmyrene merchants (Gawlikowski 1996:
142-3; Yon 2002: 263-4).
Merchants travelling from India, however, also had other opportunities. Literary
sources such as the first-century CE Periplus Maris Erythraei and the rich archaeological
records from Roman Red Sea ports attest that other ships sailing out of Indian ports took
the much longer sea journey to Berenike and Myos Hormos in Egypt (Peacock and Blue
2006; Tomber 2008: 19-25, 44-56, 71-82; Wendrich et al. 2003). From there caravans
conveyed goods across the desert to Koptos on the Nile. From Koptos Nile boats brought
eastern commodities to Alexandria for customs, processing and marketing. For the
merchants from Palmyra, the choice was easy, they were simply going back home, and
thus chose the Palmyra route. The simultaneous use of these two different corridors of
exchange between Indian Ocean and the Mediterranean, however, warrants explanation.
How did a long overland journey across the Syrian Desert compete with a shorter one
across the Egyptian Eastern Desert and downriver transport on the Nile? In this case it can
be suggested that a straightforward answer is best and arguably one can be found by
studying patterns of navigation, overland movement, flood and weather along the
alternative routes.

Table 1 lists the approximate distances of the different legs along the two routes as
measured on the map. The distances actually travelled will have been longer. Pliny the
Elder ( HN 6.102), for instance, gives the figure of 309 Roman miles = 409km for the
Berenike to Koptos road. The total distance along the Red Sea-Nile route is about a third
longer than along the Persian Gulf-Syrian Desert alternative. The overland distance to be
covered between the Gulf and the Mediterranean is, however, almost four times greater
than that between Berenike and Koptos. The cycle of the monsoon system and the
dependence on commodities subject to annual harvest seasons will probably have limited
the maritime circuit to one annual return journey from Persian Gulf ports as well as from

Table 1 Approximate distances from the Indus to the Mediterranean

Red Sea corridor Persian Gulf corridor

Barbarikon-Berenike 4500km Barbarikon-Forat/Charax 2350km


Berenike-Koptos 380km Forat/Charax-Palmyra 1100km
Koptos- Alexandria 760km Palmyra-Antioch via Chalkis 300km
Total distance 5640km 3750km
Overland distance 380km 1400km
Riverine distance 760km

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
400 Eivind Heldaas Seland

Red Sea ports, so the longer sea journey from India to the Red Sea can hardly explain
use of the Persian Gulf-Syrian Desert alternative. Recent work has challenged establis
views that the costs of land transport in the ancient world were prohibitive compare
those of maritime and riverine alternatives: people could and did convey goods overl
all the time (Adams 2007: 2-16), and almost any long-distance journey could incorpo
different means of transport (McCormick 2001: 65-6). This, however, does not mean t
ship transport was not generally a more efficient way of moving large quantities of good
as it indeed remains today; and the choice of a significantly longer overland journey in th
labile political environment between the Euphrates and the Orontes, which was also
border zone between the Roman and Arsacid (Parthian) empires, merits attention. Le
follow the two routes available to our merchants waiting in a port at the mouth of
Indus and see whether the alternatives had merits or disadvantages, which might exp
why some would choose one option, others the other.

Methodological approach

Pre-modern travel was directly influenced by climate and topography. Modern


infrastructure and transport technology started to reach the regions between the Indian
Ocean and the Mediterranean only in the late nineteenth century, and traditional
practices continued to coexist with modern alternatives well into the twentieth century.
This enables us to utilize the experiences of medieval and early modern travellers
preserved in handbooks, travel descriptions and ethnographic accounts to reconstruct
ancient patterns of communication. For example, camels which conveyed goods from
Spasinou Charax to Palmyra in the second century CE walked at the same pace as the
camels of the annual caravans between Basra and Aleppo in the eighteenth century. The
climate might have been slightly cooler and more humid (Issar 2003: 25-7), but literary
sources leave no doubt that the landscape between Mesopotamia and north-western
Syria was a desert, exploited primarily by nomadic camel pastoralists (Pliny HN 5.87,
5.88, 6.126). Traffic on the rivers Nile and Euphrates was influenced by inundations that
continued to take place every year at the same time of the year until the construction of
the Aswan and Tabqa dams in 1902/1964 and 1973. Navigation on the Indian Ocean
relies on the monsoons, which vary in strength from year to year, but which have
experienced only minor long-term fluctuations over the last 2000 years ago (Gupta et al.
2003). This explains why the fragmentary navigation schedules provided in sources such
as the first-century Periplus of the Erythraean Sea and Pliny the Elder's Natural History ,
fifteenth-century and seventeenth-century navigation handbooks by Ibn Majid and John
Thornton as well as the eighteenth-century travelogue by Carsten Niebuhr are similar
whenever they overlap (Seland 2008a). Sailing technology has certainly changed over
time, but recent work shows that these changes were not dramatic and had limited
impact on the performance of the ships used in this region (Whitewright 2011, in press).
Medieval, Ottoman and colonial practices cannot reveal the exact dates, speeds and
travel times of the ancient period, but arguably they can aid reconstruction of schedules
of communication to the degree to which they were influenced by topography and
annual climatic cycles.

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
The Persian Gulf or the Red Sea? 401

Indian Ocean, Red Sea, Persian Gulf: navigation patterns and travel times

The monsoons of the Arabian Sea blow steadily from the south west from May to
September and from the north east from November to March. April/May and October
are transitional periods. Pliny the Elder indicated that the best time to leave Indian
ports for the Red Sea was between the start of December and 13 January (HN 6.106).
This would bring ships to the southern Red Sea in time to catch the favourable
south-east winds prevailing there from January to March (National Imagery and
Mapping Agency 2001: January-March). These are described in modern times as
reliable up to the latitude of Jiddah and Aydhab (Facey 2004: 11), not far from
Berenike, the start of the ancient overland route to the Nile. From April to December,
the prevailing wind in the Southern Red Sea is from the north (National Imagery and
Mapping Agency 2001: April-December), so ships from India will probably have
planned their journey so as to arrive in the Red Sea by the end of March at the latest
(Facey 2004: 9-11).
The India-Persian Gulf leg is not mentioned in classical sources, but can be
reconstructed using Islamic and British sources. Ahmad Ibn Majid (d. circa 1500) gave
18 October as the start of the sailing season from Gujarat to Arabia (Tibbets 1981: 230),
which would also be the first leg for ships heading up the Gulf. Later departure would
certainly have been possible. Ibn Majid considered up until early March as safe (Tibbets
1981: 231). James Horsburgh's nineteenth-century India Directory recommends the
months of November through February for the passage from India to the Gulf
(Horsburgh 1841: 483). In July and August, only the so-called southern passage (westward
to Africa below the Equator and back north east along the coasts of Africa and Arabia)
was available (1841: 484), while the passage in September and October, before the onset of
the north-east monsoon is described as 'very tedious' (1841: 484).
Concerning arrival times, we have no 'closing time' in the Persian Gulf like that
constituted by the northerly winds prevailing in the Red Sea from April onwards. A
nineteenth-century journey from Bombay (now Mumbai) to Muscat was expected to last
only ten to twelve days in the November-February season (Horsburgh 1841: 454), and the
trip from the mouth of the Indus probably took even less time. Not all recorded
experiences were that smooth. Carsten Niebuhr needed twenty-seven days for this distance
in bad weather during December and January 1764 (Niebuhr 2003: 94-5), and another
eighteen days to Bushehr (2003: 100-1), still c. 190 nautical miles (352km) from Basra.
Pietro Delia Valle left Chaul near Mumbai on 17 December 1624 and arrived at Muscat on
13 January (Delia Valle 1843: 797-801). Leaving Muscat on 26 January 1625, he entered
the Shatt al-Arab (the river leading from the Gulf to Basra) only on 1 1 March after a
difficult journey (1843: 808-15). If ancient journeys were somewhere in-between the swift
passages anticipated by British captains sailing non-stop and with the relatively reliable
instruments and maps of the nineteenth century and the slow passage of Delia Valle,
whose ship partly navigated by following the coasts, experienced calms, anchored at times,
and at other times even utilized oars, we could perhaps hypothesize that the passage from
the mouth of the Indus to Furat at the head of the Persian Gulf would take our Palmy rene
friends about two months. This would bring ships carrying Indian goods for Palmyra to
their Persian Gulf destinations any time after mid-December.

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
402 Eivind Heldaas Seland

Arranging transport

When ships arrived in port, there will have been formalities to take care of before good
could be transported onwards. The Muziris Papyrus (P. Vindob. G. 40822/SB XVII
13167), reveals that in the second century, the main tax on Red Sea trade, the 25 per cent
tetarte, was levied in Alexandria. There has been considerable debate on the nature and
details of the agreement and accounts drawn up in the papyrus (Rathbone 2000), but it is
clear that goods were sent under seal to Alexandria via Koptos for taxation and that they
were stored in government warehouses ( apothekai demosiai) in Koptos, awaiting the
riverine leg of the journey (recto, col II: 3-7). This arrangement presupposes that an
official inventory of imported goods must have been made upon arrival in the Red Sea
ports, in order to make sure that they did not disappear on their way to taxation. While
details of Roman Red Sea taxation are not clear, nothing at all is known about taxation in
the kingdom of Mesene, but it seems likely that the kings of Mesene, too, would levy some
kind of tax on incoming commerce.
After customs had been cleared (or preferably before), transport had to be arranged.
For the routes by way of Egypt as well as Palmyra camels were the obvious choice of pack
animal. A considerable number of animals would be needed. Nineteenth-century British
army veterinary advice was not to exceed 135-80 kilos (300-4001b) for prolonged journeys
unless you wanted to ruin your animal (Leonard 1894: 204-6). This correlates well with a
normal camel load in Roman Egypt, estimated by Adams as approximately 4001b (2007:
80). The few ancient depictions we have of camel caravans also show animals relatively
lightly loaded, with two amphorae on each side. There is now a fair amount of
archaeological evidence that the merchant ships sailing out of Roman Red Sea ports were
of the same kind as those sailing in the Mediterranean (Whitewright 2007: 83-4). There
ships of 100-50 tons of burden were normal, with much larger vessels not uncommon
(Casson 1995 [1971]: 170-3). Ships carried things other than trade goods and would not
necessarily utilize their whole capacity, and most goods imported from the Red Sea/Indian
Ocean region were relatively low weight/high value. Nevertheless, it is clear that the
number of animals required to convey goods delivered by even a single ship would be
substantial. In the winter months, when most ships arrived, the demand for transport will
have increased dramatically. The main Red Sea-Nile routes were well served with
protected wells, cisterns and road stations (Sidebotham et al. 2008: 329^3), and there is
evidence of relatively small groups of travellers and animals moving back and forth
throughout the year and of medium-scale private transport firms operating, the Nikanor
family business being the most famous example (Adams 2007: 221-34). It seems likely that
merchants here would depend on such contractors to bring their consignments to the
government warehouses on the Nile.
In the Syrian Desert we find no trace of such private contractors and there was no
infrastructure comparable to that maintained by Roman authorities in the Eastern Desert
of Egypt. Transport was organized by merchants (Greek: emporoi/ Aramaic: tgr) in
caravans (, synodia/syrh ), which are well attested in the Palmyrene epigraphic record
(Gawlikowski 1996: 142-3; Yon 2002: 263-4). For the outward journey from Palmyra to
the Gulf, the caravans seem to have crossed the desert to Hit and travelled downstream on
the Euphrates from there on rafts consisting of inflated skins (Gawlikowski 1983: 168-9).

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
The Persian Gulf or the Red Sea? 403

For the return journey, the river or the river valley was not an equally attractive
alternative. In the Ottoman period, travelling with laden camels along the river during the
winter and spring months was considered impractical because the ground was frequently
muddy due to winter rains and spring inundation (Grant 1937: 42-3). Upstream
navigation depended on oars or boats being towed along the riverbank. This must have
been challenging, at least during the floods, which reached their highest levels in April-
May (Finet 1969: 37-8; Wirth 1971: 109). Camels, on the other hand, were readily
available near the Gulf ports, as the camel-herding nomadic skenitai ('tent-dwellers')
known to have inhabited the Syrian Desert in the Roman period (Strabo, Geo. 2.5.32,
16.1.8, 28, 16.2.1, 11, 16.3.1; Pliny HN 5.65, 5.87. 6.125, 6.14^5, 6.146) probably had their
animals on winter pasture in what are now the border regions between Iraq and Saudi-
Arabia ready to travel back to the Palmyra region for the drier summer months, as their
Bedouin counterparts did in the eighteenth to twentieth century (Raswan 1930). This
model admittedly builds on the parallel of Ottoman period caravans (Carruthers 1929;
Grant 1937), but it is fully compatible with the many Palmyrene inscriptions describing the
return of caravans from Spasinou Charax at the head of the Persian Gulf and the probable
arrival time for ships travelling from India to the Gulf in the winter. Jean-Bap tiste Yon
(2002: 95-7, 100-1) and Michael Sommer (2005: 170-83) have argued for a close
relationship between the nomads of the steppe and the city population in Palmyra. This
would have been instrumental in furnishing the security needed for the journey across the
steppe and can help explain how the Palmyrenes were able to organize transport without
major investment in infrastructure akin to that in the Eastern Desert of Egypt.

The overland journey

After the formalities of taxation had been taken care of and transport had been organized,
goods would be ready to move on overland. Allowing a month after ships had arrived in
port, this could take place in March-April (Egypt) or after mid-January (Mesene). The
desert road to Koptos on the Nile is described by Pliny as a twelve-day journey ( N.H .
6.102), compatible with the 380km distance and well served with road stations, cisterns and
wells. This would bring the Indian Ocean imports to the government warehouses in Koptos
in early May at the latest, but possibly as early as March. The journey across the Syrian
Desert was substantially longer. Although the region receives more rainfall than the Eastern
Desert, there was little permanent infrastructure apart from unprotected wells and cisterns.
In the Ottoman period, caravan leaders would choose their route according to the current
security situation and availability of water (Beawes 1929: 29; Niebuhr 2003: 242-3).
Recorded journeys in the eighteenth century show that the c. 1 100km distance from Taybeh
or El-Kowm, c. two days' travel from Palmyra, to Basra could be covered by merchant
caravans in about one month (Carruthers 1929: passim). As provisions for people and
animals other than camels were not available under way, it is not to be expected that
caravans spent more time on the road than necessary. This means that if Palmyrene
merchants left Indian ports as soon as the monsoon allowed in November they could be
home by February or March, while their colleagues who travelled to Berenike would be
likely to be still engaged with formalities and transport on the Red Sea coast.

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
404 Eivind Heldaas Seland

Reaching the Mediterranean

The crucial difference between the Persian Gulf and the Red Sea route was, however,
last leg to the Mediterranean. No clear evidence reveals the westward direction of Ind
Ocean imports once they had reached Palmyra. The Damascus road seems unlikely.
would bring goods southward again and leave caravans with the Antilebanon a
Lebanon ranges to cross before the coast was reached. The military road in the direct
of Emesa (Homs) is well documented by milestones of the Severan emperors and th
common opinion has been that this was also the main westward direction of Palmyr
commerce (Sommer 2005: 203; Will 1992: 83). It has, however, also been argued tha
goods from Palmyra were taxed in Antioch (Seland 2008b; Young 2001: 149) and, if thi
correct, a route by way of Chalkis, near modern Aleppo, would have been more
convenient. Whether the caravans reached the coast by way of Emesa ( c . 200km), Apamea
( c . 250km) or Chalkis-Antioch ( c . 300km), the distance was easily covered in less than ten
days, all of the route going through territory under Roman control. This means that the
Indian and Chinese commodities that left the harbour of Barbarikon in November could

reach the Mediterranean by late spring, even allowing time for demanding bureaucracy
and slow pace of travel.
This was not equally straightforward in the case of Egypt. When commodities had been
stored in the government warehouses in Koptos, they would have to await onward
transport along the Nile. John P. Cooper's (2008, 2010) work on medieval and early
modern Nile navigation provides insight into the mechanisms and rhythm of movement on
this river before the building of the Aswan dams that is also relevant to the ancient period,
for which fewer sources survive.

Cooper dispels the image of Nile navigation as easy and documents the high risk of
grounding (2008: 74-5, 79-85), especially during the season of low Nile, which lasted
approximately from February to June (2008: 76-7). In the nineteenth century, only boats
with less than 0.5m draught/6t carrying capacity were able to navigate the upper Nile year
round, boats of 1.5m draught and 40t capacity had a nine-month season, 1.9m/100t boats
sailed for seven months and larger boats only five months (Cooper 2008: 79-81). Nile
navigation became increasingly difficult in early summer, not only due to decreasing water
levels, but also because of periodic gales in the Delta region (Cooper 2008: 93-5), which in
the April-May period made navigation even with larger rafts hazardous (Cooper 2010:
356). The season commenced only as the inundation took full effect in August-September
(Cooper 2008: 75-8, 79-81).
The implication of this is that goods imported from India by way of the Red Sea would
reach the Nile in the months of increasingly difficult navigation conditions. Roman Nile
boats came in all sizes, just as the boats studied by Cooper did. No ancient data on
draught is preserved, but I. J. Poll's survey on the size of Nile boats in Ptolemaic and
Roman papyri includes vessels with capacity from 50 to 18000 artabas (1996: 135-8). The
artaba was a dry measure generally used for wheat, and, using the Roman-period artaba of
approximately 401itre/30kg (von Reden 2010: 149), this equals circa 1.5-540 metric tons,
reflecting the nineteenth-century range discussed by Cooper. It is likely that relatively large
boats would be preferred for the transport of eastern goods. Literary and archaeological
evidence clearly shows that the volume of imports was high (Tomber 2008: 19-56).

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
The Persian Gulf or the Red Sea? 405

Figure 1 Upper: the reconstruction of ancient routes in the Persian Gulf and the Red Sea. Lower:
seasonal departure and arrival times for commodities.

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
406 Eivind Heldaas Seland

Utilizing large boats on the Nile and waiting for the main navigation season was not o
the most cost-effective mode of transport, but also sensible for avoiding the dange
grounding with expensive cargoes during the period of low Nile and difficult win
conditions. It thus seems likely that goods imported by way of the Red Sea left Kop
only in August-September. Cooper's data show that the Nile journey from Qus, close
ancient Koptos, to Alexandria could be expected to take about fifteen to twenty d
(2008: 99, 107). This would bring Indian goods to Alexandria in mid-late September at
earliest and then there was taxation to attend to, as the Muziris-papyrus shows. T
means that goods imported by way of Egypt would be ready for transit to Mediterranean
ports of call just as the main sailing season, which lasted from late May through m
September (Casson 1995 [1971]: 270-2), was ending.

The answer?

Long-distance movement in the pre-modern world was intimately linked to seasonal


changes in winds and precipitation. This does not mean that goods were not moved 'out of
season': it was perfectly possible to travel on the Nile in early summer using a small boat,
but by waiting for the inundation a safer passage on a bigger boat became an option. It
was certainly an option, although far more dangerous, to attempt to cross the Indian
Ocean outside seasons, ignoring the cycle of the monsoons and utilizing day breezes and
local winds, but that does not mean such conduct was usual or advisable. Mediterranean
sailors sailed in winter if they had to, but we know that they avoided this if they could. A
camel caravan from the Persian Gulf to Palmyra could be arranged in the early autumn,
but access to animals would be better in late winter/early spring when nomads habitually
moved their herds northwards. Such patterns provide insight into the factors travellers had
to take into account when making rational choices concerning safety of life and property.
The routes by way of the Persian Gulf and the Red Sea were in use during the same period.
The reconstruction of passages from the mouth of the Indus to the Mediterranean shows that
commodities transported by way of the Persian Gulf and the Syrian Desert were likely to
arrive at Antioch or another Mediterranean port by late spring, while goods going by way of
Egypt would probably reach Alexandria in the early fall (Fig. 1). In other words, Palmyrene
trade with India reached the Mediterranean at the start of the sailing season, Egyptian trade
at the end. The employment of multiple routes ensured that Mediterranean markets would
receive Indian Ocean imports at different times of the year and to multiple entrepots. This
would reduce risks connected to weather and political conditions on the Red Sea and
Euphrates borders and is also likely to have had a balancing effect on prices. In sum, this
seems to have been a strong incentive to keep up the route by way of the Syrian Desert, even
though the overland journey required was about four times longer.

Acknowledgements

Thanks to Julian Whitewright for advice and for sharing forthcoming work on Arab
navigation on the Red Sea and Indian Ocean, to John P. Cooper for making available his

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
The Persian Gulf or the Red Sea? 407

dissertation on Nile navigation and to Richard H. Pierce and two anonymous reviewers
for their comments on the manuscript. An outline of this article was presented as a lecture
at the Silk Routes and Eastern Contacts Seminar, Department of Archaeology, University
of Oxford and the finished result benefited greatly from the following discussion. This
research is a part of the project 'Palmyrena: city, hinterland and caravan trade between
Orient and Occident', funded by the Norwegian Research Council.

University of Bergen
Eivind.Seland@ahkr.uib.no

References

Adams, C. 2007. Land Transport in Roman Egypt: A Study of Economics and Administration in A
Roman Province. Oxford: Oxford University Press.
Beawes, W. 1929. A journey from Aleppo to Basra in 1745. In The Desert Route to India (ed. D.
Carruthers). London: The Hakluyt Society, pp. 1-40.
Butcher, K. 2003. Roman Syria and the Near East. Los Angeles, CA: Getty Publications.
Carruthers, D. (ed.) 1929. The Desert Route to India. London: The Hakluyt Society.
Casson, L. 1995 [1971]. Ships and Seamanship in the Ancient World. Baltimore, MD, and London:
Johns Hopkins University Press.

Cooper, J. P. 2008. The medieval Nile: route, navigation and landscape in Islamic Egypt. Doctoral
dissertation, Faculty of Laws, Arts and Social Sciences, School of Humanities, University of
Southampton.
Cooper, J. P. 2010. Humbler crafts: rafts of the Egyptian Nile, 17th to 20th centuries AD. Journal of
International Nautical Archaeology , 40(1): 1-17.
Delia Valle, P. 1843. Viaggi di Pietro Delia Valle il Pellegrino descritti da lui medesimo in letter e
familiari all erudito suo amico Mario Schipano. Vol. 2. Turin: A. Fontana.
Facey, W. 2004. The Red Sea: the wind regime and the location of ports. In Trade and Travel in the
Red Sea Region (eds P. Lunde and A. Porter). Oxford: British Archaeological Reports International
Series 1269, pp. 7-17.
Finet, A. 1969. L'Euphrate, route commerciale de la Mesopotamie. Annales Archeologiques Arabes
Syriennes , 19: 37-48.

Gawlikowski, M. 1983. Palmyre et l'Euphrate. Syria, 60: 53-68.


Gawlikowski, M. 1996. Palmyra and its caravan trade. Annales Archeologiques Arabes Syriennes , 42:
139^4.

Grant, C. P. 1937. The Syrian Desert: Caravans, Travel and Exploration. London: A. & C. Bl
Gupta, A. K., Anderson, D. M and Overpeck, J. T. 2003. Abrupt changes in the As
southwest monsoon during the Holocene and their links to the North Atlantic Ocean. Natur
421: 354-7.

Horsburgh, J. 1841. The India Directory, or Directions for Sailing to and from the East Indies, China,
Australia, and the Adjacent Ports of Africa and South America, 5th edn. London: W. H. Allen.
Issar, A. S. 2003. Climate Changes during the Holocene and their Impact on Hydrological Systems.
Cambridge and New York: Cambridge University Press.
Leonard, A. G. 1894. The Camel: Its Uses and Management. London: Longmans, Green.

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
408 Eivind Heldaas Seland

McCormick, M. 2001. Origins of the European Economy: Communications and Commerce AD 300
900. Cambridge: Cambridge University Press.
National Imagery and Mapping Agency 2001. PUB 109 Atlas of Pilot Charts for the Indian Ocean
Culver City, CA: NIMA.
Niebuhr, C. 2003. Rejsebeskrivelse fra Arabien og andre omkringliggende lande. Vol. 2. Copenhagen:
Forlaget Vandkunsten.
Peacock, D. and Blue, L. (eds) 2006. Myos Hormos - Quseir Al-Qadim: Roman and Islamic Ports o
the Red Sea. Vol. 1, Survey and Excavations 1999-2003. Oxford: Oxbow.
Poll, I. J. 1996. Ladefahigkeit und GroBe der Nilschiffe. Archiv fur Papyrusfors chung, 42(1): 127
38.

P. Vindob. G. 40.822. 1987. Hypotheken-Urkunde eines Seedarlehens fur eine Reise nach Muziris
und Apographe fur die Tetarte in Alexandria (zu P. Vindob. G. 40.822) (ed. G. Thlir). Tyche , 2: 229-
45.

Raswan, C. R. 1930. Tribal areas and migration lines of the north Arabian Bedouins. Geographical
Review , 20(3): 494-502.
Rathbone, D. 2000. The 'Muziris' papyrus (SB XVIII 13167): financing Roman trade with India.
Bulletin de la Societe Archeologique d'Alexandrie , 46: 39-50.
Seland, E. H. 2008a. The Indian ships at Moscha and the Indo-Arabian trading circuit. Proceedings
of the Seminar for Arabian Studies , 38: 283-8.
Seland, E. H. 2008b. Trade routes of Palmyra: with special notes on western routes in the Palmyrene
trade. In Jebel Bishri in Context: Introduction to the Archaeological Studies and the Neighbourhood of
Jebel Bishri in Central Syria (ed. M. Lonnqvist). Oxford: British Archaeological Reports
International Series 1817, pp. 89-97.
Sidebotham, S., Hense, M. and Nouwens, H. M. 2008. The Red Land: The Illustrated Archaeology of
Egypt's Eastern Desert. Cairo and New York: The American University in Cairo Press.
Sommer, M. 2005. Roms orientalische Steppengrenze: Palmyra - Edessa - Dura-Europos - Hatra:
Eine Kulturgeschichte von Pompeius bis Diocletian. Stuttgart: Franz Steiner Verlag.
Teixidor, J. 1984. Un port romain du desert, Palmyre et son commerce d'Auguste a Caracalla.
Semitica , 34: 1-127.

Tibbetts, G. R. 1981. Arab Navigation in the Indian Ocean before the Coming of the Portuguese being
a Translation of Kitab al-Fawaid ft usul al-bahr wa'l-qawa'id of Ahmad b. Majid al-Najdi. Reprint of
the 1971 edition. London: The Royal Asiatic Society of Great Britain and Ireland.
Tomber, R. 2008. Indo- Roman Trade, from Pots to Pepper. London: Duckworth.
von Reden, S. 2010. Money in Classical Antiquity. Cambridge: Cambridge University Press.
Wendrich, W. Z., Tomber, R. S., Sidebotham, S. E., Harrel, J. A., Cappers, R. J. T and Bagnall, R.
S. 2003. Berenike crossroads: the integration of information. Journal of the Economic and Social
History of the Orient , 46(1): 46-87.

Whitewright, J. 2007. How fast is fast? Technology, trade and speed under sail in the Roman Red
Sea. In Natural Resources and Cultural Connections of the Red Sea (eds J. Starkey, P. Starkey and T.
Wilkinson). Oxford: British Archaeological Reports International Series 1661, pp. 77-87.
Whitewright, J. 2011. The potential performance of ancient Mediterranean sailing rigs. The
International Journal of Nautical Archaeology , 40(1): 2-17.

Whitewright, J. In press. Sailing with the Mu'allim: the technical practice of sailing in the medieval
Red Sea. In Navigated Spaces, Connected Places: Proceedings of the Fifth International Conference on
the People of the Red Sea, Exeter 2010 (eds D. Agius, J. Cooper and C. Zazarro). Oxford: British
Archaeological Reports International Series.

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms
The Persian Gulf or the Red Sea? 409

Will, E. 1957. Marchands et chefs de cara vanes a Palmyre. Syria, 34: 262-17.
Will, E. 1992. Les Palmyreniens: La Venise des sables (Ier siecle avant - IHeme siecle apres J.-C.).
Paris: Armand Colin.

Wirth, E. 1971. Syrien, eine geographische Landeskunde. Darmstadt: Wissenschaftliche Buchge-


sellschaft.

Yon, J. -B. 2002. Les Notables de Palmyre. Beyrouth: Institut Fransais d'Archaelogie du Proche
Orient.

Young, G. K. 2001. Rome's Eastern Trade: International Commerce and Imperial Policy 31 BC - AD
30. London and New York: Routledge.

Abbreviations

Inv.= Inventaire des inscriptions de Palmyre . 1930-1949. Damas: Direction Generate des
Antiquites de Syrie.
PAT= Hillers, D. R. and Cussini, E. 2003. Palmyrene Aramaic Texts. Baltimore, MD:
Johns Hopkins University Press.
SB = Sammelbuch griechischer Urkunden aus Agypten. Wiesbaden: Harrassowitz.
Ancient authors and their works have been abbreviated in accordance with Liddell, H. G.,
Scott, R. and Jones, H. S. 1940. A Greek English Lexicon (with a revised supplement of
1996), 9th edn. Oxford: Clarendon Press.

Eivind Heldaas Seland works as a postdoctoral fellow at the Department of Archaeology,


History, Cultural Studies and Religion, University of Bergen, Norway. His main academic
interest is in trade between the Indian Ocean and the Mediterranean in the pre-Islamic
period and its relationship to political power.

This content downloaded from


65.88.89.49 on Sat, 09 Oct 2021 13:41:38 UTC
All use subject to https://about.jstor.org/terms

You might also like