TRACING NEW DIMENSIONS IN THE ROMAN MILITARY ORGANIZATION OF THE EASTERN LIMES

Abstract

SYGIS – the Finnish archaeological survey and mapping project of Jebel Bishri-is tracing new dimensions for the Roman eastern frontier or the so-called eastern limes in Syria. It has been noted in several Roman frontier studies that the Euphrates region is insufficiently studied to clearly define the development of Roman policy and the extent of the military presence in the east. Jebel Bishri, a mountain between the Euphrates and Palmyra, has largely remained as an empty spot on the maps representing the Roman military presence. This is an illusion due to the fact that the mountain in its central parts has not been earlier practically studied by archaeological means. The surrounding Euphrates Valley and the Strata Diocletiana have been studied, e.g., by A. Poidebard in his early aerial prospections. However, new satellite image prospections, field surveys and mapping have traced, documented and identified earlier little known or unknown Roman military installations and networks on the side of the Euphrates and in the central parts of the mountain. Especially the prospections with QuickBird satellite images offering good spatial resolution have made it possible to trace new sites which would have been difficult to recognize on the ground. The information gained from the new prospections and mapping enhances our understanding of the Roman military organization on the Euphrates and the fact that the military installations penetrated deeper in the desert-steppe areas and mountain of Jebel Bishri than earlier thought.

Full text

________CIPA 2005 XX International Symposium, 26 September – 01 October, 2005, Torino, Italy________
TRACING NEW DIMENSIONS IN THE ROMAN
MILITARY ORGANIZATION OF THE EASTERN LIMES
M. Lönnqvistª - K. Lönnqvistª - M. Stout Whitingª - M. Törmäb - M. Nunezc - J. Okkonenc
ªminna.lonnqvist@helsinki.fi, kenneth.lonnqvist@helsinki.fi, mwhiting@cc.helsinki.fi, University of Helsinki, FINLAND
bmarkus.torma@hut.fi, Helsinki University of Technology, FINLAND
cmilton.nunez@oulu.fi, jari.okkonen@oulu.fi, University of Oulu, FINLAND
KEY WORDS. Archaeology, Cultural Heritage, Remote Sensing, Surveying, Mapping
ABSTRACT
SYGIS – the Finnish archaeological survey and mapping project of Jebel Bishri - is tracing new dimensions for the Roman eastern
frontier or the so-called eastern limes in Syria. It has been noted in several Roman frontier studies that the Euphrates region is
insufficiently studied to clearly define the development of Roman policy and the extent of the military presence in the east. Jebel
Bishri, a mountain between the Euphrates and Palmyra, has largely remained as an empty spot on the maps representing the Roman
military presence. This is an illusion due to the fact that the mountain in its central parts has not been earlier practically studied by
archaeological means. The surrounding Euphrates Valley and the Strata Diocletiana have been studied, e.g., by A. Poidebard in his
early aerial prospections. However, new satellite image prospections, field surveys and mapping have traced, documented and
identified earlier little known or unknown Roman military installations and networks on the side of the Euphrates and in the central
parts of the mountain. Especially the prospections with QuickBird satellite images offering good spatial resolution have made it
possible to trace new sites which would have been difficult to recognize on the ground. The information gained from the new
prospections and mapping enhances our understanding of the Roman military organization on the Euphrates and the fact that the
military installations penetrated deeper in the desert-steppe areas and mountain of Jebel Bishri than earlier thought.
1. THE ROMAN LIMES IN SYRIA
It has been noted in several Roman frontier, or the so-called
limes studies, that the region of the Euphrates in Syria is still
insufficiently studied and known to say anything definite about
the Roman military organization there (see, e.g., Parker 2000,
134-135). SYGIS, a Finnish archaeological survey and mapping
project, since 2000 (see Lönnqvist and Törmä, 2003) has been
tracing new features and dimensions of the Roman military
presence in the mountainous region of Jebel Bishri in Central
Syria. Jebel Bishri flanks the southern side of the Middle
Euphrates Valley and extends deep towards Palmyra and the
Syrian desert. The mountain has remained as an almost blank
spot on the maps of the Roman military organization
demonstrating the Roman eastern limes. Prospecting with
remote-sensing methods, surveying, recording, documenting
and identifying on the ground the little or hither-to-unknown
Roman remains brings out new features and information to
better understand the Roman border in the Euphrates region.
The city of Palmyra in the middle of the Syrian desert south-
west of Jebel Bishri was a Roman ally with changing degrees of
autonomy and had the responsibility for controlling the desert
after Syria was annexed as a Roman province in 64 B.C. by
Pompey’s conquest (Plut. Pomp., 39). The eastern limit of the
Palmyrene control is described as having extended to the
Euphrates (Appian, the Civil Wars 5.1.9.). The so-called limes
interior passed Palmyra and consisted of areas directly under
Rome. However, the ’Desert limes’, the limes exterior, that lies
south and east of the Roman limes with its few desert castles
was not enough to control the nomads. (Musil 1928, 248;
Chapot 1907, 245-249). Later, the struggle for Palmyra’s
independence from the Roman hegemony and the search for the
status of its own empire led to the revolt of Palmyra by queen
Zenobia against Rome in AD 272. The limit of Palmyrene
power materialized in two fortresses on the Euphrates: those
bearing names of Zenobia and Zalabiya (e.g., Stoneman 1992).
Fig. 1. The Roman limes in Syria, including main Roman roads
and routes. The mountain of Jebel Bishri is located in the centre
to the east from the Strata Diocletiana (after Stoneman 1992).
After the struggle by Palmyra for its independence, the securing
of the limes by the Romans became a priority. The empire
especially started to strengthen its eastern limit towards the
nomads of the desert and the Parthian border zone of the
Euphrates. The work is particularly associated with the Emperor
Diocletian in inscriptions (CIL III, Suppl. 2, 14380) dating to
AD 306 and by historical narratives from the Late Roman and

________CIPA 2005 XX International Symposium, 26 September – 01 October, 2005, Torino, Italy________
Byzantine periods. Zosimus (2.34) describes Diocletian’s vast
building operations on different limes zones. In 1907, V. Chapot
already described differences in the eastern limes compared to
the differences in the visible line of ramparts and fosses or
continuous walls of the imperial limes in Europe. (Chapot 1907,
245-247). The Strata Diocletiana which was a continuation of
the Via Nova Traiana extending from Transjordan northwards,
is a route characterized by milestones, castra and castella. This
fortified line leads from Azraq through Palmyra to the
Euphrates. However, John Malalas (Chron. 12.308) mentions
that the Strata Diocletiana even extended from Egypt to the
Euphrates, but the material remains commemorating the
operation are attested only approximately from northern
Transjordan to the Euphrates.
In the western piedmont area of Jebel Bishri, military
installations of the Strata Diocletiana have long been known in
a line consisting of oases such as Taibe, Al-Kowm, Qdeir,
Rasafa and Sura. Oriza, a legionary base mentioned in
Ptolemy’s Geography and in the Tabula Peutingeriana, has
been identified with the oasis of Taibe that offers remains from
the Roman period. Beside Oriza, Sura is a well-known
legionary base of the Late Empire (Parker 2000, 122-126). Near
the slopes of Jebel Bishri there also exists Qasr al-Hair ash-
Sharqi which has been identified with ancient Adada (cf. A-
didi). (See, e.g., Musil, 1928, 233; Chapot, 1907, 329-330).
The Finnish project SYGIS surveyed in the region in the year
2000 (See Fig. 2).
During Diocletian’s reign the Habur river became the Roman
boundary in the east. At the junction of the Euphrates and the
Habur, according to Ammianus Marcellinus (23.5.2), Diocletian
built the fortress of Circesium. It also became necessary to
secure the area from Palmyra and Oriza to Circesium on the
right bank of the Euphrates (Musil 1928, 254). This article tries
to elucidate, through the collected evidence, how the area
between Oriza and Circesium was better secured than generally
thought. We concentrate on the areas prospected and surveyed
by the Finnish project both on the Euphrates side and the inner
regions of the mountain.
2. REMOTE SENSING THE FRONTIER
A. Poidebard made the first aerial surveys and prospections in
the area of the Strata Diocletiana and the Euphrates in the
1920s and 1930s. He published the studies in his classic work
La Trace de Rome dans le desert de Syrie, Texte and Atlas
(Paris, 1934) and later in R. Mouterde and A. Poidebard in Le
Limes de Chalcis, Texte and Atlas (Paris, 1945). Sir A. Stein
carried out similar kinds of studies in his Limes Report 1941
(see Kennedy 1982). D. Kennedy and D. Riley have especially
complemented Poidebard’s remote sensing studies in Syria, e.g.,
in Rome’s Desert Frontier from the Air (London, 1990). The
General Organization of Remote Sensing (GORS) in Syria has
also produced a fine archaeological space atlas (GORS, 2002)
of major archaeological sites and periods, including the Roman
military bases in Syria.
The Finnish project has used LANDSAT satellite images,
CORONA declassified satellite photographs, QuickBird
satellite images and the SRTM mission 2000 DEM (Digital
Elevation Model) data. The latter DEM data has been partly
received thanks to the projects membership in the DLR (the
German Aerospace Centre) projects and NASA’s world
monitoring program.
The spatial resolution of the QuickBird images has been
especially suitable for prospecting ancient structures in the
desert-steppe environment of Jebel Bishri, but LANDSAT-7
Fig. 2. Jebel Bishri and Roman remains in Le Limes de Chalcis
by R. Mouterde and A. Poidebard (1945).
ETM images and CORONA satellite photographs (KH-4A
mission, the size of the photos being 2.25 x 29.8 inches and the
resolution 2.7 m, see http://edc.usgs.gov/
products/satellite/declass1.html) have been used for general
mapping and visualization. The panchromatic channel of the
LANDSAT image reaches a resolution of 15 m, whereas the
spatial resolution of the QuickBird image is 0.6 m.
In his Atlas, Poidebard (1934) marked the fortresses and forts of
Zenobia, Mambri, Tabus and Qreiye along the Euphrates and
Jebel Bishri on the way from Sura to Circesium and also
indicated possible Roman roads linking these ancient fortresses.
The fortresses and forts formed a defensive line towards the
Parthian and later Persian border on the Euphrates.
Archaeological space atlas of Syria (GORS, 2002) mentions the
fortress of Zenobia as well as Mambri and Birtha (Qreiye) along
Jebel Bishri, but not the fort at Tabus which belongs to the
current survey area of the Finnish project. The tell at the village
of Tibne, which we earlier detected in the CORONA satellite
photographs (S FWD 1034-2 28 JUN 66; Lönnqvist and Törmä,
2004), is to be identified with the fort of Mambri. Mambri is
mentioned by the ancient historian Procopius (Buildings II, 8.7)
as having been built by the Emperor Diocletian and situated c. 8
km from Zenobia. The marble walls of a large compound are
still protruding from the tell, but the site has remained
unexcavated. Qreiye situated in the village of Ayyash, was
originally presented in Poidebard’s Atlas (1934, Pl. LXXXVII).
It is identified with Birtha Arupan in the Parthian version listing
the cities and forts conquered in the spring of AD 253. A
German archaeological expedition of the DAI (the German
Archaeological Institute) led by Dr. Markus Gschwind has been
surveying, mapping and excavating the site since 2002.
(http://www.dainst.org/print.php?id=742). The fort of Qreiye is

________CIPA 2005 XX International Symposium, 26 September – 01 October, 2005, Torino, Italy________
very visible on the CORONA photographs. The corner towers
of the triangular shaped fort of Tabus between Tibne and Qreiye
can also be detected in the CORONA photographs (S-FWD
1034-2 28 JUN 66), but the remains are hardly recognizable
before visiting the area on the ground. (See Fig. 5). We shall
here present the results of the field surveys and mapping
surrounding Tabus, elucidating the network of the forts on the
Euphrates, as well as remote-sensing studies in the inner parts
of Jebel Bishri with QuickBird satellite images. A closer field
study of the ruins at Tabus (H 19: UTM coordinates 0586948,
3925414; see Lönnqvist and Törmä 2004), situated at the edge
of Jebel Bishri looking over the Euphrates Valley 25 km north-
west of the city of Deir ez-Zor), is presented elsewhere in this
volume (see Lönnqvist et al. 2005b). In Le Limes de Chalcis
(Mouterde and Poidebard 1945, 130-32), it is mentioned that
there was an ancient route crossing over Jebel Bishri from
Rasafa-Sergiopolis through the desert wells “birs” such as Bir
Rehub and Bir Siqri to the Roman military post of Qseyibe from
which there was a connection to Circesium through the desert.
The attached map does not, however, show any forts or
fortresses along the route on the mountain (see Fig. 2). This is a
cartographic illusion which is due to the fact that the area was
earlier largely unexplored.
3. THE STRATEGY AND METHODS OF THE FIELD
SURVEY 2004
In the field season 2004 of the Finnish project, two working
groups operated in different regions: one on the Euphrates side
(situated c. 206-300 m a.s.l.; see Figs. 3 and 4: Plateau 1) on the
north-eastern edge and piedmont, the other in the area called
Nadra in inner districts on the mountain (situated ca. 490-600 m
a.s.l., see Fig. 2), because we wished to see how the remains
differed in type and periodically in two different environmental
areas.
Fig. 3. The field survey areas (Nadra and Plateau 1) of 2004
marked on the TPC G-4C Air Information Map (1:500 000). ©
Military Survey, Ministry of Defence, United Kingdom, 1998.
After the survey areas were chosen, the survey transects were
defined in the terrain according to natural borders such as the
edges of the alluvial terraces, the mountain edge, hills or wadis.
This approach enabled us to evaluate how the archaeological
sites and finds were located in the terrain according to natural
and environmental conditions, and how the environmental
phenomena may have affected the human activity and settling in
the region over time. Survey transects were studied in both
areas by field walking at 15 m intervals wherever the terrain
allowed, in groups consisting of 2-3 people. In the survey, all
the periods from prehistoric times to modern Bedouin tent bases
were taken into account. In this article we, however, limit
ourselves to the sites associated with Roman remains.
With a GPS (Global Positioning System, UTM zone 37), each
site was provided with UTM coordinates and the mean height,
m a.s.l. On the Euphrates side, recording was done with an
EDM (Electro-optical Distance Measurer) which was operated
by Jari Okkonen, PhD. The EDM was used, for instance, to
produce digital maps with ArcView program of the ancient
roads, graveyards and for mapping the fort at Tabus. A GPR
(Ground Penetrating Radar) was used by topographist Josep
Pedret Rodes, MSc., also on the Euphrates side to investigate
the geological formation of the main research area called
Plateau 1, the construction of an ancient road and tombs around
Plateau 1 and the fort at Tabus (see further in this volume
Lönnqvist et al., 2005b).
BW photographs, colour slides and digital images were taken.
Manual recording and drawing in the computerized field forms
were used. Accurate drawings to scale were prepared of the
most important archaeological sites and structures on millimetre
paper. Geographical directions and lines of archaeological
structures were determined by using digital and calibrated
international compasses for Near Eastern latitudes and military
technology-based compasses. Associated artefacts, such as
pottery and flints, were collected and all the artefacts from the
survey areas were later photographed, measured and stored at
the Palmyra Museum.
4. NETWORKING ON THE EUPHRATES
During the process of the field survey 2004, we identified two
ancient roads next to Plateau 1 and the fort at Tabus on the
north-eastern edge of Jebel Bishri along the Euphrates. The first
identified road (H 13) leads from E (UTM 0588873, 3924616;
c. 209 m a.s.l) to W (UTM 0587800, 3924989; c. 216 m a.s.l.)
and is preserved over one kilometre’s length on an alluvial
terrace (see Figs. 4 and 6).
Fig. 4. The Euphratine research area recorded with EDM and
mapped with ArcView program. Jari Okkonen 2004.
The alignment of this road can also be detected in a CORONA
satellite photograph (S-FWD 1034-2 28 JUN 66, see Fig. 5),
which reveals a longer continuation of the road, apparently to

________CIPA 2005 XX International Symposium, 26 September – 01 October, 2005, Torino, Italy________
Qreiye and to the west over a basalt bridge (H 14) which is now
in ruins damaged by fluvial currents and wadi streams.
Waypoints were taken with a GPS along the road for the
purpose of the mapping on a rectified satellite image from the
LANDSAT-7 ETM panchromatic channel (see Fig. 6). The road
has been constructed of two layers of angular marble stones
varying from 10–30 cm in section and in between which there is
a 15 cm thick layer of coarse sand; the overall thickness
recognizable from an eastern section being c. 35 cm. The top
stone layer forms the smoothed pavement of the road. The
stones of both layers consist of white, greenish and greyish
marble. The road has once been covered by a layer of asphalt
which has eroded away.
Fig. 5. The forts of Tabus and Qreiye along the Euphrates river
with a connecting ancient road (H 13) detected on a CORONA
satellite photograph intermingling with the modern Deir ez-
Zor–Aleppo road. The distance between the forts is c. 16 km.
Fig. 6. GPS points lining the road (H 13) on LANDSAT-7 ETM
satellite image (© Eurimage 2000), the road is also seen in the
CORONA photograph on Fig. 5. Mapping Jari Okkonen 2004.
In association with the road, a piece of Roman pottery and a
fragment of a marble statue (a wrist with a bracelet?) were
discovered and recorded. It became clear that the Pleistocene
alluvial terrace was chosen for allocating the aligning of the
road because, in that way, the irrigated and often flooded area,
could be avoided. The fort of Tabus, especially the eastern
tower above the road, may have functioned as a watch tower for
the road and its traffic (see Fig. 7).
Fig. 7. Straight alignment of the marble paved road (H 13)
towards the fort of Tabus taken from SE. Photo: Eivind Seland.
Fig. 8. GPR studies of the marble paved road (H 13) by Josep
Pedret Rodés and Elena Garcia-Guixé 2004.
The location, the 5 m width, the structure, the straight alignment
of the road and the associated finds are comparable to the
information pertaining to the Roman highways (see passim
Chevallier 1989) and Poidebard’s prospections (1934). The
structure of the section is reminiscent of the Legio–Scythopolis
Roman road in northern Judaea (cf. Isaac and Roll 1982, 40, 41,
Fig. 5), which was in use during the Late Roman and Byzantine
eras. However, the visible thickness and marble as paving
material in this road (H 13) are generally differing from the

________CIPA 2005 XX International Symposium, 26 September – 01 October, 2005, Torino, Italy________
Roman roads and their foundations. The Roman road
foundations are usually 50–65 cm thick - a depth to which our
GPR with a 500 MHz antenna could not penetrate in the clayish
alluvial terrace (see Fig. 8). The marble also poses another
question because of its softness as a paving material. Further
studies are thus needed. However, it has to be taken into
account that the Romans always used the available local
materials. Marble was in plentiful supply on the mountain of
Jebel Bishri, just a few dozens meters from the preserved road,
where we had earlier identified an ancient marble quarry (see
Lönnqvist and Törmä 2004). Quarries are typically connected
with Roman roads (Chevallier 1989). The associated finds of
the road (H 13) also refer to the Roman period, although the
road was covered with asphalt for more recent use, perhaps in
the 20th century, and led to a basalt bridge (H 14) apparently
built during the Ottoman rule. The location of the bridge, if the
road dates to the Roman period, points to the existence of a
Roman bridge point at the site.
The other road (H 15) discovered differs from the former one as
it is cut higher (ca. 236-250 m a.s.l.) into the natural rock of
Jebel Bishri and leads up towards (UTM 0587778, 3925051 to
0587276, 3925041) Tabus situated to the southeast. It does not,
however, enter the area of the fort itself but passes it, continuing
into the desert from the south-eastern side. This road consists of
a c. 3.5 m wide exposed marble bedrock into which rails or
wheel marks and transverse grooves to prevent slipping have
been carved to ease the climbing of the animals, such as
donkeys and horses pulling a cart up a slope. These kinds of
surfaces are common among Roman roads, but similar type of
roads may also have been in use in earlier periods (Chevallier
1989, 89). Beneath the rock-cut road at Jebel Bishri, there
appeared a c. 11 metre wide dam construction (H 22: UTM
0587192, 3925188) in a wadi, which was clearly the water
harvesting site for the fort of Tabus (H 19). From the dam, a
path led up to the fort at its eastern end.
Fig. 9. Hand-drawing of the road (H 15) with carved wheel
marks and transverse grooves cut into the marble rock and
climbing up to the mountain beneath the fort of Tabus.
After the recording and documentation of Tabus (see the
detailed description of Tabus in this volume in Lönnqvist et al.,
2005b) we studied and compared the architectural features,
surface finds and executed a GIS viewshed analysis between the
forts of Mambri, Tabus and Qreiye. It is probable that the site of
Tabus, used for guarding and even defence, was already
occupied during the heyday of Palmyra, e.g., in the time of
Zenobia, when Palmyra had to secure its territory and caravans
using the Euphrates road. However, in its present form, the fort
may have been rebuilt by the Emperor Diocletian and/or
Justinian I. The GIS viewshed analysis indicated good
intervisibility of the forts so that they seem to have functioned
together as a defensive line with a road (see Lönnqvist et al.,
2005). The evidence of the roads and forts with their
intervisibility brings new information about the strategic
function of the military organization along the Euphrates.
5. FILLING THE STRATEGIC GAP OF JEBEL BISHRI
It was discovered during the field survey that Nadra in the inner
regions of Jebel Bishri (see Fig. 3) offered sites with Late
Roman and Byzantine pottery showing that the plains had been
sparsely inhabited (e.g., site I 7 at Nadra: UTM 0551964,
3910907) during those periods. The people lived there along the
water ways. Through our prospections with QuickBird images it
has become clearer, how the central parts of Jebel Bishri a few
kilometers NW of Nadra and to the south from the Rasafa –
Circesium desert track (cf. Fig. 2), seem, in fact, to have been
militarily secured by a network of roads and forts that in their
extent may have housed several units of Late Roman legion
size. We also executed Sobel-analyses for contouring the forts
from the QuickBird images.
Fig. 10. Fort 1a on Jebel Bishri. QuickBird (© Eurimage 2003).
Fort 1a is a square, NE-SW oriented structure ca. 110 m x 110
m in size (see Fig. 10). The visible 5-7 equal-sized barracks
arranged around the inner face of the S parallel long wall fill up
the space south of the via principalis, which enters the fort from
SE through the porta principalis dextra. The structure is very
similar to the auxiliary fort of Eining on the Danube dating to
the Late Empire (cf. Southern and Dixon 2000, 134, Fig. 61).
However, this fort on Jebel Bishri with thick walls is larger in
size and there exist 5-7 more barracks to the north, so it seems
to have comprised altogether up to 14 barracks. The main road
of the fort is turning to the north into a second road just before
reaching the principal gate leading to another fort. There exists
an adjoining Fort 1b that is also NE-SW aligned, but it is almost
twice as large as its neighbours, being ca. 120 m x 220 m
covering an area of close to 3 ha. The adjoining fort appears to
contain 5-6 equal sized large barracks arranged in NW-SE
oriented rows. Similar arrangements are known, for instance,
from Ain Sinu I in Iraq (Kennedy and Riley 1990, 214) and Tell
Brak in Syria (Kennedy and Riley 1990, 215). At Ain Sinu there
is even an adjoining fort placed in the same way as Fort 1 on
Jebel Bishri. There appears also to be the remains of a third
structure (1c), aligned also NE-SW, but without any visible
room arrangements and survived towers. The outer walls are
visible and they cover c. 150 x 270 m, or about 4 ha.
Fort 2 to the east of Fort 1a-c is a large parallelogram (see Fig.
11). The size of the fort covers c. 90 m x 295 m, or c. 2.5 ha.
(The sizes of comparable forts with external towers are Ain
Sinu II, 3.7 ha, and Umm er-Resas, 2.2 ha). The via principalis
enters the centre of the long northern wall. In the centre of the
fort there are apparently the remains of the praetorium or
principia, in which the via principalis also ends. No remains of
barracks are visible. However, the fort has three well-preserved

________CIPA 2005 XX International Symposium, 26 September – 01 October, 2005, Torino, Italy________
external towers projecting from each corner. This fort type is
typically Late Roman and comparable with, for instance, Ulcisia
Castra or Castra Constantia on the Danube (cf. Southern and
Dixon 2000, 134). Examples of comparable towers are known,
for instance, from Betthorus (el-Lejjûn) in Jordan (Parker 2000,
128-130) and Qasr Khabbaz in Iraq (Kennedy and Riley 1990,
212) and appear to date to the 4th century AD (Diocletian–
Constantine?), judging from the architectural style. It is evident
that the forts were connected with Diocletian’s building
programs linking the Strata Diocletiana with Circesium through
the desert-steppe.
Fig. 11. Fort 2 on Jebel Bishri with visible defence towers on
the NE and SE corners. QuickBird (© Eurimage 2003).
The unpaved (surface cleared and marked by lines of stone)
roads leading to the forts have been bridged, the remains of
which are still visible across the wadi. The direction of the road
from Fort 1 appears to connect to the fort of Qebaqeb in the
south-eastern piedmont of Jebel Bishri (see Fig. 2). This Roman
road system went SE to the stronghold of Circesium. Forts 1-2
are an estimated 40 km from Circesium or equal to a 2-3 days
march for a normally equipped Roman infantry soldier.
6. CONCLUSIONS
The most important objective of this paper has been to draw
attention that the area of Jebel Bishri was more organized by
military installations and road networks in the Late Roman
period than most modern studies and encyclopedias of the
ancient world indicate. Security and defence works as well as
transporting were already begun by the Palmyrenes through the
western oases and along the Euphrates. According to the visible
remains, the major fortification projects on the way from Sura
to Circesium took place in the Late Roman and Byzantine
periods from the 4th century AD to the reign of Justinian I in the
6th century AD. According to our remote sensing studies the
central area of Jebel Bishri, although generally being referred to
as terra incognita, is also, in fact, an area offering
archaeological structures appearing to date to the 4th century
AD. The evidence shows that the way from the Strata
Diocletiana to Circesium through the desert was more militarily
secured by forts than earlier thought. The documentation of the
hitherto poorly known or the identification of unknown remains
underlines the need for redefining the significance of the
remains on Jebel Bishri for the defensive and military history of
the Roman eastern limes and the Roman army in Late Antiquity.
ACKNOWLEDGEMENTS
See, the acknowledgements in Lönnqvist et al., 2005b in this
volume. We would also like to thank Dr. Tønnes Bekker-
Nielsen for his comments.
REFERENCES
Chapot, V., 1907. La Frontiére de l’Euphrate de Pompée a la
conquéte arabe. Paris.
Chevallier, R., 1989. Roman Roads. Batsford Ltd. London.
GORS, 2002. Syria, Archaeology from Space. Syrian Arab
Republic, Prime Ministership, General Organization of Remote
Sensing. Damaskus.
Isaac, B. and Roll, I., 1982. Roman Roads in Judaea I, The
Legio-Scythopolis Road. BAR International Series 141. Oxford.
Kennedy, D., 1982. Archaeological Explorations on the Roman
Frontier in North-East Jordan. BAR International Series 134.
Oxford.
Kennedy, D. and Riley, D., 1990. Rome’s Desert Frontier From
the Air. London.
Lönnqvist, M. and Törmä, M. 2003. SYGIS – The Finnish
Archaeological Project in Syria. In: The ICOMOS & ISPRS
Comittee for Documentation of Cultural Heritage. Proceedings
of the XIXth International Symposium CIPA 2003. New
Perspectives To Save Cultural Heritage. Antalya (Turkey) 30
September–04 October, 2003. The International Archives of the
Photogrammetry, Remote Sensing and Spatial Information
Sciences, Istanbul, Turkey, Volume XXXIV, Part 5/C15, pp.
609-614.
Lönnqvist, M. and Törmä, M., 2004. Different Implications of a
Spatial Boundary, Jebel Bishri between the Desert and the Sown
in Syria. Proceedings of the XXthInternational Congress.
Istanbul 12th-23rd July 2004. The International Society for
Photogrammetry and Remote Sensing, The International
Archives of the Photogrammetry, Remote Sensing and Spatial
Information Sciences, Istanbul, Turkey, Volume XXXV, Part B,
Proceedings, Commission 3, pp. 897-902.
Musil, A., 1928. Palmyrena, A Topographical Itinerary.
American Geographical Society of Oriental Explorations and
Studies No. 4, Ed. J.K. Wright, Czech Academy of Sciences and
Arts. New York.
Parker, T. S., 2000. Roman Legionary Fortresses in the East. In:
Roman Fortresses and their Legions. Ed. Richard J. Brewer.
The Society of Antiquaries of London and National Museums &
Galleries of Wales, Trowbridge, pp. 128-130.
Poidebard, A., 1934. La Trace de Rome dans le Désert de Syrie.
Le limes de Trajan a la conquête Arabe. Recherches aériennes
(1925-1932). Haut-commissariat de la République Française en
Syrie et au Liban. Service des Antiquités et des Beaux-Arts.
Bibliothèque Archéologique et Historique, Tome XVIII, Vol. 1:
Text and Vol. 2: Atlas. Paris.
Poidebard, A. and Mouterde, R., 1945. Le Limes de Chalcis:
organization de la steppe en Haute Syrie romaine: documents
aériens et épigraphiques. Bibliothèque Archéologique et
Historique. Délégation général de la France au Levant 38. Paris.
Southern, P. and Dixon, K. R., 2000. The Late Roman Army.
Cambridge.
Stoneman, R., 1992. Palmyra and its Empire. Ann Arbor,
Michigan.