An inspection of the Limes of Africa Proconsularis

In March 2014 I was visiting Tunisia with a group led by an experienced archaeologist (he has excavated in Italy and Libya). Our time was much taken up with visits to Roman towns and cities. However, thanks to UK Foreign Office advice regarding a few areas considered dangerous to visit, we had to rearrange our route slightly. Luckily, this did allow us to search out the 'Limes' (pronounced Leem-ez) - the Roman frontier installations- and do a little bit of on the hoof archaeology. The following pages feature the places we visited with notes and pictures.

After I returned, I was able to consult an excellent book on the subject: 'Tripolitania' by David J. Mattingly (son of the famous Harold Mattingly) which provided more information on these southern Limes.

Tunisia, along with parts of Libya, occupies roughly what was once the Roman province of Africa Proconsularis. It can be divided into two halves. The north is green and fertile and mountainous to the east, watered by weather systems from the Mediterranean. To the south is desert. The changeover point is the east-west, Cherb range of hills/mountains (see map). South of those mountains are the 'Chotts'. These are vast salt lakes or pans that stretch from Algeria, across Tunisia almost to Gabes on the coast. The area of the largest, Chott el Jerid', is 7000 sq. km. They mostly resemble vast areas of flat, wet sand. There is more open water in the winter when the evaporation is lowest.

Part of the Chott el Jerid with the Cherb range in the background.

Although the Chotts are sometimes passable with caution, they would have been a formidable barrier in ancient times. As well as the Cherb mountains to the north, there is a range to the south of Chott el Fejej called Jebel Tebaga, and a north-south range, which includes Jebel Demmer and the Matmata range, that separates the coast from the desert. To the west is the great sand sea, the Grand Erg Oriental.

The collective name for the Roman walls, ditches, roads and forts in this area and stretching down into Libya, is the Limes Tripolitanus. The function of these devices was not to present an impassable barrier to movement but to control movement between the desert and the populated and fertile areas including the coast as well as creating a hegemony over tribes living on the margins of the desert. Many of these tribes were nomadic, travelling from oasis to oasis with their herds most of the year. But at the end of summer they would transfer their herds to the Chott el-Djerid area, where they worked gathering olives.


Map of Tunisia

This is a close-up map of central Tunisia, at the northern end of the Limes Tripolitanus

The numbers refer to the places visited, which are described in individual posts below.

1. The walls at Bir Oum Ali

2. The Fort Benia Guedah Ceder

3. Anonymous desert lookout tower

4. The walls at Degache

The walls at Bir Oum Ali

The Cherb range of mountains provide a barrier between the fertile north of Tunisia and the desert south and in Roman times also separated the territory of the Capsitani and the Nybgenii. To control movement between north and south, the Romans built walls, known as Clausurae (meaning 'the fortification in a defile), in passes and wadis at several places (but not continuously). The best preserved are at Bir Oum Ali, on the road from Gafsa to Kebili. The fact that this is a modern road means that much of the wall and the gateway that allowed passage through the walls have been removed. However, there are remarkably preserved sections of wall to the north and south of the road as we discovered.

Southern section of wall from the south side.

The wall stands up to 6m in places and is approximately 1.5m wide. It consists of two faces of small neat masonry with a rubble and mortar fill. The 'bricks' for the masonry would have been available all around.

Southern section of wall from the north side.

A section through the wall showing the 'walkway'. The walkway is 0.75m wide and the parapets 0.6m high. Apparently the floor and walls of the walkway were once plastered. Clearly, as a defensive feature, the walkway wouldn't have provided much protection and it has been suggested that it may have had another purpose, perhaps to channel water, although it is difficult to see where from.

Higher up the steeper slopes, the wall is topped by a triangular cap.

The section of wall to the north and the modern road in the foreground. A gate once stood where the road is and a possible cistern for water storage has been identified.

An archaeological conversation.

There are up to 9 other remains of clausurae in the Cherb range, dating to the beginning of the second century. Rather than being defensive, it is believed that the walls and gates controlled movement between the pastoral desert and the settled agricultural north.

The Fort Benia Guedah Ceder

Between the range of mountains called Jebel Tebaga and the Matmata Hills is what is known at the 'Tebaga Gap'. It provides an easy passage from the desert to the coast. Even in modern times it has had strategic importance. In 1943 it was part of the Mareth Line, defended by Germans and Italians against the Allies. In March 1943 the Tebaga Gap was breached by New Zealand and British forces.

In Roman times, the gap was closed by more clausurae, about which more below. In the middle of the gap is a well-preserved fort, called Benia Guedah Ceder. Like most ancient remains in Tunisia, it was first discovered (from a European point of view) during the French colonial period. Some excavation and possibly some reconstruction has taken place.

View of the fort looking east.

The style of the fort indicates fourth century, but pottery shards seem to indicate an earlier date, possibly third century. The size is 60 x 40m. The walls are built of good quality ashlar masonry, 0.60 - 0.80m thick. There are square bastions projecting at the south, east and west corners.

Interior of the fort. One of the interior rooms was a stable, with water troughs.

Plan of the fort.

The only gate into the fort on the southeast side. There was a chicane upon entry (see plan) which is clearly evident. Note the lintel supported by one small brick!

Internal doorway looking somewhat precarious.

View looking west(ish)

Another view.

The Jebel Tebaga clausura has been recognized as a 17km linear earthwork from Gebel Tebaga to the foothills of the Gebel Melab. This consisted of walls on the scarps of the Jebels and ditch and bank across the valley floor. The maps in 'Tripolitania' seems to show this clausura passing adjacent to the fort. On our visit, this was not immediately obvious. We did pass through a couple ditches in our 4x4's but they appear to date from WWII. Clearly the wall in the foreground of the picture above is not part of a clausura, but an external building, possibly a Byzantine addition.

There is a sister fort close to the same clausura called Benia bel Recheb. There are also several forts stretching to the south along the Limes Tripolitanus.

Anonymous desert watch tower

Interspersed between the main forts of the Limes were manned lookout towers. These had the job of watching out for any unusual movements and reporting to the nearest fort by signal or horse.

Of course, the location of the remains of these towers is not always obvious. However, the skilled archaeologist will always ask at the local pizza joint.

This establishment does have a water closet, but you have to bring your own water.

As can be seen from the next picture, this 'tower' resembles more a pile of stones. However, the evidence is there to be found.

Firstly, note the upright square stones that form the perimeter of the tower.

The remains of a dry stone wall.

This appears to be a cistern for storing water.

The cistern again. Note the lump of dressed stone.

A piece of dressed stone similar to the one in the cistern. Probably part of a doorway.

Pieces of terra sigillata pottery verify Roman use.

There is evidence of a midden (rubbish heap) outside of the presumed entrance to the tower. Also a short distance away, there seems to be evidence of a few dwellings.

The walls of Degache

If the Tebaga gap channelled movement on the east side of the country, then there was a gap on the west side which must have provided easy passage south to north. This gap is that between the Chott el Jerid and the Chott el Gharsa and to the west of the eastern tip of the Cherb range. The main town in this gap today is Tozeur. It sits around a large oasis which today has growing in it over 200,000 date palms. The Roman town was called Tusuros. In the medieval period it grew rich by acting as the terminus for trans-Saharan caravans.

The area must have been protected during the period of Roman rule, but it doesn't appear in any accounts of the Limes and no reports from the French colonial period appear to have been made (well, I haven't scoured the archives of the Bibliothèque nationale, but as far as I know). We were told of some walls at Degache, so we set off to investigate.

Degache is just a few kilometres to the northeast of Tozeur. It has its own oasis. The nearby Roman town was called Thegis.

In the middle of the oasis we found this rather impressive wall.

It was difficult to estimate the length of the wall in the time available, but it may have been up to 100m long. Certainly too long to be part of a building. The maximum height left standing is a couple of metres. As can be seen, it was made with large blocks of cut sandstone. These would have had to have been brought from the mountains, some km away.

Typically Roman pottery shards were found in the vicinity. Our resident archaeologist was satisfied that this is indeed a Roman wall. Whether it was part of the Limes or was built for some other purpose is difficult to say.

Another part of the wall.

Yet another part of the wall

This interesting platform sits close to the wall on the 'barbarian' side. Steps seem to be cut into it. Maybe for an impressive statue? Or maybe for a tower type tomb?

Another view of the platform

Well, that is the end of this little archaeological adventure. I hope has been of some interest.