Subscribe now for full access and no adverts
At some point in the last few decades BC, Roman legionaries paused on the banks of the Mera River, to the north of Lake Como, in Italy. While the soldiers were there, some stones in the riverbed caught their eye. Although the rock is a distinctive one, which outcrops upriver in the Bergell region, the legionaries’ interest was not piqued by its geological novelty. Instead, they noticed that the natural river processes had smoothed these stones into an ideal size and shape to serve as slingshot. The soldiers were stocking up on ammunition. Even though the battlegroup was already well equipped with lead slingshot, their instincts were sound. Ahead lay a journey into hostile terrain. The soldiers would follow the Mera valley into modern Switzerland, and then make for the major Alpine pass at Septimer. We know – and the legionaries surely suspected – that they were marching towards at least one battle. That clash lay roughly 70km distant, presumably placing it a matter of days in their future as the soldiers hunted for projectiles on the riverbank. And when the fighting did erupt, legionary slingers would find themselves in the thick of the action.
This military expedition formed one part of Rome’s struggle to subdue the various groups living in the Alps to the north of Italy. These efforts started in the 30s BC and culminated in a successful campaign of conquest ordered by Augustus and spearheaded by Drusus and Tiberius in 15 BC. While this broad chronological framework is provided by the surviving ancient literature, specific details about the events playing out during this era are relatively sparse. Our knowledge of the legionary battlegroup as it moved into the Alpine region comes not from the pen of any Roman writer, but rather relics of their passing that have survived in the landscape for over 2,000 years. These telltale traces range all of the way from a handful of stray finds to sizeable military structures. Metal-detectorists and archaeologists were first alerted to the former presence of these troops a little over two decades ago. Following in their footsteps has culminated in the recently announced discovery of a well-preserved fortification straddling an Alpine ridge at a height of 2,230m above sea level. The ability to reconstruct a sizeable chunk of the route this battlegroup took, and witness a range of their activities while on the march, is little short of extraordinary.
An advance in archaeology
‘It all started in 2003’, says Hannes Flück, of the Archaeological Service of the Grisons, who is one of the three project leads, alongside Peter-Andrew Schwarz, Vindonissa-Professor at the University of Basel, and Thomas Reitmaier, head of the Archaeological Service of the Grisons. ‘A metal-detectorist with a permit to work in the Graubünden Canton of Switzerland went looking for old roads in the region. He investigated a valley called “Surses” in the local language, which provides a good connection between the lands to the south and the north of the Alps. The Septimer Pass lies at one end, providing a steep but relatively easy access point, while to the north the valley leads towards another important passage at the Lenzerheide. Searching an area of pasture at the northern end of the valley yielded two lead slingshots, which had been stamped with the names of two different Roman legions: the Third and the Twelfth. Roman lead slingshot is rare in Switzerland, so it suggested something out of the ordinary thereabouts. Even so, while examining the remainder of the pasture produced lots of interesting finds, there was nothing else from this early Roman period. Twenty years on, we still don’t know exactly why those two lead slingshot ended up there; perhaps they were simply lost on route.’
‘At around the same time that this was happening, another metal-detectorist was working illegally on the Septimer Pass. Roman material was also found there, including more lead slingshot – this time with the Third and the Twelfth Legions accompanied by the Tenth Legion – as well as hobnails from military marching sandals. Werner Zanier, an archaeologist, found out about this and a dig was undertaken by the Bayerische Akademie der Wissenschaften and the Archaeological Service in 2008-2009. This showed that when the Roman troops reached the Septimer Pass, they created a temporary camp: a type of fortification that provided secure overnight accommodation while on campaign. The camp was large enough to accommodate roughly 800 soldiers. Examining it also revealed that the soldiers were carrying both lead and stone slingshot. It is these latter that can be traced to outcrops at Bergell, in the Mera valley, where stones of about the right calibre occur in the watercourse.’
To the north of the Septimer Pass, the metal-detectorist who recovered the first pair of lead slingshot continued combing the valley. His persistence was rewarded when he encountered a concentration of militaria in an area of partially open woodland called Crestas and Burschignas. This lies to the east of a gorge, and forms the peak of a triangle of terrain that rises steeply from and dominates a narrow margin of relatively level ground running along the edge of the gorge. Today, this shelf of land is exploited by a modern road traversing the valley. The arms and armour found in the fields can be divided into two broad categories: those used by the Roman army, and those more likely to be wielded by local warriors. Among the Roman material there was once again a conspicuous quantity of lead slingshot bearing stamps representing the Third, Twelfth, and Tenth Legions. The finds were duly published by Jürg Rageth of the Archaeological Service, and linked to combat during Drusus and Tiberius’ campaign in 15 BC. Neither interpretation found universal acceptance.
By 2012, the material coming from Crestas and Burschignas had dried up and it was assumed that the site was exhausted. Not everyone was convinced, though, and a younger metal-detectorist visited the site in 2018. He realised that there was still Roman material in the ground and presented his results to the Archaeological Service, who – together with Peter-Andrew Schwarz – initiated a project to locate, excavate, and record the remaining finds. This was undertaken in collaboration with metal-detectorists organised in the AGP (Swiss Working Group for Archaeological Propection). It ran for three years from 2021 to 2023, with analysis of the results now under way. While the precise significance of the local weaponry and armour from the site is currently being discussed, pieces of scabbards, axes, lances, helmets, and shields were all recovered. As for the Roman military metalwork, roughly 3,000 pieces were found, including approximately 450 lead slingshot, nearly 30 tips from ballista bolts, up to 20 arrowheads, one or two parts of javelins, and 2,500 hobnails. Examining this material leaves no doubt that the Roman troops operating here came from the same battlegroup that camped on the Septimer Pass. It helps establish their motive, too, for barraging these fields with projectiles.
Alpine clash
‘When you look at the nature of the site and the distribution of materials we have, it is very clear that this is a battlefield’, says Hannes. ‘It was neither a fortified military camp nor a local settlement. We think that the location was chosen by local forces, who anticipated that the legions would pass along the shelf of lower ground between the fields and the gorge. The Romans could hardly leave an enemy force controlling such a position, so they would have been obliged to attack up the steeply rising ground. Either that, or the battle may have started lower down the slope, closer to the legionary line of march, with a second phase of fighting following as the locals retreated uphill.’
‘We’re not entirely sure who the local people involved in this clash were, but historians have traditionally associated this region with a group called the Suanetes. Their name does not come from any inscriptions found in the locality. Instead, it was inscribed on the Tropaeum Alpium, a victory monument that was erected in France and provides a long list of peoples defeated in the Alpine campaign, including the Suanetes. The Tropaeum was supposedly erected by the grateful Senate to celebrate Augustus’ accomplishments, but we can all guess who really gave the order to set up a monument commemorating himself. Anyway, while it is important to be cautious about assigning names to local groups in this way, “Suanetes” is a convenient label to use. We do know that people lived in the vicinity of the battlefield at around the time of the fighting. A prehistoric settlement lies just opposite on the other side of the valley at Salouf-Motta Vallac, while another site that is surely of this date is known just beyond the northern mouth of the valley at a place called Bot da Loz.’
When it comes to understanding the nature of the fighting, the lead sling-shot has so far received most study. Separating them out by legionary stamp and plotting their distribution reveals that most of those associated with the Third, Twelfth, and Tenth Legions were scattered across different parts of the battlefield. Some of these slingshot bear impact marks, showing that they had been launched, rather than simply dropped. The overall numbers reveal that the legions are not equally represented, with the Twelfth Legion responsible for both the majority of examples – some 260 – and also the most intense concentration of projectiles at the site. This occurred in an area that measured roughly 25m square and was named ‘slingshot terrace’ by the archaeologists. Compared to the slingshot, other projectiles such as ballista bolts and arrowheads were relatively rare. This is probably in part because they would have been easier to recover after the battle, except for those cases where the shaft had broken off. Experiments have demonstrated that slingshot would be much harder to find, with those that missed their targets lodging themselves deeply in the ground. It is surely no coincidence that the other finds present in large quantities, the Roman hobnails, would also be easy to miss by anyone who was looting the battlefield.
‘Most of the hobnails were found individually, where they had fallen off a shoe’, says Hannes. ‘They occur across the site, which shows that over the course of the fighting Roman troops were present in all parts of the battlefield. On top of that, to me, the distribution of the lead slingshot suggests an attempt to drive the local fighters back. One scenario we favour at the moment is that the Suanetes made a last stand on “slingshot terrace”, where they were on the receiving end of a concentrated barrage. I believe that the Romans most likely won this battle, but we will have to see what full study of the material reveals. The date is an important factor here. At the moment we have about 100 coins from the sites associated with these troops. The latest is a dupondius that shows these events couldn’t have happened before 28 BC. This makes it possible that the troops were part of the campaign of 15 BC; if so, I’m very confident that they won.’
Tightening control
We cannot be entirely sure how much further the battlegroup advanced following the fighting at Crestas and Burschignas, but they were certainly able to continue up the valley. One illustration of this comes from a further cluster of finds on the summit of Plang Ses. This striking natural feature rises sharply from the eastern side of the valley, a little to the north of the battlefield. Objects from Plang Ses include more slingshot, and even a fine example of a legionary dagger that was deposited on its summit. The distinctive style of this weapon allows it to be assigned to the period spanning the late Republican and mid-Augustan eras. Initially, the archaeologists thought that a small, fortified post had been established on Plang Ses, but excavation proved that a set of possible defences were natural features. Even so, it may be that the steep slopes leading to the summit were considered protection enough. The military interest in this feature was presumably motivated by the views that it offered, with the prehistoric settlement at Bot da Loz clearly visible to the north. That the Romans were interested in manning key points within the landscape has been confirmed by the most recent discovery associated with the legionaries’ progress.
‘A detectorist came up to us during our last season of work at Crestas and Burschignas in 2023’, Hannes recalls, ‘and said “I have to show you something”. That “something” turned out to be a standing earthwork complete with triple ditches on a mountain ridge at Colm la Runga, 1.8km distant from, and 900m above, the battlefield. Up until that summer, the LiDAR available in Switzerland only had a resolution of about 2m above a height of 2,000m. Then it was improved to 50cm, which meant that this earthwork, lying at a height of 2,230m, suddenly leapt out in the imagery. I immediately thought that it had a Roman or prehistoric vibe, because the triple ditches were so suggestive, but it could also have been First World War, Second World War, or something like that. So we decided that we needed to find out what this earthwork was. The detectorist visited the site with our co-project manager Thomas Reitmaier, and the first find was the point of a small lance or a big arrow, so it didn’t help that much. The third, though, was a Roman hobnail, and the sixth was a lead slingshot bearing the stamp of the Third Legion. At that point it was obvious that the site was connected to our project, so we decided to investigate further.’
The first step was geophysical survey, which revealed various anomalies, at least some of which are now known to be hearths. Two things that were conspicuously absent were any sign of internal buildings or a rampart. Opening trenches across the defences revealed that the latter was present, in the form of a wall created using carefully stacked turf blocks. It seems that this material was gathered immediately outside the fortification, where traces of cut marks were identified. By contrast, the grass within the rampart, which encloses an area of 700m², had been deliberately left in place. Pegs were found there, some of them still arranged in a row, indicating that the soldiers lived in tents. Excavation has commenced on a curious circular feature visible on the LiDAR, which has produced charcoal. Surprisingly, perhaps, given the range of finds from the site, it is the charcoal that the archaeologists are most excited about. They are hoping that it will be possible to get a dendrochronological date from this, identifying the specific year when all of the events described so far played out.
Raid or conquest?
Such a date should reveal whether the legionary presence amounted to a fleeting raid designed to punish troublesome inhabitants of the Alps or formed part of Drusus’ and Tiberius’ strategy to achieve a lasting conquest in 15 BC. Some initial inferences can, though, be drawn from the archaeology. In particular, it seems clear that the battlegroup had been carefully prepared for this operation, rather than rapidly assembled to respond to a surprise incursion by Alpine warriors. The lead slingshot provides one example of this. As the three different legions cast them using different moulds, we know that they were produced at unit level. At the same time, while Roman slingshot in general has produced a rich corpus of inscriptions, ranging from unit names to insults that were literally hurled at foes, the letters were normally cast in moulds. Stamping inscriptions on slingshot was much rarer, so the quirky use of this technique by all three legions making up the battlegroup would fit with the projectiles having been prefabricated in preparation for the same mission. This seems to be confirmed by the isotopic signature of the lead, which has been studied in collaboration with Southampton University. All three legions created their slingshot using lead from the same source, which probably lies in southern Spain.
Comparable care was invested in the composition of the battlegroup. ‘This force was specifically chosen for Alpine combat’, says Hannes. ‘While slingers are a feature of Late Republican armies, I think we can say with confidence that they were a good option in this terrain. Estimating the size of the force is also important. We know that three legions were involved, so a natural estimate would be 1,500 men, as that would be the equivalent of one cohort from each legion. At the same time, this number seems to be too large for the 800 or so soldiers estimated for the camp on the Septimer Pass. As a disproportionately large amount of the slingshot from the battlefield came from the Twelfth Legion, perhaps it was the only one present at full cohort strength, while the Third and the Tenth sent detachments of slingers.’
‘An alternative possibility is that only one part of the force was committed to the battle. The slingshot from the fortification on Colm la Runga just names the Third and Tenth legions, raising the possibility that some of them were providing flank protection on the ridge line, while their comrades were fighting nearby. It would be helpful to know where the construction of the fortification falls within the overall chronology of events, but we think it was probably established after the battle. Some possible loot from it was even found at Colm la Runga. If so, this raises the question of what the post was intended to achieve. It would have been time-costly to construct, something that is probably explained by it being intended for longer use than an overnight camp. There was also only space for about 80-100 men, meaning that it was far too small to accommodate the entire force.’
‘If we look for a tactical advantage that Colm la Runga offered, it must be the outstanding overview of the surrounding landscape. All of the key natural and human features are visible, allowing a careful eye to be kept on this region. It is hard to see why this would be much of an advantage for a punishment expedition. Although it sounds drastic, in those circumstances the force just had to go in, cut off heads, and leave. There would have been no need to start tightening control of the wider landscape, and establishing the post at Colm la Runga just reeks of control. That mindset is a much better fit with conquest. If so, it seems likely that these troops were involved in the 15 BC conquest, and their ultimate objective was Lake Constance, at the gateway to the lands stretching north of the Alps.’
Further Information:
• More details about the project can be found in German here: www.gr.ch/DE/institutionen/verwaltung/ekud/afk/adg/kooperationen/Seiten/CVMBAT-Surses.aspx.
• An article, also in German, about the findings is available here: https://doi.org/-10.5169/seals-1046324.
• The project Instagram is @cvmbat2025.
You must be logged in to post a comment.