On walls northern of Pompeiibetween Porta Vesuvio and Porta Ercolano, have been visible for decades ballistic impact footprints dating back tosiege conducted by Lucius Cornelius Sulla in the’89 BCduring the Social War (a civil conflict between Rome and some rebellious cities, including Pompeii). Alongside the classic circular cavities produced by spherical projectilesa group of researchers from the University of Campania Luigi Vanvitelli and the University of Bologna has identified a different type of trackdescribed in a study published in the journal Heritage: small cavities quadrangularapproximately 25–30 mm in diameter, arranged in a fan along an arc of a circle, with short and regular spacing.
These sets of footprints were the subject of the study published in February 2026, as part of the SCORPiò-NIDI projectfinanced by the Italian Ministry of University and Research. The thesis put forward by the authors is that these traces were produced by polybolos: one automatic repeating crossbowwhose invention is attributed to Dionysius of Alexandria, described in the 3rd century BC by Philo of Byzantium.
The relief was conducted with advanced digital instrumentation: terrestrial laser scanner for general context, close-up photogrammetry and structured light scanner for individual cavities. There morphology of the imprints, i.e. with a quadrangular profile, the radial arrangement, the volumes of material removed by the impact of the projectile comparable between one sign and another, distinguish the traces under study clearly from those produced by spherical projectiles (such as those of catapults) or by single darts. The attribution is not based on form alone, but oncombined geometry analysis, spatial distribution and compatibility with documented firing mechanisms.
The point of connection with the ancient sources is a passage from the Belopoeica (text describing the construction of siege engines) in which Philo describes the principal limit of polybolos: the darts they don’t dispersebut they follow a trajectory concentrated on a single point, tracing a narrow arc. This feature, presented by the ancient author as a tactical flawcorresponds to the configuration observed by researchers on the walls of Pompeii. The German general Erwin Schramm (1856-1935), who at the beginning of the twentieth century built a working prototype of the machine by replacing the wooden chain described by Philo with a bicycle chain, he had already detected this inability to disperse of bullets.
Three groups of footprints were analyzed as case studies. All three are interpreted as missed shots: the darts they did not reach the target and left their mark on the tuff blocks. The use of a rapid-fire weapon is justified precisely by the mobility of the targets (probably the Pompeian archers on the walls): a single artilleryman with a traditional weapon would not have been able to realign the shot with that precision between one shot and another, nor would it have been tactically rational to use more machines independent to hit a single target.
A historical-contextual element is added to support the hypothesis: in 96 BCSilla had held the role of governor of Ciliciaprovince close to Rhodescenter of excellence in military engineering and in artillery construction. Philo probably had contact with i master craftsmen of the rhodium arsenal. It is therefore conceivable that Sulla had access to Rhodian technological developmentsincluding a polybolos enhanced compared to the original model described by Philo more than a century earlier.
The authors also calculated the dimensions of the machine starting from the damage detected, using the proportional formula of Philo, who binds the diameter of modiolusi.e. the cylinder containing i torsion beams of the weapon, to length of the dart. The values obtained are compatible with the estimates published in the literature and with the results of simulations conducted by the mechanical engineering team involved in the project. The studio recognizes its own limits: no physical remains of the weapon were found, nor metal tips in the wall blocks, and the cavities present small dimensions measurement uncertaintiesdue to the wear and tear of time. It therefore remains a hypothesis, which will have to be verified through physical reconstruction of the machine and other ballistic tests.
