Pompeii is famous for how it ended: buried under several meters of volcanic ash when Vesuvius erupted in 79 AD, freezing the city mid-life and preserving it for nearly two millennia of archaeological inquiry.
But Pompeii has another violent history that predates the volcano by 168 years — and it has left its own marks on the city's walls.
In 89 BC, the Roman general Lucius Cornelius Sulla besieged Pompeii during the Social War, a conflict in which Rome's Italian allies revolted in an attempt to win Roman citizenship. Sulla's legions attacked the city's fortifications with artillery, catapults, and siege equipment. Pompeii eventually fell. The siege left damage still visible on the ancient walls.
Now, researchers analyzing that damage say some of the impact patterns are consistent with something remarkable: a polybolos — an ancient repeating bolt launcher capable of firing multiple projectiles in rapid succession without manually reloading between shots. The device, invented by Greek engineer Dionysius of Alexandria in the 3rd century BC, is sometimes called "the world's first machine gun."
If the interpretation is correct, it would be the first physical evidence of a rapid-fire repeating weapon used in documented ancient siege warfare.
What Is a Polybolos?
Before explaining the evidence, it helps to understand the device.
The polybolos (Greek: πολύβολος, meaning "many-thrower") was described by ancient engineers including Heron of Alexandria (c. 10–70 AD), who wrote about various mechanical devices. According to his descriptions and those of other ancient sources, the polybolos was a chain-driven repeating catapult. Rather than requiring an operator to manually insert a bolt, draw the string, and fire for each shot, the polybolos used a continuous chain loop mechanism to feed bolts, tension the bowstring, and release automatically in sequence.
The mechanism, as reconstructed by modern scholars and engineers, works roughly as follows:
- A continuous flat-link chain, driven by a rotating wheel, moves the main slider back and forth
- As the slider moves back, it draws the bowstring to the cocked position; a bolt drops from a magazine into the groove
- As the slider moves forward, it trips the firing mechanism and releases the bolt
- The process repeats with each rotation of the drive wheel
How fast could it fire? Ancient sources are vague, but modern reconstructions of the polybolos have achieved rates of several bolts per minute with a single operator turning the drive wheel — significantly faster than a conventional catapult requiring full manual reload between shots.
The polybolos was not, by modern standards, a machine gun — it was not automatic in the sense of self-powering, and its rate of fire was far slower than any modern automatic weapon. But in the context of ancient siege warfare, where individual artillery pieces required multiple crew members and significant time between shots, a device capable of firing repeatedly from the same crew position represented a qualitative tactical advantage.
What Did Researchers Find at Pompeii?
The analysis focuses on the tufa stone blocks of Pompeii's Samnite-era city walls, particularly sections in the northern and western portions of the fortification that faced the likely direction of Sulla's siege approaches.
Researchers observed clusters of bolt-impact scars — holes and impact marks consistent with Roman bolt-thrower projectiles — that share several unusual characteristics:
Density and clustering: The impact marks occur in tight clusters of multiple holes within small areas of single stone blocks. A conventional catapult or bolt thrower of the period would typically produce scattered impacts as crews repositioned and adjusted aim between shots. The tight clustering suggests rapid, repeated firing at a fixed point — consistent with a device that could fire multiple bolts in quick succession without repositioning.
Consistent trajectory angle: The angle of penetration across the clustered impacts is highly consistent, suggesting the weapon did not move between shots. Manual reloading of a standard bolt thrower typically introduces variation in aim; a mechanized repeating device would maintain a more constant firing position.
Projectile dimensions: The hole sizes are consistent with standard Roman bolt dimensions for the period — approximately 60–70 centimeters in length — but the clustering pattern does not match what researchers would expect from a battery of separate individual weapons firing simultaneously.
The researchers' conclusion: the impact pattern is most consistent with a single polybolos-type device firing multiple bolts in rapid succession at a fixed point on the wall.
Historical Context: Did Rome Actually Use the Polybolos?
The existence of the polybolos as a device is not in dispute — ancient texts describe it clearly enough that modern engineers have built working reconstructions. What is debated is how widely it was actually deployed in combat.
Several ancient sources mention various mechanical devices used in siege warfare, but direct references to the polybolos in battle are rare. Most of what we know about ancient siege artillery comes from engineering manuals (like Heron's) and from accounts of major sieges — the Macedonian siege of Tyre (332 BC), the Roman sieges of Carthage and Numantia (146 BC), the siege of Jerusalem (70 AD). In many of these accounts, the specific types of artillery are described only in general terms.
The polybolos is thought to have been invented around 270–250 BC by Dionysius of Alexandria, a military engineer working for Ptolemy II of Egypt. It appears in ancient engineering literature as a known device. Whether it was adopted by Roman military engineers for actual field use — as opposed to being an interesting experimental concept — has long been an open question.
The Social War siege of Pompeii (89 BC) would be a plausible context for its use: Sulla was one of Rome's most aggressive military commanders, known for using siege technology aggressively; the Romans had had over 150 years to adopt and adapt Greek engineering since the invention of the polybolos; and a siege of a fortified city — where the ability to concentrate bolt fire on specific sections of wall would be tactically valuable — is exactly the use case the device was designed for.
What the Skeptics Say
Not all archaeologists and ancient weapons historians are convinced. The criticisms fall into a few categories:
Alternative explanations for clustering: Impact clustering on a wall section can result from multiple causes — a crew deliberately targeting a single spot for wall-breaching purposes, an elevation where multiple weapons converged, or post-siege reconstruction that moved stones. Critics argue that the physical evidence, while interesting, does not conclusively rule out conventional explanations.
The silence of the sources: If Sulla's legions used a remarkable rapid-fire device at Pompeii, it is argued, we might expect some mention of it in the historical record. The Social War is relatively well-documented by ancient standards; its absence from any contemporary account of the siege is a point against the polybolos interpretation.
Preservation and interpretation limits: Pompeii's walls have been exposed to nearly 2,100 years of weathering, subsequent construction, earthquake damage (a significant earthquake struck Pompeii in 62 AD, 17 years before Vesuvius), and archaeological excavation since the 18th century. Impact scars can be degraded, merged, or misinterpreted. The researchers' interpretation requires assuming the current damage patterns reliably reflect the original impact patterns — an assumption that some scholars consider problematic.
Why This Discovery Matters — Even If Disputed
The debate over this specific evidence will continue. Archaeological interpretations of ancient weapon impacts are inherently probabilistic — without a written account from someone at the siege saying "we used a polybolos on the north wall," certainty is impossible.
But the broader significance of the research goes beyond whether the polybolos interpretation is ultimately confirmed.
Ancient military technology was more sophisticated than popular imagination typically allows. The standard image of Roman warfare — massed infantry with swords and shields — is accurate but incomplete. Rome's military dominance across three centuries was built partly on engineering superiority: siege towers, ballistas, onagers, assault ramps, advanced logistics. The Romans were enthusiastic adopters of enemy technology, particularly from the Greeks and Carthaginians.
The polybolos fits exactly that pattern. If a Greek engineer invented a repeating bolt launcher, and if it demonstrably worked in engineering tests, then the question is not whether Rome could have used it — they clearly could have, and had the resources and organizational capacity to manufacture and deploy it — but whether they chose to and left evidence of doing so.
The Pompeii research suggests the answer may be yes. And Pompeii itself — uniquely preserved, uniquely studied — is exactly the right place to find that answer, even if the evidence is, as archaeology so often is, frustratingly incomplete.
Somewhere in the volcanic ash that buried a city, there may be bolts that once flew faster than anyone in the ancient world could reload a weapon to fire them. The walls of Pompeii, scarred by an army that came before the volcano, may finally be telling us how.