The story is one of the durable physical-mechanical legends of classical antiquity: during the 213 BC Roman siege of Syracuse, Archimedes deployed an array of polished bronze mirrors arranged along the city’s seaward walls; the mirrors focused the Sicilian summer sun onto the wooden hulls of Marcellus’s ships at the harbour mouth; the ships ignited and burned to the waterline. The Roman naval assault was defeated by the focused thermal energy of sunlight.

The story is probably not true. It is a useful test case for the way ancient stories accumulate physical detail across centuries of retelling.

What the early sources say

The earliest surviving source on the actual siege of Syracuse is Polybius, writing approximately 70 years after the event. His detailed account of the Archimedean defences mentions the catapults, the cranes, the inverted-grapple Claws of Archimedes that lifted Roman ships out of the water and dropped them prow-down — but says nothing about mirrors or fire-from-sunlight. Livy and Plutarch, writing in the early Imperial period, give similar detail and similarly mention no mirrors. The mirror story does not appear in any source within four centuries of the event.

The first surviving reference is a single passing line in Lucian of Samosata (writing around 165 AD, approximately 370 years after the siege), in a satirical dialogue. Lucian mentions that Archimedes “set fire to the enemy ships by means of his art” — without specifying what art. The mirror reading is a later inference.

The first detailed mirror description is by the 6th-century Byzantine architect Anthemius of Tralles (the chief designer of the rebuilt Hagia Sophia in Constantinople), in a treatise called On Remarkable Mechanical Devices composed around 530 AD — about 740 years after the siege. Anthemius proposes that Archimedes used a configuration of 24 large hexagonal flat mirrors, mounted on movable supports, that were collectively aligned to converge sunlight on a single distant target point. The 12th-century Byzantine compiler Tzetzes repeated the Anthemius description with elaborated detail. The story enters the Renaissance European tradition through these Byzantine sources.

Does it actually work?

Several modern attempts to test the device have produced ambiguous results. The Greek engineer Ioannis Sakas conducted a 1973 reconstruction at Skaramangas naval base using 70 polished copper mirrors aimed at a stationary wooden mock-up at 50 metres — and successfully ignited the target in approximately 90 seconds. The 2005 MIT student project (a class exercise under Professor David Wallace) repeated the experiment with 127 1-square-foot mirrors and achieved ignition of a stationary wooden boat at about 30 metres in approximately 10 minutes of full Massachusetts October sunlight.

The MythBusters television programme tested the same scenario twice (in 2006 and 2010) and failed to produce ignition under more realistic operational conditions: a moving target, partial cloud cover, and the practical challenge of holding a large array of mirrors precisely co-aligned for the required ignition period.

The substantial physical objection is operational rather than thermal. The technique requires a large mirror array (probably hundreds), precise individual aiming, several minutes of held convergence on a single point of the target ship, and full sunlight at the right angle. Ships under attack are moving. Cloud cover during the actual 213 BC siege would have been intermittent. The substantial defenders on the Syracuse seaward walls would have been under projectile attack from the Roman ships throughout the engagement.

The catapults and the grapples that Polybius records (and that Marcellus would have remembered painfully) were operationally trivial by comparison. A defender who had the engineering capacity to build mirror arrays would have built catapults — which is what the Polybian record actually attributes to Archimedes.

What probably happened

The most likely answer is that Lucian’s brief 2nd-century mention of “art” referred to incendiary projectiles — burning pitch or naphtha, launched from the catapults that the earlier sources record. The Byzantine elaboration of the story into the mirror-array reconstruction was probably a tribute to Archimedean optical-geometric work (which Anthemius would have known through the manuscript tradition that eventually fed the Constantinople palimpsest) rather than a recovery of a genuine 3rd-century BC engagement detail.

The mirror story is a useful example of how classical technical legends accumulate. By the time Anthemius was writing, the Archimedean reputation in optics and geometry was secure; the mirror story was a plausible elaboration consistent with the secure reputation; once it entered the Byzantine compilations, it became part of the standard image of Archimedes for the European Renaissance. Whether or not it ever happened, it tells us something true about how Archimedes was remembered.