Greek fire
Greek fire was the Byzantine Empire's most fearsome secret, a burning liquid that did something no warrior had seen before: it kept burning on the surface of the sea. From roughly the seventh century onward, enemy sailors reported watching the water itself seem to catch light, as flames spread across the waves toward their hulls. The thunder that accompanied each discharge carried across any battle. The roar made it sound, in one surviving account, like thunder from heaven.
For centuries, rival powers tried to steal or copy the weapon. Bulgarians captured the metal tubes that projected it. Arabs seized an intact fireship. None of them could replicate what Byzantine crews unleashed. The formula died with the empire that made it. What ingredients burned on water? Who invented it? And why, after centuries of battlefield dominance, did the secret vanish entirely?
Kallinikos arrived at Constantinople around 672, a Jewish architect who had fled his home city of Heliopolis in Syria. His homeland had recently fallen to the Muslim conquests, and he brought with him something the Byzantines urgently needed. The chronicler Theophanes the Confessor recorded his achievement directly: Kallinikos had devised a sea fire that ignited the Arab ships and burned them with all hands.
The Byzantines were in serious trouble at exactly this moment. Long wars with Sassanid Persia had left the empire weakened, and within a single generation, Syria, Palestine, and Egypt had all fallen to the Arabs. By around 672 the Arab fleets were pressing toward Constantinople itself. Greek fire helped repel them at both the first and second Arab sieges of the city.
The historian James Partington thought it unlikely that any one person invented the weapon. He judged it was probably developed by chemists in Constantinople who had inherited the discoveries of the Alexandrian chemical school. A later chronicler, George Kedrenos, writing in the eleventh century, added a detail most scholars consider an error: he placed Kallinikos's home city in Egypt rather than Syria. Kedrenos also recorded the story, dismissed by modern scholars as implausible, that Kallinikos's descendants, a family called Lampros, meaning "brilliant," maintained the secret of the fire's manufacture down to Kedrenos's own time.
Emperor Constantine Porphyrogennetos wrote directly to his son Romanos II with a warning so serious he put it in his book De Administrando Imperio: never reveal the composition of Greek fire. His reasoning was theological. He claimed an angel had shown the formula to the first Christian emperor, Constantine, and had bound him not to prepare it for anyone but Christians, and only within the imperial city.
The warning was not merely pious. Constantine added a specific cautionary story: one Byzantine official had been bribed to hand some of the substance to the empire's enemies. As that man was entering a church, a flame from heaven struck him down.
The secrecy had structural depth beyond divine warnings. Knowledge of the entire weapon system was deliberately split among operators and technicians, each aware of only one component. That is why, when the Bulgarians captured the city of Mesembria and the town of Debeltos in 814, they took 36 of the projecting tubes and quantities of the substance itself, and still could not use either. The Arabs captured at least one fireship intact in 827, with the same outcome. A weapon that required total knowledge of every part to function was a weapon that could not be reverse-engineered through theft alone.
Byzantine sources agreed on four properties that set Greek fire apart from other incendiary weapons of the era. It burned on water. It was a liquid, not a projectile. Its discharge came with a roar and heavy smoke. And it could only be extinguished by sand, strong vinegar, or old urine.
The primary method of delivery was a bronze tube called a siphōn, mounted on the prow of a warship under the forecastle. Anna Komnene, writing in her Alexiad, described the beast-shaped projectors that Emperor Alexios I ordered mounted on each ship before a battle against the Pisans in 1099: brass or iron heads of lions and other animals, gilded over, with the fire passing through the mouths so that the creatures appeared to vomit flame.
Portable versions also existed. Emperor Leo VI the Wise claimed to have invented the hand-held cheirosiphōn, which translates literally as "hand-siphōn." It appears extensively in military documents of the tenth century and was recommended for disrupting enemy formations in the field as well as for use against siege towers. Nikephoros II Phokas specifically advised field armies to carry them.
For the largest tubes, the mechanism reconstructed by researchers John Haldon and Maurice Byrne required a bronze pump to pressurize the oil, a brazier to heat it in a sealed tank above, and a swivel-mounted nozzle. When pressure reached the right level, a valve opened and the mixture discharged, igniting at the nozzle's mouth. Haldon tested a full-scale version of this design in 2002 for a television program called Machines Times Forgot. Using crude oil mixed with wood resins, the device reached a flame temperature of over 1000 degrees Celsius and an effective range of up to 15 metres.
The first popular theory on what made Greek fire burn was the saltpeter hypothesis, anchored in the "thunder and smoke" descriptions and the apparent distance the flame could be projected. Chemist Marcellin Berthelot was among the most prominent nineteenth-century supporters of what historians called the "French school" on this question. That argument has since been rejected. Saltpeter does not appear to have been used in warfare in Europe or the Middle East before the thirteenth century.
A second theory pointed to quicklime. Some sources said water intensified the flames rather than quenching them, and quicklime reacts violently with water. The chemist Zenghelis tested this in 1932 and found the result of that reaction negligible in open sea conditions. Emperor Leo's military manual, the Tactica, also records Greek fire being poured directly onto the dry decks of enemy ships, which rules out water contact as a trigger.
Calcium phosphide was a third candidate, since it releases phosphine gas on contact with water and that gas ignites spontaneously. Extensive experiments failed to reproduce the intensity that historical accounts described.
Most modern scholars landed on petroleum, thickened with resins, as the most plausible base. The Byzantines had access to crude oil from naturally occurring wells around the Black Sea, including wells near Tmutorakan that Constantine Porphyrogennetos noted himself. A surviving Latin manuscript preserved at Wolfenbüttel in Germany, dating to the ninth century, identifies naphtha as the main component of Greek fire and describes the operation of the projecting tubes. The Praecepta Militaria refers to the substance as "sticky fire," suggesting resins were added to increase both duration and intensity of the flame.
The weapon's battlefield record stretched across several centuries. After stopping the Arab sieges of Constantinople, Greek fire secured victories during the Byzantine expansion of the late ninth and early tenth centuries. In 941, when Byzantine forces faced a Rus' fleet described as vastly more numerous, the fleet's engineers mounted projecting tubes not just on the prow but also amidships and at the stern, adapting the standard layout to meet the numbers against them. A second Rus' raid in 1043 was repelled by the same means.
Civil wars proved equally reliant on the weapon. In the revolt of the thematic fleets in 727 and in the large-scale rebellion led by Thomas the Slav between 821 and 823, the rebel fleets were defeated both times by the Constantinople-based central Imperial fleet using Greek fire.
The weapon was not, however, the naval equivalent of a decisive trump card. Naval historian John Pryor described it explicitly as not a "ship-killer" in the way the naval ram had been. Its siphōn-deployed version had limited range and required calm water and favorable wind to use safely. Muslim navies adapted over time by staying outside its effective range and soaking felt or hides in vinegar as a protective measure.
Anna Komnene's account of the 1099 battle against the Pisans stands as one of the last vivid firsthand-adjacent descriptions of Greek fire in confirmed Byzantine use. During the 1203 siege of Constantinople by the Fourth Crusade, improvised fireships appeared but no account confirms actual Greek fire was used. By that point, the empire had spent roughly twenty years in general military decline, and the Byzantines may have lost access to the regions where their primary ingredients were gathered.
The Memoirs of the Lord of Joinville preserve one of the most striking outside perspectives on a weapon called Greek fire, recorded during the Seventh Crusade in the thirteenth century. The fire was by then in Saracen hands rather than Byzantine, but Joinville's description of the effect remained vivid: the tail of fire trailing behind it was as big as a great spear, its noise like the thunder of heaven, its appearance like a dragon flying through the air, and the light it shed bright enough to see the entire camp as though it were day.
Whether that Saracen weapon was chemically identical to the Byzantine original is a separate question. A twelfth-century treatise prepared by Mardi bin Ali al-Tarsusi for Saladin records an Arab version called naft, which had a petroleum base with sulfur and resins added. The text treats it as a distinct Arab formula; any direct relation to the Byzantine compound is considered unlikely by scholars.
In the nineteenth century, an Armenian named Kavafian approached the Ottoman government claiming to have developed a new form of Greek fire. He refused to reveal the composition unless he was placed in personal command of its use in naval battles. Ottoman authorities poisoned him not long after, never having learned the secret. The episode is a near-exact echo of the original Byzantine situation, separated by over a thousand years: a weapon's power concentrated in a single person who trusted no one, and died for it.
Common questions
Who invented Greek fire and when was it developed?
Greek fire is attributed by the chronicler Theophanes the Confessor to Kallinikos, a Jewish architect from Heliopolis in Syria, who fled to Constantinople around 672. The historian James Partington considered it more likely the weapon was developed collectively by chemists in Constantinople building on the Alexandrian chemical tradition, rather than by a single inventor.
What was Greek fire made of?
The exact formula was never recorded and remains unknown. Most modern scholars agree it was based on petroleum, comparable in composition to napalm, likely thickened with resins to increase burning intensity. A ninth-century Latin manuscript preserved at Wolfenbüttel identifies naphtha as the main component.
How did the Byzantines deploy Greek fire in naval battles?
The primary method was a bronze tube called a siphōn, mounted on the prow of a warship, which projected the heated and pressurized liquid and ignited it at the nozzle's mouth. Portable hand-held versions called cheirosiphōnes were also used on land and at sea, and earthenware grenades filled with the substance were thrown by catapult.
Why did enemies fail to copy Greek fire after capturing it?
Knowledge of the weapon system was deliberately compartmentalised, with each operator or technician knowing only one component. When the Bulgarians captured 36 projecting tubes and quantities of the substance itself in 814, they still could not make use of them. The complete formula, manufacturing process, and deployment system all had to work together, and no enemy ever obtained all of it.
When did Greek fire fall out of use?
Greek fire continued to be mentioned in Byzantine sources through the twelfth century, with Anna Komnene describing its use against the Pisans in 1099. By the 1203 siege of Constantinople, no account confirms its use, possibly because the empire had lost access to its primary ingredients or because the secret had been lost.
How effective was Greek fire as a weapon?
Naval historian John Pryor noted it was not a "ship-killer" comparable to the naval ram, and its siphōn-deployed version had limited range and required calm water and favorable winds. Despite those limits, historian John Julius Norwich wrote that it is impossible to exaggerate its importance in Byzantine history, pointing to its role in repelling the Arab sieges of Constantinople.
All sources
17 references cited across the entry
- 2bookChronographiaImpensis E. Weberi — 1839
- 3webGreek FireLinda Alchin
- 5harvnbPryor, Jeffreys (2006) p. 507Pryor, Jeffreys — 2006
- 6harvnbPryor, Jeffreys (2006) p. 509Pryor, Jeffreys — 2006
- 7harvnbPartington (1999) p. 3Partington — 1999
- 8magazineThe Link: Greek Fire
- 9bookI Segreti della signoraIsabella Cortese — Giovanni Bariletto — 1565
- 10harvnbHaldon (2006) p. 297–315Haldon — 2006
- 11bookThe Brass ButterflyWilliam Golding — Faber and Faber — 1958-01-01
- 12webHonour Bright
- 13bookDemigods and Monsters Your Favorite Authors on Rick Riordan's Percy Jackson and the Olympians SeriesBenBella Books — 2013
- 15bookThe Science of Michael Crichton An Unauthorized Exploration Into the Real Science Behind the Fictional Worlds of Michael CrichtonBenBella Books — 2008
- 16news10 Real-Life Historical Connections In Game Of Thrones That You Never NoticedSam Hutchinson