What Greek Fire Really Was

by Brian Dunning
May 17, 2022

Today we're going to go way back into history, back to the Middle Ages, when Byzantine soldiers fought with their Muslim rivals. It was a time when a fearsome new weapon is said to have appeared: Greek fire, an ancient version of napalm, just as deadly and just as terrifying. Many questions surround it. What was it made of? And if it really was so effective, why did people stop using it in warfare? And finally the big question: Was it actually real, or just one of those exaggerated or misrepresented stories? Today we're going to find out exactly how much we know, and don't know, about each of these elusive questions.

First of all, let's clear up the obvious big one: Greek fire was real. It was not some legendary or apocryphal thing mentioned in some old story. Greek fire is known from reliable historical accounts, and appeared in the 7th century in naval warfare. It was sprayed onto enemy ships by a mechanical delivery system. Although its exact composition was lost to history, it was an improvement on naphtha, which was well known and had been used in warfare as an incendiary weapon for centuries.

Petroleum was well known to the ancients, as there are places where it simply comes up out of the ground. It could be found bubbling up as petroleum seeps, or as old solids that looked like solidified black lava. Either way it was simple to collect in limitless amounts, and just as easy to distill. The first fraction to boil off was naphtha, a clear, thin, yellow liquid that was highly flammable and very useful. Naphtha had been used in ancient warfare for a long time. You could dip incendiary arrows into it, or you could fill pots with it and launch them with a catapult — both of which were done for centuries. The second fraction was kerosene — very useful itself, and the third fraction was asphalt, also called tar or bitumen. Proper asphalt (actually a mixture of bitumen, sand, and limestone) had been used as a pavement as early as 625 BCE by the Babylonians, and bitumen as caulk for ships, baths, and aqueducts even earlier. But it was that volatile naphtha that captured the imagination of the ancient weapons engineers.

So, setting the stage for the appearance of Greek fire: Constantinople, during the 7th century, was the capital of the Byzantine Empire. It was a period called the Byzantine Dark Ages, when Constantinople was under constant threat of invasion and sacking by both warrior tribes and the Arabs. Control of the sea was important, as the city was defended by the famous Theodosian Walls on land and ships were needed to keep the city supplied. And always out there was the formidable Arab fleet. War was imminent.

Kallinikos of Heliopolis was a Byzantine chemist and engineer who was active during the 7th century. By all historical accounts, his was the formulation for Greek fire that was adopted by the Byzantine military in the middle of the century. But perhaps more important than the formulation was the delivery mechanism that he contrived. We know little about it, other than it was called a siphon, and squirted a stream of Greek fire from a bronze nozzle mounted on a swivel at the bow of the Byzantine dromon ships. The Greek fire was heated and pressurized, we presume below decks and using some kind of human-operated pump. Kallinikos' siphon is believed to have had a range of some 15 meters, give or take, which made it very dangerous for an enemy ship to maneuver within boarding distance. There are also records of handheld siphons that could be carried by a single soldier, operated by a single cylinder, something like a big syringe. Ships would also pour Greek fire onto their enemies' decks with buckets suspended from pivoting cranes.

It should be stressed that Greek fire and the siphons were not entirely new inventions; in fact, the emperor was already building "fire ships" to fight the Arabs at least two years before Kallinikos arrived. His were innovations and improvements, though historians (primarily Theophanes) don't tell us what the improvements were. Also, it shouldn't really be called Greek fire at all, actually. It wasn't Greek; Kallinikos having a Greek name (though he was from the Syrian city of Heliopolis, also a Greek name) is probably the reason that it stuck. The first name it was known by was Roman fire, and then various names like marine fire, liquid fire, artificial fire, and so on.

Its first test in real battle came in the First Arab Siege of Constantinople (674-678 CE). The Byzantine fleet sailed out and met the Arabs, and with their new weapon, routed them. It was a resounding success. The Kallinikos formulation was able to burn even in water; it was all but impossible to extinguish. It burned with a load roar and produced immense amounts of black smoke, both of which contributed to it being a weapon of terror which sent the Arabs fleeing. It was so effective that Emperor Constantine IV made its exact composition a state secret, and the Kallinikos family was given an exclusive contract to manufacture and supply it.

A few years later the Arab fleet tried again in the Second Arab Siege (717-718 CE), by some accounts bringing their own version of Greek fire and siphons which were relatively ineffective. The Byzantine siphons had been improved and again obliterated the Arab ships. This time they fled for good and were never again much of a threat.

Time and time again for more than five centuries, Greek fire proved decisive in battle. Historical accounts speak of its monstrous roar and smoke in battles throughout the region. And then, quite suddenly, all mention of it disappears from the accounts. Battles still happened, but historians no longer mentioned Greek fire. What happened to it?

The end of its line came during the Fourth Crusade, specifically when the Crusaders sacked Constantinople in 1204, but the exact reason is something of a mystery. Some historians have speculated that the Second Council of the Lateran of 1139, attended by something like a thousand Catholic clerics, banned the use of Greek fire as being too destructive. This would mesh relatively well with the disappearance of Greek fire from warfare accounts after the Fourth Crusade. However, nothing much like that appears in the thirty Canons that were enacted by the Council. Canon 18 prohibits arson, and Canon 29 prohibits the "murderous" use of bows and crossbows against Christians. It would take a very creative interpretation to combine the two to contrive that the Council intended to forbid the use of fire as a weapon against Muslims.

One leading theory is that since Constantinople was sacked, looted, and burned, the closely-guarded state secret of Greek fire's formulation was simply lost in the chaos. Whatever interest there may have been in recreating it was insufficient.

Another possible reason for the decline of Greek fire is simply that it aged itself out of effectiveness. Catapults were hardly precision weapons and siphons or tubes required that you get so close to your enemy that you'd have been killed by bowmen long before you got there, especially considering the improvements to the crossbow during the Crusades. It could be that ballistic weapons had advanced faster than Giant Squirtgun technology. Some authors have argued that the formulation was never really lost at all, rather than it was gradually supplanted by the gunpowder-fired cannon. However, that was still almost two centuries in the future, and two centuries of silence is a long spell to argue that one weapon system was gradually replaced by another. But of course, the absence of mention in the historical accounts doesn't prove Greek fire wasn't used, just that nobody saw fit to write of it.

Another possibility is confusion of terminology. The name Greek fire came from the Crusaders, who used the term generically for all incendiary weapons. The complex weapon system of Greek fire and the siphons and fire ships may not have been what the historians were referring to in many cases.

The secrecy surrounding Greek fire may have also been its downfall. Greek fire was a complete weapon system, not just the liquid. Some workers made the bronze nozzles; others made the pumps and heaters; others made the tubes; others made the liquid. Knowledge was compartmentalized so that if one worker fell into enemy hands, little would be given up. Some authors have argued that this complex chain of custody of the knowledge was a potential weak point in their ability to retain it, and it's possible that the knowledge was lost earlier than 1204, possibly much earlier, even a few centuries.

So, in point of fact, we don't know exactly why or when Greek fire went out of fashion. There is likely some amount of truth in each of the preceding theories. But this question pales compared to the big one we're here to answer today: What was Greek fire made of?

Countless historians and chemists have put forth their hypotheses, most notably John Haldon and Maurice Byrne who built their own hypothetical siphon for a 2002 National Geographic TV show. It fired simple crude oil mixed with wood resin. It worked, but there's no reason to suspect it was any closer to the real thing than competing ideas.

But before we get into the specifics of the formulation, I want to stress a concept we've talked about in other episodes, such as the one on who built Stonehenge and how. That's the scope of uncertainty. The scope of the uncertainty here, as it was in that case too, is actually quite small. Saying that we don't know what Greek fire was made from doesn't mean that it was some incredible, magical, powerful thing that we have no idea how it could have been done. It simply means that we don't have records telling us which of the many possible formulations it might have been. There are lots of compounds that can be mixed to create a substance with all of Greek fire's known characteristics, compounds which were all in familiar use at the time. So there is a mystery here, but it's a relatively small and unimportant one.

There are two basic schools of thought on the composition: one that it contained saltpeter, and the other that it did not. Saltpeter, as an oxidizing agent, would have made it burn more explosively, and in fact would have made it the first gunpowder weapon. All the hypotheses fall into one of these two camps. The most popular common theme to the hypotheses is that it was primarily naphtha with wood resin as a thickener. Quicklime may have been used to help it combust more spontaneously and to burn underwater, but there are also sound arguments it was not used. Sulfur was used in the known formulations of other incendiary weapons of the day, including the Arab version of Greek fire, so it may have been included here as well. It's even possible that the exact composition didn't matter too much, and more than one may have been used. The fact is we just don't know, and another fact is that it just doesn't matter all that much.

And so, there we have our Skeptoid-sized survey of what we know and don't know about Greek fire. It was indeed a fascinating system, but some of the claims about the mysteries surrounding it are definitely overblown.

By Brian Dunning
Follow @BrianDunning