Before it fell for good to the Ottomans in 1453, the city of Constantinople withstood something like a dozen sieges by foreign armies over its long history. The one successful siege, by the soldiers of the Fourth Crusade in 1204, was of course eventually rolled back when the Latin Empire went defunct in 1261. But other than that, all of them prior to 1453 were unsuccessful. Arab armies under the Umayyad Caliphate attempted two of those failed sieges. The first may have been made in 674-678 (this is generally considered the Arabs’ “first siege” of Constantinople, though it wasn’t really as cohesive as that description implies), when the Byzantines may have been saved by the advent of Greek Fire and its use in destroying an Arab naval blockade. The second probably began on this date in 717 and ended on this date in 718, and was the last time the caliphate really threatened the city.
I say “may have been made” and “may have been saved” when referring to the alleged first Arab siege of Constantinople because there are good reasons to believe that whatever happened in the 670s was not a siege of Constantinople per se and that, whatever it was, it was not ended with the first use of Greek Fire. We only have one account of a “siege,” from the 8th-9th century Byzantine historian Theophanes the Confessor, and modern historians tend to believe that he stitched a number of minor Arab incursions into one single major siege that never actually happened. Anyway, the Arabs were soon drawn into their second empire-wide civil war, so for some time the Byzantines actually were in the dominant position in this relationship and were even able to extract tribute payments from the Arabs. But then everybody’s fortunes shifted, and from the 690s through the 710s its was the Byzantines who were dealing with internal strife while Arab armies kept pushing deeper and deeper into Anatolia.
The problem the Arabs had in defeating the Byzantines once and for all was that supply routes over the Taurus Mountains were too difficult to sustain for any length of time, and the Anatolian populace wasn’t prepared to support an Arab occupation. So for most of the history of Arab-Byzantine conflict the two sides didn’t invade each other’s territory so much as they launched raids. The primary goal was to hurt the other side and bring back some nice booty, and when the campaign season started to wind down the raiders usually had no choice but to retreat back across the mountains into their own territory. This would be the state of affairs between the Caliphate (and its successor states) and the Byzantines until the late 10th century, when the resurgent Byzantines briefly won back control of northern Syria and Armenia. But then the Battle of Manzikert in 1071 opened Anatolia up to Turkish settlement. This is all well into the future in 717, of course, and so at this point the Arabs were very much looking to do away with the Byzantine Empire by capturing its capital.
This was, of course, easier said then done. There’s a reason why Constantinople survived as many sieges as it did and why it took the Ottomans, armed with massive siege cannons, to finally capture the city.
The Theodosian Walls, built in the 5th century, were so massive that it took 15th century siege guns to finally breach them. The Arabs of the eighth century, who were still new to siege warfare in general and obviously weren’t lugging around any 15th century siege guns with them, didn’t have a chance of breaching them and knew it, so the plan instead was to blockade the city by land and sea simultaneously and simply wait the Byzantines out. This was not, as we will see, a recipe for success.
It was during the reign of the Umayyad Caliph Sulayman (d. 717) that plans for a new siege of Constantinople really took shape, under the direct command of the caliph’s brother, Maslama b. Abd al-Malik (d. 738). But since a siege required a naval blockade, and a naval blockade required a navy, and the Arabs didn’t have much of a navy at this point, they were forced to construct a whole mess of ships. All that ship-building was pretty hard to keep under wraps. Byzantine agents picked up on the caliphal activities starting in about 715, and consequently the empire had lots of time to strengthen Constantinople’s fortifications and stock it with plenty of supplies. Their preparations were hampered, however, by the fact that the Byzantine Empire was still stuck in that period of instability that I mentioned earlier. Between 715 and 717, shortly before the siege began, the empire changed hands twice, from Anastasius II to Theodosius III and finally to Leo III (d. 741), who took power in March 717 and ruled for almost a quarter of a century after that. People who know something about Church history will recognize Leo as the emperor who ushered in Iconoclasm, but we’re not going down that road here. Leo is a key player in our story, though.
The Arabs knew that things were a little rocking in Constantinople, and it was by exploiting this internal strife that the Arabs hoped they could make the Byzantine Empire their own, if not directly then at least by putting a friendly puppet on the throne. Leo, who at this point was a mere general in the imperial army, apparently sent word to Maslama that he would be willing to make the empire a vassal to the Umayyads if Maslama were able to put him on the throne in place of Theodosius–who, it’s said, had been forced by the army to become emperor, and governed like a guy who had never wanted the job in the first place. Maslama presumably hoped to use Leo to create chaos inside Constantinople that the Arabs could use to take the city. Leo, it quickly became clear, intended to double-cross Maslama as soon as events warranted. Which they did in 716, when Maslama captured a fortress in central Anatolia and Leo promptly garrisoned it with his men and shut the gates to the Arabs. With campaign season almost over, Maslama had to retreat to Cilicia, the coastal pass through the mountains, and Leo was then able to make his way to Constantinople and convince the (probably thankful) Theodosius to abdicate.
The following year, Maslama, with his army and navy in synch, made the push to Constantinople, besieging the city from land on either July or August 15, 717, while his navy sailed into the harbor. Unlike the alleged siege in the 670s, we know in 717 that the Byzantines made great use of Greek Fire to destroy Arab ships, forcing their navy back out of the harbor and away to a safer distance. At that point, with the benefit of hindsight, it seems clear that the siege was over. The failure to secure a naval blockade left the Byzantines free to continue bringing in supplies by ship, which meant that a “starve them out” siege was doomed to fail. But then Mother Nature decided to compound the Arabs’ problems by making the winter of 718 particularly harsh, which meant that the Arabs ran out of food–even though they had reportedly brought a huge amount of food with them and even brought seeds and implements to farm more–and that disease ran rampant through their camps.
Arab sources relate that the army ran out of supplies not so much because of the harsh winter as because of an inexplicable con job by Leo. Leo supposedly sent word to Maslama that he wanted to uphold his end of the original deal and hand the city over to the caliphate, but he needed Maslama to prove to the people inside the city that his army meant business. Somehow, according to these sources, Leo convinced Maslama to send part of his army’s food supply into the city, to show his great generosity to the Romans, and to burn the rest, to show the Romans that his army wasn’t going to settle for a siege but was ready to launch an all-out assault. Maslama supposedly did as Leo suggested and, well, we know how that worked out.
Now, it is almost impossible to believe that any human being could be credulous enough to actually do what these sources say Maslama did. But as I say this is a story that’s related in Arab sources–sources that, although they were written much later and under the Umayyads’ successors, the Abbasids, don’t otherwise seem to have been that critical of Maslama. Roman sources don’t mention this at all, and it’s quite possible–even likely–that it didn’t happen, but in that case the question is why the Arab sources felt it necessary to include the story. Possibly they intended to portray Leo as a conniving backstabber and to explain an otherwise difficult-to-explain Arab defeat, and the affair only reads to us cynical moderns like a story intended to portray Maslama as though he were dead from the neck up. Maybe portraying Maslama as a dupe was the nicest thing Arab historians could do for him.
When the weather improved, an Arab supply convoy arrived from Egypt, but because almost every experienced Muslim sailor was involved in the initial expedition, this relief fleet was crewed by Christians with less experience in lengthy, trans-Mediterranean voyages and also less loyalty to the caliphate. Those ships in this fleet, or at least the ones that a) survived the trip and b) weren’t handed to the Byzantines by their defecting crews, were attacked by the Byzantine navy, again using Greek Fire, and most were destroyed. Likewise, a relief army sent by the new Caliph, Umar II (d. 720), was met and wiped out by a Byzantine force. The Arab position outside the walls of Constantinople became untenable. At some point they fought a battle against an erstwhile Byzantine enemy, the Bulgars, either because Leo had cut a deal with the Bulgars or because Arab foragers had made the mistake of crossing into Bulgar land during their foraging, and the Bulgars soundly defeated them at the cost of thousands of dead Arab soldiers. On August 15, 718, either one year or 13 months to the day from when the siege began, the Arabs packed up and got the hell out of Dodge.
Notice that I haven’t said anything about troop numbers here, and that’s because there’s really no reliable count. The Arabs certainly began the siege outnumbering the Byzantine defenders, but there are Byzantine accounts that have the Arab army numbering in the hundreds of thousands, which is insane for an eighth century siege army–the daily supply demands would simply have been too much to sustain. Even the casualty counts offered by European sources are in the hundreds of thousands, which is again ridiculous. The Arab army may have numbered in the six digits, but not much more than 100,000, and between the Bulgars, the hunger, and the disease it’s likely that very few of them made it back home alive.
If we can conceive of a scenario under which this siege could have been successful, which to be sure might be impossible, then we have to say that its failure was of crucial importance in shaping European history. I mean, if the Battle of Tours (732) was important in stopping the Muslim advance into western Europe, then surely the fact that the Byzantine Empire survived the eighth century was equally important in stopping a Muslim advance into eastern Europe–more important, really, because there’s no historical consensus on whether the Muslim army defeated at Tours was really an army of conquest and not just a raiding party. This 717 siege in particular was important in that it was the last time a caliphal army would seriously entertain the idea of capturing the Byzantine capital. From this point on the frontier was more or less set at the Taurus Mountains and the military conflict between the two empires become one of dueling raids. And those raids wouldn’t be carried out by the Umayyads. The failure of the 717 siege and its cost in blood and treasure helped put the Umayyads into a steep decline that ended with their eventual ouster in 750.
Hi, how’s it going? Thanks for reading; attwiw wouldn’t exist without you! If you’ve enjoyed this or any other posts here, please share widely and help build our audience. You can follow this site (and like, share, etc. its content) on lots of social media outlets. I’m now accepting guest submissions, just email me. Most critically, if you’re a regular reader I hope you’ll read this and consider helping this place to stay alive.