The Sieges of Alexander the Great Read online

  It seems likely that Alexander had scouted the city before he arrived as he immediately made camp opposite what he perceived to be the weakest point of the fortifications, and set about rebuilding the siege equipment that he had brought overland from Tyre. The siege engines were evidently capable of being flat-packed for transportation, a clever innovation which would have greatly increased the movement rate of the baggage train. Also immediately upon his arrival he constructed some form of circumvallation around the city, although the details of this will become apparent as we examine the detail of the siege. Behind his lines, and apparently unseen by the defenders, Alexander also began to dig a number of mine shafts in the direction of the city with the intention of undermining and then collapsing the walls. The geology of the area particularly lent itself to this form of siege technique, a technique that Alexander was using here for only the second time as the terrain he had encountered previously on his sieges had been largely rock and entirely unsuitable for sapping. The ground around Gaza was soft sand with no large rock outcroppings, apart from probably the mound upon which the city stood, that would impede the progress of a tunnelling operation. Despite the sand being soft on the surface, it was evidently firm and compacted a few feet below ground, enough that the tunnels did not cave in whilst shafts were being dug. The fact that sapping was begun immediately also tells us that the mound upon which the fortress stood was not of any great size, although still notable for its defensive qualities. If the tell had been as high as it is today, the tunnels would have been too deep beneath the walls for them to have been affected by their collapse once they had been fired.160

  Once Alexander arrived at the gates of Gaza he made camp and began preparations for the coming siege. His first order was for the construction of what Arrian describes as a turf wall around the city.161 This turf wall can be seen as a forerunner to the very common Roman practice of circumvallation, something that Alexander was to use whenever he had the opportunity. The most puzzling aspect of this particular offensive fortification is the question of where Alexander got enough turf to construct a circumvallation that must have been in excess of 3km in length, considering that Gaza was located in a desert and surrounded by sand. One possible answer is that the area was perhaps more fertile in the fourth century, and thus there was more turf than there is in that region at present. Upon consideration of the actual events of the siege, this seems extremely unlikely given that we hear of the siege engines sinking into the sand as they were brought up against the walls. If the circumvallation was indeed made of turf, it seems most probable that the turf either came from some distance away, or that Arrian simply made a mistake in using the wrong terminology (i.e. the circumvallation was not made of turf). The latter option seems by far the most probable to me as circumvallations in Arrian’s day would probably have been largely made of turf, particularly in Europe; this is probably a simple error by Arrian and nothing more.

  Alexander was never a man who was at ease with having men standing idle; at the same time as the circumvallation was being constructed by the troops, his engineers were rebuilding the siege engines that the army had carried from Tyre. The engineers themselves evidently advised Alexander to wait until the bulk of the siege equipment was brought by sea from Tyre, but Alexander ordered them to press on with what they had available, probably the smaller catapults that were more quickly dismantled and more easily transported by cart.

  The defensive levee, we are told, was of a size that would allow the catapults and various other siege machines to be at the same level as the city walls:162

  The plan of campaign was to enable the siege engines to be brought to bear upon the defences by ringing the town with a raised earthwork up to the level of their base, and mounting the engines upon it.

  Arrian is the only one of our sources to stress the size of the walls as being the major difficulty; Curtius chooses to concentrate his attention on the sapping operation, which is only mentioned in passing by Arrian. The reality of the situation was probably that neither ancient source is correct to place stress on only one element of the attack. Throughout his career Alexander would always attack from more than one direction. This was his most frequently used tactic and indeed one of his hallmarks. Both the walls were assaulted directly by catapults, as in Arrian, and the sapping operation was undertaken with equal vigour, as in Curtius. Each historian is simply using a different primary source that placed greater or lesser emphasis upon each of the two elements, ignoring the other almost entirely.163

  Arrian also tells us that the levee was highest opposite the southern city wall, the area perceived to be the weakest of the entire defensive circuit. When the circumvallation reached the correct height, the siege towers and battering rams were brought up for an assault on the city walls. This is a strong indication that the construction of the circumvallation took a considerable length of time as the siege towers would have been far too cumbersome to have been carried overland, and therefore must have been transported south by ship and dragged the final 4km across the desert. The siege lasted from September to November 332, and we can easily imagine the initial stage of the siege (the construction of the circumvallation, the delivery of the siege towers and the initial sapping operations) as taking up four to six weeks of that time.

  There are many problems with the accounts of the early part of the siege. Curtius must be incorrect in stating that the assault on the walls was a diversionary tactic; although this theory would certainly fit well with Alexander’s general plan of attacking at various points at once. There would, however, have been no need for a diversion at this time; the point was to collapse the walls, and the more defenders that were in that area when the walls did collapse, the better for Alexander. The attack must have been a serious attempt by Alexander to carry the siege and prevent the work required on the mining operations. The fact that Alexander did not wait for the bulk of his siege engines before commencing sapping operations is either an indication of impatience or a sign that the walls were not as strong as Arrian would have us believe.

  The second main problem with the accounts of the first stage of the siege is the circumvallation itself. We are essentially asked to believe that siege towers and battering rams were pulled up to the top of the circumvallation and then down the other side, and that this was seriously seen by Alexander as a viable way to attack the fortifications. It is possible that the mound was only constructed against the southern wall and was not a circumvallation at all.164 The only supporting evidence for this theory is that Curtius makes no reference to the mound at all during the early narrative of the siege. This is not wholly surprising, however, given that he chose to concentrate his account on the sapping operations. I think this theory can be rejected partly because of this, and partly because it was Alexander’s general practice to build a circumvallation where the opportunity presented itself. On the northeast frontier and in India circumvallations are frequently reported, even when the fortress was far less formidable than Gaza. Alexander’s primary thinking was likely the defence of his camp from possible sorties as he had experienced at Halicarnassus and Tyre.

  Whilst it is possible that the mound was only against one wall, it is contrary to Alexander’s strategy at almost every military encounter of his career, that of attacking from multiple directions simultaneously. It would also have been supremely dangerous, as the tunnel would have collapsed the mound as well as the wall. There would have been no way to collapse only a small section of the tunnel. The only possible support for the argument might be to suggest that the mound was built to protect the mining operation from the prying eyes of the defenders, and potentially to attract more of them to that stretch of the wall, so that when it did collapse, more of them would be killed, but this is an untenable argument.

  The reality of the situation is almost certainly that the circumvallation was actually built with frequent and regular gaps to allow the siege engines to be dragged along level ground between the raised sections. If this was
the case, it also explains how the entire city was surrounded with a mound as high as the city walls in probably only four to six weeks. It seems more plausible that a number, perhaps a large number, of individual mounds were built, upon which the catapults were stationed. This would have looked like a crenellated circumvallation and would be a very clever solution to a potentially difficult problem facing the besiegers of how to get the towers and rams to the walls.

  After the initial stages that any siege must undergo, preparation before the assault began. Alexander was ready to commence siege operations proper. We can only imagine the fear and anticipation that the defenders felt at the prospect of the inevitable assault as the circumvallation gained in height, and catapults became a more common site on the highest mounds.

  After weeks of preparations, the assault was finally ready to begin; Alexander was about to offer sacrifice to the gods as was the usual practice when a crow flew overhead and dropped a stone upon Alexander’s head. Aristander, Alexander’s Egyptian seer, interpreted this favourably (again, as was usual) stating that the city would fall, but also offered the warning that Alexander should take care of his own personal safety that day. The story is only of real interest because of what Plutarch reported was the fate of the bird after delivering its payload successfully: the bird became entangled in the cords of a torsion catapult.165 This gives some indication of the complexity of ancient torsion catapults if nothing else.

  Curtius’ account is the more interesting, however. He states that the bird landed on the nearest siege tower and that its wings became stuck on the surface, a surface that had been smeared with pitch and sulphur.166 The question is why would a siege tower be smeared with pitch and sulphur? These chemicals were used in the ancient world as incendiaries, not as flame retardants with we might reasonably expect Alexander to have covered his siege towers, especially after their destruction by fire at Tyre. This exact combination of chemicals had been used by the Spartans against the Plataeans during the siege of that city in 429. Alexander was something of a student of military history and was always one to recycle a good idea. Indeed, this very siege may have formed part of the model for Alexander’s attack on Gaza as the Spartans also made use of mounds of earth and built a circumvallation.167 Along with the siege of Plataea as a precedent, the Peloponnesian War perhaps also provides an explanation for the pitch and sulphur on Alexander’s siege towers. Thucydides provides this climactic description of the fall of the Athenian-held fortification at Delium in 424/3:168

  The Boeotians took the fort by an engine of the following description. They sawed in two and scooped out a great beam from end to end and fitted it together again like a pipe. They hung by chains a cauldron at one extremity, with which communicated an iron tube projecting from the beam, and this they brought up on carts to the part of the wall composed of vines and timber and inserted huge bellows into their end of the beam and blew with them. The blast passing closely confined into the cauldron, filled with lighted coals, sulphur and pitch made a great blaze and set fire to the wall and made it impossible for the defenders to remain at their posts. They abandoned their positions and fled; and so the fortifications were captured.

  It seems likely to me that Alexander still had fresh in his mind the effect of the Tyrian fire ship against his own siege towers only a few months before. He had seen at first hand how effective fire could be when used correctly and he used the existing precedents to his own advantage at Gaza.

  There is only one possible reason that Alexander would cover a siege tower in pitch and sulphur, and that was because he intended it to burn. Alexander wanted the first act of the siege to be a recreation of the fall of Delium and obviate the mines before they were complete. There is no other conceivable explanation for preparing his siege towers to burn at the slightest encouragement. In this act, Alexander shows himself capable of learning the lessons of history, and of trying something innovative.

  Despite the incident involving the raven, and Aristander’s response, Alexander finally opened the siege proper by ordering the siege engines forward. The opening of the siege turned out to be something of an anti-climax after weeks of build-up; the sand proved to be far too soft and the engines quickly became bogged down in the loose and shifting ground. This hazard should have been anticipated by Alexander’s engineers, but evidently was not and this failure left the attackers in a vulnerable position.

  It seems probable that Alexander had intended a strong initial assault on the city walls that he hoped would carry the day without the need for a protracted siege as at Tyre and Halicarnassus. With the Macedonians under constant missile bombardment from the defenders, however, and unable to respond in kind because of the difficulties of bringing enough engines close to the walls, Alexander changed his strategy and evidently held back the bulk of the quality troops.169 At the same time as holding back these quality troops, those who were in the forefront of the assault were ordered to withdraw and, wherever possible, to rescue the beached siege engines. As the Macedonians were struggling to execute the order for withdrawal, whilst also attempting to extricate their siege equipment from the desert sands, Batis ordered a sortie from the city, just as Memnon and the Tyrians had done before him.

  Alexander’s adaptability during this siege is quite brilliant and is hinted at in Arrian. Alexander quickly realized that his initial assault would fail because of the siege engines becoming bogged down, and that he would suffer heavy losses if he continued the siege without them. His change of strategy was essentially twofold: he seemingly expected, or anticipated, a counter-attack from Batis when the defenders saw the difficulties that the Macedonians were in. This form of sortie was a well-practised defensive tactic and one that Alexander had seen several times during his sieges to date. If Batis was to undertake a sortie, Alexander wanted to ensure that his troops were ready to launch their own counter-attack against them and catch them unprepared as the Gazans would no doubt be concentrating on the destruction of the siege engines. Alexander would not be caught unprepared again as he had been at Halicarnassus. The second element was that if Batis did not counter-attack, Alexander left enough forward troops to drag the siege engines to safety, out of missile range of the walls. This rapid evolution of strategy essentially saw Alexander failing in his initial plan but immediately devising a method of attempting to lure the defenders into a hastily prepared trap that would recover the difficult situation he found himself in. This would be entirely consistent with Alexander’s repeated strategy of always fighting battles on terrain of his choosing, even during a siege as here.

  If a sortie was indeed what Alexander was expecting after his initial reverse, then he was not to be disappointed. Arrian tells us:170

  Before long, however, the defenders of the town made a sortie in strength; the Arab troops endeavoured to set fire to the siege engines, and the heavy attacks with missile weapons delivered from their commanding position almost succeeded in thrusting the Macedonians back down the earthwork they had raised.

  After what Curtius described as something of an assassination attempt on the part of one of the Arab mercenaries, Alexander organized the hypaspists and attacked the Arabs who had sortied from the town.171 Alexander’s counter-attack was delivered with his usual use of tactical skill and the application of overwhelming force at exactly the correct location, and he succeeded in driving back the Arabs with heavy losses to them. During this counter-attack against the Arab mercenaries, and counter to the advice of Aristander who had warned Alexander to take care after the incident with the crow, Alexander was struck by a bolt from an arrow-throwing catapult stationed upon the walls. The bolt penetrated his shield and cuirass and struck him in the shoulder. The wound was grievous, and Alexander was carried from the field, not for the last time during his career. The defenders evidently believed that they had slain the Macedonian king, and began celebrating a great victory.172

  Alexander’s wound was serious, and seems to have taken him several weeks to recover as there appears
to have been a significant pause in the siege at this juncture. The time was not wasted, however. Further preparations were undertaken for the final stages, the mounds that formed the circumvallation were strengthened and raised and the engineers redoubled their efforts on the mines, as well as the bulk of the artillery having arrived by sea from Tyre.173 Curtius suggests that Alexander was adopting a new tactic at this point, but it seems more reasonable to suggest that he was building upon the tactic that he had already decided upon.

  Some modern historians have believed the circumvallation mounds to have been rather low, but this almost certainly underestimated the amount of earth 30–40,000 men could move by hand in a couple of weeks. Arrian went the other way and claimed the circumvallation mounds were up to c.76m high, but the modern mound upon which the city sits is only c.30m high and Arrian surely makes a mistake in his interpretation of his sources.174

  After what we can only describe as a failure to capture the city as Alexander had hoped, albeit having turned the initial mistake into a minor tactical victory in driving off the Arabs with heavy losses, Alexander was in much the same position as he had been in prior to the aborted assault. After a trip of apparently several weeks, the heavy artillery pieces and the siege towers that had been instrumental in the capture of Tyre finally arrived after having been transported first by boat and then dragged across the desert to the Macedonian camp; Arrian goes on to say:175