Mitch Williamson is a technical writer with an interest in military and naval affairs. He has published articles in Cross & Cockade International and Wartime magazines. He was research associate for the Bio-history Cross in the Sky, a book about Charles ‘Moth’ Eaton’s career, in collaboration with the flier’s son, Dr Charles S. Eaton. He also assisted in picture research for John Burton’s Fortnight of Infamy. Mitch is now publishing on the WWW various specialist websites combined with custom website design work. He enjoys working and supporting his local C3 Church.
“Curate and Compile“
In the late 15th century China was still a world leader in many areas of technology, having enjoyed advanced economic development for many centuries before the West. However, it began to suffer from worsening ossification of the central government and scholar-elite into endemic corruption and a rigid interpretation of Confucianism which ultimately was unable to adapt the rural economy to the expanding population. Late Ming China slowly withered under a baleful climate of stifling bureaucracy and self-imposed insulation from the emerging centers of world trade and technological innovation, which were shifting from China to Europe. For instance, the tendency to concentrate firearms production and casting artillery in centralized locations may have inhibited innovation in design. Political crisis also interfered with military reform and adaptation. At least on land the Xuande emperor had been a committed war leader. His son, Zhu Qizhen (Zhengtong Emperor), was not. Goaded to invade Mongolia, he was captured and lost an army of 500,000 to the Mongols at Tumu in 1449, after which the Mongols advanced on Beijing.
Battle of Tumu, (September 1, 1449)
In 1449 the Ming emperor Zhu Qizhen (Zhengtong), son of the fierce Xuande emperor, was just 21. Accepting advice from his chief eunuch, Wang Zhen, he invaded Mongolia with a huge host several hundred thousand strong and a truly mammoth supply train. Without ever encountering the Mongols the army turned around once it reached the extreme edge of its supplies. Just a few days march from a fortified town, and food and water, its rearguard was ambushed. Another was quickly formed but it too was cut off and wiped out by pursuing Mongols. Then the main body was surrounded. Weak from thirst, hunger, and overlong marches, the Ming Army stood no chance in the battle that followed. Wang Zhen was killed and Emperor Zhu Qizhen captured. As many as 500,000 Chinese may have perished in the Tumu campaign and battle. The Mongol horde then moved toward Beijing, raiding, pillaging, and raping as it passed unimpeded by any Ming army. The eight border garrisons (built by Hongwu but later abandoned by Yongle) did nothing but tend to themselves. As the Mongols were ill-equipped for a siege, after a week of plundering the outlying districts and countryside around Beijing they left, steppe ponies burdened with booty. In 1450 the Mongols released the boy emperor but in the interim his brother had claimed the throne. The Zhengtong Emperor did not regain power until he mounted a successful coup against his brother in 1457. After a long debate over appropriate strategy toward the Mongols, the Ming court decided to adopt a pure defensive posture and began construction of 700 miles of the Great Wall.
After that, the terrified Ming rebuilt old frontier fortifications and added 700 new miles of Great Wall to huddle behind in fear of Mongol raids-in short, they surrendered the old claim to rule Mongolia and shifted to a purely defensive strategy. From 1474 wall-building intensified and the number of firearms troops multiplied, with most in garrisons along the walls. Since their major enemies lacked fortifications, Chinese field tactics emphasized the use of guns mainly in defense. It was only in civil wars that Chinese gunners faced the tactical problem of overwhelming fortifications.
Construction of defensive walls began during the reign of China’s “First Emperor,” Qin Shi Huang, in 221 B. C. E. These connected sections of preexisting border fortifications of Qin’s defeated and annexed enemies, dating to the Warring States period, from which the Qin empire had emerged as victor. The building technique of this remarkable structure was the ancient method of stamped earth that employed masses of slave laborers as well as military conscripts. Some parts of the wall stood for nearly two millennia and were incorporated into the modern “Great Wall” built by the Ming dynasty following the humiliation of defeat and capture of the Zhengtong Emperor at Tumu (1449). After he regained the throne in 1457, the Ming court decided on a purely defensive strategy and began building 700 miles of new defensive walls starting in 1474, fortifying the northern frontier against Mongol raiders. The Ming system involved hundreds of watchtowers, signal-beacon platforms, and self-sufficient garrisons organized as military colonies. Infantry were positioned along the wall to give warning. But the main idea was for cavalry to move quickly to any point of alarm and stop raiders from breaking through. In that, the Ming strategy emulated Mongol practices from the Yuan dynasty. It was also reminiscent, though not influenced by, the Roman defensive system of “limes” which in Germania alone were 500 kilometers long.
The Great Wall was meant to reduce costs to the Ming of garrisoning a thousand-mile frontier by channeling raiders and invaders into known invasion routes to predetermined choke points protected by cavalry armies. This strategy was mostly ineffective. The Great Wall was simply outflanked in 1550 by Mongol raiders who rode around it to the northeast to descend on Beijing and pillage its suburbs (they could not take the city because they had no siege engines or artillery). The wall was also breached by collaboration with the Mongols of Ming frontier military colonies, which over time became increasingly “barbarian” through trade, marriage, and daily contact with the wilder peoples on the other side. Some Han garrisons lived in so much fear of the Mongols they were militarily useless; others lost touch with the distant court and hardly maintained military preparations at all. Finally, the Great Wall could always be breached by treachery or foolhardy invitation. Either or both occurred when a Ming general allowed the Manchus to enter China via the Shanhaiguan Pass to aid in the last Ming civil war in 1644, which brought the Ming dynasty to an end and put the Qing in power.
China never built a defensive wall along its Pacific sea frontier, as it felt no threat from that quarter. And yet, the main threat to its long-term stability and independence came across the Pacific in the form of European navies and marines. As with the 20th century Maginot Line in France, building the Great Wall in some ways signaled Ming defeatism rather than advertised Ming strength. The overall historical meaning of the Great Wall is ambiguous. To some, it signifies the worst features of China’s exploitative past; to others, it celebrates the longevity of China’s advanced, classical civilization.
A Case of Gigantism. The Boeing XB-15 shown here was the largest airplane in the world when it was completed in 1937. Although an important step in large bomber development, the plane was slow and underpowered.
With a wingspan of 149ft and a maximum take-off weight of 70,706lb, the Boeing XB-15 was an intimidating looking bomber for the mid-1930s which was let down by a serious lack of horsepower. The sole aircraft nicknamed `Grandpappy’, served with the 2nd BG at Langley Field and later, after being re-designated as the XC-105 transport, with the 20th TCS (Troop Carrier Squadron). The giant aircraft was retired on December 18, 1944.
The XB-15 set a number of `payload to height and speed/weight over distance’ records including the international 5,000km speed record with a 2,000kg payload.
Before the B-17 appeared on the scene, the Air Corps had stated a requirement for a bomber far advanced beyond the Flying Fortress. Known as Project A, this plane was to have a maximum range of 5,000 miles and speed of 200 miles per hour, with a 2,000-pound bomb load. It was to have the ability “to reinforce Hawaii, Panama, and Alaska without the use of intermediate servicing facilities.” In May 1934 the General Staff approved as its tactical mission “the destruction by bombs of distant land or naval targets.” Boeing began development of the plane in June 1934, and the Air Corps contracted for one Project A plane. The Army was interested enough in the Boeing proposal to issue a contract on 28 June 1934 that covered the design, wind-tunnel testing and a mock-up, all covered by the designation XBLR-1(Experimental Bomber Long Range 1). The aircraft passed the design and inspection process successfully and on 29 June 1935 a contract for a single XBLR-1 was issued; the designation was changed to XB-15 in July 1936, when the more long-winded designations were finally dropped from the Army vocabulary.
The aircraft that Boeing produced was a big, four-engined, mid-wing cantilever monoplane of all-metal semi-monocoque construction. Although it was not immediately obvious, the structural design was firmly based around that developed through the Monomail and YB-9 aircraft, the only exception to that rule being that the wing surface area aft of the main spar was covered in fabric instead of metal alloy sheet. [t had originally been intended to install four Allison V-1710 liquid-cooled V-12 engines, but before the design was finally confirmed the Allison engine was cancelled and the XB-15’s powerplant changed to the Pratt & Whitney R- 1830 air-cooled radial. The wings that the engines were mounted on were of sufficient depth to allow the aircraft engineer to enter the space aft of the engines to carry out maintenance in flight, should it be needed – this feature was later covered over to the Boeing Model 314 flying boat.
The ten-man crew was housed in a fuselage that featured heating, pressurization and ventilation as well as crew rest bunks, an in-flight kitchen and a lavatory. The aircraft had a retractable main undercarriage units and tail wheel, the former having two main wheels per leg to reduce the pavement loading factor. The intended defensive armament was the heaviest yet fitted to a combat aircraft consisting of six machine guns. These were located in the nose turret, the forward-facing ventral turret, in the top turret, one in each waist blister and in a rear-facing ventral turret. The basic bomb load was set at 8,000lb (1, 600kg), although a maximum overload of 12,000lb (5,400kg) could be carried over a shorter distance.
It was the largest and heaviest airplane built in the United States up to that time: its wingspan was 149 feet versus 104 for the B-17. It was 20 feet longer, with a gross weight was over 70,000 pounds-more than double that of the Fortress. The problem with the plane was its lack of engine power: its four motors, each producing 1,000 horsepower, were not big enough. One Boeing expert summed the XB-15’s problem nicely: “It provided an example of a typical situation, where a promising new design was handicapped by lack of the bigger power plants necessary to develop its full potential.” Only one prototype was built, and it was later converted to a cargo plane.
The plane flew for the first time in 1937, but it was too large for the power plants then available. The B-17 benefited from early work done on the XB-15, and the B-29 later was the offspring of Project A.
In service for eight years, the XB-15 served with two squadrons at Langley Field, Hampton, Virginia, the 49th Bombardment Squadron (Heavy), 2d Bombardment Group (Heavy) between 1938 and 1940, and the 41st Reconnaissance Squadron (Heavy), 2d Bombardment Wing between 1940 and 1942.
February 14, 1939, Maj. Caleb V. Haynes and his crew flew the Boeing XB-15, laden with more than 3,000 pounds of medical supplies, from Langley Field, Virginia, to Chile for the relief of earthquake victims. The flight demonstrated not only U. S. humanitarian-airlift capabilities, but also the range and payload of the new airplane.
This aircraft was then reassigned to the Sixth Air Force in the Caribbean and arrived at the Panama Air Depot (PAD), Albrook Field, Canal Zone, on 23 March 1943.
This aircraft was underpowered and was never placed in production but was used for experimental tests. Because of its cargo-carrying capacity, it was redesignated XC-105-BO on 6 May 6, 1943 and was modified by the PAD. After modification, it was assigned to the 20th Troop Carrier Squadron, Sixth Air Force Service Command, based at Albrook Field in December 1943.
During its eight year in service, the XB-15/XC-105 flew 60 combat missions including antisubmarine warfare patrols and 70 cargo trips carrying 5,200 passengers, 440,000 lbs (199 581 kg) of cargo and 94,000 lbs (42 638 kg) of mail. The XC-105 was placed in storage at the PAD in May 1944 due to structural damage. This one of kind aircraft was ignominiously shoved into the Curundu Swamp, east of Albrook Field, where it slowly sunk into the muck. It remains there to this day.
XC-105 “Grandpappy” in Panama
Technical Specifications (XC-105 [XB-15]) Type: Long- range troop and cargo transport. Manufacturer: Boeing Airplane Co., Seattle, Washington. Total military versions: 1 (AAF). Capacity: Crew of six plus unspecified number of troops or 27,350 lbs. of cargo. Powerplants: Four 1,000-hp Pratt & Whitney R-1830-11 Twin Wasp 14-cylinder air-cooled radial engines driving three- bladed Hamilton Standard variable- pitch metal propellers. Performance: Max. speed 197-mph, cruise 152-mph; ceiling 18,900 ft.; range 5,130 mi. Weights: 37,309 lbs. empty, 65,068 lbs. gross, 70,706 lbs. max. takeoff. Dimensions: Span 149 ft., length 87 ft. 7 in., wing area 2,780 sq. ft.
The Boeing Model 294, originally designated XBLR-1 (experimental bomber, long- range) but re-designated XB-15 in 1936 while still under construction, had been designed to test the “hemispheric defense” concept, which contemplated an aircraft that could deliver 2,000 lbs. of bombs from U. S. territories to targets within a 2,000-mile combat radius. However, due to the limited powerplant availability at that time, the XB-15, when flown in October 1937, proved to be woefully underpowered in terms of speed and rate- of- climb. Despite its shortcomings, the XB-15 gave Boeing valuable experience in the design and construction of the large aircraft it would be producing in the near future (e. g., Model 314, B-29, C-97). In 1943, after installation of an aft- mounted cargo door and an internal hoist, the XB-15 became a transport under the designation XC-105, with takeoff weight upped to 92,000 lbs. In its new role, the aircraft flew cargo, mail, and passengers on routes in the Caribbean until being scrapped at Kelly Field shortly before the war ended in 1945.
Agridi, 1230, in which a Cypriot army decisively defeated an army ten-times as strong
In the surprise attack at Casal Imbert, the Cypriot/Ibelin army had lost roughly 30 knights and the bulk of their horses and equipment. More important, the Genoese had lost their ships. Thus, while Filangieri struck in Cyprus, King Henry had no means of responding. Filangieri had brilliantly taken advantage of his enemy’s concentration of forces in one place, to attack in another.
Yet Filangieri had underestimated the Cypriot King. Henry had come of age on the same day that he had to flee from Casal Imbert in his nightshirt. He now proved that he had been no puppet of the Ibelins. Had he been merely their prisoner up to now, he would have abandoned their cause and turned to Filangieri to help him crush his former jailers. Instead, he used his increased stature as king to make significant concessions to the Genoese, securing their continued support, and in order to obtain revenue and fighting men through the bestowal of fiefs in Cyprus upon Syrian knights. In just one month, the Cypriot/Ibelin army was sufficiently reequipped to return to Cyprus — in Imperial ships.
The latter had been tied up in Acre idle. Henry and the Lord of Beirut appealed to the anti-Imperial Patriarch of Jerusalem, arguing that Filangieri in occupying Cyprus had committed a grave sin that threatened the safety of the Holy Land. The point was that Filangieri had attacked a Christian monarch without justification. While the Patriarch sympathized, he demurred, saying he could not interfere in secular affairs. However, he also noted that he would not stop anyone from seizing the ships. At once, the pro-Ibelin mob rushed down to the harbour, where they managed to seize 13 of the large Imperial “salanders” (apparently warships), while the remaining Imperial ships managed to escape by slipping anchor and sailing away.
In these “confiscated” ships, the Cypriot/Ibelin knights, turcopoles and sergeants sailed for Cyprus at the very end of May or the first days of June. Expecting the ports to be heavily defended, Beirut took the radical decision to beach (i.e. wreck) their confiscated Imperial galleys on the shore of an island near Famagusta. This island was connected to the mainland only at low tide. From here, some of the men took small boats into Famagusta harbor to make noise and create a diversion, while the bulk of the army crossed via the ford to the mainland at dawn without encountering serious opposition.
Indeed, by daybreak, it was clear that the Imperial forces, possibly overestimating the strength of the Cypriot/Ibelin force in the darkness, had opted for a strategic withdrawal. King Henry and his troops spent three days in Famagusta receiving the surrender of the key castle of Kantara and collecting further support before advancing cautiously toward Nicosia. Although they encountered no resistance, they found that the retreating Imperial troops had burned the granges and also vandalized the water and windmills.
On arrival in Nicosia, the Cypriot/Ibelin army found that, again, the enemy had retreated before them. With a sense of relief, they sought food and lodging — only to be called to arms at vespers. The men rushed out, mustered and marched north to face an Imperial attack. When they were beyond the walls, however, they discovered that the alarm had been rung by Beirut himself. Recognizing that they still faced an intact and formidable enemy army that might strike at any time, Beirut wanted no repeat of Casal Imbert. He ordered the collected and alerted army to camp in a defensible position near water and gardens and a watch was established.
The next morning, the army set out along the main road from Nicosia to Kyrenia. Between these two cities a dramatic mountain range with jagged peaks and deep pine forests rises up. From Nicosia, the road runs almost due north, weaving with the terrain, until it turns sharply to the right to enter a pass that runs west-east. Then having crested the pass, the road turns north again to descend toward the coastal plain and the port of Kyrenia. Just before the end of the pass, the main road to the royal castle of St. Hilarion branches off.
St. Hilarion still held for King Henry and was filled with many women and children of Ibelin supporters as well as King Henry’s two sisters. King Henry had received word that the castle was dangerously short of supplies and would soon have to capitulate if it did not receive aid. Anticipating an attempt to relieve St. Hilarion, Filangieri positioned the main body of his army inside the pass, where it was invisible from the lower part of the road, but he had deployed two advance divisions across the Nicosia-Kyrenia road just below the entrance to the pass.
The mountains separating Kyrenia from Nicosia, seen from the north looking west.
The Imperial forces on Cyprus consisted of the Cypriot traitors and the bulk of the Sicilian knights and Imperial mercenaries. Altogether, Filangieri could deploy over 2,000 horsemen and an unknown number of archers and infantry. The Cypriot/Ibelin army, on the other hand, had been decimated by the desertions, the reinforcement of Beirut, and the losses of Casal Imbert. King Henry could field only 236 knights, supported by sergeants and turcopoles of unrecorded number. The advancing Cypriot/Ibelin army was not only much smaller, it was below the Imperial army and would have to fight uphill to breakthrough.
However, King Henry knew of a steep and narrow path that ascended the mountain from a village called Agridi just less than a mile west of the main road, i.e., before the enemy positions. Beirut proposed advancing to Agridi, and under cover of darkness the next night, sending relief to St. Hilarion over the narrow path. Beirut divided his army into four divisions, commanded as follows: 1) Sir Hugh d’Ibelin (Beirut’s third son) and Sir Anseau de Brie, 2) Sir Baldwin d’Ibelin (Beirut’s second son, 3) the Lord of Caesarea (Beirut’s nephew) and 4) Beirut with King Henry.
Beirut’s eldest son Balian, who already had a reputation for prowess from earlier engagements, was publicly denied the place of honor in command of the vanguard, because he had been excommunicated for failing to set aside his wife — and cousin — Eschiva de Montbéliard. (That same lady who had provisioned and was holding the only other castle that had remained loyal to the king as described last week) Saying he trusted God more than Sir Balian’s knighthood, Beirut ordered his firstborn and heir to the rear.
Daylight, however, revealed the pathetic size of the Cypriot/Ibelin army. Immediately, the Sicilians took heart and with cheers, the first division started to descend the slope to attack. Led by Count Walter of Manupello, this division only glancingly engaged with Beirut’s rearguard before continuing toward Nicosia. Christopher Marshall in Warfare in the Latin East suggests this was a matter of incompetence; that the charge was carried out so badly that momentum swept it past the enemy doing no damage. It is equally possible, however, that the intention was to either divert some of Beirut’s troops and divide his forces or to reestablish Imperial control of Nicosia and cut Beirut and King Henry off from retreat. We know some of Beirut’s knights wanted to pursue and Beirut had to prevent them. Certainly, it was only after the Cypriot/Ibelin force carried the day that Count of Manupello retreated to Gastria to seek refuge with the Templars.
Meanwhile, however, the second Imperial division had fallen on the first division of the Cypriot/Ibelin army and pressed it so hard that it had to be reinforced by the second division. The fighting became fierce and hand-to-hand. Sir Anseau de Brie unhorsed the commander, the Count de Menope, and the Cypriot infantry closed in to kill. According to the account of Philip de Novare, no less than seventeen Sicilian knights dismounted to protect him and help him remount — only to be slaughtered by the Cypriot sergeants shouting “Kill! Kill!” Not exactly the picture of chivalry.
Yet the battle was far from won. Filangieri’s main force was still safe within the pass. Had they reinforced at this point, the Cypriot army would probably have been overwhelmed. Instead, Sir Balian, with only five knights, attacked from a point high up the slope along a rugged and steep path leading to the head of the pass, cutting off reinforcements at this choke point. He was so hard-pressed that the men around Beirut urged him to go to his son’s assistance, but Beirut insisted that his division — with the King — must continue to advance, presumably toward St. Hilarion. Without further assistance, Sir Balian’s small troops held the foot of the pass and prevented Filangieri from reinforcing his advance divisions.
Below the pass, the Cypriot/Ibelin army decimated the Imperial troops that had engaged them. No less than sixty knights — a huge number by 13th century standards — had been killed and forty more had been taken prisoner. Filangieri decided to cut his losses and disengaged, retreating up the pass and down again to Kyrenia. King Henry and the Ibelins proceeded to the successful relief of St. Hilarion.
Although the Imperial army had sustained shockingly high losses by the standards of the day, it was by no means annihilated. Yet, apparently the fight had gone out it. The Count of Manupello’s division, denied refuge by the Templars, surrendered, while Filangieri and the Cypriot traitors retreated to the fortress of Kyrenia. From here they sent appeals for help to Armenia, Antioch, and the Emperor himself, but when these yielded nothing, Filangieri and the traitors sailed away. The garrison they left behind eventually capitulated after a year-long siege. In short, Agridi had proved decisive. Frederick II never again attempted to send “orders” to King Henry, and the Pope later absolved King Henry of all oaths to the Holy Roman Emperor.
Emperor Frederick II has re-established Christian control of Jerusalem, but the Sultan brags that he will “purify” the Holy City and drive the Christians out as soon as the ten-year truce expires. The common people of the Holy Land show their contempt for the Emperor and his treaty by pelting him with offal, while the barons resist Frederick’s absolutism and demand rule of law. Filled with resentment and bitterness toward his impertinent subjects, the Emperor vows to destroy the family that embodies the independent spirit of Outremer: the Ibelins. While the Emperor’s deputies will stop at nothing to fulfill their orders, the Ibelins must gain allies at almost any cost. Yet with the Pope now on the side of the Emperor, Balian’s marriage becomes a spiritual weapon turned against his father.
Montgisard, 1177, in which a 16-year-old youth suffering from leprosy defeated an invading army tens of thousands strong under the Sultan Saladin with just 400 knights and civilian/amateur infantry.
Saladin, who had gathered his forces in Egypt to repel the impending attack from Jerusalem, rapidly learned that the invasion of Egypt been called off and that the bulk of Frankish fighting forces had moved north. It was a splendid opportunity to strike at his enemies and the Sultan seized the opportunity with a force estimated at 26,000 light horse — which leaves open the question of whether infantry was with him or not. The force also allegedly included some 1,000 mamluks of the Sultan’s personal bodyguard.
According to an anonymous Christian chronicler from northern Syria, the news of Saladin’s invasion plunged Jerusalem into despair. The king was just 16 years old, had no battle experience of his own, and his most experienced commanders (or many of them) were besieging Hama. The Constable of the Kingdom, the competent and wise Humphrey de Toron II, was gravely ill. But according to Archbishop William of Tyre, Baldwin’s former tutor now his chancellor and our best contemporary source, Baldwin rallied his forces and with just 376 knights made a dash to Ascalon, the southern-most stronghold of his kingdom.
Note: Baldwin personally led his knights to Ascalon and he did so on horseback — not in a litter as some novelists and hobby historians suggest. At this stage, Baldwin’s leprosy manifested itself only in an inability to use his right arm. Despite this handicap, he had benefited from special riding instruction as a youth. His biography, Professor Bernard Hamilton writes:
Baldwin, who was effectively one-handed, needed to learn special skills if he intended to fight, because he would have to control his mount in battle with his knees alone. The training he received was clearly first-rate because he remained an excellent rider until he became too ill to mount.
William of Tyre, who had been Baldwin’s tutor before he became king and served him later as his chancellor writes even more compellingly:
He…was more skilled than men who were older than himself in controlling horses and in riding them at a gallop.
Arriving there only shortly before Saladin himself on November 22, King Baldwin took control of the city, but then hesitated to risk open battle with the Saracens because of the imbalance of forces. Thus, while King Baldwin’s dash to Ascalon had been heroic, it appeared to have been less than wise strategically. Saladin had effectively trapped the King and his knights inside Ascalon, and nothing lay between Saladin and Jerusalem except scattered garrisons.
Rather than wasting time besieging a fortified city with a strong defending force, Saladin left enough of his army behind to maintain the siege of Ascalon and moved off with the bulk of his troops. Indeed, the Sultan and his emirs were so confident of victory that they took time to plunder the rich cities of the coastal plain, notably Ramla and Lydda. In Jerusalem, the terrified population sought refuge in the Citadel of David.
But Baldwin IV was not yet defeated. With the number of Saracen troops surrounding Ascalon dramatically reduced, he risked a sortie. He also got word to the Templars in the fortress of Gaza, and they sortied out to rendezvous with the King. Together this mounted force started to shadow Saladin’s now dispersed and no longer disciplined army. Frankish tactics, however, required a combination of cavalry and infantry, so King Baldwin could not engage the enemy until he had infantry as well. He, therefore, issued the arrière ban, a general call to arms that obligated every Christian to rally to the royal standard in defense of the realm. The burgesses started streaming to join him.
What happened next is usually depicted as a “miracle” or just “dumb luck.” However, Michael Erhlich in a reassessment of the Battle of Montgisard published in Medieval Military History [Vol. XI, 2013, pp. 95-105] argues convincingly that the Franks, in fact, very cleverly lured Saladin into marshy ground, where his superiority of numbers could not come into play. He writes:
The Franks knew the terrain much better than Saladin. Fearing his numerical superiority, they certainly discarded the option to attack him in Ramla because in this case the topography of the area presented many difficulties for them. Their move towards Ibelin indicated that they wanted to fight the Muslim army. Otherwise, they could have taken the safe road to Jerusalem, such as the route via Hebron. Therefore, the fact that they followed Saladin clearly indicated that the Frankish king believed that he could win a battle.
He continues the story:
Baldwin passed near enough to Saladin’s camp in Ramla to persuade Saladin to follow him, but also very near to the mountains where he could escape in case he lost the day…An interesting maneuver was that the Frankish army … did not use the main road [to Jerusalem] but a parallel path which was barely known to strangers.
Baldwin…succeeded in maneuvering Saladin to the place he wanted: a marshy area…[where] numerical superiority became a burden rather than an advantage. It demanded additional efforts to maneuver the trapped army, which fell into total chaos. Led by a local lord, who certainly knew the terrain better than any body else on the battlefield, the Frankish army managed to defeat the Muslim army, in spite of its initial superiority.
Although the battle was hard-fought and there were Christian casualties, the Sultan’s forces were routed. Not only that, Saladin himself came very close to being killed or captured and allegedly escaped on the back of a pack-camel. Yet for the bulk of his army there was no escape. Those who were not slaughtered immediately on the field, found themselves scattered and virtually defenseless in enemy territory. Although they abandoned their plunder, it was still a long way home — and the rains had set in. Cold, wet, slowed down by the mud, no longer benefiting from the strength of numbers, they were easy prey for the residents and settlers of the Kingdom of Jerusalem. The latter, after the sack of Lydda, Ramla and other lesser places, had good reason to crave revenge. Furthermore, even after escaping Christian territory, the Sultan’s troops still found no refuge because once in the desert the Bedouins took advantage of the situation to enslave as many men as they could catch in order to enrich themselves. Very few men of the Sultan’s army made it home to safety in Egypt.
But to what extend was this really King Baldwin’s victory?
A number of modern historians, basing their assessment on Arab sources, claim that the real commander at Montgisard was Reynald de Chatillon, the Lord of Transjordan. In fact, this is a red herring. Arab sources had absolutely no insight into the Frankish command structure. The most prominent fighter on the battlefield is not necessarily (indeed rarely) the actual commander. Furthermore, because Chatillon was a familiar figure to the Arabs, he was readily recognized; not so the other lords. Most important, Arab chroniclers were at pains to justify Saladin’s summary execution of Chatillon ten years later after the Battle of Hattin by making Chatillon into a particularly dangerous enemy of Islam. Making him the mastermind of Montgisard fit this agenda, but it proves nothing about who actually devised the strategy and led the Frankish army to victory at Montgisard.
Ehrlick furthermore stresses that the victory depended on superior knowledge and effective use of the terrain. This, he rightly contends could only have come from local lord — someone who knew not just the main roads but the by-roads and all the little swamps, creeks, forests and hillocks along the way. That was most certainly not Reynald de Chatillon, a Frenchman who had been Prince of Antioch, a prisoner of the Saracens for 15 years, and then became Lord of Transjordan far from this little piece of the kingdom. It was also not King Baldwin. It was undoubtedly the Ibelin brothers. They were fighting near their birthplace and, in Baldwin’s case, within his lordship. In accordance with custom, Baldwin claimed — and received — the right to lead the vanguard in the battle.
Yet, in the end, that does not take away from Baldwin’s right to claim this victory as his own. It was King Baldwin who made the decision to ride to Ascalon with less than 400 knights. It was Baldwin who sortied out of Ascalon to start shadowing Saladin’s army. And it was Baldwin who agreed to the plan of attack proposed by the Lord of Ramla and/or his brother.
As Professor Hamilton stresses, given his incurable and debilitating disease, it would have been legitimate for Baldwin to abdicate all command to a constable, the royal official responsible for commanding the feudal army in the absence of the king. Instead, Baldwin “led his armies in person and took part in the fighting even though he was effectively one-handed, using the skills which his Arab riding-master had taught him.” [Bernard Hamilton, The Leper King and his Heirs: Baldwin IV and the Crusader Kingdom of Jerusalem, Cambridge University Press, 2000, 108.]
The report that Sardis had been captured and burned by the Athenians and the Ionians was brought to Darius and that Aristagoras of Miletus had instigated that joint action. It is said that when Darius heard of his affair … he asked who the Athenians were. When he was he told he called for his bow. He took it and loosed an arrow into the sky saying ‘Gods grant that I shall be allowed to punish the Athenians.’ He then ordered one of his slaves to repeat the following words three times every time he dined: ‘Master, do not forget the Athenians!’
The story is amusing but fanciful, appealing to the audience perhaps but not that convincing when searching for authenticity. Darius had campaigned in Thrace and further north, had in his court many of Greek origin and was close to Hippias, the former tyrant of Athens who when expelled in 510, had gone into exile at Sigeum on the Persian side of the Hellespont. Darius needed no reminding about the Athenians and the fact that he appointed Mardonius as the new satrap for Hellespontine Phrygia and Thrace, with instructions to take the Persian presence into Greece just months after the Ionian Revolt had been quelled plainly indicates that a general subjugation of the Greek mainland including Athens had been contemplated for a long time. The battles at Marathon and Thermopylae, ten years apart, but between the same foes, certainly remain the two most easily recalled military engagements of ancient Greece, if not in all antiquity. Yet there is much less modern attention on the battlefields themselves and the historical contexts of each episode, which usually draw just a few sentences in a general coverage of the wars between the Greeks and the Persians. The campaigns that led to battles at Marathon in the summer of 490 and at Thermopylae in early August 480 have become famous partly because the expansion of Persian Empire in a westward direction was thwarted, although its resources were much greater than those available to the Greek cities, and partly because of the stature of the main source material, which again hinges on Herodotus’ Histories. Notwithstanding the fame of the work in question, its seemingly inchoate narrative regarding these battles, provided by history’s first historian, poses numerous puzzles, which for any sense to prevail about them require some critical analysis. Although there is a decade between Marathon and Thermopylae the campaign that led to the latter began almost immediately after the Persian defeat in 490 and continued down through 487 when dealing with a revolt in Egypt took precedence (Herodt. 7.1).
Darius had every intention of enlarging his provinces on the European side of the Hellespont, but the defeat at Marathon, although no catastrophe, upset his plans and before he could regain the initiative in that quarter he died in 486. It was only in 485 after the Egyptian rebellion was quelled that an invasion of southern Greece again became a primary objective of the Persians and their new king Xerxes. He wanted to redress the failure of 490, which implicitly was an affront to the dignity of his kingdom and although he massacred the Greek defenders of Thermopylae this proved to be just another minor victory in an overall campaign that became a Persian debacle. Yet Marathon and Thermopylae far more than Salamis and Plataea dominate the popular imagination, and so the focus here will be to trace the two battlefields to place events at them in a realistic and historical context. The reason for concentrating on just two battles and not the entire war is firstly that geographically they are very close, secondly that both although land battles were heavily influenced by events on sea, and thirdly both involved one or both forces of small size. These were not the great displays of manpower and military might expected of the pitched battle.
Within months of peaceful conditions being restored along the coastal fringe of western Asia Minor, Darius ordered his new satrap Mardonius to continue the work begun by Megabazus and proceed to campaign further along the northern Aegean coast with a view to subduing the entire mainland of Greece (Herodt. 6.44). Before Mardonius crossed the Hellespont he visited the cities of Ionia and Herodotus notes a most unexpected gesture on the part of this new satrap in that he installed democratic governments in the cities that had recently been reconquered. Tyranny of the kind previously favoured by the Persians in these parts was no longer to be permitted. Bearing in mind that this course of action is precisely what Hecataeus urged Artaphernes, the Lydian satrap, to do, according to Diodorus (10.25.4), it ought not to have come as such a surprise. Furthermore, while Herodotus noted that some former leaders had been restored to their cities along the Hellespont he makes no mention of any in Ionia, which suggests that the Persians recognized that the imposition of governing through single rulers simply did not have popular support. Mardonius’ decree was not entirely motivated from a desire to please the local populations, however, for he knew that in the campaign he planned to wage he would need financial and material support from these cities. Therefore to avoid further civil unrest it was of a little substance to him whether the people ruled themselves or were ruled by tyrants just as long as they were compliant to his wishes and needs. Herodotus presents it as an amazing event, but it was actually simply a matter of sensible politics and part of the planning of the logistics for a new venture in Europe.
In the early summer of 492 Mardonius moved quickly by transporting his army, which Herodotus states was impressive, from Abydos to Sestos, the narrowest point on the Hellespont. From there a Persian army marching overland had little hostility to be concerned about since Megabazus had already imposed Persian rule from the western shore of the Propontis to the Chersonese, and then in Thrace as far as the River Strymon. Therefore, it must have been clear to all that Mardonius’ objective can only have been Greece and since the Macedonian king had already made a treaty with Darius the way through to Thessaly lay open. However, things did not go according to plan. At first there was a successful occupation of Thasos, which was taken without opposition, but then the fleet that had accompanied the army was caught in a gale off Mount Athos in the Chersonese. The northern or etesian winds of the summer months can be violent and were especially dangerous to ancient shipping. On this occasion Herodotus records that it was said that three hundred ships were sunk and as many as twenty thousand men from their crews were killed, some because they could not swim, others became the victims of shark attacks, and others were caught on the rocks (Herodt. 6.44). The army too suddenly encountered a setback when the Thracian tribe the Brygi made a surprise assault by night. The Persians seem to have been taken completely unawares and Mardonius himself was injured. The general, however, refused to advance further until he had punished this tribe but the result seems to have been that the campaigning season drew to a close without any further positive results and Mardonius led his army back to the Hellespont. Herodotus states that Mardonius’ army hardly behaved in a glorious fashion (Herodt. 6.45), although fault for the disaster to the fleet could hardly have been on account of incompetence of the commander. Later Herodotus (Herodt. 6.94) confirms that Darius had relieved Mardonius of the command against the Greeks.
In the winter of the same year Darius ordered the citizens of Thasos to demolish the fortifications to their city and send their ships to Abdera (Herodt. 6.46). The Thasians had been besieged by Histiaeus some years before, but being a wealthy polis which, says Herodotus had an annual income from its Thracian gold mines of between two and three hundred talents a year, the citizens had responded to external threats by expanding their navy and strengthening their city’s walls. In 492/1, however, they recognized the futility of a war with the Persians who had occupied much of Thrace and all neighbouring islands and so obeyed the commands of the dominant power in the region. Darius also wanted to test sentiment in Greece not because Mardonius had accomplished little of note in the previous year but to avoid further losses to the Persian treasury. Darius is well remembered as being a prudent ruler, and evidently decided to try diplomatic means to obtain his goal, however at the same time, ever the realist, he gave orders for the preparation of a further military campaign and demanded that the cities of western Asia Minor have warships and transport vessels in readiness. Meanwhile, heralds were sent to the islands of the Aegean and to all the cities on the Greek mainland demanding fire and water from each of these communities as a sign of their submission. The island communities were quick to comply since most if not all of them were within hours of Persian held territory. One of the islands to offer submission to Darius was Aegina (Herodt. 6.49) situated in the Bay of Salamis and within sight of Athens itself. The Athenians appealed to Sparta to intervene in what they considered to be a hostile action by the Aeginetans who were members of the Peloponnesian League under the leadership of the Spartans.
The Spartan king, the same Cleomenes who had rejected the appeals of Aristagoras of Miletus for military aid, arrived in Aegina soon after and took hostages who were then despatched to Athens for safe keeping. This was to ensure that the Aeginetans went no further in their attempts to curry favour with Persia. The Persians would have had cause to regret not being in a position to intervene in Aegina’s internal affairs since that city had a strong fleet and its harbour would have made a useful base in the event of a Darius launching an attack on Attica and the Peloponnese. But the Spartans were not the immediate target since they had not fought alongside the Ionians like the Athenians and Eretrians, and so this opportunity to gain a foothold in southern Greece was lost, irrevocably as it turned out. Evidently, an attack against at least one ally of Sparta, however loosened the bond between the Athenians and Spartans had become since the expulsion of Hippias in 510, was also regarded as a threat to the Peloponnese. The Spartans recognized that threat and acted at once. The citizens of Aegina may have considered the Spartan king’s action high-handed and may have begun planning a retaliation but the hostage-taking had the required effect and nothing more is heard about Aegina for the next five or six years. The Peloponnese and Attica seemed united against any involvement with Persia, although elsewhere medizing as it became known was common enough.
In the meantime, in the early summer of 490 Darius ordered the rendezvous of a new army and fleet in Cilicia, near Tarsus. The land forces consisted of infantry and a large contingent of cavalry and the army was reviewed on the Plain of Aleia by the joint commanders Datis and Artaphernes, son of the Artaphernes who had been the previous satrap of Lydia. The appointment of two or more generals to a command was plainly a common enough practice among the Persians and had been employed effectively in the war in Ionia, but in this instance was also probably a conscious decision in reaction to the recent failure of Mardonius who had been granted sole command in Thrace. From there the army sailed to Samos. Herodotus describes this force as a powerful one but just how large was it? A fleet consisting of six hundred triremes (Herodt. 6.95) would require 102,000 rowers, some of whom could have been utilized as light armed troops in the field, plus a further 18,000 heavy infantry, thirty carried by each warship. Still this total of 120,000 appears to be unrealistically high and problematic in logistical terms especially supplies. A fleet of this size would have required almost as many transport ships carrying food and fodder since local communities compelled to provide material aid would have simply buckled under the strain. A fleet of 1200 in 490 is not credible nor can the Ionians and the islanders of the Aegean had delivered sufficient supplies. Another reading of the text is therefore required. Herodotus must be using the name ‘trireme’ in a loose or careless fashion forgetting that whereas by his day, this was ubiquitous ‘ship’ employed for all purposes, that was not the case in the campaign to Marathon. In 490 the trireme was still a relatively novel construction and since the historian refers to transport ships for the horses (‘horse-carrying ships’) these were almost certainly not warships. A cavalry force of as little as one thousand would have required about forty triremes, and double that number of smaller vessels, especially if there was more than one horse for each trooper. It means that of the six hundred in total, perhaps one hundred or more were smaller transport ships. Moreover, some of the warships were undoubtedly of the older bireme or pentekonter construction. Altogether a fleet comprising a mixture of shipping would reduce the total to perhaps 80,000 rowers, 10,000 infantry and 2000 cavalry. The force was certainly powerful but this was not intended for a full invasion of mainland Greece but as a punitive expedition against Athens and Eretria to cause havoc before a still more powerful force could be sent to enforce Persian rule over the broader region. Herodotus has perhaps inadvertently inflated the size and power of the Persian force, which can be corrected here but for his audience it would have sounded much more impressive than it actually was if they thought in terms of contemporary triremes. The fleet probably called at Miletus before making the short crossing to Samos, but instead of heading north towards the Hellespont and the usual crossing points between the two continents sailed out in a south-westerly direction across the Icarian Sea. Herodotus affirms that this route had been chosen since the Persians were still shaken by their severe losses around Mount Athos in the previous summer and had decided to avoid that route altogether. The transportation of an army, especially one with cavalry units across the open sea, even keeping close to the islands was yet another innovation by the Persian generals, and perhaps of Darius himself.
The ‘Marathon Campaign’ began almost as a carbon copy of the Naxos expedition, and indeed Naxos was one of the first objectives since the fleet sailed west from Samos. The Naxians will surely have been alerted to this imminent threat yet unlike their spirited defence against the Persian attack, led by Megabates and Aristagoras, they offered no defence at all. The size of this latest expedition may have been just too intimidating for the Naxians who apparently abandoned their city and fled into the hills. The Persians plundered and burned the city and the temples and continued on their way. The episode must have occurred over a matter of days and is given little coverage by Herodotus, although there is perhaps more here than the narrative yields to the reader. The Naxians had been confident of withstanding an attack a decade before but in 490 made no attempt to do so. This can be attributed to a number of reasons, that the attack came early in the summer before the harvest was gathered and when food supplies were at their lowest after the winter so that there were simply insufficient supplies to see out a blockade or that there had been a change in the political leadership at Naxos, which was less opposed to an entente with the Persians. Herodotus (6.49) claimed that all the islands had offered fire and water to Darius, so the attack may have been unexpected and unprovoked. Finally, the example of the fate of some of the Ionian cities was still fresh enough to make a defence the island seem a worthless proposition.
Datis also occupied the island of Delos, although the population fled before the Persians arrived. On account of the cult to Apollo and Artemis, which was also held in esteem by the Persians, the island was not plundered and its people were invited to return. The Persian fleet then had a short distance to cover before they landed on the southernmost point of Euboea at Carystus. Datis had already enforced the submission of all the islands he had visited and collected troops and hostages from each. He now demanded from the citizens of Carystus that they also join the war against their neighbours but, even in the face of what must have appeared overwhelming odds, they refused. A siege commenced and the land around the city was devastated and the people of Carystus surrendered to the Persians and the city was spared destruction. The Eretrians will have had some days’ warning that they were about to be attacked but will surely have heard reports of the Persian expedition from well before the attack on Carystus. They sent messengers to Athens requesting aid and the Athenians responded immediately by sending a force of four thousand who, according to Herodotus, were from families that had been settled on lands belonging to Chalcis some years before. Such a prompt and positive reaction was not copied by any similar action by the Eretrians who were divided in how they should meet the Persian threat. One group wanted to flee from the city and make for the safety of the surrounding hills – which they probably did – another group with their sights set on future personal gain was conspiring to turn the city over to the enemy without a fight. An Eretrian citizen named Aeschines was alerted to this treachery and he informed the Athenians who at once withdrew and crossed the straits to Oropus just in time to escape the disaster that followed.
The Persian fleet made land at a number of beaches close to Eretria (Herodt. 5.100) and they prepared to make an assault on the city, which remained well defended since many of the citizens had chosen to remain but were not confident enough to offer battle outside their fortifications. The Persians appear to have attacked the city but there is no mention of any specialist siege equipment and it is likely that they concentrated on undermining a section of the circuit walls. The fight went on for six days with heavy casualties but with no obvious conclusion in sight until certain Eretrians who were pro-Persian opened a postern gate or successfully connived to leave a section of the walls unguarded. The Persians sent troops in and opposition seems to have completely collapsed as the sack of the city began. The traitors are named by Herodotus (Herodt. 5.101) as Euphorbus and Philagrus who were no doubt well rewarded, but may not have been allowed to remain in Eretria but rather resettled elsewhere. Xenophon in his Anabasis (8.7), written after 400 BC, which describes the events of a rebellion and its aftermath against the Persian king Artaxerxes II by his brother Cyrus in which the writer participated as a mercenary, mentions a meeting between himself and descendents of a certain Gongylus of Eretria. Gongylus had participated in the betrayal of Eretria in 490 for which he had been granted lands in Mysia. His widow who was named Hellas still lived in one of these possessions in the Caicus Valley, which later became the city of Pergamum.
Eretria was neither a major settlement nor especially well defendable, although it possesses an impressive acropolis on a steep hill above its theatre. The population was probably hardly more than twenty thousand so its seizure by the Persians was predictable. Those who were caught were taken as prisoners to Asia Minor and resettled. The temple of Athena Daphnephorus was burned and plundered by the attackers in revenge for the destruction of the burning of the temple at Sardis. Datis was certainly carrying out instructions but it might have been wiser to have been more generous in his treatment of the city. In fact, the severity of the punishment meted out to the Eretrians may, like that to the Milesians, have been exaggerated by the Greek writers of history. Like Miletus, Eretria quickly recovered, its citizens, many of whom must have fled to safe havens elsewhere on Euboea, returned and rebuilt their city, although the temple of Athena appears to have been a long time in restoration. Just ten years later, in the allied Greek fleet that saved the mainland from Persian domination, the Eretrians provided the same number of warships as they had sent to aid the Ionians in 499. This is a clear indication of the dramatization of the episode in Herodotus and as it was received in the later literature.
After a few days the Persians re-embarked their troops and sailed for Attica, but there was absolutely no chance of catching the Athenians unprepared since the events at Eretria will have been keenly observed from Oropus. The Persian fleet was probably shadowed by scouts as it made its way down from Eretria, past Rhamnous and into the Bay of Marathon, where an army of almost entirely Athenian citizens was encamped and waiting. The plain at Marathon stretches for at least five kilometres (2 miles) in length between two steeply sided headlands, especially that of the Mount Pentelicon range to the southern edge. The depth of the plain is roughly two kilometres (2000 yards) from the hills that give access into central Attica from the sea. The landscape including the sea level has not altered much from the time of the battle. The tumulus in honour of the dead Athenians is as prominent today as it would have been in 490 and will have be clearly visible to travellers passing by land or by ship. Obviously today the landscape has been altered by modern developments in housing and farming but the general nature of the battlefield remains the same. The land usage in 490 probably consisted of small subsistence farms with scattered bush and trees but which was easily level enough for the effective deployment of the cavalry that had been so carefully transported from Asia.
The forces assembled by the Athenians seem hardly to have made for a strong opposition or made a protracted campaign likely. An army of roughly ten thousand drawn from each of the tribes of Attica marched out from Athens to meet the attackers, which as a force is just two thousand more than the Naxians, who had seen off the Persian attack just over a decade earlier, but who had recently surrendered without a fight. The enemy must certainly have had an overall numerical superiority especially in cavalry units, although that military arm constituted a problem in itself since the nature of the land in Attica was mostly unsuitable for large cavalry deployment. The northern and western quarters of Attica and hence the route for any force whose objective is Athens itself is particularly hilly with narrow valleys and steep sides gorges. This means that the Persians were extremely limited in the places they could effectively operate out from. Marathon on the western coast of Attica and Phaleron just to the southwest of Athens had the available space for making the superiority of the cavalry count and had the space for beaching the fleet. Otherwise, the use of cavalry could easily become a handicap and a structural weakness for any attacking army. And this is clearly what actually happened. The Persians were guided to Marathon by Hippias who knew the area well and at least was able to give some specialist advice but he must also have had qualms about the ultimate success of the venture. If he did not voice this concern it may only have been to ensure that any negative remarks were not held against him later on. Hippias like the Persian commanders knew that unless they controlled the battlefield the enemy would start with a major advantage and quite simply they allowed the Greeks with a smaller force mostly of infantry to start hostilities from higher ground while their cavalry does not appear to have been fully disembarked or brought into action.
One can also easily discern the extent to which the Marathon campaign became as much myth as history when the tale of the courier Pheidippides is encountered in the narrative. The Athenians had received reinforcements from just one of their allies, namely Plataea on the southern edge of Boeotia, a small community that probably sent most of its available manpower. The Plataean contingent numbered approximately a thousand and was to be stationed on the left wing on the northern side of the plain. The Athenian generals were also counting on the support of Sparta. If the Spartans sent troops the other cities in the Peloponnese that looked to Sparta for leadership would follow. Herodotus states that before the Athenian army had fully mustered in the city and therefore perhaps as much as a week before the battle Pheidippides was ordered to run to Sparta and appeal for aid. Why the appeal was left until the last minute when the Athenians could have sent requests some time beforehand is not explained and exposes the extent of dramatic invention in the text. The distance between Athens and Sparta is approximately one hundred and fifty kilometres (100 miles). Twice Herodotus says (Herodt. 6.107) that Pheidippides twice encountered the god Pan, either a personification of Dionysus or the god himself while on his way. The presence of Pan or Dionysus in this account is not a random event that was added for entertainment but was linked to the origin of the cult of this god at Athens and his cave on Mount Pentelicon, which rises to the southwest of the plain of Marathon. The runner is said to have met the god, a habitué of the mountainside, this time on Mount Parthenium just above the city of Tegea in the Peloponnese and on the frontier with Laconia. Pan addressed Pheidippides asking why he was not given honours in Athens when he had helped its people in the past and would again in the future. The Athenians did not forget this and when times were more favourable they built a temple dedicated to this deity beneath the Acropolis and from 490 held sacrifices and games in thanks for his intervention during this crisis. Again, myth has entered the account for Marathon when it is noticeably absent from the record of the Ionian War.
Pheidippides arrived in Sparta just twenty-four hours after he left Athens and in his appeal for Spartan aid he specifically noted that Eretria had just been destroyed. This pinpoints the episode to within a matter of days in the mid-summer of 490, and indeed Herodotus states (Herodt. 6.102) that the Persians remained on Euboea only for a few days. The Spartans are said to have been sympathetic but in accordance with their laws and because they were celebrating the festival of the Carneia celebrated between the seventh and fifteenth of the month Carneus (the Athenian month Metageitnion and roughly August) in honour of Apollo (Apollo Carneus), and since it was on the ninth day that Pheidippides addressed them they were unable to leave for another six days to join their allies. Pheidippides returned with a promise for future aid but nothing more. The runner’s mission also exposes the absence of Athenian planning and the ad-hoc nature of their preparations. The institutions of the democracy, while only recently inaugurated in Athens, tended to preclude rapid decision-making. The planning of the defence of Athens could easily have been placed in motion some months before particularly since the Athenians had known for some years that Persian revenge would come. They also had Mardonius’ campaigns of the previous year when contacts with trading partners in the Euxine had surely been affected. All in all, the myopic attitude of ancient communities to the outside world prevalent in antiquity is very plainly revealed here.
While the delay to their departure is attributed to a scrupulous observation of religious principles, there may have been suspicions that the reluctance on the part of the Spartans may also have been based on political grounds. And so a delay was also imposed on the Athenians, although there was no consensus. This is again clear from Herodotus who gives a glimpse of infighting among the ten generals and perhaps the rather ambiguous or very cautious approach of the person in command, Callimachus the polemarch (Herodt. 6.109), whose home town, Aphidnae, was just on the other side of the mountains from Marathon. Among the eleven was the same Miltiades who had fled from the Chersonese three years before and who had acquired the office of general on account of his exploits and family background. Herodotus writes that Miltiades, supported by four generals – there was deadlock about the best action to take – was for an immediate engagement with the enemy. This made some sense since the Athenians already held the higher ground and the Persians had to disembark.
The Athenians and their allies are said to have already encamped among the hills to the south of the bay. The Persians having rounded the northern headland, Cape Cynosura, into the bay of Marathon beached their enormous fleet, approximately two kilometres away from their enemy who must have been in full view of the attackers. Herodotus’ account is not coherent and some guesswork is needed to understand the events of the next few days. The Persians evidently disembarked and although the plain might have been suitable for employing cavalry units it would have taken a great deal of time to offload the horses and supplies and form them up into effective units. This will account for several days since not all the ships will have been able to beach at the same time and some complex schedule would have been enforced besides making an encampment for the troops and sending out foragers to meet all the needs of soldiers and animals alike.
The Athenians and their allies must have watched all these proceedings from their vantage point. The problem was one of waiting for the Spartans to arrive and thereby having battle-hardened troops among the front line. The Athenian citizen hoplites will have had very little recent experience of a battle, especially against a force that had obtained recent victories across the Aegean and on Euboea. Miltiades was the leading advocate, or so Herodotus claims, of an immediate engagement and this must be connected with not allowing the invaders to become comfortable in their new bridgehead. He persuaded Callimachus to vote against delaying any further and seems to have been concerned that some of the generals were secretly in contact with the Persians (Herodt. 6.109). It made good sense to catch the Persians and their allies unsettled and unprepared but there was also the adoption of some interesting strategy, attributed by modern scholars to Miltiades but in fact probably one that was discussed at length by the commanders, that of weakening the centre while adding extra troops to both wings of the army. This would result in the centre being deliberately allowed to withdraw in the face of superior weight from their opponents but also allowed the right and left wings of the army to rout their opposition and then sweep round to attack the enemy’s main concentration of troops from the rear.
This plan was put into effect in a most shattering manner and gave the battle of Marathon the fame it has enjoyed ever since. The Athenians occupied the centre while Callimachus as the senior commander occupied the right wing, which was closest to Mount Pentelicon, while the Plataeans were assigned the very end of the left wing, which was probably supplemented with some of the Athenians or possibly other allies. These duly advanced down into the plain where no pronounced gradient is noticeable and once the level ground was reached the Greeks did not possess an advantage in terms of a slope down towards their opponents. Hence all speed was necessary. The Greek army numbering 11,000 opposed an opponent of about the same number, although the Persians probably had additional light armed skirmishers. It is evident that with these numbers they could not possibly have filled the entire plain as is often illustrated on maps in discussions about this engagement. A single horseman or heavily armoured infantryman might occupy as much as a metre each but neither side on this occasion stood in such a thin line. Each side will have formed up its troops in ranks that for infantry by the end of the century varied from eight to sixteen deep. On a plain that extends in length for some five kilometres the battle was confined to just the southern end near to where the tumulus in honour of the dead Athenians was erected. This is closer to the Pentelicon range of mountains where a tumulus to the dead Plataeans was also erected and which still stands near the site of the modern museum. It shows clearly enough that the battlefield was on the small rather than on the grand scale. While a Persian encampment to the northern end of the bay where the hills behind offered protection Herodotus appears to describe the Greeks attacking ships that can only have been beached at the southern end of the bay, although once again a dramatic element may be present.
At this point, it seems appropriate to voice some scepticism about the usual perception of the battle in that Marathon is represented as a great victory for the hoplite or heavily armoured infantry over the Persians whose force must have in some measure consisted of cavalry, or Herodotus would not have made the point about the horse transports. Descriptions of Marathon tend to dwell on the prowess or training of the hoplite but whether this was a vital factor at all is questionable. The terrain was ideal for cavalry being employed interspersed with light armed troops, which was indeed a feature of battles elsewhere in the Ancient World. The Persian invasion force was clearly a mix of cavalry, heavy and light armed infantry, and such a combination would surely have been most successfully met by a similar composition in the Athenian led defence. Furthermore, the Persians had with them Greeks from Asia Minor and the islands and therefore the ethnic composition of the two sides would not have been that discrete. The landscape quite clearly indicates that the battle took place on more or less a level space and this may well indicate, although Herodotus makes no mention of this, that the Athenians made use of cavalry among their ranks as well. To counteract any superiority their enemy possessed the Athenians could naturally draw on their own cavalry especially since they were well informed about the nature of the invaders long before they arrived from not only reports from Eretria and Oropus but also observation of the Persian fleet from Rhamnous. Finally, some strength in cavalry also then makes the account of the Athenians moving rapidly back to Phaleron to intercept any Persian attempt at landing there afterwards more understandable. This would have been more easily accomplished by cavalry units than by exhausted hoplites.
The battle according to Herodotus was hotly contested, although it is likely that a later and more heroic interpretation that became the tradition crept into the account of what was probably a short and sharp encounter. The Greek attack caught the Persians and their allies on the back foot, which is surprising if the latter had superiority in cavalry but comprehensible if they were evenly matched with similar troops moving rapidly towards them from the southwest. A late afternoon assault by the Greeks, as Munro suggests, would have placed the sun in the eyes of the enemy. Yet if they moved out of the Vrana Valley in the direction of the beach, a distance of not less than eight stadia states Herodotus (6.112) or roughly one thousand six hundred metres (a little under one mile) would such a long approach have caught the Persians in disarray? The Persians would have been caught unprepared only if they had not secured a bridgehead, although that seems improbable. If that were the case then Herodotus’ account of the Athenians waiting at Marathon becomes meaningless since a sudden attack would imply that the Athenians met up with their allies closer to Athens, marched down the Vrana Valley and went immediately into battle. The Persians would have had too short a time to fully organize their forces and were particularly badly beaten on their right and left wings by the unexpected superior numbers sent against them by the Athenians. In such a situation neither army could have formed up in much of a fashion if the Greeks were charging across the plain while their opponents hastily drew up their lines. This is not what Herodotus recounts since he is categorical about the components in each line, which suggests the usual drawing up of army formations prior to engagement. It is true that Herodotus claims (6.112) that the Greeks possessed no cavalry or archers but since this is contained in a section that is mostly devoted to material that is of dubious historicity it can probably be dismissed. Munro suggested that the Greeks advanced at a rush to avoid Persian archers but that tactic would only be effective if there had been none in position and this supposed charge in full armour was for nearly a mile. As regards the Athenian centre, right and left wings, Van Wees is rightly sceptical about the tactics ever working, as they are described by the Herodotus, simply because in the confusion of battle to have the two wings acting in perfect unison is at best implausible.
Once the Athenian wings had supposedly turned their opposites to flight but had not pursued them and instead turned together to attack the Persian centre from the rear the rout in which the enemy casualties are given precisely as six thousand four hundred began. The fugitives were pursued to the beach and attempts were made to capture or destroy the enemy ships. It was during the fighting around the ships that Callimachus and two of the other generals, Stesilaus and Cynegirus, were killed. If this is an accurate account by Herodotus why, if there had been a really serious rout with a substantial number of the enemy dead, were the Persians able to disengage with very little trouble? It is more likely therefore that the Persians put up a stiffer resistance on the field and at their ships, most of which they launched successfully from the beach. The senior Athenian commanders were killed in fierce fighting, which was certainly not against terrified fugitives who had thrown away their weapons. The loss of Persian ships was minimal, seven out of six hundred, and even if the fatalities given appear to be high in fact the Greeks probably inflicted more casualties on non-combatants and unarmed camp followers among the Persians trying to evacuate the beach than military personnel. This would naturally enough have not been remembered by any of the participants or entered the later historical accounts. Whereas large numbers of casualties were to be expected whenever a rout took place in the later stages of a battle, unlike in modern conflicts camp followers intermingled with the troops during the fighting in the hope of finding good plunder. It is these who are more likely to have been killed. The fact that the cavalry was clearly not much affected by the defeat has led some to believe that the Persians were about to sail and were in the process of loading their horses on to the waiting transports. This is an equally odd interpretation of the events since it presupposes such a lengthy delay between the beaching of the invasion force and the battle that the Persians rather than risk a battle on prepared ground were instead prepared to venture into less well chartered territory, even if they had Hippias with them who obviously knew Attica well enough. If the cavalry escaped largely unscathed then it must mean that the Athenians did not, in fact, manage to cause much damage to this section of the army.
To sum up, it seems probable that the Athenian attack was staged in the late afternoon brought on possibly by seeing the Persians about to sail off. The Athenian command may have assumed that the Persians either intended to return to Euboea or sailing on to Athens. If the first, the intention must have been to at least gain some credit before their enemy departed, if the second, then to disrupt any further invasion of southern Attica. Thus the Greeks erupted out of the Vrana Valley mostly in two columns of a rapidly moving combination of cavalry, heavy and light armed infantry. This meant that the Persians did not have the time to draw up their entire force especially since the cavalry may well have been already on the transports. They then paid the penalty for not taking the offensive when they could have marched south into Central Attica before the Athenians arrived or decided against forcing an assault on higher ground where the Athenians are said to have camped if they were there already waiting for the allies to arrive.
With very little certain detail to go on it is possible to suggest that, contrary to Herodotus’ statement (6.102), that the Persians remained in Eretria for only a few days, and that this period extended over some weeks. The Persians could have believed that the Athenians like the Eretrians would place their trust in their city’s fortifications rather than risk a battle. The fact that both sides made for Marathon shows that the Athenians had been forewarned of the Persian arrival and that it was one of two places that could accommodate the invader’s fleet and cavalry. Similarly, the Persians would probably have also been happier with an early battle before the Spartans arrived and had purposely landed during this Laconian festival because they had been informed about its significance by Hippias. Assuming that the Athenians were already encamped on Mount Pentelicon when the Persians beached at Marathon and refused to do battle they effectively debarred the Persians from attempting a march inland and once that route was denied then Datis and Artaphernes had to make for the alternative, which was the bay at Phaleron. Before they could make for southern Attica the Athenians had a successful engagement behind them, which may not have weakened the Persians by much but a defeated side was psychologically less willing to risk a second battle so soon after the last. A retreat would have seemed sensible in order to plan a more comprehensive second assault, perhaps the following spring.
The Athenian victory at Marathon was therefore no seminal point in the history of ancient warfare, but it provided and excellent opportunity for propaganda, as the subsequent offering to Apollo at Delphi and which was proclaimed on an inscription erected at the Treasury of the Athenians, just below the temple (see plate). The tropaion or trophy consisted of armour taken from the Persian dead and was arranged on a triangular ledge on the lower side of the treasury building. The shape of this pediment probably intended to resemble the prow of one of the captured enemy ships, albeit that there were very few of those. Moreover, the victory became in later times attributed to the heavy infantry, although the evidence, such as it is, hardly supports that contention. And so the historical tradition clearly shows that this version became canonical but quite contrary to what was, in reality, just a moderate success and merely a setback for the enemy. The tumulus erected to the dead Athenian soldiers also provided another propaganda coup for the Greeks who admitted to just one hundred and ninety-two killed. Although Herodotus gives no figure for the Plataean dead these were also honoured in a similar fashion at the entrance to the valley of the Vrana River. However, it is likely that many other Greeks combatants died but were not honoured in the same way as those who possessed an elite status. The difference between the two casualty lists is probably far less than is presented by Herodotus.
The prominence in the deliberations before the battle given to Miltiades and the heroism attributed to the polemarch Callimachus (Herodt. 6.108; cf. Pausanias, 1.15.1) stems from the fame of one of the three frescoes in the Stoa Poikile (‘The Painted Colonnade’) in the Athenian agora which was commissioned by Cimon, the son of Miltiades, whose wife was also of the same family as Pericles. Although he died in disgrace soon after the battle, his relationship with Cimon and Pericles, the most prominent Athenians in the generation following the defeat of Xerxes, goes far to explain the fact that his fame endured. But Callimachus was also honoured with a stele, or dedicatory inscription, on the Acropolis, which suggests that for contemporaries of the victory he was regarded as a more significant figure than Miltiades, and that this memorial was in part erected to counter any alternative version put about by the latter. Herodotus has little to say about the two thousand Spartans who were sent out to help the Athenians after the end of their festival. Although they hurried to take part in the defence of Attica and reached Marathon just three days after leaving their home, they were still too late (Herodt. 6.120), but they nonetheless wished to view the scene of the hostilities. After they had viewed the dead and complimented the Athenians on their victory they returned to the Peloponnese. There is no mention about the identity of the commander of this force and whether or not it was one of the kings as would have been the standard practice. Indeed, there is also no clarification about whether this number represents the Spartiates alone, which if it did would mean that a substantial army had taken the field or whether this was the total number in this force. Also left unclear is whether or not Sparta had summoned its allies in the Peloponnese to provide troops to join this expedition.
Having made the point that cavalry were an important element in the Persian force Herodotus then makes no mention of their use by either side at Marathon. Yet there is a source that does highlight the importance of the cavalry (hippeis) in Athenian society and since that evidence is a commemoration of the Persian wars, especially Marathon, then it must surely indicate that the Athenian army had a strong element of mounted troops in its army. This would go some way to explain why the Persians were so at pains to transport their cavalry such a long distance from Cilicia. They had been warned what to expect from the Athenian exiles such as Hippias. Thus while the Athenian force is usually considered to have consisted of hoplite heavy infantry there is no mention of hoplites on the frieze of the Parthenon on the Acropolis that commemorates the victories over the Persians.
Besides Herodotus’ account there is little of substance about Marathon in later accounts. Diodorus covered this event in Book 10 of his history but a single fragment (10.27.1–3) remains in which an incident is related to be placed just prior to the start of the battle where Datis sent heralds with a message to the Athenians demanding their surrender. In this message he claimed that all Athenians were descendants of Medus, king of Media, just as he was and that he had come to reclaim his ancestor’s land from which he had been expelled. Miltiades, speaking on behalf of the Athenians generals, rejected Datis’ claim to Attica and stated that, in fact, the Medes ought to swear allegiance to Athens. Once rebuffed Datis prepared for battle. The episode is probably an invention but it does add certain points of interest, not least the prominence again given to Miltiades in the affair and reflects probably his dominating place in Diodorus’ overall coverage of this battle, which he obtained from Ephorus and not from Herodotus who does not retell this tale. The preparation for battle suggests that Ephorus or perhaps Diodorus did not believe the Athenians had taken their enemy unawares and that some time did elapse between the arrival of the Persians at Marathon and their departure to Phaleron.
Pausanias, writing still later, adds a further element to the mythical aspects surrounding this battle when he describes the Sanctuary (temenos) of Nemesis at Rhamnous and a costly error of Datis. He says that the deity was depicted on a block of Parian marble that the Persians brought with them to inscribe their victory over the Athenians and is therefore another instance of hubris (pride) meeting nemesis (downfall), which is so evident throughout Herodotus’ text. The origin of the marble block is interesting since Paros was occupied by the Persians after they had subjugated Naxos and Delos. Herodotus merely says that after they left Naxos the Persians took the remaining islands on their way to Euboea, but next after Delos is Paros. Yet if this block already contained the image of Nemesis then it may well have been pillaged from Rhamnous, which is not recorded, but would show that the Persians were active around the Bay of Marathon for some time. The authenticity of the statement is much less certain since Datis had taken particular care not to offend Apollo at Delos so why succumb to sacrilege elsewhere? Yet, with Hippias in his entourage Datis would have listened to any advice and a shrine to Nemesis almost visible from where the Persians were proposing to land would surely have made any show of arrogance in the vicinity, an act to avoid at all costs. The tale given by Pausanias is probably an invention designed to appeal to the Roman tourist of such cult sites in the second century AD rather than a reflection of Datis’ character. Meanwhile, Herodotus who had a predilection for such drama would have found it irresistible for inclusion in his narrative had he known it.
Datis immediately sailed for Athens hoping to find the city undefended but probably took several days to sail around Cape Sounion. Plutarch (Arist. 5.4) adds some interesting evidence about the Persian intention and of the date and timing of the battle. In his life of Aristides who was one of the generals at Marathon he states that when the Athenians saw the Persians setting their ships towards the south and clearly about to aim for Athens itself, the tribe commanded by his subject was left to patrol the battlefield to prevent looters and to bury the dead. This in itself was normal practice but he also claims that the Persians were obliged to head south because of the prevailing wind, which must have been blowing from the north. The etesian winds blow mid-May to mid-September and could make sailing very difficult, even impossible, in the Aegean hence the Persian command may have had no choice in their destination. Moreover, the winds are at their most severe in the afternoon and tend to die away overnight picking up again in mid-morning. Strong gales are more likely in the middle of the summer than towards autumn, which again places the battle in August rather than in September, but also in the afternoon. Datis arrived at Phaleron early in the morning perhaps and stayed only a short time in order to beat the afternoon winds, which would have hindered his easterly course towards Asia.
In 490 the broad crescent of Phaleron Bay, to the southwest of the city, was the main beaching place for vessels either bringing merchandise to Athens or of Athenian ships and even its war fleet. In 493/2 the archon Themistocles had supported a measure to relocate much of the city’s commercial activities to the Piraeus, which offered more secure harbour facilities and which could be fortified and protected in time of attack. This was mainly in response to the raiding of the coast of Attica by their neighbours the Aeginetans but since Aegina had recently been pacified at roughly the same time the work of the Piraeus was unfinished. However, the Athenians’ war fleet, by then at least forty triremes, had probably been moved to the new harbour especially when the Persians were reported to be moving down the coast from Marathon. For the Persians Phaleron would have been the obvious point for disembarking mounted and infantry troops. From there it was a short distance to the city’s fortifications which, without the army sent to Marathon, would have been inadequately defended and unable to withstand an assault. The walls of Athens were no doubt reasonably secure but it is worth remembering that in 490 they did not consist of the elaborate defensive circuit, which was to be completed only fifty years later by Pericles. Furthermore, any sympathizer of the Pisistratids or the Persians would have been ready to open a gate as had occurred at Eretria and may well have been planned in advance with Hippias and Datis. Indeed, the drawing away of the Athenian army to Marathon has an element of planning about it even if the Persians suffered a defeat. This was clearly not a disaster and perhaps been factored in as a tactical reverse to win the ultimate prize. It was Hippias who advised the plain of Marathon as the best suitable place for collecting the invaders’ forces before these advanced towards Athens from the northwest. Herodotus (6.115) was also aware of some reports that gained some credibility then or later that the family of Cleisthenes and Pericles possibly favoured a Persian occupation of the city and which understandably as an admirer of the latter he dismissed out of hand. The Alcmaeonids, descended from tyrants and probably not averse to desiring outright power themselves, may have been pro-Persian while at the same time hardly sympathetic towards their great rivals Hippias and Miltiades. For a generation, Pericles was a quasi-tyrant and his rule was benign with all the acceptable trappings of senior statesmanship than of an absolute ruler. Invective about his family’s activities during past events such as Marathon still became an inevitable accompaniment to the possession of power and influence from those who envied Pericles’ position or who wished to emulate him. Still, in 490 there were surely others in Athens who might have been or could have been persuaded to betray the city by a promise of wealth or power. This is indeed hinted at by Herodotus through the words he gives to Miltiades when speaking to Callimachus when urging him to attack the Persians. In ancient sieges at a time when siege machinery was either not transported with an army or was not available the easiest way to capture a town or city was by bribing a resident to allow the enemy entry. There are numerous examples of such conclusions to siege events, the capture of Athens might also have been accomplished in this fashion.
Herodotus (6.116) states that the Athenians reached Phaleron and made their camp at Cynosarges near the precinct of a temple dedicated to Heracles before the Persian fleet rounded Cape Sounion. The Persians seeing the Greeks in possession of the beach chose not to attempt a landing, and perhaps rode the waves until such time that they were able to sail away. Where did they go? The first news of the Persians’ whereabouts comes in Herodotus’ account of another supernatural event, this time a vision or dream only when Datis had reached the island of Mykonos. Yet it is very unlikely that the Persians can have gone far from Phaleron before they would have been obliged to beach in order to rest the crews and attend to the horses. Carystus seems a likely harbour on their route to Ionia, which was not at too great a distance from Phaleron assuming that the coast of Attica was avoided altogether. However, depending on the time of day the Persians reached Phaleron then a halt in Attica perhaps near Sounion seems probable. During the retreat from Phaleron when he was at Mykonos Datis is said to have had a dream, the details of which were never spoken of, but as a result of this he ordered a search of the ships in the fleet. As a result of this search an image of the god Apollo was discovered (Herodt. 6.118) on a Phoenician vessel and he heard that it had been stolen from Delos. Datis immediately set sail for the island where he returned the image to the Delians, who had taken up their former residence but he requested that they transport it to its original home, which was at Delium on the Greek mainland opposite Chalcis. The Delians failed to fulfil this request probably because they knew that Datis was in no position to enforce it and wanted the image for their own cult centre.
There is a great deal of fabricated material in accounts of Marathon yet there is no mention of an oracle in connection with the battle. The Ionian War possesses neither oracles nor mythical material. It would seem therefore that there was a great deal of scientific evidence available for events in Asia Minor between 500 and 493, but much less secure information for Marathon. This again contrasts a great deal with the account of the invasion of Greece by Xerxes where oracular material is highlighted as indeed is the physical presence of the Persians at Delphi. Immediately after the end of his coverage of the fight at Marathon Herodotus goes on to describe the last year of Miltiades’ life in which he led an expedition against Paros, a campaign lavishly funded by the Athenians on the promise of great gains from the attack but which ultimately yielded very little. Miltiades’ hubris or pride, which led to a loss of popularity, and death soon after from an infected wound was, however, foretold by the oracle to the Parians (Herodt. 6. 135) and gave them the resolve to resist any Athenian attack on their island. It is therefore unlikely to have been merely by chance that Herodotus chose instead to dwell on more supernatural elements that he regarded in some way as more relevant to Marathon than for either the preceding or the following wars. Instead it is the presence of the gods themselves whose enmity is on occasion incurred or whose possible future anger must be avoided at all costs that is the more telling aspect. Thus the fear of and respect of deities is perhaps meant to reveal something about the beliefs of the time or the places which feature in the war. The god Pan puts in a physical appearance, Heracles and his cult centres are alluded to on more than one occasion, the goddess Nemesis was offended by the actions of Datis – according to Pausanias – who was concerned to ensure the goodwill of Apollo both at Delos and at Mykonos, while ghostly figures are also said to have joined the fight at Marathon. The battle of Marathon achieved mythic substance and the ingredients of myth filled up the gaps of an otherwise brief and unremarkable event.
The attempted submission of Aegina in the late 490s to the Persians ought not to have caused the surprise Herodotus claims since the Aeginetans had been hostile towards Athens over a long period of time. They were close neighbours and their commercial interests and ambitions clashed so the goodwill of Persia would have given Aegina the edge. The intervention of the Spartans denied the Persians a base on the island in 490 and while a section of the population in Aegina which would have wanted revenge on Athens for holding hostages from among its citizens, the Athenians wanted to ensure that the Aeginetans would not be a source of trouble and possible enemy in the future. Intervention in the affairs of Aegina occurred in the middle years of the next decade and the defeat of Aegina came as a result of the Athenians investing in a much larger war fleet than they had possessed up to that time in order to strike at least a parity with their neighbour. In about 487 Themistocles persuaded the people to vote to build an extra one hundred triremes to be funded from the silver bullion produced from mines in Laurion near to Cape Sounion in which rich deposits had recently been discovered.
Naval matters must also have been prominent in the thoughts of Darius following his latest failure to move the Persian frontiers westward. It was quite clear that the failure of Mardonius’ campaign in 492 was not the result of the fighting with the Brygi but had everything to do with the loss of ships at Mount Athos. This point is emphasized in Herodotus’ account, although he does not explain that it impacted negatively on supplying a large army when the movement was meant to be rapid and even when there were allied states along the route. Climatic conditions also draw no comment from ancient writers yet one poor summer, or worse a series of cold summers, resulting in a failure in the harvests and hence the depletion in local supplies placed heavy or impossible burdens on communities who had to supply armies on the move. A lack of locally produced supplies placed a heavier reliance on an accompanying fleet and when that was severely damaged it put paid to Mardonius’ plans, and explains Darius’ decision to launch a seaborne expedition instead in 490. After Marathon, Darius is said to have planned (Herodt. 7.1) a new expedition against the Greeks, which would have involved a far greater army and fleet. Xerxes’ decision to make use of Darius’ plans meant also that the logistics had to be solved before he actually took command of his army and fleet. This, of course, meant that newly levied forces had to be moved from Asia to Europe and, while this could have been accomplished by ferrying men, animals and materials and many camp followers across in transport ships as Mardonius had done, the ambitious Xerxes chose another spectacular and ultimately memorable way. He ordered the bridging of the Hellespont. It is often forgotten that, in fact, he was not the first to bridge the waters between Asia and Europe since Darius had ordered this to be done at the Bosphorus thirty years before in his campaign against the Scythians. On that occasion the architect or engineer had been a Greek from Samos named Mandrocles (Herodt. 4.87). The first bridge at the northern end of the channel joining the Aegean to the Euxine was in effect an elongated pontoon construction probably of pentekonters or biremes lashed together with some sort of covering to allow the crossing of men and animals. The Bosphorus is a relatively narrow channel, and is today bridged at Istanbul, and probably in extent not that much larger than Darius’ bridging of the lower Danube later in the same campaign. Herodotus gives the impression that Xerxes, clearly wished not only to emulate but also to surpass Darius by instructing engineers to erect two bridges over the Hellespont (the Sea of Helle), which opens into Aegean (The Chief Sea or Archipelago).
For Herodotus the immensity of the project, which matched or was made to match the ego of the king, was, even so, successfully accomplished only at the second attempt. He says that the Hellespont was spanned by two bridges of almost fantastic design, and modern scholarship is divided between those rather sceptical of the historicity of this episode and those who see elements of historical fact admittedly mixed with the historian’s desire to denigrate the character of the Persian king. The scepticism arises from the scale of construction in an age when such technical knowledge did not exist. Large buildings could be raised on land using stone with little or no use of concrete, bridges over streams and rivers could be accomplished but the Hellespont or the modern Dardanelles is a rather different proposition. The depth of the channel and the strong current plus the length of the construction all combine to make the venture extremely hazardous with a good chance of failure. Of course, seventy-five years later in September 413 BC the Syracusans threw up a barricade of ships across the entrance to their Great Harbour in just three days, according to Diodorus (13.14.2). These ships were almost certainly aging triremes, numbering about forty and linked bow to stern by chains and planks of wood and the entire cordon was garrisoned with troops. In contrast to the Hellespont, the main harbour at Syracuse has quite a gentle swell with little obvious current and in the event of a storm the ships could easily be disengaged and moored in safety in the smaller harbour to the north of Ortygia. This does not seem to have been necessary during the fairly short period that the boom was in place and which effectively prevented the escape of the Athenian besiegers. The entrance to the harbour at Syracuse is approximately fifteen hundred metres (just short of a mile), and forty triremes each a little less than forty metres or roughly forty yards in length uses up that space in a single boom (40 × 40 = 1600 metres). The Hellespont between Abydos and Sestos at the point stipulated by Herodotus (Herodt. 7.36) was roughly seven stadia across (1400 metres, about three quarters of a mile) and he states categorically that the bridges were brought from the Asia side of the channel to a headland between Sestos and Madytus (Herodt. 7.33–34). Herodotus (Herodt. 7.36) is also firm in claiming that the more northerly of the two bridges required three hundred and sixty vessels of both pentekonter and trireme design, while the southern bridge required three hundred and fourteen ships. Unlike the fortified boom that the Syracusans built across their harbour mouth this Persian bridge cannot have had the ships linked bow to stern on account of the strong current running into the Aegean, which means that the ships must have been joined by their beams facing the Bosphorus. Triremes had a beam of five metres and pentekonters a metre less, so the northernmost bridge if using all three hundred and sixty triremes measures 1800 metres (roughly the same in yards), the southern bridge with three hundred and fourteen pentekonters 1256 metres. This means that Herodotus’ figures are either incorrect or that a bridging point a little to the north of Abydos must be sought. This is also difficult since good stretches of beaches were also needed on both sides and where these exist, the current is markedly less but the distances at 3.87 and 3.39 kilometres for the two bridges would require more than double the number of vessels specified. In the end Herodotus’ position probably comes fairly close to the actual bridging point but neither the specific situation nor the number of ships employed in the construction can have been accurate.
The number of ships claimed for the bridges is remarkable since the figure must be double that stated by Herodotus since the first attempt to bridge the Hellespont failed. This is assuming that a large number of the original six hundred or so ships were destroyed in a violent storm and that even if some were salvaged for the second attempt then the total number of ships must exceed one thousand. Where did Xerxes’ engineers lay their hands on so many warships? According to Herodotus (6.8–9) there were about six hundred Persian warships and three hundred and fifty-three warships manned by the Ionian Greeks and their allies in the battle of Lade in 494. A number of these vessels would have been destroyed in the fighting but the majority would have remained operational. Xerxes had access to massive resources but he does not appear to have ordered the building of new ships as the foundation for the bridges, whereas he certainly ordered cables to be manufactured to join the vessels together. Therefore the ships were already in service or in harbour and must have been requisitioned from the cities in the region. Ships in any age have a limited working life but with limited technology the use of wooden vessels was fairly brief. A trireme probably served as a warship for no more than ten years, perhaps less, and after that as a transport for men or horses. Triremes or pentekonters would probably not have been re-employed as merchant ships because of the lack of space in the hold and the need for so many rowers. Twenty years’ use would have been followed by decay on a beach or reuse as firewood. But decayed ships would have been useless for the bridging project since they would be unstable in the water – especially in the strong current of the Hellespont – and highly susceptible to damage in adverse conditions. The engineers appointed by Xerxes for the second attempt, perhaps Ionian Greeks, at the bridging point must have sought out ships in good condition and either of relatively recent construction but not much more than ten years old. In 481 BC this points to warships from the 490s and a little later. Many of the warships used at Lade were no longer available since Mardonius lost at least three hundred at Mount Athos in 492. Any that survived had probably been retired from service. The Persian fleet that went to Marathon came back virtually unscathed and consisted of at least six hundred warships. Warships built after Mardonius’ Thracian campaign and any built on the orders of Darius in the months following the defeat at Marathon were therefore the building blocks of the bridging project and not much older. Xerxes will also have ordered new warships to be built for his campaigns in Egypt from 487 and still more once he had decided on the new invasion of the Greek mainland. This all points to almost frenetic activity in the regions closest to the future theatre of war, but also indicates that there were financial benefits from funding a new war and the investment this brought from the centre of the empire. The communities in Asia Minor especially benefited from Xerxes’ territorial ambitions. These same communities in the 490s had suffered from the lengthy war with Persia but would now have seen a full and rapid recovery in their wealth. It is hardly surprising that among Xerxes’ war fleet that accompanied him to Greece between a quarter and a third came from Greek-speaking cities.
Still, it is extraordinary that Xerxes was able to dispose of about a thousand warships of trireme or pentekonter design that were deemed too old or ruinous, but were still usable for bridging the Hellespont. Besides these he also had in service another twelve hundred warships, mostly triremes of recent construction for his fleet. The disparity in the size between the two ship types indicates that if there were two bridges then the triremes placed beam to beam comprised one, the pentekonters beam to beam the other. The difference in height and width using the two types would have resulted in an unusable structure. The trireme with its five-metre beam was at least nine metres in height to take account of its three tiers of rowers, while the pentekonter was a much smaller vessel with a single line of rowers not that much shorter at twenty-eight to thirty three metres, with a beam of four metres, but its height was not much more than three metres. A bridge consisting of a haphazard combination of both types seems quite impossible. It has been argued that the Hellespont, which today has a depth in places of up to one hundred metres, was in antiquity about one and half metres lower at the Abydos to Sestos crossing. Yet it is also clear that the sea has retreated quite dramatically further south in the Troad; at Troy the citadel once at the beach is now five kilometres inland. If the channel further north had a lower level in antiquity then the crossing distance would also have been somewhat less than the seven stadia ascribed to it by Herodotus.
It is possibly fairest to conclude that as far as the bridging of the Hellespont is concerned that, although there are clearly invented elements in Herodotus’ account, the assertion that the channel was crossed by Xerxes seems irrefutable simply because it is stated so firmly. Still, it may well be that there may have been just one rather than two pontoon bridges available to Xerxes and his army for crossing the channel in the late spring of 480, and that the crossing was a combined operation using bridge and ferries. This seems plausible if only because of the waste of resources that building two bridges would have consumed. Thus one should not forget that quite explicit in Herodotus’ account is the glorification of Xerxes’ ambitions employed by the historian to elevate his subject’s pride so that the victory by the Greeks can be equally enhanced. Finally, Herodotus states (Herodt. 9.114) that when the Greek fleet arrived in the Hellespont in the summer of 479, by then victorious over the Persians whom they had fought and defeated on the beaches around Mount Mycale just across the Latmian Gulf from Miletus, ‘they found the bridges broken up’, but provides no further details.
Regarding Xerxes’ war fleet, Barker, for example, believes that Herodotus’ figure of one thousand two hundred and seven triremes (Herodt. 7.89) included those six hundred and seventy-four that had been used in the construction process of the bridge, hence leaving approximately another six hundred to accompany the invading army. However, as Wallinga has shown this will not do, and the six hundred ascribed to the Persian fleet in 481/0 is drawn from a comparison with the fleet of Darius in his Scythian expedition and the Persian fleet at Lade, but it is clear that Xerxes’ fleet was very much larger and meant to be greater than any employed in this region before. And indeed Herodotus is at pains almost to describe how Darius ordered the building of not just warships but also of other vessels (Herodt. 7.1) when he began to make plans for another attack on the Greeks in the aftermath of Marathon. Herodotus’ numbers are extraordinarily precise and he must have drawn these from an earlier source, perhaps one of the early Ionian writers of history who may have remembered details and recorded them. Xerxes’ fleet suffered considerable losses en route to Attica, especially at Artemisium (see below). Even if some of these were exaggerated by Herodotus to glorify the achievements of the Greeks there were still over six hundred ships in the Persian fleet at Salamis. Hence the starting total was not far from the figure given by the historian.
The total numbers in Xerxes’ army have also been the subject of apparently endless debate and while clearly there is again much elaboration in Herodotus’ narrative, the lowest possible totals for his army and navy make extremely impressive reading and as reports and rumours emerged about the numbers to the Greeks highly intimidating. The crews of even a thousand triremes add up to about two hundred thousand men and these were mostly rowers not expecting to be involved in land battles. This figure would not be much less if there a sizeable number of the smaller pentekonters in the fleet. The fleet moving in tandem with the army west from the Hellespont also had to beach at the end of every day’s progress for the ships’ hulls to dry off and for the crews to eat and rest. And this number does not take into account the huge number of civilian craft that would have accompanied the army in the hope of economic gain along the way. Merchant shipping may have relied more on sail than rowers but the crew of another thousand vessels may well have added another one hundred thousand to the sea borne element alone. The army was clearly a large one and Herodotus gives a breakdown, taken presumably from a source that either remembered this occasion or drawn from knowledge of the various regions that came under Persian rule. The list of the various contingents may appear impressive (Herodt. 7.61–81) but simply illustrates the wide variety of armed troops that were available and could be levied by a Persian king. It says nothing about their military capability. Herodotus surely provides the list for the amazement of an audience who would not have seen such a sight, but he was also consciously emulating Homer’s catalogue in the Iliad of the participants in the protagonists of the Trojan War. Homer’s Trojan War was the portrayal of the supreme conflict, Herodotus’ war between the Greeks and the Persian Empire was meant to equal or even to surpass that ancient myth with one of his own time and perhaps crafting. The Persian army is therefore given a fabulous total number but it is worth remembering that at the final land battle at Plataea in the summer of 479 the invaders must have at least have equalled the Greek armed forces, which are said to have numbered one hundred thousand. The crossing of the Hellespont in the late spring of 480 and the end of the next summer casualties and the division of the army after Salamis between Xerxes and Mardonius must surely indicate armed forces somewhere in the region of two hundred thousand. Mardonius was left with the task of securing Greece by Xerxes so required sufficiently large numbers to overcome any opposition while the king could not be seen to return without a suitable bodyguard. Put the land forces total with the total from the fleet and merchant shipping and the number approaches half a million, not to mention camp followers, slaves and servants, plus all the pack animals and wagons and one can easily see why the total might be inflated further since such a force would blot out the landscape and appear almost like a swarm of locusts.
Because of the difficulties Mardonius had encountered in 492 and perhaps as a solution to the problem of sailing with Mount Athos first directly ahead and then to the right with strong tailwinds the decision was taken to bypass the peninsula altogether (Herodt. 7.24). The plan was to excavate a canal across the isthmus leading from the city of Acanthus to the peninsula on which Mount Athos is the dominating natural feature. As with his detailed description of the bridging of the Hellespont Herodotus is particularly precise about this as can be gauged from his narrative (Herodt. 7.22–24):
The Persians Bubares, the son of Megabazus, and Artachaees, the son of Artaeus, supervised the construction. For Athos is a great and famous mountain and comes down to the sea and is inhabited. At its furthest limit on the side of the mainland the mountain is a peninsula (‘chersonese’) and the isthmus at this point is twelve stadia.
The Persian engineers chose what may already have been an existing causeway and since the distance is less than two and half kilometres (about a mile and a half) and land mostly level, the construction ought not to have been that formidable. Yet Herodotus states (Herodt. 7.22) that work on the canal began three years before Xerxes’ army crossed the Hellespont. There was no shortage in the numbers of men available and many of these were drawn from the armed forces and transported to Athos by ship. The local communities were also obliged to provide labourers. Gangs of workers seem to have been collected along ethnic lines since the Phoenicians are said to have shown particular skill in the segment of the canal assigned to them. Finally, at the entrances to the canal breakwaters were built to prevent silting from the prevailing sea currents (Herodt. 7.37). Remains of the canal have been excavated but whether it was much more than the existing causeway with some water flowing through parts of it is questionable. Furthermore, the number of ships that passed this way when Xerxes had seen the canal is likely to have been a fraction of the ships in the fleet. Most captains would have taken the risk of sailing around the headland in mid-summer when the chance of gales was not as great as later in the year. It was, however, another instance of the overweening pride of the oriental despot, as Herodotus claims that the building of a canal was quite unnecessary (Herodt. 7.24) because ships could be easily dragged across the isthmus. Still, in the narrative it adds to the drama leading to the ultimate Persian failure and the humbling of Xerxes. The evidence suggests that the canal remained in use for a very short time, mainly because there was no city nearby to maintain the facility. It is surprising perhaps that the Athenians never showed the slightest interest in the canal when they were major importers of grain from the Euxine and when the shipping route from there to Attica was a very busy and lucrative one. This too possibly points to an unfinished or temporary project rather than one that was intended to last or that could easily have been utilized later.
The question of supply dumps in the campaign to Thermopylae is a very important issue and has seldom if at all been addressed by modern scholars. Herodotus places great emphasis on the incredible size of the Persian armament, both on land and on the sea; and, while some element of exaggeration is obvious in order to enhance the later victory of the Greeks, it is clear that the Persians came in force. But the campaign required intricate planning, which evidently took four years before the king even set foot in Europe and preceded the crossing of the Hellespont. Herodotus notes the placement of supplies along the march probably collected at the end of the summer in 481 since much of the material stored was perishable grain for making bread. Other supplies such as meat and fodder for the horses and other pack animals could be requisitioned from local towns and cities. Merchants accompanying the column sold all kinds of goods, and bought others, chiefly slaves. This enormous body of troops and camp followers is in modern parlance easily comparable to a moving hypermarket/superstore on foot! Xerxes was advised that there must be the placement of supplies along any route that initially lay through the Persian satrapy of Thrace and from there, via Macedonia whose king had made a treaty with the Persians, into Thessaly. The use of supply dumps seems an innovation, although Darius may have employed such measures on his march to the Danube in 513/12. Its use may be related to the fact that there were few wealthy cities along the route of advance through southern Thrace and the western border of Macedonia. Xerxes’ army and fleet marched or sailed from Cilicia, some of the ships or land forces from very much further afield, but there is no mention of the same exercise elsewhere because the route possessed sufficiently wealthy communities for requisition or market places. Herodotus names five places where supplies were brought by ship from Asia: Leuce Acte, Tyrodiza, described as ‘of the Perinthians’, Doriscus, Eion at the mouth of the River Strymon, and in Macedonia. Leuce Acte is said to have been chosen as the base for most of the equipment, raw materials or foodstuffs, which were transported by ship mostly from Asia Minor (Herodt. 7.25). Strabo in his geographical survey of the region lists the places along the western shore of the Hellespont from its southern entrance, and Leuce Acte is given as the fifth town after Aegospotami, which is two hundred and eighty stadia north of Sestos. The first and most important supply dump was therefore perhaps no more than four hundred stadia (80.8 kilometres, roughly 50 miles) distance from the Persian bridging point of the channel. If the intention had been to situate the supply depots along the route intended for the invasion force then Tyrodiza seems out of the way since Perinthus was situated towards the centre of the Propontis and lies six hundred and thirty stadia (127 kilometres or roughly seventy-five miles) from Byzantium and about twelve hundred stadia (250 kilometres or a little under 150 miles) from Eleus (Strabo, fr. 57). Ideally a supply depot should have been sited towards southern Thrace, but Herodotus’ text suggests that Tyrodiza was a harbour of Perinthus. Barker has made the attractive suggestion that Tyrodiza actually lay at the head of the Melian Gulf (Gulf of Saros) near or at the mouth of the Melas River. There is no evidence to support this conjecture but it is attractive in terms of the logistical placement of Persian supplies; and Xerxes’ army had to cross the Melas before moving into southern Thrace. Incidentally, Herodotus says (Herodt. 7.58) that the Melas River, like the Scamander River in the Troad (Herodt. 7.43), ran dry on account of the number of Persian troops and animals who camped there, which shows that it was one of the main halting points in this early part of the expedition. This point would also make tactical sense since it lies further along the direct route from Leuce Acte. No major urban centres are attested in this area, although Strabo (7.7.6) states that Greeks had settled there. Perinthus was a colony of Samos and therefore the Samians may also have been active at the Melas River. After the Gulf of Melas the next supply base is said to have been at Doriscus on the Hebrus River. Doriscus was clearly an ideal site since the beach and the surrounding country were flat (Herodt. 7.59). The fleet could be hauled ashore to dry off and there were sufficient land for an encampment of the entire army. That being the case, Herodotus claims that Xerxes insisted on carrying out a tally of his forces (Herodt. 7.60; cf. 7.87), which is said to have been one million and seven hundred thousand infantry and eighty thousand cavalry not to mention mounted camels and teams of chariots. The fleet consisted of three hundred warships from Phoenicia, two hundred from Egypt, one hundred and fifty from Cyprus, one hundred from Cilicia, thirty from the Pamphylians, fifty from Lycia, seventy from Caria, thirty from the Dorians living in Asia, one hundred from the Ionian Greeks, seventeen from the islands, and sixty from the cities of Aeolia; the cities of the Hellespont provided another one hundred but the citizens of Abydos were instructed not to accompany the fleet but were to guard the bridge (Herodt. 7.89). Besides this total figure, Herodotus states that there were another three thousand smaller vessels and transports. Diodorus (11.3.9 cf. Herodt. 7.184) claims that there were eighty hundred and fifty horse transports and about three thousand smaller vessels he names as triaconters or thirty-oared ships. Like Doriscus, Eion on the Strymon River, was in Persian hands and had been fortified by them since Darius had ordered Megabazus to occupy the region after the expedition against the Scythians (Herodt. 7.108). After Eion, Xerxes spent some time at Acanthus before moving on towards Macedonia where he ordered his fleet to await his arrival at Therme (Herodt. 7.121).
Compared with the difficulties encountered by Mardonius in 492 Xerxes’ advance across southern Thrace and into Macedonia went without any major mishaps. Still, between Acanthus and Therme, Herodotus notes (Herodt. 7.125–126) that lions attacked camels in the baggage train but not the usual pack animals. He is at a loss to explain their partiality but at the same time provides evidence for lions being common all the way between Abdera in southern Thrace to Acarnania in western Greece and almost to the northern shore of the Corinthian Gulf. The supply depot in Macedonia, and there may have been more than one, is not attested but is likely to have been at the coast rather than inland. The ruler of Macedonia at this time was Alexander I who had a treaty of friendship with the Persians and so was expected to provide help to Xerxes whose route lay through his kingdom. If the Persians sent supplies here in advance it may indicate that Macedonia was unable to provide for the numbers that were anticipated to gather there before moving south into Thessaly. Moreover, at this time the Macedonians had limited access to the sea since the best harbours in the region at Pydna and Methone were independent states. The fact that Xerxes was able to encamp his entire force at Therme (Herodt. 7.127) shows plainly enough that the final supply depot, although not initially specified, must have been at or near Therme. And the Persian forces occupied the coast from Therme all the way to the Haliacmon River, which marked the southern boundary of Macedonia at this time. The distance from Therme to the Haliacmon River was about fifty kilometres (30 miles) and must have been filled with the various camps of the contingents in the army and the ships of all sorts drying off in the summer heat.⁹⁰ The Persians evidently remained in Macedonia for a number of days if not weeks (Herodt. 7.131) for during this stay Xerxes went on ahead of his forces to explore the various possible routes by which he might enter Central Greece through the mountains, which mostly barred his way. Mount Olympus divides Macedonia from Thessaly and only one or two passable routes linked the two regions. The most feasible of these was the road, also the most obvious, which hugged the coast and ran between Mount Olympus and Mount Ossa – the gorge at Tempe – and seemed the logical course since the army and fleet needed to move in tandem. But there was another route about which Xerxes had been informed, which lay to the west of Olympus. Xerxes therefore sailed south to the mouth of the Pineus River to view his options. He was very surprised at the size of the river and when informed by local guides that all the rivers of Thessaly joined the Pineus before it entered the sea wondered if the river’s course might be diverted by some artificial means. When the king was told that Thessaly was surrounded on all sides by high mountains and that this was the sole possible exit a river might make to the sea he is said to have declared that by just damming the river at its gorge through the mountains before it reached the coastal plain the whole of Thessaly could easily be swamped. And Xerxes observed that the Thessalians had been clever to offer themselves as his allies and subjects knowing that their land was vulnerable to conquest such were his resources. The king’s curiosity was satisfied but not his plans for it must have been clear that another bridge or number of bridges would be needed at the mouth of the Pineas and that such construction would delay the invasion further. Apparently no longer interested in the coastal road into Thessaly Xerxes returned to Therme and gave orders that the inland route south was to be taken by his forces. This also meant that army and fleet would be separated for a number of days. Herodotus adds the rather odd information (Herodt. 7.131) that the forests, presumably to the west of Therme, were cut down by one third of Xerxes’ soldiers in preparation for their advance. It is far more likely that local timber was consumed in huge amounts by the fires needed in the various camps and for repairs to the ships. It is possible, although hardly credible, that wood might be carried for fording streams and rivers when the mountains through which they were about to depart were also forested unless Herodotus means that the forests were so dense that a road had to be cut through them. While the forest clearance was taking place heralds who had been despatched by Xerxes to the Greek communities further south demanding fire and water, the traditional signs of surrender, returned from their assignments, not all by any means successful (Herodt. 7.132). Still, among the Greeks to offer formal allegiance to the Persians were those who lived in Thessaly, Locri, Magnesia, Thebes and most of the cities of Boeotia, and Perrhaebia (through which the Persians were about to march on their way into Thessaly).
Herodotus’ account at this point appears to have been derived from more than a single source since having stated that the Thessalians submitted to Xerxes while the king was at Therme. Later in the narrative (7.173) as something of an apology Herodotus also claims that the Thessalians only deserted to the Persians when no alternative remained to them. This appears as if his source wished to exonerate Thessalian activities, although in the narrative it is explicitly stated that these Greeks served Xerxes well in the subsequent war. The Thessalians had earlier summoned help from the other Greek communities intent on fighting the Persians, and the response had been a combined force of Athenians and Spartans numbering about ten thousand hoplites. This force went by sea along the cost to Halus in Achaea Phthiotis and then crossed inland into Thessaly. Themistocles was in command of the Athenian troops, a Spartan named at Euaenetus of the Peloponnesian forces, and they took up a defensive position at the gorge at Tempe. The route through the gorge was narrow and overhung with high cliffs and even without opposition would have held up the Persian advance, although at that stage Xerxes had yet to cross the Hellespont. But the Greeks are said to have remained at Tempe only for a short time and retreated after they were advised by the Macedonian king Alexander that they would be vulnerable to an attack from the rear by another route from Macedonia and ran the risk of being caught on two fronts by the invaders. The Greek force marched south again and returned to their ships and the safety of Attica and the Peloponnese. What is not clear is just how long this force remained in Thessaly and must have withdrawn before Xerxes went to reconnoitre a way south for his army. Xerxes went in a Sidonian trireme, which Herodotus states was the ship and crew the king favoured (Herodt. 7.128). He was accompanied by a sizeable force, although probably not his entire fleet as Herodotus claims, and sailed to the mouth of the Pineus near Tempe. The date of Xerxes’ visit here is probably late June some weeks after the Greeks had departed and following the submission of the Thessalians. Xerxes does not appear to have remained at the Pineus for any length of time and returned to Therme before directing his army by the inland route into Thessaly. The Greeks may have missed an opportunity of intercepting the Persian king. But why would the Greeks have withdrawn in May when Xerxes had yet to march through Thrace unless Herodotus’ chronology is suspect and the defenders actually remained at Tempe until the Persians arrived at Therme? In May there was no need for a rapid withdrawal before the Persians were actually close at hand, unless it became known to Themistocles and his fellow commanders that the Thessalians were about to medize. There are missing details in Herodotus’ account and the chronology may require some alteration to make the Greek retreat and Thessalian desertion of the defence of Greece more causally related.
Herodotus states that it took a full month for the completion of the crossing into Europe and then another three months before the Persian took Athens. The distance from the Hellespont to Athens is, however, just eight hundred and fifty kilometres (about 530 miles). This means that with the frequent halts for the pack animals to recover and for the various reviews of the land and sea forces an average daily march of a maximum of perhaps fifteen kilometres (10 miles) is indicated and probably less on most days. The slow rate of the advance had much to do with the size of the forces involved and the number of camp followers, but Xerxes seems to have also wanted to display his power to these new subjects who, says Herodotus (Herodt. 7.44–45) suffered greatly from the financial burden placed on them for entertaining the king and his immense entourage. The importance of the fleet in supplying the land forces cannot be underestimated and was probably high on the agenda of the deliberations among the Athenians and the Pelponnesian Greeks about what was the effective means of disrupting Xerxes’ advance to such extent that he would have to withdraw from his objective. Inflicting damage to the Persian warships and merchant shipping therefore became a priority and the place best to achieve that result was thought to be Artemisium at the northern end of Eubeoa. The combined fleet of the Hellenic League was sent north to intercept the Persians moving south along the coast. The command was given to the Spartan Eurybiades but Themistocles may be identified as the main architect of the strategy. To delay the Persian land forces long enough to cause maximum disruption to the accompanying fleet, another army was sent under the command of Leonidas, one of the Spartan kings to hold the pass at Thermopylae on the mainland opposite Artemisium.
Herodotus gives precise figures for the pass (7.175–177) and had plainly visited the site, although he does not claim to have been there. He states the road that ran through Thermopylae in Trachis was just 16 metres (50 feet) wide. It was a coastal road rather than a gorge as at Tempe with high cliffs on the left looking north and the sea to the right. In some place both north and south of Thermopylae the way was even narrower and barely sufficient for a single wagon to pass safely. This seemed an ideal spot to make a defence where superior numbers would count for less, and it was only when the Greeks arrived there that they discovered a major Achilles’ heel in their strategy, which was a mountain track that would be used by the enemy to attack defenders along the road from the rear. This revelation seems remarkable and hardly credible if the southern Greeks were joined by citizens from nearby communities as Herodotus states. His figures for the Greek force are also specific: the ‘official’ total was three hundred Spartiate hoplites (the full Spartan citizens), five hundred from Tegea, five hundred from Mantinea, one hundred and twenty from Arcadian Orchomenus, one thousand from the other communities of Arcadia, four hundred from Corinth, Phlius contributed two hundred, Mycenae eighty. These three thousand two hundred from the Peloponnesian League were accompanied by four hundred Thebans and seven hundred citizens of Thespiae and at Thermopylae were joined by one thousand Locrians, a thousand Melians, and roughly a thousand Phocians. The overall number of hoplites was seven thousand two hundred but there would also have been almost the same number of light armed troops not to mention the non-combatants and camp followers. There is no mention of any contribution of troops from Athens or Plataea.
Later Herodotus (7. 215) gives some of the history of the use of this ‘secret path’ in former times by the Melians and Thessalians. This track was hardly a best kept secret! It also illustrates that this was a known weakness in any attempt to hold Thermopylae or that the southern Greeks were really extremely careless in their strategic planning. Themistocles who is usually credited with being the architect of Xerxes’ downfall and a tactician of genius was surely aware of such an alternative route, which Herodotus reveals to have been quite a straightforward route to tackle, beginning in the valley of the River Aesopus (7.215–218). Therefore, it can probably be assumed at the root of the description of what was ultimately a glorious failure for the defenders are some negative impressions about Themistocles and praise instead for Spartan courage on land. The Phocian contingent offered to guard the track, but Leonidas must have known from the time he arrived at Thermopylae that at best his role would be of a brief delaying nature. There cannot have been any thought at all of a victory on land against the Persian forces at this stage, although Herodotus suggests that the choice of Thermopylae had much to do with limiting the use of enemy cavalry (7.177) and that some use could be made of an existing but dilapidated defensive wall that the Phocians had erected many years before in anticipation of an invasion by the Thessalians (7.176).
The Greeks had probably been present at Thermopylae for some days and although this force consisted of some well-trained hoplites in terms of numbers – perhaps ten thousand in all – they posed no real threat to the Persians who ought to have been able to sweep them aside. Xerxes clearly recognized the problem of forcing a way through the defenders on the coastal road and waited four days (Herodt. 7.210) for the Greeks to retreat in safety. On the fifth day by which time he was exasperated he ordered an attack with Median and Cissian troops but when these had suffered severe losses they were replaced by the king’s own Persian guard known as ‘The Immortals’. These too were equally unsuccessful against the Greek forces. On the next day, events repeated themselves and the Persians were repulsed with heavy losses. It was at this point that a certain inhabitant of Trachis named Ephialtes came to the Persian camp in the hope of a reward for telling of the existence of a track through the mountains that would allow the Persians to outflank the Greeks. Xerxes was delighted and sent off a column of troops after nightfall. At daybreak the Persians had arrived at the spot where the Phocians were on guard. These were quickly scattered – a half-hearted defence seems implicit in the text – and these hurriedly brought the news that the Persians would shortly be attacking from the southern end of the pass. The Spartan king instructed most of the Greeks to leave because, says Herodotus (7.220), most were unenthusiastic about facing certain death, although he also notes that some claimed that many of the Greeks simply left in a panic. Neither claim is truly appropriate since Leonidas would have been able to command his Peloponnesian allies to remain if he had wanted them to do so while it was Boeotian troops that actually stayed. It is therefore more likely that he dismissed his allies in order that they would be not be lost to the general cause while the Spartan force was to delay the enemy as long it could. This of course can hardly have been for more than a few hours especially since the Spartans are said to have abandoned the defensive walls in preference to a more open position. The day would have ended long after the Greek defeat. Twenty thousand Persians are said to have been killed. Most of the Thebans surrendered and were branded as slaves. The entire Spartan force of one thousand and the seven hundred hoplites from Thespiae were killed.
The fleet of the recently founded ‘Hellenic League’ had been sent to Artemisium, the place of the Artemision or the temenos of Artemis, on the north coast of Euboea, which allowed close contact with the land forces at Thermopylae. But here the sea also became a narrow channel (Herodt. 7.176), which would reduce the effectiveness of Persian numerical superiority. There were two hundred and seventy-one warships (Herodt. 8.2) with one hundred and twenty-seven triremes from Athens, and so the Greeks were heavily outnumbered. The choice of taking up a defensive position at this passage towards the more sheltered inner channel – the Euripus – between Euboea and the mainland is not remarkable, but it must certainly have resonated with an audience familiar with the geography that both land and sea forces took advantage of the landscape or here seascape to attempt to overcome the Persian threat. And if the Greeks could inflict some injury to their opponents’ fleet it immediately deprived Xerxes of command of the sea and made supplying his army more susceptible to disruption. Hence the Greeks would impose their own will where up to that point the initiative had lain with the Persians. In 490 before Marathon, on account of their victory at Lade three years before, the Persians controlled the sea lanes, but the vigour of especially the Athenians seems to have inspired a confidence to defeat this enemy navy. Perhaps they were spurred on by the fact that they had emerged the victors from the encounter at Marathon, although to what extent this apparent bravado is historically accurate and not written with the benefit of hindsight is difficult to judge. Herodotus possibly indicates that the Greeks were in fact far less enthusiastic about giving battle at sea than later became the tradition for when the fleet’s commanders heard of the approach of some Persian warships they immediately withdrew to Chalcis (7.183). Ten Persian triremes had been sent on in advance of the main body from Therme to reconnoitre and approached the island of Sciathos where there was a flotilla of three Greek ships. The latter turned tail but two were captured with their crews while the third beached at the mouth of the Pineus and the crew escaped overland into Thessaly.
From the Haliacmon River to Thermopylae the distance is roughly two hundred and seventy kilometres (about 170 miles) and it would have taken Xerxes’ army nearly three weeks to advance to a point where they came into contact with the defenders at Thermopylae. The army is reported to have departed eleven days earlier than the fleet (Herodt. 7.179), which seems like an accurate report. But Xerxes’ strategy had already begun to unravel because of a change in the weather. The fleet encountered no problems sailing south along the coast from the Thermaic Gulf and arrived between the town of Casthanea and Cape Sepias on the coast of Magnesia (Herodt. 7.183). There the earliest arrivals beached but the size of the fleet meant that there was no space and the majority of ships were forced to anchor offshore in eight lines (7.188). It should be assumed that the warships had precedence and were the quickest vessels in the fleet and so were brought ashore while the transports and the other ships lay at anchor. On the very next morning Herodotus claims (7.188) a storm blew up from the east known as a ‘Hellespontine Gale’, and those ships closest to the beach were brought ashore but those forced to ride out the high winds suffered many losses. Herodotus (7.190–191; cf. Diod. 11.12.3) claims four hundred triremes were reported as lost while large numbers of merchant vessels remained unaccounted for. The figure cannot be an accurate one and, as later in his narrative, is an attempt to reduce the Persian fleet to roughly six hundred warships at the start of the battle at Salamis. While a violent summer storm could wreak havoc with ancient shipping, mariners of the time would not have been taken entirely by surprise and so some exaggeration in the total losses may be expected from Herodotus especially since it does not appear to have been a cause of concern to the Persian command. The gale relented on the fourth day by which time Xerxes had issued instructions to take the war to the Greek fleet. Having observed the withdrawal of the Greeks south to Chalcis, two hundred Persian triremes were despatched to cover the southern end of the Euripus Channel perhaps aiming for Carystus or Eretria. The main body moved to Aphetae (Herodt. 7.193; Diod. 11.12.3) and arrived there two days after the Persian land forces had departed. Herodotus states categorically that the Persians arrived there early in the afternoon and that on the next evening the Greek force stationed at Artemisium went on the attack using a circle formation with their bows outwards. With this tactic, which prevented them from being rammed, they captured thirty Persian ships but how was this accomplished? The Greek ships were not in a position to grapple the enemy but they might have dashed out of the defensive circle and rammed amidships any encircling Persian vessel that came too close. However, these enemy warships were more likely to have been disabled or sunk rather than have been captured. That same night there was another violent storm that brought ashore the debris from the recent battle with the Greeks and seems to have unnerved the Persians at Cape Sepias. The following day the Persian fleet put to sea again and confronted the Greek fleet now reinforced from Chalcis but again when the sides broke contact it was the invaders whose losses are said (Herodt. 8.16) to have been severe but no figure is given except for the warships sunk by the previous evening’s storm as they made their way along on the Aegean coast of Euboea. On this second occasion the Greeks, although in a stronger position than in the first encounter, were still obliged to retreat to the safety of Salamis since reports arrived telling of the catastrophe at Thermopylae (Herodt. 8.21). Still, it is stormy weather as much as fighting that is noticeably the constant theme in the account of the battle of Artemisium, and which accounted for more casualties than the fighting itself. Herodotus ominously reports (8.14; cf. Diod. 11.13.1) that it was as if the deities, offended by Xerxes’ pride perhaps, were determined to reduce the numbers of the enemy fleet so that the Greeks would in future be more evenly matched.
Any detailed recollection of the encounter or rather series of encounters at Artemisium was quickly lost. Indeed as soon as the Greeks scored their great victories at Salamis, Plataea and Mount Mycale, all thoughts about the indecisive affair off the northern coast of Euboea ceased. Thermopylae remained as an example of heroism in the face of adversity and overwhelming odds, but the Greek inability to hold the Persian advance on sea before Attica could easily be suppressed. The defeat at Thermopylae happened when the Greek and Persians fleets were still engaged on what was probably a second day of fighting. They were warned by messengers of the defeat on land and that Xerxes was already advancing south. Tactics dependent on synchronized action between two forces were always a problem in antiquity and success could hardly be guaranteed, especially when one was on land the other on the seas. The battle at sea needed to be won so that Xerxes would think twice about venturing further into Greece, and this did not happen. Xerxes may have faced problems on land and on sea but he had accomplished his objective of penetrating the obstacles to an invasion of Attica and the Peloponnese. Still, a brief but curious interlude full of the supernatural intruded before the Persians occupied Athens their ultimate goal.
After forcing his way through Thermopylae Xerxes’ army marched south through Boeotia causing immense damage and devastation. While on the march the king’s attention was drawn to the fact that he was very close to Delphi, a place whose contents, says Herodotus (8.35), Xerxes was more familiar than he was with the contents of own treasury. Of particular interest to him were the treasures dedicated to Apollo by Croesus, former king of Lydia, whose kingdom had been conquered by the Persian king Cyrus in 545 BC. He ordered a force to attack and sack Delphi and to return to him with its treasures. There was perhaps another reason for the attack since Xerxes hardly needed the wealth of Delphi but he knew that recent oracular messages delivered especially to the Athenians urged the defence of Greece and promised victory against the Persians. The oracle regarding the defence of Athens by the employment of a wooden wall (Herodt. 7.141), which Themistocles had presented as a case for investing in the construction of a hundred new triremes as late as 483, would by then have been known well enough in Susa. The sack of Delphi and destruction of the temple of Apollo might have seemed an appropriate action to take after the victory at Thermopylae.
The Delphians were well aware of their danger and sought oracular advice of whether to hide the temple treasures or take them elsewhere for safe keeping only to be told that Apollo would look after his own possessions (Herodt. 8.36). The citizens of Delphi nevertheless were not convinced that the god would also look after them as well and so families were evacuated across the Gulf of Corinth to Achaea or into the mountains around Parnassus and Amphissa. As the Persian troops approached Delphi from the northwest, still the main route inland from Orchomenus, defence of the temenos lay in the hands of Apollo’s priest and just sixty volunteers. As the Persian troops reached the shrine of Athena Pronaia the priest noticed that at the temple of Apollo arms that were dedicated to the god were lying before his shrine. At the same moment it was claimed that loud thunder was heard and great boulders fell from the cliffs high above the Spring of Castalia crashed down and killing many of the attackers. Besides this two phantom giants were said to have emerged to fight with the Delphians who had taken advantage of this intervention by the god and scattered the invaders who fled. Later reports also spoke about two gigantic warriors – supposedly local heroes from ancient times – who also joined in the rout killing many of the Persian troops. If the attack took place at all, and there is considerable doubt about this since the political sentiment of many of the members of the Amphictionic council, which oversaw the management of the Delphic Games, second in importance only to those held at Olympia, had already joined the Persian cause, then a fortuitous earthquake, common enough in these parts, would have been a sure sign of the god’s displeasure. Historical the episode is not but here once again the narrative has that unlikely mix of fact and unreality that pervades Herodotus’ account of the campaigns to Marathon and Thermopylae.
The fame of Thermopylae unquestionably matches that of Marathon, and rather in the same way that the latter’s renown is constantly reinforced by the running of the ‘marathon race’ every four years at the modern Olympic Games, the former has remained in the public imagination through continued cinematic interpretations. The development of an elaborate modern myth takes its starting point from the epitaph to the dead Spartans for whom this was held up as the apex of not exactly famous victory but glorious defeat. Herodotus, of course, records the inscription and had almost certainly visited the scene of the fight and observed the memorial (Herodt. 7.228):
Oh stranger on the road announce to the Lacedaemonians that obeying their words we lie dead here.
Once here four thousand from the Peloponnese fought against three hundred times ten thousand.
The result of the engagement between the Greeks and the Persians at Thermopylae clearly provided another opportunity for propaganda and an entry into the historical tradition. But was it really like it was recorded by Herodotus and to what extent have modern versions altered the ancient accounts?
For such an apparently great victory Herodotus’ coverage of Marathon is remarkably brief (Herodt. 6.112–115) and is barely any longer than his account of the sack of Sardis at the start of the Ionian War (Herodt. 5.100–102). This brevity probably indicates that the writer was not in a position to retrieve any more than the barest outline of a military engagement, and that his own sources, whether oral or written, were almost completely deficient and that much of what he was obliged to provide his audience with therefore a gloss to cover this shortfall. And what Herodotus relates has long been recognized as both inaccurate and incompatible with any scientific study of the episode. He may wished to have bypassed any detailed account but could hardly ignore Marathon entirely; however, he was in possession of much more material for Xerxes’ invasion and that was where his main focus rests in the later parts of his history. The date of the battle of Marathon (mid-August 490) plainly illustrates the slow progress of the Persians and indicates that this was no lightning strike. The Persians must have left Samos only in May or June and did not arrive on Euboea until early August and so six to eight weeks were spent subjecting the Cyclades to their rule and ensuring a stable line of communications with Asia Minor. The planning up to departure had been meticulous and Datis and Artaphernes can hardly be criticized for poor management of the expedition up to their beaching at Marathon. However, the organization of the troops and especially of the cavalry detachments is open to censure for once landed there seems to have been insufficient attention to the possibility of a hard fight. It should also be remembered that the heavy emphasis on cavalry use indicates not a whim of Darius but that he had discussed the expedition with Hippias long before it set out or before the first preparations were ordered. This gives an insight into this ruler’s habit of consultation and being prepared to accept advice from specialists who knew the region and the inhabitants. The Persians might be forgiven for being over-confident following the ease with which the expedition had gone thus far, but they must surely have realized, especially with the presence of Hippias as adviser, that the encounter here would be much more fiercely contested. It was either this lax discipline or perhaps a decision to go forward to Athens itself without a fight at Marathon that proved the expedition’s undoing. The Persians may simply have decided that with the hiatus in the hostilities and the chance that the Athenians would not engage that they faced a better chance of outright victory by sailing without a further delay to Phaleron. This decision allowed them to be caught without their cavalry in a position for best utilization. The Persians clearly had well trained infantry and these were almost certainly outnumbered by their opponents, yet they did initially well in repulsing and turning the admittedly weakened Greek centre before falling into the trap designed by the Athenian commanders. Finally, the tenacity of the defence of Attica must have surprised the Persians after the ease with which they had taken the islands and especially much of Euboea. Hippias may well have given over an optimistic assessment of the reception the invaders would have received.
For the campaign to Thermopylae it is again the question of supplying an invading army and insuring an easy progress which dominates the narrative in Herodotus and later sources. Admittedly there is a great deal of fanciful and entertaining material about the hubris of Xerxes or the participation of the deity in saving his temple at Delphi or the role of oracles in foretelling defeat for Persian and triumph for Athens. As with the campaign to Marathon, the text has an almost eclectic mix of fact, pseudo-fact, and fiction, but remain the ingredients of a highly successful narrative composition. Its survival intact attested to that success! The defence of Thermopylae, for it was that rather than a battlefield in the true sense, seeing that two opposing armies were not on this occasion drawn up in the classical manner of infantry or cavalry drawn up as a centre and wings, and the tactic of encirclement being attempted by either side. The importance of holding Thermopylae is plain because the main north south route into Attica, Boeotia and the Peloponnese lay through this narrow one sided gorge. The road ran beside the sea to the left as one approached from the north with high overhanging cliffs on the right. There are others routes into Central and Southern Greece, for example via Amphissa to the north of Mount Parnassus but for an army of the size that accompanied Xerxes that was almost impassable so the Persians were forced to go by Thermopylae. And so the interface between sea and land in these campaigns continued as the protagonists regrouped after Thermopylae and Artemisium and events inexorably moved on to Salamis, Plataea and Mycale; but that is another story. And indeed the stage was set for the final showdown but the form those encounters took and the logistical problems that were overcome in undertaking such ambitious adventures were solved in the Persian expedition to Marathon and along the road Xerxes took to Thermopylae. But perhaps most important again in both Marathon and Thermopylae has been overcoming the problems of logistics for an army on the move, the placement of the battlefield in its geographical context and the affect the weather can have on the outcome of any military campaign.
493/2 Mardonius satrap of Hellespontine Phrygia and Thrace.
492 (summer) Mardonius in Thrace where his fleet was destroyed by gales trying to round Mount Athos in the Chalcidice.
491 Datis and Artaphernes appointed to joint command of an expedition to Greece.
490 July Persians attacked and captured Carystos and Eretria on Euboea.
August Persian forces landed at Marathon and were defeated.
August Persians failed to land at Phaleron.
September Persians returned to Asia Minor.
487/84 Rebellion in Egypt against Persian rule.
486 Death of Darius. Accession of Xerxes.
483 Start of construction of a canal across the isthmus to Mount Athos.
481 Xerxes ordered the ‘double bridging’ of the Hellespont.
480 May Xerxes crossed the Hellespont.
May Themistocles at Tempe.
June Persian army and fleet in Thrace and Macedonia.
July Persian victory at Thermopylae, sea battle at Artemisium. Persians attacked Delphi.