The Scythian campaign opened a series of events that caused Persia to become embroiled in a conflict with Athens.

After 520 BCE the tyrant of Athens, Hippias, helped Miltiades, a member of a rival aristocratic clan, to establish himself as tyrant of the Thrakian Chersonnesos, the peninsula that forms the western shore of the Dardanelles. A number of Athenian settlers were planted in the general area that includes the island of Lemnos which blocks the entrance to the Dardanelles. The advance of King Darius into Thrakia in 515 BCE, as the first move in his advance into Scythian, caused a number of the settlers to return to Athens. It could be that the political instability that followed the return to Athens of the dispossessed settlers who had left their city because they were not friendly to the tyrant, set the stage for the plot against the tyranny that succeeded only in killing Hippias' brother (514 BCE).

The failure of the Scythian campaign caused the Greek cities of the Bosphorus to revolt against Persia. These cities were bound for economic reasons to be dependent on the goodwill of whoever dominated the area of Scythian, with the result that once the Persian king had lost they had to throw their lot with the Scythians. These cities may have been afraid of the Scythian revenge raids into Thrakia that took place in the wake of the Persian withdrawal. In 513 BCE the Persian army did not return to Asia by crossing the Bosphorus, as it had been done in the advance, but was forced to move south and cross at the Dardanelles where it could count on the support of Miltiades. The following year the Scythians took their vengeance on Miltiades by forcing him to abandon his possession of the Chersonessos and return to Athens. The Scythians made overtures to Sparta for an alliance, and it could be that this contributed to the Spartan decision to intervene militarily in Athens in 510 BCE in order to overthrow the tyranny. The expelled tyrant Hippias withdrew to Sigeion in the Troad, on the shore facing the former possession of Miltiades, which indicates that Hippias was an ally of Persia.

Up to the moment of his retreat from Scythian, King Darius had been satisfied with a rather loose control of Thrakia and of the Greek cities of the Aegean Sea, because if he had succeeded in conquering Scythian, Thrakia would have been safely in his grip and the Greeks of the Aegean Sea, whose economic life depended in great part on the trade with the Black Sea, would have been at his mercy. The failure of the campaign compelled King Darius to readjust his plans. The revolt of the Greek cities of the Bosphorus was put down, but it had given a taste to the Greeks of the possibility of revolting against Persian domination. When the King withdrew to Asia, he left in Europe the satrap Megabazos with a force of 80,000 men. Megabazos performed well his task of firmly establishing Persian power in Thrakia. He was so effective that he could force the King of Makedonia to become at least nominally a vassal of Persia.

The Paionians who lived in the area of the river Strymon, which today marks the Graeco-Turkish frontier, tried to resist but were defeated by the encircling tactics dear to the Persians. They aligned themselves along the coast in order to block a Persian advance from the direction of the Dardanelles, but the Persians moved inland through the mountains and fell upon the Paionian cities that had been left undefended. A number of Paionian tribes were deported wholesale to Asia Minor. The tyrant of Miletos, Histiaios, who had saved the bridge across the Danube for the retreating Persians, received as a gift some of the territory vacated by the Paionians; but, when he built a fortified town on the Strymon in a position that not only dominated the crossing of the river, but also an area rich in silver mines and timber for shipbuilding, the Persians concluded that this was too much power in the hands of a subject who was already master of Miletos, the richest city of Ionia. The King announced to Histiaios that he was going to make him one of his counselors and invited him to pay a visit to the capital of Susa, where he was kept as an honored guest. The actual exercise of the tyranny in Miletos was entrusted to Histiaios' nephew and son-in-law, Aristagoras.

As acting tyrant of Miletos, Aristagoras intended to serve his masters well, but was not very successful. When civil war broke out in the island of Naxos, the richest island of the Aegean Sea, and the expelled oligarchs took refuge in Miletos, Aristagoras thought that this provided a good opportunity for seizing the island. He made present to the Persians that the possession of Naxos would make possible not only to extend Persian rule to all the Kyklades islands of the central Aegean, but also provide a basis for the conquest of the island of Euboia that lies against the mainland of Greece and is separated by a narrow body of the sea from Attika (V 31). Aristagoras asked help in conquering Naxos from the Persians who sent a force of 200 triremes with a body of embarked infantry, but possibly the Persian contingent was larger than he had bargained for. A quarrel broke out between the Persian commander and Aristagoras on the question of who should be the effective commander of the expedition against Naxos, with the result that the enterprise failed (499 BCE).

Aristagoras became convinced that as a result of this affair his days as ruler of Miletos might be counted and decided that his only chance of salvation was to stir up a general revolt of the Greek cities against Persian rule. The climate for a revolt was favorable because the Persian method of dealing with the subject Greek cities was to let them be ruled by a tyrant. At that turn of time the institution of tyranny had become unpopular among the Greks; the change had been marked by the expulsion of the tyrant Hippias from Athens (510 BCE) followed by the establishment of a democratic constitution. Aristagoras rebelled against Persia and tried to enlist the support of the cities of Greece proper in the name of a crusade for the liberation of all the Greeks subjected to Persia. Only later, at the end of the general revolt of their Greek subjects, the Persians understood what was the political basis of this nationalistic slogan and permitted the Greek cities under their rule to have democratic constitutions.

Aristagoras visited Sparta where with the help of his map of the world he tried to explain that there was a possibility for a successful Greek attack against the very heart of the Persian Empire. But the Spartans were not impressed with this project that to them must have appeared megalomaniacal. He was better received in Athens where he was granted a small but effective force of ships and infantry. Athens had made a formal act of submission to Persia in 508 BCE at the moment in which a democratic constitution had been established in spite of Spartan pressures; but the Athenian policy of friendship with the Persians was wrecked by their insistence that the Athenians permit the return to his city of the former tyrant Hippias (V 96). The Athenians granted Aristagoras the help of 20 triremes which probably carried a hoplite force of about 200 men (this represented about one fourth of the Athenian sea and land strength); the city of Eretria, which as an immediate neighbor of Athens had no particular love for the Athenian cause, contributed 5 triremes with infantry because earlier Miletos had given her assistance in a fight with a neighboring city of the island of Euboia. With these allies in 498 BCE Aristagoras was able to stage a surprise raid from the coast of Asia Minor inland to the city of Sardis, the former capital of the kingdom of Lydia, which was then the capital of the most important Persian satrapy in Asia Minor. The Persian satrap with his troops was resisting in the citadel of Sardis, when the city by plan or by accident went up in flames. Thereupon Aristagoras withdrew, leaving behind the smoldering city, but the event stimulated a universal revolt of the Greek subjects of Persia all the way from the Bosphorus to the island of Cyprus.

The Athenians did not support this revolt since from their point of view the attack on Sardis had ended in disaster, the Persian cavalry having inflicted heavy losses on the Athenian raiders just when they had returned to the coast at Ephesos. The political faction led by the Pisistratids, the clan of the former tyrants, was able to convince the Athenians to adopt a policy of neutrality in the conflict between Persia and her Greek subjects. It could be that it was because of this political shift that in 496 BCE Miltiades left Athens again to reassert his personal rule in the Chersonnesos. Meanwhile the Persians were able to stage a counterattack and to reduce again to subjection the revolted cities. The combined fleet of the Ionians which gathered up to 353 triremes proved no match for the Persian fleet of 600 triremes (VI 9). The successful Persian counter-offensive culminated in 494 BCE with the siege of Miletos, which was blockaded by 600 Persian triremes, according to Herodotus (VI 9). The Greeks were able to align 353 triremes, but when battle to break the siege of Miletos was engaged, some of the Greek contingents lost heart and deserted, with the result that the Persians stormed the city and destroyed it; its inhabitants were sold as slaves. In their counterattack the Persians relied on the help of the Phoenicians, the mortal enemies of the Greeks, whom earlier they had carefully excluded from the operations in the Aegean and Black Sea area. In 493 BCE the Phoenicians went after Miltiades since they could not tolerate the domination by a Greek of the most strategic point on the line of access to the Black Sea. It must have been with great glee that the Phoenicians in the name of the Persian King burned the Greek cities of Byzantion and Kalchedon that dominated the passage of the Bosphorus (VI 33). Miltiades barely escaped with his life from the hands of the Phoenicians and returned to Athens. Miltiades' eldest son was captured by the Phoenicians who brought him to the Persians, but they treated this prisoner with great honor and gave him a Persian wife (VI 41). Upon Miltiades' return to Athens his political enemies tried to get rid of his embarrassing presence by bringing him to trial as a supporter of the institution of tyranny. Several scholars agree that it was on this occasion that Miltiades concocted the distorted version reported by Herodotus of what had happened at the bridge over the Danube about twenty years earlier; his version of the events had the purpose of proving that he had been an opponent of tyranny and enemy of Persia as far back as that. The ferocious revenge taken by the Persians on the city of Miletos had caused emotional shock in Athens and embarrassed the pro-Persian party. As a result there prevailed in Athens the policy of alliance with Sparta which, however, was not popular, since it meant subservience of Athens to Sparta, given that the latter had a much larger land army. Themistokles proposed the policy of building Athens as a great naval power, by which she would be independent both of Persia and Sparta, but this policy was adopted only after the battle of Marathon.

The year 493 BCE was used by the Persians to consolidate their position in the Greek cities of Ionia and to introduce their new policy that favored democracy over tyranny. Although the Persians had succeeded in reducing their former subjects, it became clear to them that they could not feel safe as long as some of the Greek cities were still independent. In 492 BCE the Persians initiated a campaign to conquer Europe. A large army after having crossed the Dardanelles imposed Persian will on the territories north of the Greek mainland, Thrakia and Makedonia, including the Greek island of Thasos off the coast of Thrakia. They would have continued by advancing into northern Greece, but the project had to be abandoned when the Persian fleet that was moving along the Thrakian coast in order to support the land forces was partly wrecked when caught in a storm while turning the promontory of Mount Athos. Herodotus relates that it was said that the Persians lost 300 triremes and over 20,000 men (VI 44). After this mishap the Persians decided to experiment with completely different tactics: Instead of using the navy to support a large land army, the navy would be used as the main instrument of attack. The year 491 BCE was spent in putting together and training a not-too-great (by Persian standards) but choice force of infantry and cavalry to be carried by the fleet and to be used in amphibious operations. The infantry was carried by the triremes which were in the usual number of 600 (VI 95), but the cavalry was carried on other triremes that had been specially adapted. Possibly the Persians anticipated what was done later by the Athenians who used their old triremes to transport the cavalry (Thuk. II 56).

The purpose of this naval force was to strike terror among the Greeks and in particular to convince the Athenians to break their alliance with Sparta and to adopt a pro-Persian policy. The prestige of the Persian empire required that those independent Greek cities, Athens and Eretria, that had participated in the raid against Sardis be meted exemplary punishment; furthermore, the Persians may have assumed that once these more pertinacious opponents had been disposed of, the remaining cities of Greece proper could be persuaded to accept Persian suzerainty. The Persians aimed at achieving their objectives with the most economical means. An opportunity for successful maneuvering seemed to have offered itself when the tyrant Hippias who had been expelled from Athens fled to the Persian court at Susa, where he asked for assistance claiming to have supporters among the opponents of the newly established democratic constitution of Athens. The Persians were hoping to succeed with the help of the Pisistratid party within Athens, and for this reason the Persian fleet carried along the exile Hippias who also was used as political and military advisor. Hippias was unquestionably popular within Athens, so much so that earlier the Spartans had tried to win Hippias to their side and had invited him to Sparta, proposing to restore him as a tyrant in Athens. (V 91, 93).

The Persian fleet assembled at Samos in the spring of 490 BCE and as a first step moved to attack the island of Naxos. The inhabitants of this island fled to the mountains, abandoning the city of Naxos to the Persians, who put it on fire. The Persians continued subduing other islands so as to close a circle around Athens; the key island of Aigina in the Saronic Gulf in front of Athens had already been an ally of Persia for a few years. Finally the Persians landed at Karystos at the southern tip of the island of Euboia, and after a brief siege forced the people of Karystos to join the Persian side. Having acquired this foothold in Euboia, the Persian fleet, after setting up three bases on the coast of the territory of Eretria, disembarked the cavalry and infantry. The plan to move against the Greek mainland by seizing the island of Naxos and then the island of Euboia had been already considered by the Persians in the unfortunate expedition of 499 BCE on the suggestion of Aristagoras. The people of Eretria withdrew within the city walls, but they lacked the full determination to resist; after six days of Persian siege, on the seventh day two of the leading citizens opened the gate to the enemy. The male citizens of Eretria were made prisoners and taken to the small island of Aigileia (the present Stira) which was one of the bases of the Persian fleet on the coast of Eretria and faces the bay of Marathon in Attika.

As is stated by Platous (Menexenos| 240 C; Laws 698 D), all the Greek cities were paralyzed with fear before the Persian forces. There was no naval force in Greece that could challenge the Persians at sea (at the time the Athenian navy had not more than 80 triremes) with the result that the Persians could freely move to land their cavalry and infantry where they chose. No other Greeks came to the help of Eretria except for 4000 Athenian settlers on the island of Euboia who were instructed by their mother country to help the Eretrians, and these too saved themselves before the battle was over. The technical superiority of the Persian amphibious force was such that it could strike anywhere with impunity. No Greek city would help another because nobody could tell for sure where the blow would fall next.

The cavalry added to the navy completed the principle of absolute mobility on which the strategy of this operation was founded. There are critical historians who consider that this was not true. For instance, De Sanctis and Guilio Giannelli claimed that Herodotus' mention of a Persian cavalry force carried on triremes fitted for the transport of horses (VI 94, 95, 101, 102) is preposterous, because it would have been impossible to transport horses on a long sea voyage. Giannelli, who has written a special essay on the battle of Marathon, added that the Persian triremes were 100 and not 600, so that they would not have been an overpowering threat to the Athenian navy or other Greek navies. That horses could be transported on triremes is evidenced by Thukydides (II 56, VI, 43) and later Athenian documents. Thirty horses could be loaded on a trireme by reducing the oarsmen to 60, that is, one third of the regular number. Whoever invented the trireme reckoned sexagesimally, since in principle there were three rows of 30 oarsmen each on each side, with a total of 180. The lesser fighting ship used earlier by the Greeks, the penteconter, had been conceived by decimal reckoning, since it had two rows of 25 oarsmen each on each side with a total of 100. Probably the 30 horses were placed crosswise in the triremes, one for each bench of oarsmen. It has been argued that when horses were loaded on triremes all the regular seats were removed and the 60 oarsmen sat on the stormdeck that covered the seats of the regular three rows of oarsmen. The assertion of some critical historians that horses could not have been carried by the Persians in their island-hopping operation is gratuitous when we know that in 415 BCE the Athenians sent all the way to Syracuse in Sicily a force of 100 triremes which was fitted out for naval combat and carried 4000 hoplites and 300 horses (Thuk. VI 31, 43).

According to the account of Herodotus (VI 110), the Persians waited a few days after the fall of Eretria and then moved against Athens, confident that the Athenians would behave like the Eretrians. The Persians landed at Marathon which is a few miles across the sea from Eretrian territory and was considered the area of Attica "most suitable for the landing of cavalry" (VI 102). The landing at Marathon must have been conceived as being part of the same general operation as the landing at Eretria, because Herodotus states that when the Athenians learned of the landing at Marathon, the part of Attika nearest to Eretria, "they too came to the rescue" (VI 103). This sentence refers to the fact that when the Persians landed at Eretria, the Athenians had not answered the call of the Eretrians "to come to the rescue" (VI 100), except by so instructing the Athenian settlers who were already in the island of Euboia. Marathon is at the northern limit of Attika at a distance of 42 kilometers from Athens.

The aim of the Persian tactics has been clarified by the discovery in Athens in 1933 A.D. of a stone that contains a most important inscription. There are reasons to believe that this stone was part of the monument that listed the names of the 192 Athenians who died in the battle of Marathon; the monument had been erected on the Athenian Akropolis and was destroyed by the Persians during their occupation of Athens in 480 BCE The inscription contains the text of an epigram of four lines honoring the heroes of Marathon. Several lines are missing, but the essential meaning of the epigram is clear; it can be paraphrased as follows: these men must have had a heart of steel to fight outside the city gates and align themselves in combat formation against an enemy that planned to burn the coastal regions; thereby they saved the city, for they compelled the Persian force to withdraw. The text of the epigram agrees with the expressions used by ancient historians, but adds the new information that the Persians planned to burn the coastal region. But, in truth, this piece of information agrees with the statement of Herodotus that the Persians, "having subdued Eretria, waited a few days and then sailed to Attika, ravaging the land a great deal and believing that they would do to the Athenians what they had done to the Eretrians" (VI 102). This is the reading indicated by the manuscripts of Herodotus, but some editors have changed it because they did not understand the Persian tactics. They have altered the verb kategazw, which in the language of Herodotus means "to ravage" a country.

The Persians hoped that the Athenians would behave like the Eretrians, that is, withdraw within the city walls. If this had taken place the Persians would have been free to ravage the countryside all along the shore, destruction that would have been particularly painful to the Athenians because it was the month of September, the month in whih grapes, olives, and fruits are gathered. The Persians planned to follow the method used later by the Spartans in the Peloponnesian War when they systematically ravaged the Attic countryside while the Athenians kept within the walls; but the Persians could have been much more effective than the Spartans because they also controlled the sea and could block all supplies. The Persians hoped that by these tactics they could force the Athenians to change their foreign policy. The return of Hippias to Athens would have been the guarantee that this city would definitely stay on the Persian side. As Fritz Schachermeyr has pointed out, the Persians did not want to destroy Athens, but on the contrary were eager to have the Athenians as allies and to use them as their chief instrument of penetration in Greece and, I would say, the Mediterranean. The Persians did not give up the hope of an alliance with Athens even after the battle of Marathon and conducted negotiations with Themistokles to this end.

The Persians had established their base at Marathon because they assumed that the Athenians would never dare to attack them that far from Athens (some 28 miles) since to move the army there would have meant to leave the city unprotected. But Miltiades convinced the Athenians to take this risk and to march to Marathon. Sources different from Herodotus report that a decree was put before the popular assembly sanctioning the policy that "it is necessary to go out," and not to stay within the walls. The decree of Miltiades agrees with the text of the mentioned epigram, the wording of which is paraphrased in a statement of Cornelius Nepos (Miltiades V) that Athenienses copias ex urbe eduxerunt locoque idoneo castra fecerunt. The Athenians set up camp at the very margin of the plain of Marathon in an area that was higher and rocky so that the Persians could not use their cavalry against them. Practically all scholars agree that the Athenians aligned themselves across the valley of Vrana, the very bottom of which is about 2000 meters from the shore of Marathon. The Persians were aligned along the shore in front of their ships that were either beached or moored on the sandy beach of the bay of Marathon. The result of the extremely daring maneuver of Miltiades was that the Persians could not move from their position; they had landed in Attika but they had found themselves pinned against the shore. Their strength was the cavalry but they could not use it against the enemy positions, and if they tried to move either north or south along the shore they would be exposed to a flank attack. If they tried to embark their forces they could be exposed to an attack while they were in disarray. The infantry could be embarked quickly, but the embarking of the horses must have been a lengthy operation.

Miltiades followed the strategy recommended by the best German generals when the Allies were planning their landing in France: one should try to stop the Allies within the first two kilometers from the shore or not try at all. According to those generals the alternative was to withdraw all the way to the Siegfried line. Between these two alternatives that were similar to those considered by the Athenians, the Oberkommando der Wehrmacht and Hitler chose a middle course, which proved unfortunate.

The surprise move of Miltiades had achieved a victory without a fight. The choice of Marathon as a landing place had proved an error for the Persians; Miltiades took advantage of this error by leading his troops to the valley which is today called Vrana where he could align his army in a position protected on one side by Mount Agrieliki and on the other side by Mount Kotroni, both steep and rocky hillocks. The problems is that of explaining why the Persians, usually so well informed in matters of geography, chose such an unfortunate location. Herodotus states that this was the area of Attika "most suitable for the landing of the cavalry" (VI 102). Most interpreters understand differently the verb enippeusai and translate "most suitable for the deployment of cavalry" and then continue by observing that Herodotus was mistaken because the place was most unsuitable, whereas there were other suitable places. It is the usual technique of interpreting the text of Herodotus so as to let him appear as absurd as possible. As it is observed, the plain of Marathon was limited to the north by a large lagoon and to the south by a smaller one and was further cut by torrents, so that the Persian army was restricted to a relatively small area. This proves that the Persians chose a location where they could easily defend themselves if they were attacked while landing the cavalry. Furthermore, the long sandy beach of Marathon would permit to pull the triremes on shore; this must have been necessary in order to discharge the horses from the triremes, which must have been a long and complex operation. The time needed to disembark and embark the horses may have been the greatest single cause of failure in the campaign of Marathon. It is significant that in the following campaign of 480 BCE the Persians loaded their horses on "small merchant ships" (VII 97) instead of triremes.

Herodotus explains that the Persians landed at Marathon on the direction of Hippias (VI 107). Immediately following this statement Herodotus relates that Hippias had a dream followed by a coughing which he interpreted as indicating that the operation would fail; any modern psychoanalyst would also interpret this coughing as the expression of a subconscious desire for defeat. The coughing which caused Hippias to lose a tooth on the sand of the beach followed a dream of intercourse with his own mother. The Oedipal dream can be understood as significant because Marathon had been chosen by Hippias' father, Peisistratos, for a landing which opened the way for a victorious march on Athens and the establishment of tyranny (I 62). It would seem that Herodotus intends to give a psychoanalytic explanation of why Hippias, who obviously knew the terrain to perfection, made a bad choice in advising the Persians. Hippias wanted to do as well as his father, but his wish took the character of an Oedipal wish to outdo the father and possess the mother, with subsequent guilt and self-defeat.

The Persian strategy and that of Miltiades resulted in a stalemate because the Persians could not move beyond Marathon as they had planned, being unable to use the cavalry on rocky ground, while the Greeks, because of the Persian cavalry, could not attack the Persians that were encamped on the shore of the plain. The Athenians had sent a messenger to Sparta asking for help and were waiting for the arrival of the Spartan forces. The Persians were apparently waiting for dissensions to break out among the Athenians, since they had banked on Hippias' promise that some of his former subjects would rally to his side.

The stalemate had lasted through the tenth day after the Persian landing at Marathon, when on the morning of the eleventh day the Athenians went to the attack. Herodotus (VI 112) asserts that the Greeks resorted to a method of warfare never used before: instead of marching towards the enemy, they engaged in a race (dromos) for the entire distance that separated the two armies, not less than 8 stadia (probably artabic stadia of 8 to a Roman mile, or 1500 meters). The Greek line was strong on the wings and weak at the center. This distribution of the forces was to be expected since the wings had to withstand the strong Persian cavalry force; in ancient warfare the function of the cavalry was to protect the wings and to harass the enemy wings. In the absence of the cavalry, the Persian wings were unusually weak. The Persian center broke though the enemy lines, but the Greek wings closed behind the Persians. The Persians had no other recourse except to try to run back to their ships. According to the report, the Athenians killed approximately 6400 Persians at a loss of 192 of their own men, but they did not succeed in getting hold of the Persian ships before they could take off from the beach; only seven ships were captured (VI 115).

All scholars agree that Herodotus' account contains "not a few patent contradictions." Bury regrets that at the time of Marathon there was not "a contemporary historian lice Thukydides to ask searching questions and record the truth." (44)

Macan declares that the story of the Athenian advance against the Persians is "probably genuine," as long as we assume that it was a march at double speed and not a race, but "the rest is distortion, exaggeration, inconsequence, glorification." (45)

Most scholars are only somewhat less critical than Macan. A group that is more radical than Macan claims that actually it was the Persians who went to the attack; their argument is that, since Herodotus does not mention the participation of the Persian cavalry in the battle, it can be inferred that the Persians had decided to attack the Athenians on the hills.

Several other explanations for the failure of the Persian cavalry to participate in the battle have been offered. Grote suggested that the Athenians caught the Persian horsemen by surprise so that they did not have time to get on their mounts. Among the recent writers, H. G. L. Hammond claims that the cavalry was pasturing further north and did not arrive in time for the beginning of the battle. By the time it arrived, it could not be deployed because the armies were fighting at close quarters. (46)

Some scholars claim that the Persian cavalry had not yet arrived from Eretria, even though Herodotus states that the Persians had landed at Marathon a few days after the capture of Eretria and that the battle took place on the eleventh day after the landing. In order to explain why the cavalry was still at Eretria, Munro adds the further suggestion that the Persians had landed at the same time at Karystos, Eretria, and Marathon with the result that their forces were scattered in three separate actions. (47)

I have already mentioned the opinion that Herodotus is completely wrong when he states that the expeditionary forces sent to Greece included horsemen. At the opposite extreme there are the critics, such as Johannes Kromayer (48) and Hans Delbrueck, (49) who claim that Herodotus is in error when he assumes that the Persian cavalry did not participate in the battle. (50)

Among the minority of scholars who do not assume that the Persians were wanton in their military actions and that Herodotus is fanciful in his report, there prevails the opinion that the cavalry was absent from the battle because it had been embarked, since the Persians were planning to withdraw from Marathon and to land at the Phaleron, the outer harbor of Athens. (51)

The withdrawal of the Persian cavalry is mentioned in the dictionary of Suidas where he explains the meaning of the idiom xwris ippeis "without cavalry, the cavalry is off": "As Datis who had landed in Attica was retiring, the Ionians by climbing on trees signalled to the Athenians the cavalry is off." "As Miltiades learned in this way of their withdrawal, he engaged battle and won. Hence, this expression is used proverbially to refer to those who are breaking their military formation." According to this text the Ionians who were serving in the Persian fleet betrayed their Persian commander by informing their fellow Greeks; the withdrawal must have taken place at night because otherwise the Athenians encamped above Marathon would have seen by themselves what was taking place.

We may disregard the opinion of those, such as Schachermeyr, who question the account of Herodotus by claiming that the Persians never planned to land at the Phaleron after the withdrawal from Marathon (fantaisies had said Hauvette of this [52]).

An opposite position is taken by Anton E. Raubitschek who claims not only that the Persians planned to land at the Phaleron, but actually landed and were defeated there in a battle with the Athenians; (53) neither Herodotus nor any other Greek source hints at the occurrence of this repetition of the battle of Marathon.

Among the other more recent writers on the subject, A. W. Gomme was willing to accept Herodotus' account as having some value. Gomme gave an explanation of what happened at Marathon that to my mind is convincing and in agreement with the texts; but, since to assume that Herodotus said something sensible is a serious offense for modern scholarship, before presenting his views Gomme engaged in elaborate expiatory rites. He began his article thus:

Everyone knows that Herodotus' narrative of Marathon will not do. Many improvements have been suggested: some good, some bad. . . . My theme is rather this: if we reject Herodotus, are we justified at all in correcting, or adding to, his narrative, or ought we just to sit back, and say nothing, because correction is arbitrary? (54)