Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

Marathon and the myth of the same-day march
Holoka, James P
Greek, Roman and Byzantine Studies; Winter 1997; 38, 4; ProQuest
pg. 329

Marathon and the Myth 
of the Same-Day March 

For R.D.G. & J.L. 
O"I1t1cpOv oiOpov 

James P. Holoka 

In every description of a battle there is a neces­
sary lie, resulting from the need of describing in 
a few words the actions of thousands of men 
spread over several miles, and subject to most vio­
lent moral excitement under the influence of fear, 
shame, and death. 

Leo Tolstoy, "Some Words about War and 
Peaa ~ (tr. A. Maude), 1868 

T HE ANCIENT SOURCES for the Battle of Marathon are often 
confusing and contradictory. None is contemporary with 
the event. Herodotus (Book 6), the earliest and best 

source, composed his account some fifty years after the battle; 1 

Cornelius Nepos wrote more than four centuries later,2 
Plutarch nearly six. Nepos and Plutarch may have drawn on the 
account of Ephorus, writing over a century after the fact. 
Although we must rely on these sources and assess them with 
the greatest care, no reconstruction based on them will depict 
Marathon with real precision. Nonetheless, a more plausible 

I H. R. Immerwahr, "Historiography," in P. E. Easterling and B. M. W. 
Knox, edd., The Cambridge History of Classical Literature I: Greek Litera­
ture (Cambridge 1985) 427f: "The work cannot have been published until the 
early years of the Peloponnesian War. There are, however, at least three 
reasons why this date is not especially significant: (1) the work was known in 
Athens prior to publication; (2) it has a long prehistory in oral composition, 
and conceivably in partial advance 'publication' (or better, circulation); (3) at 
the time of its ultimate publication the work was something of an anach­
ronism, for it reflected an earlier world view and style of writing. Its 'ideal' 
date is closer to the 440s than the 420s." 

2 See C. Han/et, "En Marge de Cornelius Nepos: (i) Miltiade, I'heros de 
Marathon," EtClI0 (1941) 381-93; E. M. Jenkinson, "Genus scripturae leve: 
Cornelius Nepos and the Early History of Biography at Rome,» AN R W I.3 
(Berlin 1973) 709-15. 

329 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

330 MARATHON AND THE SAME-DAY MARCH 

recreation of the battle and its aftermath can be achieved by due 
appreciation of logistical issues and the extent of human physical 
endurance. The lack of such an appreciation has been most 
damaging specifically in scholarly reconstruction of the time­
frame of the Battle of Marathon and the march back to Athens. 

Plutarch's statement that the Athenians, "fearful that they [the 
Persians] might take the undefended city, hastened back to 
town ... and reached it the same day,"3 has been the corner­
stone of a persistent and very influential version of Marathon, 
particularly because it has shaped many scholars' views of 
Herodotus' account. 4 The essential elements of what may be 
called the modern vulgate are as follows: 

(1) the battle began at or very shortly after dawn, about 5:30 A.M.; 
(2) the battle was completed within three or four hours, i. e., by 

about 9:00 A.M.; 
(3) after the battle, a signal was given to the Persian fleet, encour­

aging it to sail round Cape Sunion; 
(4) the Athenian army, realizing such a run by sea to Athens was 

underway, made a ca twenty-six-mile forced march overland to 
the city on the very day of the battle; 

(5) the Persian force arrived by sea at Phalerum harbor towards 
evening the day of the battle, but did not disembark for fear of 
the Greek army, which had already arrived; 

(6) the day of the battle and march was 11 September 490. 

Up to the nineteenth century, many historians accepted this 
same-day version of these events. George Grote, in the most 
influential and widely read Victorian history of ancient Greece,s 
wrote that «on the very day of the battle, the Athenian army 
marched back with the utmost speed [to Athens] ... which they 
reached before the arrival of the Persian fleet."6 But more 

3 Arist. 5.4: !j)0~T)9EV'tEc; fl~ 't~v 1tOA.1V tpT)flOV A«~ro<H 'troY UfluvoflEvrov, 'tUt~ 
flEV EVVEU <puA.al~ l]1trlYov'to 1tpO~ 'to ao'tu KUt KU'tT]VUOUV UU9T)flEpOv. 

4 Thus e.g. N. A. DOENGES. "The Campaign and Battle of Marathon," His­
tona 47 (1998: hereafter 'Doenges') 15 n.27, writes: "Herodotus, 6.116, places 
the return of the victorious Athenian army to the city on the day of the 
battle." He does not. 

S Cf J. Vaio, "George Grote," in W. W. Briggs and W. M. Calder III, edd., 
Classical Scholarship: A Biographical Encyclopedia (New York 1990) 119: 
Grote's· History of Greece fundamentally shaped the perception of ancient 
Greek culture and history among English-speaking scholars and readers in 
the Victorian age and had a major impact on the Continent as well, especially 
in Germany ...... 

6 History of Greece2 (New York 1861 [London 1859]) IV 351. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 331 

critical scholars challenged such naive acceptance of Plutarch 
and argued that the return to Athens must have taken place the 
day after the battle. George Rawlinson, for example, observed 
that "if the Greeks performed this march, one of seven hours, 
as Herodotus has been thought to imply, the very same after­
noon, it would be about the most remarkable of the events of a 
very memorable day." He suggested that Athenian knowledge 
of the intention of the Persian force to sail for Athens" is ex­
tremely likely to have been early the next day. "7 Miiller­
Striibing argued that the same-day march was both "a physical 
impossibility" and beyond what any commander could in 
conscience demand of battle-weary troops: "es moralisch un­
moglich ist, [da6] der athenische Feldherr diesen Marsch ange­
ordnet haben solI. "8 Among Anglophone scholars, this skep­
ticism is typified by How and Wells in their still unsuperseded 
(though outdated) commentary on Herodotus: "the distance 
from Marathon to Athens ... is more than an army could march 
after a pitched battle .... Both march and voyage ... should be 
assigned to the following day."9 

But the same-day myth was given a new, much stronger lease 
on life beginning in 1959, when the first edition of N. G. L. 
Hammond's soon-to-be-standard textbook appeared. In his ac­
count of Marathon, Hammond reports the results of a personal 
field test: "I walked fast from Athens to the mound at Mara­
thon in 6 hours and returned the same day to Athens in 7 
hours." 10 The march of the Greek army was thus no "physical 
impossibility" after all. 

My purpose here is not to challenge every element in Ham­
mond's reconstruction of the Battle of Marathon; and I shall, in 
fact, follow his version of the phases of the fighting and the 
movement of the opposing forces on and around the plain of 

7 The History of Herodotus (New York 1860) III 413 n.9; see also H. Muller­
Strobing, "Zur Schlacht von Marathon," Jahrb. f cl. Phil. 129 (1879) 433--48, 
followed by G. BUSOLT, Griechische Geschichte 2 (Gotha 1895: hereafter 
'Busolt'). 

8 Muller-Striibing (supra n.7) 445. 
9 W. W. How and J. WELLS, A Commentary on Herodotus (Oxford 1912: 

hereafter 'How and Wells') II 113. J. A. R. MUNRO, "The Battle and After," 
CAH IV (1926: 'Munro') 250f, represents a half-way house: "the Athenian 
generals had no time to lose. The Persian fleet would be arriving off Phalerum 
on the morning after the battle. Plutarch is doubtless right in putting the 
march back to Athens on the same day as the battle, and Herodotus implies 
as much" [my emphases]. 

10 A History of Greece to 322 B. C. (Oxford 1959, 1967 2, 19863) 216 n.2. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

332 MARATHON AND THE SAME-DAY MARCH 

Marathon on the day of battle. Nor shall I be concerned with 
other problematic features of Herodotus' account: Philippides' 
run to Sparta and back, the possibility that prevailing Etesian 
winds in September dictate an October invasion date, or the 
presence, size, disposition, and movement of the Persian 
cavalry force. ll 

In 1968, Hammond reinforced his anecdotal evidence for the 
myth's plausibility in a long paper, which amply attests his con­
trol of the ancient sources, exhaustive familiarity with modern 
scholarly literature on the battle, and detailed knowledge of the 
topography of Attica. Republished in revised form, it provided 
the basis for his discussion of Marathon in the second edition of 
the Cambridge Ancient History.12 Hammond's imprimatur 
quickly ensured renewed and wide currency for the same-day 
myth, not only in the specialist literature, but also in books 
written for wider audiences: Peter Green, in a popular history 
of the Persian Wars, writes 

as if they had not done enough already, the Athenians once 
again achieved a near-miracle .... [They J at once set off back to 
Athens, each man for himself: 'as fast as their feet could carry 
them,' Herodotus says, and one can well believe it .... They can 
scarcely have got there before four in the afternoon, and 

lIOn the basis of Suda s.v. xwpl<; l.1t1tCl<;, it has been contended that the 
Athenians (having secretly been informed by Ionians in the Persian host) had 
to make it back to Athens the very day of the battle, because the Persian 
cavalry had been dispatched by ship for Athens the night before. P. Green, 
per ep. elec. (14 October 1997): "The knowledge of what Datis was doing, 
combined with the chance to snatch an unexpected victory, meant that every­
thing was done with maximum speed, and as soon as possible." But this scen­
ario assumes a great deal: that the cavalry was "away" on board ships to 
Athens and not simply elsewhere on the expansive Marathon plain; that the 
Persian cavalry contingent was large enough by itself to pose a threat to the 
city-in fact, we do not have reliable evidence for its numbers; and that 
ancient sources, both literary (Nep. Milt. 5; Paus. 1.32.3) and plastic (the Stoa 
Poecile painting, a sarcophagus from Brescia, and a frieze from the temple of 
Athena Nike on the Acropolis), are wrong in representing the cavalry as 
present at the battle. See, in general, G. Shrimpton, "The Persian Cavalry at 
Marathon, H Phoenix 34 (1980) 20-37; J. A. S. Evans, "Herodotus and the 
Battle of Marathon," Historia 42 (1993) 293-99; A. Santosuosso, Soldiers, 
Citizens, and the Symbols of War: From Classical Greece to Republican 
R orne, 500-167 B. C. (Boulder 1997) 33; Doenges 5, 14. 

12 "The Campaign and the Battle of Marathon," IHS 88 (1968) 13-57, re­
vised in Studies in Greek History: A Companion Volume to A History of 
Greece to 322 B.C. (Oxford 1973: hereafter 'Hammond') 170-250; CAH2 IV: 
Persia, Greece and the Western Mediterranean c. 525 to 479 B.C. (1988) 
506-17. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 

Datis's squadron may well have sailed into Phaleron roads an 
hour or less later. Yet that one hour made all the differ­
ence ... P 

333 

In what follows, I shall examine, item by item, the chief prob­
lems inherent in the durable fable convenue of the same-day 
march. 

1. Commencement of Fighting 

No ancient author states what time of day the battle began. So 
far as Herodotus informs us, it could have been 9:00 A.M. or 
noon as easily as first light. That it might have been dawn does 
not mean it must have been dawn. In fact the sole reason for 
assigning commencement of battle to so early an hour is to lend 
credibility to the same-day myth: the Athenians must have be­
gun fighting at dawn because they must have had time to march 
back to Athens the same day. Q.E.D. I do not exaggerate the 
circularity of the argument. Pritchett writes that "the accounts 
of the precipitous haste with which the victorious Athenians 
marched home in order to anticipate a Persian landing in the 
Bay of Phaleron suggest that the battle took place in the 
morning."14 So too, Hammond (210f): "if the march took 8 or 9 
hours and the army reached Cynosarges in daylight, they 

13 Xerxes at Salamis (New York 1970) 36. So too V. Hanson, The Wars of 
the Ancient Greeks (London 1999) 85f: 'To cap off the day's killing, the tired 
Athenian hop lites then trekked en masse for eight hours over the pass to save 
their unprotected city from the retreating Persian fleet, displaying remarkable 
endurance and confidence, and so capturing for ever the collective 
imagination of the West." Also F. Chamoux, La civilisation grecque a 
l'epoque archa'ique et classique (Paris 1963) 98: "Miltiade et les strateges 
ramenerent l'armee vers Athenes Ie jour meme et arriverent a temps pour 
prevenir une tentative de debarquement au Phalere. Voyant que la cote etait 
defendue, Datis et Artapherne n'insisterent pas et regagnerent I' Asie." Cf A. 
R. BURN, Persia and the Greeks: The Defense of the West, 546-478 R.C. (New 
York 1962: hereafter 'Burn') 251f; J. V. A. FINE, The Ancient Greeks: A 
CriticaL History (Cambridge [Mass.] 1983: 'Fine') 285; A. B. Lloyd, Marathon: 
The Story of Civilizations on Collision Course (New York 1973) 20lf, who 
apparently thinks the Greeks marched all night: «it was a long tramp for 
muscles chilled in the open hills and stiffened on the battlefield, but the ranks 
were in good heart and the necessity urgent." Similarly C. Meier, Athen: Ein 
Neubeginn der WeLtgeschichte (Munich 1995) 253: "Nach kurzer Ruhepause, 
vielleicht noch in der Nacht, brach die Armee in Richtung Athen auf. Sie 
muBte vor der Persern dort sein." 

14 W. K. Pritchett, "Marathon," CPCA 4 (1960) 173. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

334 MARATHON AND THE SAME-DAY MARCH 

started about 9 a.m. or 10 a.m.; for darkness fell soon after 6.30 
p.m .... The shield signal then was made before 9 a.m .... It fol­
lows that the battle started very close to dawn, i.e. very close to 
5.30 a.m." 

In fact, there is reason to think the Athenians did not attack at 
first light. Though ancient testimony for this mundane element 
of daily routine on campaign is lacking, it is very likely they 
took a meal. 15 Although Marathon may not, like some hoplite 
battles, have taken all day,16 according to Herodotus the fighting 
did take "a long time."17 Any responsible commander will have 
seen to it that the troops fortified themselves with nourishment 
before going into action. 

The army's leaders must also have attended to indispensable 
religious obligations prior to the battle. Greek armies typically 
prepared for a day's combat by consulting the omens and pay­
ing homage to the gods by sacrifice. 18 And Herodotus (6.112) 
indeed says that the army charged into battle after favorable pre­
liminary sacrifices ('tel O"q>ayw: £ytv£'to 1(o.1.,a). Pritchett has dis tin-

15 One thinks of Odysseus' admonition to Achilles at Il. 19.157ff, 162f: oiJ1e 
6A.iyoll xpOvov Eo'tal q)'UA01tle;. dl't' av 7tpoo'tov Of!lA"OWOl q>uAayyee; av5poov. EV 
5£ eEl>e; 7tVEUOll f!EVOe; af!q>o'tEpOtOlV .... ou ya.p aviJp 1tpo1tav ~f!ap Ee; TjEAlOv 
Ka'ta5uv'ta aKf!T\VOe; OtWlO 5\)vTjo£'tat av'ta f!uX£oSat ("For the combat will 
last no little while once the masses of men meet in battle and the god breathes 
strength into both sides .... No man unnourished by food will be able to fight 
his foe all the livelong day to the sun's setting."). 

16 See the table showing duration of land battles according to Greek histori­
ans in W. K. Pritchett, The Greek State at War I-V (Berkeley 1971-91) at IV 
47-50. 

17 6.113: f!aXOf!EVWII 5i: 'tip MapaSipvl XPOVOC; EytV£'tO 7toAAD". 
18 Cf Socrates at Xen. Oee. 5.19: Kat wi>" f!EII Ell 'tip 7tOAEf!CP opi!." O{f!Ol 7tpO 

'toov 1tOA£f!lKooV 1tPU~£WV E~apeoKof!EIIOU" 'toue; ewu" Kat E1tepW'tooV'tae; 
eUOtale; Kat OiWVOle;. 0 'tl 'tE XPl) 7tOl£lV Kat 0 'tl f!r1 (" And you observe, I sup­
pose, that men engaged in war try to propitiate the gods before taking action; 
and with sacrifices and omens seek to know what they ought to do and what 
they ought not to do," tr. E. C. Marchant (Loeb: Cambridge [Mass.] 1923). 
Cf Onas. 1025: f!r1't£ 5E de; 1topdav E~aYE'tw to O'tpUt£\)f!a f!TJ't£ 7tPOC; J.LUXT\V 
't<X'ttE'tW, f!l) 7tpon:pov Suouf!£VOC; ("The general should neither lead his army 
on a journey, nor marshall it for battle without first making a sacrifice," tr. 
Illinois Greek Club, Aeneas Tacticus, Aesclepiodotus, and Onasander (Loeb: 
Cambridge [Mass.] 1923). M. H. Jameson, "Sacrifice before Battle," in V. Han­
son, cd., Hoplites: The Classical Greek Battle Experience (London 1991) 197: 
"E very stage in the process that led up to the clash of hoplite phalanxes on the 
field of battle was marked by attention to the gods," and passim. Cf Lloyd 
(supra n.D) 185ff and, on Macedonian practice, his • Philip II and Alexander 
the Great: The Moulding of Macedon's Army," in A. B. Lloyd, ed., Battle in 
Antiquity (London 1996) 184. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

Goo 

SOfOS 

• 

Athens ? 
via I 

Pallene \ 

(a) Map of Marathon 

HOLOKA PLATE 1 

o 
1 

miles 

(b) Athenian Tetradrachm, ca 480-69, after 
C. M. Kraay, Greek Coins (New York 1966) p1.119.360 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 335 

guished two ritual sacrifices in the typical preparations for 
battle: the first in the camp, "'in the course of which ... the 
omens had to be interpreted as favorable before the action 
could begin," and the second, '" of a supplicatory and propi tia­
tory nature, performed immediately before the action began. "19 

In addition, it must have taken longer than usual to array the 
troops in line of battle, a line over a mile long,20 arranged not in 
the standard phalanx formation eight ranks deep but shallower 
in the center to ensure sufficient breadth and strength on the 
wings to forestall envelopment by the Persians' superior num­
bers.21 Miltiades could not have made the decision so to array 
his troops without first observing the deployment of enemy 
forces confronting him. He waited for the Persians to leave 
their encampment some three miles away (see Plate la), cross 
the plain, and form up in battle line; he then made the necessary 
tactical arrangement or rearrangement of his own line in 
response. Evans maintains that Miltiades "stretched his battle 
line to make it equal to the Persian front, the length of which he 
could have estimated only after he saw the Persian army 
advancing in battle order across the plain." So too, Lazenby 
points out that Herodotus "says that on the day of battle, the 
Athenian centre was formed few ranks deep 'in order to make 

19 Pritchett (supra n.16) I 110. 
20 J. F. LAZENBY. The Defence of Greece: 490-479 B.C. (Warminster 1993: 

hereafter 'Lazenby') 64: "If the figures given by most of the secondary [sci!. 
ancient] sources for the Greek army are roughly correct (i.e. 9000 Athenians 
and 1000 Plataians), and we allow about 90 centimetres per man, we would 
get the following frontages: left wing 315 metres, centre c.1000, right wing 
c.300 metres." So Pritchett (supra n.16) IV 143f and Hammond 178, who notes 
that modern estimates have ranged from 1,000 meters (Kromayer) to 2,750 
meters (Myres). 

21 The question of Persian numbers at Marathon is much vexed. The con­
sensus of scholarly opinion is that the force was in the neighborhood of 
20,000-30,000 combatants. See, e.g., Santosuosso (supra n.ll) 29; J. A. G. van 
der Veer, "The Baltle of Marathon: A Topographical Survey," Mnemosyne 
SER. 4 35 (1982) 309; Doenges sf. Hans Delbriick, History of the Art of War P, 
tr. W. J. Renfroe (Berlin 1920; rpt. Lincoln [Nebraska] 1990) 70, contended 
that "nothing forces us to assume that the Persians had numerical superiority 
at all at Marathon and Plataea; it is completely possible, in fact even 
probable-in my opinion, certain-that the Greeks were stronger"; this view 
has found almost no supporters, chiefly because the figure of 6,400 Persian 
dead implies a force considerably greater than 10,000. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

336 MARATHON AND THE SAME-DAY MARCH 

the army equal to the Median army' (6.111.3), and this surely 
implies that it was the Persians who were 'first in the field'."22 

The meal, the two ritual procedures, the movement of the 
Persian forces from their camp and their formation for engage­
ment, Miltiades' assessment of their dispositions and his own 
deployment of 10,000 men in line of battle-all this took, at a 
reasonably conservative guess, at least two hours. Fighting 
could have commenced no earlier than 7:30 A.M. 

II. Duration of the Battle 

It is true that the typical hop lite battle of classical Greece was 
often a short, sharp engagement 23 that yielded a victor-that is, 
possessor of the field-within two or three hours or even less. 24 
This was not, however, invariably the case: the battle of Himera 
in 480 took "from dawn to late evening" (Hdt. 7.167: £~ ,;ou~ '" 
~£XPl ()dAT\~ 6'ViT\~) and the fighting on Sphacteria in 425 leading 
up to the Spartan surrender lasted "most of the day" (Thuc. 
4.35.4: tll~ ,,~£pa~ to 1tA£tcrtOV). G. L. Cawkwell, adducing 
other examples of lengthy engagements, concludes that "battles 
were not short in duration .... Generally ... the historians 
contented themselves with the phrase 'for a long time: which 
some might be disposed to dismiss. But we know what the 
Greeks knew as a long battle .... There is no reason to think of 
battles as quickly over. "25 And, too, Marathon was emphatically 
not a typical case. Greek forces were outnumbered two to one, 
perhaps three to one. Persian troops were not equipped or 
trained to fight quite like hoplites. 26 The Greeks had to adjust 
their tactics and their techniques of hand-to-hand combat to 

22 Hdt. 6.111.3: 'to otpatOltEOOV E~tOOU~vov 'tip MT\OU(ro o'tpatOltEOcp, to IlEV 
au'tou IlEOOV EyivEtO Eltt tu~tae; 6Hyae;, lCat 'taUttl ~V o.09EvEo'ta'tov to 
otpa'tolt£OOV, 'to O£ lCEpae; h:UtEPOV EPPro'tO ltl..f)9El.; Evans (supra n.11) 285; 
Lazenby 62. 

23 Burn 51: "The time of a hand-to-hand melee must be counted in minutes 
rather than hours.» Cf Munro 249: "the 'long time' of conflict described by 
Herodotus ... is to be measured in minutes, not in hours.» 

24 So Veg. 3.9, endorsed (tacitly) by Pritchett (supra n.16) IV 46, 49ff. 
25 • Orthodoxy and Hoplites," CQ NS. 39 (1989) 376. 

26 For example, they relied more heavily on missile weapons than did Greek 
infantry. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 337 

their enemy's style of fighting, the weaknesses in his protective 
armor, and the threat posed by his offensive weaponryY 

Another anomaly is that the combatants ranged over an un­
usually large battle ground in the course of the fighting. Con­
sider: we know that the two battle lines collided after the Athen­
ians had crossed an intervening distance of eight stades (ca one 
mile), and that the Greeks likely sustained most of their casual­
ties in the vicinity of the Soros, or burial tumulus. 28 After the 
Persian forces broke and fled, there was fighting later in and 
near the Great Marsh some three miles (Hammond 211) 
northeast of the Soros, as seen in the famous painting by Micon 
and Panaenus in the Stoa Poecile (built ca 460; Paus. 1.15.4). 
There was also heavy fighting some distance29 from the marsh, 
probably at the Persian encampment and again by the ships 
along the 3,OOO-yard length 30 of the Schoenia beach up to Cape 
Cynosura, about four miles from the Soros, and in the sea itself 
as the Persians struggled to embark. Here the polemarch Cal­
limachus fell. 

All told, Greeks forces ranged over a distance of six miles 
while engaged in heavy fighting, each man burdened by seven­
ty pounds of armor and weapons under the hot sun of a late 
summer's day. The same-day myth has them then retrace their 
steps another three to five miles back across the plain of Mara­
thon, depending on where the troops were on the beach after 
the Persian departure. Hammond imagines the army leaving 
from Schoenia "three miles east of the [burial] mound," but this 

27 On the diversity of Persian armament see W. W. How, • Arms, Tactics 
and Strategies in the Persian Wars," JHS 43 (1923) 123; Santosuosso (supra 
n.11)29. 

28 Hammond 178: • honour, as well as convenience in collecting the corpses, 
required that the memorial was made at the spot where most of the dead had 
fallen. In practice, too, the presence of Persian arrowheads in the fill of the 
mound puts the matter beyond any doubt at all." But see Doenges 13 n.21: -It 
is likely in view of the extensive area over which the battle took place that the 
Athenians suffered casualties not just at the Soros .... The dead from all over 
were gathered and cremated, and the mound was erected in a prominent 
position on the plain of Marathon which had more to do with the road 
system through the plain than the site of the battle." For a clear general 
discussion of topographical issues, see van der Veer (supra n.21) 310-18. 

29 Hammond (233) places the Persian encampment near Kato-Souli more 
than two miles from the shore; the Greek hop lites pursuing the Persians will 
have thus traversed a distance up to two miles in moving from camp to shore 
along the intervening marsh. 

30 Hdt. 6.113-15; cf Hammond 219f. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

338 MARATHON AND THE SAME-DAY MARCH 

is reckoning from the western end of the Schoenia, where the 
marsh extended nearly to the shore. There is no reason to think 
that fighting did not occur at other points further east along the 
whole length of the beach. J1 Merely to march eleven miles in 
armor would require at least four hours. But the army did not, 
of course, cover this ground unopposed. Herodotus reports 
that the Greek hoplites killed 6,400 enemy combatants-not en 
masse in moments by a hail of missiles, like the Spartans at Ther­
mopylae (Hdt. 7.225), but by pursuing and attacking the enemy 
at close quarters. The killing was hard work, requiring the 
hoplites to parry the blows of their opponents and to pierce 
their protective armor, slashing and stabbing with edged metal 
in hard-fought hand-to-hand engagements. The Greeks also cap­
tured seven Persian ships. If we allow the hoplites two hours (a 
cautious estimate) for fighting and killing the enemy at sites on 
and around the plain, near the marsh, and along the shore, we 
arrive at a total of six hours for the duration of the battle. 
Fighting ceased at 1 :30 P.M. at the earliest; all in all, a good day's 
work. As Herodotus says, it had taken "a long time." If we­
again, conservatively-allow another hour for reassembling of 
Greek forces in the vicinity of the Soros from scattered 
positions on the battlefield, we arrive at 2:30 P.M. At most, only 
four hours of daylight remained for a march to Athens. 

Hammond, however, crams the battle into three and a half 
hours and asks us to believe that it was all over by 9 A.M. and that 
by 9:30 the Greek force had regrouped and refreshed itself 
sufficiently to begin the march to the city. In this, he follows 
Munro (249), who makes the battle seem like a sporting 
diversion: "the battle from first to last was a brief affair, a 
morning's work before luncheon." 

III. The Signal 

"In Athens the Alcmaeonidae were accused of suggesting this 
move [round Sunion to Athens]; they had, it was said, an 
understanding with the Persians, and raised a shield as a signal to 

31 Hammond 210, 220, 181 fig. 11; see also E. C. May et aL, Ancient and 
Medieval Warfare (=The West Point Military History Series [Wayne (New 
Jersey) 1984]) 18 map 3b. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 339 

them when they were already on board" (Hdt. 6.115).32 The 
post-battle signal is problematic. As it is hard to imagine that the 
raised shield was discernible from the Persian fleet standing out 
to sea, let alone intelligible as conveying any precise message, 
some historians have conjectured that the Greek traitor was 
contriving a heliograph and that the signaling apparatus was a 
polished metal disk reflecting the light of the sun from some 
high point, perhaps Mt Agrieliki or Pentelikon, adjacent to the 
Marathon plain (Hammond 210). This so strains the evidence 
that other historians have simply discounted this detail, 
especially as Herodotus is at pains to expose the explicitly anti­
Alcmaeonid propaganda in the story. Lazenby sensibly argues 
(pace Munro) that "it seems unlikely that there was such a 
signal, and [likely] that it was one of the rumours used to 
discredit the Alkmaionidai and their associates in the early 480s, 
when several of them were ostracized, probably primarily for 
alleged 'medism'."33 Busolt (593f) traces the story to the imagina­
tion of some overexcited combatant: "Es scheint jedoch auch 
dieses Zeichen nur in der erregten Phantasie einiger Krieger 
existiert zu haben." Hammond, however, finds here support 
for his very-eady-morning time frame: the signal must have 
been given before 10:00 A.M. because afterward the angle of the 
sun would have prevented the signaler from redirecting its rays 
in the right direction. This presupposes a traitor-signaler who 
either (a) knew in advance when the battle would take place and 
how long it would last or (b) had so little competence as a heliog­
rapher as to be unaware that useful transmissions from his 

f,0sition would be impossible after 10:00 A.M. Is it not more like­
y that the putative signaler thought to equip himself with a 

32 ui'ti" OE tcrXE EV 'AO"vuiot(H E~ 'AA.Kj.lErovtOErov j.l"xuvTj<; uu'tou<; 'tuu'tu 
E1ttVO"OTjvUt· 'tou'tou<; yap crUVOEj.lEVOU<; 'to'icrt nEP<JTIcrt avuOE~at acr1tlOa EoUcrt 
ilo" EV 'tDcrt v"ucri; A. de Selincourt, tr., Herodotus: The Histories, rev. A. R. 
Burn (Harmondsworth 1972) 429f. 

33 Lazenby 73; cf R. Sealey, A History of the Greek City-States, ca. 700-338 
B.C. (Berkeley 1976) 202f: -The tradition reaching Herodotus said that this 
signal was given by the Alcmaeonidae; the historian denied this indignantly, 
asserting that the Alcmaeonidae had constantly opposed the Peisistratidae, 
but he insisted that a signal was flashed .... If political animosity ... mis­
represented the shield as a signal to the Persians and blamed the Alcmae­
onidae, this merely indicates the prestigious and enviable position which the 
clan had attained." So, too, Fine 286; T. Buckley, Aspects of Greek History: 
750-323 B.C. A Source-Based Approach (London 19%) 154ff; Doenges 15. 
Those who think Alcmaeonid complicity likely include Burn (252) and others 
cited in Lazenby 73 n.65. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

340 MARATHON AND THE SAME-DAY MARCH 

second polished disk, so that he could redirect the reflected 
rays of the sun from the first to the second and thereby send a 
signal in any direction at any time during daylight hours? If we 
may conjure up one polished disk, why not two? 

IV. The Forced March to Athens 

Herodotus says that the Athenian army, having intercepted 
and correctly deciphered the shield (or heliographic) signal, 
proceeded to Athens "as fast as their feet would carry them. '" 34 
He does not say that they did so the same day. That detail 
comes from Plutarch. To preserve the Plutarch an version, 
Hammond initially proposed the shorter, but more arduous, 
route via Cephisia: that is, a total of twenty-two miles from the 
area of the Marathon burial monument to the precinct of 
Herac1es at Cynosarges, where Herodotus says the Athenian 
army encamped on arrival at Athens. In CAW, however, he 
changed this to the much likelier twenty-five-mile route via 
Pall ene, through lower elevations and along more level 
ground. 35 But to these twenty-five miles, we must add, to reach 
the total distance for the day, the ten miles or so covered during 
the battle itself. Further, though it encamped at Cynosarges,36 
the army must have proceeded to the beach at Phalerum, an 
additional three-mile march, to prevent the disembarkation of 
the Persian troopsY We thus reach a daunting total of thirty­
eight miles for this action-packed day. 

We must bear in mind, too, that the army had to convey 
nearly 600,000 pounds of armor (seventy pounds per man times 

H Hdt. 6.116: 'A9"vulm Oe 00<; ltoOrov dxov taxltJ'tu E!30,,9wv E<; to natu. 
}5 Hammond, CAH2 (supra n.12) 512: "the victorious army marched ... the 

40 km [24.8 miles] to Athens. That evening they were in position in the pre­
cinct of Heracles at Cynosarges, when the leading ships arrived off Phal­
erom"; earlier (507), Hammond distinguishes two routes, ·one of about 35 
kms [21.7 miles] over the hills between Mt. Parnes and Mt. Pentelicus [sci!. via 
Cephisia] ... and the other of about 40 kms over the low pass between Mt. 
Pentelicus and Mt. Hymettus by Pallene and ... along the coast." See also 
Doenges 7. 

}6 Not at the base of the south slope of Lykabettos (Busolt 594), but in the 
deme Diomeia, south of the city: see E. Honigmann, • Kynosarges," RE 12 
(1925) 33; Munro 251; and T. Martin, Ancient Greece: From Prehistoric to 
Hellenistic Times (New Haven 1996) 118 plan 2. 

}7 See J. BALCER, The Persian Conquest of the Greeks: 545-450 B.C. 
(Konstanz 1995: hereafter 'Balcer') 221. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 341 

8,500 men)38 over the entire distance. As the hoplites were 
responsible for procuring their own equipment, which invol­
ved considerable expense, there will have been no cache or 
armory of extra panoplies at Athens. Further, the use of slave 
armor-bearers or wagons drawn by oxen or mules will have 
retarded rather than expedited the army's movement. The men 
also required food and plentiful water to avoid heat prostration 
and utter physical collapse: at a minimum, two pounds of food 
and two quarts (about four pounds) of water for each man.39 

Modern attempts to replicate the conditions of ancient troop 
movements have yielded informative results. During recent 
field-tests in England and the Netherlands, though participants 
held up well under daily march-rates of twenty-five miles, they 
were carrying only shields and spears, slung over their shoul­
ders on a strap, and walking on good roads in mild tempera­
tures. 40 

Hammond, a "famously fast walker" (Lazenby 74), endorses 
the same-day myth, because as a young man (in 1930)41 he hiked 
from Marathon to Athens and back in one day, "taking the 
direct route over M t. Pentelicus" -that is, twenty-two miles via 
Cephisia42-evidently unencumbered by armor or weapons 

38 The initial 10,000 less (a) 192 Athenian dead, (b) another 350-400 (?) too 
seriously wounded to march, (c) 20 dead and 40 (?) wounded Plataeans, and 
(d) the ca 900 of the tribal regiment (the Antiochis) under Aristides left be­
hind (Plut. Arist. 5) to care for the wounded and bury the dead. 

39 See D. W. ENGELS, Alexander the Great and the Logistics of the Mace­
donian Army (Berkeley 1978: hereafter 'Engels') 123-26 . 

• 0 D. Atkinson and L. Morgan, -The Wellington and Nijmegen Marches," 
in Roman Military Equipment: The Accoutrements of War (Oxford 1987) 99-
1 07. E. McGeer, a Byzantinist, describing similar field trials in which research­
ers wore or carried simulated ancient Roman military equipment during six 
days of marches, stresses the deleterious effects of lack of water and salt on the 
participants (Sowing the Dragon's Teeth: Byzantine Warfare in the Tenth 
Century [Washington [DC] 1995J 339): 

Without adequate intake of either, they were initially prone to spells of tem­
porary blindness, disorientation, and dehydration.... Each participant 
required half a pint of water every hour .... The toll of heat and thirst on 
Byzantine soldiers ... during campaigns in Syria or the Balkans will certainly 
have been far heavier by comparison .... The author of the De re miLitari ... 
insited that ... soldiers proceeding through waterless regions in the heat of 
summer would not last to the midday meal but would succumb to thirst 
along with their horses .... 

• 1 While - a student at the British School of Archaeology at Athens": 
Hammond 210 . 

• 2 See supra n.l 0, 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

342 MARATHON AND THE SAME-DAY MARCH 

and not bleeding or suffering from exhaustion or heat stroke. 
The Athenian hoplites, however, were not all young men: they 
ranged in age from twenty to sixty years old. 

Hesiod's description of a good ploughman in his Works and 
Days, a vigorous well-fed fellow of forty years, who could 
drive a straight furrow and was not to be distracted by his 
companions like his younger fellows, evokes the hoplite as 
well as the farm labourer. Soldiers were not always, or 
typically, young men, but men in the prime of life, even 
approaching old age.43 

Xenophon reports that a mercenary commander once 
boasted of the physical superiority of his men to the soldiers of 
city-state armies because "armies made up of citizens must 
include some men who are already past and some who have not 
yet reached their prime. And there are very few people in each 
city who keep constantly in good physical training" (Hell. 
6.1.5 ).44 As Hanson sensibly concludes, "he seems to imply that 
the aged within the ranks of most hoplite armies were a limiting 
factor on the mobility ... of the phalanx. "45 

The length of the battle of Marathon-both in elapsed time 
and distance covered-so depleted the stamina of a middle-aged 
citizen-soldier as to make a trek back to Athens on the same 
day at best foolish, at worst lethal. The opening phase of the 
battle, in the vicinity of the Soros, took place in two stages or 
movements. In the first, the Persian center pushed the Greek 
center back some distance, inflicting most of the casualties the 
Greeks would suffer during the battle, while the Greek wings 
drove the Persian wings into flight. In the second, the Greek 
wings either turned on the Persian center in a pincer move­
ment or, on Hammond's interpretation,46 reversed field, re­
formed, and attacked the Persian center, driving it in flight 
toward the marsh. 

43 S. Mitchell, "Hoplite Warfare in Ancient Greece," in Lloyd (supra n.18) 
96. 

44 o.'AlJJ. ta f,L£V £1, troy ltOM:COY OtP<XtE\)f,L<Xt<X tou~ f,L(v ltpoE'AT)A:\)eOta~ i\OT) 
t<Xl~ i]'At1d<ll~ £Xu, 'tou~ 0' OUltCO o.Kf,L&.SOV't<X~. OCOf,L<XOKOUOt yE f,LiJv f,Lo.Aa <lA-tYOt 
'ttY£~ tv EKuo'tl1 ltOAEt, tr. R. Warner, X enophon: History of My Times 
(Hellenica) (Harmondsworth 1966) 248. 

45 V. H ANSON, The Western Way of War: Infantry Battle in Classical Greece 
(New York 1989: hereafter 'Hanson') 90: "after all, thirty of forty-two age 
classes liable to military service were composed of men over thirty years of 
age." 

46 Hammond 226; cf Santosuosso (supra n.ll) 36 fig. 2.4. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 343 

If the fighting had ceased at this point, we would have to do 
with a short, sharp engagement, though rather longer than typi­
cal hoplite vs hoplite battles, because of the change of direction 
and the regrouping during the struggle. But, as we have seen, 
the fighting did not end there. Delbriick stipulates that "a rather 
long pause occurred between the two phases of the battle," be­
cause, he maintains, the Athenians were preoccupied with 
caring for dead and wounded and looting the fallen Persians. 47 

But the pause more likely had to do with fatigue and the hop­
lites' need to rehydrate themselves and to remove their armor 
for a respite to avoid hyperthermia before following up their 
victory by determined pursuit of the enemy.48 The Greeks then 
chased the Persians across the plain and inflicted heavy losses 
on them in difficult, wetland terrain. In particular, fighting on 
damp or muddy ground must have been both slow going and 
especially taxing. Hanson reports that "modern efforts to dupli­
cate the difficulties faced by men in hoplite armor have shown 
that ground that is sandy or merely loosely packed-not to 
mention wet or muddy-requires a 20-25 percent increase in 
oxygen demands. "49 Further, Sage notes: 

in general, pursuit was carried out only for a short distance in 
the absence of effective light armed and cavalry support in the 
major hoplite battles. Hoplite equipment was not suited to 
swift or lengthy pursuit and the basic concept in Greek 
warfare was domination of the battlefield and not the exter­
mination of the enemy.50 

Nonetheless, the Marathon fighters pursued the Persians down 
to the shore and into the surf, where there was again especially 
fierce fighting. 

The special exertions demanded of the men of Marathon 
were, to be sure, extraordinary. But hoplite battle, under any 
conditions, was always a severe test of the combatant's physical 

47 Delbriick (supra n.21) 82; cf C. Hignett, Xerxes' Invasion of Greece 
(Oxford 1963) 72. 

48 Hanson remarks (60): "There ... seems to have been a special reluctance 
on the part of the Greek infantryman to put on his body armor, strap on the 
shield, and don his helmet until the last possible moment before battle. This 
expresses his sensible aversion toward wearing arms and armor until their life­
saving potential was more significant than their inherent discomfort." 

49 Hanson 81; sec also W. Donlan and J. Thompson, "The Charge at Mara­
thon," CJ 71 (1979) 420. 

50 M. Sage, Warfare in Ancient Greece: A Sourcebook (New York 1996) 95. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

344 MARATHON AND THE SAME-DAY MARCH 

endurance. Merely moving any distance, let alone engaging in 
protracted combat, while encased in the full panoply of helmet, 
breastplate, greaves, and shield, and wielding spear or sword 
demanded a great expenditure of a soldier's reserves of physical 
strength. Hanson reports (56): 

My own students at California State University, Fresno, who 
have created metal and wood replicas of ancient Greek and 
Roman armor and weapons, find it difficult to keep the 
weight [of the armor] under seventy pounds. After about 
thirty minutes of dueling in mock battles under the sun of 
the San Joaquin Valley they are utterly exhausted. 

Before their march back to Athens, the men of Marathon had 
fought a long time in difficult conditions; all were exhausted by 
the physical exertions entailed in killing 6,400 of the enemy in a 
series of engagements on an exceptionally expansive and varied 
battlefield; during the battle, all had suffered the effects of expo­
sure to the summer sun, risking dehydration and heat stroke 
inside their helmets 51 and heavy body armor, which had no 
provision for ventilation (Hanson 79): 

the solid, continuous plate of metal upon the body gave little 
relief from either heat or cold. In the summer, perspiration 
must have soaked the hoplite's inner garment: the shiny 
bronze which could dazzle the enemy across the battlefield 
could just as well act as a solar collector of sorts that would 
make the entire surface hot to the touch .... We often hear, 
then, of hoplites who came near collapse from dehydration, 
or became delirious as a result of heat prostration-surely a 
likely phenomenon for armored men in a country where it is 
so hot in the spring and summer. 

Most men had raw sores from the chafing of armor worn too 
long; 52 some were suffering from minor wounds and perhaps 
loss of blood. And the same-day march back to Athens would 
have spanned the hottest portion of the day.53 

51 Cf Hanson 72f: "Besides the loss of perception, the helmet was uncom­
fortable because of its weight (five or more pounds) on the neck and because 
of the heat it generated around the eyes, mouth, nose, and ears .... " 

52 Cf Hanson 76: "like the awkward leggings of the First World War infan­
trymen, the greave was likely to chafe when running or even simply walking." 

53 The mean monthly temperature maximum in September in Athens is 34° 
c (93°p); the overall mean monthly temperature is 24°c (76°p); the average 
relative humidity is 42%. The climate of the area is classified as "subtropical" 
(i.e., over 9°c for 8-12 months). See W. Holl, ed., World-Climates (Stuttgart 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 345 

No ancient army ever covered thirty-eight miles on foot on 
the day of a major battle. This conclusion is corroborated by 
many investigations of the rate of marching-speed of foot 
soldiers in other historical eras. 

Engels' meticulous study of march rates for the Macedonian 
army has shown that 19.5 miles per day (mpd) was the fastest 
rate achieved by Alexander's army;54 the overall average, exclud­
ing rest days, was 15 mpd. Granted, the Macedonian infantry in 
Asia was initially four times larger than the Greek force at Mara­
thon, but it had a deserved reputation for speed of movement. 

Other historical studies of march rates for non-mechanized 
foot armies reach conclusions similar to Engels'. For example, 
the average rate for Roman armies was no more than eighteen 
mpd. 55 Byzantine armies, too, appear to have traveled at most 

1981) 182; M. Conway and L. L. Liston, edd., The Weather Handbook (rev. 
ed.: Atlanta 1990) 405. 

54 Engels 155f: -it is doubtful whether this rate could ever be surpassed un­
der any conditions. Notably, when the army achieved this rat.e in the Sinai, it 
was provisioned by the fleet, and hence its transport problems were reduced.» 
But see B. Bar-Kochva, The Seleucid Army: Organization and Tactics in the 
Great Campaigns (Cambridge 1976) 95: "the five days' march of the huge 
Ptolemaic army under the scorching sun of the Sinai desert from Pelusium to 
Raphia (Polybius 5.80.1-3) ... works out at 36 km [22 miles] a day." 

ss G. Bodinier and J. Childs, "Transport," in A. Corvisier and J. Childs, 
edd., A Dictionary of Military History and the Art of War, tr. C. Turner 
(Oxford 1994) 819: "On the march, the average daily stage [for Roman 
legionaries] was about 18 miles .... » G. Horsmann, Untersuchungen zur mili­
tarischen Ausbilding im republikanischen und kaiserzeitlichen Rom (Bop­
pard am Rhein 1991) 121, maintains that "Durchschnittsleistungen wie die 27 
km [16.75 miles] pro Tag bei Caesars siebzehntagigem Marsch von Corfinium 
nach Brundisium diirften eine Ausnahme sein .... » A commonsensical mind­
fulness of the limits of human endurance dictates rejection of such prepos­
terous claims as Livy's (27.45-50, 28.9) for the marches of the army of Claudi­
us Nero between Apulia and the Metaurus River in 207 (240 miles in six days) 
or Caesar's (BG 7.40-41) for his march of twenty-four miles with four legions 
and all his cavalry from his camp at Gergovia to confront the rebellious Aedui 
and back again, with only three hours rest, before sunrise. We have to do in 
such accounts with patriotic exaggeration and bald self-promotion, respective­
ly. Veg. 1.9 (militari ergo gradu XX milia passuum quinque horis dumtaxat 
aestivis conftcienda sunt. pleno autem gradu, qui citatior est, totidem horis 
XXIII milia peragenda sunt) may seem to be evidence for faster march-rates, 
but note well G. R. Watson, The Roman Soldier (Ithaca 1969) 54f: "It will be 
noticed that no mention is made by Vegetius of halts on the march. It is hard 
to believe that a route march of five hours during initial training would be 
made without a halt of any sort, and we must assume that such halts were in 
fact made. It is probable that the five hours are exclusive of halts. At any rate 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

346 MARATHON AND THE SAME-DAY MARCH 

about sixteen mpd. 56 An investigation of the march rates of cru­
sader armies concludes that they "should have [had] no difficul­
ty in averaging 18 miles a day or slightly better on level roads. 
Speeds would have decreased in the Byzantine region but not 
too sharply."S7 

Armies have marched farther on foot than twenty-five mpd, 
but only at a very high price in loss of men from physical ex­
haustion. During the War of the Spanish Succession, the Duke 
of Marlborough's infantry penetrated the ne plus ultra lines by 
covering thirty-six miles in a sixteen-hour march to Arleux. 
However, "men marched until they dropped, fainting or dying 
by the road. The track was lined with stupefied soldiers, of 
whom scores expired from their exertions. Little more than half 
stayed the course." 58 

The average daily march for Napoleon's Grand Armee was 
between ten and twenty-two miles. It has been calculated that, 
even with the aid of horse-drawn troop transports, Napoleon's 
troops could make only twenty-four to thirty mpdY During 

the marching speed, if due allowance is made for the greater length of the 
Roman hour in summer and the shorter distance of the Roman mile, is at the 
military pace very similar to that which is normally adopted in the British 
army, where three miles are covered in the hour, with one halt of ten minutes 
included. " 

56 McGeer (supra n.40) 341: "That he considered sixteen miles (24 km [sic]) 
a very long distance ... for an army to travel in one day, and a journey likely 
to wear out men and horses, implies that Phokas would not have expected 
more of his own soldiers in ordinary circumstances." 

57 J. W. Nesbitt, "The Rate of March of Crusading Armies in Europe: A 
Study and Computation," Traditio 19 (1963) 180, with focus on the itineraries 
of Peter the Hermit and Godfrey of Bouillon. So too D. L. Smith, "Muscovite 
Logistics, 1462-1598," Slavonic and East European Review 71 (1993) 39 n.10: 
"The average man cannot march more than 12-18 miles a day and carry more 
than 80 pounds with him." 

58 W. S. Churchill, Marlborough: His Life and Times, Book 2 (London 
1947) 849f; cf G. W. L. Nicholson, Marlborough and the Spanish War of 
Succession (Ottawa 1955) 149: "the last ten miles under the hot morning sun 
imposed a terrific strain. The numbers who dropped out of sheer exhaustion 
ran into the hundreds, and then the thousands." 

59 So Bodinier and Childs (supra n.55) 820. Although John Elting, Swords 
around a Throne: Napoleon's Grande Armee (New York 1988) 463, claims 
(in my opinion, overestimating) that by doubling the etapes, the army could 
reach a peak rate of thirty to thirty-five mpd, he acknowledges that ·such 
marches, especially when made in bad weather on short rations, killed and 
crippled as many young soldiers as a battle ... " [my emphasis]. Thus when 
General Louis Friant marched his division (of Davout's III Corps) the seventy 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 347 

the American Civil War, a rate of about two miles per hour 
(mph) marked the top end for armies on foot, and sustained 
marches at that speed took their toll on the troops.60 

British infantry in World War I were expected to be capable 
of marching twelve to fifteen miles in an eight-hour day.61 
British and American military manuals currently recommend 
march rates of 1.5 mph when traveling cross-country, 2.5 mph 
in good weather on optimum surfaces,62 though a rate of 3 mph 
is possible.63 

miles from Vienna to Austerlitz in thirty-six hours, "hardly one man out of 
twenty was still with their [sic] eagles at the end '" " (Elting 463). In its 
earliest days, even motorized transport did not achieve rates much over thirty 
mpd: J. Huston, The Sinews of War: Army Logistics 1775-1953 (Washington 
1966) 298, notes that "a test run in 1912 of two Army trucks and a privately 
owned company truck from Washington, D.C. to Fort Benjamin Harrison at 
Indianapolis, covering 1,524 miles in forty-eight days [31.75 mpd], did not give 
wholly satisfactory results but did show the trucks had some promise." Cf 
266: "After making about fourteen miles the first day, [George] Custer's men 
rode hard the next two days, covering over thirty miles a day .... " 

60 S. Sears, Landscape Turned Red: The Battle of Antietam (New Haven 
1983) 285f: "[General A. P. Hill's] Light Division's remarkable march from 
Harper's Ferry-seventeen miles in less than eight hours-rivaled the best 
marks of Jackson's famous foot cavalry .... To be sure, uncounted hundreds 
were left by the roadside exhausted .... " Cf S. Foote, The Civil War: A Narra­
tive, I: Fort Sumter to Perryville (New York 1958) 464, on Stonewall Jackson's 
Shenandoah Valley campaign: "Since March 22, the eve of Kernstown, his 
troops had covered 646 miles of road in forty-eight marching days," that is, an 
average of 13.5 mpd. 

61 D. Winter, Death's Men (London 1978) 70-74. Cf Huston (supra n.59) 
388: "Marching two corps abreast, the divisions moving on parallel roads 
when possible, the Third Army advanced across Lorraine and across Luxem­
bourg to reach the German frontier .... While on this move, men marched an 
average of about twelve miles a day." 

62 Dept. of the U. S. Army, FM 21-18: Foot Marches (Washington 1990 
<http://www.adtdl.army.millcgi-bin/atdl.dll/frnl21-18/fm2118.htm> ) 4.20: "a 
normal foot march day is 8 hours, for a distance of 32 km [19.87 miles] at a 
rate of 4 kph [2048 mph]. The maximum distances recommended for forced 
marches are: 56 km [34.78 miles] in 24 hours; 96 km [59.62 miles] in 48 hours; 
or 128 km [79.49 miles] in 72 hours. A forced march usually exceeds this 
distance by increasing the hours marched rather than by increasing the rate of 
march"; the manual then advises a period of 21.84 hours, including time for 
rest and meals, to cover 52 km [32.29 miles]. See also Nesbit (supra n.57) 175 
n.29; McGeer (supra nAO) 341 n.12. 

63 R. S. Rush, per cp. elec. (20 December 1997), describing the training of a 
cohort battalion of newly recruited soldiers: "Age ranged from 18 to 44 with 
the average about 24 .... We marched 50 minutes and took a 10-minute break. 
Average speed was 3 miles per hour. ... (Eight hours [for 20 miles] was the 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

348 MARATHON AND THE SAME-DAY MARCH 

At 2.5 mph, the Marathon hoplites required ten hours for the 
route via Pallene to Phalerum: at 1.5 mph over the more dif­
ficult terrain via Cephisia, they needed over fourteen hours. 
Even if we improve these rates by an additional 1.0 mph owing 
to the exigent tactical situation and patriotic fervor, they still 
req uired over seven hours via Pallene, almost nine hours via 
Cephisia. They had no more than four hours. Moreover, as we 
have seen, the Greeks at Marathon were in no shape for a faster­
than-average march. 

No Athenian strategos was stupid enough or cruel enough to 
put badly battle-fatigued men through the ordeal of a same-day 
march back to Athens under such conditions. Had he done so, 
he would have fielded a force so reduced in numbers and 
broken down with fatigue that it could not have prevented the 
disembarkation of the Persian army, which may still have en­
joyed numerical superiority. The same-day version of the battle 
and the march, if it existed in some source earlier than Plutarch, 
likely resulted from the tendency of battles perceived to be 
turning points in history to inspire romanticizing and mytho­
logizing retellings and of victorious commanders, like Miltiades, 
to undergo concomitant lionizing.64 

V. The Persian Voyage to Athens 

The same-day Persian voyage is as implausible 65 as the Athen­
ian march overland. Again, Hammond imagines a best-case 
scenario: the Persian forces, despite their truly catastrophic 
defeat, had sufficient presence of mind and determination to 

standard that could not be missed; the time was usually around seven hours.) 
Other standards: 12 miles in 3 hours (that was cooking, and soldiers needed 
several hours to recover from the speed march). The longest I can remember 
was about 35 miles; it occurred in the Donau region in Germany (very hilly), 
began about 1200 [hours] and ended about 0800 the next day. We also 
accomplished one 100-mile march each year (5x20-mile marches in 5 days)." 

,. For this phenomenon in the case of Marathon see the astute analysis by 
K.-J. H6lkeskamp, CMarathon-von Monument zum Mythos," in D. Papen­
fuss and V. M. Strocka, edd., Gab es das griechische Wunder? Griechenland 
zwischen dem Ende des 6. und der Mitte des 5. }ahrhunderts v. Chr., forth­
coming; cf Fine 287: KBefore long the traditions of the battle became en­
shrouded in legend"; Balcer 221: cFor the Athenians, the Battle of Marathon 
... rapidly unfolded as the initial step in the Great Event in Western 
Historiography, embellished, gilded, and glorified beyond measure." 

'5 It is flatly called ·unglaubwiirdig" by F. Schachermeyr, KMarathon und 
die persischer Politik," HZ 172 (1951) 29. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 349 

make a dash for Athens round Cape Sunion, a distance of about 
seventy miles. 66 This disregards the horrendous dimensions of 
the military disaster suffered by the Persians and the attendant 
dire effect on morale, for they had lost between a third and a 
fourth of their combatants-that is, more than thirty times the 
Greek casualties. Typically, defeated armies suffered only about 
three times the casualties of their victorious opponents. Krentz 
has shown that the defeated side in hoplite battles averaged 
losses of 14%, as against 5% for the victors.67 Low morale alone 
will have unfitted the Persian forces for any further en­
gagement in their immediate future. 68 

Hammond also downplays the detour to Aigilia (Hdt. 6.115) 
to pick up Eretrian prisoners, whose embarkation will have 
occupied some time. He reckons a sailing rate of five knots, in 
part because of a favorable wind from the northeast-favorable 
for the journey from Marathon to Sunion, unfavorable for the 
run up the west coast of Attica from Sunion to Phalerum. The 
truth is that we do not know that Persian vessels could have 
made such speed. Casson has compiled statistics showing a two­
to three-knot rate of speed as the norm for ancient sailing 
vessels. 69 

In fact, the Persian fleet could not have made the journey 
around Sun ion to Athens on the same day as the battle. 7o If they 
set out before first light on the following day, they might just 
have made it to Phalcrum by dusk. If it is true that the Athenian 
army and the Persian fleet arrived at Athens the same day, it 
was the day after or, more likely, the second day after the battle 
(Balcer 221). 

66 So Sealey (supra n.33) 191; cf Hammond 221: "fifty-eight nautical miles" 
[ca 67 terrestrial miles]. 

67 P. Krentz, "Casualties in Hoplite Battles," GRBS 26 (1985) 13-20. Persian 
losses exceeded even those sustained in the more lethal atmosphere of modern 
warfare: cf J. McPherson, Battle Cry of Freedom: The Civil War Era (Oxford 
1988) 472: "The casualty rate in Lee's army was 20 percent in its major battles 
and campaigns .... The figure for Ulysses S. Grant's troops was 16 percent .. " 

68 Cf Doenges 16: "It is ... not clear why Datis after embarking the rem­
nants of his defeated army decided to sail to Athens .... At that point he 
lacked the military force to seize and hold the city. His purpose almost 
certainly was reconnaissance only." Morale as well as numbers was a critical 
deficiency. 

69 L. Casson, ·Speed under Sail of Ancient Ships," TAPA 82 (1951) 136-49. 
70 See, in general, the meticulous calculations in A. T. Hodge, "Marathon: 

The Persians' Voyage," TAPA 105 (1975) 155-73. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

350 MARA THON AND THE SAME-DAY MARCH 

VI. The Date of the Battle 

The dating of the battle to 11 September is based on two items 
of information and a plausible guess. The first item is that the 
Athenians took the decision to defend their city against the 
Persian invasion force at Marathon on the festival day of 
Artemis Agrotera. 71 This we know to have been the sixth day of 
Boedromion (roughly, the last day of August). The second is 
that the Spartan relief force requested by the Athenians did not 
leave immediately owing to a religious observation. It arrived a 
day late for the battle (PI. Leg. 698 E) after a march of three days 
(Isoc. Paneg. 87), having set out from Sparta only after the 
appearance of the full moon. The plausible guess is that the 
Spartan religious festival was the Karneia, which occurred in 
September. As the full moon rose shortly after sunset on the 
ninth of September,72 the Spartans therefore left on the tenth, 
marched a second day on the eleventh, and arrived on the 
twelfth, a day late for the battle. Thus, we arrive at an absolute 
date for the battle: 11 September, that is, the day before the 
Spartan arrival. 

This reconstruction must be rejected on logistical grounds: 
the Spartan force of 2,000 hoplites and their entourage (likely 
one or more helot armor-bearers and other attendants per 
hopliteF3 cannot have marched from Sparta to Marathon (via 
Athens according to Hdt. 6.120: ~KOV E~ 'ta~ 'A8"va~) in three 
days. Curiously, even historians who mistrust the same-day 
Marathon-to-Athens march blithely accept the utterly improb­
able Spartan three-day march,74 though How and Wells allow 
that "the feat is wonderful"; Herodotus says only that the 
Spartans were "in Attica" on the third day out. 75 The distance 
involved is about 176 miles (about 150 from Sparta to Athens, 

71 See Hammond 217 for a discussion of the sources (especially Ar. Eq. 660, 
Xen. An. 3.2.12). 

72 See Hammond 216 n.3 for comparative tabulations of moonrise/moonset 
and sunrise/sunset; on 9 September 490, the sun set at 6:26 and the moon rose 
at 6:32. 

73 Though known for their spare life-style at home, the Spartans enjoyed 
the luxury of personal servants while on campaign. At Plataea in 479, for 
example, the Spartan force of 5,000 was attended by 35,000 helots (Hdt. 9.29); 
the intent here, however, may have been to forestall revolt at home. 

74 E.g., most recently, Doenges 16. 
75 How and Wells 115; Hdt. 6.120: 'tpnulol h l:1t<l.p'tT]<; EyivoV'to EV 'tU 

'A't'tlKfi· 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 351 

an additional twenty-six to Marathon from Athens).76 Even if 
the force had come on horseback, 77 it could not have made the 
journey in less than four or five days. A reasonable guess for an 
army moving on foot is eight days at least; twenty-two mpd 
under heavy loads is no mean feat. 78 Fifty-eight mpd is more 
than "a heroic piece of marching, "79 it is an utter impossibility. 80 

The battle of Marathon thus took place at the earliest on 17 
September, that is, no less than eight days after the full moon. 

Such a revised date both avoids the glaring adynaton of the 
three-day march and better fits another piece of evidence. 
Sometime after Marathon, Athens began to issue tetradrachms 81 

76 Had the Spartans proceeded directly to Marathon, and not first to 
Athens, they would have saved only five or six miles. 

77 This would have required 4,000 mounts, allowing one helot armor-bearer 
to each hoplite. If the typical rider weighed at least 150 pounds and helot and 
hoplite each carried half the seventy-pound panoply, we reach a total weight 
of 185 pounds per horse, leaving out of account any provisions or other sup­
plies the army may have needed en route (including ten pounds of grain per 
horse per day). Two hundred pounds represents the maximum carrying 
capacity of a horse. Such heavily burdened animals could not have managed 
the forty-five or fifty mpd that Engels (153-56) has calculated as the fastest 
rate achieved by Alexander's Macedonian cavalry. Cf M. Speidel, Riding for 
Caesar: The Roman Emperor's Horse Guards (Cambridge [Mass.] 1994) 117: 
·Cavalry could travel up to 65 km (40 miles) a day without overtaxing the 
horses. Longer marches ... wore out the hooves of the often unshod horses." 

78 Exceeding the fastest daily march rate achieved by Alexander's infantry 
moving on foot; see supra n.54. 

79 Burn 253; cf How and Wells 115. Munro (251) thinks "the rapidity of 
[the Spartan] march is remarkable," and suggests, without any ancient testi­
mony, that the three days may refer only to the time taken to march from 
frontier to frontier, that is, 108 miles; he claims "thirty-six miles a day is hard 
marching, but not unparalleled," but cites no parallels. In another context, R. 
D. Milns, "Alexander's Pursuit of Darius through Iran," Historia 15 (1966) 
256, says that a "march rate ... of 36 miles per day is a physical impossibility." 

80 Milns (supra n.79: 256) calls fifty-two mpd "an absurdity." The examples 
adduced against Milns by C. Neumann, • A Note on Alexander's March­
Rates," Historia 20 (1971) 196ff, are convincingly disposed of by Engels 154ff. 

81 Illustrations in C. M. Kraay, Greek Coins (New York 1966) pI. 119. The 
dating of the introduction of coins of this type remains uncertain. C. T. Seh­
man, Greek Coins: A History of Metallic Currency and Coinage down to the 
Fall of the Hellenistic Kingdoms2 (London 1955) 91, says "in the autumn of 
490 B.C.," followed by Hammond 215f: "commemorative coins struck from 
490 B.C. onwards showed a waning moon behind the owl of Athena." G. K. 
Jenkins, Archaic and Classical Greek Coins (London 1976) 81, prefers a date 
"around 480 B.C." C. M. Kraay, Archaic and Classical Greek Coins (London 
1976) 62, cautions that "it is perhaps unlikely that the designing and the 
minting of a new coinage was given as high a priority by the returning 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

J52 MARATHON AND THE SAME-DAY MARCH 

that (apparently) commemorate the Athenians' victory 82 by the 
inclusion of a waning crescent moon beside Athena's owl on 
the reverse (see Plate 1 b). 83 Hammond thinks this is an indirect 
allusion to the date of the battle and offers a complicated 
scenario involving Persian miscalculations in pre-dawn cavalry 
movements owing to the time of moonset the night before the 
battle.84 All others who believe the coin is commemorative see 
the lunar symbol as a more direct allusion to date. Seltman, for 
example, argues as follows that the battle commemorated is in 
fact Marathon, not, as others have maintained, Salamis: 

The tiny waning moon, now introduced for the first time 
upon the coins, is ... a reference to the date [of Marathon], 
which impressed the average Athenian mind far more even 
than Salamis .... While Marathon was probably fought with a 
moon already into its third quarter, Salamis was certainly 
fought with a full moon and could, therefore, not have been 
the occasion which caused the adoption of a waning moon as 
part of the Athenian coinage.85 

The coin designer has clearly shown a waning crescent, not a 
waning gibbous moon. If it is in fact commemorative, the coin 
type indicates a battle following the full moon by at least a week, 
not the day or two of the traditional dating scheme, which is 
based on a patently impossible rate of march for the Spartan 
army of reinforcement. 

Athenians [after 480] as it is given by many numismatists. Shelter, food and 
defence were more urgent, so that the appearance of the new coins may have 
been delayed until near 475." 

82 So Sehman (supra n.81) 91; cf Burn 255f. 
83 Some numismatists doubt that the lunar symbol has any topical 

reference. See Kraay (supra n.81 [1976]) 61£: «The moon has been declared to 
be a waning moon and to refer to the phase under which Marathon or 
Salamis were [sic] fought .... More probably the moon was added to the larger 
£lan of the tetradrachms alone, simply as a reference to the nocturnal cha­
racter of the owl"; also E. J. P. Raven, «Problems of the Earliest Owls of 
Athens," in C. M. Kraay and G. K. Jenkins, edd., Essays in Greek Coinage 
Presented to Stanley Robinson (Oxford 1968) 51£; Jenkins (supra n.81) 81£. 

84 Hammond 216f, 227; cf. CAl-P (supra n.12) 511. 
85 Seltman (supra n.81) 92; cf. his Athens: Its History and Coinage before 

the Persian Invasion (Cambridge 1924) 10Jf. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

JAMES P. HOLOKA 353 

A really precise knowledge of the events of the battle of Mara­
thon and its immediate aftermath remains beyond our grasp. 
We can, however, demythologize accounts of the battle and 
reduce the proportions of the necessary lie in our inferences 
from the ancient testimonia by a realistic assessment of logistical 
exigencies and the limits of human physical endurance. This will 
result in not only the correction of a chronological distortion 
but also a more accurate historical grasp of the achievement of 
the men of Marathon. To imagine that the Greek forces could 
have fought and marched on the same day is both to misrep­
resent the nature of hoplite warfare and to belittle the victory 
won through the astounding physical effort that 10,000 infantry 
exerted against the superior numbers of an unfamiliar op­
ponent. 86 

EASTERN MICHIGAN UNIVERSITY 

NO'Vember, 1999 

86 I am most grateful to Eastern Michigan University for a Research Award 
in support of my work on this essay, preliminary versions of which were read 
at annual meetings of the Michigan Academy of Science, Arts, and Letters at 
Calvin College, Grand Rapids, on 21 March 1997, and the American 
Philological Association in Chicago on 29 December 1997; and at the 
Geburtstag Symposion for Joachim Latacz in Basel on 18 May 1999, and the 
15 October 1999 meeting of the Military Studies Group in the University of 
Michigan History Department. The following friends and colleagues patiently 
fielded many inquiries and offered constructive criticisms: George Cassar, 
Robert Citino, Peter Green, Karl-Joachim Holkeskamp, Donald Lateiner, 
Jonathan Marwill, Reinaldo Perez, and David Potter. On military logistics, I 
have had the benefit of comment by Lieutenant Colonel Brian Hoxie and 
Command Sergeant Major (ret.) Robert S. Rush, U. S. Army. The anonymous 
reader for GRBS made salutary recommendations at the penultimate stage. 
My son Christian was instrumental in producing both map and illustration. 
The dedication of these pages to Richard D. Goff and Joachim Latacz is a 
token return on lavish investments of advice and encouragement.