Jump to ContentJump to Main Navigation
Feeding the DemocracyThe Athenian Grain Supply in the Fifth and Fourth Centuries BC$

Alfonso Moreno

Print publication date: 2007

Print ISBN-13: 9780199228409

Published to Oxford Scholarship Online: January 2008

DOI: 10.1093/acprof:oso/9780199228409.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 17 September 2019

The Fruits of Empire

The Fruits of Empire

(p.77) 3 The Fruits of Empire
Feeding the Democracy

Alfonso Moreno (Contributor Webpage)

Oxford University Press

Abstract and Keywords

This chapter presents the evidence of Athens' economic exploitation of its Aegean empire in the fifth century bc, particularly the colonization of overseas territories and the division of land into an allotment system (cleruchies) in some cases clearly designed to produce grain surpluses for shipment to Athens. Special attention is given to Euboea and its role as Athens' chief fifth‐century granary, protected by a system of fortifications largely disregarded by previous scholarship. The Peloponnesian War and the political crisis of 411 bc which led to the loss of most of the island are examined in detail. The Athenian method of obtaining grain in the form of taxes is studied in the light of a new interpretation of the recently published Grain‐Tax Law of Lemnos, Imbros, and Scyros of the year 374/3 bc.

Keywords:   colonization, cleruchies, Euboea, Lemnos, Imbros, Scyros, tax, Peloponnesian War, fortifications

I. Athens and Euboea

Around the year 380, Isocrates wrote his Panegyricus in praise of Athens. His aim was to cast the city, whose tyranny a new generation of Greeks had not experienced first hand, as the deserving leader of a future Panhellenic conquest of Persia. This radical makeover demanded an eloquent denial of earlier Athenian imperialism, including the most deeply resented of all its manifestations—the official confiscation and distribution of foreign land to Athenian cleruchs:

For these reasons it befits all thinking people to be deeply grateful to us, much more than to reproach us for our cleruchies, which we sent into the depopulated states for the protection of their territories and not through greed. And here is the proof: we had in proportion to the number of our citizens a very small territory, but a very great empire; we possessed not only twice as many warships as all other states combined, but these were strong enough to engage double their number. Just at the very doorstep of Attica lay Euboea, which was not only naturally adapted to the rule of the sea, but also surpassed all the islands in every other advantage; we had greater control over it than over our own country, and besides we knew that both among the Greeks and among the barbarians those have the highest reputation who, having made their neighbors refugees, obtain for themselves a life of affluence and ease. Nevertheless, none of these factors tempted us to do the island wrong; instead, we alone of those who have obtained great power allowed ourselves to live more poorly than those accused of having been our slaves.1

These eloquent words no doubt rang extremely hollow. Euboea, the “large and wealthy island” (νη̑σος μεγάλη καὶ εὐδαίμων) that (p.78)

                      The Fruits of Empire

Map 2. Euboea.

(p.79) (p.80) allegedly tempted Persia to conquer the Aegean in 500, had long been a prime target of Athenian ambition.2 Did Isocrates expect his audience to forget the “horse‐feeding” Hippobotai, the Chalcidian aristocracy, whose fertile lands had become the first conquest of the post‐Peisistratid democracy, expropriated and distributed in kleroi to 4,000 Athenians in 506?3 The shackles of the enemy prisoners and a monumental bronze quadriga, Athena's share of the spoils, still recalled the defeat of Chalcis to visitors of the Acropolis at the time of Pausanias.4 The quadriga that had been destroyed or removed by the Persians had been rededicated, probably in 446, and Athens thus rubbed salt on a wound made even deeper.5 For in that year Pericles had quickly and ruthlessly crushed the revolt not just of Chalcis, but of the entire island of Euboea, exiling the Histiaeans, expropriating their land for Athenian cleruchs, and pacifying the rest “by agreement” (we shall see below what this meant).6 Plutarch amplifies this report by adding the forces that Pericles took (fifty ships and 5,000 hoplites), his justification for making refugees of the Histiaeans (they allegedly killed the crew of one Athenian ship), and the fact that, for the second time, he expelled the Hippobotai.7 Not until 411 did Euboea make a successful break from Athens, a loss that according to Thucydides terrified the Athenians even more than the Sicilian disaster, for “on Euboea they were more dependent than on Attica”.8

Given this historical background, not to mention Athens' all too emphatic protest in 378/7 that her leadership of a Second League would not revert to any form of occupation of allied land,9 it is doubtful that Isocrates found anyone to believe his words.10 Is it perhaps more (p.81) understandable that the role of Euboea in the fifth‐century Athenian grain supply is currently undervalued? Direct literary evidence for grain exports from the island is non‐existent outside a few scattered references in Thucydides and Aristophanes, and these are usually taken to imply that Athens relied on grain from Euboea only during the relatively narrow period of the Peloponnesian War.11

The chief aim of this chapter is to refute this modern view by means of a variety of evidence, most of it long known but never systematically applied to our question. We will see that Euboea played an important role as a producer of Athenian grain from 506, and was Athens' main granary from 446 to 411. Comparative evidence from the cleruchies of Lemnos, Imbros, and Scyros in the fourth century will also be adduced. At the outset, however, we must reacquire a sense of the significance and potential of Euboean agriculture. This was well understood by earlier scholars and travellers, whose works now tend to encounter tenacious academic recalcitrance.12

II. The Big Picture: Euboean Agriculture, Geography, and History

Geyer long ago compiled a thorough collection of the Classical references (mostly textual but some numismatic) to the proverbial wealth of Euboea in grain, oxen, swine, sheep, goats, hens, hounds, horses, wine, olives, figs, pears, apples, forests, nuts, fish, purpura (p.82) shells, ores, marble, and salt.13 Philippson systematically studied Euboea's topography and geography, and assembled all earlier work from antiquity to the twentieth century.14 Bakhuizen has most recently compiled the existing evidence on the sizes and limits of the territories of Classical Histiaea,15 Chalcis,16 and Eretria (see Map 2).17 It is not my intention to retrace their steps much further than my footnotes indicate, or even to focus on the famously rich Lelantine plain between Chalcis and Eretria, that rich volcanic vineyard which Geyer vividly describes as a vast garden.18

It is more impressive to focus on the less famous territory of Classical Histiaea, which stretched over the northern third of the (p.83) island.19 The beauty and fertility of this country have never failed to elicit the lavish praise of European travellers, who variously compare it to the Swiss landscape or to the Arcadia of Poussin.20 Strabo himself believed that its two well‐watered rivers (an abnormality in insular Greece) were magic: sheep drinking from the Cereus turned white, and from the Neleus black.21 Such views may be impressionistic or romantic, but they strike at a fundamental truth: Euboea is not a typical Aegean island. It is in fact hardly an island at all, as was well recognized in antiquity, being separated from the mainland for its entire length by an average of only 12 km of water, which at Chalcis narrow to a mere 72.5 m.22 Geographically, geologically, and climatically Euboea belongs much less to the world of the Aegean than to that of continental central Greece, from Boeotia to Thessaly.23 The peculiar fact that a herd of red Devon cows could be imported to this island in the 1960s, “to improve the dairy farming in the area”, almost invites belief in Strabo's magic rivers.24 The territory of Histiaea, like those of Eretria and Chalcis, turns out to have little in common with arid Attica.25

In AD 1833, while he still could, the fleeing Ottoman grandee Hadji Ismail Bey sold his feudal estate located in the territory of ancient Histiaea, at Achmetaga (mod. Prokopi), to Edward Noel and Fredrick Fellenberg. These were men educated in the (then‐famous) (p.84) agricultural and philosophical academy at Hofwyl, Berne, and intent on starting a similar institution in the newly independent Greece.26 Although they had little money, they engaged immediately to pay the asking price: 10,000 pounds sterling, all in cash, which they considered a bargain.27 From the price alone it is clear that the buyers knew good alluvial lands when they saw them.28 In a letter of that year, Noel writes to his brother in England:

You cannot have an idea of the richness of the soil and advantages to be derived in Greece. The content of the possession I have cannot be short of 15,000 acres, more than half of which is covered with fine forests of pine and other timber, which, if felled and sent to Syra for ship‐building, would be more than adequate to pay for the purchase money.29

Today the Noels still own their estate, perhaps the last surviving bastion of Frankish feudalism left in Greece.30 The agricultural use of Achmetaga under these English and Swiss owners is surprisingly suggestive of conditions in antiquity. Timber felling was for a time the principal industry of their estate. Besides the ships (Noel writes), “many houses in Athens will yet, I trust, be built of Achmetaga produce”.31 (This recalls the wealthy Athenian Meidias of Anagyrous exporting from Euboea (along with cattle) fences, door‐posts, and pit‐props for his silver mines in 348.32) The exporting of Achmetaga grain took more (p.85) effort to arrange: Fellenberg was dead and Noel ill from malaria only months after their arrival. The Cereus river valley, with its multiple springs and low water table, is marshy mosquito‐country that needs to be dammed, drained, and irrigated before it can be farmed.33 But eventually, Noel and his new Swiss associate, Karl von Müller, began to export grain along with timber. Significantly, they did so from the nearby site of ancient Kerinthos, once a deme of Histiaea.

…at the mouth of the river Boudoros caiques could load the timber and grain off the cliffs. Above this small natural harbor on the ancient site of Kerinthos, Noel and Müller had built a warehouse to store their merchandise awaiting shipment. The ruins of it are still to be seen among the scattered remains of the ancient town which supplied the stone for its construction. Today the harbor is silted up and the anchorage is gone.34

We have slight, but fortunately some, record of labour and grain yields at Achmetaga at this time. It comes from the disappointed comments of the American Henry Baird (the ultimate source evidently being Noel himself):

Though [Euboea's] dimensions are almost precisely those of Long Island, the population, according to Mr. N., is but seventy‐five thousand. So sparse a population is insufficient to cultivate the island to any considerable extent with the agricultural implements now in use. The fields are said not to yield much more than a third as much grain as those of an equal extent in England: and this, although Euboea was once the granary of Athens! All the land is divided into two categories, one half being sown with wheat and the other lying fallow according to the popular notion, that is, cultivated with Indian corn (maize).35

By the standards of the grain yields obtained in England (during what has recently been called its agricultural “golden age”), it is not surprising that Achmetaga paled.36 (The average yield of wheat in England in the 1850s, about 27 bushels per acre, or 1,800 kg per hectare, was more than twice as great as in Ireland.37) Thanks to Noel's general figure of “a third”, we can see that the land of Achmetaga in his day (p.86) must have yielded on average about 600 kg of wheat per hectare. This would surely have been an above‐average yield for the whole of Euboea in the 1850s, obtained by capable Western agronomists.38 Nevertheless, it is a figure fully compatible with Greek and Euboean yield‐trends into the twentieth century, and useful to us because it was obtained intensively but pre‐industrially.39

How much grain could the entire island produce? Euboea has an area of 3,653 km2, of which approximately one third is cultivable and was used in the years 1970 and 1979 as follows (all figures in km2):40






















In 1979, Euboea produced a total of 24,173 tons of wheat, 11,977 tons of barley, and 3,226 tons of oats. Assuming for the sake of argument that the land was used in the same proportions in antiquity, we can apply the yields attested for Achmetaga in the (p.87) 1850s to estimate an ancient yearly production of about 16,000 tons of grain (equivalent to about 490,000 Attic wheat medimnoi). But with higher proportions of arable land under grains (say 500 km2, still allowing for fallow), one can reasonably expect up to about 30,000 tons (equivalent to 910,000 wheat medimnoi).41

If anything from “over two‐thirds” to the whole of Euboea were under Athenian control in 421, as Andocides and Aeschines claim, would we be right in estimating that Athens controlled at least 610,000 wheat medimnoi per year from Euboea in the period 446–11?42 I have shown in Chapter 1 that rigid statistics do not take us very far: they are helpful only in outlining what Braudel calls “the limits of the possible”.

The experience of another maritime empire in Euboea may be even more instructive than our English and Swiss example. Venice's control of the grain supply of the Aegean and Black Seas, and their passage through the ports of “Negroponte”, began in the aftermath of the Fourth Crusade in 1204. Euboea itself needed to import grain (principally from Crete), at least until its official annexation by Venice in 1366.43 (This fact reminds us of a similar phenomenon in the Black Sea, which by the time of Polybius had gone from famous exporter to occasional importer of grain.44) From 1366 until its loss to the Turks in 1470, Euboea was the “apple of the eye” (la pupilla) of the Serene Republic and one of its principal sources of grain, making the island “the key and foundation of our affairs” (clavis et fundamentum rerum nostrarum) (1454) and the “most important part of our (p.88) disposition” (maxima pars status nostri) (1458).45 A large number of towers (fifty‐five on the latest count), usually sited on the island's most fertile plains, is the most enduring physical manifestation of Venice's agricultural exploitation, and has been the object of recent archaeological study.46 It has been shown that these towers belonged to great Venetian owners of huge agricultural tracts on Euboea.47

In contrast to our meagre Classical documentation, the surviving Venetian evidence for the island's grain production must be plentiful, but has never, as far as I know, been systematically studied. Nevertheless, the general picture is clearly one of Venetian dependence on grain from the island until the end of the sixteenth century.48 Braudel quotes a letter from Piero de'Medici to Cosimo I dated 14 October 1559:

…these Signiori [the Venetians] are on the point, with their manoeuvres, of receiving Negroponte as a fief from the Turk: they are offering to pay a huge tribute, so great that it is doubtful whether the island would yield so much revenue. And all this is in order to have enough grain for their needs, without passing through France or Spain.49

Documents like these are tantalizing, especially since the population of Venice (120,000 in 1509, 200,000 by 1600) was comparable to that of Classical Athens, and grain consumption rates were similar (about 2 quintals, or 200 kg, of wheat per man per year).50 Many of Venice's Euboean towers also happen to be sited directly over, or in the near vicinity of, surveyed Classical remains.51 This remains a fruitful opportunity for future collaboration between historians and archaeologists of the Classical and Early Modern Periods.

(p.89) III. Patterns of Athenian Land‐Holding and the Athenian Cleruchies On Euboea

Our study of Athenian control of Euboea must begin with the basic question: what sort of Athenians owned land there, and what did they do with it?

We have our clearest evidence of Athenian land‐ownership on Euboea in the fragmentary records of the property confiscated and auctioned after the aristocratic scandals of the Hermokopidai and the Mysteries in 415.52 Oeonias of Atene (“one of the richest men known in Athens at any period”53) is recorded as owning unharvested crops (including figs, grapes, and olives) in the Lelantine plain,54 as well as landed property in Lelanton, Diros, and Geraestus, the collective sale of which is recorded at a massive 81 talents, 2,000 dr.55 There are other fragmentary references to wealthy Athenian overseas properties (ὑπερορία) on Euboea: at Diros belonging to Nicides of Melite,56 and at Eretria.57 Because we cannot trace legal ownership further back than Oeonias and his fellow aristocrats, it is impossible to say whether this was in origin privately acquired or cleruchic land. It has, however, been convincingly shown that, within certain broad parameters normal to the Greek law of landed property, cleruchic land was alienable.58 Men like Oeonias could, in other (p.90) words, amass their collections of Euboean properties by purchasing kleroi from other Athenians.

Another inscription from the 420s, a record of the lease of sacred property on Euboea, features one Panaitios as the lessee of property “in Orobiae, in [the territory of] Histiaea”.59 He appears to pay rent of 20 dr. per year for an olive grove and land under cereals and vines (γη̑ ψιλή).60 The inscription features three other lots in Chalcis (ll. 3, 11, 22), one in Eretria (l. 14), and one in Aigale “in the [the territory of] Eretria” (l. 9), but the lessees' names have not survived. Following Raubitschek's sound suggestion, we can identify Panaitios as the man ([Π]αναιτίο) whose property appears on the confiscation stelai as ἐν τȏι[ἀ]γρ[ȏι]| τȏι ἐν ᾽Ισθ[μȏι.]and ἐν ᾽Αρ—.61 Erxleben identifies the first of these locations as the Isthmus of the Kenaion peninsula (west of Histiaea), and guesses that ᾽Αρ—may be Argura (on the Psachna plain northwest of Chalcis).62 If at least the first hypothesis is correct, Panaitios joins Oeonias in demonstrating that individual Athenians could normally own (or rent) a collection of scattered agricultural properties on Euboea. Significantly, this mirrors the pattern of aristocratic land‐holding that we find in Attica itself.63

Literary sources supplement our knowledge of Athenian land‐holdings on Euboea. Lysimachus of Alopeke, the son of the famous Aristides, is said by Demosthenes to have received, as a gift from the Athenian demos, 200 plethra of land in Euboea (half under cereals and vines, half with fruit and olive trees (πεϕυτευμένη)), 100 mnai in cash, and 4 dr. a day for life.64 Hierocles the oracle collector (p.91) (χρησμολόγος), whose oracles were punctiliously obeyed by the demos after the Euboean revolt of 446,65 also seems to have received land on the territory of Histiaea (probably at Elymnion), according to Aristophanes in 421.66

Significantly, there is also at least one aristocratic property “in Oropos” and another at “the sacred harbor at Oropos” (ἐν ᾽Οροπȏ[ι]) and [χο]| ρίο ἐν ᾽Οροπȏι ἐν ἱερ[ȏι λιμένι]).67 The latter was the “sacred harbour” serving the Amphiareion, a place called Delphinium on the coast 5 km east of Oropus.68 The attestation of Oropian properties is principally important because of the role of Oropus as the vital link between Athens and its Euboean cleruchies.69

The sort of men who owned these properties are well represented by the aristocrat Charmides (cousin of the infamous oligarch Critias), who in Xenophon's Symposium (set in 421) recalls the days “when I was a rich man in this city”. He used to be worried by robbers and sycophants, he complains, and plagued by having to finance state projects (δαπανᾶν ὑπὸ τη̑ς πόλεως), “but now that I'm deprived of my overseas properties, and get nothing out of those that I have here, and the contents of my house are sold, I sleep happily and fully (p.92) relaxed…Then I used to bring tribute to the demos, but now the state imposes a tax and supports me”.70 This passage may be nothing more than a collage carelessly pieced together with references to various later events (the confiscations of 414, the Decelean War from 413, and the diobelia of Cleophon from 410), but it nevertheless preserves a historical sketch of the type of Athenian who owned overseas real‐estate and was obliged by the demos to reciprocate through liturgies.71

Xenophon's fictional Charmides mirrors the real person whose estate is the subject of Lysias 32: this Diodotus belonged to the liturgic class thanks in part to an annual rent income in grain, originating from an investment of 2,000 dr. in the Thracian Chersonesus.72 The plight of Charmides under the sycophants in turn recalls the comic image of Cleon's eisphorai as the bane of similarly rich Athenians “from Chersonesus”.73 Significantly, the eisphora is the very tax imposed by Athens on its cleruchs at Histiaea: it was evidently expected that Athenians settling there would already be, or soon be counted, among the city's plousioi.74 This evidence pointing to the existence of rich cleruchs and their liability to Athenian liturgies and taxation is impressive. The recently discovered Grain‐Tax Law of 374/3 now considerably expands our knowledge of this (p.93) phenomenon: as I interpret this law, it was the highest Athenian property class resident on the island who were to provide the demos with large quantities of public grain (σῖτος ἐν τῳ̑ κοινῳ̑).75

The distribution by lot of kleroi ensured, of course, that all Athenian citizen classes could benefit from cleruchies. Xenophon provides the example of Eutherus, who after returning to Athens after the Peloponnesian War says to Socrates: “Since we were deprived of our foreign property, and my father left me no property in Attica, I have been forced to take up my residence here and to earn a bare livelihood by manual labour”.76 This case reveals the democratic nature of the system, but cannot establish that Eutherus was the typical Athenian cleruch. A passage from Aristophanes' Wasps will in fact suggest the opposite: that the typical Euboean cleruch was more likely to be someone like Oeonias.77 The aim of fifth‐century Athenian cleruchies, despite Plutarch's oft‐quoted statement, was certainly not that of relieving Athens of a lazy and impoverished urban mob, and dispatching it overseas as some kind of imperial garnisaire.78

A. The Cleruchy at Chalcis and Cleruchic Mobility

The Athenian cleruchy established at Chalcis in 506 seems not to have outlasted the Persian Wars, and we last hear of it in the prelude to the Persian sack of Eretria.79 No explanation has ever been offered for its apparent dissolution or for why Pericles had to re‐expel the Hippobotai in 446.80 This is surprising, for Herodotus is fairly clear on the reason. At the request of Eretria, he says, the Athenians “gave (p.94) in assistance the 4,000 men who were cleruchs on the land of the Chalcidian Hippobotai”.81 When these cleruchs arrived at Eretria, Aeschines son of Nothon, one of the most prominent Eretrians, revealed to them that the situation there was hopeless and urged them to “go back to their homes”. Herodotus shows us what this meant: “The Athenians followed the advice of Aeschines, and by crossing to Oropus they saved themselves”.82 The implication is that most or all of the Athenian cleruchs resided in Athens, and not in the cleruchy at Chalcis.83

The recently published roster of 250 Council members in the Athenian cleruchy at Samos (first established in 365/4) serves to confirm this interpretation of Herodotus. The publishers of this relatively well‐preserved and unique inscription were able to identify a considerable fraction of the cleruchs (with varying degrees of certainty, from the absolute to the likely) with men present in Athens from the 360s to the 320s.84 Samos therefore presents an almost paradoxical case of durability and mobility, for on the one hand Athens outfitted its cleruchy with a full set of permanent civic institutions (a half‐Council of 250, and a half‐board of five generals, alongside the normal ten tribes, and nine archons), but on the other hand the cleruchs seem to have moved freely and frequently between Samos and Attica, and indeed to have lived their lives in both places.85

If this was true for Samos, it would be truer still for the cleruchs at Chalcis, who could easily travel the short distance between Euboea and Athens. They would, in other words, have been absentee owners, many or most of them no doubt rentiers like the Athenian cleruchs at Lesbos in 427 described by Thucydides:

Afterwards tribute was not imposed upon the Lesbians; but all their land, except that of the Methymnians, was divided into three thousand allotments, three hundred of which were reserved as sacred for the gods, (p.95) and the rest assigned by lot to Athenian cleruchs, who were sent out to the island. With these the Lesbians agreed to pay a rent of two minae a year for each allotment, and cultivated the land themselves.86

The new Samian document now lends overwhelming weight to the well‐known arguments of Brunt and A. H. M. Jones, to the effect that many or most of the cleruchs at Lesbos returned to live in Athens. Brunt explained: “Athens was an imperial city, governed by the citizens themselves, even the poorest, and there they enjoyed the pleasure of power, and its material perquisites, which they could vote to themselves”.87 This is correct as long as we are careful not to associate cleruchs with “the poorest”: obtaining a rent of two minae a year from Lesbos meant ipso facto that each of these 2,700 men was now of hoplite status (at least).88 Some of the bouleutai at Samos (like Euetion of Sphettos, appearing also in Athens as a mining‐lessee and trierarch) were in fact quite wealthy men.89

These cases of mobile and wealthy cleruchs underline that it is absurd to think of cleruchies as either military colonies or democratic charity‐affairs. While it is obvious that Athenian interests private and public were served by any cleruchs who chose to remain abroad, the idea that cleruchs were officially forced to remain overseas and garrison their cleruchies has absolutely no foundation.90 Athenian (p.96) absentee rentierism sufficiently explains, in short, why a new generation of Hippobotai apparently faced no Athenian opposition in returning to, and establishing themselves at, Chalcis by 446.

B. The Periclean Settlement

The settlement of the Euboean revolt in 446/5 drastically redesigned Athens' relations to the island. Pericles not only successfully established Athenian domination of Euboea on a secure and permanent footing, but also performed what was regarded as a democratic masterstroke. Although Tolmides (as is now practically certain) had imposed a cleruchy on Carystus in 453/2 or 452/1,91 the conquest of Euboea was still remembered decades later as the joint success of Pericles and the Athenian demos. Geometry appears in Clouds as the “democratic and useful scheme” (σόϕισμα δημοτικὸν καὶ χρήσιμον) with which the cleruchic land was measured out92—seeing the island on a map then prompts Strepsiades to say that the Athenians under Pericles had defeated and performed this intellectual exercise on the Euboeans.93 Athenian popular control over Euboea seems to have become a kind of comic topos in the fifth century, with another poet saying that thanks to Pericles “the demos became like a wild horse ‘and no longer dared to obey, but bit Euboea and leapt on the (p.97) islands’”.94 But comedy simultaneously reflects considerable ambivalence and even popular resentment towards a settlement which (as we have seen in the cases of Oeonias and others) undoubtedly enriched and benefitted Athens' already wealthy political class. In 422 Bdelycleon attacks these men as follows:

But whenever they're scared, they promise you Euboea and get set to supply you with fifty‐medimnoi rations of grain. But they never give it to you, not counting yesterday when you got five medimnoi, but only after narrowly escaping a challenge to your citizenship, and then it was barley in one‐quart installments. Which is why I kept you locked up: I wanted to feed you and didn't want these blowhards to make a chump of you.95

The passage clearly refers to habitual fears and promises, but particularly (as the scholiast thought, referring to Philochorus) to events of the year 424/3, when Athens seems to have reinforced its military control on Euboea.96 Especially if compared with the enthusiasm of Strepsiades, Bdelycleon's hostility gives the idea that Pericles' democratic project on the island had been betrayed: the land of Euboea and its grain, instead of a public good, were in the hands of a greedy few capable of clinging to it only by calling the Athenian demos to arms.

Of course, these are only statements of perception, useful on a general level but as historically imprecise as statements of the kind that fifth‐century Athens “held more than two‐thirds of Euboea”, or indeed the whole of it.97 Let us therefore turn to Pericles' settlement in more detail. The principal consequence of the Athenian revolt in 446, as Thucydides notes, was the depopulation of Histiaea and its resettlement with Athenian cleruchs.98 The number of cleruchs is variously recorded as 1,000 or 2,000.99 In order to make this (p.98) settlement permanent (i.e. to avoid the sort of dissolution just seen at Chalcis through cleruchic homecoming), Histiaea was provided with the institutional permanence of an Athens in miniature. Attested on the fragmentary inscription of her regulations is possibly her own council and court, and certainly a number of “travelling judges”.100

It is in reference to the fact that the Athenians at Histiaea “lived away” but nevertheless retained these “same institutions”, that Thucydides calls Histiaea (along with Lemnos, Imbros, and Aegina) a colony (ἀποικία) of Athens:

The Athenians themselves, who were Ionians, went of their own free will against the Syracusans, who were Dorians; they were followed by the Lemnians and Imbrians, and the then inhabitants of Aegina, and by the Histiaeans dwelling at Histiaea in Euboea: all these were their own colonists (ἄποικοι), speaking the same language with them, and retaining the same institutions.101

Thucydides here does not say that Histiaea, Lemnos, Imbros, or Aegina were apoikiai and not cleruchies, and his statement does not exclude other Athenian cleruchs serving in the Syracusan campaign.102 Thucydides is interested in the cleruchs that he lists simply because they were exceptional: they lived (or mostly lived) “on‐site”, and yet were still linked to Athens in speech and institutions (i.e. they had neither “gone native” nor become independent colonies). The Athenian cleruchic phenomenon was indeed remarkable and worth Thucydides' attention.103 The corollary of Thucydides' comment is (p.99) that fifth‐century Athenian cleruchs normally resided in Attica, and from there drew an income (in produce or rent) from their overseas properties.

We can now turn to aspects of Pericles' subjection of Euboea covered by Thucydides' term “by agreement” (ὁμολογίᾳ). Despite considerable resistance,104 the weight of the evidence strongly favours the case for the establishment of new cleruchies at Chalcis and Eretria. First the Tribute Lists: despite having been frustrated in their rebellion, Chalcis and Eretria experienced a lowering of tribute (at Chalcis from five(?) to three, and at Eretria from six to three talents), a moderate but necessary measure to compensate for the loss of part of their land to Athens.105 The stele recording the oaths of the settlement with Chalcis also mandates an exemption from Chalcidian taxes for ένοι who pay Athenian taxes, and these are probably to be identified with Athenians.106 Much more secure (in fact a direct parallel to the later Athenian settlement of Lesbos) is the evidence from the list of leases of Athenian sacred lands on Euboea.107 This document agrees with a garbled passage by Aelian, which should be interpreted as saying that the lands of the Chalcidian Hippobotai were now divided into 2,000 lots for Athenian cleruchs and the goddess Athena, presumably in the same proportion of 10:1 as in Lesbos.108 (p.100) The surviving leases show that this process of land distribution took place not only on the territory of Histiaea (which had been depopulated),109 but also on that of Chalcis,110 and Eretria.111 The unusually high numbers of Attic white ground lekythoi and choes from Eretria (the only one of these three sites which has undergone excavation), also suggest an Athenian cleruchy, especially since “there is a noticeable increase about the middle of the fifth century and these objects seem to be present in quantity right down to 411”.112 And probably at this time (since it is a direct corollary to Pericles' citizenship law of 451) the right of intermarriage (ἐπιγαμία) with Athenians was also given to all Euboeans.113 This law, which effectively freed the hereditary transfer of all Euboean private land into Athenian ownership, only makes sense as part of a full legal panoply giving Athenians the means of appropriating Euboean territory.

The future safety of Athenian interests on Euboea constituted a different facet of the settlement of 446. The Athenians entrusted this to their generals under a broad mandate: “As to the protection (ϕυλακή) of Euboea, the generals shall have the responsibility, to their best ability, that it be as excellent as possible for the Athenians”.114 The principal physical manifestation of this command to (p.101) “guard” or “watch” was the massive fortification of Rhamnous,115 a position that, as Ober noticed, “commands a fine view over the Euripus straits to Euboea, but has limited outlook inland in any direction”.116 As a fort protecting the Athenian cleruchies on Euboea, Rhamnous (and its view) had at least one predecessor. The much humbler fort on the edge of the Lelantine plain at Vrachos has now been excavated and dated by pottery to between the end of the sixth century and c.480. It is therefore contemporary, and in all likelihood associated, with the Athenian cleruchy established at Chalcis in 506. Its excavators describe the view from this fort:

To the west there is a clear view across the river and the fertile Lelantine Plain to the hill of Vathrovouni which conceals Chalcis, and the southward view over the coastal plain and the straits of Euboea is also excellent. But to the east the ground is broken, and an enemy could with care approach unseen.117

Significantly, the pottery at Vrachos does not seem compatible with an Athenian (or even Eretrian) occupation of the fort.118 This fact considerably puzzled the excavators, who hold to the fort's role in (p.102) protecting the Athenian cleruchy, but seem surprised to conclude that it was manned by a large troop (of up to 200) non‐Athenian mercenary archers.119 The puzzle is solved if (as we discussed above) Athenian cleruchs chose to stay on their plots or go back to Athens as rentiers, but in any case were never intended or officially used as garnisaires. We will return to Vrachos and Rhamnous below, but at this point we must note that they were probably not the only Athenian forts guarding the island. Archaeological survey strongly suggests that the Athenian guard of Euboea, especially after 446, entailed the construction of numerous other forts. As we shall see, this was in effect a permanent siege, by land and sea, of the entire island of Euboea and its major cities: “the Athenians in conjunction with Pericles besieged it, especially the Chalcidians and Eretrians”.120

C. Cleruchic Taxation

From the creation of the first documented Athenian overseas settlement at Salamis and throughout the Classical Period,121 the principle was the same: “cleruchs remained Athenians and were liable to Athenian taxes and to service in the Athenian forces”.122 Again, the latter condition does not mean that they were the garnisaires of their cleruchies, but simply that having a kleros qualified them ipso facto as hoplites,123 and that they could therefore be marshalled into the (p.103) Athenian army or cavalry in groups designated as, for example, “Histiaeans”, “Lemnians”, “Imbrians”, etc.124 Obviously, just as the ownership of the kleros provided the basis for Athenian military service, it also provided the basis for Athenian taxation, hence the evident signs of state concern regarding the kleros' transfer and rent.125

The problem of cleruchies and taxation, previously informed only by stray references to taxation at Salamis, Chalcis, and Histiaea, can now be better understood thanks to the well‐preserved Grain‐Tax Law of 374/3.126 Since the publication of Stroud's editio princeps and commentary, it is generally agreed that this law converts two previously existing taxes in cash (a dodekate, or one‐twelfth, and a pentekoste, or one‐fiftieth), originating in Lemnos, Imbros, and Scyros, into the same taxes in grain.127 Second, it is agreed that the payers of the taxes are the Athenian cleruchs themselves, since (as Miltiades and Cimon ensured, even more ruthlessly than Pericles at Euboea) all three islands lacked native populations by 476/5.128 Third, it is agreed that a fourth‐century nomos such as this was automatically endowed with a superior legal status and had perpetual effect.129 Fourth, it is agreed that the supply of grain envisaged by the law would be large and its management complex.130 In short, it is clear that, at least from (p.104) 374/3, the Athenian cleruchs of Lemnos, Imbros, and Scyros were directly responsible for supplying a considerable quantity of public grain to Athens.

This conclusion is obviously interesting in itself, but there still remain many serious problems in our understanding of the inscription. It seems likely that the pentekoste is, as Stroud thinks, a tax “levied on the cargoes of grain that left the harbours of the three islands”.131 But the rest of the inscription remains, on the whole, enigmatic. What role is played by the additional (and larger) dodekate? What is the function of the merides of 500 medimnoi, and 3,000 medimnoi (ll. 8, 32)?132 What, if any, is the significance of a rate of one‐twelfth? Were all or only some of the cleruchs on the islands liable to pay? On what basis was the tax sold in Athens (was the successful bidder he who gave the highest estimate of one‐twelfth of the following year's harvest)? Did Athens sell the tax to one or more tax collectors? How did he (or they) collect the tax? What is the role of a symmory of six men? Are these six tax‐collectors, or taxpayers? None of these questions has been satisfactorily answered, and Stroud himself concludes his commentary saying: “I publish this inscription knowing that I have left unresolved many of the important issues it raises”.133

The document is, however, fairly clear once we understand the nature of the tax itself, a dodekate. Stroud notes that this is “our first and only evidence for an Athenian tax at the rate of 8⅓%, ἡ δωδεκάτη”.134 This is only partly true, and crucially overlooks the fact (long ago demonstrated by Böckh) that the yearly income (p.105) defining the taxable capital of the first Solonic class, the pentakosiomedimnoi, was itself a dodekate.135 The tax of 500 medimnoi (100 of wheat and 400 of barley) imposed by Agyrrhius is thus simply the whole of the notional income that defined a Solonic pentakosiomedimnos. As Böckh also showed, the proportion between an income of 500 and a taxable capital of 6,000 (both again being obviously notional figures, officially assumed without regard for actual variations) must have been set by Solon on the basis of a net rent (i.e. income) from land of one‐twelfth, which he probably obtained by dividing the earlier prevailing one‐sixth (the old Athenian hektemorage) by half.136 As well as a tax, the dodekate can therefore also be (p.106) seen as a notionally payable “rent”: every pentakosiomedimnos was liable to pay this rent/tax if he held any Attic land (including land in cleruchies), because all such land must have notionally belonged to the Athenian polis.137

That Agyrrhius' law refers to a class of Solonic pentakosiomedimnoi on Lemnos, Imbros, and Scyros finds additional support in (or explains the presence of) the complete term for the members of this class in the genitive plural (πεντακοσιομεδίμνων) thirteen years earlier in the badly mutilated decree concerning the Athenian cleruchy at Lemnos.138 Equally suggestively, the same decree has a reference to the regulations of the cleruchy at Salamis established in Solon's day.139

Once the essential fact is clear that this is a tax on pentakosiomedimnoi, the full meaning of Agyrrhius' law emerges. By 374/3, when the drachma:medimnos relation was no longer 1:1, but approximately 9–6:1 for wheat, and 5–3:1 for barley,140 only an exceptionally wealthy Athenian would still be qualified to belong to the pentakosiomedimnic class and pay its tax (τελεῖν πεντακοσιομέδιμνον) in Solon's sense, i.e. to produce a taxable 500 medimnoi worth approximately 2,900 to 1,800 dr. And yet this is what Agyrrhius' law, in suitably democratic language, is asking his taxpayers to do (ll. 5–10). (p.107) The law was a democratic masterstroke not only because it provided a public supply of grain for the Athenian demos, but also because it turned a tax on the wealthy cleruchs of Lemnos, Imbros, and Scyros from a dodekate in cash (a relatively painless yearly payment of 500 dr.) into a much more onerous dodekate in grain, equalling a yearly 500 medimnoi from each cleruch.141

If Lemnos, Imbros, and Scyros had ever been offshore tax havens for the rich, Agyrrhius' law ensured this would no longer be the case.142 It is not surprising that Demosthenes called him “a good man, a democrat, a zealous champion of our masses on repeated occasions”.143 In effect his Grain‐Tax was a new and considerable liturgy that, although costing only about half the price of a trierarchy, had to be performed year after year.144

Let us now examine the sale and collection of the tax. The Athenian state would have a record of how many pentakosiomedimnoi held plots of land (kleroi) in its cleruchy, and could thus predetermine (even before the harvest) exactly how much grain was to be collected: it was simply a matter of multiplying 500 medimnoi by a number of kleroi.145 Now enter the prospective tax‐collector. He would be a single man, and (by necessity) would be extremely wealthy and well‐connected himself.146 For not only would the expense of arranging (p.108) and transporting the entire tax from the three islands to Piraeus, and from there to the asty, rest on his shoulders (ll. 10–15), but in order to buy the tax he would need to present two solvent guarantors for each 500 medimnoi in his bid (ll. 29–31).147 Significantly, the law refers to him as an emporos (l. 26). The sale of the tax is evidently a winner‐takes‐all contest between these individual shipping tycoons, each striving to outbid the rest in bringing back to Athens the greatest portion of the predetermined tax. The successful buyer would return to Athens with this contracted amount (the net tax), gaining the right to keep the difference between it and the full predetermined amount (the gross tax). This difference, minus the expenses incurred in collecting the tax, would be the tax‐collector's profit.

But how was the tax collected? Once the tax‐collector arrived in Lemnos, Imbros, and Scyros, he would find the pentakosiomedimnoi there neatly arranged in symmories to pay the tax, and each symmory would be responsible for handing to him a meris of 3,000 medimnoi (ll. 31–3).148 The Athenian state would hold each of the pentakosiomedimnoi and their symmories severally and individually liable for the transfer of this batch of grain to the tax‐collector (ll. 33–6). We can immediately see the logic of this arrangement: the symmories simplified the sale of the tax in Athens (since the payload of the average merchant ship was 3,000 medimnoi,149 each prospective buyer effectively bid by the numbers of ships that he could profitably afford to mobilize). By setting the tax burden on six pentakosiomedimnoi (p.109) instead of one, the symmories also lowered the risk of incapacities in paying the grain (e.g. in cases of harvest failure or unforeseen dowry payments). Finally, the symmories simplified both the collection of the tax by the tax collector, as well as the enforcement by the Athenian state in cases of non‐compliance with the law.

From the arrangement of the pentakosiomedimnoi into symmories we can easily deduce the minimum and maximum amounts of grain tax from the islands, and this in turn reveals the entire land‐holding structure of Lemnos, Imbros, and Scyros.150 For, on the one hand, the law clearly envisages that each island will have at least one of these symmories. And, on the other hand, the law also assumes that every cleruch on the islands will be a pentakosiomedimnos capable of producing at least enough grain to pay the tax (both from a cadastral and a democratic perspective, it would have been impossible for the Athenian state to grant pentakosiomedimnic kleroi only to some of its cleruchs). It might be asked in this connection whether this position conflicts with the fact that Athens kept a cavalry force, commanded by a hipparch in Lemnos during the fourth century. In fact, Hyperides shows us these hippeis as individually liable to pay a tax (termed a misthos), a useful fact to notice in connection with our inscription.151 But as Ste Croix shows, there is no contradiction in belonging to the tax class pentakosiomedimnoi and at the same time serving as a cavalryman.152

Now Scyros, the smallest island, has an area of 202 km2, of which approximately one‐quarter is cultivable; Imbros has 225 km2, of which approximately one‐half is cultivable; Lemnos has 477 km2, of which approximately three‐quarters are cultivable.153 The (p.110) distribution of the cleruchs would obviously have depended on the approximate proportion of cultivable land on each island, expressed roughly in the ratio 1:2:7. In other words, if Scyros had only one symmory, Imbros would have two, and Lemnos would have seven: and this total of ten symmories (i.e. sixty cleruchs) would have paid a yearly tax of 30,000 medimnoi of grain.

However, Lemnos, Imbros, and Scyros were obviously not occupied by just sixty cleruchs. The islands cannot have been so empty of people. Nor can one imagine each cleruch having a kleros of 870 ha (almost 10,000 plethra), since this would provide income next to which a dodekate of 500 dr. (or even 500 medimnoi) would be impossibly low.

Let us then imagine the converse scenario in which each kleros was only large enough to produce Agyrrhius' tax of 500 medimnoi. Assuming possible yields of 650 kg/ha and the use of biennial fallow, each kleros would need to measure about 50 ha (575 plethra).154 We should reserve one‐tenth of the cultivable land for sacred (and thus tax‐free) kleroi, following the example of Lesbos in 427.155 Dividing the islands in this fashion would result in 18 symmories for Scyros, 36 for Imbros, and 126 for Lemnos. This total of 180 symmories (i.e. 1,080 cleruchs) would have paid a yearly tax of 540,000 medimnoi of grain. But this again is an impossible scenario, not only because the dodekate cannot have equalled the entire production of the islands, but also because a yearly 540,000 medimnoi never came to Athens from Lemnos, Imbros, and Scyros. According to Demosthenes, 400,000 medimnoi came to Athens yearly around the mid‐fourth century from the Bosporan kingdom, and another 400,000 from “all other places” put together.156


Table 3. Estimates based on the Grain‐Tax Law of 374/3

Cultivable area

Number of kleroi of 1,000 plethra

Number of six‐person cleruch symmories

Total medimnoi produced (yield = 650 kg/ha, biennial fallow)

Medimnoi taxed as dodekate by Agyrrhius

Medimnoi remaining to Cleruchs after tax


60,000 plethra = c¼ of 202 km2

60 (54 for cleruchs, 6 sacred)






120,000 plethra = c.½ of 225 km2

120 (108 for cleruchs, 12 sacred)






420,000 plethra = c.¾ of 477 km2

420 (378 for cleruchs, 42 sacred)






600,000 plethra

600 (540 for cleruchs, 60 sacred)





The actual size of the cleruchies and their grain tax must therefore lie between these two extreme hypotheses, for example as in the scenario in Table 3. This is put forward only as a set of estimates, but we are now at least in the realm of the possible. Kleroi of 1,000 plethra would be very large indeed, but fully compatible with the largest Athenian land‐holdings attested.157 The estimated populations are also of acceptable size. On Scyros, a population of 60 families (54 cleruchs and six lessees of sacred lands) may seem especially low, but this does not take into account non‐cleruchs (e.g. metics or persons engaged in pasturage in the mountainous southern half of the island).158

Cleruchic slave‐ownership may have been unusually large, not only due to the wealth of a pentakosiomedimnos, but also (and (p.112) especially) because their labour would have been vital to the exploitation of very large estates. Assuming that in Greek antiquity one man could plough between 0.2 to 0.4 hectares and harvest 0.2 hectares per day, our 1,000 plethra (87 hectares) under biennial fallow would require 109 to 218 man days to plough and 218 man days to harvest.159 Guessing a force of forty to fifty slaves per plot, each task could be accomplished in four to five days.160 In sum, the number of cleruchs, non‐cleruchs, and slaves would result in a total population for the cleruchy at Scyros that is comparable to that of the modern island (c.2,700 people).161 We must remember that the total size of Scyros is 202 km2, so that (if our approximations are sound) population density on this cleruchy would have been thirteen inhabitants per square km. This level of habitation tells us something very interesting about the way that the Athenian grain supply from the cleruchies functioned, and how it compared to settlement patterns and the economy of Attica. Contrast Euonymon's calculated minimum population of 2,300 and its total size of c. 15 km2: a population density of at least 153 inhabitants per square km. This stark difference in settlement density between area of production and area of consumption resulted in large surpluses that could be shipped and sold at Athens, or taxed by Agyrrhius' dodekate in grain.

Our calculations of total grain production for Lemnos, Imbros, and Scyros put us in the realm of probability, since the First Fruits offered in 329/8 must have been very small fractions of a total of approximately this order. It is true that our estimates are, on average, about 25% higher per island than those extrapolated from IG II2 1672 by Garnsey and others, but this should not surprise us: as we have seen in Chapter 1, these scholars may well have used the wrong fractions to calculate total production; or 329/8 may have produced a bad harvest; or (according to A. H. M. Jones) Athenian farmers may (p.113) have systematically underestimated their crops.162 (In relation to this last possibility, we must observe that the dodekate, being a flat rent/tax amount of 500 medimnoi, was impossible to evade by underestimating production.) More importantly, our calculations allow for each of our cleruchs to pay a yearly tax of 500 medimnoi in grain, while retaining just as much for individual consumption and sale. This fits with the assumption, surely correct, that it was not Agyrrhius' aim, by switching the dodekate from cash to grain in 374/3, to kill the goose that lay the golden egg. Finally, our calculated tax of nearly 300,000 medimnoi fits well with Demosthenes' information that 400,000 medimnoi came to Athens from “all other places”. As grain sources for fourth‐century Athens, our three islands put together will have ranked only below the Bosporan Kingdom itself.163

The Grain‐Tax reveals much about the power and wealth of Athenian emporoi. For in order to collect even the minimum tax of 30,000 medimnoi, a single tax‐collector would have had to mobilize ten ships to Lemnos, Imbros, and Scyros. And at the likely tax level of about 270,000 medimnoi, ninety ships would be required. In order to collect this much grain, the tax‐collector would undoubtedly have to arrange sub‐contracts with a multitude of smaller emporoi, and call on a very large group of solvent guarantors (varying according to (p.114) what fraction of the gross tax he bid to bring back to Athens). Clearly the rules for the sale of this tax are drafted to favour men endowed with extensive maritime contacts and the support of a network of solvent friends. The Grain‐Tax therefore makes untenable the old view that all Greek emporoi were humble men.164

Let us now ask if there could have been an earlier dodekate in grain applied to the Athenian cleruchs at Histiaea, Chalcis, Eretria, and Carystus. Two considerations show this to be a very likely possibility. First, as we have seen, the tax and its computation as a rent in kind are Solonic, just as the first Athenian cleruchy at Salamis. But a second piece of evidence is decisive, and applies directly to Euboea. We have seen Aristophanes' Bdelycleon expressing angry disappointment with the politicians who “promised Euboea, but never gave it”.165 If Aristophanes is seeking in this passage to portray the demagogues as tight‐fisted crooks, the modern reader is inevitably confused to find a grain distribution “in five‐medimnoi amounts” (κατὰ πέντε μεδίμνους) being criticized, since this would entail each recipient's receiving a free choinix of barley every day for eight months, a lavish gift! How should we understand this statement? There is no error in the text: the five‐medimnoi amounts are a tenth of the fiftymedimnoi amounts (κατὰ πεντήκοντα) that the demagogues had originally promised one line earlier. It is the crucial evocation of the fraction (a tenth) that gives the joke its point: because it assumes that even the supposedly generous promise of fifty medimnoi fell one‐tenth short of the amount (i.e. 500 medimnoi—a dodekate) that Aristophanes and his audience must have perceived as being due from the cleruchic taxpayer on Euboea. In short, by enshrining in his law the promise that each share would be 500 medimnoi (ἡ δὲ μερὶς ἑκάστη ἔσται πεντακόσιοι μέδιμνοι), Agyrrhius was distributing grain to the Athenian demos precisely as Bdelycleon had wished, in 500‐medimnoi amounts. The Grain‐Tax Law of 374/3 would, under this theory, not be Agyrrhius' invention, but would have a direct predecessor in Athenian‐controlled, fifth‐century Euboea.

A Euboean dodekate in grain, if we follow Aristophanes' language, would have been habitually proposed by politicians, most recently (as the scholiast suggests) in 424/3 at a time of special danger to the (p.115) island, when Athenian troops were mobilized there.166 It was therefore not a yearly tax, but an eisphora, which incidentally is the term that we find in the mutilated regulations for the cleruchy at Histiaea.167 The Grain‐Tax of 374/3 itself is probably best considered as an eisphora, but (being embodied in a nomos) one to be paid every year in perpetuity. Agyrrhius' law thus becomes the first known yearly eisphora, a mode of taxation previously thought to have begun in Athens only during the last half of the fourth century.168

If Athens imposed eisphorai in grain on wealthy cleruchs on Euboea by 422, just as in Lemnos, Imbros, and Scyros in 374/3, our knowledge of the Athenian grain supply is radically transformed. The evidence shows that the importance of these islands as grain sources was based on surplus production from the land‐holdings of a relatively small number of wealthy citizens. Wealthy Athenians could transport their supply of grain to Athens and sell it there, but they could also be directly taxed for a considerable part of it by the demos. The problem of how this delicate fact should be handled undoubtedly shaped the political relationship of the Athenian demos and its aristocracy. Chapter 5 of this book will examine in greater detail the textual evidence for this interesting clash of economic and political ambitions.

D. The Decelean Route (446–13 BC)

Along with their regulations concerning the eisphora from Histiaea, the Athenians included regulations concerning the cattle (“horses, asses, and sheep”) belonging to their cleruchs, and the rates for ferrying passengers along the main routes in the Euboean gulf: Chalcis–Oropus, Oropus–Histiaea, and Chalcis–Histiaea.169 Lines 72–7 are most tantalizing. Lewis edits them thus: (p.116)

  • έσθο τέτταρας ὀβολό[ς..........21 .......... . ]
  • [. ]μὲν hοι πομπέυοντε[ς .........18......... πομ]‐
  • [π]εύεται, τελέτο τὸ hέ[μισυ. ἐὰν δὲ hο πορθμεύον μὲ]
  • [ἐ]θέλει ἄγεν τὸν πο[μπεύοντα ...7.... κατὰ τὰ γεγ]‐
  • [ρα]μμένα,. Ε. [..............29 ...............]
  • ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

The inscription is so fragmentary that it is impossible to know securely what is missing, but I think that this very specific clause, showing the Athenian state ensuring that those engaged in a certain kind of πομπή from Euboea pay only half‐price to the ferryman, is much more likely to refer to the conveyance of grain (πομπή or ἐκπομπὴ του̑ σίτου, or σιτοπομπία) than to a festival or mission.170 Whatever the specific content, the inscription clearly demonstrates that the sea‐traffic between Attica and Euboea was important and required the direct oversight of the Athenian demos.171 The mass naval evacuation of flocks and beasts of burden from Attica, sent “to Euboea and the nearby islands” in 431, was therefore not without long and significant precedent, both logistical and legal.172

The inscription also clearly shows the important role of the Oropus–Decelea route for transport between Athens and Euboea as far back as 446, and that such transport was not a measure instituted during the Peloponnesian War.173 Oropus (mod. Skala Oropou) had provided Athens with its principal “bridge” to Euboea since 506.174 (p.117) As Thucydides says, “Oropus, facing Eretria, while held by the Athenians could not be other than a serious annoyance, both to Eretria and to the whole of Euboea”.175 There was an Athenian ϕρούριον there for precisely this purpose.176 Describing the main route between Oropus and Athens, via Decelea, Ober concludes: “The Decelea road was clearly a major access route and definitely could be used by even the largest, most heavily encumbered armies”.177 There is no question that loaded mules and carts could easily cover this route of 48 km.178

E. The Sunium Route and the Loss of Euboea (413–411 BC)

It was the occupation of Decelea by the Peloponnesian army from 413 that cut the Oropus–Athens route and required the new and expensive step of sending all traffic of provisions from Euboea around Sunium.

Provisions, which had been formerly conveyed by the shorter route from Euboea to Oropus and thence overland through Decelea, were now carried by sea round the promontory of Sunium at great cost. Athens was obliged to import everything from abroad, and resembled a fort rather than a city.179

Although the Peloponnesians now “prevented the Athenians from enjoying the use of their land”, practically destroying Attica's agriculture and husbandry,180 this is clearly not the meaning of “at great (p.118) cost” in this context.181 Thucydides is here saying simply that sea transport from Euboea around Sunium was more expensive than transport by sea from Euboea to Oropus, and thence by land through the Decelean pass—an important exception to the modern dogma that land transport in antiquity was necessarily more uneconomical and inefficient than sea transport.182 To what in particular does Thucydides' “great cost” then refer?

The answer lies in the logistical aspects of securing the new Sunium route. Modern sailing times suggest that instead of a one‐hour sail between Oropus and Eretria,183 the Athenians now had to deal with journeys lasting over nine hours in the best conditions,184 longer given the notoriously hazardous winds and difficult waters past the narrows off Cape Ay. Marina, through the channel between Attica and the island of Helena (mod. Makronesos), and around Cape Sunium.185 A short crossing previously left to individuals using the local ferrymen, and in any case easily guarded from mutually visible shores,186 was thus turned into a massive state‐directed effort to engage, coordinate, and protect convoys of merchant ships using the new route.187

Thucydides gives the most important example of these centralized preparations: during the winter of 413/12, the Athenians “fortified Sunium for the protection of their grain‐ships on their sailing‐round (περίπλους) [from Euboea] to Athens”.188 Archaeology can amplify (p.119) this passage, although the crumbling cliffs of Sunium have taken some of this important fortification into the sea.189

Wrede carried out the fundamental study of the oldest remaining fortification at Sunium, a wall of poros blocks running northwest to southeast for about 110 m, and measuring up to 2.6 m in height.190 The wall could safely be dated to the last half of the fifth century based on independent criteria,191 and Wrede specifically attributed its unusually hasty and poor execution to the emergency described by Thucydides in 413/12.192 The wall of the fort, originally about 400 m long, with towers, controlled land access to the cape from north and east, while the sea‐facing cliffs, about 20 m high, completed the circuit to the south and west.193 The two ship‐sheds situated outside the walls to the northwest, and believed also to date from 413/12, were apparently built to house guard‐ships about half as long as triremes.194 The Athenians must have paid dearly for this project, but they received a clear return: the ground at the top of the fortress, 63 m above sea level, has uninterrupted visibility over the Aegean for about 225°, northeast to west.195

Archaeology has shown that the other pivotal Attic coastal fortresses on the route from the Euboean gulf to Piraeus (“le complément nécessaire du Sounion au Sud”) was Rhamnous.196 An outer circuit of towers and walls, about 800 m long and at least 6 m high, surround a fortified citadel 40 m above sea level.197 The site commands the Euboean gulf before it narrows dangerously at Cape Ayia (p.120) Marina, 5.25 km to the southeast. Visibility is excellent from the area of the narrows, all the way to the territory of Eretria at Amarynthos to the northwest. Just below the fort down its east and west slopes are two convenient harbours.198 The size and finish of the complex leave a decidedly different impression from the work at Sunium, and there is a simple reason why Thucydides does not group the two: the most recent excavations at Rhamnous have dated the fortification definitively to c.450.199 This fact is confirmed epigraphically: by 429/8, Rhamnous was secure enough for its sanctuary of Nemesis not to have to transfer its treasures to Athens, whereas, for example, Poseidon at Sunium did make such a transfer.200

The mid‐fifth‐century date provides the essential clue to the nature of this Attic fort. If we accept 446 (or shortly thereafter) as the likeliest date of its construction, we find that Rhamnous is as closely linked to Euboea historically as it is geographically (at a mere 8 km across the straits). We have, in other words, a fort built specifically to secure Athenian control of that island. We shall have opportunity below to explore its companion forts (largely unexplored and probably contemporary) ringing Euboea like so many iron bars.

We must now examine a third site, at Thoricus, where the Belgian Mission led by H. Mussche excavated a fort on the neck of the Ayios Nikolaos peninsula.201 The fort's walls, about 800 m long and surviving up to 2.7 m in height, enclosed an isthmus between two sheltered bays, Frankolimani and Port Mandri, which are strategically situated halfway between Piraeus and Rhamnous, and supplied (p.121) with spring water.202 The top of the fort, where the saint's chapel now stands, is at 31 m above sea level and commands the channel of Makronesos.203 Despite the intrinsic uncertainty in dating any ancient wall by its style of construction, Wrede's analysis at Sunium provided Mussche with the basis for dating the fort at Thoricus to the winter of 413/12—and this despite the explicit testimony of Xenophon, who dates it to the spring of 410/9.204 In the fort at Thoricus Mussche claimed to identify not only a design similar to Sunium's, but also the same hasty and poor masonry work.205 Should we accept Mussche's arguments?

Although Mussche does not note this, Xenophon's alleged mistake in post‐dating could be placed alongside another (much more serious) flaw in the earliest section of the Hellenica: the recall of the Syracusan Hermocrates, wrongly put in the summer of 410/9 instead of 411/10.206 If Xenophon was indeed mistaken and Thoricus was fortified in 413/12, it would, however, be much more difficult to account for Thucydides' silence. We should surely expect him to say “they fortified Sunium and Thoricus for the protection of their grain‐ships…”. It is, in short, more likely that Thoricus was not fortified in the same year as Sunium.

But the archaeological insights of Mussche can be combined with another, much better alternative date. In midsummer of 411, before the revolt of Euboea,207 we find the Four Hundred fortifying Eetioneia (p.122) and the makra stoa in the Piraeus.208 The leaders of the operation, according to Xenophon, were Aristoteles, Melanthius, Aristarchus, and their fellow generals.209 Thucydides describes their activities:

They also walled off the largest stoa in the Piraeus and the nearest to the new fortification, which it joined; this they controlled themselves, and ordered that the grain be seized, both what was already inside and what came in by sea, and that it be taken for sale from there.210

It is clear from this passage that the Four Hundred were attempting to exercise a strict control over Athenian grain stocks and supplies.211 This should cause the attentive reader no surprise, and even less suspicion, or else Thucydides (who has a well‐known antipathy for this group) would hardly have missed the opportunity to raise it.212 This was Agis' third year at Decelea,213 Oropus had been lost to the Boeotians less than six months before,214 and Euboean rebels had made two known overtures to Sparta over the past eighteen months.215 Athenian finances were in a shambles, and (as we have seen) the transport of grain around Attica was extremely costly.

Even before a Peloponnesian fleet appeared in the Saronic gulf, it was therefore clear that the summer of 411 would be decisive for the fate of Euboea and its role as a source of grain for Athens. But now Agesandridas and his forty‐two ships were at hand, ready at any moment to sail along the coasts of Attica to Euboea and raise the (p.123) island in revolt.216 It was in these urgent circumstances that the Four Hundred will have decided to supplement Sunium and Rhamnous, the two principal forts securing the passage of grain from Euboea, with a third fort at Thoricus and a fourth at Piraeus itself, on Eetioneia.

Unfortunately, Thucydides does not tell us any of this, at least not directly. He will simply not allow the leadership of the Four Hundred or their generals to speak for themselves. Instead, Thucydides is happy to report only the inflammatory criticism of their dissident colleague Theramenes:

Theramenes insisted that these ships were intended, not for Euboea, but for the party who were fortifying Eetioneia, and that if the people were not on the alert, they would be undone before they knew where they were.217

And again:

Theramenes insisted that if they had been on their way to Euboea they would never have gone up the Saronic gulf to Aegina and then turned and anchored at Epidaurus but that someone had invited them for the purposes which he had always alleged; it was impossible therefore to be any longer indifferent.218

Both of these abbreviated speeches reveal something that Thucydides does not wish to present explicitly, namely the explanation of the Four Hundred themselves for their policy of fortification. If we read between the few lines assigned to Theramenes, we can see plainly enough that they were appealing to the existing danger to Euboea. Their previous actions amply confirm this: only eight days after the inauguration of their new Council on 22 Thargelion, the Four Hundred had sent out to Eretria one of their own members, Polystratus of Deirades, as its new garrison commander (a fact which we learn, not from Thucydides, of course, but from pseudo‐Lysias!).219 Let us therefore consider what the danger to Euboea had to do with a fortification at Piraeus, and the solution is clear: with the Athenian navy at Samos in open insurrection against the new regime, Piraeus (p.124) was dangerously unprotected, its main harbour and grain stores vulnerable to a Peloponnesian raid—or worse. All shipping into Piraeus was likewise open game. This fact, and not some nefarious concert with the Four Hundred, is what must have brought Agesandridas into the Saronic gulf in the first place, and retained him there for some time. Years later, one early morning in the spring of 387, Teleutias with twelve ships showed just what Agesandridas will have hoped: slipping into Piraeus, he seized merchants and shipowners from the Deigma,220 towed away their ships, and sailed back out to intercept fishermen, ferries, and merchantmen bringing in grain and other goods from the direction of Sunium.221 In the face of this very real and dangerous possibility, it made sound sense for the Four Hundred to fortify not only the entrance to the harbour, but also the makra stoa with its stores of grain. They also needed to reinforce the route from Euboea against Peloponnesian piracy, which meant fortifying Thoricus as quickly as possible. The hasty and poor masonry work discovered at Thoricus reflects Thucydides' description of the work at Eetioneia (“they worked eagerly on the fortification and wanted to have it finished in time”), and we might have comparable archaeological remains in Piraeus today had the building not been completely dismantled at the instigation of Theramenes.222

It is clear why Thucydides, who wants to guide his reader as unambiguously as possible to his own version that Eetioneia was nothing but a stratagem to betray Athens to Sparta, never makes clear this concern of the Four Hundred for the safety of Piraeus or Euboea. His wholehearted espousal of Theramenes' conspiracy theory regarding Eetioneia is barely concealed behind a sly façade of distanced objectivity. “The charge was not a mere calumny, but had some foundation in the disposition of the ruling party”.223 Thucydides leaves unmentioned the fortification of Thoricus because doing so would have given the lie to his case, and revealed the honest concern of the Four Hundred, his villains, for the safety of the Athenian grain supply.

Xenophon probably had more innocent reasons for placing the fortification of Thoricus in 410/9. As Mussche showed, his interest in (p.125) that fortification is directly linked to his long‐held conviction of the importance of the Laurium mines.224 Xenophon refers to Thoricus once again (and at greater length) in Ways and Means, but only in connection with its use as a base for cavalry and patrols protecting the mines against enemy incursion, and as a place to secure the mines' slave personnel.225 Here Xenophon is undoubtedly looking back to the Decelean War,226 and his memory of 410/9 in connection with Thoricus (if not an outright mistake, his second misplaced event from the summer of 411) may perhaps relate to its role (even its re‐fortification?) against an otherwise undocumented threat from Decelea to Laurium in 409. It is clear why Xenophon would have disregarded Thoricus' role as an important naval station on the route from Euboea: if its fortification by the Four Hundred is correct, that role will have turned out to be negligibly short. When Euboea rose in revolt in late August or early September 411, the fort had probably been in use for not much longer than a month.

We can reflect that had the partisan opinion of Theramenes against the fort at Eetioneia not prevailed, and had the Four Hundred been given the opportunity to act on their legitimate interest in securing the Euboean grain‐route, Athens might well have avoided the final calamity of losing Euboea. Instead, Athenian forces spent much of the precious summer of 411 absorbed in paranoia, fearing and ineffectually dismantling a useful fortification. It was not until Agesandridas had finally sailed around Sunium that the majority finally woke up to the imminent loss of their principal granary, and scrambled to piece together some semblance of a response. Disingenuously, as if the Four Hundred had not long anticipated this by sending out Polystratus, Thucydides now says: “the matter was vital and urgent: Euboea was all in all to them now that they were shut out from Attica”.227 The historian makes the pitiful defeat of the thirty‐six Athenian ships finally assembled at Eretria almost sound like a surprise: (p.126)

When the news of the battle and of the defection of Euboea was brought to Athens, the Athenians were panic‐stricken. Nothing which had happened before, not even the ruin of the Sicilian expedition, however overwhelming at the time, had so terrified them. The army at Samos was in insurrection; they had no ships in reserve or crews to man them; there was a revolution at home—civil war might break out at any moment: and by this new and terrible misfortune they had lost, not only their ships, but what was worse, Euboea, on which they were more dependent for supplies than on Attica itself. Had they not reason to despair? But what touched them nearest, and most agitated their minds, was the fear lest their enemies, emboldened by victory, should at once attack the Piraeus, in which no ships were left; indeed they fancied that they were all but there. And had the Peloponnesians been a little more enterprising they could have easily executed such a plan…228

The paragraph is a masterpiece: Thucydides here prepares his reader to assume uncritically that the deposition of the Four Hundred, to which his narrative turns immediately, was simply a natural consequence of their failure in avoiding this disaster, and their negligence in bringing Athens so close to total ruin.229 For had only the Spartans exploited this victory by besieging Piraeus, he says, they would have anticipated Lysander by seven years.230 Only a writer as skilful as Thucydides could avoid the fact that the Four Hundred had been right all along, and in the end land them squarely with the blame.231

IV. Forts and the Imperial Territory

The “Old Oligarch”, that shrewd observer of fifth‐century Athenian imperialism, argued that thalassocracy was the fundamental factor in giving the Athenians control of the natural resources from Sicily to (p.127) Egypt to Pontus. Island allies, he tells us, were kept obedient by the threat of starvation (λιμῳ̑),232 while mainland cities were controlled either through fear (διὰ δέος) or need (διὰ χρείαν), depending on their size: “for there is no city which does not have to import or export, and these activities will be impossible for a city unless it is obedient to the rulers of the sea”.233 This naval supremacy was responsible, according to him, for giving Athens access to crops (οἱ καρποί) (2.6), luxury goods (σμικρότερα; εὐωχίαι; ὅτι ἡδύ) (2.7), naval matériel—timber, iron, copper, flax, wax—and wealth generally (ὁ πλου̑τος) (2.11). And, while enemies could be excluded from these resources, “I, without doing anything, have all this from the land because of the sea”.234

In this connection (cf. ἔτι δέ), the Old Oligarch mentions what he believes is the indispensable tool of Athenian naval enforcement:

…every mainland has either a projecting headland or an offshore island or some strait, so that it is possible for a naval power to put in there and to injure those who dwell on the land.235

Making no claim to being exhaustive, we can point to the offensive use by the Athenians of five such locations in the pages of Thucydides.236 The first of these is on Atalante (“formerly an empty island off the coast of Opuntian Locris”) from 431;237 the second at Boudoron (“a promontory on Salamis facing Megara”) from 429 to 427;238 the (p.128) third on the island of Minoa (“in front of Megara”) from 427;239 the fourth (and most elaborately narrated by Thucydides) at Pylos from 425 to 409;240 and the fifth at Delphinium (on Chios) from 412 to 407.241

It is clear, however, that a naval fort could also be used defensively, as in the case of the Megarians at Minoa before the loss of the island to the Athenians.242 Or a fort could serve offensive and defensive functions simultaneously. This was the case with the island of Atalante on the Euboean straits (see Map 2), fortified by the Athenian general Cleopompus in 431 “in order to prevent pirates (λῃσταί) sailing from Opus and other places in Locris and plundering Euboea”,243 but also as an “ἐπιτείχισμα of Locris for waging war on the inhabitants of the country”.244 As the otherwise uncertain regulations for Histiaea suffice to show, piracy around Euboea had been a concern since the foundation of the cleruchy.245 Thucydides adds that Cleopompus had been sent “for the protection (ϕυλακή) of Euboea”.246 This is probably the official wording of the mission, or close to it, since its last clause recalls the final lines of the Chalcis decree of 446/5.247 The type of Athenian protection exercised at (p.129) Atalante in 431 perfectly suits the concepts that interested the Old Oligarch: by protecting Euboea, maritime forts would have provided Athens with the resources of the land through the control of the sea.

This section will present evidence suggesting that other Athenian maritime forts besides Atalante ringed Euboea. This survey work necessarily provides only a preliminary footing. Nevertheless, its cumulative weight shows that the scholiast to Aristophanes is to be literally believed when he says that the Athenians under Pericles laid siege to the island (ἐπολιόρκησαν).248 And it seems that this encirclement lasted well beyond 446. The best‐known segment of this ring of fortresses is, as we have seen, at Rhamnous. The archaeological case for the rest must remain circumstantial until excavation is undertaken, but we can make considerable headway by selectively examining the five examples offered by Thucydides. They will help us understand not only the general purpose, but also the physical location and construction of these forts.

Thucydides explains that on the promontory of Boudoron “there was a fort (ϕρούριον) and a guard (ϕυλακή) of three ships to prevent anything entering or exiting Megara by sea”.249 The fort on the island of Minoa, located directly before Megara, was obviously expected to accomplish this blockade much more efficiently.250 Delphinium (15 km to the north of Chios) had a different, more complex purpose: it was used to wreak havoc on the economy of an enemy territory,251 while its small double harbour also offered easy control of the sea‐traffic along the coast of Asia Minor.252

The location of these forts was determined not only by strategic considerations, but also by the access to supplies and the possibility of anchoring or beaching ships.253 Access to building materials for (p.130) the fort was also crucial, as Demosthenes insisted at Pylos.254 Construction was to be quick, easy, and cheap, out of the available wood and stone. Thucydides offers the fullest description in the case of Pylos:

…at length the soldiers, who were standing about idle, were themselves seized with a desire to fortify the place forthwith. So they put their hands to the work; and, being unprovided with iron tools, brought stones which they picked out and put them together as they happened to fit; if they required to use mortar, having no hods, they carried it on their backs, which they bent so as to form a resting‐place for it, clasping their hands behind them that it might not fall off. By every means in their power they hurried on the weaker points, wanting to finish them before the Lacedaemonians arrived. The position was in most places so strongly fortified by nature as to have no need of a wall…In six days the Athenians finished the wall on the land side, and in places towards the sea where it was most required; they then left Demosthenes with five ships to defend it.255

It is interesting to compare this description with the reported excavations on the fortification wall on the acropolis at Delphinium:

The thickness of the wall averaged 2.10 m and the construction was of rough stones quarried from the hill with no attempt at facing. There is hardly a stone in it which could not have been moved by one man, and everything points to hasty building. Such a wall could hardly have stood higher than about 3.0 m.256

 We have no information on the number of men stationed in these garrisons, and we can expect that they varied from case to case. Diodorus, however, gives “about five hundred” as the number of the Athenian garrison at Delphinium surrendering to Callicratidas in 407.257 A number of two to five guard‐ships for each maritime fort seems the standard, as the cases of Atalante, Boudoron, and Pylos show.

(p.131) After the fortification of Atalante in 431, we hear no more of it until the great earthquake and tsunami of 426, which “carried away a part of the Athenian fort and dashed in pieces one of two ships which were drawn up”.258 The full description of that event (evidently famous) by the Hellenistic historian‐seismologist Demetrius of Callatis specifies that these ships were triremes, and shows that they were drawn up on slips (νεώρια) located just outside the Athenian enceinte.259 Atalante lay some 20 km across the Euboean gulf from Orobiae. But the preliminary survey of Euboea carried out by Sackett et al. (from 1939 to 1965) suggests a dozen other similar sites much closer to the island, of which the best example is Ayios Vasileios, on the coast north of ancient Kerinthos (nos. 13–14 on Map 2). The survey describes it as:

…once an island and now a rocky promontory separated from the main island by a sandbar, and having abundant evidence of a Classical fortified settlement. Walls, tiles, and many sherds including good fifth‐century black glaze are thickly scattered on the promontory and on the slopes opposite…260

This short description, written in the 1960s, still holds perfectly. Autopsy on 26 March 2000 revealed a desolate and still unexplored site, c.400 m (northeast–southwest) by 200 m (northwest–southeast). The rocky nature of this chalk islet, like the rugged shore of the mainland directly opposite, makes it distinctly useless for an agricultural settlement. On the other hand, the enormous amount of ceramic debris from black‐glaze vessels, tiles, and amphorae (the place having, as far as I could tell, no water‐source of its own), shows that the place was long occupied. The stone socles of those buildings that are visible above ground (especially on the north side) recall the masonry work described at Pylos and Delphinium, and make the use of this site as a (p.132) maritime fort very probable. At least one area with the right slope and size for hauling up guard‐ships exists on the islet itself,261 but the site's main advantage consists in its control of the enormous, 6 km‐long sandy beach stretching south from modern Angale to ancient Kerinthos.262 It also commands a wide panoramic view along much of the northeastern coast of Euboea, and as far as the islands of Sciathos and Peparethos (mod. Skopelos) to the north.

Sackett et al. treat Ayios Vasileios as a representative site, suggesting that it formed part of a system of naval fortification together with other Classical sites along the eastern coast of Euboea.263 From Oreoi/Histiaea to Carystus additional possible fort sites were identified at (with reference to Map 2): Elliniko (no. 11),264 Vasilika (no. 12),265 Kotsikia Paralia (near Kerinthos),266 Tambouri (near Kerinthos),267 Kerinthos (nos. 13–14),268 Pilion (p.133) (Kastri) (no. 15),269 Vlakhia (Asprokhorto),270 Limnionas (?) (northeast of Mt. Dhirfys),271 Lamari (?),272 Cape Kyme (no. 83),273 Ano Potamia (no. 81),274 Philagra (no. 89),275 and Geraestus (no. 92).276 A complementary function would have been served by similar sites on the opposite (and more accessible) side of the island, at Khironisi (no. 4),277 (p.134) Likhas Kastri (ancient Dion?) (no. 2),278 Yialtra Kastelli (ancient Athenae Diades?) (no. 3),279 Orobiae (no. 19),280 Atalante,281 Politika (ancient Aigai?) (no. 26),282 Chalcis,283 Eretria,284 and Amarynthos.285 East of Amarynthos a ship would enter the line of sight of the fortress at Rhamnous, but additional fort sites may be located at Aliveri‐Mylaki (ancient Porthmos?) (no. 65),286 Nea Styra (no. 88),287 and Marmari (no. 90).288

Whether acting as shelters for shipping, forts (ϕρούρια) against pirates and other enemies,289 or stations (ὁρμητήρια) and border bases (ἐπιτειχίσματα) against non‐Athenians on Euboea (or adjacent (p.135) territories),290 these sites will have been the bases for the enormous Athenian naval power (probably dozens of triremes) mentioned by Thucydides as guarding Euboea during the Peloponnesian War.291 This force will have been occasionally supplemented by special missions, like the thirty ships sent to ravage Locris and protect Euboea in 431.292 But the creation of permanent facilities, like those described at Atalante or evident from the massive amount of surface remains at Ayios Vasileios, indicates constant garrisoning. This fact lends support to Thucydides' testimony that the Athenians had 16,000 men in the reserves in 431. He divides these men into two groups: those in the forts, which were hardly all in Attica, and those on the battlements (i.e. in Athens itself).293

It was precisely as its own base to counter the Athenian phrouria guarding Euboea that in 426 Sparta took the radical step of sending out a colony to Heracleia in Trachis, an event which (as Hornblower shows) Thucydides treats with special emphasis and with significant narrative parallels to his description of Decelea.294 Besides helping the Dorians and Trachinians: (p.136)

…they also thought that the situation of the new city would be convenient for carrying on the war against the Athenians: there a navy could be equipped if they wanted to attack Euboea, which was quite near, and the station would be handy for the conveyance of troops to Chalcidice.295

But defending the colony from its own neighbours proved more trouble than the Spartans had foreseen, and their plans to use Heracleia against Euboea were not successful. Nevertheless, the foundation can certainly be seen as an important precursor to Sparta's role in aiding Euboea's defection from Athens in 411. This was the crucial first attempt by the non‐naval power actually to deprive Athens of its overseas grain‐supply.

Athenian forts not only ringed Euboea, but also were probably coordinated with other forts controlling its interior. Thucydides mentions one of these inland forts as the shelter for the Athenian crews defeated at Eretria in 411. It must have been not far from the coast:

Those of them who took refuge in the city of Eretria, relying on the friendship of the inhabitants, fared worst, for they were butchered by them; but such as gained the fortified position which the Athenians held in Eretrian territory escaped, and also the crews of the vessels which reached Chalcis.296

Even after the Euboean revolt, this fort continued in Athenian hands, subsisting financially from sums it managed to raise from the territory of Eretria, but reported to Athens as a loan (ἀνομολόγημα) due to the goddess.297

These testimonies of a fort in Eretria are considerably supplemented by two inland forts that have actually been excavated and studied. (p.137) These are Vrachos and Ano Potamia, the former in the territory of Chalcis, the latter in that of Eretria. On Vrachos (near Phylla on Map 2), fortification walls and two buildings have been dated to the late‐sixth to early‐fifth centuries from pottery evidence, and Coulton reconstructs its Building 3, a large (112‐m‐long) row of twenty identical rooms, as an army barrack capable of housing around 200 men.298 We have already seen that the position of this fort, suitable only to observing movement from Chalcis over the Lelantine plain, fits well into the context of the Athenian cleruchy at Chalcis in 506.299 The pottery from this fort is, as we have also seen, particularly valuable in proving that Athenian cleruchs were never required to garrison their own overseas lands.300 Vrachos heads a list of other (still unexcavated) fort sites on the largest plains on the island, namely that of Histiaea in the north,301 and that of Psachna (at Ayios Ilias (no. 30), Pirgos (no. 31), and Ayia Paraskevi (no. 33)), just north of Chalcis.302

Ano Potamia (no. 81) was excavated by A. Sampson from 1976 to 1978. The fort, on a 267‐m‐high hill overlooking the bay of Kyme, consists of two co‐centric walls encompassing an area (p.138) c.120 m (east–west) by 80 m (north–south). The upper fortification (Section VI) could be dated to the mid‐fifth century from pottery, especially lekythoi and fine black glaze.303 Within the fort on its lower western side are two stone‐lined cisterns and a metallurgical workshop: the pottery and architectural evidence for these structures presented by the excavator are meagre but fully compatible with a similar mid‐fifth‐century date.304

In the territory of Histiaea, Sackett et al. suggest that Athenian cleruchs were responsible for hill fortifications located inland at Prokopion/Achmetaga (no. 16)305 and Dhafni Stefaniou (near no. 18).306 The former controls the Cereus, the latter (together with contemporary hill forts at Strofilia (no. 17)307 and Dhafni Kastro (no. 18)308) the Neleus river valley. In the Classical Period, a convenient land‐route across the northern third of the island followed the Boudoros to the Neleus, ascended gently to the 300 m Mourties‐Misipetri Ridge,309 and descended to the Euboean gulf at Limni (ancient Elymnion or Aigai?310) via contemporary sites at Ayia Kiriaki (near (p.139) no. 22)311 and Limni Kastria (no. 21).312 The survey suggests other cross‐island routes further north connecting Orobiae (mod. Rovies) with the plain of Histiaea. One led over a Classical site on the 800 m pass near Galatsadhes over Mt. Telethrion,313 another via the lower (400 m) pass between Mts. Telethrion and Xiron through Palaiokhori.314 Athenian forts in the territory of Histiaea may have been necessary despite the expulsion of the Histiaeans, since (as the fragmentary regulations of the cleruchy show) the Ellopians and the people of Dium were allowed to remain there, and the Locrians opposite were considered to pose a threat.315

In the cleruch‐occupied territory of Eretria, a long valley running almost at sea‐level connects the Classical fortifications at Cape Kyme316 with Aliveri‐Mylaki (ancient Porthmos),317 via the Classical hill sites above Ano Potamia (ancient Oichalia or Kyme/Komi?),318 Avlonarion (ancient Oichalia or Tamynai?) (no. 72),319 and Lambousa. (no. 71).320

When properly excavated and studied, many of these sites will probably prove to be similar to those excavated by Sampson and by Sapouna‐Sakellaraki et al. If so, the Athenian phenomenon of fortification for territorial defense, so well‐known from fourth‐century Attica, will have been pioneered in the fifth‐century cleruchies.321 (p.140) Of course, some of these sites will also turn out not to have been just forts (or even forts at all), but either variants of fortified hill‐towns like Dystos, whose exceptionally well‐preserved enceinte and buildings date to the fifth century,322 or individual farmsteads like those near Sunium, studied by Young.323 In any case, the number, preliminary dating, and character of even a fraction of these sites join our other evidence in making clear that Athenian protection and exploitation of Euboean grain was a phenomenon long pre‐dating the Peloponnesian War.

V. Conclusion

This chapter has presented evidence for the importance of fifth‐century Euboea in supplying Athens with grain. We have stressed the role of wealthy and mobile Athenian cleruchs in this phenomenon, and moved to the purely physical mechanisms that gave Athens control of the island and its agricultural resources. We have seen that all aspects of the cleruchic system that provided grain for Athens, from its foundation, to its taxation, to its protection, were addressed by Athenian law and ranked as vital interests of the Athenian state.324

The Grain‐Tax of 374/3 now provides a valuable fourth‐century parallel. For although Euboea was the granary of the Athenian Empire after 446, its history as such cannot be divorced from Lemnos, Imbros, or Scyros, or indeed from that of earlier cleruchies. The history of the Athenian cleruchy begins with Salamis in the time of Solon,325 but the origin of the great overseas Athenian land‐holder is really to be found in the north Aegean, beginning with Phrynon at Sigeion c.600.326 In this tradition fits the Peisistratid control of (p.141) Sigeion and Lampsacus, and finally Philaid control of the Thracian Chersonesus, Lemnos, and Imbros.327 The ultimate purpose of these projects was not to settle or to feed the Athenian poor. It was instead to provide the abundant lands and resources for an Athenian elite. It was also to enable these aristocrats to continue to live (and indeed to rule) as princes over serf‐like populations of Thracians or Pelasgians or the like—a luxury denied to them over Athenian citizens after the Solonian reforms. Most important, however, is the fact that for these men a desire for overseas resources was far from the wish to sever ties with their city. Political life and competition with their peers in a common arena (at Athens) was still the focus of their ambitions and their desire for wealth. The result was that peculiar Athenian form of overseas colonization, in which colonies never became autonomous, but were perceived and treated as parts of an overseas Attic territory (χώρα).328

Along with the Cleisthenic reforms, the Athenians approached this colonizing program on democratic but equally aggressive terms: to Salamis were added the lands of the Chalcidian Hippobotai in 506. The process may have acquired a democratic veneer—all Athenians were now equals and companions in the project—but the desire to raid and reap the fruits of overseas lands remained essentially the same. In fact, the new democracy drank so deeply and so early from this sort of exploitation that it became second nature. Difficult as it is not to admire the Athenians' effort against Persia, especially in 490, it must be remembered that, just as freedom burned brightest in their hearts, they followed Miltiades to enslave their fellow Greeks at Paros. He “promised to enrich them if they would accompany him”.329 But when Miltiades failed, he was impeached before the demos. The prosecutor was Xanthippus, the father of Pericles, and the proposed penalty was death. It was only by pleading his services (p.142) at Marathon and his conquest of Lemnos that Miltiades was fined fifty talents and allowed to die from his own wound.330 The rapacious ethos of the aristocratic buccaneer was, however, now firmly enshrined in Athenian democratic politics.331

Miltiades dead, his son Cimon tried his hand after 479. Following firmly in the footsteps of the Philaids he not only re‐established Athenian control in the north Aegean, but also took Scyros in 476/5 and gave it to the demos to colonize. Claiming to have cleared the pirates' nest on that island, Cimon resembled Thucydides' Minos, who had done the same—and like Minos he had made the sea safe “so that its revenues would increasingly converge on him” (του̑ τὰς προσόδους μᾶλλον ἰέναι αὐτῳ̑).332 The analogy runs even deeper, for Minos:

…having expelled the pirates when he colonized the greater part of the islands (οἱ γὰρ ἐκ τω̑ν νήσων κακου̑ργοι ἀνέστησαν ὑπ᾽ αὐτου̑,ὅτεπερ καὶ τὰς πολλὰς αὐτω̑ν κατῴκιζε), the dwellers on the sea‐coast began to grow richer and to live in a more settled manner; and some of them, finding their wealth increase beyond their expectations, surrounded their towns with walls. The love of gain (κέρδος) made the weaker willing to serve the stronger, and the command of wealth (περιουσία) enabled the more powerful to subjugate lesser cities.333

How easy it is to confuse the mythical Minoan with the fifth‐century Athenian empire of revenues and colonies. There could be no mirror more explicit of the Athenian empire as Thucydides knew it, a system that von Reden has characterized as “the application of local policies of power and patronage on a large scale”.334 And so, Pericles and Tolmides followed Cimon (and Minos), taking Athenian cleruchs to all corners of the Aegean.335 And Alcibiades, fresh from taking a mere 500 cleruchs to Melos, quickly moved in 415 to the next and far more grandiose level—the conquest and colonization of Sicily.336 Why not? (p.143) By the time of the Peloponnesian War, the chain of Athenian cleruchies stretched from Pontic Amisus (renamed Piraeus!), near Colchis, to Piraeus. Along the way lay Sinope, Astacus, the Thracian Chersonesus, Lemnos, Imbros, Scyros, and Euboea.337 This was not yet a grain‐route to Crimea, but itself the grain‐source of Athens: an empire of cleruchies with Euboea as its crown jewel.

The Athenians who went as cleruchs to these places were either wealthy land‐holders, or (as probably at Brea) were allotted the lands that made them wealthy land‐holders. In the latter case the system was indeed democratic. What is important is that the interests of all Athenian cleruchs either were or became those of an elite. And so their duties. As Agyrrhius' Grain‐Tax shows, the Athenian demos could expect to be supplied by men receiving great surpluses as settlers on what was technically Athenian public land. We may say that the gradual collapse of the Athenian Empire, beginning in 411 with the loss of all of Euboea except Histiaea, and culminating in 404 at Aigospotamoi, transformed this system only superficially. The cleruchies on the route to the Black Sea, once lost, had nevertheless given to their old owners deep personal and economic connections to that distant world. The experience of the repatriated cleruch offered the possibility of building a new system with the same elite at the helm.

It should therefore not surprise us, as we will find in the next chapter, that a group of Athenians discovered special affinities of political outlook, and even artistic taste, with the kings of Panticapaeum. In 404 the Athenian empire was lost, but the shape of the city's grain‐supply experienced, fundamentally, a smooth transition into the fourth century.


(1) Isoc. 4.107–9.

(2) Hdt. 5.31.

(3) Hdt. 5.77. On the cavalry aristocracies of Chalcis and Eretria, see Arist. Pol. 1289b35–40. Compounds of ἵππος as personal names are very prevalent at Eretria: see the prosopographical analysis in Wallace (1947), 128–30.

(4) Hdt. 5.77, who records the dedicatory epigram; Paus. 1.28.2.

(5) See GHI 15 (=IG I3 501): the preserved fragments of both the first and the later inscription.

(6) Thuc. 1.114.

(7) Plut. Per. 23.

(8) Thuc. 8.96.

(9) GHI II 22, ll. 25–46. “Disavowal of ownership of real property in allied territories is the most thoroughly spelled‐out promise of the decree of Aristoteles” (Cargill (1981), 146). As Gauthier says (1972), 170, this clause gives “comme une vision ‘en négatif’ de la 1re Confédération athénienne”.

(10) Isocrates, like any good lawyer, makes no factual misrepresentations in the passage quoted: Histiaea was in fact ἐρημουμένη and depopulated (because it had been made so by the Athenians!), and the expression οὐδὲν τούτων ἡμᾶς ἐπη̑ρε περὶ τοὺς ἔχοντας τὴν νη̑σον ἐξαμαρτεῖν is suitably vague. There are no modern scholars even seemingly prepared to take Isocrates at his word, except (surprisingly) Meiggs (1972), 567. See Conclusion, pp. 315–16.

(11) So Garnsey (1988), 132–3; Westlake (1948), 2–5.

(12) See e.g. Westlake (n. 40 below); see also Picard's work on fourth‐ to first‐century Euboean history (1979), 338–9: “Les historiens modernes ont cru que l'Eubée était alors une grande région exportatrice de blé; l'île aurait été le grenier traditionnel d'Athénes; c'est par convois entiers que le blé eubéen serait arrivé dans la ville. Je ne sais comment une île aussi montagneuse aurait pu donner des récoltes de cette importance…rien ne permet de faire de l'Eubée une terre à blé fortement exportatrice.” Among the earlier historians criticized by Picard are Gernet (1909), 308–10; Jardé (1925), 194; and Michell (1940), 260–2.

(13) See Geyer (1903), 14–19. The comic poet Sopater (4th–3rd c.) called Eretria “the city of white barley‐meal”: Ἐρέτριαν ὡρμήθημεν εἰς λευκάλϕιτον (ap. Ath. 4.160b). Theophrastus (HP 8.4.4, plagiarized by Pliny NH 18.70) discusses light wheats from Euboea and especially from Carystus. The soil types of Chalcis, Kerinthos, and Eretria also find frequent ancient references, see Geyer (1903), 18 n. 4.

(14) See Philippson (1950–9), vol. 1.2, 561–637.

(15) Histiaea occupied the end of Euboea stretching west of the range running W–E, at opposite ends of which stand Mts. Aegae (mod. Kandhíli) and Pyxaria: see Bakhuizen, p. 127. It included the towns of Orobiae and Kerinthos. The “farming districts here” consist of the valleys of the Cereus and Neleus, and the Histiaean plain. The total area of the territory is approximately 1,000 km2; cf. IACP 656. At the western tip of the island, Cape Kenaion (probably the territory of the small polis of Dium) seems also to have been part of the Athenian cleruchic settlement, undergoing the foundation of Athenae Diades: see Strab. 10.1.5: ᾽Αθη̑ναι αἱ Δίαδες, κτίσμα ᾽Αθηναίων, ὑπερκείμενον του̑ ἐπὶ Κυ̑νον πορθμου̑ with IG I3 41, ll. 101–2.

(16) See Bakhuizen (1985), 127: “The Chalcidian territory proper is marked by natural geographical boundaries…It has roughly the shape of half a circle (radius c. 17 to 20 km) [i.e. 907–1,256 km2, so up to about half the total area of Attica]. It is a basin which is enclosed by an almost unbroken succession of mountain ranges to the north, east and south‐east, but which is open to the Euboean Sea along the base of the half circle (length: c. 34 km).” Bakhuizen (1985), 130–2, divides this into five “farming districts”: (1) the foothills of Mt. Aegae (mod. Kandhíli); (2) the alluvial Psachna plain; (3) the land between the Psachna and Lelantine plains; (4) the Lelantine plain; and (5) the coastal strip from the Lelantine plain to the eastern border with Classical Eretria (near modern Malakondas).

(17) This territory stretched east–southeast from the range running southwest–northeast, at opposite ends of which stand Mts. Voudhókhi (anc. Olympus) and Aloktéri; the southeast border was with Styra, which seems to have been independent: see Bakhuizen (1985), 123; Wallace (1947), 115 n. 2, and 129. The “farming districts” here consist of the Eretrian plain, the Aliveri valley, the coastal valleys near modern Kyme, and the inland valleys at Avlonarion, Lepoura, and Dystos. The total area of this territory is approximately 900 km2.

(18) Geyer (1903), 15–16 (“ein wahrer Garten…eine Stunde breit”); the ancient testimonia on this plain are on p. 16 n. 1; see IACP, 643–63.

(19) See n. 15 above; see also Bakhuizen (1985), 140 n. 33, 141 n. 42; see more detailed maps in Sackett et al. (1966).

(20) Many of these accounts are collected in Noel‐Baker (2000), 29–30, 47, 50–1, 182, 184, 191, 264–5, 326–7; Lancaster (1947), 106–9: “…across a wide and fertile valley, through groves of plane trees, reputed to be even larger and older than those in the Vale of Tempe itself…”; Buchon (1911), 47–8: “…le plus beau ravin que j'aie encore vu…une véritable retraite pour un poëte”; Teller (1880), 129: “Nur wenige Gebiete des heutigen Hellas können sich in bezug auf landschaftliche Schönheit mit diesem Eilande messen, das in seinem engen Rahmen die vershiedenartigsten Bilder umschließt, Gemälde von wilder Großartigkeit, wie sie dem höheren Kalkgebirge eigen sind, und Szenerien von weicherem freundlicherem Charakter, wie sie die tiefen, schattigen, im herrlichsten Vegetationsschmuck prangenden Talkessel der Schieferregion oder die fruchtbaren, gartenähnlichen Tertiärbecken darbieten”.

(21) Strab. 10.1.14.

(22) Ephorus FGrHist 70 F119 = Strab. 9.2.2. The bridge across was built in 410 with Boeotian help (Diod. 13.47.3): see Bakhuizen (1985), 48–49.

(23) See Philippson (1950–9), vol. 1.2, 568.

(24) Noel‐Baker (2000), 338.

(25) See e.g. Buchon (1911), 47–8: “…l'andrachni (arbousier à tronc de corail) qui n'est qu'un arbrisseau dans l'Attique, croît ici à l'égal des plus grandes arbres”.

(26) The former was a gentleman cousin of Lady Byron, the latter an aristocrat, son of Hofwyl's founder, Philipp Emanuel von Fellenberg. See Noel‐Baker (2000). Control of Euboea had been formally transferred to the Greek state under the London Protocol of 1830 but Turkish ownership was respected: see Philippson (1950–9), vol. 1.2, 635.

(27) Edward Noel in Noel‐Baker (2000), 50–1: “I was obliged to decide before hearing the advice of friends, as every day the land rose in price and if not purchased directly from the Turks, who are to leave the country, would fall into the hands of Greek speculators, who would make you pay threefold”. The purchasing power of £10,000 in 1833 is approximately equivalent to that of £637,000 in 1998: Twigger (1999).

(28) For the precise composition of the soil, see Zvorykin (1939); a scientific description of the landscape appears in Philippson (1950–9), vol. 1.2, 585–9.

(29) Noel‐Baker (2000), 51.

(30) See Buchon (1911), 54–5, who discusses a rent in kind collected by the Noels in 1841 from their peasant tenants at a warehouse at Drasi: the rent was a third of their harvest, compared to the fourth then being levied by the Greek government for the use of public lands.

(31) Edward Noel: letter dated 23 November 1836 (Noel‐Baker 2000, 89). Cf. Theophr. HP 5.2.1.

(32) Dem. 21.167; Meiggs (1982), 194, 206 suggests that Attica and Euboea would have been the source for Athenian rebuilding after 480/79, and normally for fuel.

(33) See Noel‐Baker (2000), 62–3, 179; Philippson (1950–9), vol. 1.2, 585. Sallares (2000), 17–42, makes the case (supported by the latest microbiological evidence) for the presence of malaria in Greece as far back as the eighth century.

(34) Noel‐Baker (2000), 126–7.

(35) Baird (ap. ibid. 184).

(36) Turner et al. (2001), 225.

(37) Ibid. 129, 161, 218.

(38) Bursian (1862–72), 406–7 describes the Noel estate as: “…jetzt, Dank den Bemühungen aus dem westlichen Europa eingewanderter grosser Grundbesitzer, welche eine rationelle Landwirthschaft und eine rationelle Forstcultur eingeführt haben, der am besten angebaute Theil der Insel”; Buchon (1911), 45–6: “J'entrai dans une délicieuse petite vallée dont le centre et tous les flancs étaient couverts des blés les plus beaux, les plus vastes, les plus purs de toute herbe parasite, et entourés de tous côtés de forêts verdoyantes; au milieu coulait un ruisseau bordé d'une double haie dont le vert foncé contrastait agréablement avec le vert plus tendre des blés. C'était pour moi un espectacle nouveau, celui du travail intelligent de l'homme pliant à son utilité les beautés et les richesses de la nature.”

(39) For Greece, see Garnsey (1988), 95 nn. 18–19. The average yield of wheat per hectare in Euboea was 640 kg in 1939, 960 kg in 1951, 936 kg in 1957, 1,193 kg in 1960, and 1,463 kg in 1979. The corresponding figure for Greece as a whole in 1979 was 2,235 kg/ha (i.e. about one third higher than in Euboea): Settas (1984), 118–19; Jardé (1925), 203, has the figure of 835 kg of wheat per hectare for Euboea in 1921: compare how this ranks with e.g. Arcadia (1,153 kg), “Macedonia” (978 kg), Crete (929 kg), “Attica and Boeotia” (697 kg), Chios (494 kg).

(40) Area: Settas (1984), 22 (Philippson (1950–9), vol. 1.2, 564) gives the area of Euboea as 3,530 km2). My table is based on figures in Settas (1984), 121–2. Cf. the following misleading statement by Westlake (1948), 4: “The cultivable parts of the island are very fertile, but they amount to only about one‐fifth of its area, the remainder being mountainous country fit only for pasturing sheep”.

(41) Receipt records for the first Ottoman tax (in grain) on Euboea have been published (Balta (1989), see esp. 35, 156–7). They would indicate that total grain production on the island was 4–5,000 tons. But such receipts leave us with similar problems to those already seen in relation to IG II2 1672 (including the lack of a secure coefficient—was the dîme one‐eighth?): was this a good or bad harvest? How efficient or corrupt was this first report and collection of the island's produce for the Ottoman invader? Conversely, what was the impact of Venetian withdrawal?

(42) Andoc. 3.9; Aeschin. 2.175. Andocides in particular, it must be remembered, was no competent student of Athenian history, as the ridiculous mistakes peppered throughout his speech immediately demonstrate (see e.g. 3.5, on the fortification of Piraeus). Aeschin. 2.172–6 is, in turn, a plagiarism of Andoc. 3.3–12, pace Harris (2000). I use the term “controlled” narrowly, not meaning to imply that Athens imported this entire amount from Euboea (a local population needed to be fed!).

(43) Peyer (1950), 108; Aymard (1966), 47 n. 4; Philippson (1950–9), vol. 1.2, 635.

(44) Polyb. 4.38.

(45) Lock (1996), 108, 110; Aymard (1966), 47.

(46) See Lock (1996), 111.

(47) See ibid. 110; Philippson (1950–9), 1.2, 704.

(48) See Aymard (1966), 3: “Durant ce demi‐siècle [1550–1600], Venise importe moins, et souvent de moins loin. Depuis qu'elle a au XVe siècle étendu sa domination dans la basse et riche plaine padane, elle tend à demander à ses territoires de Terre Ferme une contribution de plus en plus importante.”

(49) In Braudel (1972), 592. The Venetians evidently thought they could do much better than the Ottoman tax collector: see n. 41 above.

(50) Aymard (1966), 15–20; see also Braudel (1972), 420–1.

(51) See Sackett et al. (1966), 58, 70, figs. 10–11.

(52) For the date, see Pritchett (1953), 232–4.

(53) APF, 419.

(54) Pritchett (1953), 233, 252 (Stele II, ll. 177–8).

(55) Pritchett (1953), 254, 261 (Stele II, ll. 311–14). On Oeonias, see also Andoc. 1.12–13: like Nicides, he was accused of profaning the Mysteries, fled Athens, and was sentenced to death in absentia. He appears as the owner of two further properties, perhaps at Oropus (see n. 67 below), and at another, unknown, location (Stele X, l. 33).

(56) Probably referring to a house and land: see Pritchett (1953), 263 (Stele IV, ll. 15–21); (1956), 271.

(57) Pritchett (1953), 251 (Stele II, ll. 90–5), 274 (Stele VI, ll. 150–1, may refer to ἐπιμελ[ηταί ἐν] Ἐρετρία[ι—]according to Lewis (1966), 190 n. 48, Green and Sinclair (1970), 525 n. 45, and if so would probably refer to slaves (χωρίς οἰκου̑ντες, cf. the other slaves on Stele VI, ll. 20–2) rather than to Athenian officials, as assumed by Lewis (1966), 184).

(58) The topic is too complex to be discussed here, but see Erxleben (1975), 84–5; Cargill (1981), 192–7; cf. the unfounded doubts by Salomon (1997), 166 nn. 553, 554. See also Finley (1973), 6–7, 122, 147–9, 151 on the horoi evidence (from Lemnos and Scyros), now updated for Lemnos by Salomon (1997), 171–5. Gauthier's general view that cleruchic land was inalienable (1966, 70, 88) is held only by a minority of scholars, as he himself admits (1973, 163; cf. 173 n. 3), and causes insurmountable difficulties in the case of Euboea.

(59) IG I3 418, ll. 6–7 with Raubitschek (1943), 28–33.

(60) Raubitschek (1943), 31; see Pritchett (1956), 263 for the definitions of γη̑ ψιλή and πεϕυτευμένη.

(61) Raubitschek (1943), 31; Erxleben (1975), 88, with Pritchett (1953), 272 (Stele VI, ll. 66–8). Panaitios is also mentioned in Pritchett (1953), 252 (Stele II, ll. 170–1).

(62) Erxleben (1975), 88; Khironisi (n. 277 below) is a Classical site on this isthmus, while the tower site at Ay. Paraskevi (n. 302 below) is tentatively identified as Argura.

(63) See Pritchett (1956), 275–6; Osborne (1985), 47–63.

(64) See Dem. 20.115; cf. Plut. Arist. 27.2 (who only records the γη̑ πεϕυτευμένη). APF, 48–53, has exploded the ancient myth that Aristides “the Just” and his family were impoverished, but it does not follow (as he thinks) that this decree was a fourth‐century forgery. It is obvious that this very large grant of land was intended as an honour (Erxleben (1975), 87: “eine Auszeichnung in Würdigung der Verdienste des Vaters”), and would provide Lysimachus with much more than bare maintenance. And, even if based on a forged decree, the story would be based historically on the amounts of land actually owned on Euboea by fifth‐century Athenian aristocrats, confirmed by the Attic Stelai. If genuine, the decree belongs soon after 453/2 or 446 and is probably to be attributed to Alcibiades II, the grandfather of the famous general: see Erxleben (1975), 87–8 with APF, 15–16.

(65) GHI 52 (= IG I3 40), ll. 64–9; on Hierocles and oracle collectors, see Parker (2005), 111–15.

(66) Ar. Pax 1046–7, 1125–6. On Elymnion, see n. 310 below. On Hierocles, see GHI, 143.

(67) Pritchett (1953), 286 (Stele VIII, ll. 4–6). As Erxleben thinks (1975: 85 n. 11), this property may have belonged to Oeonias, who appears again in ll. 8–9.

(68) Strab. 9.2.6. Lolling (1885: 352–4) identified this παλαιὰ Ἐρέτρια with Amarynthos, probably rightly (see Sackett et al. (1966: 65); he also identified Delphinion with the closest point on the coast to the Amphiareion, the shallow bay at Kamaraki (Mandraki), with submerged remains (1885: 351–2). Two ancient walls lining the bay have now been surveyed by Cosmopoulos (2001), 59–60, 90–1, figs. 39, 54 and (1989), 273–6, figs. 1–4, but remain insecurely dated. Thucydides' statement ἀπέχει δὲ μάλιστα ὁ ᾨρωπὸς τη̑ς τω̑ν Ἐρετριω̑ν πόλεως θαλάσσης μέτρον ἑξήκοντα σταδίους, which agrees with the distance in Strabo, may suggest that it was this bay that Agesandridas used as his base (in the territory of Oropus) before attacking the Athenians at Eretria.

(69) See pp. 116–17 below.

(70) Xen. Symp. 4.30–2.

(71) See APF, 331.

(72) Lys. 32.6 (δισχιλίας δὲ ὀϕειλομένας ἐν Χερρονήσῳ), 15 (ϕοιτᾶν δὲ καὶ σῖτον αὐτοῖς ἐκ Χερρονήσου καθ᾽ ἕκαστον ἐνιαυτόν), 24 (liturgies). Diodotus had made a fortune in ἐμπορία (32.4: ἐργασαμένου δὲ Διόδοτου κατ᾽ ἐμπορίαν πολλὰ χοήματα…; see 32.25: his brother investing on a voyage to the Adriatic). See Cohen (1989), 210 n. 16: “Return paid not in money, but in crops, would be especially appropriate only to a loan relating to land. The recurring yearly payment, of course, suggests a calculation of yield related strictly to the passage of time, i.e. interest.”

(73) Ar. Eq. 261–5. Also see the scholion: ‹ἐκ Χερρονήσου:› Χερρόνησος τη̑ς Θράκης χωρίον καὶ πόλις, ὑποτελὴς τω̑ν ᾽Αθηναίων, εὔϕορος εἰς πυρου̑ γεωργίαν. ὅθεν καὶ ἐσιταγώγουν οἱ ᾽Αθηναῖοι, κτλ. See also: Ar. Eq. 773–6; 923–6.

(74) IG I3 41.38 with Cary (1925), 244, 250; Graham (1964), 171–2; Hansen (1999), 112–16 (for the fourth‐century evidence on εἰσϕορά); on this clause see Schmitz (1988), 89–90, and for evidence of other “prominenten” (e.g. Aristophanes, Ariston) in fifth‐century Athenian cleruchies, see pp. 84–9: “…in der Regel jeder Bürger unabhängig von seiner Zensusklasse sich an einer Kleruchensiedlung beteiligen konnte”. Among these, the case in Pl. Euthphr. 4c is particularly interesting: a wealthy Athenian family residing on a farm in Naxos with their slaves (οἰκέται) and hired labour (πελάτης). For the fourth century cleruchies see Cargill (1981), 196: “numerous fourth century Athenian settlers are from families that can only be described as wealthy”; he collects the prosopographical evidence in his appendix B.

(75) See pp. 104–6 below.

(76) Xen. Mem. 2.8.1 (Penguin trans.).

(77) See n. 95 below.

(78) Plut. Per. 11 with Brunt (1966), 71. Plutarch here probably had in mind Roman social and military conditions. We must not forget that the lines from the Brea decree which appear to exclude the two higher property classes from the cleruchy (see n. 138 below), are an amendment to the original decree, which originally (as Meiggs and Lewis rightly saw, GHI, 132) “did not restrict membership at all”.

(79) Hdt. 6.100–101; cf. Graham (1964), 169 n. 1.

(80) See n. 7 above. See e.g. Geyer (1903), 48, and most recently, Ostwald (2002), 135 n. 13: “what happened to it [the cleruchy] can only be speculated”.

(81) Hdt. 6.100 with Brunt (1966), 87–9.

(82) Hdt. 6.100–1.

(83) Pace e.g. A. H. M. Jones (1957), 175.

(84) Hallof and Habicht (1995), 293–9.

(85) Ibid. 299: “Die neue Urkunde lehrt manches über die Mobilität der Kleruchen…Es zeigt nämlich, daß die Entscheidung des Einzelnen, sich in einer Kleruchie anzusiedeln, nicht unwiderruflich war, der Betreffende vielmehr früher oder später nach Attika zurückkehren konnte”; and p. 303: “Die etwa 250 Namen umfassende Liste athenischer Ratsherren auf Samos zeugt mit besonderer Eindrücklichkeit für die Enschlossenheit Athens, annektierten fremden Besitz für die Dauer zu behaupten”.

(86) Thuc. 3.50 (Crawley trans.) with Brunt (1966), 81–4.

(87) Brunt (1966), 72 n. 8; see A. H. M. Jones (1957), 174–6.

(88) See [Arist.] Ath. Pol. 7.4: ζευγίσιον δὲ τελεῖν τοὺς διακόσια τὰ συνάμϕω ποιου̑ντας, with [Dem.] 43.54: ὁ δὲ ζευγίτης ἑκατὸν πεντήκοντα (sc. δραχμάς); see also n. 135 below. Probably by the early fifth century, the terms of the Solonic requirement changed to being expressed in drachmas instead of in medimnoi, on a 1:1 basis: see Busolt (1926), 822 n. 1, 837, 1272, 1276; Rhodes (1993), 142–5; cf. the assessment in [Arist.] Oec. 1347a18–24, where the sum of 200 dr. per person again appears, though in a context difficult to interpret.

(89) See Hallof and Habicht (1995), 294.

(90) A. H. M. Jones (1957), 174–5, rightly dismisses the two main sources for this view: Plut. Per. 11 (see n. 78 above) and Isoc. 4.107 (see n. 1 above); in contrast, he emphasizes that no military participation is recorded of Athenian cleruchs in key cases during the Peloponnesian War (Lesbos in 424 and 411, and Euboea in 411) when we should expect it had they been present in the cleruchies. The exception proves the rule: in 411 the whole of Euboea revolted, except Oreoi/Histiaea which was held by the Athenians (Thuc. 8.95: καὶ ὓστερον οὐ πολλῳ̑ Εὔβοιάν τε ἅπασαν ἀποστήσαντες πλὴν ᾨρεου̑ (ταύτην δὲ αὐτοὶ ᾽Αθηναῖοι εἰ̑χον)…). Cf. Salomon (1997), 120–50 (rightly criticized by Osborne (1999), 207), who takes the extreme and very doubtful position that Athenian cleruchies were rotating garrisons temporarily settled on public lots; see also the further doubts cast by Faraguna (1999), 69–73. Geyer (1903), 50, 71, long ago held the same mistaken view (e.g.: “so war die Kleruchie in Chalkis keine Ackerbaukolonie, sondern eine Garnison…”, and: “es [the Eretrian apoikia] war eine Ständige Garnison, die im Verein mit der chalkidischen, histiaiischen und karystischen die Insel bewachen und zugleich in Botmässigkeit erhalten sollte”). But the phenomenon of Militärkolonien, very different from cleruchies, is distinguished and discussed, e.g. in Busolt (1926), 1279.

(91) Paus. 1.27.5 mentioning (and Diod 11.88 dating to 453/2) Tolmides' cleruchy “to Euboea and Naxos”, with IG I3 259, col. II, l. 16 showing Carystus' tribute as 12 talents in 454/3 (this is lowered to 7.5 talents in 450/49). The number of cleruchs is reported by Diod. 11.88 as a total of 1,000 for Euboea and Naxos. See Erxleben's discussion (1975: 85–7) against the version of events reconstructed, before the fragment was known, in ATL III, 294–9.

(92) Ar. Nub. 202–5.

(93) Ibid. 211–13. I take the verb παρατέμνω here to operate in a triple pun: to lie geographically, to lay low (i.e. defeat), or to lay out (for measurement into kleroi).

(94) Plut. Per. 7.6.

(95) Ar. Vesp. 715–21 (Loeb trans.).

(96) See n. 114 below.

(97) These statements are made by Andocides and Aeschines respectively: see n. 42 above. See also the Ravenna schol. on Ar. Nub. 213: ἐκληρούχησαν δὲ αὐτὴν ᾽Αθηναῖοι κρατήσαντες αὐτη̑ς.

(98) Thuc. 1.114 (Ἑστιαιᾶς δὲ ἐξοικίσαντες αὐτοὶ (sc. οἱ ᾽Αθηναῖοι) τὴν γη̑ν ἔσχον); also Philoch. FGrHist 328 F118: Περικλέους δὲ στρατηγου̑ντος καταστρέψασθαι αὐτοὺ πᾶσάν ϕησιν Φιλόχορος, καὶ τὴν μὲν ἄλλην ἐπὶ ὁμολογίᾳ κατασταθη̑ναι, Ἑστιαιέων δὲ ἀποικισθέντων αὐτοὺς τὴν χώραν ἔχειν.

(99) Diod. 12.22 (1,000); Theopomp. FGrHist 115 F387 (2,000); 2,000 is taken as the more likely figure for the cleruchs in IACP, 656 (citing Figueira (1991), 258–60); see Cary (1925), 248.

(100) IG I3 41, ll. 58, 89–110 with Cary (1925), 249, who follows von Gärtringen's restoration of thirty δικασταί, just as in Athens ([Arist.] Ath. Pol. 26, 53). Lewis restores the number as [hεπ]τά, but thirty now seems a more likely guess if the recently published list of Athenian councilmen in the cleruchy at Samos from 365 to 321 is any guide. Here are nine archons, as in Athens, while the Council (250) and Generals (5) are precisely half: see Hallof and Habicht (1995), 288–91. On “travelling judges” (δικασταὶ κατὰ δήμους) see Andrewes (1982 b), 407.

(101) Thuc. 7.57 (Jowett trans.).

(102) Modern efforts to deny that Histiaea was a cleruchy, based largely on this sentence (e.g. Figueira (1991), 12–13, 69–70, 223), are unconvincing (“the nature of the distinction is not to be determined philologically”: Brunt (1966), 73, 77); see also Graham (1964), 170–2.

(103) See Haloff and Habicht (1995), 301: “Die Aussendung von Kleruchien war ein typisch athenisches und fast allein auf Athen beschränktes Phänomen”. Indeed, the phenomenon of overseas colonization by a thalassocrat is a major theme in Thucydides, sounded first at 1.4 (Minos) and returning thereafter in his work: Kallet (2001), 25–6.

(104) See especially Figueira (1991), 258–60; Meiggs (1972), 565–70.

(105) This depends on the restorations proposed in ATL III, 294–5.

(106) IG I3 40, ll. 52–7: τὸς δ| ὲ χσένος τὸς ἐν Χαλκίδι,hόσοι οἰκȏντες| μὲ τελȏσιν ᾽Αθέναζε, καὶ εἴ τοι δέδοται h| υπὸ τὸ δέμο τȏ ᾽Αθεναίον ἀτέλεια, τὸς δὲ ἄ| λλος τελε̑ν ἐς Χαλκίδα, καθάπερ hοι ἄλλο| ι Χαλκιδέες, with Ostwald (2002), 141 (restating ATL III, 297): “The objection that ‘an Athenian decree would not call Athenians ξένοι [in GHI, 143]’ has only limited validity once we assume that the present Athenian decree echoes the language of the original Chalcidian request for a ruling about the obligations of aliens in their midst. Support for this assumption can conceivably be derived from the prominence of the phrase τὸς δὲ χσένος τὸς ἐν Χαλκίδι which introduces this clause (52–3), and whose position parallels the introductory phrase περὶ δὲ τȏν hομέρον in line 47. There is, accordingly, no reason to deny Athenians a presence in Chalcis by arguing as Gauthier does, that Athenian allies among resident aliens are meant here…”. See also n. 114 below.

(107) IG I3 418.

(108) Ael. VH 6.1 with ATL III, 295–6 (whose interpretation of the passage I accept), and Raubitschek (1943), 30. One of the stones marking Athena's land was found at Péï, some 5 km northeast of Chalcis (IG XII.9.934: [τ]έμενος| [᾽Α]θηναίης); cf. IG I3 418.2–3: […c.5 ..]ς τεμένε| [ἐν Χαλκί]δι παρὰ τὸ τη̑ς ᾽Αθεναίας προσ[τȏιον ?—] (restoring [᾽Αθηνᾶ]ς on l. 2 seems possible despite the alternative form on the following line )).

(109) ἐν Ἑστιαίαι—: One lot: l. 6: ᾽Οροβίασι.

(110) ἐν Χαλκίδι—: three separate lots: ll. 3: παρὰ τὸ τη̑ς ᾽Αθεναίας προσ[τȏιον ?—], 11, and 22.

(111) ἐν Ἐρετρίαι—: two separate lots: ll. 9: Αἰγαλε̑θεν, and 14.

(112) Green and Sinclair (1970), 522–7.

(113) Lys. 34.3.

(114) IG I3 40 (= GHI 52), ll. 76–9 (446–5 BC) (περὶ δὲ ϕυ| λακε̑ς Εὐβοίας τὸς στρατηγὸς ἐπιμέλεσ| θαι h ος ἂν δύνονται ἄριστα,h όπος ἂν ἔχε| ι h ος βέλτιστα ᾽Αθεναίοις) (Fornara trans.). The down‐dating of this decree to the Euboean expedition of 424/3 (recorded by the scholiast to Ar. Vesp. 718 with Philoch. FGrHist 328 F130: τὰ περὶ τὴν Δὔβοιαν δύναται καὶ αὐτὰ συνάιδειν ταῖς διδασκαλίαις· πέρυσι γὰρ ἐπὶ ἄρχοντος ᾽Ισάρχου ἐστράτευσαν ἐπ᾽ αὐτήν, ὡς Φιλόχορος), as proposed in H. B. Mattingly (1961), 124–32, is unconvincing. One would need to assign impossible perversity to Thucydides in failing to record a large expedition to Euboea, especially given his own clear belief in the island's importance to Athens (esp. in 8.95–6, but elsewhere in his work from 1.114 on). The expedition is missing from Thucydides simply because it was relatively minor: it was probably recorded by Philochorus as no more than a reinforcement of the forts securing the island, intended to stem the tide of allied rebellion after Delium and Amphipolis (see Lewis (1992), 427 n. 144; Erxleben (1975), 87: “Viel eher handelt es sich bei diesem Feldzug um Sicherungs‐ und Schutzmaßnahmen”). That these took place habitually in any case is implied by Aristophanes' own phraseology on the lines (Vesp. 715–18) that prompted the scholiast's reference to Philochorus: ἀλλ᾽ ὁπόταν μὲν δείσωσ᾽ αὐτοί, τὴν Εὔβοιαν διδόασιν | ὑμῖν, καὶ σῖτον ὑϕίστανται κατὰ πεντήκοντα μεδίμνους| ποριεῖν. To deny the historicity of Philochorus' report, on the other hand, would be dangerous given the absence of other events from Thucydides' narrative, e.g. the absence of an engagement at Spartolus c.424/3, made evident only in the treaty quoted in Thuc. 5.18.5, which now appears recorded in a cavalry casualty list from that year (Parlama and Stampolidis (2000), 396–9: showing the battle sequence Megara, Tanagra (i.e. Delium), Spartolus) discovered north of Kerameikos in the Athens Metro excavations. This is one more event demonstrating “the likelihood that the narrative [of Thuc.] is not as all‐embracing as it may appear” (Lewis 1992: 380). I am grateful to Prof. E. Badian for personally communicating this insight on the date of the casualty list.

(115) See pp. 119–20 below.

(116) Ober (1985), 137.

(117) Sapouna‐Sakellaraki et al. (2002), 1.

(118) Ibid. 62: “Gesamthaft gesehen ist die auf dem Vrachos von Phylla gefundene Keramik nicht besonders qualitätsvoll, dürfte in einem Militärkamp allerdings auch nicht zu erwarten sein…Das ebenfalls in die spätarchaische Zeit zu datierende Material von der eretrischen Agora, das vielleicht in den dortigen Läden zum Verkauf aufgeboten wurde, ist dagegen reich an schwarzfiguriger Keramik und zeichnet sich auch durch ihre qualitätsvollere Fabrikation aus.” Cf. p. 114: “there is in fact little Attic pottery for an Athenian force, and the low standard of the single letters incised on some cup bases contrasts with the degree of literacy suggested by painted and incised Attic inscriptions of this period…”.

(119) Sapouna‐Sakellaraki et al. (2002), 87 (bronze arrowhead), 114–15. The excavators tentatively suggest Athens' corps of Scythian bowmen.

(120) Schol. vet. Ar. Nub. 213a (ἐπολιόρκησαν αὐτὴν ᾽Αθηναῖοι μετὰ Περικλέους καὶ μάλιστα Χαλκιδέας καὶ Ἐρετριέας): see n. 93 above.

(121) IG I3 1 = GHI 14 (NB the analogous language in [Arist.] Ath. Pol. 55.3: Faraguna (1999), 86); most scholars would restore κλερόχος in the first line: see Cargill (1995), 2 n. 5. The duties of these Athenian settlers are set out as paying taxes and giving military service to Athens (ll. 2–3: [᾽Αθένε]| σι τελε̑ν καὶ στρατ[εύεσθ]αι). Renting of land is restricted, with a tax to be paid to the Athenian treasury (ll. 3–7). The tenant must be capable of providing arms worth 30 dr. for military service (ll. 8–10). An archon of Salamis is charged with overseeing these conditions (ll. 7–8, 11). The resettlement of Lemnos with Athenian cleruchs in 386 seems to look back explicitly to these regulations (IG II2 30 fr. b, l. 34: καθάπ]ερ τοῖς ἐς Σαλαμ[ῖνα; see Stroud (1971), 172–3 who rightly stresses that details of the rules need not be the same, and GHI, 27); see also Faraguna (1999), 84–5, who discusses a new fragment published by Matthaiou that supports the above interpretation.

(122) Meiggs (1972), 121.

(123) See n. 88 above.

(124) See e.g. n. 101 above, also Thuc. 4.28 on the Lemnian and Imbrian forces taken by Cleon to Pylos. It is in this sense that the “cleruchs at Chalcis” are sent (from Athens, as I think) to help the Eretrians: see pp. 93–4 above. See also the example of the Lemnian cavalry, below, p. 109.

(125) See GHI, 27; these still very obscure rules on the alienability of cleruchic land cannot be interpreted automatically as rules prohibiting the alienability of land, as does Salomon (1997), 170.

(126) A synopsis of the following arguments is published in Moreno (2003), 97–106.

(127) Stroud (1998), 80, 109; cf. Harris (1999), 272, who thinks that the dodekate is a transit‐tax (collected in cash) on grain trade passing through the three islands (this is correctly dismissed by Rhodes and Osborne (GHI II, 123), Bresson (2000), 207, Fantasia (2004), 515, and Engels (2000), 114: Harris' transit‐tax would have shifted overseas grain away from Athens rather than towards it).

(128) I follow Badian (1993), 99, in dating the capture of Scyros. On the expulsions, see IACP, 742, 756, 774; Stroud (1998), 31–2 with Cargill (1995), 5–6.

(129) See Stroud (1998), 15–16, with Hansen (1999), 161–77.

(130) The Athenian Aiakeion would have to be converted into a warehouse (ll. 14–16, with Stroud (1998), 97–8), and ten men would have to be elected by vote (and not by lot, as normal magistracies) to supervise an evidently complicated process of weighing, storing, and selling the grain (ll. 36–55 with Stroud (1998), 70–1).

(131) Stroud (1998), 37–8. This is to be distinguished from the πεντεκοστὴ του̑ σίτου collected in cash at Piraeus, as mentioned in Dem. 59.27 (see Stroud (1998), 37 nn. 75–8 for references to scholarship on this tax). Another πεντεκοστή was decreed between 336 and 330 BC to be collected on imports and exports probably at Oropus (called “Nea”), as appears from IG II2 334 and SEG xviii 13 as interpreted by Robert (1960), 189–203. Still another was the πεντεκοστή levied as an ad valorem tax on all merchandise entering or leaving Piraeus, and collected by Agyrrhius and Andocides shortly after the Peloponnesian War: since Andocides claims to have collected more than 36 talents in 401/0, this would indicate that the Piraeus handled more than 1,800 talents worth of merchandise even in the depressed conditions following the war. For the Athenian empire as a revenue‐generating mechanism, see Kallet (2001), 195–205.

(132) The meaning of μερίς as a portion of land, as proposed by Faraguna (1999), 92–3, seems to me very unlikely.

(133) Stroud (1998), 120.

(134) Stroud (1998), 27, 109; cf. Faraguna (1999), 65 n. 2 for reference to the epigraphic attestation of a twelfth at Iasos, and Engels (2000), 106 n. 38.

(135) The Solonic value of 500 medimnoi was one‐twelfth of one talent (6,000 dr.) (see Plut. Sol. 23.3 = Ruschenbusch F77), and one talent (according to the section of Pollux's Onomasticon dealing with the Athenian constitution) was how much a member of the first Solonic class “expended upon the state” (ἀνήλισκον εἰς τὸ δημόσιον): Τιμήματα δ᾽ ἠ̑ν τέτταρα· πεντακοσιομεδίμνων, ἱππέων,ζευγιτω̑ν, θητω̑ν. οἱ μὲν ἐκ του̑ πεντακόσια μέτρα ξηρὰ καὶ ὑγρὰ ποιεῖν κληθέντες· ἀνήλισκον δ᾽ εἰς τὸ δημόσιον τάλαντον· οἱ δὲ τὴν ἱππάδα τελου̑ντες ἐκ μὲν δύνασθαι τρέϕειν ἵππους κεκλη̑σθαι δοκου̑σιν, ἐποίουν δὲ μέτρα τριακόσια, ἀνήλισκον δὲ ἡμιτάλαντον. οἱ δὲ τὸ ζευγήσιον τελου̑ντες ἀπο διακοσίων μέτρων κατελέγοντο, ἀνήλισκον δὲ μνᾶς δέκα· οἱ δὲ τὸ θητικὸν οὐδεμίαν ἀρχὴν ἠ̑ρχον, οὐδὲ ἀνήλισκον οὐδέν (Poll. Onom. 8.129.6–131.1). Böckh brilliantly interpreted this in 1817 to mean that 6,000 medimnoi/drachmas was the taxable capital and the dodekate (= 500 medimnoi) the tax (1828 Eng. trans., vol. 2, 270–1). It has long been held (and tacitly accepted) that this interpretation was “exploded” by Beloch (see Rhodes (1993), 140, citing Hignett), but any reader of Beloch's attack (1885), 245, can safely dismiss the basis of this explosion as a harmless (and, since 1890, obsolete) argumentum ex silentio: “Und überhaupt ist er sehr misslich, eine Einrichtung so verwickelter Natur und so zweifelhaften Nutzens, wie eine Progressivteuer, in die Zeit der Kindheit der Finanzwissenschaft verlegen zu wollen, um so mehr als das ganze spätere Alterthum von einer solchen Steuerform nichts weiss, oder doch wenigstens nicht über die rohesten Ansätze dazu herausgekommen ist”. It should be noted that Ste Croix (2004), in the most thorough modern study of this evidence, actually follows Böckh, despite arguing against him(!), 57–8: “I would now accept the passage in Pollux as probable evidence that the census classes were intended to be used as the basis for assessment for direct taxation, in such a way that the three top classes paid at the same rate (whatever it was on each occasion) but on a notional assessment which was different for each class, so that if the levy was 1%, for example, each Pentakosiomedimnos would pay 60 drachmae, each Hippeus 30 drachmae and each Zeugites 10 drachmae…unlike the complexities which Böckh's theory involves, such a system was an admirably simple one” (emphasis in original). But Ste Croix's is precisely the system that Böckh envisages (see [1817] 1828, 271–2). Ste Croix also sees no major obstacle for the institution of the tax by Solon.

(136) Cf. Stroud (1998), 32: “We do know, however, of rents in kind at this period, and 8% is (suggestively?) a very common rate for rent paid by lessors of public land in Attica”. He pursues this insight no further.

(137) Faraguna (1999), 81–9, presents a series of convincing cases supporting this, and suggests that the Athenian equivalent of the cleruchic tax was the ἐγκτητικόν. The taxable capital of the other classes was a tenth, a fifth, and nothing, respectively: see n. 135 above. The polis' notional title of ownership need not have affected the private ownership or alienability of Attic or cleruchic land any more than the “Radical Title” of the Crown affects land in England today: see the fuller discussion of this Athenian phenomenon on pp. 304–6 below.

(138) Agora XIX L.3, l. 12, with Stroud's commentary (1971), 172: “It appears that the property‐qualifications of the new cleruchs were the subject of this part of the decree”. In his commentary on the grain tax, Stroud (1998), 43 n. 88, again misses the point: “The latest attestation of πεντακοσιομέδιμνοι in an Attic inscription known to me is (significantly?) the decree concerning Lemnos, …387/6 BC”. Since my interpretation of the grain tax proves the presence of πεντακοσιομέδιμνοι on Lemnos, Luria's interpretation (ap. Stroud (1971)) of the word as barring them from the cleruchy (restoring [πλὴν ἱππέων κα]ὶ πεντακοσιομεδίμνων, on analogy with the Brea decree GHI 49, ll. 39–42: ἐς δὲ| [Β]ρέαν ἐχ θετȏν καὶ ζε| [υ]γιτȏν ἰέναι τὸς ἀπο| [ί]κος), is to be decisively rejected. See n. 78 above.

(139) These regulations were inscribed towards the end of the sixth century. See n. 121 above.

(140) For the prices and the evidence, see Pritchett (1956), 186, 196–8, and Appendix 1.

(141) On the switch from 500 medimnoi to 500 dr., see n. 88 above.

(142) See Engels (2000), 108 on a recorded revenue of 3.5 talents from Lemnos and Imbros in 305/4. I use the term “tax haven” quite intentionally: cleruchs were otherwise not liable for payment of Athenian taxes or service on symmories (Dem. 14.16).

(143) Dem. 24.134.

(144) See APF, pp. xxi–xxii on the costs of trierarchies.

(145) I assume the operation of a legal system that somehow ensured that the number of kleroi remained the same over time, perhaps similar to that envisaged by Plato in Leg. 740b–c, addressing the problem of partible inheritance. Alternatively, see Foxhall (2003), 85–6, on ways in which elites circumvent the problems of partible inheritance.

(146) The singular ὁ πριάμενος is used five times (in ll. 11, 18, 22, 27, 30). Stroud is badly misled into thinking of several collectors by the use of the plural twice (ll. 21, 47), and by his erroneous interpretation of the role of the symmories (ll. 31–6: see n. 148 below). The two appearances of ὁ πριάμενος in the plural are most likely an imprecision in the law's drafting, as for example in the sentence: “Presidents of the United States are elected to serve a 4‐year term of office”. Let us remember (Andoc 1.133) that Agyrrhius himself had been the single buyer of the pentekoste (in cash) for thirty talents in 402/1, and that Andocides outbid him the following year (again, individually) by offering thirty‐six talents. In each case, the individual was no doubt the ἀρχώνης of a large group of speculators who remained in the background. The same kind of anti‐competitive bidding practices will of course have been possible with Agyrrhius' Grain‐Tax.

(147) Since the law does not prohibit it, we should assume that the same pair of guarantors, depending on their wealth, could stand as surety for more than one batch of 500 medimnoi.

(148) Cf. Stroud (1998), 65, goes wrong in assuming that a symmory refers to a group of tax collectors: “Agyrrhius appears to provide for joint speculation by six priamenoi possibly on a smaller scale than some of the other emporoi, with shared risks and expenses”. Faraguna (1999), 66 and Engels (2000), 104, 108 follow this mistake. Theirs would be the first and only known instance of an Athenian tax‐collector symmory, whereas taxpayer symmories are well attested: see Hansen (1999), 113: “Symmories were first established in 378/7 in order to systematize the payment of eisphora…”.

(149) See Casson (1954), 183–99.

(150) As Faraguna shows (1999), 77–80, the possibility that there existed a community of (non‐Athenian) Lemnians alongside the Athenians on Lemnos can be discarded at the outset.

(151) Hyp. Lyc. fr. IVb cols. 13–15; see Faraguna (1999), 81.

(152) Ste Croix (2004), 25–6.

(153) See Philippson (1950–9), vol. 4, 53, 221, 224; IACP, 742, 756, 774, gives the size of Scyros as 223 km2, Imbros as 275 km2, Lemnos as 478 km2 (I intentionally choose the lower figures to make my results conservative). Garnsey (1988), 100, is unacceptably pessimistic on the cultivable extent of Lemnos: see Fredrich (1906 a), 242: “So ist Lemnos die flachste und landschaftlich langweiligste, obwohl die grösste der Inseln im thrakischen Meere [Philippson (1950–9), vol. 4, 227, glosses: ‘…auch die fruchtbarste’]”. Fredrich also gives the following valuable description: “Fruchtbarer ist der nicht vulkanische Teil, besonders um die Golfe von Mudros und Purnia und die Ostküste. Hier kann man in Mai und Juni wirklich eine Stunde und länger durch Getreidefelder reiten. Nur als Weide brauchbar ist der ganze Nordwesten und der Phakos. Baüme sind überall auf der Insel selten, und Fruchtbaüme fehlen fast völlig; auch der Weinbau muss gegen das Altertum sehr zurückgegangen sein.”

(154) Foxhall (1997), 130, table 10.2, thinks that the minimum holding for a pentakosiomedimnos would have been “c. 20+ – 34+ ha” for wheat and “c. 17–28 ha” for barley (i.e. 229–389 plethra and 194–320 plethra respectively), but she arrives at this figure using the considerably heavier weights for grain believed before 1998; and assumes too high a range of yields (650–1,000 kg/ha).

(155) Thuc. 3.50.

(156) Dem. 20.31–2. We have every reason to trust Demosthenes' information: see discussion in Chapter 1, p. 32 n. 185.

(157) Compare Alcibiades' 300‐plethra at Erchia in the Attic Mesogaia ([Pl.] Alcib. I 123c); or Lysimachus' 200‐plethra in Euboea (see n. 64 above); on these, see Finley (1973), 58; on Alcibiades, see also APF, 20. See also Ste Croix's classic (1966) study of the estate of Phaenippus: “I suggest that we ought to think in terms of no more than three or four hundred acres [= 1,390–1,850 plethra] at the very most, and probably only a hundred or two [= 460–920 plethra]” (p. 110).

(158) See Leake (1835), 106–11; Fredrich (1906 b), 257–9; Philippson (1950–9), vol. 4, 53–60.

(159) See Foxhall (2003), 81; cf. Braudel (1981), 337.

(160) Given the paucity of our evidence this cannot be more than a guess, grounded in Athenian reality only by the statement of Socrates (in Plato) that a wealthy man in the Greek cities of his day would have “fifty slaves or more” (Resp. 578e). I fancy the input of agricultural labour in an Athenian cleruchy to have been similar to that described in Tolstoy's Anna Karenina (part 3, chapter 4), where Levin joins his forty‐two liberated serfs in harvesting his estate. I thank Oswyn Murray for bringing this passage to my attention.

(161) See R. Barber (1988), 423.

(162) A. H. M. Jones (1957), 77; on evasion, see Stroud (1998), 34 n. 69. I consider the specific proportions of 1/600 for barley and 1/1200 for wheat, used by Garnsey to extrapolate total production form IG II2 1672, as uncertain in the extreme (see Chapter 1, pp. 13–14).

(163) If the grain tax of Lemnos, Imbros, and Scyros was in the vicinity of 300,000 medimnoi, as the evidence shows must be the case, Stroud's identification of the Aiakeion as the Rectangular Peribolos in the Athenian agora is to be rejected. Even the minimum tax of 30,000 medimnoi required by the law almost equals the c.31,000 medimnoi which Stroud admits would hardly fit into this building of 821 m2. (See Stroud (1998), 98: fitting 31,000 medimnoi of grain in the Rectangular Peribolos would require piling it up to a height of c.2 m and leaving no room inside for work or access.) How large would the Aiakeion then need to be? Clearly Athens envisaged that only the net tax would be stored there (obviously, the tax‐collector could not bid to deliver to Athens the gross tax, leaving himself no profit or means to fund his operation). If we therefore imagine that the Aiakeion would fit a net tax of, say, half of 270,000 medimnoi, this would require a building at least four times the size of the Rectangular Peribolos, and about as large as the Square Peristyle on the east side of the agora (dated, however, c.300). While I agree with Stroud that the Aiakeion is somewhere near the agora, perhaps it still lies undiscovered to its north or east. The evidence for a location in Kollytos is an epigraphic restoration, and I would thus prefer it not to dictate where we should look for the Aiakeion.

(164) See discussion in Chapter 5, pp. 242–99.

(165) Ar. Vesp. 715–18.

(166) See n. 114 above.

(167) See n. 74 above.

(168) See Hansen (1999), 112 n. 318.

(169) IG I3 41, ll. 60, 67–76 (446/5 or soon after). The restoration of the fares and routes in IG I3 is as follows (I also include those in ATL III, 301–2, and McGregor (1982), 105, in parentheses): Chalcis to Oropus, 3 obols (2 obols); Oropus to Histiaea (or to Dium?), or vice versa, 7 obols (1 dr.); Chalcis to Histiaea, 4 obols. Note only the Oropus–Histiaea route is price‐controlled in each direction, and also that these transactions, and indeed the entire monetary economy of Euboea under the Athenian Empire, were carried out using Athenian currency: see Kraay (1976), 89.

(170) Pace e.g. McGregor (1982), 109. If the subject were a festival (of which Athens had very many), how could a ferryman normally hope to ascertain that a passenger was in fact going to it? Calculating a smaller stoichos, Hiller von Gärtringen had: ἔσθο τέτταρας ὀβολός. [ἐὰν δὲ πλέοσι| ν]μὲν hοι πομπέυοντε[ς, αὐτὸς δὲ μὲ πομ| π]εύεται, τελέτο τὸ hέ[μισυ. ἐὰν δέτις μ| ὲ]θέλει ἄγεν τὸν π[λέοντα κατὰ τὰ γεγ| ρα]μμένα (see Cary (1925), 246–7). For forms of πομπή see: Thuc. 3.51: ἐκπομπή; 4.108 (of Amphipolis): ἡ πόλις αὐτοῖς ἠ̑ν ὠϕέλιμος ξύλων τε ναυπηγησίμων πομπῃ̑; also Dem. 18.87, 241, 301; 19.123; 23.155; Theopomp. FGrHist 115 F292; Philoch. FGrHist 328 F 162: σιτοπομπία; IG II2 212.15: ἐκπομπὴ του̑ σίτου; GHI II 100, ll. 219–20: ἐμπορία οἰκεία καὶ | [σιτ]οπομπία…

(171) Ziebarth (1929), 123.

(172) Thuc. 2.14. Notice, with HCT II, 48, the use of the aorist διεπέμψαντο, denoting “a single, organized operation, aided perhaps by the state for the transport by sea”, in contrast to the previous imperfects. The evacuation was not, of course, total: see Thuc. 7.27.

(173) See Ziebarth (1929), 123, against the view held e.g. by Jardé (1925: 197 n. 4).

(174) The date is based on the Athenian campaign against Boeotia and Chalcis in Hdt. 5.77, see Petrakos (1997), 489. Given our absolute lack of evidence, the likeliest guess for the date of the first annexation of Oropus by Athens is c.519, i.e. coinciding with the Plataean alliance in Thuc. 3.68 (Beloch (1912), 391 n. 3). Pace ACT I, 279 (“It cannot have been Athenian in 507 because it is not a Kleisthenic deme”), Oropus was, like Salamis and Eleutherae, a Cleisthenic possession not included in the Cleisthenic system (see Rhodes (1993), 253, 773). The acropolis of Classical Oropus seems to have been on Loumperdi hill (80 m. above sea level): see Cosmopoulos (2001), 58; Frazer (1913), 465; with HMGS (1990).

(175) Thuc. 8.60 (Jowett trans.). It was land, as Thucydides says (2.23), ἣν νέμονται Ὠρώπιοι ᾽Αθηναίων ὑπήκοοι.

(176) Thuc. 8.60; see also Lys. 20.6 with Rhodes (1993), 611.

(177) Ober (1985), 115.

(178) See NSID, 209.

(179) Thuc. 7.28 (Jowett trans.) with Ehrenberg (1962), 116; pace Michell (1940), 262; Westlake (1948), 3 n. 1.

(180) Thuc. 7.27, with Westlake (1948), 3–5, and Ehrenberg (1962), 116 n. 3, challenged by Hanson (1995), 153–66.

(181) Pace Westlake (1948), 3–4.

(182) See ibid. 3.

(183) Buchon (1911), 74: “D'Oropos à Eretria, il n'y a qu'une heure de navigation par un bon vent”.

(184) See Noel‐Baker (2000), 45, describing a caique journey (“with good wind in her sails”) from Piraeus to Chalcis in autumn.

(185) Pouilloux (1954), 19–20; Mussche (1961), 178; Semple (1932), 629; Lawrence (1979), 174.

(186) Thuc. 8.60: the Athenian garrison at Oropus (Ἀθηναίων ἐμϕρουρούντων); Thuc. 8.95: the Athenian fort on the territory of Eretria (τὸ τείχισμα τὸ ἐν τῃ̑ Ἐρετρίᾳ) with HCT V, 319–20.

(187) See Thuc. 7.28: τω̑ν τε πάντων ὁμοίως ἐπακτω̑ν ἐδεῖτο ἡπόλις, καὶ ἀντὶ του̑ πόλις εἰ̑ναι ϕρούριον κατέστη.

(188) Thuc. 8.4. The περίπλους is clearly that around Attica from Euboea, not just that around the Sunium promontory: see HCT V, 11 (connecting this with Thuc. 7.28); this is also made clear by the fact that the “intended preparations” detailed in this chapter (παρασκευάζοντο .ὥσπερ διενοήθησαν…) look back to 8.1 to their initial description, among which is:…καὶ τὰ τω̑ν ξυμμάχων ἐς ἀσϕάλειαν ποιεῖσθαι, καὶ μάλιστα τὴν Εὔβοιαν…(“…they would make sure of their allies, and above all of Euboea…”: Jowett).

(189) Mussche (1964), 430; map in Travlos (1988), 408.

(190) Wrede (1933), 10–11, pls. 26–7, with plan in Mussche (1964).

(191) Another wall of the same masonry style stands as a reinforcement of the inner wall of the Poseidon temenos, finished c.440: see n. 192 below.

(192) Wrede (1933), 10–11, 19, pls. 26–7, 45–7, with Mussche (1964), 424, and Travlos (1988), 404.

(193) Ibid. 405.

(194) Blackman (1968), 185, against the Hellenistic date proposed by Kenny (1947), 197 (and followed e.g. by Osborne (1985), 31, and Hornblower (2002), 126). Kenny relied on the fact that “the ship‐sheds are outside the fifth‐century wall”. However, the passage from Demetrius of Callatis, quoted in n. 259 below, shows that Athenian ships at the fort on Atalante were hauled up outside the walls.

(195) HMGS (1988).

(196) See Pouilloux (1954), 58.

(197) Petrakos (1999 a), 47–51, ills. 8–9; map, 16. Frazer (1913), 449, gives a good description of the site.

(198) Petrakos (1999 a), 161–2, ills. 101–2.

(199) Ibid. 26, 79. The key to the dating is the circular tower at the entrance to the citadel, dated by ceramic finds to 480–50, and providing a terminus post quem for an orthogonal tower enclosing it (ibid. 145–8, ills. 92–4). The masonry of the later tower is analogous to that of the south wall of the outer enceinte (ibid. 51–61, ills. 10–17), and of the retaining walls of the sanctuary of Nemesis outside the fort; the in‐fill of the latter, on which the great temple stood, yielded ceramics no later than the mid‐fifth century (ibid. 210–15, ills. 106, 126–8). There is now no reason to follow Pouilloux, Wrede, and Scranton in dating the fort to 412 (see Pouilloux (1954), 43–60).

(200) IG I3 383 (accounts of the Treasurers of the Other Gods, pursuant to the 434/3 decree of Callias, GHI 58A), ll. 106–11, 319, 330–1, with Eliot (1956), 200; Petrakos (1999 a), 79; GHI, 146.

(201) See the map in Travlos (1988), 433.

(202) Mussche (1961), 177–80 (the spring emerges underwater in Port Limani, and Mussche reports its use by local fishermen); see Frazer (1913), 408. Notice the length of the walls is similar to that at Rhamnous, and also to that at Decelea (“the perimeter exceeded 800 m in length and the carefully built rubble was about 2 m thick”: Lawrence (1979), 175).

(203) Mussche (1961), 178, with HMGS (1988); on St Nicholas, the patron saint of sailors, often found taking over havens like Thoricus from pagan predecessors, see Semple (1932), 623–4, 629.

(204) Xen. Hell. 1.2.1: Τῳ̑ δὲ ἄλλῳ ἔτει… ᾽Αθηναῖοι μὲν Θορικὸν ἐτείχισαν.

(205) Mussche (1961), 204; (1964), 431.

(206) Cf. Xen. Hell. 1.1.27 with Thuc. 8 and HCT V, 281–5, 441–2 (referring to the period around 411/10 as “the years about which he [Xenophon] knew least”).

(207) If one follows the chronology set out by Rhodes (1993), 405–11, based on [Arist.] Ath. Pol. 32.1–33.1, then the Four Hundred dismissed the democratic Council on 14 Thargelion (= 9 June) 412/11, and fell by the end of the second month of 411/10 (i.e. late August or early September), so that their rule spanned four (Attic) calendar months: Thargelion, Skirophorion, Hekatombaion, and Metageitnion.

(208) This is probably the same as the stoa ἀλϕιτόπωλις built by Pericles; see [Dem.] 34.37; Ar. Ach. 548 with schol 548a; σιτικὸν ἐμπόριον: [Arist.] Ath. Pol. 51.4. Judeich (1931), 448, locates it on the northern side of the Great Harbour, on the peninsula that separates the latter from the Κωϕὸς Λιμήν. See also Garland (1987), 152–3.

(209) Xen. Hell. 2.3.45. Aristarchus, the only one of these three men who appears in APF, may have been from Decelea (see APF, 48). Did a personal interest in the old grain‐route contribute to his fortification activities in 411? It would also be interesting to know the demotics of the other two.

(210) Thuc. 8.90. My translation follows the comments in HCT V, 306–7.

(211) See Rosivach (2000), 47: “an effort to prevent hoarding by middlemen”.

(212) There is, for example, no suggestion whatsoever that the Four Hundred were planning to “starve” Athens into acquiescence to their rule, as Ferguson (1958: 337) thinks. On Thucydides' antipathy for the Four Hundred, see Westlake (1989), 181–200.

(213) Thuc. 7.19.

(214) Thuc. 8.60.

(215) Thuc. 8.5 (early 412: Euboean envoys to Agis at Decelea); 8.60 (early 411: Eretrian envoys to the Peloponnesian navy at Rhodes).

(216) Thuc. 8.91–2.

(217) Thuc. 8.91 (Jowett trans.).

(218) Thuc. 8.92 (Jowett trans.).

(219) [Lys.] 20.14, 17; see APF, 467–8. The inauguration date (in [Aris.] Ath. Pol. 32) equals 17 June, that of Polystratus' departure 24 June; see n. 207 above.

(220) On the Deigma, see Judeich (1931), 448.

(221) Xen. Hell. 5.1.21–4.

(222) Thuc. 8.92, 94.

(223) Thuc. 8.91.

(224) Mussche (1961), 202–4: “Xénophon…n'envisage que le point de vue du tacticien terrestre”.

(225) Xen. Vect. 4.43–8.

(226) Cf. the report of 20,000 escaped Athenian slaves in Thuc. 7.27, and the similar phenomenon at Chios (Thuc. 8.40).

(227) Thuc. 8.95 (Jowett trans.).

(228) Thuc. 8.96 (Jowett trans.).

(229) Thuc. 8.97.

(230) Thuc. 8.96.

(231) Thanks especially to Badian (1993), 125–62, we have begun to appreciate that a historical context in which rhetoric, dispute, and persuasion counted for so much produced Thucydides the advocate: a writer who may tell the truth and nothing but the truth, but often (and especially in matters of politics) not the whole truth. Needless to say, many historians still refuse to remove Thucydides from the pedestal of historical objectivity.

(232) [Xen.] Ath. Pol. 2.2.

(233) [Xen.] Ath. Pol. 2.3.

(234) [Xen.] Ath. Pol. 2.12.

(235) [Xen.] Ath. Pol. 2.13. Pace Kalinka (1913), 222, who considers this passage a non‐sequitur, or an inept repetition of 2.4 (on ravaging by the thalassocrat, generally), it is indisputable that economic and military considerations are closely inter‐linked in this and previous sections of the text.

(236) Additional examples include: (1) on Cythera from 424 (Thuc. 4.54: taken and garrisoned by Nicias; 118: kept by Athens under truce of 423; 5.14: used with Pylos to ravage Laconia (λῃστευομένης τη̑ς χώρας ἐκ τη̑ς Πύλου καὶ Κυθήρων); 18: required to be given up under Peace of Callias; but, 7.57, evidently still under Athens in 413); and (2) in Laconia opposite Cythera, on an isthmus‐like site (Thuc. 7.26: ἐς τὰ καταντικρὺΚυθήρων τη̑ς Λακωνικη̑ς .ἰσθμω̑δές τι χωρίον), probably on the bay of Boiai (HCT IV, 399–400), from summer to winter of 413/12 (8.4).

(237) Thuc. 2.32.

(238) Thuc. 2.93–4; the site (ἐπὶ δὲ τη̑ς Σαλαμῖνος τὸ ἀκρωτήριον τὸ πρὸς Μέγαρα ὁρω̑ν) was identified on the south shore of the Perama promontory by McLeod (1960), 316–23.

(239) Thuc. 3.51; 4.67, 118; Legon identified the site (ἣκεῖται πρὸ Μεγάρων) as the modern promontory (and former island?) of Teichos; Thucydides implies that it superseded Boudoron as the main Athenian ϕρούριον against Megara.

(240) Thuc. 4.2–6, 8–23, 26–41, 118 (mostly echoed by Diod. 12.61–3). The end of this fortification, after the Spartan victory over its Messenian garrison, is in Diod. 13.64.5–7. Unsurprisingly, Pylos has received the most frequent attention of modern scholars, although (from a topographical point of view) it is the least amenable to a definitive solution, given the complex series of changes to the coastline since antiquity that Pritchett highlighted in (1965), 6–17, and revisited in (1994); see also Wilson (1979), 47–84.

(241) Thuc. 8.38, 40, 55. The end of this fortification is in Xen. Hell. 1.5.15; Diod. 13.76.3–4; the site was published by Boardman (1956), 41–9.

(242) Thuc. 3.51.

(243) Thuc. 2.32.

(244) Diod. 12.44.

(245) See IG I3 41, l. 39: [—.]ν ἐὰμ μὲ λειστȏν—; see Cary (1925), 250: “Fortunately the two main inferences from the inscription, that Histiaea was plagued with privateers and property‐tax, stand beyond the range of doubt”.

(246) Thuc. 2.26; Diod. 12.44 has: τήν τε Εὔβοιαν παραϕυλάττειν καὶ Λοκροῖς πολεμεῖν. Before fortifying Atalante, the general Cleopompus and his thirty ships had ravaged the Locrian countryside and taken hostages from Thronium.

(247) IG I3 40, ll. 76–9: see pp. 100–1 above.

(248) Schol. vet. Ar. Nub. 213a.

(249) Thuc. 2.93. Since the fort was in easy communication with Piraeus by means of fire signals (ϕρυκτοί) (Thuc. 2.94) it does not necessarily follow that “three ships were able to intimidate the Megarians” (McLeod (1960), 323), but the number seems common for forts of this type. Brasidas was nevertheless able to attack this fort by sailing out of Nisaea at night with forty ships.

(250) Thuc. 3.51.

(251) Delphinium became a magnet for slave desertions, like Decelea: see Thuc. 8.40 and cf. 7.27.

(252) See Boardman (1956), 41–5.

(253) See ibid. 42–3.

(254) Thuc. 4.3 with HCT III, 439.

(255) Thuc. 4.4 (Jowett trans.). Diod. 12.61.1 gives twenty days instead of six (.ἐν εἴκοσιν ἡμέραις ἐτείχισε τὴν Πύλον), but is probably wrong: HCT III, 441. The description of such forts conforms perfectly to Polybius' famous summary of Greek practice in fortification: Polyb. 6.42.2–3.

(256) Boardman (1956), 47.

(257) Diod. 13.76.4.

(258) Thuc. 3.89.

(259) Strab. 1.3.20 (citing the extended description of this seismic event by Demetrius in his work on earthquakes). Outside the walls, the ships could be protected by an outer stockade, e.g. at Pylos (Thuc. 4.9 with HCT III, 444; Lawrence (1979), 161); and at Chios (Thuc. 8.40). For the logistics of building these slips and hauling ships onto them, see Coates and Shaw (1993), 87–90. See also n. 261 below.

(260) Sackett et al. (1966), 44–5. The survey also reports slag and signs of local smelting (see 45, 110). This site (as well as the River Boudoros) is misplaced in Fossey and Morin (2000) (Barrington Atlas Map 55): cf. HMGS (1989), and ABSA 61, pl. 8.

(261) The slipways at Sunium (dry length: 18 m) are considerably smaller and steeper than those at Zea in Piraeus (dry length: c.37 m), and show that guard‐ships need not have been full‐sized triremes: see Kenny (1947), 194–6, pl. 31–4; Blackman (1968), 184–5.

(262) See Philippson (1950–9), vol. 1, 579 with HMGS (1989).

(263) See Sackett et al. (1966), 75 n. 123.

(264) Ibid. 42: “…two Archaic‐Classical sites in the small coastal plain of Elliniko;…25 m above the sea…a small terrace is thickly scattered with small fine archaic sherds;…the second site is the lower and southern of two ridges…and there are considerable traces of fortification walls which look earlier than Classical;…enough black and brown glaze of fine thin fabric was found to suggest that the fortifications may date from the sixth century BC.”

(265) Ibid. 42 n. 34: “A hill site c. 220 m above the beach, oval in shape north to south…On top of the hill is the ruin of a Frankish tower and there are walls and much masonry to the south and toward the spring. The scatter of sherds over the whole hill suggests a settlement of almost a square kilometer…Classical: black glaze sherds and tiles, pyramidal loom‐weight.”

(266) Ibid. 45: “…in a little bay with a sandy headland, there are Classical walls, tiles, and sherds (?remains of an Athenian fort)…”; see Philippson (1950–9), vol. 1, 579 (under Port Lutró).

(267) Sackett et al. (1966), 44: “Classical brown and black unglazed tiles and traces of buildings on the summit”; Philippson (1950–9), vol. 1, 579: “antike Trümmer”.

(268) Sackett et al. (1966), 43–4: “…a cliff site at the south end of a sandy beach;…less than a kilometer from the sea there is another hill, Ayios Ilias;…a small enclosed plain is thus formed by the two hills and looks as if it were once a bay, now silted and dried up by a change in the shape of the estuary, which has in fact changed since 1939, when the estuary was too deep to wade even in summer. Classical: many black glaze sherds from cups, skyphoi, kantharoi, kraters, lekythoi. Fine Attic glazed ware including Red Figure fragments, one with part of draped figure and basket, probably Attic import of the third quarter of the fifth century. Also lamp and casserole fragments, tiles, and pyramidal loom weights;…traces of the approach to the town from the estuary and of a wall on the west side of the hill are still visible…The town area must have been considerable since the hill, about 300 × 60 m., was completely inhabited. It seems to have been abandoned before the Roman period.”

(269) Ibid. 45: “…two hills:…a fortified rocky spur overlooking the village…traces of walling preserved between outcrops of rock and rising up to three courses and over 2 m high; the walled summit area is very confined (c. 20 × 25 m); the second site is the adjacent spur;…sherds and tiles from both hills and from the valley slopes between include Classical (black glaze skyphoi fragments), etc…”.

(270) Ibid. 75 n. 123 (there confusingly called “Kastri (Vlakhia)”). This is suggested by the survey as a possible coastal fortified site, but no ancient remains are reported.

(271) Bakhuizen (1976), 49–51, attempts to identify a site on this bay with Elymnion; see also Bakhuizen (1985), 127–8, 140 n. 33. But Sackett et al. report nothing from the site except signs of smelting in Roman or Medieval times (1966), 110.

(272) See Sackett et al. (1966), 75 n. 123: the site is probably near the mouth of the Lamaris river, commanding Paralia Khiliadou. This is suggested by the survey as a possible coastal fortified site, but no ancient remains are reported.

(273) Ibid. 76: “…a much disintegrated wall roughly constructed of the local limestone, running for about 150 m in a north‐west to south‐east direction and ending against the cliffs at north and south. It defended an acropolis of c. 50 × 130 m from the landward side, and measured up to 2.0 m thickness, and stood 1.0–1.50 m high in places. Powell's report of late or post‐Classical graves and pithoi from here suggest that this was a fourth‐century stronghold like Philagra and Kastri…”; Sampson (1981), 54, is attracted to an early Hellenistic date. Khili is a separate site, c.1.5 km northwest from Cape Kyme (photo in ibid. pl. 96): “…strong walling high on the rocky spur overlooking the hamlet of Khili from the south, but this cannot be dated”.

(274) Sackett et al. (1966), 75; Sampson (1981), 54. See below, pp. 137–8.

(275) Sackett et al. (1966), 80, and n. 132: “A strongly fortified hill site…commanding the small beach, which is the only landfall for miles on a forbidding rocky coastline…Sackett found only late Classical material in 1964. These include two narrow skyphos bases with dull black glaze (fourth century BC). The fortification wall, with two towers and other protruding bastions, runs for over 500 m along the southern and eastern sides of the summit, and is preserved in places up to 15 courses and 5.0 m high…”.

(276) Ibid. 81–2; see Thuc. 3.33, Strab. 10.1.7.

(277) Sackett et al. (1966), 38–9: “…a small headland facing north to the Oreoi channel, above 30 m above the sea with a sandy beach on either side. The headland has a good view of the entrance to the Gulf of Volos, to Oreoi, and west to Phthiotis;…a fair‐sized Classical site;…certain Classical pottery, slag, and crucible fragments;…An ancient wall crosses the headland between the beaches and there are traces of walls on the top of the site.”

(278) Ibid. 37: “…a low hill about 60 m above the sea, and about 30 m by 20 m in area at the top. On the lower slopes are Classical to Roman, higher up are Early Helladic to Geometric sherds fairly widely scattered;…Its size and position suggest an important settlement fairly continuously inhabited from Early Helladic onwards.”

(279) Ibid. 37–8: “A fine natural acropolis rises sharply from a narrow valley and slopes gradually from the shore, with traces of a winding road from the beach to the top of the site. There are remains of walls and buildings and the whole area is thickly scattered with sherds from Neolithic to Roman”.

(280) Ibid. 46–7: “The site is a long hill about 50 m high and wide and about 300 m long, at right angles to the shore…Classical sherds, lamp fragments, painted plaster, and tiles are scattered on the top of the hill…”.

(281) See p. 131 above.

(282) Sackett et al. (1966), 53: “A low mound c. 200 m in from the sea;…Classical and later material, including a number of inscribed tombstones;…The sherd scatter continues in the adjacent fields for over 50 m…”.

(283) See Bakhuizen (1985), passim.

(284) See Sackett et al. (1966), 62–3; Sapouna‐Sakellaraki (1995).

(285) Sackett et al. (1966), 65 and n. 92.

(286) Ibid. 69: “Many Classical, Hellenistic, and Roman finds…lekythoi, rhytons, lamps, figurines, loom‐weights, glass bottles…”; IG XII.9.90–123, 190.

(287) Sackett et al. (1966), 78–9: “…a small hill with good anchorage;…remains of a harbor mole;…the sherds and tiles so far discovered on the slopes are of Classical date. To the south and east of the hill three areas have been noted by Hankey as having similar remains, with glazed tiles and Classical sherds. At 3 which is a higher ridge, there are also the remains of a Classical‐Hellenistic building c. 9 m square with masonry preserved three courses high in a style like that at Platanistos, and at 5 also traces of walling. These may have formed part of the harbor town for the cipollino quarries on the hill of Ayios Nikolaos above Styra.”

(288) See ibid. 80.

(289) See Thuc. 2.32, 3.89.

(290) See Dem. 19.219, 326; 18.71; 8.36 (respectively) for the use of these terms in connection with Euboea.

(291) See Thuc. 3.17: the Athenian fleet at its height consisted of “a hundred ships which guarded Attica, Euboea, and Salamis, and another hundred which were cruising off Peloponnesus, not including the ships employed in blockading Potidaea and at other places; so that in one and the same summer their fleet numbered two hundred and fifty”. Steup's criticism of this passage as a late interpolation is rightly rejected by Gomme (HCT II, 272–7) and Hornblower (2002), 400–1. The number of ships was drastically reduced by 411: on the revolt of the island from Athens, Thucydides (8.95) says that the number of ships sent to Euboea, added to the number already there, totalled thirty‐six (see also Thuc. 8.74, 86).

(292) Thuc. 2.26.

(293) Thuc. 2.14: χρήμασι μὲν οὐ̑ν οὓτως ἐθάρσυνεν αὐτοὺς, ὁπλίτας δὲ τρισχιλίους καὶ μυρίους εἰ̑ναι ἄνευ τω̑ν ἐν τοῖς ϕρουρίοις καὶ τω̑ν παρ᾽ ἔπαλξιν ἑξακισχιλίων καὶ μυρίων. On this passage, see HCT II, 34–9; ACT I, 255–7. Ober (1985), 193 n. 7, rightly takes Thucydides to refer to forts both inside and outside of Attica. Attempts to explain away the manpower figures (e.g. Beloch's (1886), 66, emendation of 16,000 to 6,000 reservists), like the numbers of ships above (n. 291), fail to grapple with the sheer defensiveness of Periclean strategy as far as Euboea was concerned: Athenian interests in the island meant that the revolt of 446 could not be allowed to happen again.

(294) ACT I, 501.

(295) Thuc. 3.92; see also Thuc. 3.93: “While the new colonists were collecting at Heraclea, the Athenians grew alarmed; the scheme appeared to be aimed at Euboea, for Cape Cenaeum on the opposite coast is within a short sail” (Jowett).

(296) Thuc. 8.95. The idea that this fortification may date from as late as 413 or 412–11, put forward by Brunt (1966), 88, is unconvincing. The sorry state of Athenian finances after the Syracusan disaster meant that Athenian plans (mentioned by Thuc. 8.1) to ensure the obedience of Euboea above all their other allies never amounted to any major reinforcements: see n. 291. The fort was clearly not in the asty of Eretria, as shown by Steup (see HCT V, 319–20). Cf. ATL III, 295 and n. 100.

(297) GHI 84, ll. 16–18 (3,740 dr., 1.25 obols on the sixth prytany of 410/9 to the general Eukleides at Eretria). See GHI, 259, ATL III, 365, and HCT V, 319–20.

(298) Coulton in Sapouna‐Sakellaraki et al. (2002), 40–3, and generally 111–16.

(299) Hdt. 5.77, 6.100; Sapouna‐Sakellaraki et al. (2002), 113–15.

(300) See pp. 101–2 above.

(301) Sackett et al. (1966), 39–40 n. 29 (Classical finds): “The mound is about 30 m high and about 100 m by 135 m along the sides. It is ringed by a fortification, Byzantine and later, using Classical blocks (relaid in mortar) as a base. On the south side are remains of two rectangular towers, the better preserved about 4.0 m high, jutting 4.5 m from the wall…The size and position of Oreoi and its range of remains…point to a large settlement from early times.”

(302) Ibid. 54–6 n. 61: Ay. Ilias, “…the most prominent hill in the low‐lying plain of Psachna;…it holds a commanding view over the whole plain to the west and south as far as Politika and Chalcis, and to Mikrikapa, Krases, Steni, and Mt. Dhirfys to the east. The chapel of Ay. Ilias on the summit is partly founded on ancient blocks. These may perhaps be connected with the Ionic temple reported by Bursian. Sherds were found scattered over the entire western and southern slopes of the hill…dated Middle Helladic, Late Helladic, Geometric, and Classical. This was a strong and central site of some importance in its area, easily fortified at the summit”; Pirgos: “…a low mound marked by the remains of a medieval tower. Sherds of the Early Helladic, Middle Helladic, Classical, and Medieval periods have been found here scattered widely over the fields for more than 300 meters to the south and 100 meters to the west”; Ay. Paraskevi (“Pyrgos ruin” = ancient Argura?): “…scattered sherds and tiles of the Geometric(?), Classical, and Hellenistic periods; also traces of ancient walling at the northern limit of the ploughland…”.

(303) Sampson (1981), 29–31 (for the date see esp. 30 n. 40), figs. 54–73, with maps.

(304) Ibid. 18–22. From pottery and coin finds it is evident that the fort continued in use into the Hellenistic period.

(305) Sackett et al. (1966), 45–6: “A hill site defending the natural route along the Cereus valley, c. 2 km north of the village and immediately west of the road…There are traces of walls on the hill and good black glaze sherds, indicating a substantial settlement in the fifth century: fragments from kylix, olpe, cups and bowls (one ribbed), lekanis, kraters, amphorae, and tiles; also fifth‐century lamps and a third‐century acroceraunus.” On 26 March 2000, I observed scattered surface fragments of Attic Red Figure from the southern slope of this site, which is almost inaccessible through thick woods and brush.

(306) Ibid. 46: “A fortified site…on a rock outcrop below the sheer cliff of Misipetri (525 m). The circuit wall can be seen from the top of Misipetri but is hard to trace on the site because of thick scrub. Inside the walls Classical sherds and tiles and remains of buildings were found in 1964.”

(307) Ibid. 46: “…a fortified hill‐site c. 1 km west of the village on a ridge which runs south. Classical walls on the east and west sides of this ridge and Early Helladic, Classical, Roman, and Medieval sherds suggest that the site was intermittently used.”

(308) Ibid. 46 n. 40: “…a low hill (Kastro), where one course of large stones running round the circular crest appears to be the remains of a fortification wall. Other masonry and architectural remains (which have since disappeared) were noted at the road cutting in 1939. One stone, still remaining at the roadside, has the inscription ΕΠΙΔΟΣΙΣ. Sherds include ?Protogeometric, Geometric, and Classical.”

(309) Ibid. 51: “…?Geometric, Archaic, and Classical sherds, similar to those from Dhafni Kastro…”.

(310) Ibid. 49. Cf. n. 271 above for an alternative identification of Elymnion.

(311) Sackett et al. (1966) 50 n. 51: “…black glaze and tiles…”.

(312) Ibid. 49 n. 52: “…glazed tiles and sherds similar to those from Prokopion Kastri. Also loom‐weights…”.

(313) Ibid. 49: “…indeterminate prehistoric, Classical, and Roman sherds…”.

(314) Ibid. 48–9: “…traces of a wall follow the contours of the hill with conglomerate blocks here and there;…slopes are strewn with sherds;…A few tiles were found on top of the hill;…On the east slope lower down are later sherds, Classical to Roman, and remains of walls with blocks in position.… The quality and range of finds point to a settlement on a fairly large scale from Early Helladic to Late Helladic and again from Classical to Roman.”

(315) IG I3 41, ll. 101–2; see Erxleben (1975), 90.

(316) See n. 273 above.

(317) Sackett et al. (1966), p. 69: “Many Classical, Hellenistic, and Roman finds…”.

(318) Ibid. 75 and n. 120. See in detail pp. 137–8 above.

(319) Ibid. 71–3: Palaiokastri, “a high and imposing conical hill, with remains of a fourth‐century (or Hellenistic) fortification much destroyed and in most places little more than a raised ridge in the fields. The towers reported in 1902 are no longer to be seen”; Itea (fourth‐century temple of Apollo at Tamynai?); see also Sampson (1981), 52–3.

(320) Sackett et al. (1966), 71.

(321) Sapouna‐Sakellaraki et al. (2002), 114. Cf. Ober (1985), 195: “It was only after the [Peloponnesian] war that new economic and military conditions and the growth of the defensive mentality favored the development of a strategy of preclusive defense based on a system of border fortifications”.

(322) See Wiegand (1899), 458–67.

(323) Young (1956), 141–2; cf. Osborne (1985), 31–4.

(324) Contrast Garnsey (1988), 132: “The people received their share of the revenues of empire in the form of cash, not grain”.

(325) See Andrewes (1982 a), 372–3.

(326) See ibid. 373–4.

(327) See Andrewes (1982 b), 403–5.

(328) See Ehrenberg (1946), 142–3; the passage most clearly illustrating this perception is Hdt. 6.139–40; more on this in Chapter 5, pp. 304–6.

(329) Hdt. 6.132. The passage significantly echoes Aristagoras' promise to Artaphernes to help him win the Cyclades and eventually Euboea: Hdt. 5.31; see further Ehrenberg (1946), 137–42; Byron, Don Juan, canto III.86 (The Isles of Greece 12): “The tyrant of the Chersonese / Was freedom's best and bravest friend / That tyrant was Miltiades! / Oh! that the present hour would lend / Another despot of the kind! / Such chains as his were sure to bind.”

(330) Hdt. 6.135–6.

(331) Cf. Humphreys (1978), 167–8; Murray (1993), 50–1.

(332) Thuc. 1.4; 1.98; Plut. Cim. 8; Kallet (2001), 25–6.

(333) Thuc. 1.8 (Jowett trans.); Minos' colonization “of most of the Cyclades” appears in Thuc. 1.4; cf. [Xen.] Ath. Pol. 2.

(334) von Reden (1995), 129.

(335) See n. 91 above with Plut. Per. 11; note also the contiguous placement of the statue of Pericles and Phidias' Athena Lemnia on the acropolis (Paus. 1.28.2).

(336) Melos and Alcibiades: Thuc. 5.84, 116 with Meiggs (1972), 345; on Sicily, see Kallet (2001), 25 n. 18 (who follows Avery (1973), 8–13), stressing Thucydides' presentation of the colonial history of Sicily (Book 6.1–5) in connection with the sending of the cleruchy to Melos: “Thucydides implicitly presents the expedition to Sicily as a colonizing venture…”.

(337) See Theopomp. FGrHist 115 F389–90; Diod. 12.34.4–5; Plut. Per. 19.1–2. Most of these places are archaeologically unknown. The recent work of Isaac on the Thracian Chersonesus (1986), 159–97, is unique in coordinating the history of this important Athenian possession with the meagre archaeological work so far carried out.