Rivers as Weapons in Ancient War: Babylon, Mantinea and Kirrha

A city wall can be battered, mined, or starved into surrender. A river cannot be argued with. Ancient commanders knew that many cities depended on a stream that ran beside or through them, for drinking water, for defense, or for trade, and a handful of Greek and Roman writers preserved stories of generals who turned that dependence against the people inside. Some of these stories rest on near-contemporary testimony. Others were collected centuries later, attached to famous names, or borrowed from myth.

This article looks at the main cases and asks how much weight the evidence for each can bear, from the punished river of Herodotus to a Spartan siege recorded by a man who was alive at the time.

A king who punished a river

Herodotus, writing in the fifth century BCE, opens his account of Cyrus the Great’s march on Babylon with a detour. At the river Gyndes (identified by his translator A. D. Godley as the modern Diyala, a tributary of the Tigris), one of the king’s sacred white horses tried to swim across and was swept away. Herodotus says Cyrus was so angry that he threatened the river “that he would make it so weak that women should ever after cross it easily without wetting their knees.” He then halted the campaign, marked out 180 channels on each bank, and set the army digging. The work took the whole summer, and only the next spring, “when Cyrus had punished the Gyndes by parting it among the three hundred and sixty canals,” did he move on.¹

As military history this is hard to credit. A field army does not usually spend a campaigning season avenging a horse, and Herodotus offers no witness for the episode. Yet it carries a practical idea that Herodotus returns to a few chapters later: a river can be divided, weakened, and made fordable by digging.

Babylon and the road along the riverbed

Herodotus prepares the ground for his Babylon story with a Babylonian queen he calls Nitocris. He says she made the Euphrates wind so that the approach from Media would be slow, dug a vast lake basin upstream, and then, to build a bridge in the city, turned the whole river into that basin until its old bed lay dry.¹ Herodotus gives no source for these works, which are best read as Greek tradition about Babylon rather than as an engineering record.

When Cyrus reached Babylon, Herodotus continues, the Babylonians lost a battle outside the walls, withdrew, and sat behind stores laid in for many years. Cyrus was stuck. Then came the solution, introduced with a historian’s caution: “Whether, then, someone advised him in his difficulty, or he perceived for himself what to do, I know not.” Cyrus posted troops where the river entered and left the city, drew the Euphrates off into the old basin, and when the water had dropped to “the height of the middle of a man’s thigh,” his men waded in along the channel. The city was so large, Herodotus was told, that people in the center went on celebrating a festival while the outer districts fell.¹

Xenophon’s Cyropaedia, written in the fourth century BCE, tells a fuller version. Cyrus digs a great trench around the city and builds towers on palm-trunk foundations, a show of preparation for a long blockade that the defenders, with supplies for more than twenty years, find amusing. On a festival night the trench mouths are opened, the water pours in, and “the river-bed formed a highway into the heart of the town.” Xenophon gives Cyrus a speech that begins, “My friends, the river has stepped aside for us.”² The Cyropaedia is a work of political and moral instruction built around an idealized Cyrus, not a history in the modern sense, so its agreement with Herodotus shows that the river story circulated among Greeks; it does not independently confirm it.

What the Babylonian sources say

Babylonian documents tell a different story, or at least a much shorter one. The Nabonidus Chronicle, a cuneiform record of the reign of the last native king of Babylon (556 to 539 BCE), reports that Cyrus fought the Babylonian army at Opis on the Tigris, that “Sippar was captured without a battle,” and that days later Ugbaru, governor of Gutium, and the army of Cyrus entered Babylon “without battle.” Cyrus himself arrived a little over two weeks later, and the chronicle stresses that the temple rites were not interrupted.³ The Cyrus Cylinder, a royal inscription composed for the new ruler, says that the god Marduk “made him enter his city Babylon without fighting or battle.”⁴

Neither text mentions a siege of years, a diversion, or soldiers in the riverbed. Both are interested documents: the Cylinder in particular presents Cyrus as Marduk’s chosen king, welcomed by the city. But the chronicle is a dry year-by-year record, and its account of a quick capture after a defeat in the field is hard to square with Herodotus’s long siege.

Modern scholars have taken this in different directions. A 2024 student paper in the Journal of Military and Strategic Studies summarizes the range: Paul-Alain Beaulieu treats Herodotus’s story as popular myth, Muhammad Dandamayev notes that there seems to be no direct archaeological evidence for a siege, and Pierre Briant finds a voluntary surrender of so well fortified a city hard to believe. The paper’s author argues that the two accounts can be reconciled, with the basin of Herodotus placed near Sippar, and that a covert entry along a drained channel is consistent with a capture “without battle,” since Babylonian chronicles record outcomes rather than methods.⁵ That is one proposal, not a consensus. What can be said firmly is that the river stratagem at Babylon is attested only by Greek writers who were not there, and that the Babylonian record does not contradict a peaceful entry but gives no sign of a drained Euphrates either.

Mantinea, 385 BCE: a dam against mud brick

The best-documented case comes from mainland Greece. Xenophon, a contemporary of the events, describes in his Hellenica how the Spartans ordered the Arcadian city of Mantinea to pull down its walls. When the Mantineans refused, King Agesipolis led an army against them. He ravaged their land and then surrounded the city with a trench and a wall, but he learned that the Mantineans had a large store of grain from the previous year’s harvest. Rather than wear out Sparta and its allies in a long blockade, he dammed the river that flowed through the town at the point where it left the walls.⁶

The water backed up above the foundations of houses and walls. Xenophon describes the lower courses of brick becoming saturated and failing to hold the rows above, the wall cracking, and the defenders trying to prop it with timber before they gave in. Sparta then imposed harsher terms: the walls came down and Mantinea was broken up into four separate villages. Xenophon ends the episode with a dry moral, that mankind had learned one lesson, “not to conduct a river through a fortress town.”⁶

Pausanias, a travel writer of the second century CE who described Mantinea and its monuments, names the river as the Ophis and explains why the trick worked. The walls were of unburnt brick, which he says stands up to siege engines better than stone, “but it crumbles under the action of water just as wax is melted by the sun.” He adds that the idea was not new: “Agesipolis only copied an established custom, and one celebrated among the Greeks,” first used, he says, by the Athenian general Cimon against the Persians holding Eion on the river Strymon.⁷

Here the evidence is about as good as ancient military history allows. Xenophon was a contemporary, Pausanias wrote centuries later with a knowledge of the ground and adds the name of the river and the building material, and the method they describe fits that material. The Mantinea siege shows a river used not to slip soldiers inside or to cut off water, but as a solvent applied directly to the enemy’s defenses.

Kirrha: diversion and poison

A darker story concerns Kirrha, which Pausanias identifies with the Crisa of Homer, the town below Delphi that Greek tradition says was destroyed in the First Sacred War of the early sixth century BCE. Pausanias reports that the Amphictyons, the league that oversaw the sanctuary, made war on Kirrha for seizing land belonging to Apollo, with Cleisthenes, tyrant of Sicyon, in command and the Athenian Solon as adviser. In his account Solon first diverted the river Pleistus, which fed the town by a channel. When the defenders held out on well water and rainwater, he steeped hellebore roots in the river and turned it back into its course. The defenders drank, fell ill with persistent diarrhea, abandoned the walls, and the city was taken.⁸

The Roman writer Frontinus, in his first-century CE handbook of stratagems, tells the same trick but gives it to Cleisthenes, who “cut the water-pipes leading into the town of the Crisaeans” and then restored the supply laced with hellebore. His translator notes that a third writer, Polyaenus, credits it to someone else again, a man named Eurylochus.⁹

Three authors, three different inventors. That alone suggests a story that floated free of any firm record. The historical basis of the war itself has also been doubted. Noel Robertson’s 1978 article in The Classical Quarterly, pointedly titled “The Myth of the First Sacred War,” observes how much detail the sources supply for a conflict of around 590 BCE and puts that tradition under scrutiny.¹⁰ Kirrha is best treated as a case of how later Greeks imagined river warfare, including the deliberate fouling of drinking water, rather than as a securely documented event.

Roman handbooks and a legendary labor

Frontinus gathered his water stratagems in a single short chapter, “On Diverting Streams and Contaminating Waters.” Alongside Kirrha it includes Publius Servilius cutting off the stream that supplied Isaura, Julius Caesar diverting the springs of a besieged Gallic town, and a Lucius Metellus in Hither Spain who “diverted the course of a river and directed it from a higher level against the camp of the enemy, which was located on low ground,” then ambushed the men who fled the flood. Bennett, his translator, identifies this commander as Quintus Caecilius Metellus Macedonicus, campaigning around 143 to 142 BCE, and notes that the better manuscripts give the first name as Lucius.⁹

The same chapter shows the limits of such collections. Frontinus credits the draining of the Euphrates at Babylon to Alexander the Great, although, as Bennett points out, Herodotus, Xenophon and Polyaenus all give it to Cyrus. He then adds that the legendary queen Semiramis “is said to have done the same thing.”⁹ A handbook of examples rewards a good story more than a checked one, and a successful stratagem tends to collect famous names.

Greek myth had its own river engineer. In Apollodorus’s handbook of myths, the fifth labor of Hercules was to clear the dung of King Augeas’s cattle in a single day. Hercules breached the foundations of the yard and, “diverting the courses of the Alpheus and Peneus, which flowed near each other, he turned them into the yard,” flushing it out through another opening.¹¹ It is not a war story and not history, but it shows that the idea of redirecting a river to do in hours what human labor could not was familiar enough to become a hero’s feat.

Read together, these sources sort into three groups. Mantinea is history, well attested and physically plausible. Babylon is a contested Greek tradition set against Babylonian records that describe a quick and bloodless entry. Kirrha, Semiramis and Augeas belong to legend or to stories that cannot now be traced to a reliable witness. What runs through all of them is a recognition that a city’s water, whether moat, supply, or foundation, was also one of its weak points.

Footnotes

1. Herodotus 1.185-191, trans. Godley (1920), with Godley’s note on the Gyndes. ↩ ↩² ↩³

2. Xenophon, Cyropaedia 7.5.7-20, trans. Dakyns. ↩

3. Nabonidus Chronicle (ABC 7) iii.12-22, trans. Grayson (Livius). ↩

4. Cyrus Cylinder, line 17 (Livius translation after Cogan and Schaudig). ↩

5. Mehraban 2024, JMSS 23(4), introduction and conclusion. ↩

6. Xenophon, Hellenica 5.2.1-7, trans. Dakyns. ↩ ↩²

7. Pausanias 8.8.7-9, trans. Jones (1933). ↩

8. Pausanias 10.37.5-8, trans. Jones. ↩

9. Frontinus, Stratagems 3.7.1-6, trans. Bennett (1925), with notes 35-37. ↩ ↩² ↩³

10. Robertson 1978, CQ 28(1), 38-73 (title and abstract only). ↩

11. Apollodorus, Library 2.5.5, trans. Frazer. ↩