Equations are not being displayed properly on some articles. We hope to have this fixed soon. Our apologies.

Paradeisopoulos, I. (2016). Changing names, fitting views: The conversion of river Arsanias into the Euphrates of Xenophon. PHILICA.COM Article number 582.

ISSN 1751-3030  
Log in  
  998 Articles and Observations available | Content last updated 20 February, 17:52  
Philica entries accessed 2 795 485 times  

NEWS: The SOAP Project, in collaboration with CERN, are conducting a survey on open-access publishing. Please take a moment to give them your views

Submit an Article or Observation

We aim to suit all browsers, but recommend Firefox particularly:

Changing names, fitting views: The conversion of river Arsanias into the Euphrates of Xenophon

Iordanis K. Paradeisopoulosunconfirmed user (University of Patras)

Published in histo.philica.com

In ancient Greek, Roman, Armenian, Arabic, Byzantine, and Ottoman texts, sources of the Euphrates are those of the river Frat (Karasu, currently West Euphrates) which rises to the north of Erzurum. But in the literature on the Anabasis, usually as the Euphrates of Xenophon, crossed by the Ten Thousand not far from its sources, is taken the Arsanias (Murad, currently East Euphrates). We examine whether this identification rests on evidence.
It seems that in the texts examined there is no reference to the Arsanias as Euphrates. Moreover, it seems that the wrong interpretation of a corrupted passage in the Geography of Ptolemy had resulted in the omission of this southeastern tributary of the Euphrates from the European maps of the early modern era. The omission lasted until the eighteenth century, when the Arsanias appears at last on the maps, now as the Euphrates of Xenophon, simultaneously with attempts for the identification of the route of the retreat of the Ten Thousand.

Article body

The Euphrates and its tributary before Xenophon’s era[1]

Euphrates (Εὐφράτης) is the traditional name of the river known as Perath in Hebrew, Purat in Assyrian, and Ufratush in Persian cuneiform. From this last name derives probably the Greek Euphrates, known as Al-Furat in Arabic, Yeprat in Armenian, and Fırat or Frat in Turkish.[2] Its southeastern tributary -currently known as the Eastern Euphrates or Murad çay - was known at least since the Neo-Assyrian Empire as Arsania.[3]

Xenophon and the Greeks in the army of Cyrus the Younger met for the first time the Euphrates during their anabasis, on their way from Sardes to Babylonia and the battle at Cunaxa. They arrived at the river at Thapsacus (Anab. 1.4.11); they crossed to its eastern bank (1.4.16), and marched along it to its confluence with the river Khabour (1.4.19); to the deserted town Corsote (1.5.4); to Pylae (1.5.5); opposite to the city of Charmande (1.5.10); and to Babylonia (1.7.1). The death of Cyrus in the battle of Cunaxa (1.8.1-29) started their retreat. After their conflict with the Persians of Tissaphernes, their plan was to cross the big rivers near their sources because ‘as you approach there, they become passable, without even wetting your knees, even though they are impassable at a distance.’ (3.2.22) Indeed the Ten Thousand crossed the Euphrates in the retreat ‘wetting up to the waist’; they were told that its sources were not too far from there (4.5.2).

It is rational to assume that Xenophon uses for the river the name Euphrates according to the in­for­mation he had.[4] However, as still there is no consensus[5] on the solution to the complex puzzle of the route, the parasangs, and the chronology of the retreat of the Ten Thousand, views that Xenophon was mistaken are not infrequent,[6] arguing that we cannot easily transfer on modern rivers the river names offered by Xenophon; or that the concerned region of the upper Euphrates was completely unknown to Xenophon; and, additionally, that we do not know what information he obtained on the field and what was still present later, when he wrote down his campaign memories.

At Xenophon’s time, the only written informa­tion available to him was that of Herodotus but it does not offer clues. Before the entry of the Euphrates into Mesopotamia, Herodotus mentions only that the Euphrates flows from the land of the Armenians and empties into the Erythraean Sea (1.180).[7] Thus Xenophon re­lied upon oral information concerning the name of the river he calls Eu­phrates in the retreat. Did he mean the ‘West Euphrates’ (Frat, Karasu) or the ‘East Euphrates’ (Arsanias, Murad)? The sources of these rivers (i.e. the probable area of the crossing of the Ten Thousand) are at a distance. Therefore, if Xenophon is taken to mean that he had crossed the Arsanias (Murad) as ‘Euphrates’, either in the vicinity of its sources[8] or a few hundred kilometres below (to the west),[9] then evidence must be produced that this river was ever known as ‘Euphrates’; moreover because the north-western tributary (Karasu) still bears this name (Frat).

The Euphrates and its tributary in the ancient sources after Xenophon

In the ancient sources there is a distinction between the Euphrates (Frat, later Karasu) and the Arsanias (later Murad). According to Strabo (11.12.3), the Euphrates, with its sources in the northerly region of the Taurus, flows west through Greater Armenia, to Lesser Ar­menia, having the latter on its right and Acilisene [Ἀκιλησηνή, i.e. the area of Erzincan] on the left. This means that his Euphrates is the Frat/Karasu. A second description in Strabo (11.14.2) could be misinterpreted, and it did in the eighteenth century: “both the Euphrates and the Araxes flow from Mt. Abus, the former towards the west and the latter towards the east.” Mt. Abus extends from the neighborhood of Erzurum in the west towards the area to the south of Artaxata in the east.[10] Pliny (see below) says that the sources of the Araxes and the Karasu are close. Probably this is also what Strabo means here, i.e. that his Euphrates is again the Frat (Karasu). Others understood that the Euphrates rises from the same mountain range but from its other end to the southeast; hence they interpreted here that Strabo’s Euphrates is the Murad.[11] However, once again, describing Pontos and Paphla­gonia, Strabo (12.3.8) refers clearly to the Karasu as the Euphrates: King Mithridates seized a well-watered mountain near Da­steira[12] in Acilisene; nearby, also, was the Eu­phrates which separates Acilisene from Lesser Armenia.

According to Pliny (HN 5.83), the Euphrates is the Karasu because it “rises in Caranitis, a prefecture of Armenia Major, as two of those who have seen it most recently have recorded. Domitius Corbulo places it on Mount Aga and Licinius Mucianus at the lower reaches of a mountain that they call Capotes, twelve miles above Zimara; at its origins it is known as the Pyxurates.”[13] Pliny (HN 6.26) speaks of the neighbouring sources of the Araxes and the Euphrates, i.e. the Karasu: ‘The Araxes rises in the same mountains as the Euphrates, at a distance from it of six miles only’. Solinus agrees, without specifying a distance: ‘But the Araxes takes its head at a short distance from the Euphrates and then is carried into the Caspian Sea.’[14] Strabo said earlier that on the same mountain as the Euphrates rises also the Araxes: ‘Above Mount Masius far to the east along Gordyene is the Niphates, then the Abus, from which flow both the Euphrates and the Araxes, the former to the west, the latter to the east.’[15] This is why Stephanus Byzantius calls Araxes ‘the brother of the Euphrates’,[16] i.e. the brother of the Frat/Karasu.

Pliny (HN 5.83–84) says also that the Euphrates (Karasu), which rises in Caranitis[17] in Greater Armenia, receives in its course the rivers Lycus, Arsanias [i.e. the Murad, currently East Euphrates], and Arsanus. In a third passage (HN 6.128) Pliny makes again a clear distinction between the Euphrates (Frat, Karasu) and the Arsanias (Murad): [The water] of the Arsanias, being lighter, floats on the surface of the Tigris for a distance of nearly four miles, after which they separate, and the Arsanias flows into the Euphrates.

Plutarch (Luc. 31) says that Tigranes on the fourth day encamped over against the Romans, keeping the river Arsanias between himself and them. Obviously Arsanias here is the Murad. Tigranes had encamped to the south of the river, towards his capital at Tigranocerta.

Plutarch (Pomp. 32) says also that Pompey overtook [Mithridates] near the Euphrates river, and en­camped close by; and fearing lest the king should get the advantage of him by crossing the Euphrates, he put his army in battle array and led it against him at midnight. Euphrates here is the Karasu, because the battles between Pompey and Mithridates, the king of Pontos, were fought in the kingdom of Mithridates.

Appian (Mithr. 101) testifies also to the escape of Mithridates through the headwaters of the Euphrates (Frat, Karasu) towards the river Apsarus (Çoruh) and Colchis.

According to Tacitus (Annales 15.15), [the Roman] Pætus threw a bridge over the river Arsanias (Murad), which flowed by the camp, apparently with the view of facilitating his march.

The Euphrates in Ptolemy

Ptolemy describes both rivers, but also applies the name Euphrates to the former, to the river from Erzurum.[18] However, until recently Ptolemy’s (5.13.7) incomplete text concerning the Arsanias was taken traditionally to mean not a river but mountains [ὄρη]. Maps 1a and 1b are copies of the pages which illustrate the course of the Euphrates in a fifteenth-century Latin manuscript of Ptolemy’s Geographia.[19] Euphrates is the Karasu (Frat) rising near Erzurum. It bends to the south towards Syria, receiving from the west the Melas (Μέλας, Black) of the ancients,[20] the river Tohma.[21] There is no Arsanias (Murad) in the Maps 1a and 1b because, according to the wrong interpretation of Ptolemy’s 5.13.7, the eastern tributary is not mentioned in his text.


Map 1a: The Euphrates in a manuscript of Ptolemy’s Geographia

(on line at Bibliothèque-médiathèque de Nancy, France)


Map 1b: The Euphrates in a manuscript of Ptolemy’s Geographia

(on line at Bibliothèque-médiathèque de Nancy, France)

The tradition of interpreting Ptolemy 5.13.7 as referring to mountains (ρη) instead of referring to the river Arsanias (Murad) continued for centuries. According to Nobbe 1845 ‘there are also other important mountains.’[22] Later, the only complete up to now English translation of 1932[23] takes this passage as referring to a river: ‘There is another noted river which empties into the Euphrates.’ Recently a new edition of Ptolemy’s text was based on the Codex Seragliensis found in 1927, i.e. after the edition of Nobbe. The Greek text here differs from Nobbe’s edition in more than 1,000 passages.[24] This time there are no mountains and Ptolemy reads:

στι δ κα τρα ξιολογωτρα π το Εφράτου ποταμο κτροπ, ς τ μν συνάπτον τ Εφράτ ποταμ πρας πχει μορας 71°30' 40°30' τ δ κατ τς πηγς πρας 77° 41°.[25]

Thus the geographical coordinates of the joining point and of the sources were assigned to the ‘other more notable divergence’ (τρα ξιολογωτρα κτροπ) from the Euphrates, i.e. to the Arsanias. Ptolemy had not omitted the tributary. Almost three hundred years earlier, the eighteenth-century cartographers were convinced that this was meant by Ptolemy. Probably some of them had read in Oriental sources the details of the courses of the Euphrates (Frat, Karasu) and its tributary (Arsanias, Murad).

The Euphrates and its tributary in late-antiquity sources

The Tabula Peutingeriana (Peutinger’s Map), an illustrated schematic road map in a parchment scroll, dates from the fifth century a.d,[26] and is distorted, especially in the east-west direction. Map 2 is an exhibit depicting (highlighted) the rivers Euphrates and Tigris.[27] Notwithstanding the distortion, this map illustrates clearly that Euphrates here is again the Frat/Karasu rising from the Pontic range, i.e. from the Paryadres (Παρυάδρης) mountains (Mons Parverdes in the map). A comparison with Map 16 later in this article shows that the Arsanias (Murad) should be depicted to the north of the line from Melitene (near the Euphrates) to Tigranocerta (near the sources of the East Tigris). Melitene is depicted as Melentenis, and Tigranocerta as Triganocarten. However, instead of the river Arsanias a mountain is shown along its way. Probably this is evidence to the Ptolemaic origin of the Tabula Peutingeriana as well as to the fact that Ptolemy 5.13.7 was misinterpreted already from the fifth century.


Map 2: No Arsanias in the surviving copy of the Tabula Peutingeriana

(annotated extract from the scalabe map on line at peutinger.atlantides.org)

The Geography once attributed to the Armenian fifth (?) a.d. cent.[28] historian Moses of Choren (Movsēs Xorenaci)[29] speaks of the Karasu (Frat) as the Euphrates.[30] Xorenaci himself speaks of the Karasu (Frat) as the Euphrates in his History.[31]

Procopius refers also to the Frat (Karasu) as Euphrates in the area of Erzurum.[32] He makes clear that the Euphrates (Frat, Karasu) receives the tributary Arsinus (Arsanias, Murad).[33]

In an Armenian manuscript which elaborates on the scripture of Genesis 1.10-14,[34] there is a Notice on the Four Rivers of Paradise and of Forty-Two Other Famous and Major Rivers. It distinguishes clearly between the Euphrates and the Arsanias.[35]

In the History of the Armenians, attributed to Faustos of Byzantium, a fifth-century a.d. Armenian,[36] there are references to the Euphrates.[37] In the nineteenth century, one of these references was taken to mean that Faustus referred to the river Murad (Arsanias) as the Euphrates.[38] Here the identification of the river is associated with that of the city Mcurn in the text. This city still remains mysterious for the scholars.[39] It was destroyed by an earthquake in antiquity. Already by the fifth century a.d. Mcurn was only a vague memory.[40] There are several proposals for its location. Its relation with the Euphrates is discussed later in this article.[41]

In the tenth century, the Euphrates is still Εφράτης and the Arsanias is Ἄρσινας (Arsinas) or Ἀρσίνης (Arsines) in the chronicle of the continuators of Theophanes, narrating the campaign of the Emperor Basil I the Macedonian (r. 867–886).[42]

The Arab geographers in the next section show that the Arsanias retained its name down to the thirteenth century. An examination of the Arab prose and poetry of the era might prove useful. For example, the Arab poet Abu Firas[43] is cited as alluding to a battle that took place near the Arsanias, and to ‘Greek blood flowing into the river.’[44]

The Euphrates and the Arsanias of the Arab geographers

According to the tenth-century Arab scholar Ibn-Serapion (Suhrab),[45] the source of the Euphrates [Frat, i.e. the Karasu] is at a spring in Mount Akradkhis.[46] Then[47] it passes the city of Kamkh [Kemah] and next, after flowing past Malatya, which is two miles distant from its bank, it comes to the city of Sumaysat [Samosata], to Jisr Manbij [Zeugma, Ζεγμα], to Balis [Barbalissos] … For the Arsanias Ibn-Serapion says that a river called Arsanas falls into the Euphrates, and this is the river of Shamshat [Arsamosata]. Its source is in a mountain in the limits of the country of Taron.[48] It flows by the gate of the city of Shamshat, and then passes near the gate of a fortress called Hisn Ziyad [Kharput]. Finally it falls into the Euphrates about two marches above Malatya, and on the eastern bank. [49]

The tenth-century Arab historian and geographer al-Masudi[50] says[51] that the Euphrates originates in the territory of Kalikala [Erzurum],[52] a border town of Armenia. It stems from the mountains of Afradohos,[53] one day’s walking distance from this city. It crosses the country of the Rum [Romans, Byzantines] before reaching Malatia. One of Masudi’s co-religionists, who was a prisoner among Christians, assured him that the Euphrates, in his course across the country of the Rum, receives several tributaries, including a river that flows out of Lake el-Marzeboun, the largest lake in this country. Probably the Murad (Arsanias) is mentioned here, if Lake el-Marzeboun is Lake Van (which is the largest lake in this area), although the Murad, which runs parallel and to the north of Lake Van (see Map 16), does not flow out of this lake.

In the eleventh century the Andalusian Muslim geographer and historian Al-Bakri[54] gives to the Arsanias the pronunciation Arasnas.[55]

The Frat (Karasu) of Erzurum is also the Euphrates of the twelfth-century Muslim geographer and cartographer al-Idrisi[56] who says that the Euphrates issues from the very heart of the territory of the Rum [Romans], not far from Cazala, from the mountains of Calicala [Erzurum]. Then, traversing the Rum [Roman] districts, it flows as far as Cameh [Kemah], and thence to Malatia, so as to be only two miles distant from it. It then descends to Samosat [Samosata], whence it is navigable to Baghdad.[57]

Idrisi testifies also to the name still retained by the Arsanias in his time. He identifies a place called Tell-Arsanas on the banks of this river, which he distinguishes clearly from the Euphrates: ‘From there [Malatia], it is 12 miles to Tell-Arsanas, a place located on the banks of a considerable tributary of the Euphrates, which flows into the river below Simsat [Arsamosata].’[58] The Arsanias is also called ‘River of Arsamosata’ both by Ibn-Serapion in the tenth century as shown, and by the fourteenth-century Arab geographer Abulfeda later in this section.

In his Tabula Rogeriana, the world map prepared by Idrisi in 1154 on the commission of the Norman king Roger II of Sicily, there is a relatively accurate depiction of the Euphrates and its tributaries.[59] As shown in Idrisi’s detail (Map 3), this is one of the last references to the river Arsanias (Murad, Arsas in the map).


Map 3: Al-Idrisi: Detail with the names of Euphrates (al frat), Arsanias (arsas) and Tigris (digla), with the north to the bottom

(annotated extract from the scalabe map on line at the Library of Congress)

In the thirteenth century the Arsanias was still called Arsanas as testified by Yaqut al-Hamawi[60] in his Mu'jam ul-Buldān [61] who refers to the extreme coldness of its waters.[62]

According to the fourteenth-century Arab geographer Abulfeda,[63] the Euphrates [Nahr Al-forat] is again the Frat (Karasu) originating in the Erzurum area.[64] It passes near the town of Malatya and arrives at Samosata. From there, it heads towards the east, passing from the Castle of Rum,[65] name of a fortified place in the southwest of the river. Abulfeda says that the Euphrates receives many rivers, and gives naissance to several canals. Among the rivers that flow into the Euphrates, he mentions first the Arsanias (Murad) under the name Arsamosata river [Nahr Şemşat], passing Şemşat [Arsamosata], then Zyad-Hisn [Castle Zyad], otherwise known as Khart-Bert [Kharput], and flowing into the Euphrates, above Malatya.

The absence of the Arsanias (Murad) from early modern maps in Europe

The wrong interpretation of Ptolemy 5.13.7 was the reason why the Arsanias (Murad) was not showing on early modern maps drawn in Europe. In the 1489 Map of the World by Henricus Martellus (Map 4),[66] which was based on Ptolemy’s work, the Euphrates (Frat, Karasu) is depicted as rising in the area of Erzurum, close to the sources of the Araxes, and traversing Syria and Mesopotamia on its way to the Persian Gulf. The Tigris (both its Western and Eastern branch) is also depicted. The Murad (Arsanias) is missing. Mountains are shown along its route.


Map 4: No Arsanias in the world map by Henricus Martellus (1489)

(annotated extract from the map on line at the British Library)

In the Map of First Asia (Turkey) by Martin Waldseemüller,[67] again there is no Arsanias (Map 5.1). The Euphrates (Frat, Karasu) bends to the south in the area of Erzincan and flows towards Syria. Mountains are depicted where the Arsanias (Murad) should appear.

The leading cartographers of the sixteenth century did not depict the river Arsanias (Murad) in their maps. This is the case in Gastaldi’s (1564) map of Asia Minor, a variant of which was published in Venice in 1570 (Map 5.2).[68] Further, there is no Murad in the map of Anatolia by Abraham Ortelius (Map 5.3), [69] as published in 1572/73 in a German edition of his Theatrum Orbis Terrarum.


Map 5: No eastern tributary (Arsanias, Murad) in the leading cartographers of the 16th-17th century

(annotated extracts from on line at zvab.com, raremaps.com, raremaps.com, Gallica, bergbook.com, and turkeyinmaps.com respectively)

It has been asserted that the ‘Euphrates’ (taken to be the Murad) shows on the maps of Gastaldi and Ortelius but later it was omitted by Mercator and Hondius.[70] This is not the case. There is no Murad in the map of Ortelius. On the other hand, the river showing on Gastaldi’s map as flowing into the Frat (Karasu) in the area of Erzincan is an assumed tributary of the Karasu.[71] The Murad (Arsanias), if shown on this map, should appear further south, in the area of Kharput.

The omission of the Murad (Arsanias), based on the wrong interpretation of Ptolemy, persisted in the cartography of the seventeenth century. This is evident (Map 5.4) in the 1623 edition of the 1606 map of Jodocus Hondius (1563-1612).[72] The same understanding (Map 5.5) underlies the 1632 map of Gerard Mercator (1512-1594).[73] It is repeated (Map 5.6) in the 1635 map of Joan Blaeu (1586-1673).[74]

However, from time to time appeared also maps depicting an eastern tributary of the Euphrates; they were based on Pliny (HN 5.83–84) or on the correct interpretation of Ptolemy (5.13.7). Thus, in the 1535 map of Michel Servet (1511-1553) both a mountain and a river are depicted along the course of the Arsanias (Map 6.1) but the river does not have a name.[75] This is also the case (Map 6.4) in the Map of the Third Asia (1579) from Gerard Mercator’s Geographia.[76] In the 1655 map of Nicolas Sanson (1600-1667), the royal geographer to Kings Louis XIII and Louis XIV, we have a depiction of the Euphrates and its tributaries Lycus, Arsanus, and Arsanias (Map 6.2) based on Pliny.[77] In the 1720 map published by Christopher Weigel (1654–1725) a long tributary of the Euphrates is shown without name (Map 6.3), flowing into the main stream to the south of Melitene (Malatya).[78] Finally (Map 6.5), at the bottom of the map of Henri Liébaux (1729), the Arsanias flows into the Euphrates.[79]


Map 6: Depictions of the Arsanias in 16th-18th century maps

(annotated extracts from on line maps at Gallica, Wikimedia, Gallica, geographicus.com, and Gallica respectively)

Jacobus Golius (1596-1667) in his Notae in Alfergan does not shed light on the upper Euphrates and its tributaries. He discusses Albufeda’s Şemşat (Arsamosata) and its river (i.e. the Arsanias/Murad) but he does not go beyond saying that this river does not seem to him to be the Melas of the ancients.[80]

The Euphrates of the Ottoman geographers and travelers

Djihan-numa (Mirror of the World) is the work of the Ottoman historian and geographer Hadji Khalifa (1609-1657). The first version circulated in manuscripts and reached Europe in the eighteenth century. The printed version in Ottoman text was published in Constantinople by Ibrahim Müteferrika in 1732, and its Latin translation was published in Europe in 1818.[81] Hadji Khalifa distinguishes between the rivers Murad (Arsanias) and Frat (Euphrates):[82]

River Murad has two origins. One of these is in Mount Ala [Ala Dağ], where the river is born, but then increases in size, and with its tributary, which is called Carmur, divided into four channels, passes under a great stone bridge called Gedamenşah, and is united with the stream of Melakerd [Malazgirt]. The other [origin of the Murad] is in the large summer lakes, whence it bursts towards the south joined by rivers, the most noticeable of which is the [‘small’] Kara Su rising in the meadows of Musch [Muş].[83] Passing from Kenedsch [Genç], Tjajatjur [Çamlıyurt?], and Palu, [the Murad] joins the Euphrates at Raschvan [Keban?]. River Euphrates is born in the Schugni valley between the mountains of Kalikala [Erzurum],[84] and passing first from Tergjan [Tercan], Erzengjan [Erzincan], Kemak [Kemah], Kurutjaj [Kuruçay], Akin [Ağın], and Raschvan [Keban?], where it is united with the river Murad, proceeds passing also near Hakim Chani [Hekimhan], and receiving the Kyrk Getjid [Tohma][85] river from Malatya, flows through the Nushar passage…[86] On the other hand at the sources, at the origin of the Euphrates, located near Erzurum, those who wash themselves in the springs, would not be liable to disease throughout the year. Worth mentioning is also the lake Deşt Erzen … which has fresh water and many fishes.[87]

Thus the Karasu (Frat) of Erzurum is also the Euphrates of Hadji Khalifa. Map 7 underlines the principal difference with the European perceptions in this era: [88] the eastern tributary (Arsanias) is depicted and named (Murad).


Map 7: Annotated extract from an edition of Müteferrika

(extract from a map on line at parssea.com and Wikimedia)

The second of the ten volumes of Seyahatname (Book of Travels), written by the Ottoman traveler and writer Evliya Çelebi (1611– after 1682), contains the following passage on the Euphrates:

The great river Euphrates flows through the middle of the plain of Erzerum. Its source is at the bottom of the pilgrimage of Dumlubaba… it flows towards the west … joins the Murad (the name of which it assumes),[89] and passes like a sea in the neighbourhood of Malatia to Samosat. … In the plain of Erzerum its water is very sweet and palatable, well worth being recorded in the Koran by the verse ‘And we gave you to drink of the water of the Euphrates.’… Whoever bathes therein three times, may be certain of being cured of many diseases[90]… The Euphrates freezes in the winter …, but it never freezes south of Erzerum; it is a sweet clear water, a great comfort to the inhabitants of Erzerum…[91]

From antiquity down to Evliya Çelebi, as Euphrates (Frat) was always known the river originating in the area of Erzurum. The sources of the Euphrates in the texts are invariably the sources of the Frat/Karasu. How did it happen that as sources of the Euphrates we take nowadays the sources of the Arsanias (Murad)? Who coined the terms ‘West’ and ‘East Euphrates’ for these rivers? Who proposed a march of the Ten Thousand from the area of Muş towards the sources of the Arsanias (Murad) in the east, instead of a march towards the sources of the Frat (Karasu) in the north, which was indeed a march towards the Black Sea?

The Euphrates of the French cartographers in the eighteenth century

An interesting map was drawn circa the year 1653 by Pierre du Val (1618–1683), geographer of the Kings of France Louis XIII and Louis XIV. [92] Its theme was the Itinerary of the Ten Thousand. Map 8 is a detail in the area of the Euphrates and its tributary. Xenophon’s Teleboas (Teleboe fluvius) is depicted along the course of the Arsanias (Murad). Also, usual until that time but explicitly depicted in the map, Xenophon’s sources of the Euphrates (Fons Euphratis) are those of the Frat/Karasu (Euphrates fluvius). The sources of the East Tigris are also depicted (Fons Tigris), as well as a northerly route from the sources of the latter river to those of the former, crossing the ‘Teleboas’ (Arsanias, Murad).

However, it is not clear whether this depiction of the Arsanias (Murad) on the map was documented, and constituted something more than an artistic impression. The latter is probable; first, because Duval in his Geography of the World does not discuss the courses of the Euphrates and its tributaries; [93] and, second, because in his map Turkey in Asia[94] different courses of the tributaries are depicted.


Map 8: Detail from Du Val’s (c. 1653) Itinerary of the Ten Thousand

(annotated extract from a map on line at Gallica)

The rediscovery of the Arsanias (Murad), the south-eastern tributary of the Euphrates, by Guillaume Delisle (de Lisle, 1675-1726), the leading cartographer in his era, probably was not associated with recourse to Arabic or Ottoman geographical texts.[95] Delisle’s initial view is illustrated in his 1701 Map of Turkey, Arabia, and Persia (Map 9).[96] Here the Euphrates (the Frat, Karasu) rises as usual in the area of Erzurum. It receives to the north of Malatya a river from the west, named Kara. It is not a confusion with the Karasu (Frat) but a reference to the ancient Melas, the river Tohma.[97] The Murad (Arsanias) is also depicted on the map, flowing from the east into the Euphrates to the south of Kharput (Carputh on the map) and to the north of Arsamosata (Ximxat). However, the Murad is depicted with caution, without name, shorter and thinner, less significant than the main stream.


Map 9: Annotated extract from Delisle’s (1701) Turkey, Arabia, and Persia

(extract from a map on line at davidramsey.com)

On April the 23rd, 1721, Delisle presented to the Academy of France his dissertation on the route of the anabasis of Prince Cyrus the Younger and the retreat of the Ten Thousand.[98] He explained straight away the reason of this engagement:

The King, in the progress of his studies, has arrived at the reading of Xenophon. The illustrious people who care for the education of his Majesty asked me for a map, with which he could follow this historian more accurately than with the ordinary maps. I have tried to fulfill their intention and I have had the honor to present to the King this map, of which I give here the outline, and I will report to the Academy the reasons which have driven me to adopt views away from the generally believed ones.


Map 10: Delisle’s (1721) map of Xenophon’s Anabasis, attached to his dissertation at the Academy of France

(map on line at Gallica)

Delisle’s map is depicted above (Map 10).[99] It became the standard for the route of Xenophon’s Anabasis, and still illustrates relevant books and articles. Limiting our scope to the Euphrates and its sources, we may quote in Delisle’s own words the rediscovery of the eastern tributary:[100]

Having examined ancient authors after Xenophon, in order to see if there was anything equivocal about the sources and the courses of these rivers, I found that Strabo puts the sources of the Euphrates and those of the Tigris at a distance of 2500 stadia from each other, while Xenophon puts this distance at 1350 stadia. Mr. de Tournefort distinguishes two sources of this river which form two different branches, between which lies the city of Erzurum, but they are too close to each other for use in this explanation. However Ptolemy, in his description of Armenia, describes a third and very long branch of the Euphrates and he puts the remote source at about 50 leagues to the southeast of the first sources. It is this last branch that seemed to me to be the one in question in Xenophon, especially because it is also to the north of the Carduchians, as this author suggests.

After two years (1723), Delisle published a detailed Map of the Retreat of the Ten Thousand (Map 11), probably similar to (or identical with) the one he had presented to the King.[101] As in the map attached to his dissertation at the Academy of France (Map 10), Delisle calls Euphrates all branches; first, the Frat (Karasu) near Erzurum, then the Murad, and finally the river after the confluence. He has the Ten Thousand passing near the sources of the Murad, depicted as the sources of the Euphrates (Fons Euphratis). They had passed from the sources of East Tigris (Fons Tigridis), and had crossed the local stream of Muş (the ‘small’ Karasu) as Xenophon’s Teleboas. Probably this is the earlier illustration of the prevailing view for this leg of the route of the Retreat of the Ten Thousand, and is repeated (with variations) in almost all proposals until the publication of Lendle’s book in 1995.[102]

Map 11: Annotated extract from Delisle’s (1723) Retreat of the Ten Thousand

(extract from a scalable map on line at davidramsey.com)

The same denominations of the rivers appear after Delisle’s death (1726) in his map of Alexander’s Empire and Expedition (1731).[103] Delisle must be credited with the rediscovery of the ‘lost’ eastern tributary (Arsanias, Murad) of the Euphrates because he realized that Ptolemy 5.13.7 was referring to a tributary (divergence, ἐκτροπ) and not to mountains. However, he applies to it without evidence the name Euphrates. Ptolemy is clear: ‘There is another more notable divergence from the Euphrates, which joins the river Euphrates at 71°30' 40°30', and has its sources at 77° 41°’. Thus the Murad (Arsanias) is not Ptolemy’s Euphrates. Delisle’s argument is not convincing: ‘this last branch [the Murad] seemed to me to be the one in question in Xenophon, especially because it is also to the north of the Carduchians.’ But after crossing the Centrites (Bohtan), all rivers in the Anabasis were to the north of the Carduchians.

Jean Baptiste Bourguignon d' Anville (1697-1782) belonged to the next generation of prominent European cartographers. His maps of ancient geography have influenced Xenophon’s commentators. In his 1779 book on the Euphrates and the Tigris, d’Anville refers to a Geography manuscript translated from Turkish to French. He had read Hadji Khalifa’s Djihan-numa.[104] A French translation was available in Paris since 1744.[105] However, concerning the descriptions and denominations of the rivers in the area of Xenophon’s Armenia, d’Anville’s assumptions do not seem always to be justified:[106]

“To the Euphrates, which has been already mentioned as having its origin near to Arz-roum, is added another branch, whose sources, called in the country Bing-gheul, or the Thousand Foun­tains, form a river which appears to have been that named Lycus.”

D’Anville is wrong here in his identification of the Bingöl river with Pliny’s Lycus (HN 5.83–84, see above). This river does not flow to the west but to the east. It is not a tributary of the Euphrates (Frat, Karasu) but of the Araxes. He continues:

“The river, of which the union of these two streams makes the com­mencement, is particularly called Frat. But there is still another Euphrates, which having its fountains further distant, becomes more considerable than the precedent at its junction. This Euphrates is that which, precisely un­der this name, the ten thousand passed in re­turning; and the same that Corbulon, charged with the conduct of the war in Armenia under Nero, makes issue from a district called Caranites, according to the report of Pliny”.

Corbulo and Pliny do not refer to the Murad (Arsanias) as the Euphrates, but to the Frat (Karasu), i.e. to the river originating in Caranitis (Strabo’s Καρηντις) which is the area of Erzurum (Arm. Garin). As shown later, Rennell followed d’Anville in this erroneous assumption. D’Anville continues:

“There are circumstances that authorize the application to it of the name Arsanias, which another river decidedly claims. This is what the Turks name Morad-tsai which signifies the Water of Desire”.

D’Anville means here that the Arsanias was not the Murad, i.e. his ‘Euphrates’ in the retreat of the Ten Thousand, but a different river (see Map 12 below). He continues:

“Ptolemy recognizes a twofold Eu­phrates, concerning which modern literati manifest an embarrassment which, a further knowledge of the country will remove”.

After Delisle, this is the second modern explicit acknowledgment that Ptolemy has recorded the flow of this tributary of the Euphrates. However, as shown above, Ptolemy did not apply the name ‘Euphrates’ to it. This passage is interesting because d’Anville believes that his view for a ‘second Euphrates’, which he takes to be the proper one, mentioned by Xenophon, would cause embarrassment to the scholars of his era. Indeed he knew that he was renaming the rivers in the area but the scholars were not embarrassed.

D’Anville’s initial views are depicted in his 1764 Map of the Eastern Roman Empire (Map 12)[107] which includes: the detachment of Caranitis from Theodosiopolis (Erzurum), and its displacement to the east, towards the sources of the Murad; the flow of the Bingöl river into the Euphrates; the rediscovered second Euphrates (Murad) in the place of the ancient Arsanias; and an independent river Arsanias to the south of the Murad. These views confused Saint-Martin, as shown later in the article.


Map 12: Annotated extract from d’Anville’s (1764) Map of the Eastern Roman Empire

(extract from a scalable map on line at davidramsey.com)

After some years, d’Anville improved this perspective in his Turkey in Asia (Map 13).[108] This is the first close-to-reality map of the upper Euphrates drawn by Europeans in the modern era but still with significant errors. It depicts a westerly flow of the Bingöl river towards the Euphrates (Frat, Karasu). It also depicts wrongly Malazkerd (Malazgirt) and its river to the north of the Murad instead of the south. Above all, the denomination ‘Morad chai or Euphrates’ remains arbitrary.


Map 13: Annotated extract from d’Anville’s (1794) Turkey in Asia

(extract from a scalable map on line at davidramsey.com)

The confusion caused by this new definition of the Euphrates is evident in the 1841 dictionary[109] occupying a part of the second volume in the French translation of Movsēs Xorenac‘i’s History of Armenia. It starts saying that the Euphrates of the ancients, the Frat of the Turks, has its source in the mountains of southern Armenia, near Diyadin, under the name of Murad. Then it quotes Xorenac‘i 3.59 whο refers to the Frat/Karasu as the Euphrates. It continues with the use of a different source, and the outcome is that the Euphrates (which is the Frat/Karasu but is taken to be the Murad) seems to receive along its way the river Murad. To maximize the confusion, the same author, in the same book, and in a different section, says that the Euphrates rises from the Bingöl Mountains near Erzurum.[110]

Later, this new definition of the Euphrates was reflected in the extended 1907 article on the Euphrates in the German Encyclopaedia of Classical Archeology.[111] Discussing the headwaters of the Euphrates (1196-1199), the author points (1198 60ff.):

The same is true of the numerous tributaries sent to the river by the mountains, including the largest, the Arsanias, which the newer Geography regards as the other source of the current Euphrates.

He had interpreted Ptolemy (5.13.7) correctly (1199 8ff.):

The unnamed [by Ptolemy] eastern tributary can be none other than the Arsanias; Ptolemy has its spring 10 minutes of the degree south of Mount Abas.

Earlier, challenging the view that as sources of the Euphrates were meant by some ancient authors those of the Arsanias (Murad), he notes (1198 18ff.):

The name Caranitis is in Karin, the Armenian name obtained by the neighborhood of Erzerum. Theodosiopolis should be searched either at Erzerum itself or in its immediate vicinity. The distance indicated by Procopius (42 stadia) fits quite well. By contrast, the mountains of the sources of the Murad-Su, in their relation with the sources of the Tigris, either here or at Ala-Dagh, do not make sense.

The generalized conversion of the Murad into Xenophon’s Euphrates

Delisle and d’Anville offered to the Europeans the modern geographical perception characterized by a valuable innovation: the rediscovery of the lost Arsanias tributary of the Euphrates, and its placement on the map. However, their identification of the Arsanias (Murad) with Xenophon’s Euphrates was not based on solid grounds. It rather reflected a subjective interpretation of the text of the Anabasis, without further justification. In the following years, in the early nineteenth century, multiple attempts were made for the identification of the route of the retreat of the Ten Thousand in its leg related with the crossing of the Euphrates ‘not far from its sources’ (Anab. 4.5.2). The novel views and maps of Delisle and d’Anville were utilized by the prominent commentators of the Anabasis in this era.

Rennell’s probable responsibility for the confusion which still prevails, has been discussed.[112] Initially he says that “this river [the Teleboas] answers to the Arsanius of Plutarch, to which Lucullus came, on the fourth march from the northern foot of Taurus” (207 note). Thus he identifies the Teleboas with the Murad (Arsanias). But later (210–211) he confuses the headwaters of the Murad (according to Hadji Khalifa) with Pliny’s headwaters of the Euphrates (the Karasu). Rennell goes on to assert that “the Murad is also the Euphrates of Strabo. So the ancients applied the name of the confluent stream to the eastern branch, as the natives do Frat to the western” (211).

Kinneir and his company travelled in the area in the year 1814 on their way from Erzurum.[113] They crossed the Murad between Karaağıl [Karagool] to the north and Erentepe [Leese][114] to the south of the river. They swam the horses across and themselves passed on a raft supported by inflated sheep skins (378). Kinneir believes that the Greeks crossed the Murad further east, closer to its sources, because in the point he passed, the Murad was nowhere fordable (489). However, not long after Kinneir’s time the fording of the Murad to the north of Muş, to the west of Kinneir’s crossing, has been reported in at least four publications.[115] An old bridge stood (and still stands) there,[116] indicating the layout of the north-south caravan road.


Map 14: Annotated extract from Arrowsmith (1813)

(extract from a scalable map on line at davidramsey.com)

These views of Kinneir for the route of the retreat and the subsequent naming of the rivers are depicted in his map drawn by Arrowsmith and published in 1813 (Map 14).[117] The river after the confluence is named Frat or Euphrates. The Karasu is named West Frat, and the Murad answers to East Frat. Sources of the Euphrates are the sources of the ‘East Frat’ (Murad, Arsanias). This kind of naming still continues, and not only in works associated with the retreat of the Ten Thousand. Now, in any publication, the rivers before the confluence are called West and East Euphrates respectively. Also, as the Murad is longer than the Frat/Karasu up to the confluence,[118] the total length of the Euphrates (ca. 3000 km) is calculated starting from the sources of the Murad. In a European context, this would be the equivalent of naming Eastern Rhine the river Main; or naming Southern Danube the river Inn.[119]

The ‘evidence’ on the Murad as the real Euphrates of the Armenians

Notwithstanding that the prevailing views on the route of the retreat in the nineteenth century, indeed until now, were taking the Arsanias (Murad) as Euphrates, still the evidence for this selection was missing. The commentators thereafter accepted that the river Murad was meant here by Xenophon, and, apart from Rennell mentioned above, they did not engage in the justification of their selection. According to Bunbury, Xenophon’s Euphrates was evidently the Murad Chai, or eastern branch of the river because ‘Armenian writers apply the name of Euphrates to both arms, and the same usage probably prevailed among the natives in the time of Xenophon.’[120] Bunbury does not mention the names and works of these Armenian writers. His view was reproduced by Tozer, still without names.[121] Half a century before Bunbury and Tozer, Smith had noted that this river [the Frat, Karasu] was considered the proper Euphrates by the Greek and Roman writers, but ‘the Armenians give that honour to the Murad-chai’.[122] Smith offers a reference: ‘St. Martin, vol. 1, p. 42.’

Despite his responsibility for the confusion on the Murad as the real Euphrates discussed hereafter, Saint-Martin’s contribution was unambiguous. Apart from his work on Armenia, he initiated the research into the Urartu civilization,[123] unknown to his contemporaries, but unknown also to the commentators of Xenophon’s Anabasis, who confused (and probably still confuse) the Khaldians (Urartu) of Book 4 with the Chalybes of Book 5. However, it seems that Saint-Martin was himself misguided by the denominations of the prominent geographers in his era, Delisle and d’Anville. Actually Smith’s reference is not to Saint-Martin (I, 42) but a little farther. Here (I, 42) Saint-Martin says:

Also in the mountains surrounding Arzroum have their sources some of the rivers which serve to form the Euphrates. This river is called Ep'hrad in Armenian; it is formed by the union of a large number of rivers coming from various parts of Armenia. The northern part of the river Euphrates has its source in the canton of Garin, near the present city of Arzroum. The mountains in the vicinity of this city are now called, as we have already said, Bing-Gueul, that is to say, of a thousand lakes, because of the large amount of springs and small lakes found there. It is these various clusters of water which form the beginning of a part of the Euphrates.

After a discussion of the views of Strabo and Pliny, Saint-Martin continues (I, 45-46):

All Oriental geographers agree in placing the sources of the Euphrates in the neighborhood of Erzurum. … The Euphrates [Frat, Karasu] passes from the cities Arzendjan [Erzincan], Kamakh [Kemah], and Akin [Ağın]. Then it receives a very considerable river which comes from the eastern side, from the center of Armenia, a river which seems to be Pliny’s Arsanias. The Turks call it Mourad-chai, and the Armenians consider it to be the real Euphrates (3).

Saint-Martin’s footnote (3) reads: ‘Schamir, chapter vi, p. 134; Djihan-numa, p. 426-27.’

Schamir was an Armenian merchant who published in 1775 at Madras, India a book of 148 pages with the title The remaining of the history of Armenia and Georgia (Le restant de l'histoire de l'Arménie et de la Géorgie). It contained Mesrob Eretz’s life of St. Nerses, patriarch of Armenia, and the short history of Georgia and Armenia of the Armenian thirteenth-century chronicler Stepanos Orbelian (c. 1250–1305)[124] which is incorporated into Saint-Martin’s book, in Armenian with French translation.[125] It has no references to rivers whatsoever. Indeed Saint-Martin does not refer to this book of Schamir which contains Armenian medieval sources because in an earlier footnote (I, 34) we read:

(1) Exhortation to the Armenians by Eleazar Schamir, in Armenian; Madras, 1772. In addition to an exhortation to the Armenians, to excite them to shake off the yoke of their oppressors, the book also contains a historical summary of Armenia with a geographical description.

Thus there was a previous publication by Schamir.[126] Probably the geographical description of Armenia in his 1772 book contained a reference to the river Murad (the Arsanias of the ancients) as Euphrates. It either originated in West European contemporary sources or, which is more probable, Schamir had incorporated into his book a text similar to the Armenian ecclesiastical geography written in the year 1721 if not the text itself.[127]

Hadji Khalifa’s Djihan-numa (Mirror of the World) is Saint-Martin’s second reference. In a footnote (I, 33) previous to the one discussed here (I, 46), Saint-Martin refers to the Constantinople 1732 edition of Ibrahim Müteferrika. As mentioned, the Latin translation of this work was published in 1818, in the same year with Saint-Martin’s book. Saint-Martin could not count Hadji Khalifa, whom he cites frequently, as one of the Armenians who consider the Murad to be the real Euphrates; first, because Hadji Khalifa was not an Armenian; and, second, because Hadji Khalifa not only does not say that the Murad is the Euphrates, but his Euphrates is unambiguously the Frat/Karasu rising in the area of Erzurum. In order to understand Saint-Martin’s reference to Hadji Khalifa, an examination of all references to this river in his book was necessary. Thus in an extended passage Saint-Martin (I, 50-52) uses Hadji Khalifa’s description of the Murad, and it mixes it with d’Anville’s erroneous early assumptions, as well as with geographical details which cannot be traced in the works of the Armenian writers translated and incorporated into Saint-Martin’s two volumes. The extended quotation of his text (I, 50-52) hereafter is necessary because Saint-Martin is the only ‘evidence’ of the proponents of the Murad as the real Euphrates of the Armenians:

In these mountains originates the river that the Armenians regard as the true Euphrates, which appears to be the Arsanias of Pliny; currently it is known by the Turks under the name of Murad chai.

This is Saint-Martin’s second assertion that the Armenians regard the Murad as the true Euphrates,[128] still without justification. He continues with a description of the Murad which originates in Hadji Khalifa, despite some misunderstandings:

[The Murad] is formed by several small rivers that flow from various points, the two principal of which from the north and the east are united in the plain of Mush. The first, which has more particularly the name Mourad-chai, rises from Mount Dzaghgé[129] in a place called Osgik'h in the province of Dzaghgodn.[130] Initially it is formed by various streams which, after meeting, are divided in a place called Tcharmour into four arms, which join again and which, after traversing a long distance, mingle with the river of Melazkerd, which comes from the north, from the mountains of Bingöl. (1)

Hadji Khalifa has told us above that the one of the two origins of the river Murad, together with its tributary which is called Carmur,[131] divided into four channels, pass under a great stone bridge and are united with the stream of Melazkerd (Malazgirt). However, this stream does not flow into the Murad from the north but from the south. Even in less detailed maps (see Map 15) it is obvious that the location of Malazgirt is to the south of the Murad. Obviously Saint-Martin was misled here by the error in d’Anville’s map (Map 13) which shows Malazkerd (Malazgirt) to the north of the Murad, and the river of Malazkerd flowing into the Murad from the north. Footnote (1) in Saint-Martin[132] refers to the History of Armenia written by the Patriarch John VI Catholicos (Yovhannēs Drasxanakertc‘i, 1203–1221),[133] who mentions that the Christians living at the sources of the Euphrates with their leader Oski were put to death.[134]

Saint-Martin accepts the association of this place with Osgik'h placed by Vartan at the sources of the Murad, and taken to show that the Murad was the Euphrates in the history of the Patriarch.[135] He continues:

This second river, which flows from north to south, bears also among the Armenians the name Euphrates. It descends from the area around the ancient city of Medzourk'h into upper Armenia and into the canton of Mendzour. (I, 50-51)[136]


Map 15: The cantons of Upper Armenia in 387 a.d.

A careful examination reveals that Saint-Martin’s unknown source did not refer here to Malazgirt but to Mazgirt.[137] Although the name of the river near Mazgirt is Munzur and not Euphrates, first, it rises from the Bingöl mountains;[138] second, it traverses Upper Armenia, the nine cantons of which are depicted in the map;[139] and, third, it passes from the canton of Mendsour (no.3 in Map 15).[140] Saint-Martin continues with a description of the ancient city of Medzourk'h by Faustus of Byzantium, already mentioned earlier in this article:

Here is how Faustus of Byzantium, Armenian historian of the fourth century, talks about the surroundings of this city: ‘Hair, the chief eunuch [Martbied] came to the holy places, descended to the bank of the river Euphrates in a plain covered with a thick forest, at the confluence of two rivers, into a grove of plum trees, in the place where formerly existed a city that had been founded by King Sanadroug, and which was called Medzourk'h.’ This river is united with the other arm of the Euphrates near the town of Mandzgerd, or Melazkerd in Turkish. (I, 51)

Saint-Martin thinks that this passage attests that in the fifth century Faustus of Byzantium was calling Euphrates the Murad. He thinks so, because as seen below in the continuation of his text, he still believes that the tributary referred to above was not the river of Mazgirt but that of Malazgirt. His mistake is obvious because he continues describing the Murad receiving further to the west the ‘small’ Karasu of Muş (see Map 15).[141] In the direction of the flow of the Murad, Muş is after Malazgirt and before Mazgirt. If Saint-Martin had realized that the river of Mazgirt and not that of Malazgirt was the one rising from the Bingöl mountains, and descending from the area around the ancient city of Medzourk'h into Upper Armenia, and into the canton of Mendsour, then he himself would locate the ancient city of Medzourk'h near the upper reaches of the river of Mazgirt, close to the Frat (Karasu), i.e. the Euphrates to the bank of which descended Hair Martbied according to Faustus of Byzantium. Saint-Martin continues with a description which repeats the initial error of d’Anville (cf. Map 13) that the Murad (‘Euphrates’) and the Arsanias (Aradzani) were two different rivers:

The Euphrates or Murad chai flows then westward, to join the other Euphrates which comes from Erzurum. It [the Murad] receives in its course, after passing the ancient fort of Oghnagan, the Aradzani [Arsanias] river which is perhaps the [‘small’] Karasu of today’s people. The Aradzani is considerably swollen by the waters of several rivers which descend from the mountains of the Kurds and cross the Daron country, presently Mush. There is much talk of this river [Aradzani, Arsanias] in the history of Armenia: the main rivers  it receives,  are the river Meghdi, commonly called Meghraked, which comes from the city of Meghdi in the country of Daron, and the river Aldzsan, which waters a valley of that name, on the borders of the same country. (I, 51-52)[142]

Thus, according to Saint-Martin here, the ‘much talk of the river Aradzani [Arsanias] in the history of Armenia’ was about the local stream of Muş. It is sad, but Saint-Martin continues with an even more impossible argument: the ‘Euphrates’ (Murad) loses its name when it receives the [different] Arsanias, but receives again this name after its confluence with the Euphrates (Frat, Karasu):

It appears that the Aradzani [Arsanias], after joining the Euphrates [Murad], communicates its name to the latter, which it retains, until it joins the northern branch of the Euphrates, and it appears to answer to the Arsanias of Pliny, which joins the same river approximately at the same point. Before the meeting of these two rivers, the Murad chai or Aradzani receives on his left side the Mamouschegh, currently called Mamousch, which comes from the south, from the borders of Sophene. (I, 52) [143]

Unfortunately, this is Saint-Martin’s understanding of the Euphrates and the Murad (Aradzani, Arsanias) in this part of his book (I, 50-52). Later however (I, 171), he returns towards his earlier view (I, 47, see above), and asserts:

In order to go to Artaxata, which is located in the center of Armenia, it is necessary to cross the river Arsanias, which corresponds to the Aradzani of the Armenians, and is the southern Euphrates or Murad chai.

Saint-Martin was attempting to compromise two views: the description of Hadji Khalifa in Djihan-numa which was clear and unambiguous, with the Euphrates (Frat) descending from the north and receiving the Murad from the east; and d’Anville’s views[144] which were less than clear (cf. Maps 12 and 13), and were biased towards a Murad (‘Euphrates’) which could fit in an easterly route of the retreat of the Ten Thousand.

It may well be that Saint-Martin’s complicated views about the Euphrates and its tributaries are not convincing. However, a question is still pending. Were there any Armenian writers who believed that the Murad (Aradzani, Arsanias) was the real Euphrates? Saint-Martin himself has provided Armenian evidence to the contrary. He has translated and published Movsēs Xorenac‘i. He has translated and cited Faustus of Byzantium, even with the mentioned misunderstanding. He has translated and published the Armenian manuscript with the Rivers of Paradise. Up to now, the only reference to an Armenian probably believing that the Murad was the real Euphrates is to the mentioned eighteenth-century merchant-publisher in Madras. However, we have mentioned that Saint-Martin translated and published the work of a second Armenian, who could possibly be the reference of the first. The Geography of Vartabied Vartan (1721) in Saint-Martin (II, 435), taking for granted the association between a place called Saint-Oski (Osgik'h) at the sources of the Murad on the one hand, and the adobe of the first-second a.d. century blessed Oski and his followers at the sources of the Euphrates on the other, as reported in the thirteenth-century history of the patriarch, reads:

Next comes the river Euphrates which descends from the province of Egéghéats [Acilesene/Erzincan, see above]. But the real Euphrates is the one which comes out of Saint-Oski, passes from Daron [Mush, see above], and goes to Syria.

It seems that this is the evidence implied by Smith, Bunbury, and Tozer. It is the only ‘evidence’ of the proponents of a march of the Ten Thousand towards the sources of the Murad, although nobody bases anymore explicitly on Saint-Martin’s confused passage (I, 50-52) the selection of this river as Xenophon’s Euphrates. A solution has been invented: the renaming of the Euphrates (Frat, Karasu) and the Arsanias (Murad) into West and East Euphrates. Usually Xenophon is taken to mean that the Ten Thousand crossed near the sources of the ‘East Euphrates’. A crossing near the sources of the ‘West Euphrates’ could be possible, if someone preferred it. After all, with this version of rewritten history and ancient geography, both rivers are supposed to be known for centuries under this name.

The Murad (Arsanias) and the border between the two satrapies of Armenia

Rereading Xenophon’s text without biased views on which of the two rivers (Frat/Karasu or Murad/Arsanias) was (or was not) the Euphrates, it would not be irrelevant to keep in mind that during the last three hundred years we base our views on a map produced by Delisle, probably hastily, on the request of the courtiers, in order to facilitate the lectures given to King Louis XV on the Anabasis.

In Xenophon’s words (Anab. 4.3.1), the southern border of Armenia was the Centrites identified with the river Bohtan. This river was separating Armenia from the country of the Carduchians. At daybreak the Ten Thousand saw horsemen and foot-soldiers across the river, ready to deny them the passage. It was the army of Orontas (4.3.3-4), a son-in-law of the Great King (2.4.8), usually regarded as the ancestor of the Orontids of Armenia.[145] The Ten Thousand crossed the river, dispersed their enemies, and entered Armenia on 8 February 400 b.c.[146] Orontas does not appear again in the text of the Anabasis. They went on northwards for five parasangs, and reached a large village with a palace. They were at Siirt.[147] From there they marched ten parasangs, until they passed the headwaters of the [East] Tigris (4.4.3). They were at Bitlis.[148] They continued for 15 parasangs and arrived at the river Teleboas. This was a beautiful river, though not a large one, and there were many villages about it (4.4.3). They were in the area of Muş.[149] The region was called Western Armenia (4.4.4). Its vice-satrap (παρχος) was Tirivazos, a friend of the Great King.[150] Tirivazos concluded a treaty with the Greeks (4.4.6).

Thus, from Xenophon’s description, the southern border of Armenia was the Centrites (Bohtan). Also, near Teleboas was the border between Armenia (governed by Orontas) and West Armenia (governed by Tirivazos) because Tirivazos enters the scene near this river, and, after the treaty, follows the Greeks at a distance with his army (4.4.7).


Map 16: The two satrapies of Armenia around year 400 b.c.

The river Teleboas is not mentioned by other writers. Most commentators agree that Xenophon means here the ‘small’ Karasu, the local stream of Muş which empties into the Murad (Arsanias) a few kilometres to the north of this town. If this small river constituted a part of the border between the two satrapies of Armenia, then West Armenia was not to the west and south of the ‘small’ Karasu, because the Greeks had arrived from this direction, after crossing the Centrites and dispersing the army of Orontas (see Map 16). Thus the country to the southwest of the ‘small’ Karasu, was the Armenia of Orontas. Consequently, if the ‘small’ Karasu was the border, then the West Armenia of Tirivazos was to its northeast. With this perception of the border, according to most proposals,[151] the Greeks, after marching for a while along the right (eastern) bank of this river, turned to the east, towards the sources of the ‘Euphrates’ (Murad). They were followed by Tirivazos, the satrap of West Armenia, who was marching in his own territory. Map 16 was drawn according to the prevailing view that the ‘small’ Karasu was part of the northern border of the Armenia of Orontas. The details of this view are depicted[152] but the pink area, taken to belong to West Armenia, is differentiated by color. It seems that this area to the north of Lake Van and to the south of the Murad (Arsanias) is taken to belong to West Armenia just in order to justify the passage of the Greeks from there, followed by Tirivazos, in their march towards the sources of the ‘Euphrates’ (Arsanias, Murad). Otherwise, as Vaspurakan, the area around Lake Van, is the historical heart of Armenia, it would be rational to assume that it belonged to the Armenia of Orontas. In this case, the border between the two Armenias was the Arsanias (Murad, see Map 16), and the proposals for an easterly march over the assumed territory of West Armenia towards the sources of the Murad do not hold. However, the route towards the sources of the Euphrates of all historians and geographers, the Frat/Karasu, is not affected by the satrapal jurisdiction over the pink area.[153]

The Greeks never passed to the eastern bank of the ‘small’ Karasu. They crossed into the territory of Tirivazos near its confluence with the Murad. The two possible interpretations of Xenophon’s Teleboas do not affect the route and have only to do with which of the two rivers (Murad or ‘small’ Karasu) is meant by this name: either Xenophon’s Teleboas was the Murad itself[154] forded to the north of Muş near the confluence of the ‘small’ Karasu; or Xenophon does not mention the Murad and refers as Teleboas to the ‘small’ Karasu at its confluence and next to the fording.[155] In the nineteenth century the Murad was forded to the north of Muş. To the east of this fording still stands a fourteenth-century bridge. Its construction along the north-south route, near Muş, testifies to the layout of the ancient route.


Map 17: Kiepert’s (1857) retreat of the Ten Thousand

(annotated extract from a scalable map on line at davidramsey.com)

If this caravan route crossed traditionally the Murad somewhere else, for example at Yoncali, to the north of Bulanik, as proposed by Lendle after Kiepert (see Map 17),[156] then the bridge would have been there.[157] We have quoted a detailed description of the fording of the Murad near its confluence with the ‘small’ Karasu:[158]

[The Murad] appeared to be flowing in two channels through a bed having a width of 200 yards or more. After fording the first of these branches … we made our way over a beach to the second branch … the water reaching to the horses’ knees … We prepared to say good-bye to the Murad. What was our surprise to meet a third and magnificent river, sweeping towards us in an independent bed! … The confluence of the Kara Su, the stream which collects the drainage of the plain of Mush, is situated some little distance above the ford.

This ‘magnificent river’, ‘some little distance above the ford’, is probably Xenophon’s ‘beautiful river, though not a large one’ (4.4.3). Its confluence with the Murad marks the point of the crossing into the West Armenia of Tirivazos, and allows for the continuity of the northern border of (East) Armenia to the east, along the Murad (see Map 16). The ‘small’ Karasu was not part of this border. Precisely, the only part of the ‘small’ Karasu which constituted part of the northern border of (East) Armenia was its confluence with the Murad.

Xenophon’s retreat to the sources of the Euphrates

The Ten Thousand crossed the Murad near its confluence with the ‘small’ Karasu, and entered West Armenia. Their march (followed at a distance by Tirivazos) was not to the east of Muş, towards Bulanik, Malazgirt and beyond. It was to the north of Muş, towards Hınıs and Erzurum, over the territory of West Armenia (Map 16). The Greeks were not heading to the sources of the Murad (Arsanias) in the east but to the sources of the Euphrates (Frat, Karasu) and to the Black Sea in the north. Their misorientation underlying an easterly march is not justified by Xenophon’s text. After crossing the Centrites and entering Armenia, they found themselves in favourable con­ditions: they passed from populous places; from Siirt; (4.4.2); from Muş (4.4.3); later from Hınıs (4.4.7); thus they had more than once the opportunity to confirm the route. They had made a treaty with Tirivazos (4.4.6); thus they were marching through friendly country.[159] Later, in their march through the snows of Armenia, they had many guides.[160]

The details of a proposal for the march of the Ten Thousand after crossing the Teleboas and the Euphrates (Frat, Karasu), until their arrival at the Armenian villages with the subterranean houses (4.5.22) have been offered.[161] There is no indication that the Ten Thousand had lost their way. Led by many guides, they arrived where they wanted to arrive: at the foot of the mountain Kop, at or near the point from which the summer road from Erzurum ascended towards Bayburt and Trapezus.[162] However, they arrived there in late February.[163] The Kop mountain road towards Bayburt was blocked by the snow. They numbered more than ten thousand. After a stay of eight days, they had to leave. The villages were covered by snow and, probably, were running out of food. Two options were available: to follow the caravan highway to the west, in the valley of the Euphrates (Frat, Karasu); or to follow the same caravan highway to the east, towards Erzurum and beyond, in the valley of the Araxes (Phasis, Aras). They opted for the latter.

Here starts their easterly march towards the Phasians; the Taochians; the Khaldians (Urartu); the river Harpasos (Arpaçay); the Scytheni (Scythians) in present-day Armenia; and the city of Gymnias (Gyumri). This easterly march had nothing to do with the crossing of the Arsanias (Murad) near its sources as ‘Euphrates’. Indeed they crossed the Euphrates (Frat) of all known historians and geographers near its sources, in the vicinity of Erzurum, where all these historians and geographers, from antiquity until the seventeenth century, place the sources of the only river known as Euphrates, i.e. the Frat/Karasu.


In all ancient Greek, Roman, Armenian, Arabic, Byzantine, and Ottoman texts, sources of the Euphrates are those of the Frat (Karasu) which rises near Erzurum. There is no evidence that the river Arsanias (Murad) was ever known as Euphrates. This river was made into Euphrates in the eighteenth century, for the sole purpose of fitting certain views on the route of the retreat of the Ten Thousand.

Justification of this renaming was sought afterwards. Despite references to Armenian sources, the only relevant relates to an ecclesiastical geography of Armenia, written in the year 1721. It takes a place Saint-Oski near the sources of the Arsanias (Murad) as marking the adobe of the disciples of apostle Thaddeus and their leader Oski in the first-second century; it combines it with a passage in the History of the thirteenth-century Patriarch of the Armenians John VI Catholicos, according to which the disciples were living at the sources of the Euphrates; and it asserts that the Arsanias (Murad) is the ‘real Euphrates’ of the Armenians.

This subjective identification is in itself inadequate; the more so when balanced against the evidence offered in scripts, from the Neo-Assyrian Empire of the ninth century b.c. down to the Ottoman Empire of the seventeenth, including Armenian texts. The perpetuation of the view that Xenophon’s Euphrates was the Arsanias (Murad) is not justified. Unless adequate evidence to the contrary is provided, Euphrates was always (and still is) the Frat (Karasu). Probably the proposals on the route of the retreat of the Ten Thousand towards the sources of the Euphrates have to be re-examined. Also, besides the modern reference to the Frat (Karasu) and to the Murad (Arsanias) as the ‘West’ and ‘East Euphrates’, it does not seem that the people who lived and/or live in the area had ever adopted this arbitrary renaming.

March 2016



[1] The author is not affiliated with the University of Patras, Greece. He is an independent researcher who graduated from this University long ago (1977).

This article would not manage to be considered for publication in peer-reviewed journals. The reasons are understandable. First, it spans over thirty centuries but the personalized scientific knowledge, hopefully or not, is fragmented; there are orientalists, classicists, medievalists, and so on. Second, it demonstrates that a big river (the Arsanias, currently Murad, a tributary of the Euphrates) was ignored for centuries in the West because of a wrong interpretation of Ptolemy’s 5.13.7. Third, it shows that the rediscovery of the ‘lost’ river in the eighteenth century was confirmed by recourse to Oriental texts. Fourth, it demonstrates that contrary to any evidence, the ‘rediscovered’ tributary was arbitrarily given the name of the river Euphrates. Fifth, it shows that to this tributary (the Arsanias, Murad) was given the name Euphrates just in order to fit in certain views on the route of the retreat of Xenophon’s Ten Thousand. Sixth, it shows that this arbitrary renaming has resulted not only in incomprehensible proposals on the route of Xenophon’s Retreat; it has also changed the centuries-long geographical nomenclature in the area. Thus the article offers a good opportunity for post-publication reviews. Objections are welcome. The author has spent considerable time trying to understand why the Arsanias (Murad) is still considered to be Xenophon’s Euphrates, precisely the East Euphrates. Many thanks to Philica for publishing and hosting this article.           

[2] Cf. A. Negev and S. Gibson, Archaeological Encyclopedia of the Holy Land (New York, 2001) 169; Christopher Woods, “On the Euphrates”, Zeitschrift für Assyriologie und Vorderasiatische Archäologie, 95.1-2 (2005) 7–45.

[3] The Black Obelisk, discovered in 1846 by the archaeologist Henry Layard during his excavations of the site of the ancient Assyrian capital Kalhu (Nimrud), now at the British Museum, glorifies the achievements of King Shalmaneser III (reigned 858-824 B.C.). In D. Luckenbill, Ancient Records of Assyria and Babylonia, I (Chicago, 1927) 219 a part of its cuneiform text reads: ‘[604] From Enzite I departed. The river Arsania I crossed. To the land of Suhme I drew near. Uashtal, its stronghold, I captured’. The cuneiform text on the Black Obelisk refers also to the Euphrates as a river different from the Arsania: ‘[558] In my first year of reign I crossed the Euphrates at its flood… [559] … the Euphrates I crossed at its flood… [560] I crossed the Euphrates…’: Luckenbill, Ancient Records, 201, 202, 203, etc.

[4] I. K. Para­deisopoulos, “Route and Parasangs in Xenophon’s Anabasis,” GRBS 54 (2014) 220–254 at 222 [hereafter “Paradeisopoulos 2014”].

[5] A contribution to the solution of this complex puzzle has been offered in the following three articles: I. K. Para­deisopoulos, “A Chronology Model for Xenophon’s Anabasis,” GRBS 53 (2013) 645–686 [hereafter “Paradeisopoulos 2013”]; Paradeisopoulos 2014 (see previous note); and I. K. Para­deisopoulos, “Analogies between Xenophon’s Parasang and Hamilton’s Post-hour”, GRBS 55 (2015) 353–390 [hereafter “Paradeisopoulos 2015”].

[6] There is a long-standing tradition of assigning mistakes to Xenophon, and of basing proposals for the route, the parasangs, and the chronology of the Anabasis upon personal assumptions. An example: the proposal of C. Sagona, “Did Xenophon take the Aras High Road? Observations on the Historical Geography of North-East Anatolia,” in A. Sagona (ed.), A View from the Highlands (Louvain 2004) 299–333, at 302, Map, is not compatible with Xenophon’s distances. The distance between Erzincan and Gümüşhane is only 131 km (approx. 23 parasangs, 4-5 days’ march), but Xenophon narrates a march of 32 days from the Armenian villages to Gymnias. Also this proposal implies that Xenophon had misinterpreted the names of all the rivers in the region.

[7] In the following centuries, this view was repeated, among other, by Polybius [9.43; W. R. Paton (ed.), The Complete Histories of Polybius (2009) 307]; Dionysius Periegetes [976-982; Godfred Bernhard (ed.), Geographi Graeci Minores I (Lipsiae, 1828), 53-54]; Priscianus [Periegesis, 903-910; N.E. Lemaire (ed.), Poetæ Latini Minores IV (Paris, 1825) 381]; and Nicephorus Blemmydes [Τοῦ σοφωτάτου Νικηφόρου τοῦ Βλεμμίδου, Γεωγραφία Συνοπτικὴ [Karl Müller (ed.), Geographi Graeci Minores II (Paris, 1851), 466-467.]

[8] As proposed by Col. A. Boucher, L’Anabase de Xénophon (Paris 1913); C. F. Lehmann-Haupt, “Zum Rückzug der Zehntausend,” in J. Kro­mayer (ed.), Antike Schlachtfelder IV (Berlin 1931) 243–260; and V. Manfredi, La Strada dei Diecimila: topografia e geografia dell’Oriente di Senofonte (Milan 1986).

[9] As proposed by H. Kiepert [in Xenophons Anabasis erklärt von F. K. Hertlein mit einer Karte von H. Kiepert (Berlin 1857); and O. Lendle, Kommentar zu Xenophons Anabasis (Darmstadt 1995).

[10] W. Smith, Dictionary of Greek and Roman Geography I (London 1856) 7.

[11] See later this early view of d’Anville which influenced Rennell.

[12] Da­steira is in Pontos: Barrington Atlas 87 D4.

[13] T. J. Cornell (ed.), The Fragments of the Roman Historians, I (Oxford 2013) 1032-1033. In vol. III (Commentary) 612-613 of this work we read: Caranis itself, the centre of Caranitis, or rather Karenitis, is the present-day Erzerum, near which the shorter, northern branch of the Euphrates rises (Frat Su, Kara Su). The southern branch (Murat Su) emerges further east, near Diyadin. One question is which of these two branches each of the two authorities [Strabo and Pliny] meant; it would be satisfying to have a disagreement as to which branch was the true source of the river, rather than a more trivial difference as to where the water of one branch came from. For Mount Aga, in the Barrington Atlas placed south of Satala and west of Elegeia, rather west of Caranitis, Aba, from Ptol. Geog. 12.5.2 (βας ρος) has been suggested, while Mount Abos is said by Strabo 11.527 to be the source of the Araxes and the Euphrates. R. Syme, Anatolica (Oxford, 1995) 39 accepts Abos. Baumgartner, RE I, 107 identifies Abos with Bingöl Dağı, south of Erzerum; this would fit with the southern source of the river. Mucianus' description sub radicibus also fits the configuration of rivers at the southern edge of Bingöl Dağı, but the two should not be the same mountain, and Mucianus gives his a different name, Capotes. J. G. C. Anderson, Handy Classical Map of Asia Minor (London, 1912), places it north-east of Elegeia, but that may have been on the basis of this passage; nothing else is known. Pliny puts it twelve miles above Zimara, which is established on the south-flowing Euphrates in the southernmost part of Armenia Minor (RE 10A. 447, citing the Ant. Itin. 208.5 and Ptol. 15.7.2). Hence perhaps the siting of Capotes in the Barrington Atlas, west of the Euphrates. That is implausible as a source; possibly Sinara on one of the tributaries of the northern branch, attested otherwise only in late antiquity, is meant, but Zimara also occurs in the next sentence but one as a fixed point.

[14] Solin. 15; http://www.thelatinlibrary.com/solinus2a.html.

[15] Str. 11.14.2.

[16] Antonius Westermann (ed.), Stephani Byzantii θνικν quae supersunt (Lipsiae 1839) 127.

[17] Καρηντις, i.e. in the area of Erzurum; cf. Strab. 11.14.5.

[18] Cf. J. Rennell, Illustrations (Chiefly Geographical) of the History of the Expedition of Cyrus (London 1816) 221.

[19] La Cosmographie de Claude Ptolemée. Latin manuscript dated between 1401 and 1500. Bibliothèque-médiathèque de Nancy, ms. 354; on-line at bmn-renaissance.nancy.fr. Map 1a is from sheet 193 and Map 1b from sheet 197 of the manuscript. The respective web addresses are http://bmn-renaissance.nancy.fr/viewer/show/2483#page/n0/mode/1up and http://bmn-renaissance.nancy.fr/viewer/show/2485#page/n0/mode/1up.

[20] Ptolemy, Geog. 5.6.8: ὅθεν ὁ Μέλας καλούμενος ποταμός ῥέων συμβάλλει τῷ Εὐφράτποταμῷ.

[21] Barrington Atlas 64 E3, after T.B. Mitford, “Cappadocia and Armenia Minor: historical setting of the Limes”, ANRW II 7.2 (1980), 1161-1228 at 1185.

[22] K.F.A. Nobbe, Claudii Ptolemaei Geographia II (Leipzig 1845) 50.

[23] E. L. Stevenson, The Geography by Claudius Ptolemy (New York 1932) 124.

[24] Johannes Engels’ review in Aestimatio 8 (2011) 101–109 at 103.

[25] A. Stückelberger and G. Graßhoff, Ptolemaios Handbuch der Geographie, II (Bern 2006) 458.

[26] Leo Bagrow, R. A. Skelton, History of cartography2 (London, 2010) 37-38.

[27] From the Tabula Peutingeriana; on-line at peutinger.atlantides.org. The web address is http://peutinger.atlantides.org/map-a/. Click to launch viewer and zoom in the area shown in Map 2.

Cf. R. Talbert, Rome's World. The Peutinger Map Reconsidered (Cambridge 2010); M. Rathmann, “The Tabula Peutingeriana in the mirror of ancient cartography: Aspects of a reappraisal”, in K. Geus & M. Rathmann (eds.), The Survey of Oikumene. TOPOI Berlin Studies of the Ancient World, 14 (Berlin 2013) 203-222.

[28] Several modern scholars do not agree on the century in which Movsēs Xorenac‘i lived; thus it is safer to put a question mark after the 5th a.d. century.

[29] The Armenian Geography (Ašxarhac‘oyc‘) once attributed to Movsēs Xorenac‘i is now more likely to Anania Širakac‘i.

[30] Géographie attribuée a Moyse de Khoren, in Antoine-Jean Saint-Martin, Mémoires historiques et géographiques sur l'Arménie, II (Paris, 1818) 319-377: ‘[355] First Armenia is to the east of the first Cappadocia, and adjacent to the third Armenia; the Euphrates is the border to the east. … The second Armenia is to the east of Cappadocia, extending lengthwise to the Euphrates. … [359] Greater Armenia is to the east of Cappadocia and Lesser Armenia, on the banks of the Euphrates…. To the north it borders on Albania, Iberia and Colchis, until the place where the Euphrates turns toward the south’.

[31] Moïse de Khorène, Histoire d’Arménie, II (Paris, 1841) 159; Saint-Martin, Mémoires, I, 42: ‘When General Anatolius had received the orders of the Emperor [Theodosius], he came to our country and traveled in many areas, until finally he chose to lay the foundations of a town [Theodosiopolis] in the province of Garin, which is extremely fertile, abundant, well supplied with water, and seems to be in the middle of the plains, not far from the places of some of the sources of the Euphrates, which flows slowly in the beginning, increases in size as it proceeds, and forms a sort of a lake full of fish…’

[32] Procopius, Aed., 3.5: ‘These works he [Justinian] built in the Armenia which is on the right [west] of the Euphrates [Frat] River; and I shall go on to tell what was done by him in Greater Armenia [to the east of the river]. When Theodosius, the Emperor of the Romans, took over the dominion of Arsaces, as I have just related, he built on one of the hills a fort which was easy for assailants to capture, and he named it Theodosiopolis [Erzurum].’

[33] Procopius, Wars, 1.17: ‘From Tauric Armenia and the land of Celesene [Ἀκιλησηνή] the River Euphrates, flowing to the right of the Tigris, flows around an extensive territory, and since many rivers join it and among them the Arsinus, whose copious stream flows down from the land of the so-called Persarmenians…’

[34] Genesis 2.10-14: ποταμὸς δὲ ἐκπορεύεται ἐξ Ἐδὲμ ποτίζειν τὸν παράδεισον· ἐκεῖθεν ἀφορίζεται εἰς τέσσαρας ἀρχάς. ὄνομα τῷ ἑνὶ Φισῶν· οὗτος κυκλῶν πᾶσαν τὴν γῆν Εὐιλάτ, ἐκεῖ οὗ ἐστι τὸ χρυσίον· τὸ δὲ χρυσίον τῆς γῆς ἐκείνης καλόν· καὶ ἐκεῖ ἐστιν ἄνθραξ καὶ λίθος πράσινος. καὶ ὄνομα τῷ ποταμῷ τῷ δευτέρῳ Γεῶν· οὗτος κυκλῶν πᾶσαν τὴν γῆν Αἰθιοπίας. καὶ ποταμὸς τρίτος Τίγρις· οὗτος προπορευόμενος κατέναντι Ἀσσυρίων. δὲ ποταμὸς τέταρτος Εὐφράτης. Cf. Eusebius (in Lagarde, Onomastica Sacra, 252); Hieronymus (in Lagarde, Onomastica Sacra, 5); and Theodoret of Cyrus, The Questions on the Octateuch, I (Washington, DC, 2007) 66-69.

[35] Saint-Martin, Mémoires, II, 398-400: ‘The Euphrates, pouring also out of the Paradise, comes down to earth and issues from Armenia, towards the land of Pakrevant and towards the mountains of Garin, which are adjacent to the city of Ardzn [Erzurum]. … The Aradzani [Arsanias] issues from the central part of Armenia and joins the Euphrates…’

[36] The Epic Histories Attributed to Pawstos Buzand; translation and commentary by Nina G. Garsoian (Cambridge Mass., 1989).

[37] Pawstos Buzand, Epic Histories, 3.4, 4.14, 5.3, 5.4, 5.43, 6.16.

[38] Saint-Martin, Mémoires, I, 51. According to Pawstos Buzand, Epic Histories, 4.14, Hayr mardpet [chief eunuch] descended to the banks of the Euphrates, where in ancient times Sanatruk the king had built the city named Mcurn.’

[39] Nina G. Garsoian, “The Early-Mediaeval Armenian City: An Alien Element?” in Church and Culture in Early Medieval Armenia (London, 1999), 67-83 at 72.

[40] Nina G. Garsoian, “The Early-Mediaeval Armenian City”, 73.

[41] See E. Guidoboni, G. Traina, “A new catalogue of earthquakes in the historical Armenian area from antiquity to the 12th century”, Annals of Geophysics, 38.1 (1995) 85-147 at 112-113. H.A. Manandian, The Trade and Cities of Armenia in Relation to Ancient World Trade, (Lisbon, 1965) 82-84, noting that Movsēs Xorenac‘i identifies Mcurn with Mcbin-Nisibis, rebuilt by Sanatruk after a disastrous earthquake, says that “the information as to the earthquake and the destruction of Nisibis was probably invented by Movses himself.” The fifth-century (?) a.d. Armenian historian Movsēs Xorenac‘i (Moïse de Khorène, Histoire d’Arménie, I, 237) thinks that Mcurn is Mcbin-Nisibis and reports the reconstruction of this town by Sanatruk. It seems that this is confirmed by the seventh-century a.d. historian Sebeos (The Primary History of Armenia, Chapter 1, on-line at rbedrosian.com), who mentions a palace of king Sanatruk at Mcbin-Nisibis. Saint-Martin, Mémoires, I, 51, discussed later in this article, places Mcurn to the north of Malazgirt (Manzikert) and the Murad. In H.A. Manandian, The Trade and Cities of Armenia, 83-84, “according to Markwart, Mcurn stood near the Arsanias and its left tributary the Kara-su.” Tubinger Atlas des Vorderen Orients, B VI 14: Armenien und Georgien. Christentum und Territorialentwicklung vom 4. bis 7. Jahrhundert, edited by R. H. Hewsen and Ch. Bandomer (Wiesbaden, 1987), places Mcurn at Hosnek, to the west of Mazgirt and to the east of Keban.

[42] Emmanuel Bekker (ed.), Theophanes Continuatus (Οἱ μετΘεοφάνην) 5.40 (Bonn, 1838) 269: οτω δτν Εφράτην περαιωθες τφρούριον εθς ὅ Ῥαψάκιον λέγεταιξεπόρθησεν. ἰδίδτος Χάλδους καΚολωνιάτας τν μεταξχώραν Εφράτου καὶ Ἀρσίνου κελεύσας καταδραμεν

[43] Abu Firas al-Hamdani (932–968) was captured twice and spent some time in Constantinople, where he composed the collection of poems titled al-Rumiyyat (the poem of the Rum, i.e. Romans, Byzantines).

[44] James Howard-Johnston, East Rome, Sasanian Persia and the End of Antiquity (Aldershot 2006) 251.

[45] Ibn Serapion, who was probably of Persian origin, produced a work on mathematical geography, the Kitāb ̒Ajā’ib al-aqālīm al-sab̒a [Book of the Wonders of the Seven Climes]. Cf. C. E. Bosworth & M. S. Asimov, History of civilizations of Central Asia, IV (Delhi 2003) 217.

[46] According to Guy Le Strange, The Lands of the Eastern Caliphate (Cambridge 1905), the name is written Afradkhis by Masudi. We see in the next page that the French translation of Masudi calls the mountain Afradohos, and Golius, Notae in Alfergan (Amsterdam, 1669) 247 calls it Fredaches.

[47] Guy Le Strange, “Description of Mesopotamia and Baghdad, written about the year 900 a.d. by Ibn Serapion,” The Journal of the Royal Asiatic Society (London, 1895) 1-76 at 46-47.

[48] Strabo’s 11.14.5 Taronitis (Ταρωντις). The Muş Province of Turkey.

[49] Guy Le Strange, “Description of Mesopotamia” at 54.

[50] ʿAlī ibn al-usayn ibn ʿAlī al-Masʿūdī, ca.896–956.

[51] Maçoudi, Les Praires d'or, I (Paris, 1861) 214.

[52] J. Williams, Two Essays on the Geography of Ancient Asia (London 1829) 291, says that “Calicala, as proved by Golius [see later in this article], is Erze-Rom, the great bulwark of Adherbijan and Armenia, in the days of Edrisi”. Cf. Idrisi’s text and map below, and Hadji Khalifa in n.72.

[53] Fredaches in Golius, Notae in Alfergan (Amsterdam, 1669) 247.

[54] Abū ʿUbayd ʿAbd Allāh ibn ʿAbd al-ʿAzīz ibn Muammad ibn Ayyūb ibn ʿAmr al-Bakrī, c. 10141094.

[55] Guy Le Strange, “Description of Mesopotamia” at 56.

[56] Abu Abd Allah Muhammad al-Idrisi al-Qurtubi al-Hasani al-Sabti, 11001166.

[57] Géographie d'Édrisi, II (Paris, 1840) 137; cf J. Williams, Two Essays on the Geography of Ancient Asia (London 1829) 291. Idrisi means here to Baghdad via canals towards the Tigris because the Euphrates does not pass from Baghdad.

[58] Géographie d'Édrisi, II, 314.

[59] On-line at the Library of Congress (memory.loc.gov), in Konrad Miller's 1927 consolidation and transliteration. The web address is https://www.loc.gov/resource/g3200.ct001903. Zoom in the area shown in Map 3.

[60] Yāqūt ibn-'Abdullah al-Rūmī al-Hamawī, 1179–1229.

[61] Dictionary of Countries.

[62] Guy Le Strange, Collected Works, I (Croydon, 2014) 115.

[63] Abu al-Fida or Abu Al-fida' Isma'il Ibn 'ali ibn Mahmud Al-malik Al-mu'ayyad 'imad Ad-din, 1273-1331.

[64] Géographie d’Aboulféda, II (Paris, 1848) 64-66.

[65] Rum-kale, a powerful fortress on the Euphrates, 50 km west of Şanlıurfa.

[66] Currently at the British Library; on-line at prints.bl.uk. The web address is http://prints.bl.uk/art/581002/world-map-by-martellus. Click on the map to enlarge.

[67] Tabula Prima Asie; one of the 27 woodcut maps by Waldseemüller, illustrating the Strasbourg 1513 print of Ptolemy’s Geography; on-line at zvab.com. The web address is http://www.zvab.com/servlet/FrameBase?content=/de/imagegallery/imagegallery.shtml?images=http://pictures.abebooks.com/GOETZFRIED/17601222156.jpg.

[68] Il Vero Disegno della Natolia, et Carmania, con gli confini della Soria, Romania &c. dell’ Arcipelago di Giacomo Gastaldo cosmografo Venetiis MDLXX; on-line at raremaps.com. The web address is http://www.raremaps.com/gallery/detail/24533/Il_Vero_Disegno_della_provincia_della_Natolia_et_Carmania_con_gli/Gastaldi-Zaltieri.html.

[69] Abraham Ortelius, Natoliae, Quae Olim Asia Minor, Nova Descriptio. Antwerp, Platin 1572/73; on-line at raremaps.com. The web address is https://www.raremaps.com/gallery/detail/0060gh/Natoliae_Quae_Olim_Asia_Minor_Nova_Descriptio_with_Aegypti_Recentior/Ortelius.html.

[70] Sonja Brentjes, “Patchwork: The Norm of Mapmaking Practices for Western Asia in Catholic and Protestant Europe as well as in Istanbul between 1550 and 1750?” in Feza Günergun and Dhruv Raina (eds.), Science between Europe and Asia: Historical Studies on the Transmission, Adoption, and Adaptation of Knowledge (Dordrecht, 2011) 77-102 at 93.

[71] Even in the following centuries they believed that the Bingöl Su (which is a tributary of the Araxes) was a tributary of the Euphrates (Frat, Karasu). See later in this article.

[72] Natoliae sive Asia minor. Bibliothèque nationale de France; on-line at gallica.bnf.fr. The web address is http://gallica.bnf.fr/ark:/12148/btv1b8592785x/f1.item.r=hondius%20Natoliae%20sive%20Asia%20minor.

[73] Natolia sive Asia Minor; on-line at bergbook.com. The web address is https://www.bergbook.com/images/20853-01.jpg.

[74] Natolia quae Olim Asia Minor; on-line at turkeyinmaps.com. The web address is http://www.turkeyinmaps.com/Large/Blaeuanat.html.

[75] Michel Servet (1535), Asiae Tabula III: Colchis, Iberia, Albania, Armenia major. Bibliothèque nationale de France; on-line at gallica.bnf.fr. The web address is http://gallica.bnf.fr/ark:/12148/btv1b5963983m/f1.item.r=Michel%20Servet%20Asiae%20Tabula%20III.

[76] Gerardus Mercator (1579), Tabula Asiae III. On-line at Wikimedia. The web address is https://upload.wikimedia.org/wikipedia/commons/d/d4/Gerard_Mercator._Tabula_Asiae_III_%28Armenia%2C_Georgia%2C_Turkey%2C_etc.%29._1579.jpg.

[77] Nicolas Samson (1655) Armenia maior, Colchis, Iberia, Albania / ex conatibus geograph. Bibliothèque nationale de France; on-line at gallica.bnf.fr. The web address is http://gallica.bnf.fr/ark:/12148/btv1b5963985f.r=Sanson%2C+Nicolas+Armenia+maior.langFR.

[78] Christopher Weigel (1720), Armenia Vtraque, in Scena Historiarum Orientaliss Quinti Seculi; on-line at geographicus.com. The web address is http://www.geographicus.com/P/AntiqueMap/Armenia-weigel-1720.

[79] Henri Liébaux (1729), Carte de la Colchide, de l'Ibérie, de l'Albanie et des pays circonvoisins. Bibliothèque nationale de France; on-line at gallica.bnf.fr. The web address is http://gallica.bnf.fr/ark:/12148/btv1b59639915.

[80] Notae in Alfergan, 249. Golius (1596-1667), professor of mathematics and oriental languages at Leiden, Holland, translated and commented extensively the Introduction to Astronomy, written by Alfergan, a ninth-century Arab astronomer. His work Muhammedis, filii Ketiri Ferganensis, qui vulgo Alfraganus dicitur, elementa astronomica Arabice et Latine, Opera Jacobi Golii (Amsterdam, 1669) is usually cited as Notae in Alfergan. Concerning the upper Euphrates, one might assume that there would be interesting information in the Notae in Alfergan, moreover because d’Anville’s initial views in the eighteenth century were thus criticized with respect to Golius: ‘Almost everything that is valuable in the Tigris and Euphrates of d'Anville has been extracted from Golius, and that what is wrong is d'Anville's own.’ (Williams, Two Essays, 291).

[81] Gihan Numa Geographia Orientalis, ex Turcico in Latinum versa a Matthias Norberg (Londini Gothorum, 1818). Londini Gothorum is Lund, Sweden. In the preface of this Latin edition we read: ‘Published, and indeed edited, by Ibrahim Efendi, the year 1145 A.H., 698 pages in Constantinople folio, as it is called’ (…typisque, et quidem curante Ibrahim Efendi, anno Hegirae 1145, pag. 698 in folio Constantinopoli id vulgatum). Anno Hegirae 1145 is the year 1732 a.d., and Ibrahim Efendi is the editor, Ibrahim Müteferrika.

[82] Gihan Numa Geographia Orientalis, I, 629-630.

[83] Hereafter we call ‘small’ Karasu the local stream of Muş, in order to distinguish from the Karasu (Frat, Euphrates) rising in the area of Erzurum.

[84] T.A. Sinclair, Eastern Turkey: An Architectural & Archaeological Survey (Irthlingborough 1989) I, 96: Erzurum was known to the Arabs as Kalikala.

[85] In Gihan Numa, II, 413, Hadji Khalifa mentions the river Kirk Getjid on the road from Sivas to Malatya, to the south of Alacahan (Alage Chan). It is the river Tohma, the Melas (Μέλας) of the ancients. See next footnote.

[86] According to d’Anville, L’Euphrate et le Tigre (Paris, 1779), 5, the Euphrates “exiting from Melitene … and immediately below the confluence of the Melas [Black in Greek] which retains its name as the Kara-su [the river Tohma] in Turkish language, a very narrow pass, opened by the river, is called the Nushar Passage in the Geography manuscript translated from Turkish to French.” Thus d’Anville knew of and had used the manuscript of Hadji Khalifa, because in its Latin edition (Gihan Numa Geographia Orientalis, II, 19) we read: “But opposite to the passage of Euphrates, which is called Nuschar, and between the two sides is the jurisdiction called Siverek.”

[87] Cf. Movsēs Xorenac‘i earlier in this article.

[88] The map from a Ibrahim Müteferrika edition is on-line at parssea.org and Wikimedia. The web addresses are resopectively http://parssea.org/wp-content/uploads/2013/04/Basra_bahrefars.jpg and https://commons.wikimedia.org/wiki/File:Basra_bahrefars.jpg.

[89] The vague translation of the original text here means that, after the confluence, the Murad assumes the name of the Euphrates. A different interpretation is not justified by the preceding and the following text. Also, no one refers ever to the united stream after the confluence as Murad.

[90] Cf. Hadji Khalifa’s statement above.

[91] Narrative of Travels in Europe, Asia, and Africa in the Seventeenth Century by Evliya Efendi, translated by Joseph Von Hammer (London, 1850) II, 110.

[92] P. Duval, Tabula itineris Decies Mille Graecorum sub Cyro contra fratrem suum Artaxerxem regem Persarum; Bibliothèque nationale de France; on-line at gallica.bnf.fr. The web address is http://gallica.bnf.fr/ark:/12148/btv1b530367231.

[93] Pierre Duval, La géographie universelle (Paris, 1676); De la Turquie en Asie; 183-207 at 194-195.

[94] Pierre Duval, La Turquie en Asie in his Géographie universelle (1676).

[95] R. Forster, “Geographical Dissertation”, in The Whole Works of Xenophon, translated by Ashley Cooper, Spelman, Smith, Fielding and others (Philadelphia, 1843) 259-280, at 270: “[Delisle] had discovered from Ptolemy that the most easterly branch of the Euphrates [the Murad] rised fifty leagues to the southeast of the springs [of the Frat/Karasu] above Erzerum.”

[96] Guillaume de Lisle (1701), Turquie, Arabie, Perse. The David Ramsey Map Collection; on-line at davidrumsey.com. The web address is http://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~2901~300037:Turquie,-Arabie,-Perse-?sort=Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No&qvq=q:de%2BLisle%2BTurquie%2C%2BArabie%2C%2BPerse;sort:Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No;lc:RUMSEY~8~1&mi=1&trs=2.

[97] Cf. Ptolemy, Geog. 5.6.8 and n.20 above.

[98] “Détermination géographique de la situation et de l'étendue des pays traversés par le jeune Cyrus dans son expédition contre son frère Artaxercès, et par les dix mille Grecs dans leur retraite”; par M. Delisle. Mémoirs de l'Academie royale des sciences pour 1721 (Paris, 1723) 56-68 and map.

[99] The web address is http://gallica.bnf.fr/ark:/12148/bpt6k35238/f184.image.r=.

[100] Delisle, Détermination géographique, 65.

[101] Guillaume de Lisle (1723), Retraite des Dix Mille Tabula. The David Ramsey Map Collection; on-line at davidrumsey.com. The web address is http://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~2902~300038:Retraite-des-Dix-Mille-Tabula-?sort=Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No&qvq=q:delisle%2BRetraite%2Bdes%2BDix%2BMille%2BTabula;sort:Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No;lc:RUMSEY~8~1&mi=0&trs=1.

[102] Probably O. Lendle, Kommentar zu Xenophons Anabasis (Darmstadt 1995) realized the error in having the Greeks marching to the east, towards the sources of the Murad, in the area of Bayazid and Diyadin, followed at a distance by the satrap of Western Armenia (Tirivazos), marching not in Western Armenia but in the Armenia of Orontas. He proposes a march to the north from the area of Bulanik, but he does not avoid the mistake. First, he has the Greeks crossing the Murad as ‘Euphrates’ in a place where there is no evidence that the river was fordable (cf. Kinneir later in this article); and, second, he has them crossing the Murad as ‘Euphrates’ not near its sources (according to Xenophon) but 200 km downstream. Nevertheless, Lendle’s approach does away with an incomprehensible retreat from the area of Muş towards the southeastern extremity of modern Turkey, but by avoiding altogether the eastern loop of the retreat, leaves unexplained gaps in the chronology and the parasangs of the Anabasis. Cf. Paradeisopoulos 2014, 226-234.

[103] Guillaume de Lisle (1731), Alexandri Magni Imperium et Expeditio. The David Ramsey Map Collection; on-line at davidrumsey.com.

[104] See n.85 above.

[105] Vivien de Saint-Martin, Description historique et géographique de l'Asie Mineure II (Paris, 1852), 751: “Djihan Numa, or the Mirror of the World, is a Universal Geography written in Turkish in the year 1648, printed for the first time in Constantinople in 1732 and translated into French by Armain in 1744 from this edition of Constantinople. Armain’s translation, same as the Constantinople edition from which it was made, includes only Asia, which forms the first part of Djihan Numa. This part is in fact the only that has interest for us. Hadji-Khalfa has described in great detail the various Asian provinces of the Turkish Empire, including Asia Minor. This description of Asia Minor is printed for the first time on the translation of Armain in the Appendix to this volume.” Vivien de Saint-Martin means ‘printed for the first time’ in French, because a Latin translation was printed thirty-four years earlier (1818).

[106] Compendium of Ancient Geography by Monsieur D'Anville; translated from the French, I (London 1791) 356-357.

[107] Jean Baptiste Bourguignon d' Anville (1764), Orbis Romani, pars Orientalis. The David Ramsey Map Collection; on-line at davidrumsey.com. The web address is http://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~3041~410040:Orbis-Romani,-pars-orientalis-?sort=Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No&qvq=q:Anville%2C%2BOrbis%2BRomani%2C%2Bpars%2BOrientalis;sort:Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No;lc:RUMSEY~8~1&mi=0&trs=2.

[108] Jean Baptiste Bourguignon d' Anville (1794), Turkey in Asia. The David Ramsey Map Collection; on-line at davidrumsey.com. The web address is http://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~31562~1150041:Turkey-in-Asia-?sort=Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No&qvq=q:Anville%2C%2BTurkey%2Bin%2BAsia;sort:Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No;lc:RUMSEY~8~1&mi=1&trs=3.

[109] Dictionnaire historique, géographique, philologique et critique pour servir d’annotations a l’Histoire d’Arménie. In Moïse de Khorène, Histoire d’Arménie, II (Paris, 1841) 75. The numbering of the pages in this Dictionary starts again from 1, after the end of Xorenac‘i’s History in this volume.

[110] Coup d’Oeil sur l’Arménie ou Géographie Sommaire précis de l’Histoire d’Arménie. In Moïse de Khorène, Histoire d’Arménie, II (Paris, 1841) 6. The numbering of the pages for this Géographie Sommaire in the volume starts again from 1.

[111] Article by F. Weissbach, in the Realencyclopädie der Classischen Altertumswissenschaft (RE) VI.1 1907, 1195 to 1215.

[112] J. Rennell, Illustrations (Chiefly Geographical) of the History of the Expedition of Cyrus (London 1816)]. See Paradeisopoulos, 2014, 226 note 26.

[113] M.D. Kinneir, Journey through Asia Minor, Armenia, and Koordistan (Lon­don 1818).

[114] Liz (Kinneir’s Leese, and Liz in Lendle’s Kommentar) was renamed Erentepe after 1928. Cf. Index Anatolicus, Erentepe-Bulanık-Muş.

[115] James Brant and A. G. Glascott, “Notes of a Journey through a Part of Kurdistán, in the summer of 1838,” JRGS 10 (1840) 341–434, at 345–349. Viscount Pollington, “Notes on a Journey from Erz-Rúm, by Músh, Diyár-Bekr, and Bíreh-jik, to Aleppo, in June, 1838,” JRGS 10 (1840) 445–454, at 445–447. The Penny Cyclopædia 25 (London 1843) 472. H.F.B. Lynch, Armenia: Travels and Studies II (London 1901) 175–176.

[116] The Sulukh bridge to the north of Muş, a fourteenth-century “ancient bridge of fourteen arches” (Brant and Glascott, “Notes of a Journey”, 349).

[117] Aaron Arrowsmith (1813), Map of the Countries lying between the Euphrates and Indus. The David Ramsey Map Collection; on-line at davidrumsey.com. Map from the book John Macdonald Kinneir, A Geographical Memoir of the Persian Empire, accompanied by a Map (London, 1813). It is interesting that the publication of this map (1813) precedes Kinneir’s journey in the area (1814). See n.111. The web address is http://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~233572~5509711:Composite—Map-of-the-Countries-lyi?sort=Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No&qvq=q:Map%2Bof%2Bthe%2BCountries%2Blying%2Bbetween%2Bthe%2BEuphrates%2Band%2BIndus;sort:Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No;lc:RUMSEY~8~1&mi=0&trs=5.

[118] The two rivers before the confluence have lengths of ca. 650 and 450 km respectively; cf. Matthias Finger et al. (eds.), The Multi-Governance of Water: Four Case Studies (N. York, 2006) 103.

[119] Until the confluence at Mainz, the Main (529 km) is longer than the Rhine (ca. 485 km from the Lake of Constance to Mainz via Basel). Also until the confluence at Passau, the Inn has greater average flow, and it has almost the same length (517 km) with the Danube (553 km from its sources at Donaueschingen, until Passau).

[120] E.H. Bunbury, History of Ancient Geography, I (London, 1879) 353 and n.1.

[121] H. Tozer, Turkish Armenia and Eastern Asia Minor (London, 1881) 207 n.1

[122] Eli Smith, Missionary Researches in Armenia, I (London 1834), 55.

[123] Mack Chahin, The Kingdom of Armenia (Abingdon 2001) 45-46: “In 1823, a French scholar, Jean Saint-Martin, chanced upon a passage in the History of Armenia by the fifth century a.d. Armenian historian, Moses of Khoren which aroused his curiosity. … On his suggestion, the French government commissioned a young German professor, Friedrich Eduard Schultz, to go to Van to investigate. At Toprak Kale, a hill about 400 feet high near Van, Schultz discovered, carved upon the smooth wall of an ancient citadel, long cuneiform inscriptions, the Vannic inscriptions. … Although he could not read cuneiform, Schultz was able to make precise copies to take back with him to Paris. … Schultz’s copies of the Vannic inscriptions were published in Paris, in the Journal asiatique, in 1840. We now know that the modern town of Van is on the site of Tushpa, the capital of the Kingdom of Urartu (or Ararat).”

[124] Cf. A General Bibliographical Dictionary from the German of Frederic Adolphus Ebert, IV (Oxford, 1837) 1675.

[125] Saint-Martin, Mémoires, II, 56-176.

[126] Illuminator newsletter, St. Gregory’s the Illuminator Armenian Apostolic Church of Philadelphia, July-August 2012, 6: “The pioneer of the Armenian printing in India was Shahamir Shahamirian [Saint-Martin’s Schamir] who with his son Hagop opened a publishing house in Madras in 1772. The first book he printed was of a political nature, … called ‘Nor Tetrag vor Gochi Hortorag’ (New Booklet called Exhortation.)’” The article contains also a photo of the first page of Schamir’s ‘Exhortation’ in Armenian.

[127] The Géographie du Vartabied Vartan (1721) is available because it was translated and published in Saint-Martin, Mémoires, II, 407-453.

[128] The first in Saint-Martin, Mémoires, I, 46, initiated this examination.

[129] The only reference to Mount Dzaghgé in Saint-Martin’s Index (II, 497) is to the text above (I, 50).

[130] Dzaghgodn was a canton in the province of Ararad according to the Geography of Vartabied Vartan (1721) in Saint-Martin, Mémoires, II, 367.

[131] The Penny Cyclopædia 25 (London 1843) 471B mentions the junction of the Murad with the Char-buhur river “at about 39o N. lat., 41o30’ E. long.”

[132] Footnote (1) in I, 51 reads: “Jean Patr. Hist. d’Arm., ch.8, p.55 (ms. Arm. no.91). See hereafter the Geography of Vartan. Djiahn-numa, p.426, 427”.

[133] Histoire d'Arménie, par le patriarche Jean VI, dit Jean Catholicos, traduite de l'arménien en français par M. J. Saint-Martin; ouvrage posthume (Paris, 1841).

[134] The Histoire d'Arménie, par le patriarche Jean VI, at 29 (which is page 55 of the manuscript), in its English translation by Rev. Fr. Krikor Maksoudian (on-line at rbedrosian.com) reads: “Forty-three years after the death of the blessed apostle Thaddeus, in the days of Artashes king of Armenia, the holy apostle's disciples who lived at the sources of the Euphrates river and whose leader was called Oski proselytized and baptized certain Alans who were related to queen Sat'enik wife of Artashes. Because the baptized became day by day more resolute in their faith in [Christ] the Word of Life, the son of Sat'enik was irritated by them, and put the blessed Oski and his saintly companions to the sword.”

[135] Saint-Martin’s reference to Osgik'h in his Index (II, 510) is to the Geography of Vartabied Vartan (1721) in Saint-Martin, Mémoires, II, 435 where it is called Saint-Oski.

[136] The association of Medzourk'h with the canton of Mendzour is unreferenced. The only reference to Medzourk'h in Saint-Martin’s Index (II, 508) is to the passage above (I, 51). As for the canton of Mendzour, Saint-Martin’s reference in his Index (II, 508) is to the Geography attributed to Movses Khorenatsi (Géographie Attribuée a Moyse de Khoren) in Saint-Martin, Mémoires, II, 319-377 at 361. See below n.137 and Map 15.

[137] Cf. N. Adontz, Armenia in the Period of Justinian. The Political Conditions based on the Naxarar System II (Lisbon 1970; orig. 1908) 210 for Malazgirt (Mazgirt, Manazkert) and Manazkert (Manzikert, Malazgirt).

[138] According to the Encyclopaedia of Islam, Second Edition (Leiden, 1960-2005), Bingöl Dağ is the name of a mountain massif which stretches south of Erzurum across the vilayets of Erzurum, Mush and Bingöl (see Map 15). This has nothing to do with Malazgirt (Manzikert).

[139] According to the Geography attributed to Movses Khorenatsi (Géographie Attribuée a Moyse de Khoren) in Saint-Martin, Mémoires, II, 319-377 at 361, the Upper Armenia contained the following nine provinces; descriptions come from Alan Sinclair, Eastern Turkey: An Architectural and Archaeological Survey, II (Irthlingborough 1989): Taranaghi (to the west of the Frat, Euphrates), Arhioudz, Mendsour, Egeghéats (Acilesene, the area of Erzincan), Mananaghi (the valley of the river Tuzla), Terdchan (Tercan), Sber (Ispir), Schaghagom (the high district between the Bayburt and the Aşkale-Erzurum plains) and Garin (Erzurum). The borders of these cantons in Map 15 were taken from a map in Armenian available on-line at rbedrosian.com.

[140] The fourth-century a.d. canton of Mendsour in Upper Armenia became consequently the canton Muzuron (κλίμα Μουζουρῶν) in the Roman Empire of the early seventh century (Georgii Cyprii descriptio orbis Romani [Leipzig, 1890] 49).

[141] This mistake is repeated in Saint-Martin’s posthumus Fragments d'une Histoire des Arsacides I (Paris 1850) 138: “The places towards the sources of the Euphrates were under the domination of Sanadroug, according to Faustus of Byzantium; this prince founded the city of Medzourk'h in the mountainous and covered by forests cantons, in the area of the sources of the southern part of Euphrates, currently named Mourad-chai.”

[142] Oghnagan, Meghdi, Meghraked, and Aldzsan, cannot be found in any of the Armenian texts incorporated into Saint-Martin’s two volumes. Indeed Oghnagan and Aldzsan are not even contained in his Index, whereas the only reference in this Index to Meghdi and Meghraked (II, 508) is the page of the above passage (I, 52).

[143] Mamouschegh and Mamousch are unreferenced. The only reference in Saint-Martin’s Index (II, 507) is to the page of the above passage (I, 52).

[144] The fact that Saint-Martin had read d’Anville’s views is evident in his passage (I, 44), where he criticizes d’Anville’s detachment of Caranitis from Theodosiopolis (Erzurum), and its displacement to the east, towards the sources of the Murad (cf. Map 12 above): “The learned d’Anville was therefore wrong to think that Pliny wished to mention the lower part of the Euphrates, and he accordingly erred in placing the Caranitis region in the central part of Armenia, to the sources of the other Euphrates.”

[145] For the career of Orontas, see R. D. Wilkinson, “Orontes, Son of Artasyras”, Revue des études arméniennes (VII, 1970) 445-50; and M. J. Osborne, “Orontes”, Historia (22, 1973) 515—51.

[146] The derivation of the dates is discussed in Paradeisopoulos 2013.

[147] According to A. Schachner, “Xenophons Überquerung des Kentrites: ein archäologischer Nachtrag”, Istanbuler Mitteilungen 58 (2008), 411-417 (with H. Sağlamtimur): Map 5 at 414, the point of crossing the Centrites was 2 km to the north of Çattepe. The distance from this point of crossing to Siirt is 31 km, i.e. 5.4 parasangs at 5.768 km per parasang; cf. Paradeisopoulos 2013, 655.

[148] The discrepancy between the distance reported by Xenophon and the actual distance of Siirt and Bitlis (99 km or 17.2 parasangs) has resulted in proposals for a direct northerly march from Siirt to Muş. Layard 1853: 32-34 objects this alternative; cf. Paradeisopoulos 2013, 655. However, Kinneir travelled from Bitlis to Siirt not along the path of the principal road, but along a more direct north-south route. He passed from Aşağıölek (Lower Ölek) which is his Eulak; from Şeyhcuma (his Scheck Jama), renamed Çeltikli after 1928 (Index Anatolicus, Çeltikli-Bitlis); from Epr (his Serpa), renamed Dikme after 1928 (Index Anatolicus, Dikme-Bitlis); from Ürek/Örek (his Eurak), renamed Gözlüce after 1928 (Index Anatolicus, Gözlüce-Şirvan-Siirt); from Tasil, renamed Karaca köy after 1928 (Index Anatolicus, Karaca-Şirvan-Siirt); from Miskin (his Tiskin), renamed Soğanlı after 1928 (Index Anatolicus, Soğanlı-Şirvan-Siirt); from Helesan (his Halasni), renamed Kapılı.

[149] The distance between Bitlis and Muş is 14.4 parasangs (83 km); cf. Paradeisopoulos 2013, 655.

[150] Frequently Tirivazos is considered a subordinate of Orontas, who was the satrap of all Armenia. Cf. Bruno Jacobs, Die Satrapienverwaltung im Perserreich zur Zeit Darius' III. Beihefte zum Tübinger Atlas des Vorderen Orients. (Wiesbaden 1994); also his Achaemenid satrapies (on-line at iranica.com).

[151] A. Boucher, L’Anabase de Xénophon (Paris 1913) ; C.F. Lehmann-Haupt, “Zum Ruckzug der Zehntausend,” in J. Kro­mayer (ed.), Antike Schlachtfelder IV (Berlin 1931) 243–260; V. Manfredi, La Strada; Lendle, Kommentar; etc.

[152] Jacobs, Achaemenid satrapies (on-line at iranica.com) defines thus the borders of the Achaemenid satrapy of (East) Armenia governed by Orontas. Its southern border was the river Centrites/Bohtan. Beyond the junction of the Centrites and the Tigris, the southern border continued westward to meet the Euphrates [see Map 16]. The Euphrates was the western border of (East) Armenia. The northern border of this satrapy was marked first by the Murad [from the point of its confluence with the Euphrates up to the point of its confluence with the ‘small’ Karasu], and then by the ’small’ Karasu (Teleboas). After the ‘small’ Karasu, from the northeastern corner of Lake Van, the northern border ran to the north, toward the Caucasus. This means (see Map 16) that in the intermediate part, i.e. from the sources of the ‘small’ Karasu until the northeastern corner of Lake Van, the northern border of (East) Armenia was formed by the northern shore of this lake.

Similarly, according to Jacobs the east and south sides of the West Armenia of Tirivazos fit into the angle left open by the satrapy of (East) Armenia. In the north, the Caucasus was the border of West Armenia, and in the northwest the Black Sea. North of the confluence of the Euphrates (Frat, Karasu) and the Murad, the western border of West Armenia ran north-northwest toward the Black Sea, reaching the coast immediately west of Cotyora.

[153] Cf. Paradeisopoulos 2013, 655-659; 2014, 227-234.

[154] Before the conversion of the Murad into the ‘Euphrates’, Xenophon’s Teleboas was thought of to be synonymous with the Arsanias (Murad); cf. William Hazlitt, Classical Gazeteer, Dictionary of Ancient Geography, Sacred and Profane (London, 1851) 51: “Arsanius (Arsanias, Arsametus, Teleboas), a river of Armenia Major, rising in Mount Abus, and falling into the Euphrates from the left, below Arsamosata”; cf. Duval’s (c. 1653) Map 8 above.

[155] Cf. Paradeisopoulos 2014, 227-228.

[156] H. Kiepert, Asia Citerior, on-line at davidrumsey.com. Kiepert, a prominent cartographer of the ancient world, proposed that the Ten Thousand crossed the Arsanias/Murad as ‘Euphrates’ far from its sources. This view of Kiepert (1857) was repeated later by Lendle (1995). The web address is http://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~34460~1180072:Asia-citerior-.

[157] Paradeisopoulos 2014, n.38 in 230-231.

[158] Lynch, Armenia, II, 175–176.

[159] According to Lendle, Kommentar, 229, it is unlikely that Tirivazos really planned a raid on the Greeks; cf. Anab. 4.4.21: οἱ δὲ βάρβαροι ἀκούσαντες τὸν θόρυβον οὐχ ὑπέμειναν, ἀλλ᾽ ἔφευγον.

[160] Anab. 4.5.1: συσκευασάμενοι δ᾽ εὐθὺς ἐπορεύοντο διὰ χιόνος πολλῆς ἡγεμόνας ἔχοντες πολλούς.

[161] Paradeisopoulos 2013, 655-659; 2014, 227-234.

[162] Cf. Paradeisopoulos, 2013 Map 2 (669); Paradeisopoulos 2014 Map 3 (227), Map 4 (229), Map 5 (233).

[163] We have proposed 27 February 400: Paradeisopoulos 2013: 659.

Information about this Article
This Article has not yet been peer-reviewed
This Article was published on 9th March, 2016 at 02:26:29 and has been viewed 1364 times.

Creative Commons License
This work is licensed under a Creative Commons Attribution 2.5 License.
The full citation for this Article is:
Paradeisopoulos, I. (2016). Changing names, fitting views: The conversion of river Arsanias into the Euphrates of Xenophon. PHILICA.COM Article number 582.

<< Go back Review this ArticlePrinter-friendlyReport this Article

Website copyright © 2006-07 Philica; authors retain the rights to their work under this Creative Commons License and reviews are copyleft under the GNU free documentation license.
Using this site indicates acceptance of our Terms and Conditions.

This page was generated in 0.6034 seconds.