The World of the Huns. Chapter IX. Language
O. Maenchen-Helfen

10. Tribal names


The literature about the name of the Hunnic people, which in Priscus occurs as  and , and in Jordanes as Acatziri, is extensive. Tomaschek was the first to suggest a Turkish etymology, which has won wide acceptance; he thought Acatziri was agac-ari, "forest men." [463] This etymology seems to be supported by Agaj-eri in the Turko-Arabic dictionary of 1245 [464] and in Rashid-'d Din, who refers to the Mongol synonym hoi-in-irgan. [465] Sinor [466] called attention to yis-kisi, as some Turks in the Altai are named; it, too, means "people of the wood." The names of the Russian Drevlyane and the Gothic Tervingi in the Ukraine have often been adduced as parallels to agac-ari. The Drevlyane are said to have received their name "because they lived in the woods," [467] and Tervingi is supposed to have the same meaning:—"forest man." [468]

The Turkish etymology was rejected by F. W. K. Müller, Henning, and Hamilton. Müller [469] maintained that agac means "tree," not "wood,


forest." Henning [470] regards the usual derivation of the name as "scarcely better than a popular etymology." Hamilton [471] finds Agaj-eri as strange a name as Qum-eri, "man of sand," Turuk-eri, "Turk," or Rum-eri, "man of Rum," also listed in the Turko-Arabic dictionary. He maintains that no such names exist except in the contemporary Rashid-'d Din. Pelliot, [472] although he eventually accepted the usual etymology, confessed to some doubts. He pointed out that agac occurs only in the Altai and some western dialects; the Turfan texts have igac, Kashgari has yigac and yigac, and Turki yagac: Ainsi, au cas où  serait bien Agac-eri, nous devons admettre que, dès le milieu du Ve siecle, les principales caractéristiques qui séparent les divers dialects turcs s'étaient déjà partiellement établies.

Unless one is convinced that in the fifth century all Turks, or even all "Altaians," as some scholars believe, spoke the same language, Pelliot's doubts carry little weight. Hamilton's suspicion that Agaj-eri of the Mongol period was a book word is not justified either. Agac-eri, named together with the five Uigur, occur in the Čagatai version of the Oguz-name; [473] there were Agac-eri in Anatolia, [474] and there still are Agac-eri in Khuzistan. [475]

These names, undeniably, have some resemblance to Acatziri. But whether there is more to it, whether the Acatziri actually lived in woods as their name supposedly indicates, is a question which neither dictionaries nor analogies but only the texts can answer. The interpretation of Drevlyane and Tervingi is anything but certain. Tretyakov thinks that Drevlyane is a distortion of an unknown name, an attempt to give it a meaning. [476] According to Hermann, ter- in Tervingi does not mean "tree" but "resin, resinous wood" and, possibly, a kind of pine. [477] The Greutungi, those Goths who allegedly were named after the "sandy" steppes in the Ukraine, bore this name when they were still living in Scandinavia. [478] It is strange how scholars on the hunt for etymologies of Wörter are apt to forget the Sachen; Ammianus Marcellinus called the "sandy" land of the Greutungi Ermanaric's "fertile country," uberes pages. [479]


And now to the texts:

Jordanes speaks of the sites of the Acatziri and their way of life in the much-discussed chapter V of the Getica. To take up the complex problem of the chapter's composition is not germane to my purpose. [480] It is evident that Jordanes did not simply copy Cassiodorus' Gothic History. He is indebted to Cassiodorus for a good part of the description of Scythia, [481] but he adapted his source to his own work. He wrote, as Cassiodorus could not have written: In Scythia medium est locus; indomiti nationes, and so forth. A number of passages are undoubtedly his own. He speaks of the Bulgars supra mare Ponticum, quos notissimos peccatorum nostrorum mala fecerunt. Peccata is a specifically East Roman word, meaning "neglect, failure" (on the part of the emperors, generals, and so forth). [482] Not Italy but the Balkan provinces were raided and devastated by the Bulgars who crossed the Danube almost every year. It is unlikely that Cassiodorus in Ravenna was even aware of the existence of Noviodunum in Scythia minor, not to speak of the Lake Mursianus, [483] the lagoon of Razelm which the Moesian Jordanes must have known very well.

Although it is impossible to distinguish in each case between Jordanes' text and the shorter or longer borrowings from Cassiodorus, the passage in which we are interested can be assigned to its authors with a fair degree of probability:

Introrsus illis [sc. fluminibus] Dacia est, ad coronae speciem arduis Alpibus emuniia iuxta quorum sinistrum latus, qui in aquilone vergit, ab ortu Vistulae fluminis per immensa spatia Venetharum natio populosa consedit (Getica 34). (Within these [rivers] is Dacia, fortified with steep Alps in the form of a crown, next whose left side, which inclines northward, from the source of the Vistula through immense distances, dwells the populous nation of the Venethae.)
The passage has a strong Cassiodorian ring. [484] The following may also go back to the Gothic History:
Quorum nomina licet nunc per varias familias et loca mutentur, principaliter tarnen Sclaveni et Antes nominantur. (Whose names, though


perhaps now changed through different families and places, are chiefly called Sclaveni and Antes.)
But now the tone changes:
Sclaveni a civitate Novietunense et laco qui appellatur Mursiano usque ad Danastrum et in boream Viscla tenus commorantur. (The Sclaveni dwell from the city of Novietunum and the lake which is called Mursian as far as the Dniester and northward as far as the Viscla.)
Both the words in Roman type and the content give the passage to Jordanes. Note in particular the switch from Vistula to Viscla.

Proceeding eastward, the author describes the sites of the other group of the Venethae:

Antes vero, qui sunt eorum fortissimi, qua Ponticum mare curvatur, a Danastro extenduntur usque ad Danaprum, quae flumina multis mansionibus ad invicem absunt. (The Antes, indeed, who are the strongest of them, extend from the Dniester to the Dnieper, where the Pontic sea is curved. These rivers are many days' journey apart from each other.)
The words in Roman type point definitely to Jordanes. Suddenly we listen again to Cassiodorus, evidently carefully copied:
Ad litus autem Oceani, ubi tribus faucibus fluenta Vistulae fluminis ebibuntur, Vidivarii resident, ex diversis nationibus adgregati; post quos ripam Oceani item Aesti tenent, pacatum hominum genus omnino. (But on the shore of the ocean, where the streams of the River Vistula are discharged by three mouths, dwell the Vidivarii, compounded of several nations, after whom again the Aesti hold the shore of the ocean, a race of men wholly pacified.)
The allusion to Tacitus, Germania 45, 3, [485] and the change from Viscla back to Vistula points to Cassiodorus. And now the crucial passage:
quibus in austrum adsidet gens Acatzirorum fortissimo, frugem ignara, quae pecoribus et venationibus victitat. (To the south of them [quibus] resides the most mighty race of the Acatziri, ignorant of agriculture, which lives upon its herds and upon hunting.)
Mommsen listed victitare as typical for Jordanes. [486] To what does the relative pronoun refer — to the Antes or the Aesti? Getica 23-24 is in this respect rather instructive:


The Suetidi are of this stock and excel the rest in stature. However, the Dani, who trace their origin to the same stock, drove from their homes the Heruli who claim preeminence among all nations of Scandza for their tallness.

Sunt quamquam et horum positura Granii, Augandzi, Eunixi, Taetel, Rugi, Arochi, Ranii. quibus non ante multos annos Roduulf rex fuit, qui contempto proprio regno ad Theodorici Gothorum regis gremio convolavit, et, ut desideravit, inuenit. (However, there are in the place of these people the Granii, Augandzi, Eunixi, Taetel, Rugi, Aprochi, Ranii. Over these, not many years ago, Rudolf was king, who, spurning his own kingdom, fled to the bosom of Theodoric king of the Goths and found the refuge he desired.)

As the text stands, quibus refers to the seven peoples named just before. Yet Rodvulf was not their king but king of the Heruli. [487] After the short digression Jordanes returns to the nation of which he had spoken before. He was, thus, quite capable of referring by quibus not to the Aesti in the quotation from Cassiodorus but to his own Antes. That quibus must, indeed, be understood in this way is shown by the following part of the catalogus gentium:
Ultra quos [sc. Acatziros] distendunl supra mare Ponticum Bulgarum sedes, quos notissimos peccatorum nostrorum mala fecerunl. hinc iam Hunni quasi fortissimorum gentium cespes bifariam populorum rabiem pullularunt. nam alii Altziagiri, alii Sauiri nuncupantur, qui tarnen sedes habent divisas: iuxta Chersonam Altziagiri, quo Asiae bona mercator importat, qui aestate campos pervagant effusas sedes, prout armentorum invitaverint pabula. hieme supra mare Ponticum se referentes. Hunuguri autem hinc noti sunt, quia ab ipsis pellium murinarum venit commercium: quos tantorum virorum formidauit audacia.
(Beyond them [the Acatziri] extend above the Pontic sea the territories of the Bulgars, whom the punishments of our sins have made notorious. After these the Huns, like a cluster of mighty races, have spawned twofold frenzied peoples. One people are called the Altziagiri, the other the Saviri. These hold separate territories; near the Chersonese the Altziagiri, where the merchant imports the goods of Asia. These in summer wander through the plains, scattered territories, as far as the pasture of the flocks invites them. In winter they withdraw again to the coast of the Pontic sea. After these are the Hunugiri, well known because from them comes the trade in ermine. Before them the courage of many brave men has quailed.)


Even if one or the other flosculus should go back to Cassiodorus, [488] the passage as a whole must be attributed to Jordanes. The Bulgars have their sites ultra, that is, east of the Acatziri, and supra, that is, north [489] of the Black Sea; from there the "Huns sprouted out into two savage hordes." As Schirren recognized more than a century ago, Jordanes' Bulgars and Huns in this chapter of the Getica are but two names of the same people [490] Schirren thought that Jordanes simply followed Cassiodorus, who in Varia VIII, 10, 4, likewise identified the Bulgars with the Huns. But in the sixth century this equation was quite common. Ennodius, for example, called a horse captured from the Bulgars equum Huniscum. [491] To Jordanes' Bulgars, Antes, and Sclavini (Romana 388) correspond Procopius'  (VII, 14, 2) and   (Anecdota 18, ed. Comparetti, 122).

Saviri and Hunuguri, too, denote in Jordanes one, and only one, people. After describing the sites and economy of the Altziagiri, he turns to the other of the bifaria rabies. One expects that he would deal with the Saviri. Instead the speaks of the fur trade of the Hunuguri. [492] Jordanes' identification of the two peoples is quite understandable. Although Priscus, Agathias, Menander, and Theophylactus Simocatta clearly differentiate between them, their accounts show that from the 460's to the end of the seventh century the Onogurs (Hunuguri) were the closest neighbors of the Sabirs. They lived north of the Caucasus, on the eastern shores of the Black Sea, in the Kuban area. [493] In the list of nations in the appendix to the chronicle attributed to Zacharias of Mitylene, written in 555, [494] the Onogur are named first, the Sabir third.


The Bulgars — nomads, mounted archers, entirely dependent on their horses — lived, of course, in the steppe. Although no one ever doubted that, no one seems to have drawn from it the necessary conclusions as to the sites of the Acatziri. The discussion about them centered almost exclusively on their supposed proximity to the Aesti. [495] Because the Altziagiri, the western branch of the Bulgars, "were near Cherson" in the western Crimea and "in the winter betook themselves north [supra] the Black Sea," they must have roamed over the plain east of the Dnieper. To a region not far from the Roman frontier point also the frequent raids of the Bulgars across the Danube. This, in turn, permits an approximate localization of the Acatziri. West of the Bulgars, south of the Antes, leads to the lower course of the Bug and Dniester, perhaps as far west as the Pruth.

The southern border of the forest steppe in the Ukraine runs from the northern edge of the Beltsk steppe in Bessarabia to Ananyev, the upper course of the Ingul, Kremenchug on the Dnieper, Poltava, Valuiki, Borisoglebsk, to the Volga north of Saratov. [496] "It was believed," wrote the eminent Soviet geographer Berg, "that at one time the steppe was covered with forests which were destroyed by the nomads. This view is mistaken." [497] Of Herodotus' Hylaea, Minns rightly said that it hardly required many trees to attract attention in the bare steppe land. [498]

The sites of the Acatziri were south of the just indicated line, in the level, woodless steppe. Priscus is in agreement with Jordanes. He too places the Acatziri "in Scythia on the Pontic Sea." [499] The Acatziri were a people of the steppe, not "forest men," not agac-eri.

Being aware of the risks involved in the analysis of a text as patched up as the Getica, I do not delude myself about the fragility of some of my suggestions. But I trust that in the main point, the localization of the Acatziri in the steppe, I am not mistaken.

Now we can return to the name itself. I leave aside the question whether agac means tree or forest; in the fifth century it may have meant both. It is conceivable, though unlikely, that at the time we hear of the Acatziri they had moved from the forests in the north to the steppe, and their neighbors called them "forest men" because they came from there. Their name alone would not make them Turks; the Nez Percés in Idaho did


not speak French nor the Black Feet in Montana English. But all this is beside the point. If Acatziri had been Agac-eri, Priscus would have written . He never rendered a foreign g by kappa. He wrote k not g; his  is Recimer, Ricimer, Ricimerius, not Regimer;  is Edecon, Edica,  Anegisclus. He wrote  for Viminacium and  for Serdica but  for Margus.

After the elimination of the equation Acat(z)iri = Agac-eri, geographically untenable and phonetically unsound, [500] there remain two more attempts to explain the ethnic name. The one suggested by L. N. Gumilev need not detain us; he takes Acatziri for Turkish aka, "older," and carig, "army," [501] which is nonsense. The other explanation has been offered by a number of scholars, most recently by Henning and Hamilton. They assume that the Acatziri were the "white Khazars," aqxazar. Before turning to Henning, we have briefly to deal with Hamilton's excursion into Chinese. He thinks he found the name Xazir in a list of the T'ieh-le tribes in the north, preserved in Sui shu, chapter 84: "North of the kingdom of K'ang on the river A-te are ho tieh ho chieh po hu pi ch'ien chu hai ho pi hsi ho yang su pa yeh woi k'o ta and others, with more than thirty thousand soldiers." [502] To Hamilton's ear A-te, ancient a-tək, sounds very much like Atil, the Turkish name of the Volga. Ho tieh, ancient xa-d'iet, is, he believes, a transcription of Adil, again the Volga, and ho chieh, ancient at-dziet, seems to transcribe Xazir. [503] In this way Hamilton arrives at "on the Atil are the Adil Xazir." Xazir and Xazar are, in his opinion, the same: L'alterance a/i dans le suffix aoriste- était en turc ancien des plus banales. He refers to Armenian Xazirk. The Chinese name and the Armenian support in his view the equation Akatzir = *Aq-Qazir or *Aq-Xazir.

Hamilton's equations are unconvincing. Ho tieh is clearly the name of a tribe, not of a river. Atil cannot in one short passage be transcribed in two different ways. And the Volga does not flow north of K'ang = Samarkand.

In identifying the Acatziri with the Khazars, Henning follows another line of reasoning. Like Hamilton, he refers to Armenian Xazir = Khazar, which, however, he does not take for a variant of the name but of its original form. He stresses that no nation was as close to the Khazars as the


Armenians. This is certainly true, but Pelliot nevertheless called Xazir peu concluant [504] for the reconstruction of an original with i instead of a as in all other scriptions of the name. Henning thinks Xazir is supported by , a name he found in Moravcsik.  undoubtedly stands for Khazar. But the writings in which this name occurs (the Notitia Episcopalian and the glosses of an unknown scholiast) abound with corruptions. In the discussion of the names Acatziri and Khazar,  can be disregarded.

The link between Acatziri and Khazar is, in Henning's opinion, KSR in the mentioned list of nations north of the Caucasus. [505] KSR can be Xasar and Xasir; in Khazarian *Xacir may have become Xazir and later Xazar. Whether such a development in the practically unknown language of the Kasars was possible or not has little interest to us, for  — Acatziri was definitely not *Aq-Xacir.

The ethnic name occurs in Priscus six times: (1) EL 1308 (Cantabrigiensis, Bruxellensis, Escorialensis) and  (Monacensis, Palatinus); (2) EL 13026 (all codices); (3) EL 1363 (all codices); (4) EL 13923 (the same); (5) EL 58612 (the same); (6) EL 58810 (the same).

Of all the scholars who struggled with the problem of the Acatziri only Markwart realized the importance of the textual tradition. He discussed it in a work where one should not expect it, the posthumously published Entstehung der armenischen Bistümer. [506] Markwart showed that Priscus in all probability wrote , which later scribes "emended" to . They did it the first two times only. Later they let the name stand as it was, perhaps expecting that the reader would now correct it himself. Suidas, quoting Priscus, has . [507]

It is unlikely that tz in Priscus was meant to render -ts- or -c-, as Markwart thought. Since he wrote  (EL 1214), what could have prevented him from writing  for Akacir or Akatsir? There are three possibilities, as far as I can see, to account for the difference between  and Jordanes' Acatziri. The first could be the change from ti to tsi in vulgar Latin. Second, the same change may have occurred in the language of the Acatiri in the eighty years which separate Priscus from Jordanes. And, third, one could think that Priscus "reconstructed" the name; he might have heard Akatsir and still have written .

The last possibility seems somewhat far-fetched. Against the second one, in itself not exactly likely, speaks the spelling  in Agathias


versus Ultzinzures in Jordanes; Agathias wrote his history after the Getica. Only the first possibility remains: Jordanes changed  to Acatziri as he changed Scandia to Scandza or Burgundiones to Burgundzones. In my opinion, the "emended" forms in Priscus go back to the spelling in Jordanes. All manuscripts of the Excerpta de legationibus are copied from one codex. The apographs were made by Andreas Damarius and other scholars in the later part of the sixteenth century, at which time already three printed editions of the Getica existed. [508] It seems more probable to me that the sixteenth-century scholars, following Jordanes, "corrected" the Priscus text rather than Greek scribes of the sixth century as Markwart thought.

Henning's historical arguments for the identification of the Acatiri with the KSR are based on Priscus, fragments 30 and 37. In about 463, the Saraguri subdued the Acatiri after many battles; they themselves had been driven out of their country by the Sabirs who had been set in motion by the Avars, and the Avars in their turn by peoples which fled from man-eating griffins coming from the ocean (fr. 30). In 466 the Saraguri, after their attack on the Acatiri and other peoples, , marched against the Persians, crossed the Caucasus, ravaged Iberia, and overran Armenia (fr. 37). Combined with Priscus' statement about the sites of the Acatiri in Scythia on the Pontic Sea, the two fragments are supposed to prove that the people lived in the steppes between Kuban, Don, and Volga, thus in about the same region as the KSR = Xasir or Xasar.

I am unable to accept Henning's conclusions. It must be emphasized that the two fragments were shortened by the scribes who put them together for the collection of Constantine Porphyrogenitus. For fragment 30 this has been proved by Moravcsik. [509] A comparison of fragment 30 and the beginning of fragment 37 shows clearly that the latter was also abbreviated. It refers to the battles in fragment 30, for it would be absurd to assume that in 463 the Saraguri subdued the Acatiri and three years later, before marching against the Persians, attacked them again. Fragment 30, as it stands, says nothing about the war between the Saraguri and other peoples beside the Acatiri, briefly referred to in fragment 37. In other words, the original Priscus text contained considerably more about the many fights of the Saraguri against the Acatiri and other peoples. It may also have been more specific about the region where those fights took place, although Priscus was apparently not well informed about


the events in the vastness of European Sarmatia. Or should we really believe that the Herodotean griffins came from the river which encompassed the earth? Be that as it may, there is no reason to believe that in 463 the Acatiri had moved from the sites they had held in the last years of Attila's reign "in Scythia on the Pontic Sea," and after a few years under Attila's yoke had regained their freedom. For two or three years they were the subjects of the Saraguri, but by the middle of the sixth century Jordanes knew them as fortissima gens, subject to no one. Moravcsik thought it self-evident that at that time they were still where they were in the 460's, which, as I tried to demonstrate, was west of the Azov Sea.

Only by mistranslating the passage in fragment 37, which I quoted from Priscus' Greek text, could it be argued that the Acatiri followed the Saraguri on the march against the Persians. The editors of the Bonn edition and C. Müller in Fragmenta Historicorum Graecorum IV, 107, translated the passage cum Acatiris aliisque gentibus coniuncti. Doblhofer has im Bunde mit den Akatziren, [510] and Altheim zusammen mit den Akatziren, [511] although Moravcsik pointed out years [512] ago that these translations were wrong. Gordon correctly says "having attacked the Akatiri and other races." [513] I want to stress that Henning's views are not based on this mistranslation.

The East Romans tried to conclude an alliance with the Acatiri about 445 and actually concluded one with the Saraguri after the newcomers from the East had conquered the Acatiri. It is most unlikely that the Romans had contact with barbarians between Kuban, Don, and Volga, far beyond the ken of the government in Constantinople. In the wild melee of the early 460's, the Saraguri obviously pushed for a short time beyond the Don and Dnieper. All this speaks against the identification of the Acatiri with the KSR north of the Caucasus.

For Akatir I have no etymology to offer. , the name of one of their rulers in the late 440's, [514] is possibly Turkish. Justi listed it as Iranian, probably because he relied upon a third-century inscription from the Crimea [515] which Latyshev restituted in analogy with Kuridachus. [516] Ce nom, says Sinor, a une consonance turque, mais je n'ai pas réussi a l'identifier. [517] Kuridachus might be qurtaq, qurt, "wolf," and the diminutive


suffix q; compare Gothic Wulfila. Németh maintained that qurt, "worm," acquired the meaning "wolf" only in recent times. [518] How recent is "recent"? In Qazwini's Nuzhat al-qulub, written in 1339, qurt already means wolf. [519] Qurt was apparently a general term for living beings, used for the wolf when its actual name was taboo. [520]

As long as the discussion of Akatir had to be, as briefly can the other tribal names of the Huns be dealt with.


Ultzinzures, , is composed of Ultin and čur. Whatever Ultin may mean, it is probably as Turkish as il in ilčur.

Four names end certainly, one almost certainly, and another one possibly in -gur.

1. . Although the exact form hidden behind the various readings [521] cannot be determined, the name ends even in the most aberrant variants in -gur.

2. Hunuguri, [522] , is Turkish On-Ogur, "ten Ogur." Hamilton tried to prove that Onogur is a poor transcription of On-Uyghur. The Byzantines allegedly were unable to render the diphthong uy in their script. [523] But they wrote  and  [524] there was nothing that could have prevented them from writing .

3. . Except in two inferior codices, the readings, as manifold as those of Kutrigur, [525] all end in -gur.

4. The Bittugures, one of the tribes who acknowledged Dengizich as their leader, joined the Ostrogoths on their trek to Italy; Ragnaris was one of them. Agathias'  is . [526]

5. . Priscus (EL 1215) has dat, Jordanes (Getica 9012) Tuncarsosacc, *Tuncursos. Markwart emended  to read . [527]


In Koibal ton means "people"; see Pritsak (1952, 56) with reference to Castren. The Tongur, Dongur, Tongul were "bones" of the Altai Turks (Grum-Grzhimailo 1930, 19). If I remember correctly, Tongur was mentioned to me as a "bone" on the Dzhakul River in the Khoshun Kemchik in Tannu-Tuva.

6. . After signing the peace treaty of Margus, Bleda and Attila went to war against the Sorosgi, a people in Scythia. [528] Kulakovskii took the name, a hapax legomenon, for misspelled Saragur, [529] which was not a good guess. When one thinks of  in Nicephorus Callistus, [530] which goes back to Theophylactus' , one is tempted to take the sigma at the beginning of the name for a dittography:  is possibly a distortion of . Priscus'  are misspelled . [531] The possibility that  stands for  cannot entirely be ruled out.



The name occurs only in Getica 1283: angiscirosacc, angiscires. The Angisciri were one of the four tribes which remained loyal to Dengizich. Vasmer took angi- to mean the same as OE eng, "grassland"; [532] the Angisciri would, thus, be the "grassland Sciri." As, however, the other three tribes, Ultzinzures, Bittugures, and Bardores, have Turkish names, it seems more likely that Angisciri is also Turkish. The scribes may have assimilated an unfamiliar name to that of the Sciri, who in the Romana are named twice and in the Getica five times. Angis- is reminiscent of äŋiz ,"field." [533] Angisciri might be Angisgiri.



The name [534] is evidently compounded of var, as in , and -dor. This dor-dur is not only the second element in Ultzindur; it occurs also in Bayundur, the name of an Oghuz tribe, and the tribal names listed by Pritsak (1952, 77). [535] Cegitur is a Kirghiz clan. [536]


This people is named together with Unugur and Sabir. [537] Markwart (1924, 324) took Barselt to be an Ossetic plural: Barsel-t. He identified the *Barsel with Menander's , Basil-k' in Ps. Moses Chorenac'i; B'grsyk [538] in the list of 555; , the old home of the Khazars; and Barula, one of the three tribes of the Volga Bulgars. All this is highly hypothetical. For more speculations on these names, see Minorsky 1958, 94, and Oriens II, 1958, 125-126; Artamonov, index 498, s.v. Barsily. K. F. Smirnov (KS 45, 1952b, 95) thinks the graves at Agachkalinsk near Buinaksk on the northwestern shore of the Caspian Sea can be assigned to the Barsil. Barsel occurs in the legend of a Volga Bulgarian coin of the tenth century (see S. A. Oanina, MIA 111, 1962, 187, n. 41).


John of Antioch (EL 1396) is the only author to call the Cadiseni a Hunnic people. Seventy years ago Nöldeke proved that the Cadiseni, repeatedly named by Byzantine, Syriac, and Armenian authors, had nothing to do with the Huns. [539]


Mentioned only by Menander (EL 4439) as a Hunnic tribe. It is difficult not to think of Ptolemy's  in European Sarmatia [540] and Pliny's Salae in Colchis, [541] but see .


On the various forms, see Moravcsik, BT 2, 262. For years the Sabirs were the favorite objects of name hunters. Pelliot was inclined to accept Németh's Turkish etymology; [542] the Sabirs were "the wanderers." [543] Henning (1952, 502, n. 5) thought he found the Sabirs in the Sogdian Naf-namak, which would place Sabir in the neighborhood of Turfan long after the fifth century. Moór offers a particularly unconvincing Iranian etymology. [544]


After the collapse of Attila's kingdom, "Sciri and Sadagarii and certain of the Alans . . . received Scythia minor and Moesia inferior." [545] The Sadagarii cannot be separated from those Sadages who, at the same time, still loyal to Dengizich, "held the interior of Pannonia. [546] This has been pointed out by Zeuss as early as 1837 [547] but did not prevent serious scholars, as well as a host of dilettantes, from offering the fanciest Iranian and Turkish etymologies. They either divided Sadagarii into Sada-garii [548] or Sadagarii; [549] Abaev preferred the reading Sadagarii because it gave him the chance to suggest an Ossetic etymology. [550] It is, or should be, obvious that -es is the Greek and -arii the Latin ending. [551] The name is obscure. It is not even known whether the tribe formed a part of the Hunnic confederacy in the narrow sense or was only loosely connected with it.

[Previous] [Next]
[Back to Index]

463. Zeitschr. f. d. Österreich. Gymnasien 23, 1872, 142.

464. Houtsma 1894, 23, 49.

465. Quoted by Pelliot 1950, 210.

466. Sinor 1948, 3.

467. "Zane sedosha v lesekh," Povest' vremennykh let 1.

468. Schönfeld 1911, 222.

469. F. W. K. Müller 1915, 3, 34.

470. 1952, 14/3, 506.

471. 1962, 58.

472. 1950, 213.

473. Ibn Fadlan 1939, 147-148.

474. Pelliot 1950, 212, n. 1.

475. Barthold, Encyclopedia of Islam 2, 838; W. B. Henning 1952, 506, n. 8.

476. Vostochnoslavianskie plemena, 249.

477. Abh. Göttingen 3:8, N.F., Fachgruppe 4, 1914, 271-281. Cf. also the controversy between H. Rosenfeld and F. Altheini in Beiträge zur Namenkunde 7, 1956, 81-83, 195-206, 241-246; 8, 1957, 36-42.

478. Cf. W. Krause 1955, 12; Rosenfeld 1957b, 246. Cf. also Ernest Schwarz 1951, 34.

479. XXXI, 3, 1.

480. For Getica 30-35, see L. Hauptmann, Byzantion 4, 1927-1928, 138-139.

481. Cf. Cipolla 1892, 23.

482. Cf. J. Friedrich, "Über einige kontroverse Fragen im Leben des gotischen Geschichtsschreibers Jordanes," SB München 1907, 405-407.

483. On this name, see F. J. Mikkola, Symbolae grammaticae in honorum Ioannis Rozwadowski 2 (Cracow, 1928), 533; G. Nandris, The Slavonic and East European Review 18, 1939, 144; H. Łowniański, Opusculum C. Tymienicki (Poznan, 1959), 211-224. Another name of the lagoon is  (Analecta Bollandiana 31, 1926, 216).

484. Cipolla 1892, 23.

485. As in Variae V, 2, first pointed out by Schirren 1846, 49-50.

486. In his edition of Jordanes, index 199.

487. Mommsen 1882, 154.

488. For pullullare Mommsen (1882, 63, n. 2) referred to Variae III, 6. But, as Cipolla (1892, 23) rightly remarked, la fraseologia non può dare sufficente guarentigia di sicura altribuzione, perchè tra scrittori più o meno contemporanei è cosa agevole trovare riscontri di sifatte specie.

489. For ultra and supra in geographical descriptions, see H. Sturenberg 1932, 199ff.

490. Schirren 1846, 50.

491. MGH AA VII, 169.

492. Pellium murinarum commercium. Mns means any of the numerous species of small rodents, from ermine and marten to squirrel and mole; cf. Stein, PW 14, 2398. The "mice" of the Hunuguri were apparently the "wild mice" of whose skins, according to Hesychius, the Parthians used to make their coats; in the Parthian language they were called , i.e., samōr, "sable." Cf. E. Schwentner, "Ai. samura-ṣ, samuru-ṣ und die pontischen Mäuse," Zeitschrift für vergleichende Sprachforschung 71, 1953, 90-94. Turkish samur, "sable," is an Iranian loanword.

493. Moravcsik 1930.

494. To the translations listed in Moravcsik BT 2, 219, add F. W. K. Müller, Ostasiat. Zeitschr. 8, 1919-1920, 312, and Pigulevskaia, VDI 1 (6), 1939, 107-117. The story about the Hun Honagur in Movsēs Dasxurançi (1961, 63-65) is as confused as its chronology; its historical value is nil.

495. W. B. Henning (1952, 503) at least considers the possibility of locating the Acatziri with the help of the data on the Bulgars, but in his opinion it is "by no means clear where precisely one is to imagine their seats."

496. Berg 1950, 68.

497. Ibid., 108.

498. Minns 1913, 15.

499. .

500. The spelling Agazari (Chazaros . . . Iordanis Agaziros vocat) in the anonymous geographer of Ravenna IV, 1 (Cuntz 1940, 2, 44) is of no consequence; cf. J. Schnetz, SB München 6, 1942, 34.

501. In M. I. Artamonov, SA 9, 1949, 56.

502. Hamilton 1962, 26-27. The same list in Pei Shih, ch. 99, has some different scriptions. In most cases it is impossible to decide whether the names are binoms or trinoms.

503. Hamilton 1962, 53, n. 14; 57, n. 47.

504. 1950, 207, n. 3.

505. For a new view of the list, see K. Czeglédy, AOH 13, 1961, 240-251.

506. Orientalia Christiana 27 (Rome) 1932, 208-209.

507. A. Adler, ed., I, 413, 7713,14.

508. Mommsen in the preface to his edition of Jordanes, lxx.

509. 1930, 55-65.

510. 1955, 74,

511. Altheim 1962, 4, 277.

512. 1930, 60, n. 1.

513. Gordon 1960, 12.

514. EL 13015,19,23.

515. Justi 1895, 167. Cf. also Zgusta 1955, 111, 133.

516. IOSPE 1, 218.

517. Sinor 1948, 2, n. 1.

518. KCsA 2, 44.

519. BSOAS 6, 1931, 565.

520. A. M. Shcherbak in Istoricheskoe razvitie leksiki turetskikh iazykov, 132-133.

521. Moravcsik, BT 2, 171-172.

522. Getica 63. Hunuguri is not Hun-uguri, in which case Jordanes would have written Hunnuguri, but Un-Uguri.

523. 1962, 38.

524. Menander, EL 45229.

525. Moravcsik, BT 2, 238.

526. Zeuss 1837, 709.

527. Markwart 1911, 11, note. Later (1932, 208), he changed his mind and took the name to be *Tunčur. But sigma could not render č.

528. Priscus, EL 12222.

529. 1913, 1, 265.

530. PG 147, c. 385c.

531. Moravcsik, BT 2, 227, 238.

532. Arkiv. f. nord. filol. 58, 1944, 87-88.

533. Kāshgharī, 22; Malov 1951, 206: 8, 15.

534. Getica 12823; v. l. bardares.

535. Apparently none of them belongs to the group of names with the imperative suffix -dur, which L. Rásonyi discussed in AOH 15, 1962, 233-243.

536. Vinnikov 1956, fig. 16.

537. Moravcsik, BT 2, 87.

538. Restored as B'RSYLQ by Markwart; W. B. Henning (1952, 504, n. 4) suggests B'RSYGQ, Armenian Barsilk'.

539. "Zwei Völker Vorderasiens," ZDMG 33, 1897, 157-163.

540. Geog.

541. HN VI, 14.

542. 1950, 232.

543. MNy 25, 1929, 81-88.

544. UJb 31, 1959, 205-206.

545. Getica 265.

546. Ibid., 272-273.

547. Zeuss 1837.

548. Vasmer 1923, 49; Arnim 1936, 348-351; Harmatta 1947, 7-28.

549. Markwart 1903, 44, and Izv. russk. arkheol. inst. v Konst. 15, 1911, 13, note.

550. 1949, 179-180.

551. Zgusta 1955, 263, § 533.