Toponymy and ethnic Realities at the Lower Danube by Brezeanu Steli

Toponymy and ethnic Realities at the Lower Danube

in the 10th Century.

"The deserted Cities" in the Constantine Porphyrogenitus'
De administrando imperio


Stelian Brezeanu,

University of Bucharest



De administrando imperio, the most important work of
Constantine Porphyrogenitus, has been clearly subdued to the
most ample investigations among the historian's works.
Nevertheless, it contains some passages still obscure that has
not been satisfactorily analyzed by the modern scholars. Among
these passages, a special part is taken by the one referring
to "the deserted cities" from the Lower Danube.



"Isteon, oti enqen tou DanastrewV potamou proV to
apoblepon merosthn

Boulgarian eis ta peramata tou autou potamou eisin erhmokastra
kastron prwton to onomasten para twn Patzinakitwn Aspron dia
to touV liqouV autou fainestai

kata leulouV, kastron deuteron to Touggatai, kastron triton to
Kraknakatai,

kastron tetarton to Salmakatai, kastron pempton to Sakakatai,
kastron ekton [to] Giaioukatai. En autois de tois twn
palaiokastrwn ktismasin euriskoutai kai

ekklhsiwn gnwrismata tina kai stauroi laxeutoi eiV liqouV
pwrinouV, oqen kai tineV paradosin ecousin, wV Rwmaioi pote
taV katoikiaV eicon ekeise"[1].



In translation:



"On this side of the Dniester river, towards the
part that faces Bulgaria, at the crossings of this same river,
are deserted cities: the first city is that called by the
Pechenegs Aspron, because its stores look very white; the
second city is Toungatai; the third city is Kraknakatai; the
fourth city is Salmakatai; the fifth city is Sakakatai; the
sixth city is Giaioukatai. Among these buildings of the
ancient cities are found some distinctive traces of churches,
and crosses hewn out of porous stone, whence some preserve a
tradition that once on a time Romans had settlements there".



The savant-emperor's text raises some problems that
are difficult to be interpreted and that have discouraged the
modern scholars to approach it.

First, while the first among the six "deserted
cities" is not difficult to be identified - since Aspron means
"white" in the Pecheneg language, as it resulted also from the
text, it could only be Rom. Cetatea Alba or Sl. Bielograd, on
the right bank of the Dniester, on the river mouth to the
Black Sea -, the other five seem to be enigmatic, difficult if
not impossible to be deciphered. Consequently, it should not
be surprising that the Romanian historians has not paid any
attention to them[2], while the historians outside of Romania
that have dealt with the Byzantine historian's text simply
confined themselves to mention them as they are[3].

The Constantine VII's specifications around the
ancient cities' settlement are also difficult to be
interpreted. What does "on this side of the Dniester river, in
the side that regards to Bulgaria, on this river's passings
(ta peramata)" mean? Should we understand that all the six
cities are to be found out in the immediate proximity of the
river's right bank? There is nothing to forbid us to suppose
that they were by the Bulgarian region, which has the Danubian
line as frontier with Patzinakia, as the historian informs us
on other occasion[4]. In addition, we should not surpass the
possibility of some inaccuracies in their placement, because
of either the informer or the way in which the information was
interpreted by the cabinet savant Constantine Porphyrogenitus,
who was never passing in the described region. This larger
interpretation of the text referring to the cities' placement
is also imposed by the fact that the Roman domination in the
region, whether it did exist, did not penetrate in the depth
of the Northern Pontic territory, the empire confining to
control the sea's shore.

Thus, here is the most difficult point raised by the
text: the existence of a Christian Roman domination on the
Dniester's right bank or of a kind of control that is to
explain the Christian remnants in the region during the first
decades of the 10th century. It is especially because the
author expresses some doubts in connection with the existence
of such a control, when he considers that there are "some
[persons]" (tineV) to promote the tradition of the Roman
presence in the region.



a. The Southern Moldavia and the Roman Imperial Policy

We are to begin with the matter of the Roman
presence in the Northern Pontic area and especially in the
region of the Dniester's right bank. The information in
connection with this presence are more numerous and more
conclusive than it could be noted at a first sight. For the
Christian period, there are two texts to clarify this aspect:
The Ecclesiastical History of Evagrius and Chronographia of
Theofanes the Confessor.

Evagrius Scolasticus, the author of the first text,
was born to 536 and lived by the first years of the 7th
century. He lived the last part of his life in Constantinople,
where he wrote his work. Well informed and having Tucydides as
pattern, it covers the period between 431 and 594. The author
describes the Northen Pontic realities on the occasion of the
Avars' coming in Europe in 558, event that provoked an
impressive echo in the 6th century Byzantium. "After they had
left the shore of the Pont called Euxinus", Evagrius notes,
"where there were all the different kinds of barbaric nations,
while the Romans had established cities (poleiV), military
camps and some stations for the veterans and for the colons
(apoikiwn) sent by the emperors [emphasis mine], they [the
Avars] opened a pass and fought against all the Barbarians
encountered in their way, since that they achieved the Istrus'
banks and sent envoys to Justinian"[5]. The text of the
ecclesiastic author indicates a very complex reality in the
space between Crimea, named as the Cymmerian Bosphorus by the
Byzantine authors, and the Lower Danube. Beside the "Barbarian
nations", very different by their origins, languages and even
political interests, there were also Roman establishments,
having an essentially military functions, in order to preserve
a political stability in the region, in the sense of assuring
the Christian New Rome's security on the Bosphorus. The author
does not specify since when the "cities", the "military camps"
or Roman "colonies" has been dated. Probably, they were not
the exclusive work of the 6th century emperors. On the other
side, although the text refers to the "Barbarian" opposition
against the Avars, it permits us to suppose that the local
"Romans" also opposed to the newcomers.

The second text has the same importance for our
investigation. Its author, Theophanes the Confessor, writes to
815 and inspires himself from an important number of Byzantine
sources from the 7th-8th centuries that has not been preserved
by now[6]. The 9th century chronicler depicts another event,
the second as importance for the Northern Pontic area, after
the coming of the Avars: there is the coming of the
Protobulgarians led by Asparuch to 679. Theophanes describes
the succession of the events that preceded the establishment
of the newcomers at the Southern Danube: the retirement of the
Asparuch's clan towards the Danube because of the Khazars, its
establishment in the Oglu region (somewhere between the
Northern Pontic rivers and the Danube's mouths), the
unfortunate expedition of the Emperor Constantine IV to the
Istrus and its failure, followed by the river's passing by the
Protobulgarian clan. On this occasion, the chronicler delivers
one of the most important information for our investigation,
specifying that the emperor found out that the Asparuch's
Barbarian nation "settled in Oglu, beyond the Danube and,
invading the territories neighbor to the Danube, it devastates
the country now dominated by them, but on the Christian
domination on those times [emphasis mine]"[7]. Although the
passage seems to have especially Scythia Minor into account,
the expedition's stage demonstrates the Empire's interest also
at the North of the Danube's mouths at least. As a
consequence, after more than a century after the Avars'
invasion in Europe, the territory continued to be under the
attention of the "Christians" (Romans). The strategic
importance either of the territory remained the same for
Constantinople, since even under dramatic conditions for the
empire - the first great Arab siege on the New Rome (674-678)
just came to an end - the emperor gathers the last resources
for an expedition on the Danube against the new danger
appeared there. The Theophanes' text has a double importance
for the present investigation. First, it demonstrates a
continuity of the New Rome's military presence during at least
two centuries in the territories of the Danube's mouths,
because of the their strategic importance, despite the new
situation occurred in the Balkans during the 7th century.
Secondly, the Byzantine defeat in front of the new nomads come
on Danube closes a particular epoch for those territories: for
three centuries, Constantinople abandons politically and
military the region that, when Theophanes writes his
chronicle, was already under the control of the khans in
Pliska. This fact explains why the domination of the
"Christians" in the region remains explained as a "tradition"
by Constantine VII, who, in 950, seems to have some doubts
about it. On the other hand, the abandonment of the Northern
Pontic cities by the Roman emperors was to happen on the
occasion of the coming of Asparuch's Protobulgarians.

While the two texts presents an undeniable
importance in attesting a Byzantine presence in the region and
in establishing a terminus ante-quem for its ending, the
analyzing, even summarily, of the beginnings of this presence
in this territories is not lacking of significance.

The Northern Pontic territories, from Chersones to
the Danube's mouths, were being in close contact with the
Mediterranean world since the first centuries of the 1st
millenium a. Chr. First, it was through the agency of the
Greek colonies that established the link between Chersones and
the colonies in Scythia Minor and that had an essential part
in draining the hinterland's huge wealth towards the Greek
world. In the Roman period, the interest for the region grows,
but the military preoccupations predominates on the economic
ones and they are in order to stop the torrent of the nomadic
populations come from Euro-Asian steppe, which warned to
overflow on the Danubian territories, especially after Trajan
had created his province at the Northern of Danube just as a
spear jabbed in Barbaricum. At the beginning of the 20th
century already, V. Pârvan foresaw the entire importance of
the Roman action in the region, beginning with Trajan, who
raised the numerous castra on the Sereth Valley that makes the
junction between Scythia Minor and his new province. The
action culminated with Trajan's successors, when "the entire
Wallachian field and the Southern Bessarabia with the region
to Cetatea Alba (Tyras) were brought in that moment from the
Dacian way of life to the Roman one"[8]. The post-war
researches definitely confirmed the emminent archeologist's
intuition and the results were synthetized by Radu Vulpe in
some studies dealing with the Roman military and political
presence in these territories in the 1st-4th centuries[9]. For
Eugen Lozovan, the central point of the Rome's interest was
represented by the Southern Moldavia, a real "connection point
between the imperial authority solidly implanted in Scythia
Minor and in the Transcarpathian Dacia and the Roman camps
dispersed in the Northern Pontic steppe"[10].

While the abandonment of Dacia by Rome meant for
some decades a retirement of the imperial positions on the
Danube line, the strategic importance of the territories from
the Danube's mouths becomes vital after the transfer of the
imperial metropolis on the Bosporus shores by Constantine. In
the vision of Constantinople, Scythia Minor and the neighbor
Northern Pontic territories constitute an outpost in front of
the migratory waves that could overflow to the South through
this region. There are numerous news that indicate the
imperial authorities' care at the Lower Danube since the reign
of the New Rome's founder. First, it is about the Constantine
the Great's military campaigns on the North of the river that
are to transform the Wizigoths in empire's foederati, action
renewed four decades later by Valens. The same care for the
area's fate and for the Roman population in the region should
be also observed in the creation of the Bishopric of Gothia,
which titular is mentioned among the participants at the first
ecumenical council at Nicea, in 325[11]. It is certitude that
this population was formed by prisoners transferred there by
the Goths during their raids in the Balkans and in Minor Asia,
by Roman merchants, but also by the descendants of the Romans
brought by Trajan and his successors for military reasons. The
same care to consolidate the empire's positions at the North
of the river explains also the conversion to Christendom of
the Goths in the Wallachian Field and in the Southern
Moldavia, work promoted by Wulfila after 340[12]. At least
the same importance is represented by the archeological
testimonies that demonstrate the 4th century empire's efforts
not only to consolidate the Danubian boundary especially on
its portion in Scythia Minor, but also to install at the North
of the river a defensive system of castra and earthen walls,
the most famous being undoubtedly "Trajan's wall", which the
most important part is dated in this period[13].

In the 6th century, after their diminishing
consequently to the storm unleashed by the Huns in 376, the
news regarding the empire's care for the territories at the
Danube's mouths are more numerous. They owe all their
importance to the information later delivered by Evagrius and
Theophanes the Confessor. The most important data are in
connection to the activity promoted by Justinian on the
Danubian border. Procopius of Cesarea illustrates the
emperor's effort to consolidate the Danubian limes, effort
related by the historian with one carried on by his
forerunners, conferring its entire proportion in the 6th
century. "The former emperors", he writes in De aedificiis,
"covered with fortifications all the river's bank, not only on
the right side of the river, but they also built small cities
(polismata) and cities (frouria) on the opposed side ...
Later, when Atila rushed with a large army, he destroyed these
fortifications, without any difficulty, and laid waste the
greatest part of the Roman territory without any resistence.
But the Emperor Justinian rebuilt the destroyed
fortifications, not as they were previously, but much
stronger; and he repaired many of them and also he renewed
them. In this way, he gave the lost assurance back to the
Roman empire"[14]. In more sober terms, the same information
is to be detected in one of the Justinian's Novel in 535,
referring to the jurisdiction of the new Archbishopric of
Prima Justiniana, created by the emperor. Not only the
metropolitan churches and bishoprics on the South of the
Danube was to be under its titular's authority, but also the
left right of the river's eparchies. It was "because nowadays,
with the God's assistance, our state grew, so that the both
banks of the Danube are inhabited with our cities and both
Viminacium, and also Recidiva and Litterata, which are beyond
the Danube, were again subdued under our domination"[15].
Indeed, the document indicates the Danubian cities on the left
side of the Danubian limes that were under the jurisdiction of
the new created Archbishopric. Still, there is no doubt that
Justinian's action to recover the Northern Danubian cities had
also the right side of the Roman frontier at the Danube's
mouths into consideration. This care is proved by the
emperor's decision in 536, mentioned by John Lydos, to create
a military prefectura of Scythia Minor, having Odessos
(nowadays Varna) as residence. It was to have not only the
Lower Danube in its obedience, but also other three naval
provinces: Cyprus, Caria and the island in the Archipelago.
While the purpose of this decision is already clear, being
connected to the assurance of the Constantinople's and the
straights' security in front of an possible peril coming from
the Northern Pontic steppes, John Lydos puts it under the
circumstances of the recovering by Justinian of the
territories once conquered by Trajan and then lost by the
empire. It was because that "not desiring to be somehow
inferior to Trajan, [the Emperor] decided to preserve for the
Romans the Northern region that once get out of the yoke"[16].
Whether we left aside the imperial propaganda's aims that are
natural in the text of a Justinian's high magistrate as was
John Lydos, we are to remark the Constantinople's care for the
Northern Pontic regions, which strategic importance for its
security was undeniable.

In order to distinguish the New Rome's strategic
conception at the Lower Danube during the 6th-7th centuries,
it is necessary to also make referrals to another episode in
the Menander Protector's work that takes the
Slavic-Avar-Byzantine combats into account. The author
narrates the success of the imperial diplomatic action during
the reign of Tiberius II that counteracted the Avars against
the Slavs in the Southern Moldavia and in the Eastern
Wallachia. An imperial high office worker transferred the
Avars of the Khan Baian from the Northern to the Southern of
the Danube in the region of the Roman Pannonia. Afterwards,
the Avars crosses the imperial territory on the road to
Scythia Minor. They passed again the Danube in order to attack
the Slavs, the emperor's enemies. Surprised, the Slavs were
defeated by the momentary allies of the Byzantine
sovereign[17]. The evolution of these events makes obvious the
concern of Constantinople to control also the cities on the
Northern bank of the river, in order to be able to advance
offensive actions against the migratory nations in the region
(as it occurs in the case of the event presented by Menander
Protector, or the ones promoted by the empire in the last
decade of the 6th century in Banat against the Avars and the
Slavs). However, the New Rome continues to regard the Northern
Danubian territories, once dominated by the empire, as a land
belonging de jure to the empire, only temporary submitted to
the Barbarians. Among other arguments that sustain this
conception of Constantinople, there is also a detail in The
Wars of Procopius. Confronted with the Slav danger, Justinian
makes to the Slavs the proposal to occupy to Turris, a city
once built by Trajan but abandoned by the Romans because of
the Barbarian attacks. The Byzantine historian adds the fact
that the emperor promised also the territory around the city
to the Slavs, "because it was belonging to the Romans since
the very beginning"[18].

On the most occasions, the Roman sources of this
centuries mention about fortifications and cities in which
shadow the people that brought the stone and the iron into
life animate. The modern archeologists neither make many times
exception from this rule, especially when they notice the
cities' abandonment under the pressure of the invasions. The
mentioning of the sources about "the cities" and "the
colonies", as it is the case of the Evagrius' text that is
mentioned above, are still rare. There are to be attached
other two texts, although they do not present the same
testimonial value about the human realities in the region. A
Justinian's Novel in 538/539 makes referrals to the law
sanctions in case of the abuses of the military commanders in
connection with the theft from the fiscality. The punishment
regarded not only the guilty ones, but the entire military
unit, which "will be transferred from the region and ordered
beyond the river of Istrus or the Danube, in order to guard
those boundaries"[19]. The imperial document attests the
presence of some military forces in the Byzantine cities to
the North of the river. Moreover, it also demonstrates that
the guard mission in this region was regarded as one of the
most difficult ones for the soldiers. The other text belongs
to Cosmas Indicopleustes, the author of Christian Topography
and tireless traveler that also visited the Northern Pontic
regions to the middle of the 6th century. Among the
territories where he met and saw "churches and bishops,
martyrs, hermits, monks, in all the places where the Christ's
gospel had been announced", there are also the ones "towards
the North, belonging to the Scythians ... to the Bulgarians"
and to other peoples[20]. The Scythians' regions in the
Cosmas' text could represent the Scythia Minor, which
religious life in the 6th century is attested by a very rich
sources. In exchange, the Bulgarians' ones could not be
identified in other sources than with the Northern Pontic
ones and even with the Danube's mouths in 550. From that
region, the steppes' nomads organized robbery raids in the
empire's Balkan province. In this case, the Christians that
are referred in the text are not the Bulgarians, but the
populations under their hegemony or in community with them.

Such different by their nature, all these sources
impose the conclusion that the territories at the Danube's
mouths was characterized by a Roman military presence during
six centuries, from Trajan to Constantine IV, with an
interruption of a century in the context of the Hun invasion.
This presence was materialized by the cities, the earthen
walls, but also by troops, which goal was to secure the right
flank of the Roman front at the Lower Danube. It also suppose
a human permanence, represented by soldiers, but also by their
families, by manufacturers and merchants, indispensable for
the military activity in the region, as it is clearly
expressed in the Evagrius' relation. At the same time, the
human permanence means the existence of the Christ's faith
among the soldiers and the other Romans and implicit of the
churches and the other Christian symbols, especially after
Constantine the Great. The different Christian vestiges would
be the ones mentioned by Constantine Porphyrogenitus after
some centuries.

Still, the historian's mission connected with the
detection of information referring to the Rome's military and
human presence in the Northern Pontic region from the
Dniester' left side, which explains also the "deserted cities"
existence in the 10th century, is not to finish here. It is
necessary to investigate also the medieval sources, especially
the Romanian ones, respecting these ancient vestiges. The
Romanian medieval data are indeed very numerous and presented
in various sources, from the official acts to historical works
of the 17th-18th Moldavian scholars. There is a crowd of
information about these traces in the official acts and even
in toponymy. The most of them refers to the "troiene", the
earthen walls attributed to Trajan by the Romanian medieval
tradition and that crossed the Wallachian Field from Severin
to the Dniester, identified by the commoners with "Trajan's
wall". Certainly, the ruins of the antique cities are
interesting here. The most important testimonies about them
are specified by Miron Costin and by Dimitrie Cantemir.

Educated at the University of Liov, Latin speaker
and expert in Roman history, the former offers many
information about some "naruite / destroyed" cities, included
a "devastated city" in the Southern Bessarabia, on the
Cogâlnic river, considered by him as being Greek[21]. Still,
the author pays attention to the vestiges between Pruth and
Sereth rivers, where there are "naruiturile ... cum ieste mai
sus de Galati, ce-i zic Gherghina, si pe Milcov, mai sus de
Focsani, de care pomeneste Ureche - vornicul, ca o cheama
Craciuna / the ruins [...] as it is farther than Galati,
called Gergina/Gherghina, and on the Milcov [river], farther
than Focsani, which is mentioned by Ureche-the VORNIC as being
called Craciuna"[22]. While the referrals to Craciuna stops
here, the ones concerning the city near Galati are more ample
and they interest the present investigation. "La naruiturile
cetatii de la Galati, din sus, unde cade Bârladul în Dunare /
At the ruins of the city in Galati, farther than it, where the
Bârlad [river] flows into the Danube", it was found out, as
Miron Costin points out, "o piatra mare adusa la Galati, la
biserica Dii, mai mult nu s-au putut întelege, far' de atâta,
latineste: Severus, imperator Romanorum, iar româneste: Sever,
a Râmului împarat / a big stone brought from Galati, at the
church of Dii, it could not be deciphered, just that, in
Latin: Severus, imperator Romanorum, that means in Romanian:
Sever, emperor of Rome"[23].

Whatever the lecture of the inscription made by the
humanist scholar is correct or not, it is interesting here the
Latin feature of the writing on the stone discovered in the
city of Gergina near Galati. A particular importance is
suggested by his mention that he considers that these cities
had been raised by the Dacians and by the "râmleni / the
Romans", "cum iaste deschis la Cetatea-Alba / as it is clear
at Cetatea Alba"[24].

Dimitrie Cantemir's information from Descriptio
Moldaviae confirms the Miron Costin's news and offers more
precision. First, he also refers to "orasele frumoase de
odinioara, cum o arata ruinele unor vechi cladiri (veterum
aedificiorum ruinae) / the beautiful former cities, as it is
proven by the ruins of some ancient buildings"[25]. The most
of these ancient cities are settled in the Southern
Bessarabia, so that in the territory between Dniester and
Pruth. Some of them are recent, built by the Moldavian princes
or by the Turks. Others are ancient, rebuilt by the Prince
Stephen the Great. Thus, the scholar notes, "pe râul Ialpug
... nu departe de gurile lui, sunt urmele altei cetati mai
vechi, numita obisnuit Tint. Dupa ce cazuse în ruina, Stefan
cel Mare a refacut-o; mai târziu însa turcii au facut-o una cu
pamântul / on the Ialpug river [...] not far from its mouths,
there are the remnants of another more ancient city, usually
called Tint. After it had fallen into ruin, Stephen the Great
rebuilt it; yet, the Turks definitely destroyed it later"[26].
Cetatea Alba is specially mentioned by Dimitrie Cantemir:
"numita odinioara de romani Alba Iulia, de greci Moncastron,
de poloni Bielograd / formerly named Alba Iulia by the Romans,
Moncastron by the Greeks, Bielograd by the Poles"[27]. Still,
Cantemir also offers the most details in connection to the
city of Barbosi, near Galati. "Nu departe de aici", he writes,
"la gurile Siretului se vad ruinele unei cetati foarte vechi,
care astazi este numita de locuitori Gherghina. Ca dovada ca
aceasta a fost întemeiata pe vremea lui Traian sunt monedele
dezgropate în timpul nostru din darâmaturile ei si de asemenea
o piatra de marmura cu aceasta inscriptie: Im. Caesari. Div.
Filio. Nervae. Traiano. Augusto. Germ. Dacico... / Not far
from here, at the Sereth's mouths, one could see the ruins of
a very ancient city, which now is called Gergina by the
inhabitants. As a proof that it had been founded on the times
of Trajan, there are the coins dug out in our times from its
remnants and also a marble stone with this inscription: Im.
Caesari. Div. Filio. Nervae. Traiano. Augusto. Germ.
Dacico..."[28]. The inscription is not the same as the one on
the stone depicted by Miron Costin, but it offers solid basis
for authenticity[29].

Consequently, the Romanian medieval sources confirm
the existence of some earthen walls and of some cities in the
Southern Moldavia, between Dniester and the Eastern
Carpathians. All of them are of Latin origin. Among the
cities, there are explicitly mentioned Cetatea Alba, the city
of Tint on the Ialpug river, the city of Gergina near Galati,
and the city of Craciuna. Also, the sources specify the
tradition of their Roman origin, argued either by the Latin
inscriptions near Galati, or by the Latin name of Cetatea
Alba.



b. Galati. The Origin and the Evolution of a Toponym

As a consequence of these data from the antique and
medieval sources, there is surpassed the first difficulty in
the Constantine Porphyrogenitus' text that deals with the
Roman presence in the Southern Moldavia between Sereth and
Dniester, from Trajan to the end of the 7th century. Thus, it
is removed any doubt about the possibility that some Roman
Christian vestiges in the "deserted cities" in the region in
the middle of the 10th century exist, as palpable stains of
the Rome's military and human presence. The task of
identification of these cities, as it is written to the year
950 by the cabinet savant Constantine Porhyrogenitus seems to
be much more difficult.

The Byzantine historian's work has a special place
not only among his other works, but in the whole New Rome's
historical-political literature. As it has been remarked, it
is not a work of imperial propaganda, destined to a large
public, but a confidential document that was supposed to be
read exclusively by a restraint circle of the high dignitaries
in Constantinople, involved in the state's foreign policy[30].
This is the explanation for the presence in the work of some
news coming from different secret ways in the metropolis, as
there are the ones regarding the Northern territories of the
Pont and of the Lower Danube, where the post-900 events was
characterized by a specific dynamic that vitally interested
the empire. At the same time, the fact explains also the
concrete feature of the information that clearly presents the
toponymy, retaken from the alive speaking of the populations
in the region, and not in its formal expression, borrowed from
the antique sources. The Byzantine historian explicitly refers
to the manner of collecting the information through imperial
agents (1, 18-20), to his envoys' contacts with the
Pechenegues at Chersones, Dnieper, Dniester and Danube (6,
3-5; 7, 3-8 and especially 8, 5-9) and to the presence of some
Pechenegue hostages at Chersones and Constantinople (1, 18-20;
7, 5-6). These data suppose also his information's actuality.

Still, the manner of collecting the information, of
its sending to Constantinople and its annotation in a written
form in Constantine VII's working office suppose also the
possibility of some errors or at least of different
modifications due to these successive linguistic mediations.
On the other side, the extremely heterogeneous ethnical
landscape of the region also presuppose the existence of a
borrowed toponymy. It could be possible even a translated
toponymy by the newcomers, from the native inhabitants, just
as the steppe's conquerors could impose some toponyms to the
dominated population. Just as an example, the city of Aspron,
which meant White City (Cetatea Alba) in the language of the
nomads, as the author himself informs us, knew special shapes
for each populations in the region during the middle ages:
Cetatea Alba for the Romanians, Belograd or Bielograd for the
Slavs, Maurocastron for the Greeks, Moncastron for the
Italians, and Akkerman for the Turks. While the last three
seem to rely on the late Greek form of Maurocastron, meaning
"The Black City", the Romanian and the Slavonic forms have one
and the same meaning with the Pechenegue toponym of Aspron.
Since the toponym was translated from one language to another,
there should be put the natural question, which is the
original and which are the copies? Actually, the attested age
of the Pechenegue form does not represent a decisive argument
in the favor of its acceptance as original.

Other possible errors from the toponyms' shape in
the Porphyrogenitus' text could originate in the manuscript
transmission. These errors are well known by the modern
historians, especially when it is about foreign toponyms and
anthroponyms, unknown by the Greek copiers[31]. In this
context, it should be noted that the work has not been
conserved in the Constantine VII's original manuscript[32].
The editors established that it was copied by a scribe to 980,
in an also lost manuscript. The work's most ancient manuscript
dated from the 1059-1081 period and is the working result of a
certain Michael, "servant of the Cesar John Dukas", the latter
being the Emperor Constantine X Dukas' son and the Emperor
Michael VII Dukas' brother. Just that this manuscript, which
relies on the one in 980, presents corrections, additions and
modifications belonging to the 11th-14th centuries. As the
editors consider, they come from "six different hands"[33],
and they much altered the 10th century manuscript's text[34].

These observations impose more prudence and risks
for the modern historian. Taking them into consideration, it
should be noted the names of the six "deserted cities"
(eremokastra) in the Byzantine text. They are: Aspron,
Tungatai, Cracnacatai, Salmacatai, Sacacatai and Giaiucatai.
The last five seem to be composed by two letter groups,
between which the last one is a constant, catai. Even Tungatai
contains the same letter group. Undoubtedly, it is about a
Pecheneg term that could only mean "city", since it is about
"abandoned cities" and the term is also utilized in the
explanations about the "city" of Aspron and is retaken in the
name of the city on the Dniester in Romanian, Slavonic or
Greek. Still, this Turkish term is well known in the Eastern
Europe and the Middle East. It present the form of kala / kale
or kalat / kalaat. This is the term that the toponym of
Caracal, "the Black City" originates in, probably taken by the
Romanians from the Cumans, coming from kara, "black" + kale,
"city". The other form, kal'at / kalaat, with the long
stressed syllable, is to be detected in tenths of toponyms in
the region of the Latin Kingdom of Jerusalem, meaning "city",
"fortress", "castle": Kal'at Sanjil (= Château St. Gilles),
Kal'at Jahmar (= Chastel Rouge) and others[35]. They are
created by the Seldjouk Turks, deriving from the French or
Latin names of the fortresses raised by the crusaders, just as
the name of Galata of the ancient district of Pera in
Constantinople should also have a Turkish origin, provided by
the neighbor Seldjuks or Ottomans, and should not be put in
connection with a hypothetical memory of the antique Celts,
Gallatae[36].

Consequently, the form of catai in the Constantine
VII's text could be a corrupted form of the Turkish cale /
calat. The deformation is due to one of the manuscripts'
copiers or even to Constantine VII himself, who could very
well make a confusion between one letter or another from the
informative notes. Actually, the most recent editor of the
work, that is Gyula Moravcsik remarks the many errors
committed by the copier Michael, some of them being close to
our investigation, as there are the substitution of the - e -
vowel with - ai -, and especially by the copiers that
transcribed his manuscript because of the particular forms of
the letters utilized by Michael[37]. Therefore, the - t -
letter could very well be confounded with - l -, so that - t -
with - l -, just that the - ai - ending in catai could be read
as - e -. These corrections specified, it should be passed
from the catai in the text to kale, word that mean "city" in
the Turkish languages. Still, the editor notices another
particularity in the copier Michael's writing: the rising of
the - t - letter over the other letters, just like the - i -
vowel, but also the writing of the - i - vowel in the ending
position in the form of - ï -. This could provoke the
confusion between - ï - and - t -[38], fact the allows the
reading of catai by kalat, meaning the other Turkish term for
"city".

The proposed hypothesis - the correction of catai in
cale or calat - should be verified by analyzing the
Pechenegue-Cuman toponymy in the Romanian space. The form of
cale is already detectable in the toponym of Caracal and in
some other toponyms. More important seems to be the toponym of
Galati, which is largely distributed in the Romanian space,
fact that kept the linguists' and historians' attention.
Beside the Galati toponym on the Lower Danube, settled in the
proximity of the Roman-Byzantine cities on the both sides of
the river and especially near the city of Barbosi, there are
known other five homonymous settlements in the medieval
Transylvania. They are settled on a circle arch that is spread
in the internal Carpathian side, from Bistrita to the Banat.
There have been three etymologies proposed for this toponym.
For G. Weigand and G. Kisch, it has a Celtic origin, a later
remembrance of the antique name of the Galats tribe, which
crossed the Dacian space and that was to borrow the name to
the later Constantinople's Galata[39]. On the contrary, N.
Draganu, I. Iordan, C. C. Giurescu and others consider it as
having a Slavonic origin, derived from the anthroponym of
Gal[40]. Other followers of the Slavonic origin connect the
name of the Danubian city with a supposed domination of the
Galitian Principality towards the river's mouths, so that its
name mean "the small Galici" - Galic[41]. Surely, under this
new etymology, the toponyms in Transylvania remain
unexplained. Al. Philippide and E. Lozovan propose a Cuman
origin, from the term of kalat "city", "fortress", with the K
/ G alternation, frequent in the medieval sources[42], which
would confirm our hypothesis, at least to a certain extent. It
is because, while the Danubian toponym could be explained
through the Cuman way, the five homonymous Transylvanian
toponyms could not be connected with any presence and even any
domination of the Cumans in the Romanian territories on the
internal Carpathian side.

A Celtic origin is difficult to be admitted, since
it supposes the maintenance of the Celtic tribe's memory in
the Romanian space during two milleniums. Also, in the case of
other toponyms, such as Galata, this etymology was put under
question mark. Neither a Slavonic etymology could not be
admitted, since the a > o transformation is not present in the
case of the most toponyms, while it exists in the case of the
two toponyms in the Banat (Goliecz, Golecz) and in the
Bistrita area (Golaz, Goloz, Galoz). Still, even in this
latter case, there is not a convincing explanation between
"word" and "thing", between Wörter and Sachen, essential in
the explanation of the toponyms. Henceforth, there remains to
examine the Pechenegue origin of the toponym, which would
supposed the presence of the nomad clan in all the territories
that the toponym is present.

For the toponym on the Danube, this presence should
not be demonstrated anymore, being beyond any doubt. In
Transylvania, the first of the five toponyms, settled in the
Bistrita region and present under the form of Galaz[43], is
surrounded by some settlements which names contain the
ethnonym of Besseni that designate the Pechenegues in the
Latin medieval sources. Not far of the Galati in the Bistrita
region, there is attested a villa Paganica, while in 1432 the
village would return to the form of monte Besenew alias
Heidendorff[44]. There are also two toponyms on the internal
side of the Eastern Carpathian, which are derived from the
ethnonym of Besseni[45]. The second toponym of Galati[46] is
attested in the Fagaras area, in front of the city of Fagaras,
on the right bank of the Olt river. There is a village named
Bessenbach[47] ("the river of the Pechenegues" in German) and
in the same area was undoubtedly the silva Blacorum et
Bissenorum in the Andrew II's Golden Bull in 1224[48]. The
third toponym of Galati is attested in the Hateg region in
1443, when is described as a possessio valachalis[49]. In
neighborhood, in the Hunedoara area, near another Galati[50],
there is a certain Bezenew (1509) that is then mentioned as
Oláhbeseniö (1620)[51], meaning "the Romanian Besseniö" in
Hungarian. The last toponym of Galati, in Banat[52], is
surrounded by some more toponyms that prove the Pechenegues'
presence in the region, such as terra castri Boseneu (1213),
Beseneu (1230), forum Byssenorum (1390)[53], Pechenezka
(1540)[54] and others[55]. Thus, all the five toponyms of
Galati in Transylvania are located in regions with toponyms
that derive from the etnonym of Besseni. Nevertheless, the
toponym of Galati in the territories inside of the Carpathians
rises some important questions.

First, it is always or should be near an ancient
city. For the toponym in Banat, the presence of a castrum or a
forum is explicitly attested. The same is for the Galati in
the Fagaras region, in which proximity a Romanian toponym is
present, that is "Cetatea Veche / Ancient City"[56]. The two
examples, just as the Danubian Galati, impose the idea that
all the homonymous toponyms in the Romanian space are
connected to the existence of some "ancient cities", named
Kalat by the steppes' horsemen. Secondly, it is about the
milieu in which the toponym had been preserved. Since the very
beginning, one could notice that the most of these toponyms
are settled in ancient "Romanian countries", like Fagaras,
Hateg, Hunedoara or Banat. Even in the region of Bistrita,
during the middle ages, there is also present a concentrated
Romanian population, which, according to Simon of Keza, had a
coexistence with the newcomer Szeklers in the 13th
century[57]. The medieval sources bring into light the
appearance of the toponym in a Romanian area; in Fagaras, a
Romanian-Pechenegue symbiosis is attested in 1224 in the
toponym of silva Blacorum et Bissenorum, while the Galati in
Hateg is identified with a possesio valachalis, and the one in
Alba with an Oláhbeseniö. The hypothesis is also sustained by
Romanian phonetics of the toponym of Galath-Galati, despite
the linguistic expression of the medieval sources that mention
it (Latin, Hungarian, German or Slavonic). Although, in
Bistrita and especially in the Banat, it appears in a
"Slavonized" form - Golaz / Goloz, respectively Goliecz /
Golez -, that supposes either the presence of a Slavic
population in the region, or the adaptation to the Slavonic
phonetics through the offices, at least in the Banat.

Another question is connected to the origin of the
cities that are linked with the Pechenegue-Romanian toponym of
Kalat / Galati. At the Danube area, they are clearly antique
Roman. For those in Transylvania, a Hungarian origin is out of
question, since the toponym is absent in the areas of Hungary
where the Hungarian-Pechenegue coexistence is attested. The
non-Hungarian origin of the city is clear in Fagaras, where
the "Cetatea Veche / Ancient City", in the neighborhood of the
newer city of Fagaras, is previous to the Hungarian and
Saxonian presence in the region. It also dates from the period
of the Romanian-Pechenegue symbiosis in silva Blacorum et
Bissenorum. As to the toponym in the Banat, here is mentioned
a castrum or a forum Byssenorum that excludes a Hungarian
origin of the "city". Then, should it be accepted a Pechenegue
origin of the cities? Still, the steppe's horsemen were never
rising any city anywhere. Even in the Latin East, the toponym
of Kal'at is connected to the Frankish fortifications. Most
probable, this toponym should be associated with the presence
in Transylvania of some ancient Roman or Dacian cities, like
at the Lower Danube, without excluding the possibility of some
Romanian earthly cities. In the case of the toponym in the
Banat, where one could detect a castrum Bissenorum, it could
be about the Pechenegues' settlement around or inside of such
an ancient city.

Finally, another question raised by the toponym of
Galati in Transylvania is the moment of the Pechenegue
element's penetration in the Romanian population area. At the
middle of the 10th century, Constantine Porphyrogenitus
indicates that there is a distance of four days between
Patzinakia and Tourkia (meaning, Hungary) (DAI, 37/48). While
the Pechenegue domination was extended towards the Sereth line
or even the Eastern Carpathian one in the West, the Hungarians
did not surpassed Crisana at their Eastern limit. It was
especially because the tribes of Arpad, still nomadic, were
not conversant with the mountainous areas. The region inside
of the Carpathian Mountains that covered the four days walking
between "the Pechenegues" and "the Turks" was a kind of no
man's land between the two rules of the steppe's nomads and
was previously avoided by the two Turanic clans because of its
relief and landscape. By the middle of the next century, the
position of the two rules would not be essentially modified.
It would be only to 1050 when the Pechenegue clan would move,
pressed by the coming of the Uzzes, which dislocate
Patzinakia. The largest number of the Pechenegue forces
penetrates to the South of the Danube, where it is definitely
defeated only in 1091 by Alexius Comnenos at Lebounion[58].
Meanwhile, groups of Pechenegues entered in Transylvania by
the Carpathian gorges, and organize robbery raids in the
Arpadian Kingdom. It would be only in 1068, when the
Pechenegues would be definitely defeated by the Hungarians.
The vanquished groups would be colonized at the Western
frontiers of the kingdom, paid by the Hungarian Royalty to
defend the boundaries against the German attacks. Actually, in
the area there had been installed horsemen groups of the same
race with them, still beginning with the 10th century, which
left there a toponymy of Pechenegue origin[59].

However, other Pechenegues settled in the middle of
the Romanian population in the Transylvanian "tari / terrae /
countries", much before the effective Hungarian domination in
the region, materialized in a royal administration under the
form of the counties and installed only beginning with the
12th-13th centuries. The Pechenegue elements probably
constituted in real 'leaders' of the Romanian society, fact
that is to explain the prestige of a toponym such as Kalat >
Galati, which could very well translate the Romanian toponym
of "Cetatea / the City", as it seems to be the case of the
Galati in Fagaras area, where the two forms of Galati / Cetate
are attested. The Romanian-Pechenegue symbiosis is clearly
proved in the Fagaras country, where silva Blacorum et
Bissenorum is previous to the coming of the Saxons and the
Hungarians in the region. It is possible that the Pechenegue
leaders to organize the Romanian population's resistance
against the Arpadian penetration in Transylvania. Also, it
could not be excluded the possibility that, in a later period,
after the constitution of a Hungarian ruling administration in
the province, to exist Pechenegue groups in the service of the
royalty, as some toponyms in Crisana or even in the Banat,
connected to the Pechenegue names seem to attest[60]. The
process of the inclusion of the Pechenegue element inside of
the "Romanian countries" is attested in the sources. Thus, the
Galati in the Hateg is a possesio valachalis, while the one in
Sebes is a Oláhbeseniö, not before suggesting their presence
in the toponym of Galati or in the ones that have their names
as derivation. Still, this process of assimilation of the
Pechenegues was slow, whether it is observed their presence in
the Hungarian armies in the 13th century[61], their mention in
Fagaras in 1224, in silva Blacorum et Bissenorum, or the fact
that the Saxons created some toponyms that derive from the
ethnonym of Beseni, under the form of "the pagans' village"
(Heidendorff) or, in Latin form, of villa Paganica. It would
be only after their christianization, probably in the
13th-14th centuries, the Pechenegues' assimilation in the
Romanian milieu in Transylvania would be faster.

The relationship between "name" and "thing" and
between the Romanians and the Turkish clans in the Northern
Danubian space in the clarification of the toponym of Galati
is also clear in the case of another Romanian toponym, that is
Calafat, although this latter should be put into connection
the Romanian-Cuman relationship. The new clan of the Cumans
that substitutes the Pechenegues at the Lower Danube in the
second half of the 11th century extend its hegemony towards
the West to the river of Olt, so that the Wallachian Field
becomes a Cumania before the Tartar invasion. The toponym of
Caracal - Cara + cale, "the Black City" is into connection
with the Cumans. It belongs to the same semantic family of
cale / calat "city". The Cumans, opponents to Constantinople
and allies of the Wallachian-Bulgarians in the South of the
Danube, passed the Danube in their robbery expedition through
a ford in front of the city of Vidin, the ancient Roman
Bonnonia, where the toponym of "Vadul Cumanilor / the Cumans'
ford", nowadays Comana, is attested on the left side of the
river[62]. There are nowadays two Romanian toponyms, Cetatea
and Calafat near this ford. In the perimeter of the village of
Cetatea, it was discovered some Roman vestiges belonging to
the 2nd-3rd centuries[63]. It is to be supposed that the
Romanian toponym is associated with the presence here of a
Roman fortification, the pair of the much more known antique
city of Bonnonia, on the right side of the river. The
existence of some pairs of Roman-Byzantine cities on the two
banks of the Danube is a frequent phenomenon. The other
toponym, that is Calafat, which has not satisfactorily
explained, could only originate in the Turkish word, come from
Cuman way, of Kalaat, received by the Romanians under the form
of "Calafat". The Cuman only retook the Romanian in their own
language the toponym of "Cetatea", existed among the natives
by nowadays, they preserving also the Cuman name of the place.
Undoubtedly, it is not excluded that the steppe's horsemen to
build here a fortress in order to control the traffic on the
Danube, "pe drumul Diilui / on the way of Diiu" in the
Romanian medieval documents. Anyhow, Galati / Caracal /
Calafat belong to one and the same semantic family and are
toponyms preserved by the Romanians from the Pechenegue-Cuman
language.



c. "The deserted Cities"

We already established the inseparable connection
between the toponym of Galati and the Pechenegue presence at
the Lower Danube and in Transylvania in the 10th-13th
centuries. Consequently, the toponym originates in the word of
calat, also present in the case of the "deserted cities"
(eremocastra) in Moldavia, although there could not be
definitely excluded the correction of the word catai to cale,
the latter and the Turkish calat / calaat being semantically
alike. Henceforth, we have the right to read the six cities in
the Constantine the Porhyrogenitus' text as Aspron, Tung,
Cracna, Salma, Saca and Gieiou. Let us make an attempt to
identify them as far as possible, in the light of the ancient
and medieval sources.

"The City of Aspron", perhaps Asprokalat in the Pech
enegue language, does not present any identification problem.
It is settled on the Dniester's Moldavian bank, where it is
placed by the Byzantine author, and it does not represent
anything else than Cetatea Alba for the Romanians, Belgorod /
Bielgorod for the Slavs, Maurocastron for the Byzantines,
Moncastron for the Italians, Akkerman for the Ottoman
Turks[64]. The river of Aspros is also mentioned by
Constantine VII (DAI, 9/91) in its proximity, still the town's
name comes from the antique city's walls, as the Byzantine
historian explicitly indicates. It is not difficult to
conclude that the city's name has the same meaning for the
Pechenegues, Romanians and Slavs, that is "the white city",
while it takes the meaning of "the black city" for the Greeks,
Italians and Turks. In the latter case, it is probable that
the city be renamed by the Greeks after the 10th century, the
meaning being then retaken by the Italians and the Turks. On
the contrary, the endeavor of the name giving to the medieval
city remains unsolved. The city at the Dniester's mouths is
known by the Greeks and the Romans in the Antiquity, because
of its settlement in the contact area between the Northern
Pontic steppes and the sea. Its antique remnants, Greeks and
Roman, have been discovered and researched[65]. Its strategic
position explains the importance in the Moldavian defensive
system in the 15th century and later in the Ottoman one. Its
impressive fortifications built by the Romanians and the
Turks, preserved by nowadays, stands as testimony.

Among the other five cities' names, Cetatea Saca is
the most important. The toponym of Saca / Seaca is present all
around the Romanian medieval period. It is not anything else
than the Romanian adjective of "sec / seaca", meaning "dry",
belonging to the same word family like the verb of "a seca",
meaning "to drain" - to dry a river's or a lake's water or the
tree's sap. It originates in the Latin sicco, -are, just like
the adjective of siccus. In Romanian, it often appears in the
toponyms of Valea Seaca / Saca, Apa Seaca or Râul Sac: a river
and the neighbor village, another village and so on[66]. In
Wallachia, it is present in many toponyms in the form of
Seaca, but also in the name of the village of Seaca / Saca,
regarded as "the deserted village", with the conservation of
the diphthong of [ea]. On the contrary, in Transylvania there
is the same form like in DAI and in the Moldavian toponymy of
Saca / Zaca, with the S / Z interchange, known in the Latin
and Hungarian sources[67]. The same toponymic family also
includes Secatura / Sacatura / Secatura, and in the
Transylvanian toponymy there is also Zakatura[68]. Having an
exceptional frequency on the two rages of the Carpathian
Mountains, from Bukovine towards the Banat, this toponym,
together with the one of Runc, also of Latin origins, defines
a cleared land by the draining of the forest by the human
being[69].

What is the meaning of Saca in the toponym of
"Cetatea Saca"? The connotation could only by the one of
"deserted", "abandoned", "emptied", "waste" city.
Nevertheless, the remarkable fact is the identity between the
meaning of the Romanian toponym and the Greek term for
"deserted", "waste" cities (eremocastra) in the Constantine
VII's text. The fact allows us to suppose that the Byzantine
historian simply translated the Romanian toponym. It is clear
that the Pechenegues simply retook the city's name from
Romanian, which is present in the scholar emperor's text in a
Romanian-Pechenegue mixed form, that is Sacacalat, "Cetatea
Saca", that is "the Deserted City". We are to emphasize below
its identification.

The third on the list, the city of Cracna is
difficult to be identified, because of the form that the
toponym presents in the text. It could be the Craciuna in the
Moldavian sources, transmitted in the form of Crac[iu]na,
especially because it is retaken in a close form in a
Moldavian chancellery's act dated 1416, that is Crac[u]na[70].
While the solution of the manuscript transmission seems to be
satisfactory, the difficulty comes from another point. The
later Romanian city Craciuna, which was for a long time the
dispute object between Wallachia and Moldavia, was located on
the Milcov river, too far from the Danube's mouth, although it
was in the proximity of a Roman "troian". By its geographic
position and the Romanian medieval sources' testimony, the
identification with the city of Barbosi near Galati, the
medieval Gherghina, seems more acceptable. It is also present
in the Latin of Dimitrie Cantemir, transcribed as Gergina. The
transformation from the Romanian Gergina to the Greek Krakna
looks possible, whether the G / K interchange in the Byzantine
historian's text is taken into consideration. Among other
cases, this interchange is present in the name of an Armenian
prince, that is Grigorios / Krekorikios (DAI, 43, 7). It is
also detected in the numerous deformation of the human and
places' names in the work, either due to the errors of
transmission from the informer to the author's working
cabinet, or to the successive copies of the Constantine
Porphyrogenitus' manuscript. Therefore, the Romanian
transcription of Gergina / Gherghina, frequent in the medieval
anthroponymy and toponymy, to Krakna looks possible.

The solution is sustained by the presence in the
surroundings of the city of Galati of the antique relics -
Latin inscriptions, Roman coins -, remarkably documented in
the Romanian medieval sources and the modern archeological
discoveries. It is also sustained by a toponymic argument. The
present day name of the city is Barbosi, some centuries ago
attested. The toponym is very spread in the subcarpathian
regions in Moldavia and Wallachia. Marele Dictionar Geografic
al României [The Great Geographic Dictionary of Romania],
issued a century ago, mentioned some tenths of them[71].
Barbosi is nothing more than a translation of the Hungarian
toponym of Sakall / Zakall, which, at its turn, relies upon
the Romanian Saca that we dealt on other occasion[72]. The
presence of the Hungarian toponym in these regions, where many
Hungarians, Romanians and Szeklers from Transylvania was
established during the middle ages, is connected to the
Hungarian Kingdom's interests in the corridors in the
extracarpathian space that permitted them the acces towards
the Danube's mouths through the Buzau and Sereth valleys.
Louis of Anjou's privilege accorded to the merchants in Brasov
in 1358 attests the presence of the Transylvanian businessmen
at the Sereth's river mouth to the Danube, so that at Galati.
Here the continuity Rom. Saca > Hung. Sakall / Zakall > Rom.
Barbosi is thus documented. In this case, there is remarkable
the presence in the Romanian medieval toponymy of the city of
Gergina / Cracna's name in the 10th century, but also of the
toponym of Saca, changed in the present day in Barbosi through
the Hungarian Sakall. Moreover, the Romanian medieval city of
Galati was built in their vecinity, and it should be connected
with the Pechenegue toponym of Kalaat, "the City". Also here,
the Pechenegue name is nothing else than a retaking of the
ancient Turris, "the Tower", "the City", borrowed by the
steppe's people from the descendants of the ancient Roman
population at the Danube's mouths.

The second in the Byzantine historian's list, the
city of Tung or Tunc seems to be the same with Tint, mentioned
in the Cantemir's work and settled at the Ialpug river's
mouths, at the river mouth in the Black Sea. As we already
noticed, for the Moldavian erudite, it is an ancient city,
rebuilt by Stephen the Great and entirely destroyed by the
Turks, when they conquered the Bugeak after 1538. Its memory
is retaken during the 18th century in many documents. The most
important document is dated 1759, in connection with the
estate of "Tentil". The latter extended "de lânga troian
[valul lui Traian], pe Cahul, despre rasarit / from the trojan
[the Trajan's wall], on the Cahul, towards the East" and that
also comprised the village of "Barbosi" on Ialpug in its
enclosure[73]. The identification between the city of Tintil
and the "deserted city" in the Byzantine historian's work is
supported by the presence of the village of Barbosi on Ialpug
in the enclosure of the 15th-16th centuries Moldavian
fortification. As in the case of its homonym near Galati, the
toponym of Barbosi relies on the evolution Rom. Saca > Hung.
Sakall > Rom. Barbosi. The Hungarian influence in the Southern
Bessarabian toponymy should be connected with the Hungarian
Kingdom domination at Chilia and the surrounding area during
the 15th century. The value of this testimony is determined by
the fact that it attests the city's existence on the way
between Cetatea Alba and the Danube's mouths, near the earthen
wall built by the Romans in the Southern Moldavia for
defensive purposes. It is difficult to specify the toponym's
meaning that does not seem to have Romanian origins. Anyhow,
we are not to know whether the Pechenegue or the Byzantine
form be original, which should suppose a deformed transmission
to the Romanians, or, on the contrary, a transcription error
of Constantine VII.

The last two cities, Salma and Gieou, raises other
kind of problems in their identification. While the three
"deserted cities" that we proposed an identification are in
the Southern Moldavia, between Dniester and Sereth, the two
seem to be settled on the Danube's right bank, in the North of
the Scythia Minor. Salma could be the ancient Thalamonium,
identified with the city at Nufarul, on the river's Southern
branch, that is St. George, taking also the medieval Th / S
interchange into account, which could lead to the form of
Salamonium. The difference between Salma and Salamonium could
be an objection. Still, it is necessary to do not regard the
form in the antique Latin and Greek texts, but the one that
was in use in the inhabitants' way of speaking in the 10th
century. In the same region of Scythia Minor, the city of
Carsium was spelled as Cars, as it is often mentioned in the
sources and as it represents the basis for the Slavized modern
form of Hârsova. Therefore, the ancient city's name could be
in use the inhabitants' spelling under the form of Salama or
something, fact that would explain the toponym transmitted as
Salma in the Greek text.

Transcribed as Gieou, the other city could a
corrupted form for Aegyssus (the present day Tulcea), another
ancient city on the Danube's same branch. In the natives'
language, the toponym could be in use under the form of Igis
or Egis. In the Byzantine historian's transcription, the
initial vowel fell and the toponym took the Genitive form of
Gieou, as it is present in the name of the city of Axiopolis >
Axioupolis[74]. This identification is supported by the
material remnants brought to light by the archeologists as the
massive walls of the ancient cities[75].

Anyhow, the major obstacle is represented by the settling of
the two cities on the river's right bank. Still, the
difficulty is diminished whether some details connected to the
limit between Patzinakia and Bulgaria in the Byzantine text
are taken into consideration. On the one hand, the historian
affirms that the Pechenegues' domination extends towards the
neighborhood of the Bulgarian city of Silistra on the Danube
(DAI, 42/20-21), on the other hand, he asserts that there is a
half day distance between the two rules (DAI, 37/48). The fact
made the experts confused. Still, the deadlock could be
surpassed whether we admit that this no man's land of a half
of day distance is settled in the North of Scythia Minor,
having a totally different relief and landscape than those of
steppe in the Northern half of the region. The last two
"deserted cities" were to be found in this no man's land
between Patzinakia and Bulgaria, in a territory not entirely
unknown for the Pechenegues. It should be added that the
maintaining of the two ancient toponyms in the 10th century is
not to be a singular case, whether we take into account the
city of Carsium > Cars > Hârsova, which name has been
preserved by now in the region's toponymy, or the name of the
more distinguished antique city of Durostorum / Darstor /
Silistra.

In connection to the six "deserted cities", it is
necessary to specify the place where Cetatea Saca was located.
The most plausible version is its identification with the
medieval Isaccea, also settled on the Danube's bank in
Dobroudja. The medieval city was situated on the antique
Noviodunum's settlement, having a very important strategic
position in Scythia Minor, since it controlled the passage way
on the Danube's most important ford in the mouth river's area.
The antique city was abandoned during the 7th century, no
later than once with the arrival of the Asparuch's
Protobulgarians. Thus, at the middle of the 10th century, it
was an "deserted city" or, in the Romanians' language, a Saca.
After 971, when the Byzantines return at the Danube as
military power, the city is rebuilt and has the same
importance in the New Rome's defensive system. The new
fortifications and the huge quantities of Byzantine coins in
the region stand as testimonies[76]. The name of Satza, a
leader of the revolt in Paristrion against Byzantium on
1072[77], is to be probably regarded as the toponym's Greek
form retaken from the Romanians: Rom. Saca > Gr. Satza. Under
the circumstances of the Constantinopolitan power's decay at
the Danube after 1204, the medieval city fails in importance,
but a century later, to 1300, the Tartar Khan Nogai and one of
his sons establish here their residence and a coinage
workshop. The city is mentioned in the Eastern sources as
Saqcia / Sacdji[78], that probably relies on the Romanian
toponym of Saca, also present in the Constantine
Porphyrogenitus' text. The later Ottoman form of Isaccea
supposes an original Romanian Saca, which the Turks took the
present form, on the pattern of Gr. Smirna > Tk. Izmir, Gr.
Nicaea > Tk. Isnik or Gr. Vlachia > Tk. Iflak.

Whether the six "deserted cities" in the Byzantine
text are attentively regarded, there are three of them settled
on the North of the Danube, while the other three are to be
detected on the South of the river. The first three - Aspron /
Cetatea Alba, Tunc / Tintil and Cracna / Gergina - certainly
belong to the Pechenegues' domination area, while the <br/><br/>(Message over 64 KB, truncated)