Archaeological Survey on the Lower

Danube: Results and Perspectives

Sven Conrad
More than 40 years of extensive archaeological excavations carried out in
the legionary camp of Novae (near Svitov, Bulgaria) and the late Roman
fortress of Iatrus at the mouth of the Jantra River (ancient Iatros), about 15
km downstream from Novae, have produced important contributions to the
history of the Roman frontier on the lower Danube.
1
In the mid-1990s the
Roman-Germanic Commission Frankfurt/M. of the German Archaeological
Institute developed an interdisciplinary programme to investigate the links
of both fortresses to their surrounding territories. The programme was car-
ried out in collaboration with the Archaeological Institute of the Bulgarian
Academy of Sciences in Sofia, the Historical Museum in Ruse, the Archaeo-
logical Institute in Bucharest, and other institutions, with the support of the
German Research Foundation. The aim of this paper is to give a summary
for the periods from the later Iron Age up to the early Middle Ages after the
end of the fieldwork, including a short preview on perspectives of further
similar research enterprises in the Eastern Balkans and the Black Sea region
as well.
2
The research area lies on both sides of the Danube in north central Bulgaria
and south Romania, about 180 km from the western Black Sea coast (Fig. 1).
The pick-up surveys covered an area of about 1,200 km
2
between Svitov
and Ruse along the lower Danube with a length of about 50 km and a width
up to about 30 km on the Bulgarian side. On the Romanian side we carried
out a short survey along the first terrace and also in the lowlands along the
Danube. The borderlines of the area originated more or less by chance; they
are restricted in some cases by marked geographical units.
The investigations started with a short explorative survey in the autumn
1997 followed by six seasons of intensive field surveys, limited geophysical
prospections, a site specific grid survey, air photography, palynological and
soil analyses. During the fieldwork we used the geographical method, i.e.
the dividing of daily work according to geographical units (valleys, hilltops,
etc.). We decided to undertake the survey in an extensive way for two rea-
sons: 1. We had only a vague notion of what we could expect; 2. We wanted
to cover an area which allowed reliable conclusions regarding the settlements
systems in different historical periods.
The best season for fieldwalking is from the beginning of March until
310 Sven Conrad
the beginning of April. This is due to the fact that the snow has melted and
the earth is broken up by the frost, and there is still no vegetation. After the
sprouting of the winter grain and the beginning of drilling, effective work is
impossible. The autumn period from mid October until late November can be
used for control surveys mainly, but the visibility is limited after the harvest
and the new cultivation of soil.
The research area is influenced by continental climate. It is mainly char-
acterised by the wide river valley of the Danube and the Jantra River run-
ning from the Balkan range through the hilly north Bulgarian lowlands.
In earlier days, the lowlands beside the rivers were marshy, now they are
drained by amelioration and dikes. In the late glacial epoch an up to 60 m
thick layer of loess soil was deposited over the limestone rocks, which ap-
pears very often along the river valleys. This not only provided flints but
also the main building material for the later historical periods. These loess
hills with precipitous slopes to the north and slightly inclined slopes to the
south rise up to an altitude of about 150 m near the Danube and up to 300
m south-east of Bjala.
The soil is very fertile; in the west and north-east of the research area we
mostly find calcareous ernozems, in the eastern part leached and eroded
ernozems. The insignificant differences in the soil fertility seemed to have no
influence on the behaviour of occupation during the historical periods. The
western part includes wide open valleys with slightly inclined hill slopes, and
Fig. 1. The position of the research area within the Roman defense line at the lower Danube
(K. Ruppel, RGK Frankfurt/M.).
311 Archaeological Survey on the Lower Danube
in the east, one finds deeply incised narrow rocky valleys and up till today
partly preserved woodland areas. Probably for this reason the Jantra River
functioned as the western border of the Eneolithic culture of Kodadermen/
Karanovo VI/Gulmenia what was proved by our expedition.
3
Indeed, most
of the territory is woodless nowadays. We hope to settle the vegetation his-
tory by the palynological analyses, which are still in process.
The survey on the lower Danube including several methods of prospection
is the up to now most extended project in the eastern Balkan and western Black
Sea region. In the last decade several research enterprises were undertaken,
but on a smaller scale and mainly focused on special scientific problems or
selected historical periods.
4
STATISTICAL SUMMARY OF RESULTS
The primary research conception mainly focused on the historical periods
relating to the fortresses (Roman and Medieval periods). The project was
among other things inspired by the desire to
compare the well-known building periods of Iatrus and Novae with the
envelopment of the settlement networks in their hinterland and to char-
acterise the interaction between frontier zone and rural country-side;
compare the results of a site-based survey programme in the region of
Nicopolis ad Istrum carried out by a British-Bulgarian team;
5
reconstruct the ancient landscape and environment;
draw conclusions about the society and economical situation in each his-
torical period.
Additionally, during the practical fieldwork it became obvious that there
were excellent conditions to investigate the prehistoric periods from the late
Palaeolithic onwards as well.
6
Total number of
recorded sites
Tumuli Settlements, for-
tifications,
necropoleis
Technical sites
(roads,
bridges, pipe-
lines, etc.)
Number % Number % Number %
Before 1997 300 45 208 80 92 24 ---
1997-2001 364 55 52 20 289 76 23
Total 664 100 260 100 381 100 23
Table 1
312 Sven Conrad
Before we started our survey, only a few but non-systematic surveys and
information about finds in the research area were published.
7
A total of 300
sites were recorded in the Archaeological Maps of Bulgaria and Romania
(Table 1). The majority of the sites registered were tumuli (208); to this can
now be added further 52. The most important and impressive result is the
addition of 289 new sites (76%), which can be related to settlements, fortifi-
cations and necropoleis. 23 new sites belong to infrastructural and industrial
installations like roads, pipelines, bridges and quarries.
The major part of these up to now registered 381 find spots (settlements,
fortifications, necropoleis) were occupied in more than one period (Figs. 2
and 3). If we classify and distinguish these sites according to periods, we get
a total number of 1183 (before 1997: 193). The increasing number up to the
early Medieval times can probably be explained by the higher concentration
of settlements, the growth of population, and better circumstances for sur-
face finds for the more recent periods. On the other hand, with the increasing
chronological distance there are a higher number of sites covered by alluvial
and eroded material.
It is very difficult to identify individual settlements on the north bank of
the Danube and we failed to differentiate between settlement types. Hence
the following data mainly focuses on the southern (Bulgarian) part of the
research area.
THE ARCHAEOLOGICAL LANDSCAPE FROM THE LATER
IRON AGE UP TO THE EARLY MIDDLE AGES
Later Iron Age
(Hallstatt C/D La Tne; about 600 BC-early 1st century AD)
The drop of the number of settlements in the early Iron Age as compared
to the Bronze Age (Fig. 2) can be attributed to the very short duration of the
Hallstatt period in our chronological system (i.e. Hallstatt A-B) and probably
to the migration of peoples during the transition phase from the Bronze to
the Iron Age at the end of the 2nd millennium BC. We distinguished earlier
and later Iron Age finds mainly by the grey wheel-made pottery that appears
already at the end of the 7th century BC. A reliable dating of the handmade
pottery is still not available.
The settlement system of the later Iron Age especially during the
so-called Classical Thracian period is dominated by central fortifications
on hilltops and on high riverbanks (Fig. 4). We identified the most impres-
sive site to the north of the town Bjala on the right bank of the Jantra River
(Fig. 5). This site was recorded in the Archaeological map as a late Iron Age
settlement and a late Roman fortress
8
but our survey proved it only as a
late Iron Age site. It consists of an almost regular stone wall square with a
313 Archaeological Survey on the Lower Danube
side length of about 200 m with corner towers. The fortification has been
destroyed by stone robbery. There are large amounts of pottery fragments
on the surface, mostly of good quality; among them fragments of Greek
amphorae. In our research area similar fortifications probably existed on
the hilltop Kaleto in Svitov
9
which is covered by a Medieval fortress, and
near Zimnicea (north of the Danube).
10
Small and up to now unpublished
excavations of Kaleto from the early 1990s produced pottery finds similar
to those found in Bjala.
The extended tumular necropoleis concentrated in the wider vicinity of the
fortifications contain numerous conspicuous tumuli higher than 3 m which
can be referred to the upper class of the tribes (Fig. 6). About 5 km east of Bjala
near Borovo, in the area of a tumular necropolis, the famous gold treasure of
Borovo was found. It is assumed to be a diplomatic gift of the Odrysian king
Kotys I to a local ruler.
11
To each fortification presumably belonged a limited domain since no set-
tlements existed within a radius of at least 3 km (Fig. 4). Most of the larger
settlements with an area of about 10,000 m
2
are arranged in an approximately
regular network with average intervals of 4-5 km (Fig. 4). There is a striking
concentration of settlements of this size north and west of the Bjala fortifica-
tion. Small settlements (less than 5,000 m
2
) are situated mostly in the hilly
landscape north of Bjala and in the valleys south of Svitov, partly without
18
59
94
74
178
49
13
226
69
49
15
136
203
0
50
100
150
200
250
N
e
o
lit
h
ic
(
a
p
p
r
o
x
. 6
0
0
0
-
4
8
0
0
B
C
)
C
h
a
lc
o
lit
h
ic
(
a
p
p
r
o
x
. 4
8
0
0
-
3
2
0
0
B
C
)
B
r
o
n
z
e
A
g
e
(
a
p
p
r
o
x
. 3
2
0
0
-
1
2
0
0
B
C
)
H
a
lls
t
a
t
t
(
A
-
B
)
(
a
p
p
r
o
x
. 1
2
0
0
-
6
0
0
B
C
)
L
a
t
e
r
Ir
o
n
A
g
e
(
a
p
p
r
o
x
. 6
0
0
-
b
e
g
. C
1
s
t
)
P
r
e
h
is
t
. s
it
e
s
(
u
n
d
e
f
.)
R
o
m
a
n
(
b
e
g
. C
1
s
t
-
b
e
g
. C
4
t
h
)
L
a
t
e
R
o
m
a
n
(
b
e
g
. C
4
t
h
-
m
id
C
5
t
h
)
E
a
r
ly
B
y
z
a
n
t
in
e
(
e
n
d
C
5
t
h
-
e
n
d
C
6
t
h
)
E
a
r
ly
M
e
d
ie
v
a
l (
C
8
t
h
-
b
e
g
. C
1
1
t
h
)
L
a
t
e
m
e
d
ie
v
a
l (
C
1
2
t
h
-
C
1
4
t
h
)
O
s
m
a
n
ic
(
C
1
4
t
h
-
C
1
9
t
h
)
M
o
d
e
r
n
Periods
T
o
t
a
l

s
u
m

o
f

s
i
t
e
s
Fig. 2. Sites north and south of the Danube according to periods (settlements, fortifications,
necropoleis; without tumuli and technical sites; total number = 1183).
314 Sven Conrad
direct access to water. They produce only small quantities of finds. This is
an indication of a short-time occupation and can probably be attributed to a
half-nomadic way of living.
We can interpret these fortifications as tribal centres of the Thracians.
According to ancient sources (Hekat. FGrH 1 F 170), the territory between
the Jantra River and the chora of Odessos at the Black sea coast was settled
by the tribe of the Krobyzoi.
12
But the function of, for example, the Bjala
fortification as a main or a subordinate centre of this or another tribe can be
proved by archaeological excavations only. An important fortress which is
also assumed as a tribal centre of the Getes has being excavated for several
years near Isperih in the Sborjanovo region (north-east Bulgaria).
13
A lim-
ited survey proved a concentration of simultaneously existing settlements
in its vicinity.
14
Fig. 3. Total number of sites recorded before and after 1997.
315 Archaeological Survey on the Lower Danube
Chr. Popov who recently examined the few up to now recorded proto-ur-
ban centres of the early and late Iron Age in Bulgaria suggests a circulation
of the small, temporary settlements around these fortifications which prob-
ably functioned as cult centres as well.
15
The settlement systems in the Roman (early 1st-early 4th century AD)
and late Roman Periods (early 4th-mid 5th century)
The history of the Roman limes in this section of the Danube begins with the
foundation of the province of Moesia in AD 44 and the dislocation of the legio
VIII Augusta in Novae (Fig. 7) simultaneously or within a short time after that.
The territory of the Moesian province and its border along the Danube became
Fig. 4. Sites of the later Iron Age (end 7th BC-beg. 1st century AD).
316 Sven Conrad
extended to the Jantra River. The organisation of the military protection along
the Danubian bank east of the Jantra is unknown for the early Roman period.
Generally, a dislocation of troops is accepted in this area in the Flavian period
as supported by archaeological evidence. Military garrisons were permanently
dislocated along the Danube in the eastern part of Moesia after the Dacian
wars of Emperor Trajan (102-106).
16
The closed military system of the frontier fortifications had a great influ-
ence on the settlement system, especially as an economical factor (Fig. 8). There
was a fundamental and persistent need for the supply of the maintaining mili-
tary installations both from the agricultural hinterland as well as from further
afield. The map we devised according to the finds obtained from the survey
shows a relatively dense and regular settlement system in the hinterland of the
limes and reflects very likely the situation in the 3rd century.
17
Sites of larger
extent (vici or large villae) were found mainly along the great river valleys and
the wide-open valleys south of Novae. Small places like villae are situated in
the areas of water sources in small side-valleys. Especially in the area east of
the Jantra characterised by deep narrow valleys the sites are spread at greater
intervals (average interval about 5-6 km). In any case the water supply or the
possibility for access to water is a precondition for settlement.
Fig. 5.
Later Iron Age fortifi-
cation north of Bjala.
Rough drawing (right) of
S. Stefanov (1956, 14 fig.
6) and aerial view of the
mid-1940s (left) (Cour-
tesy of the Military Car-
tographic Service of the
Bulgarian Army).
b
a
317 Archaeological Survey on the Lower Danube
The consolidation of the Roman frontier after the wars of the 3rd century
under the reigns of Diocletian and Constantine I was a supposition for the
continuation or the renewal of most settlements at the beginning of the late
Roman period. As on the middle of the Danube small garrisons and observa-
tion posts were converted into fortresses. The Iatrus fort at the mouth of the
Jantra River was erected (Figs. 9 and 10) and the limes road improved.
18
The
eastern Balkans were of crucial importance in the late Roman period since
the new Imperial capital Constantinople was situated not more than 500 km
south-east of the Danube. With this background the lower Danube conse-
quently enjoyed Imperial support more than other parts of the later empire.
But probably due to the width of the river the defence system on the lower
Danube obviously did not reach the level and density of forts and burgi in
Pannonia.
19
Fig. 6. Tumuli of the later Iron Age (c. 600 BC-beg. 1st century AD).
318 Sven Conrad
In any case, the settlement network was affected by the war; however, the
previous opinion that there was a continuous decline of the settlements in the
rural territories during the 3rd century and first half of the 4th century can not
be sustained (Fig. 10). In fact, the need to deliver supplies to the reinforced
military troops at the border caused a revival of the settlement network. In
some of the settlements, we established a reduction in the areas inhabited;
but altogether continuity from the Roman to the beginning of the late Roman
period can be surmised.
Roman pottery fragments were also found on several sites on the left bank
of the Danube in the Dacia libera. Bricks and carefully treated stone material
characterising Roman settlements south of the Danube is missing. These finds
can be related to the trade across the frontier mentioned by ancient authors
(Them. Or. 10.136b; Amm.Marc. 20.11.8; 21.9-11). Roman soldiers seem to
have profited in this trade with the Goths.
20
Caused by the Gothic wars in the second half of the 4th century and the
Hunic assaults in the first half of the 5th century, the extent and intensity of
the settlements declined, and later on all settlements in the remote frontier
areas were apparently abandoned. The remaining people settled in or nearby
the fortifications; and new fortresses were erected, for example in Svitov
(Theodoroupolis?), Polsko Kosovo and Koprivec. This situation corresponds
Fig. 7. Aerial view of the legionary camp of Novae. In the foreground the valetudinarium
(2nd-3rd century) overlaid by a late Roman villa urbana; in the background the late Roman
Christian basilica.
319 Archaeological Survey on the Lower Danube
with the Iatrus fort in the first half of the 5th century when all free areas be-
tween existing buildings and in the ruins of horrea I and VII were filled with
dwelling houses.
21
A higher density of mountain fortresses can be found in
the northern foreland of the Balkan range, a suitable retreat area in a period
of anxiety.
22
Iatrus and the other remaining fortresses and settlements were totally
abandoned by the Huns in the mid 4th century. By contract the Romans had
to give up the limes fortifications and a territory with width of three days
journeys south of the Danube.
23
Fig. 8. Roman sites (beg. 1st-beg. 4th century AD).
320 Sven Conrad
Fig. 9.
Aerial view of the Iatrus fortress from south
(G. v. Blow, RGK Frankfurt/M.).
Fig. 10. Late Roman sites (beg. 4th century mid-5th century).
321 Archaeological Survey on the Lower Danube
Land plots in the rural and suburban territories of Novae
In the rural territories south of Novae we found signs of a systematic land
partition along the brook valleys (Fig. 11), which is very similar to the hin-
terland of Speyer (Germania superior).
24
The average intervals are only 2-3 km.
According to our reconstruction of the cadastre plan, 10 of the 38 estates
investigated belong to medium-sized properties of about 200-380 ha, 24 to
small-sized plots with between 50 and 200 ha, and four with less than 50 ha
are very small.
25
It can be supposed, that veterans established a part of these
rural properties after their honesta missio.
The bigger part of sites in the suburban vicinity of Novae had most likely
another status (Fig. 12). The canabae of the legionary camp at Novae extended
approximately 1.5 km along the road to Svitov; the route is roughly identical
to the former limes road. Only several mounds interrupted the occupied area.
Parts of the necropolis were found in the south. There are also some indica-
tions that the area covered by the late Roman extension Novae II belonged
to the former civilian settlement of the castrum. Deduced from the examined
pottery, on the whole we can calculate an area of about 70-80 ha for the civil-
ian settlement of the 1st century-early 4th century and an area of about 20-30
ha for the late Roman site.
26
Fig. 11. Rural sites south of Novae (Roman to late Roman period). The distance between two
lines of the grid system is 1 km.
322 Sven Conrad
A total of 32 sites within an average interval of only 300-500 m were reg-
istered south and south-east of Novae (total of Roman and late Roman sites;
Fig. 12). The size of the land estates is mainly ascertained between 5 and 30
ha; only six plots span between 30 and 50 ha. This indicates small-scale agri-
culture, and probably there also existed a system of handicraft workshops.
Presumably veterans who settled down after their active military service
owned these properties and were, thus, available in cases of emergency. The
interest in occupying a plot near the military camp caused an enormous set-
tling constraint reflected by the average geographical data (Table 2; Fig. 13).
The long distance to fresh water, the average altitude and the predominant
position of the sites at the middle slope indicate a suburban settlement struc-
ture with an artificial water supply system.
The main part of these villa sites lies within a radius of approximately
2.22 km around the groma of the camp adequate to the ancient unit of mea-
surement leuga (approximately 1.5 Roman miles). According to I. Piso the area
within the radius of 1 leuga belonged to the ager publicus and was controlled by
the camp administration. The inhabitants of the canabae, which had a similar
legal status like the veterani et cives Romani consistentes, had no right for owner-
ship of the ground within this radius. Consequently, the extended vicus was
Fig. 12. The suburban area of Novae (Roman to late Roman period). The distance between two
lines of the grid system is 1 km.
323 Archaeological Survey on the Lower Danube

Geographical factors Roman
Period
Late Roman
Period
Villae sub-
urb. s-e of
Novae
Villae rusti-
cae south of
Novae
Number of sites 113 148 35 38
Altitude (in m; average) 90 89 99 80
Distance to waters (in m;
average) 234 260 541 173
Altitude above waters
(in m; av.) 19 21 37 6
Slope (in %; average) 4 4 3 5
Size of real estates (in
ha; average) --- --- 25 159
Table 2
Roman and late Roman rural sites
south of Novae
84%
16%
Roman Period
58%
5%
31%
2% 4%
Suburban villae south-east of Novae
31%
63%
3% 3%
1
2
3
4
5
6
Roman Period
58%
5%
31%
2% 4%
Fig. 13. Position of Roman sites south of the Danube, with a distinction of the rural territory
and suburban area south/south-east of Novae: 1 down slope/valley/lowlands; 2 terrace;
3 mid slope; 4 upper slope/ridge/summit; 5 high bank; 6 end of ridge/spur.
324 Sven Conrad
situated about 2.5 km east of the legionary camp on the Danubian bank and
thus outside the leuga radius (Fig. 12). After the bestowal of municipal status
to the canabae of a legionary camp, this injunction had to be repealed and the
legal status of canabae and vicus equalised.
27
The existence of Roman suburban
villae in the leuga radius supports the theory that the canabae of Novae received
the municipal status under the reign of Septimius Severus.
28
However, this
can only be verified following finds of new inscription.
Early Byzantine Period (end of 5th century-end of 6th century)
The reconstruction of the limes at the Danube took place at the end of the 5th
century (Fig. 14). The settling was limited to the fortresses and fortified settle-
ments and their immediate vicinity. It is impossible to state a possible trial for
Fig. 14. Early Byzantine sites (end of 5th-end of 6th century).
325 Archaeological Survey on the Lower Danube
the revival of former settlements by surface finds. In these late times of the
Byzantine presence north of the Balkan range, the fortifications aimed at the
defence, and offered possibilities for retreat and agriculture.
29
The settlement
within the Iatrus fortress shows a very simple rural character with dwelling
houses built in the half-timber technique; stone fundaments are very rare.
The most solid building under the reign of Iustinianus I was the Christian
basilica.
30
After an eventful 6th century, the history of the limes on the lower Danube
the Byzantine presence north of the Balkan range ceased with the catastrophic
invasions of the Slavs and Avars at the end of the 6th century and in the be-
ginning of the 7th century.
Early Middle Ages (8th century-early 11th century)
With the immigration of the Slavs and Protobulgarians in the 7th century and
8th century the resettling of the areas deserted 200 years earlier began. The
map compiled based on the surface finds dated between 8th century up to
the beginning of the 11th century shows with 229 sites the highest density of
settlements from all examined historical periods (Fig. 15). The system consists
of extended sites, sometimes fortified, as well as small sites resulting in very
small quantities of finds. Similar to the later Iron Age this could indicate a
half-nomadic way of living. Numerous excavations in Bulgaria of early Medi-
eval sites verify the meagre character of the settlements: The new settlers
very often settled in the remains of Roman buildings, and probably used the
building material for the construction of pit houses.
31
CONCLUSIONS
The statistical evaluation of selected geographical parameters of the settle-
ments south of the Danube allows some preliminary conclusions regarding
the social and historical conditions during the examined periods.
Only the slightly increased percentages of settlement positions at upper
slope/ridge/summit, high bank and end of ridge/spur for later Iron Age
and Medieval settlements reflect the priority of natural factors and the special
way of living that influenced the choice of a place to settle down (Fig. 16).
Additionally, the later Iron Age settlements show a preference of a southern
exposition (Fig. 17). The preference of a northern and southern exposition
is amongst other things caused by the west-east orientation of most of the
ridges and valleys.
Our analyses verified that during the Roman period, apart from natural
circumstances, economic factors had a strong impact on settlement activity.
The Roman settlement network was obviously affected by the military system
of the limes and the official control of the land assignment too. Because of the
326 Sven Conrad
great local influence of limited settlement agglomerations like the concentra-
tion of settlements around the late Iron Age fortifications and for the Roman
periods the urban agglomeration in the vicinity of the legionary fortress
of Novae, we suggest that the area examined up till now is still too small to
draw statistically valid conclusions for all geographical parameters. Future
prospection will extend the research area and thus provide further compa-
rable data for statistical analysis. Furthermore, to prove the results based on
surface finds, each survey will have to be followed up by selected excavations.
The stratigraphy obtained will, in turn, help to reconstruct the progress of
historical periods and the turn from one period into another.
Under the present state of research, the examined area can serve as a
reference territory for both the Balkans and the Black Sea region. Our inter-
Fig. 15. Early Medieval sites (8th beg. 11th century).
327 Archaeological Survey on the Lower Danube
Roman Period (113 sites)
58%
5%
31%
4% 2%
Late Roman Period (148 sites)
5%
56%
30%
5%
3%
1%
Later Iron Age (129 sites)
54%
9%
19%
9%
5%
4%
1
2
3
4
5
6
Late Roman Period (148 sites)
56%
5%
30%
5% 3%
1%
Fig. 16. Position of sites south of the Danube (Roman and late Roman period without limes
fortifications and their civilian settlements). 1 down slope/valley/lowlands; 2 terrace; 3 mid
slope; 4 upper slope/ridge/summit; 5 high bank; 6 end of ridge/spur.
Later Iron Age
0
5
10
15
20
25
30
N
NE
E
SE
S
SW
W
NW
Roman Period
0
5
10
15
20
25
30
N
NE
E
SE
S
SW
W
NW
Late Roman Period
0
5
10
15
20
25
30
35
N
NE
E
SE
S
SW
W
NW
Early Middle Age
0
5
10
15
20
25
30
35
40
N
NE
E
SE
S
SW
W
NW
Fig. 17. Exposition of sites south of the Danube (Roman and late Roman period without limes
fortifications and their civilian settlements). For the number of sites see Fig. 16 (sites with
all-side exposition: later Iron Age 3; Roman period 1; late Roman period 3; early Middle
Age 3)
328 Sven Conrad
disciplinary investigations, albeit in a limited area, display very promising
opportunities and open perspectives for the historical and archaeological re-
search in that region, considering the fact that 70-80 % of the historical sites
are still unknown.
Acknowledgement
The research programme was carried out under the auspices of the R-
misch-Germanische Kommission in collaboration with the Bulgarian Academy
of Sciences Sofia, the Historical Museum Ruse, the Archaeological Institute
and the University of Bucharest, the Historical Museum Giurgiu and the
Slovakian Academy of Sciences Nitra. It has been supported by the Deutsche
Forschungsgemeinschaft Bonn since the autumn 1998. The author is very
grateful to G. v. Blow and S. v. Schnurbein (Frankfurt/M.), R. Ganev (Ruse),
P. Donevski (Svitov), R. Krau (Berlin), D. Stanev (Ruse) and L. Vagalinski
(Sofia) for their help and support.
Notes
1 Iatrus: Iatrus-Krivina I-V (Iatrus-Krivina VI under work); Blow 1994. Prelimi-
nary reports on the results of the excavations between 1992 and 2000 in Blow
& Mileva (eds.) 1999, 140-199. Novae: Summary by Press & Sarnowski 1990;
Ivanov, R. 1997, 556-574; 599-601; 1999, 184-189; 267-269; annual reports of the
excavations in Archeologia.
2 For preliminary results, see Conrad & Stanev 1998; 2002. A monograph on the
results of the archaeological and environmental analyses is still under work. The
responsible authors for the publication of the prehistoric periods are R. Krau
(Berlin; cf. Krau 2006), of the later Iron Age D. Stanev (Ruse). I wish to thank
both authors for the allowance to present some of their unpublished results.
3 Cf. Krau 2003.
4 Cf. Lichardus, Fol & Getov et al. 1996 (the prehistoric microregion in the vicinity
of Drama, south Bulgaria); Stefanov 1997 (settlement system in the vicinity of
the late Iron Age fortification of Sborjanovo near Isperih, NE Bulgaria); Bailey,
Tringham, Bass et al. 1998; Domaradski (ed.) 1999; Domaradski (ed.) 2001 (inves-
tigations on the settlement system of the late Iron Age in SW Bulgaria); Poulter
1999a; 1999b (the territory of the Roman to early Byzantine city Nicopolis ad
Istrum, north Bulgaria).
5 Cf. Poulter 1999a; 1999b.
6 Cf. Krau 2006. 6 Cf. Krau 2006.
7 Stefanov, S. 1956; Dremsizova-Nelinova & Ivanov 1983. 7 Stefanov, S. 1956; Dremsizova-Nelinova & Ivanov 1983. Stefanov, S. 1956; Dremsizova-Nelinova & Ivanov 1983.
8 Stefanov, S. 1956, 13 figs. 6 and 7; Dremsizova-Nelinova & Ivanov 1983, 31 8 Stefanov, S. 1956, 13 figs. 6 and 7; Dremsizova-Nelinova & Ivanov 1983, 31 Stefanov, S. 1956, 13 figs. 6 and 7; Dremsizova-Nelinova & Ivanov 1983, 31
no. 39.
9 Vlov 1962; Donevski 1997. 9 Vlov 1962; Donevski 1997. Vlov 1962; Donevski 1997.
10 Cf. Nestor 1950; Alexandrescu 1974; Oppermann 1984, 182. The excavations veri-
fied an important, probably fortified settlement.
11 Ivanov, D. 1975; Marazov (ed.) 1998, 222-225. Ivanov, D. 1975; Marazov (ed.) 1998, 222-225.
12 Danov 1976, 130-131. 12 Danov 1976, 130-131.
329 Archaeological Survey on the Lower Danube
13 Popov 2002, 156-165 fig. 41 (with bibliography).
14 Stefanov 1997.
15 Popov 2002, 170-172.
16 For the history of the Roman military presence and the fortification system on 16 For the history of the Roman military presence and the fortification system on
the lower Danube, see Ivanov 1997; 1999; Zahariade & Gudea 1997. New results
obtained from the survey in Conrad & Stanev 2002.
17 For that reason a definite theory regarding the transition period from late Iron
Age to Roman times is still impossible.
18 For the limes road cf. Conrad & Stanev 2002, 676 figs. 2-4.
19 Cf. Soproni 1978; Ubl 1980. Cf. Soproni 1978; Ubl 1980.
20 Cf. Poulter 1999c, 42-43. 20 Cf. Poulter 1999c, 42-43.
21 Cf. Blow 1995a.
22 In this hilly landscape many fortified hilltop settlements were erected, cf. Raev
1982, 7-16 map 1.
23 Cf. Blow 1995a, 43-49; Martin 1995, 40-41.
24 Bernhard 2001, 60-65 fig. 12.
25 Similar sizes were recorded in 25 Similar sizes were recorded in Germania superior in the territories between the
lower Neckar, Rhine and Danube, cf. Hssen 1994, 261 (plots between 60 and
120 ha, mostly between 65 and 80 ha).
26 Cf. Conrad & Stanev 2002, 674 fig. 5.
27 Piso 1991. 27 Piso 1991.
28 For this discussion, cf. Ivanov, R. 1997, 599-601; 1999, 267-269. 28 For this discussion, cf. Ivanov, R. 1997, 599-601; 1999, 267-269.
29 Cf. Henning 1987, 35-40; 1994, 471-473.
30 Cf. v. Blow 1995b, 64-65. 30 Cf. v. Blow 1995b, 64-65.
31 Cf. the results of the excavation of the Medieval settlement overlaying the former
Iatrus fortress in Wendel 1986.
Bibliography
Alexandrescu, A. 1974. Autour des fouilles de Zimnicea: I. La ncropole date
du bronze tardif de Zimnicea. II. Agglomeration et ncropole gtiques
Zimnicea, Thracia 3, 47-65.
Bailey, D.W., R.E. Tringham & J. Bass 1998. Expanding the dimensions of
early agricultural tells: the Podgoritsa Archaeological Project, Bulgaria.
JFieldA 25.4, 373-396.
Bender, H. & H. Wolff (eds.) 1994. Lndliche Besiedlung und Landwirtschaft in
den Rhein-Donau-Provinzen des Rmischen Reiches (Passauer Universittss-
chriften zur Archologie, 2). Espelkamp.
Bernhard, H. 2001. Die rmische Geschichte der Pfalz, in: K.-H. Rothenberger,
K. Scherer & J. Keddigkeit (eds.), Pflzische Geschichte I. Kaiserslautern,
45-78.
Blow, G. von 1994. Das sptantike Kastell Iatrus am Unterdonau-Limes in
Bulgarien. Stand und Probleme der Erforschung. Vortrag zur Jahressit-
zung der Rmisch-Germanischen Kommission, BerRGK 75, 5-22.
Blow, G. von 1995a. Die Entwicklung des Siedlungsbildes von Iatrus in der
Periode B/C, in: Iatrus-Krivina V. Berlin, 29-53.
330 Sven Conrad
Blow, G. von 1995b. Die Siedlungsperiode D2 in Iatrus, in: Iatrus-Krivina V.
Berlin, 61-66.
Blow, G. von & A. Mileva (eds.) 1999. Der Limes an der unteren Donau von
Diokletian bis Heraklios. Sofia.
Conrad, S. & D. Stanev 2002. Archaeological Survey on the Roman Frontier
on the lower Danube between Novae and Sexaginta Prista. Preliminary
Report (1997-2000), in: P. Freeman, J. Bennett, Z.T. Fiema & B. Hoffmann
(eds.), Proceedings of the XVIIIth International Congress of Roman Frontier
Studies held in Amman, Jordan (September 2000) (BAR Intern. Ser., 1084).
Oxford, 673-684.
Conrad, S. & D. Stanev 1998, Ein Limeskastell und sein Umfeld, Arch. Deutsch-
land, 2, 64.
Danov, C. 1976. Altthrakien. Berlin.
Domaradski, M. (ed.) 1999. Pametnici na trakijskata kultura na gornoto teenie na
reka Mesta (Razkopki i prouvanija, 26). Sofia.
Domaradski, M. (ed.) 2001. Materiali za arheologijata na sredna Struma (Razkopki
i prouvanija, 27). Sofia.
Donevski, P. 1997. New Late Roman Fortress in Svitov to the West of Novae,
in: A. Biernacki & P. Pawlak (eds.), Late Roman and Early Byzantine Cities
on the Lower Danube. Pozna, 35-38.
Dremsizova-Nelinova, C. & D. Ivanov 1983. Archeologieski pametnici v Ru-
senski okrg. Sofia.
Henning, J. 1987. Sdosteuropa zwischen Antike und Mittelalter (Archologische
Beitrge zur Landwirtschaft des 1. Jahrtausends u.Z. Schriften zur Ur- und
Frhgeschichte, 42). Berlin.
Henning, J. 1994. Die lndliche Besiedlung im Umland von Sadovec, Nord-
bulgarien (Vit-Tal) und die rmischen Agrarstrukturen im europischen
Vorland von Byzanz (Thrakien/Niedermsien), in: Bender & Wolff (eds.)
1994, 463-503.
Hssen, C.-M. 1994. Die lndliche Besiedlung und Landwirtschaft Oberger-
maniens zwischen Limes, unterem Neckar, Rhein und Donau whrend
der Kaiserzeit, in: Bender & Wolff (eds.) 1994, 255-265.
Iatrus-Krivina I-V (Schriften zur Geschichte und Kultur der Antike, 17). Berlin
1979-1995.
Ivanov, D. 1975. Srebrnoto skrovite ot selo Borovo, Rusensko, Izkustvo,
3-4, 14-22.
Ivanov, R. 1997. Das rmische Verteidigungssystem an der unteren Donau
zwischen Dorticum und Durostorum (Bulgarien) von Augustus bis Mau-
rikios, BerRGK 78, 468-640.
Ivanov, R. 1999. Dolnodunavskata otbranitelna sistema medu Dortikum i Duros-
torum ot Avgust do Mavrikij. Sofia.
Krau, R. 2006. Die prhistorische Besiedlung am Unterlauf der Jantra Prhistorische
Archologie in Sdosteuropa 20. Rahden.
331 Archaeological Survey on the Lower Danube
Lichardus, J., A. Fol & L. Getov et al. 1996. Bericht ber die bulgarisch-deutschen
Ausgrabungen in Drama (1989-1995). Neolithikum Kupferzeit
Bronzezeit Eisenzeit Rmerzeit, BerRGK 77, 5-154.
Marazov, I. (ed.). 1998. Ancient Gold: The wealth of the Thracians. Treasures from
the Republic of Bulgaria. Catalogue accompanying a travelling exhibition opening
in Feb. 1998. With essays by A. Fol, M. Tacheva, I. Venedikov. New York.
Martin, J. 1995. Sptantike und Vlkerwanderung (Oldenbourg Grundriss der
Geschichte, 4). Mnchen
3
.
Nestor, I. 1950. Raport sumar asupra campaniei de spturi arheologice de la
Zimnicea, StudCercIstorV 1.1, 93-102.
Oppermann, M. 1984. Thraker zwischen Karpatenbogen und gis.
Leipzig-Jena-Berlin.
Piso, I. 1991, Die Inschriften vom Pfaffenberg und der Bereich der Canabae
legionis, Tyche 6, 131-169.
Popov, Chr. 2002. Urbanizacija vv vtrenite rajoni na Trakija i Ilirija prez VI-I
vek predi Christa. Sofia.
Poulter, A. 1999a. The Transition to Late Antiquity on the Lower Danube: An
Interim Report (1996-8), AntJ 79, 145-185.
Poulter, A. 1999b. Gradishte near Dichin: A New Late Roman Fortress on
the Lower Danube, in: Blow & Mileva (eds.) 1999, 207-227.
Poulter, A. 1999c. Nicopolis ad Istrum: A Roman to Early Byzantine City. The
Pottery and Glass. With R.K. Falkner on the pottery and J.D. Shepherd on the
glass (Reports of the Research Committee of the Society of Antiquaries
London, 57). London-New York.
Press L. & T. Sarnowski 1990. Novae. Rmisches Legionslager und frhbyz-
antinische Stadt an der unteren Donau, AW 21, 4, 225-243.
Raev, R. 1982. Staroblgarski ukreplenija po Dolnija Dunav (VII-X v.) Varna.
Soproni, S. 1978. Der sptrmische Limes zwischen Esztergom und Szentendre.
Budapest.
Stefanov, J. 1997. Archeologiceski pametnici i istorija na prouvanija v Isperihskija
rajon. Sofia.
Stefanov, S. 1956. Starinite po dolnija basejn na Jantra (Materiali za
Archeologieskata karta na Blgarija, 8). Sofia.
Ubl, H. 1980, Der sterreichische Abschnitt des Donau-Limes. Ein For-
schungsbericht (1970-1979), in: W.S. Hanson & L.J.F. Keppie (eds.), Roman
Frontier Studies 12 (BAR Intern. Ser., 71). Oxford, 587-611.
Vlov, V. 1962. Razkopki na Kaleto v gr. Svitov, ArcheologijaSof 4.4, 7-15.
Wendel, M. 1986. Die mittelalterlichen Siedlungen, in: Iatrus-Krivina III,
27-210.
Zahariade, M. & N. Gudea 1997. The Fortifications of Lower Moesia. Amster-
dam.